Stereoselective construction of cis-decalin framework via radical domino cyclization

Article abstract of DOI:10.1016/j.tetlet.2013.01.050

Steroids with A/B cis-decalin moiety and various bioactivities have attracted many synthetic chemists as target molecules and development of efficient methodology providing cis-decalins has been desired. A stereoselective domino cyclization via Mn(OAc)₃ and Cu(OTf)₂-mediated radical reaction was found to be effective to construct cis-decalin framework.

Full text of DOI:10.1016/j.tetlet.2013.01.050

Tetrahedron Letters 54 (2013) 1589–1592
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Stereoselective construction of cis-decalin framework via radical
domino cyclization
Eriko Suzuki a, Minoru Ueda a, Shigeru Ohba b, Takeshi Sugai c, Mitsuru Shoji c,
⇑
a Department of Chemistry, Graduate School of Science, Tohoku University, 6-3 Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan
b Research and Education Center for Natural Sciences, Keio University, 4-1-1 Hiyoshi, Kouhoku-ku, Yokohama 223-8521, Japan
c Department of Pharmaceutical Science, Faculty of Pharmacy, Keio University, 1-5-30 Shibakoen, Minato-ku, Tokyo 105-8512, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Steroids with A/B cis-decalin moiety and various bioactivities have attracted many synthetic chemists as
target molecules and development of efficient methodology providing cis-decalins has been desired. A
stereoselective domino cyclization via Mn(OAc)3 and Cu(OTf)2-mediated radical reaction was found to
be effective to construct cis-decalin framework.
Received 7 December 2012
Revised 8 January 2013
Accepted 11 January 2013
Available online 18 January 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
cis-Decalin
Steriod
Radical cyclization
Domino reaction
Stereoselective synthesis
Steroids and related compounds are widely distributed in nat-
ure and have various interesting bioactivities. Their characteristic
polycyclic skeletons have attracted many chemists and a lot of
synthetic efforts have been devoted.1 Several steroids contain A/B
cis-decalin framework and significant bioactivities. For example,
digoxin (1) is one of the most famous cardiac glycosides and has
clinically been used for the treatment of congestive heart failure.2
Another steroid, batrachotoxin (2) is an alkaloid isolated from the
skins of arrow frogs and extremely potent neurotoxin (Fig. 1).3
Some preparation methods of cis-decalins in steroidal frame-
work have been reported. Kametani and Stork et al. employed cat-
alytic hydrogenation of Wieland–Miescher ketone4 for the total
syntheses of (+)-chenodeoxycholic acid and (+)-digitoxigenin,
respectively.5,6 Deslongchamps and co-workers accomplished total
synthesis of ouabain via a double Michael addition.7 Ishihara and
Yamamoto and co-workers found highly diastereoselective con-
struction of both cis- and trans-decalin via SnCl4-catalyzed tandem
cyclization of silyl dienol ethers,8 and Snider et al. developed oxi-
dative radical cyclization of polyene possessing b-keto ester.9 Since
mild reaction conditions are desirable for polyfunctionalized sub-
strate, radical cyclizations of (E)- or (Z)-trisubstituted b-keto esters
were performed to provide trans-decalins predominantly from
both substrates.10 Then, we envisaged a radical domino cyclization
of (Z)-b-keto ester 3 with oxygen-containing functional group at
benzylic position into cis-decalin 4 as illustrated in Scheme 1. Tri-
cyclic keto ester 4 would be transformed into the ABC ring moiety
of steroids via hydrolysis of ester followed by decarboxylation
along with dearomatization of benzene ring. This strategy would
enable to produce polycyclic steroids with wide variety of func-
tional groups. Herein, we describe a radical domino cyclization to-
ward cis-decalin framework.
Preparation of the domino cyclization precursor was started
from known aldehyde 511 derived from nerol as shown in
Scheme 2. First, Reformatsky reaction on 5 with methyl bromoace-
tate and zinc was attempted, however, undesired deacetylation at
allylic position was accompanied. On the other hand, aldol reaction
of aldehyde 5 and lithium enolate of methyl acetate provided
secondary alcohol 6 in 76% yield. Protection of alcohol 6 as trieth-
ylsilyl ether followed by removal of acetyl group afforded allylic
alcohol 7. Substitution of hydroxy group with iodide furnished
(Z)-allyl iodide 8 in 94% yield.
Figure 1. Structures of digoxin (1) and batrachotoxin (2).
E-mail address: shoji-mt@pha.keio.ac.jp (M. Shoji).
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.01.050

1590
E. Suzuki et al. / Tetrahedron Letters 54 (2013) 1589–1592
Scheme 1. Domino cyclization of (Z)-b-keto ester 3 to cis-decalin 4.
On the other hand, p-tolualdehyde (9) was converted into trim-
ethylsilylated cyanohydrin and subsequent treatment with lithium
bis(trimethylsilyl)amide and allyl iodide 8 provided the alkene
possessing all skeletal carbons for the desired cyclization precursor
(Scheme 3). Removal of TMS group with potassium fluoride regen-
erated the carbonyl group with the loss of cyanide and subsequent
reduction with NaBH4 afforded secondary alcohol 10 as a diaste-
reomeric mixture in 85% yield over three steps. Protection of
alcohol 10 as triisopropyl silyl ether and selective removal of tri-
ethylsilyl group under acidic condition gave b-hydroxy ester 11.
Oxidation of alcohol 11 with Dess–Martin periodinane12 furnished
the radical domino cyclization precursor 12 in 92% yield. Other b-
keto esters 13–17 were synthesized from various benzaldehydes
by the same way.
Scheme 3. Synthesis of the domino cyclization precursor 12.
oxidation between the primary and the secondary cyclizations.9b To
our delight, cis-decalin 18c was obtained as its enol ester, which
made it easy to separate cis- and trans-decalins 18c, 18k, and 18t.
Next, the effect of substituents on the aromatic ring was evalu-
ated. When methyl group was introduced at meta position, no ke-
tones were obtained with decrease of the ratio of desired cis isomer
19c (entry 2). Cyano group at para position diminished the yield
and cis/trans ratio was almost the same with 13 (entry 3).
Electron-rich aromatic ring with para methoxy group caused sub-
stitution of the siloxy group to acetate to give 21a0 along with
21a (entry 4). The acetate would be formed via addition of acetate
to benzyl cation stabilized by electron-donating para methoxy
group that might promote the release of siloxy group. On the other
hand, meta methoxy and dimethoxy derivatives 16 and 17 afforded
ketones 22k and 23k instead of acetate (entries 5 and 6). Although
meta methoxy group does not have a resonance with benzyl cation,
it works as an electron-donating group. Thus, the electron density
of the benzylic oxygen atom was increased and cleavage of O–Si
bond under acidic conditions followed by oxidation would afford
the ketones. After the screening of the substituents on aromatic
ring, para methyl group was concluded to produce the desired cis
isomer with the highest yield and ratio. Thus, b-keto ester 12
was employed for further investigation.
Since alkenyl b-keto esters 12–17 were in hand, the radical
domino cyclization was investigated. According to Snider’s condi-
tion, the cyclization precursors were exposed to manganese(III)
acetate and copper(II) acetate in acetic acid.9
Since no reaction occurred at room temperature, all reactions
were performed at a raised temperature as high as 50 °C (Table 1).
The cyclization of 12 afforded cis- and trans-decalins 18c and 18t
along with ketone 18k, which would be formed by cleavage of the
oxygen–silicon bond followed by oxidation (entry 1). The bulky TIPS
group did not seem to affect the diastereoselectivity in the domino
cyclization. Although Z-alkene 12 was used, undesired trans-decalin
18t was predominantly obtained (cis (18c + 18k):trans
(18t) = 1:1.9). It would arise from inversion of stereochemistry of
the sp3 radical intermediate or generation of carbocation by further
The solvent effect on the domino cyclization was also examined
as shown in Table 2. The ratio of cis isomer 18c was increased with
24% recovery of 12 in ethanol (Table 1, entry 1 vs Table 2, entry 1)
while no reaction was observed in tert-BuOH. Higher temperatures
promoted the formation of undesired trans isomer 18t (entry 2).
The cis/trans ratio was increased to 2.6 in trichloroethanol (entry
3). Although trifluoroethyl ether 18f via transetherification was
formed in trifluoroethanol, the cis/trans ratio was raised to 3.2 (en-
try 4). Among the mixed solvent of alcohols and trifluoroethanol,
the best cis/trans ratio was obtained in tert-BuOH with trifluoroeth-
anol (entries 5–7).
Since Ishibashi and co-workers reported the radical cycliza-
Scheme 2. Preparation of (Z)-allyl iodide 8.
tion of a-methylthio amides mediated by Mn(OAc)3 and Cu(OTf)2

E. Suzuki et al. / Tetrahedron Letters 54 (2013) 1589–1592
1591
Table 1
Domino cyclization of b-keto esters 12–17
Entry
R
Products and yield (%)
Ratio cis:trans
1
2
3
4
5
6
p-Me (12)
m-Me (13)
p-CN (14)
p-OMe (15)
m-OMe (16)
(m-OMe)2 (17)
18c (19), 18k (6), 18t (48)
19c (18), 19t (55)a
1:1.9
1:3.1
1:3.2
1:2.9
1:3.1
1:2.1
20c (13), 20t (41)
21c (12), 21a (7), 21t (34), 21a0 (21)
22c (14), 22k (3), 22t (53)a
23c (12), 23k (9), 23t (45)
a
1,2,4-Trisubstituted benzene was formed.
Table 2
Solvent effect on domino cyclization of 12
Me
Me
Me
Mn(OAc)3·2H2O (2.5 equiv)
Cu(OAc)2·H2O (1.2 equiv)
Me
Me
O
MeO2C
+
X
OTIPS
50 °C, solvent, time
HO
18c
O
OTIPS
H
H
Y
CO2Me
CO2Me
Me
12
18t
: X, Y = H, OTIPS
: X, Y = O
18f: X, Y = H, OCH2CF3
18k
Entry
Solvent
Time (h)
Yield (%)
Recovery (%)
Ratio cis:transa
18c; 18k; 18t
1
2
3
4
5
6
7
EtOH
1
2
7
2
7
7
7
21; 0; 20
24; 0; 45
24
0
11
0
13
10
5
1.1:1
1:1.9
2.6:1
3.2:1d
1.9:1
2.4:1
3.3:1
EtOHb
CCl3CH2OH
28; 16; 17
27; 14; 18c
25; 15; 21
23; 20; 18
21; 18; 12
CF3CH2OH
EtOH–CF3CH2OHe
iPrOH–CF3CH2OHe
tBuOH–CF3CH2OHe
a
b
c
(18c + 18k)/18t.
At 80 °C.
Along with 16% of 18f.
(18c + 18k + 18f)/18t.
A 3:1 mixture.
d
e
instead of Cu(OAc)2,13 domino cyclization of 12 with Cu(OTf)2
was investigated (Table 3). Both yields of 18c, 18k, and 18t
and cis/trans ratio increased with the change of Cu(OAc)2 to
Cu(OTf)2 (Table 2, entry 1 vs Table 3, entry 1). The cyclization
in tert-BuOH afforded 18c in 47% yield (entry 2) while no reac-
tion was observed with Cu(OAc)2. Surprisingly, the cyclization
proceeded even at room temperature but longer reaction time
reduced the yield of 18c (entry 3). More loading of Mn(OAc)3
gave the best yield of 18c along with 18t in 22% yield (entry
4). Although addition of trifluoroacetic acid diminished yields
and cis/trans ratio (entry 5), trifluoromethanesulfonic acid raised
the reaction rate and the yield of 18c (entry 3 vs entries 6 and
7). Unfortunately, the cyclization in 2-propanol at 0 °C resulted
in low yield because of longer reaction time and side reactions
(entry 8). When 2.5 equiv of Cu(OTf)2 was used, production of
18c and 18t was suppressed (entry 9). It was found that trifluo-
romethanesulfonic acid accelerated the reaction rate but an
overload caused undesired side reactions (entry 10).
Since acids inclined the equilibrium to enol form from ketone,
they would accelerate the formation of the manganese(III) enolate
and the subsequent radical 3a. But, excess amount of acids would
increase the activity of Mn(OAc)3 and/or Cu(OAc)2 as oxidant,
which might cause oxidation of radical 3b to the corresponding
carbocation that might decrease cis/trans selectivity on the second-
ary cyclization.
A diastereomeric mixture of enol ester 18c was converted into
hydroxy ketones 24 and 25 by decarboxylation and subsequent
deprotection (Scheme 4). After chromatographic separation, the
stereochemistry of 24 was determined by X-ray crystallographic
analysis as illustrated in Figure 2.14
In conclusion, we found an efficient synthetic method of cis-
decalin by radical domino cyclization under mild reaction condi-

1592
E. Suzuki et al. / Tetrahedron Letters 54 (2013) 1589–1592
Table 3
Domino cyclization of 12 using Cu(OTf)2
Entry
Additive (equiv)
Solvent
Temp.
Time (h)
Yields (%)
Ratio cis:transa
18c; 18k; 18t
1
2
None
None
None
None
TFA (1.0)
TfOH (1.0)
TfOH (1.5)
TfOH (1.5)
TfOH (1.5)
TfOH (3.0)
EtOH
50 °C
50 °C
RT
RT
RT
RT
RT
0 °C
RT
RT
3
2
13
7
10
7
3
21
4
2.5
33; 3; 26
47; 2; 20
36; 0; Trace
53; 1; 22
32; 0; 17
44; 1; 14
52; 1; 15
16; Trace; 0
37; 5; 9
1.4:1
2.5:1
N.D.a
2.5:1
1.9:1
3.2:1
3.5:1
N.D.a
4.7:1
2.9:1
tBuOH
tBuOH
tBuOH
tBuOH
tBuOH
tBuOH
iPrOH
tBuOH
tBuOH
3
4b
5
6
7
8
9c
10
36; 2; 13
a
b
c
Not determined.
5.0 equiv of Mn(OAc)3 was used.
2.5 equiv of Cu(OTf)2 was used.
tions. The combination of Mn(OAc)3 and Cu(OTf)2 remarkably pro-
moted the cyclization to afford the cis-decalin predominantly. This
methodology would make it readily to synthesize A/B-cis steroids,
diterpenoids, and their derivatives.
Acknowledgements
This work was supported by JSPS KAKENHI Grant Number
24510305 and ITOH SCIENCE FOUNDATION. We also thank the Ja-
pan Society for the Promotion of Sciences for the JSPS Fellowship
for Young Scientists (E.S.).
References and notes
1. Chapelon, A.-S.; Moraléda, D.; Rodriguez, R.; Ollivier, C.; Santelli, M. Tetrahedron
2007, 63, 11511. and references cited therein.
2. Thomas, R.; Gray, P.; Andrew, J. Adv. Drug Res. 1990, 19, 312.
3. (a) Dumbacher, J. P.; Beehler, B. M.; Spande, T. F.; Garraffo, H. M.; Daly, J. W.
Science 1992, 258, 799; (b) Albuquerque, E. X.; Daly, J. W.; Witkop, B. Science
1971, 172, 995.
Scheme 4. Transformation of 18c into hydroxy ketones 24 and 25.
4. (a) Gutzwiller, J.; Buchschacher, P.; Fürst, A. Synthesis 1977, 167; (b) McMurry,
J. E. J. Am. Chem. Soc. 1968, 90, 6821.
5. Kametani, T.; Suzuki, K.; Nemoto, H. J. Am. Chem. Soc. 1981, 103, 2890.
6. Stork, G.; West, F.; Lee, H. Y.; Isaacs, R. C. A.; Manabe, S. J. Am. Chem. Soc. 1996,
118, 10660.
7. Zhang, H.; Reddy, M. S.; Phoenix, S.; Deslongchamps, P. Angew. Chem., Int. Ed.
2008, 47, 1272.
8. Uyanik, M.; Ishihara, K.; Yamamoto, H. Org. Lett. 2006, 8, 5649.
9. (a) Snider, B. B.; Mohan, R.; Kates, S. A. J. Org. Chem. 1985, 50, 3659; (b) Zoretic,
P. A.; Zhang, Y.; Ribeiro, A. A. Tetrahedron Lett. 1996, 37, 1751.
10. (a) Snider, B. B.; Kiselgof, J. Y. Tetrahedron 1998, 54, 10641; (b) Yang, D.; Ye, X.-
Y.; Xu, M.; Pang, K.-W.; Cheung, K.-K. J. Am. Chem. Soc. 2000, 122, 1658.
11. Grieco, P. A. J. Chem. Soc., Chem. Commun. 1972, 486.
12. (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4115; (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.
13. Toyao, A.; Chikaoka, S.; Takeda, Y.; Tamura, O.; Muraoka, O.; Tanabe, G.;
Ishibashi, H. Tetrahedron Lett. 2001, 42, 1729.
14. Crystallographic data for 24 have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication numbers CCDC
913695. Spectral data for compund 24: 1H NMR (500 MHz, CDCl3) d 1.48 (3H, s),
1.71 (1H, br s), 1.75 (1H, ddd, J = 14.5, 10.0, 7.0 Hz), 1.94 (1H, dt, J = 13.5,
6.0 Hz), 2.20–2.27 (2H, m), 2.28 (1H, ddd, J = 11.5, 6.0, 1.5 Hz), 2.32–2.40 (1H,
m), 2.36 (3H, s), 2.43 (1H, ddd, J = 15.0, 6.0, 1.5 Hz), 2.43–2.51 (1H, m), 2.73
(1H, dd, J = 15.0, 5.5 Hz), 4.83 (1H, br s), 7.09 (1H, d, J = 8.0 Hz), 7.15 (1H, s),
7.44 (1H, d, J = 8.0 Hz); 13C NMR (125 MHz, CDCl3) d 21.4, 27.6, 35.7, 37.0, 37.4,
38.3, 41.3, 44.8, 68.0, 126.4, 127.6, 127.8, 135.0, 137.7, 142.5, 211.4; FT-IR
(neat)
1128, 1072, 1037, 1018, 974, 823, 800, 752, 665 cmꢀ1
[M+Na]+ calcd for C16H20O2SiNa: 267.1356, found: 267.1365.
m
3413, 2931, 2868, 1703, 1462, 1419, 1381, 1252, 1217, 1171, 1157,
Figure 2. ORTEP drawing of 24. Thermal ellipsoids are drawn at 50% probability
level.
;
HRMS (ESI-TOF)