Cascades to substituted indoles

Article abstract of DOI:10.1021/jo000831n

This paper describes the synthesis of dithioindoles from the free-radical cyclizations of arylisonitriles having pendant alkynes. Also described is the synthesis of substituted indoles and spiro-fused indoles from the coupling of dithioindoles with active hydrogen-containing compounds.

Full text of DOI:10.1021/jo000831n

J . Org. Chem. 2000, 65, 6213-6216
6213
Ca sca d es to Su bstitu ted In d oles
J on D. Rainier* and Abigail R. Kennedy
Department of Chemistry, The University of Arizona, Tucson, Arizona 85721
rainier@u.arizona.edu
Received May 31, 2000
This paper describes the synthesis of dithioindoles from the free-radical cyclizations of arylisonitriles
having pendant alkynes. Also described is the synthesis of substituted indoles and spiro-fused indoles
from the coupling of dithioindoles with active hydrogen-containing compounds.
In tr od u ction
for the corresponding arylisonitrile-alkene cyclizations.12,13
Assuming that arylisonitrile-alkynes followed a similar
pathway, their cyclization would result in an unprec-
edented free-radical approach to indolenines (i.e., 4) and,
through the incorporation of nucleophiles into the reac-
tion mixture, the synthesis of highly substituted indoles
(Scheme 1).14
While our main focus was the sequence depicted in
Scheme 1 beginning with the 5-exo-dig cyclization of 2,
a potentially competitive 6-endo-dig cyclization would
provide an entry into substituted quinolines. As quino-
lines are present in a variety of interesting structures
we also found this possibility intriguing.15
Described herein is the successful demonstration of the
sequence depicted in Scheme 1 through the synthesis of
bis-thiol and 2-stannyl indoles from the radical cycliza-
tions of arylisonitriles having pendant alkynes. Also
described here are our experiments showing that bis-
thioindoles are synthetically useful through their cou-
pling with active hydrogen compounds and their conver-
sion into spiro-fused indolethioimidates.
Although Fischer first described their generation some
115 years ago, the synthesis of indoles continues to
receive attention.1,2 The presence of indole subunits in a
number of bioactive molecules undoubtedly plays a key
role in the continued activity in this area.3 From our
perspective, those approaches to substituted indoles that
have employed free-radical cyclizations2 or indolenines
from gramine fragmentations4 have stood out. This is not
only because of the high reactivity of each of these species
but it is also a result of the relatively mild and neutral
reaction conditions that are required for their synthesis.
Thus, in the course of using radical or indolenine
intermediates, one can incorporate sensitive functionality
without having to be overly concerned with unwanted
side reactions.5,6
We became interested in arylisonitrile-alkyne free
radical cascades out of a desire to incorporate both radical
cyclizations and indolenine couplings into the same
reaction sequence.7-11 This notion came from a consid-
eration of the mechanism that Fukuyama has proposed
Resu lts a n d Discu ssion
(1) (a) Fischer, E.; J ourdan, F. Ber. 1883, 16, 2241. (b) Fischer, E.;
Hess, O. Ber. 1884, 17, 551.
As it seemed probable that the alkynyl substitution
might be important in the ultimate indole:quinoline
product ratio, a variety of substituted alkynylisonitriles
were synthesized in two steps from 2-iodoformanilide
(Table 1). Sonogashira coupling reactions provided o-
alkynylformanilides 7a -e from 6.16 Dehydration of the
formanilides gave cyclization precursors 8a -e.17,18
With arylisonitrile-alkynes in hand, we subjected
them to Fukuyama’s conditions using an excess of Bu3-
SnH to ensure the reduction of any indolenine that was
formed (i.e., NuH in Scheme 1 ) Bu3SnH). Delightfully
(2) For a recent review covering approaches to the synthesis of
indoles see: Gribble, G. W. J . Chem. Soc., Perkin Trans. 1 2000, 1045.
(3) For recent reviews of indole containing natural products, see:
(a) Lounasmaa, M.; Tolvanen, A. Nat. Prod. Rep. 2000, 17, 175. (b)
Faulkner, D. J . Nat. Prod. Rep. 1999, 16, 155.
(4) For examples of the synthesis of substituted indoles from
gramine fragmentations, see: (a) Cushing, T. D.; Sanz-Cervera, J . F.;
Williams, R. M. J . Am. Chem. Soc. 1993, 115, 9323. (b) Smith, A. B.,
III; Haseltine, J . N.; Visnick, M. Tetrahedron 1989, 45, 2431. (c)
Remers, W. A.; Brown, R. K. In Indoles; Houlihan, W. J ., Ed.;
Heterocyclic Compounds; Wiley-Interscience: New York, 1972; Part
One, pp 200-203.
(5) Radical reactions are generally run under neutral conditions.
See: (a) J asperse, C. P.; Curran, D. P.; Fevig, T. L. Chem. Rev. 1991,
91, 1237. (b) Bowman, W. R.; Bridge, C. F.; Brookes, P. J . Chem. Soc.,
Perkin Trans. 1 2000, 1.
(6) For examples of the generation of indolenines in the presence of
sensitive functionality, see ref 4.
(12) Fukuyama, T.; Chen, X.; Peng, G. J . Am. Chem. Soc. 1994, 116,
3127.
(13) Aryl isonitrile-alkene cyclizations have been used to generate
complex indoles. See ref 12 and: (a) Kobayashi, Y.; Fukuyama, T. J .
Heterocycl. Chem. 1998, 35, 1043. (b) Kobayashi, S.; Peng, G.; Fuku-
yama, T. Tetrahedron Lett. 1999, 40, 1519.
(14) A possible complication with the mechanism depicted in Scheme
1 is the intermolecular addition of the radical to the alkyne rather
than the isonitrile. See: Leardini, R.; Nanni, D.; Zanardi, G. J . Org.
Chem. 2000, 65, 2763.
(7) For a review of cascade sequences in organic synthesis, see:
Parsons, P. J .; Penkett, C. S.; Shell, A. J . Chem. Rev. 1996, 96, 195.
(8) For a preliminary account of this work see: Rainier, J . D.;
Kennedy, A. R.; Chase, E. Tetrahedron Lett. 1999, 40, 6325.
(9) Curran and co-workers have developed and applied isonitrile-
alkyne radical cyclizations in synthesis. See: (a) Curran, D. P.; Liu,
H. J . Am. Chem. Soc. 1991, 113, 2127. (b) Curran, D. P.; Ko, S.-B.;
J osien, H. Angew. Chem., Int. Ed. Engl. 1995, 34, 2683. (c) J osien, H.;
Ko, S.-B.; Bom, D.; Curran, D. P. Chem. Eur. J . 1998, 4, 67.
(10) Nanni and co-workers have also investigated arylisonitrile-
alkyne radical reactions. See: Nanni, D.; Pareschi, P.; Rizzoli, C.;
Sgarabotto, P.; Tundo, A. Tetrahedron 1995, 51, 9045.
(15) Ito and co-workers have demonstrated that the anionic cycliza-
tion of arylisonitriles having pendant alkynes provides quinolines.
See: Suginome, M.; Fukuda, T.; Ito, Y. Org. Lett. 1999, 1, 1977.
(16) Thorand, S.; Krause, N. J . Org. Chem. 1998, 63, 8551.
(17) Obrecht, R.; Herrmann, R.; Ugi, I. Synthesis 1985, 4, 400.
(18) 8b,c,e,f were not stable to purification in our hands. Each of
(11) For other isonitrile radical reactions, see: (a) Ryu, I.; Sonoda,
N.; Curran, D. P. Chem. Rev. 1996, 96, 6, 177. (b) Stork, G.; Sher, P.
M. J . Am. Chem. Soc. 1983, 105, 6765. (c) References 9, 10, 12, 13,
and 27.
these showed the expected isonitrile IR stretch (i.e., 8b: 2119 cm-1
8c: 2124 cm-1; 8e: 2124 cm-1; 8f: 2130 cm-1).
;
10.1021/jo000831n CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/22/2000

6214 J . Org. Chem., Vol. 65, No. 19, 2000
Rainier and Kennedy
Ta ble 3. Th iol-Med ia ted F r ee-Ra d ica l Cycliza tion s of 8a
Sch em e 1
entry
R
indole
yield (%)
1
2
3
4
5
6
Et
Bu
Ph
11
12
13
14
15
16
86
66
49
94
60
72
CH2CH2OH
CH2CH2OTBS
CH2CH2CO2Me
Ta ble 1. Ar ylison itr ile-Alk yn e Syn th eses
exposed to the radical conditions we isolated a mixture
of indole 9d and quinoline 10d 24 in a 14:1 ratio respec-
tively (entry 4). Thus, it appears that the 5-exo-dig
cyclization of 8a is largely due to the steric destabilization
of the 6-endo-dig cyclization manifold.
Having shown that the free-radical cyclization of 8a
leads to indoles, we set out to trap the indolenine
intermediate with nonhydride (hydrogen atom) nucleo-
philes. Unfortunately, we have been unsuccessful in our
attempts to couple 4a (Scheme 1, R ) Bu3Sn, R′ ) TMS)
with oxygen or nitrogen nucleophiles in the presence of
Bu3SnH.25 Apparently, indolenine 4a (R ) Bu3Sn, R′ )
TMS) has a relatively high affinity for tin hydride as even
the use of substoichiometric quantities of Bu3SnH and
excess alcohol or amine resulted in the isolation of
mixtures of reduced indole 9a along with unreacted
starting material 8a .26
entry
R
7
yield (7) (%)
8
yield (8) (%)
1
2
3
4
5
TMS
Ph
a
b
c
d
e
93
100
100
92
a
b
c
d
e
82
a
a
100
a
Bu
t-Bu
CH2OBn
57
a
Isonitrile was not purified prior to the free-radical cyclization.
Ta ble 2. Bu 3Sn H-Med ia ted Ar ylison itr ile-Alk yn e
F r ee-Ra d ica l Cycliza tion s
We were undaunted by our inability to couple 4a (R )
Bu3Sn, R′ ) TMS) with non-hydrogen nucleophiles as we
believed that we might be able to circumvent the reduc-
tion problem by simply inducing the free-radical cycliza-
tion of 8a with a reagent other than Bu3SnH. In this
regard, thiols were attractive as they might both initiate
the radical cascade and act as nucleophiles in the reaction
with the indolenine.27 Gratifyingly, bis-thioindole 11 was
isolated in 86% yield when isonitrile 8a was allowed to
react with ethanethiol and AIBN (Table 3, entry 1). As
depicted, the use of the same conditions with other alkyl
and aryl thiols also gave bis-thiol indoles. These reactions
clearly demonstrate that the free-radical cyclization of
arylisonitriles having pendant alkynes can be used to
generate indoles. These experiments also demonstrate
that the indolenine intermediates formed during these
sequences are susceptible to attack by nucleophiles.28
We became intrigued by the synthetic possibilities
associated with bis-thioindoles 11-16. From our perspec-
tive it appeared that they might serve as versatile
entry
isonitrile
R
9:10
yield (%)
1
2
3
4
5
6
8a
8b
8c
8d
8e
8fa
TMS
Ph
1:0
2.2:1
82
41
63
60
11
18
8u
1:5.3
t-Bu
CH2OBn
H
14:1b
2:1
0:1
a
b
From TBAF hydrolysis of the 10a TMS group. 10d was
isolated as the 2-stannylated quinoline.
for us, the treatment of TMS-alkyne 8a with 2 equiv of
Bu3SnH and 10% AIBN in either benzene or acetonitrile
resulted in indole 9a 19 in 82% yield after acidic workup
to protodestannylate the 2-stannylindole product (Table
2, entry 1).20 None of the corresponding quinoline was
detected.21 In a similar fashion, phenyl substituted
alkyne 8b gave indole 9b12 as the major product. How-
ever, the overall yield was lower and 9b was obtained
along with the corresponding quinoline 10b22 in a 2.2:1
product ratio respectively (entry 2). In contrast, butyl-
substituted alkyne 8c provided quinoline 10c17 as the
major product of a 5.3:1 mixture (entry 3). In an effort
to ascertain whether silicon’s ability to stabilize R-radi-
cals23 or its steric destabilization of the 6-endo pathway
through its interactions with Bu3Sn during the transition
state were responsible for the observed selectivity of 8a
we synthesized tert-butyl alkyne 8d . When 8d was
(23) For examples of silicon’s influence on free-radical cyclizations,
see: (a) Gillmann, T.; Hu¨lsen, T.; Massa, W.; Wocadlo, S. Synlett 1995,
1257. (b) Shi, C.; Zhang, Q.; Wang, K. K. J . Org. Chem. 1999, 64, 925.
(c) Wilt, J . W.; Aznavoorian, P. M. J . Org. Chem. 1978, 43, 1285.
(24) 10d was isolated as the 2-tributylstannylquinoline adduct.
Acidic workup failed to give protodestannylated product.
(25) We have attempted to couple the intermediate indolenine with
diethylamine, aniline, benzylamine, and methanol.
(26) The reduction sequence could be the result of a heterolytic or
homolytic process. For the homolytic reduction of isoelectronic enones,
see: Enholm, E. J .; Whitley, P. E. Tetrahedron Lett. 1996, 37, 559.
For an example of the use of Bu3SnH in hydride reductions of
imminium ions, see: Palmisano, G.; Lesma, G.; Nali, M.; Rindone, B.;
Tollari, S. Synthesis 1985, 1072.
(19) Ishibashi, H.; Nishida, K.; Ikeda, M. Synth. Commun. 1993,
23, 2381.
(20) The protodestannylated products were more robust and thus
easier to purify than the 2-stannylindoles.
(21) Tamao, K.; Kodama, S.; Nakajima, I.; Kumada, M. Tetrahedron
1982, 38, 3347.
(22) UÄ beda, J . I.; Villacampa, M.; Avendan˜o, C. Synthesis 1998,
1176.
(27) (a) Saegusa first demonstrated the reactivity of isonitriles with
thiol radicals. See: Saegusa, T.; Kobayashi, S.; Ito, Y. J . Org. Chem.
1970, 35, 2118. More recently, Bachi and Nanni have independently
coupled thiol radicals with isonitriles. See: (b) Bachi, M. D.; Balanov,
A.; Bar-Ner, N. J . Org. Chem. 1994, 59, 7752. (c) Camaggi, C. M.;
Leardini, R.; Nanni, D.; Zanardi, G. Tetrahedron 1998, 54, 5587.

Cascades to Substituted Indoles
J . Org. Chem., Vol. 65, No. 19, 2000 6215
Ta ble 4. Bis-Th ioin d oles a s Alk yla tin g Agen ts w ith
Active Hyd r ogen Com p ou n d s
Ta ble 5. Sp ir o F u sed Th ioim id a tes
entry
R
39
yield (%)
1
2
Et
Me
a
b
86a
80b
a
b
5:1 mixture of diastereomers. 2:1 mixture of diastereomers.
adduct 24 in 82% yield after subjecting 11 to dimethyl-
malonate, Bu3P (50 mol %), and acetonitrile at reflux
(entry 1). Interestingly, we had not only formed the
desired carbon-carbon bond through the displacement
of the ethanethiol group but we had also replaced the
TMS group with a hydrogen. Sufficiently intrigued by
this result, we subjected other acidic hydrogen containing
nucleophiles to the coupling conditions; ketoesters 17 and
18, diethylacetamidomalonate 19, diethylaminomalonate
20, and glycine Schiff bases 2135 and 2236 gave 25, 26,
27, 29, 30, and 31 respectively. On the other hand,
acetamidoglycine 23 and cyclohexanone failed to provide
adduct giving low to moderate yields of bis-thiol 32.37 This
reaction is also amenable to thioether substitution as 15
gave coupled product 28 when it was exposed to the Bu3P
conditions and acetamidomalonate 19.
The TMS group appears to be critical for the success
of the reaction. No coupling products were observed when
bis-thiol adduct 32 lacking the TMS group was re-
subjected to the alkylation conditions; this reaction
resulted in the recovery of 32 (eq 1).
a
After reduction of the imine with NaCNBH3.
intermediates in the synthesis of other, more highly
substituted species. For example, a C-10 anion (generated
via deprotonation,29 thioether reduction,30,31 or fluoride
induced desilylation32) could be used in coupling reactions
with electrophiles. Conversely, elimination of the C-10
thioether in a gramine-like fashion would provide a C-10
electrophile that would then be amenable to coupling
reactions with nucleophiles. Although Poppelsdorf and
Holt had previously found simple 3-ethylthiomethylindole
to be unreactive to gramine alkylation reactions with
active hydrogen compounds,33 we were confident that
11-16 would be more reactive by virtue of the fact that
the C-2 thioether would aid in the elimination sequence.
We elected to pursue the use of the bis-thioindoles as
electrophiles as the success of these investigations would
circumvent our inability to trap indolenines with non-
hydrogen nucleophiles from the tin mediated radical
sequence. Therefore, our initial investigations targeted
the alkylation of pronucleophiles with bis-ethanethiol
adduct 11 utilizing Somei’s conditions (Table 4).34 We
were extremely pleased to be able to isolate malonate
Our current mechanistic hypothesis, based on the
proposal by Somei for the analogous gramine fragmenta-
tion reaction and taking into account the necessity of the
TMS group, is illustrated in Scheme 2 for 11.34 We believe
that ylide 34 is generated after formation of the phos-
phonium salt 33 and loss of TMS cation. Protonation of
the ylide and deprotonation of the active hydrogen
compound leads to 35. Carbon-carbon bond formation
occurs either via direct displacement of tributylphosphine
or via an intermediate indolenine; both of these species
act to regenerate Bu3P and provide the observed prod-
ucts.
(28) It is not clear to us at the present time whether thiol addition
to 4a (R ) thioether, R′ ) TMS) is occurring heterolytically or
homolytically. Both would lead to bis-thiol indoles.
(29) For examples of the deprotonation of thiomethylsilanes see: (a)
Ager, D. J .; Cookson, R. C. Tetrahedron Lett. 1980, 21, 1677. (b) Ager,
D. J . Tetrahedron Lett. 1980, 21, 4759.
(30) For a review on the reductive metalation of phenyl thioethers
see: Cohen, T.; Bhupathy, M. Acc. Chem. Res. 1989, 22, 152.
(31) For an example of the reductive metalation of thiomethylsilanes
see: Corey, E. J .; Chen, Z. Tetrahedron Lett. 1994, 35, 8731.
(32) For an example of the fluoride-induced heterolytic fragmenta-
tion of benzylic TMS compounds to the corresponding anions see: Ricci,
A.; Fiorenza, M.; Grifagni, M. A.; Bartolini, G. Tetrahedron Lett. 1982,
23, 5079.
(33) Poppelsdorf, F.; Holt, S. J . J . Chem. Soc. 1954, 4094.
(34) Somei, M.; Karasawa, Y.; Kaneko, C. Heterocycles 1981, 16, 941.
(35) Stork, G.; Leong, A. Y. W.; Touzin, A. M. J . Org. Chem. 1976,
41, 3491.
(36) O’Donnell, M. J .; Polt, R. L. J . Org. Chem. 1982, 47, 2663.
(37) We believe that bis-thioindole 32 comes from an irreversible
coupling reaction between ethanethiol and phosphonium salt 35
(Scheme 2).

6216 J . Org. Chem., Vol. 65, No. 19, 2000
Rainier and Kennedy
Sch em e 2
C-3 spiro compounds as are present in a number of
bioactive indole natural products including, but not
limited to, the spiroindolinone spirotryprostatin.39,40 Ex-
posure of 29 to 4 Å MS and propionaldehyde or acetal-
dehyde at room-temperature resulted in the formation
of diastereomeric spirothioimidates 39 in high yields
(Table 5). These conditions are milder than the “normal”
Pictet-Spengler conditions to â-carbolines where el-
evated temperatures are generally required.41
Con clu sion
To conclude, we have carried out the synthesis of a
number of bis-thioindoles from arylisonitrile-alkyne free
radical cascades. These sequences effectively demonstrate
that free-radical cyclizations to indolenines can be coupled
with indolenine trapping reactions. We have also shown
that bis-thioindoles are synthetically useful in the syn-
thesis of other, more functionalized, indoles. This has
been demonstrated through their coupling reactions with
active hydrogen compounds and through the formation
of spiroindoles from intramolecular Mannich cyclizations.
We are currently exploring the uses of these processes
in the synthesis of highly substituted indole containing
natural products.
Regardless of the sequence of events leading to its
formation, we set out to prove the existence of phosphorus
ylide 34. The isolation of Wittig products from the
reaction of 11 with an aldehyde would provide indirect
evidence for the existence of 34. Additionally, the success
of this experiment would demonstrate the utility of 11
in carbon-carbon bond forming reactions with electro-
philes. When bis-thioindole 11 was subjected to the
alkylation conditions in the presence of benzaldehyde we
were able to isolate styrene 37 in an unoptimized 49%
yield (eq 2). Thus, it appears that ylides are important
in the use of 11 as an alkylating agent.
Ack n ow led gm en t. We gratefully acknowledge the
National Institutes of Health (GM56677) and The
University of Arizona (Startup funding for J .D.R.) for
partial support of this work. We would also like to thank
Mr. Eric Chase for carrying out a preliminary experi-
ment. We are indebted to Dr. Neil J acobsen and Dr.
Arpad Somagyi for help with NMR and mass spectros-
copy, respectively.
Bis-thioindole 11 also acts as an alkylating agent in
the absence of Bu3P. Nitrile 38 was formed in 40% yield
when 11 was subjected to KCN and DMF (eq 3).38
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and copies of 1H and 13C NMR spectra for 7c-e, 8a ,d ,
9d , 10d , 11-16, 24-32, 37, 38, and 39a ,b. This material is
available free of charge via the Internet at http://pubs.acs.org.
J O000831N
Havingdemonstrated the utilityofthe 3-ethylmethaneth-
iol group in couplings with electrophiles, we targeted
intramolecular Mannich-type (Pictet-Spengler) cycliza-
tions of aminomalonate adduct 29. Not only would these
experiments demonstrate the utility of the 2-thiol group
but if successful they would result in the formation of
(39) Isolation: (a) Cui, C.-B.; Kakeya, H.; Okada, G.; Onose, R.;
Osada, H. J . Antibiot. 1996, 49, 527. (b) Cui, C.-B.; Kakeya, H.; Osada,
H. Tetrahedron 1997, 53, 59. Synthesis: (a) Edmondson, S. D.;
Danishefsky, S. J . Angew. Chem., Int. Ed. Engl. 1998, 37, 1138. (b)
Overman, L. E.; Rosen, M. D.; Osada, H.; Shaka, A. J .; Taylor, N. D.
219th ACS National Meeting, 2000, Org 843.
(40) Other natural products containing spiro-fused indoles include
the aspidospermine alkaloids. For example, see: Saxton, J . E. Nat.
Prod. Rep. 1987, 4, 591.
(38) Makisumi, Y.; Takada, S. Chem. Pharm. Bull. 1976, 24, 770.
(41) Cook, J . M.; J awdosiuk, M. J . Org. Chem. 1984, 49, 2699.