A new synthesis of 3-arylthioindoles as selective COX-2 inhibitors using PIFA

Article abstract of DOI:10.1016/j.tetlet.2004.03.153

The direct 3-arylthiolation of 2-substituted indoles using phenyliodine(III)bis trifluoroacetate (PIFA) in (CF₃)₂CHOH with a wide variety of benzenethiols has been accomplished. In particular, indoles bearing a 6-MeSO₂ and either a 2-methyl or 2-carboxymethyl substituent could be 3-arylthiolated in good to excellent yields to afford the corresponding 3-arylthioindoles as selective COX-2 inhibitors. In a study varying the electronic nature of the 5-substituent of 2-CO₂Et indoles, it was discovered that the yield of the reaction improved as the substituent became more electron withdrawing. This result was consistent with a proposed mechanism involving benzenethiol displacement of an intermediate 3-IPh indole complex.

Full text of DOI:10.1016/j.tetlet.2004.03.153

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 4073–4075
A new synthesis of 3-arylthioindoles as selective
COX-2 inhibitors using PIFA
Jeffrey A. Campbell,* Chris A. Broka, Leyi Gong, Keith A. M. Walker and Jin-Hai Wang
Department of Medicinal Chemistry, Roche Palo Alto., 3431 Hillview Ave, Palo Alto, CA 94304, USA
Received 26 February 2004; revised 24 March 2004; accepted 25 March 2004
Abstract—The direct 3-arylthiolation of 2-substituted indoles using phenyliodine(III)bis trifluoroacetate (PIFA) in (CF3)2CHOH
with a wide variety of benzenethiols has been accomplished. In particular, indoles bearing a 6-MeSO2 and either a 2-methyl or
2-carboxymethyl substituent could be 3-arylthiolated in good to excellent yields to afford the corresponding 3-arylthioindoles as
selective COX-2 inhibitors. In a study varying the electronic nature of the 5-substituent of 2-CO2Et indoles, it was discovered that
the yield of the reaction improved as the substituent became more electron withdrawing. This result was consistent with a proposed
mechanism involving benzenethiol displacement of an intermediate 3-IPh indole complex.
Ó 2004 Elsevier Ltd. All rights reserved.
For an anti-inflammatory program targeting potent and
selective cyclooxygenase-2 (COX-2) inhibitors, we re-
quired a mild general method for the one-pot 3-aryl-
thiolation of indoles containing a 6-methylsulfonyl
moiety. Usually 3-arylthioindoles are prepared from the
Fisher indole synthesis,1 though such methodology
proved ineffective for our target molecules. An approach
involving direct 3-arylthiolation of a suitably substituted
indole would be optimal in terms of the efficiency of
analogue generation. Traditional methods for the direct
3-arylthiolation of indoles include alkylation of a
disulfide with the C-3 organozinc2a or organosodium2b
salt of an indole, the acid catalyzed alkylation of an
indole with either mono-O,S-acetals2c;d or sulfenyl
chlorides,2e;f and the alkylation of indoles via a sulfe-
nium ion promoted Pummerer rearrangement.2g These
methods, however, were insufficient to meet our needs
either because the 6-methylsulfonyl moiety readily eno-
lized or because the method was impractical for ana-
logue generation.
complex, which subsequently leads to the formation of a
radical cation via single electron transfer (SET). The
resulting radical cation would then react with a benz-
enethiol to form the desired ortho or para substituted
arylthio anisole and iodobenzene. One of the primary
driving forces of this reaction is reduction of the iodine
moiety of PIFA from I(III) to the I(I) oxidation state. It
occurred to us that electron rich heterocycles such as
indole may be amenable to a similar thioaryl substitu-
tion reaction. Thus, an indole could also form a CT-
complex with PIFA. This complex would subsequently
lead to a cationic 3-(I–Ph) indole complex, which could
undergo a arylthio substitution and reductive elimina-
tion sequence to form 3-arylthioindoles such as 2
(Scheme 1).3d;e For our COX-2 program, the synthesis of
1A (R1 ¼ Me) and 1B (R1 ¼ CO2Me) would be required
in order to effect a transformation to the desired
potential COX-2 inhibitors 2. We have recently
described the synthesis of 6-methylsulfonyl indoles 1A
and 1B using established methodology.4–6
We thus turned our attention to the use of the hyper-
valent iodine reagent PIFA,3 reported for the arylthio-
lation of anisole.3a In the reported mechanism, PIFA
reacts with the anisole to form a charge-transfer (CT)
We were pleased to find that indoles 1A and 1B react
with a wide range of benzenethiols containing either
electron releasing or electron withdrawing substituents
(Table 1) in good to excellent yields. The scope of this
reaction is shown by the examples summarized in Table
1. It is of interest to note that the 2-pyridylthio analogue
2e was prepared in 76% yield. Oxidation of the pyridine
moiety was not observed. It is apparent that the elec-
tronic nature of the benzenethiol moiety had little effect
upon the yield. The presence of a 2-CO2Me versus a
Keywords: PIFA; 3-Arylthioindole; Selective COX-2 inhibitors; Benz-
enethiol.
* Corresponding author. Tel.: +1-908-222-7000; e-mail: jcampbell@
ptcbio.com
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.03.153

4074
J. A. Campbell et al. / Tetrahedron Letters 45 (2004) 4073–4075
-
OCOCF3
OCOCF3
OCOCF3
+
Ph
+
-
SAr
I
IPh
PIFA
ArSH
PhI
R1
R1
R1
R1
N
H
N
N
H
(CF3)2CHOH
MeO2S
N
H
MeO2S
H
3-(I-Ph)-
CT-COMPLEX
1
2
COMPLEX
Scheme 1. Proposed mechanism of formation of 2 from 1 using PIFA and ArSH.
Table 2. 3-Arylthioindole mechanism study
2-Me moiety on the indole, however, appeared to im-
prove the yield of the reaction. To explore the potential
mechanistic implications of this result, a study was
conducted varying the electronic nature of the 5-sub-
stituent of 2-carboalkoxy indole. It was discovered that
the yield of the reaction improved as the substituent
became more electron withdrawing (Table 2, products
4a–f). This would apparently rule out a mechanism
involving nucleophilic attack of the indole on a PIFA
derived thioaryl electrophile. Replacement of the indole
N–H of 5-nitro-2-carboalkoxyindole with an N–Me
showed no improvement in yield (compare 4a to 4e), but
demonstrated that protection of the indole NH was not
required.
F
S
F
(2,4-DiF)PhSH
R3
R3
PIFA
R1
R1
N
(CF3)2CHOH
N
R2
R2
3
4
Product-4
R1
R2
R3
Yield (%)a
4a
4b
4c
4d
4e
4f
CO2Et
CO2Et
CO2Et
CO2Et
H
H
H
H
NO2
Cl
89
50
29
8
H
OCH3
NO2
Cl
CO2Me
CO2Me
Me
Me
83
58
In summary, the direct 3-thioarylation of indoles using
phenyliodine(III)bis trifluoroacetate (PIFA) in (CF3)2
CHOH with a wide variety of benzenethiols has been
accomplished. Under these mild conditions, it was
demonstrated that indoles bearing a 6-MeSO2 and either
a 2-methyl or 2-carboxymethyl substituent could be
3-thioarylated in good to excellent yields to afford the
corresponding 3-arylthioindoles as selective COX-2
inhibitors.6 The biological data for these compounds
and other homologues will be reported elsewhere. In a
study varying the electronic nature of the 5-substituent
of 2-carboalkoxyindole, it was discovered that the yield
a Isolated yield using 2.7 equiv (2,4-DiF)PhSH and 2.0 equiv PIFA.
of the reaction improved as the substituent became more
electron withdrawing. This result was consistent with a
proposed mechanism involving benzenethiol displace-
ment of an intermediate 3-IPh indole complex.
References and notes
1. Hamel, P.; Girard, Y. Chem. Ber. 1963, 96, 260.
2. (a) Browder, C. C.; Mitchell, M. O.; Smith, R. L.;
Sulayman, G.-E. Tetrahedron Lett. 1993, 34, 6245; (b)
Atkinson, J. G.; Hamel, P.; Girard, Y. Synth. Commun.
1988, 6, 480; (c) Matsugi, M.; Murata, K.; Nambu, H.;
Kita, Y. Tetrahedron Lett. 2001, 42, 1077; (d) Matsugi, M.;
Murata, K.; Gotanda, K.; Nambu, H.; Anilkumar, G.;
Matsumoto, K.; Kita, Y. J. Org. Chem. 2001, 66, 2434; (e)
Anzai, K. J. Heterocycl. Chem. 1979, 16, 567; (f) Raban,
M.; Chern, L. J. J. Org. Chem. 1980, 45, 1688; (g) Jain, S.;
Shukla, A.; Mukhopadhyay, A.; Suryawanshi, S. N.;
Bhakuni, D. S. Synth. Commun. 1990, 20, 1315.
3. (a) Kita, Y.; Takada, T.; Tohma, H. Pure Appl. Chem.
1996, 68, 627; (b) Pohnert, G. J. fuer. Praktishe Chemie.
2000, 342, 731; (c) Varvoglis, A. Tetrahedron 1997, 53,
1179; (d) Moriarty, R. M.; Vaid, R. K. Synthesis 1990, 431;
(e) Moriarty, R. M.; Ku, Y. Y.; Sultana, M.; Tuncay, A.
Tetrahedron Lett. 1987, 28, 431.
4. Allais, A. Eur. J. Med. Chem. Chim. Ther. 1975, 10, 187.
5. (a) Adams, R. E.; Press, J. P.; Deegan, E. G. Synth.
Commun. 1991, 12, 675; (b) Semerth, S. J. Heterocycl.
Chem. 1981, 18, 1373.
6. Broka, C. A.; Campbell, J. A. (F. Hoffmann-La Roche Ag,
Switz.). PCT Int. Appl. WO 0329212, 2003; p 46.
7. General PIFA promoted indole 3-thioarylation proce-
dure: Preparation of 2h, 3-(2,4-difluorophenyl-sulfanyl)-6-
methanesulfonyl-1H-indole-2-carboxylic acid methyl ester.
Table 1. 3-Arylthioindole syntheses
SAr
ArSH
PIFA
R1
R1
N
H
N
H
MeO2S
MeO2S
(CF3)2CHOH
1A R1=Me
1B R1=CO2Me
2
Product-2
R1
Ar
Yield (%)
2a
2b
2c
2d
2e
2f
Me
Me
4-F(Ph)
2,4-DiF(Ph)
2-Cl(Ph)
76a
67a
64a
52a
76a
62a
83b
88b;c
74b
84b
74b
74b
Me
Me
2-Cl,4-F(Ph)
2-Pyridyl
2-F(Ph)
Me
Me
2g
2h
2i
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
4-F(Ph)
2,4-DiF(Ph)
2-Cl(Ph)
2j
4-Me(Ph)
4-EtO(Ph)
4-MeO(Ph)
2k
2l
a Isolated yield using 2 equiv ArSH and 1.5 equiv PIFA.
b Isolated yield using 2.7 equiv ArSH and 2.0 equiv PIFA.
c Experimental example provided (see Ref. 7).

J. A. Campbell et al. / Tetrahedron Letters 45 (2004) 4073–4075
4075
To a stirred solution of 2-carbomethoxy-6-methylsulfony-
lindole (750 mg, 2.8 mmol) in dry (CF3)2CHOH (10 mL)
under Ar was added 2,4-difluorothiophenol (1.0 g,
7.56 mmol), followed by PIFA (2.40 g, 5.60 mmol) at room
temperature. After stirring overnight, the dark solution was
partitioned between CH2Cl2 (100 mL) and NH4OH
(10 mL). The CH2Cl2 layer was washed with more NH4OH
(3ꢀ10 mL), water (30 mL), dried (MgSO4) and concen-
trated to ꢁ20 mL volume. To this mixture was then added
30 mL of hexane, which precipitated the 977 mg (88%) of
the product, 3-(2,4-difluorophenylsulfanyl)-6-metha-
nesulfonyl-1H-indole-2-carboxylic acid methyl ester, 2h,
as a white solid: mp 202.5–203 °C; 1H NMR (CDCl3): d
3.09 (s, 3H), 3.97 (s, 3H), 6.72–7.04 (m, 3H), 7.63 (d, 1H,
J ¼ 8:5 Hz), 7.78 (d, 1H, J ¼ 8:5 Hz), 8.21 (s, 1H), 11.83 (s,
1H); 13C NMR (CDCl3) d 49.9, 57.5, 108.9, 109.3, 109.6,
117.1, 117.2, 119.1, 123.9, 127.0, 136.6, 136.7, 138.0, 138.2,
140.4, 142.2, 166.0; MS m=e 397 (Mþ); Anal. Calcd for
C17H13F2NO4S2.0.60 H2O: C, 50.02; H, 3.51; N, 3.43.
Found: C, 50.01; H, 3.21; N, 3.49.