Novel 5-aryl-1,3-dihydro-indole-2-thiones: Potent, orally active progesterone receptor agonists

Article abstract of DOI:10.1016/S0960-894X(03)00129-X

During the course of our studies on 3,3-disubstituted-5-aryloxindoles derived progesterone receptor (PR) antagonists we discovered that changing the amide funtionality to a thio-amide resulted in compounds displaying potent PR agonist activity. In this communication, the synthesis, structure activity relationships (SAR) and in vivo activity of various 5-arylthio-oxindoles will be discussed.

Full text of DOI:10.1016/S0960-894X(03)00129-X

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1317–1320
Novel 5-Aryl-1,3-dihydro-indole-2-thiones:
Potent, Orally Active Progesterone Receptor Agonists
Andrew Fensome,a,* Marci Koko,a Jay Wrobel, Puwen Zhang,a Zhiming Zhang,b
Jeffrey Cohen,b Scott Lundeen,b Kelly Rudnick,b Yuan Zhub and Richard Winnekerb
aChemical Sciences, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA
bWomen’s Health Research Institute, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA
Received 20 September 2002; accepted 16 January 2003
Abstract—During the course of our studies on 3,3-disubstituted-5-aryloxindoles derived progesterone receptor (PR) antagonists we
discovered that changing the amide funtionality to a thio-amide resulted in compounds displaying potent PR agonist activity. In
this communication, the synthesis, structure activity relationships (SAR) and in vivo activity of various 5-arylthio-oxindoles will be
discussed.
# 2003 Elsevier Science Ltd. All rights reserved.
The progesterone receptor (PR) is a member of the
steroid receptor sub-family of the nuclear hormone
receptor super-family, a group of nuclear transcription
factors (Chart1 ).1 Progesterone 1, the endogenous
ligand for the PR, is involved in the control of ovulation
and preparation of the uterus for implantation and
maintenance of pregnancy. Clinically, progestin ago-
nists are mainly used in contraception and hormone
replacement therapy, typically co-administered with an
estrogen.
progesterone. One of the most studied is medroxy-
progesterone acetate (MPA) 2.2
More recently, increasing interest has focussed on iden-
tifying and characterizing non-steroidal PR agonists,3
(Chart2 ).
Chart 1. Steroid PR agonists.
Over the last few decades considerable synthetic effort
has been aimed towards preparing new steroidal PR
agonists which are more potent and bioavailable than
Chart 2. Non-steroidal PR agonists.
During the course of SAR studies with oxindole derived
PR antagonists I,4 we discovered that transposing the
oxindole carbonyl group to a thio-carbonyl moiety II
*Corresponding author. Fax: +1-484-865-9398; e-mail: fensoma@
wyeth.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00129-X

1318
A. Fensome et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1317–1320
turned the molecules into potent selective PR agonists
(Scheme 1).
Moving the 30-chlorine functionality to the 40-position
also caused a severe decline in potency (EC50>2000
nM, 11), whereas the 20-chloride, 10, was only
slightly weaker (EC50=5.3 nM) than the parent thio-
oxindole 8.
Chemistry
The target thio-oxindole derivatives II, were synthesized
by treatment of the parent oxindole antagonists I with
Lawesson’s reagentin refluxing otluene ( Scheme 1).
Reactions were carried out in parallel in sealed tubes on
a 100 mg scale. The products were then isolated by
normal phase HPLC.
The 3,3-dialkyl substitution pattern on the thio-oxindole
core was examined next. Replacing the spirocyclohexyl
group with a 3,3-dimethyl functionality 12 had little
effect on potency, whereas the 3,3-diethyl derivative 13
was 100 fold weaker (EC50=5.3 nM and 200 nM,
respectively). There is an apparent trend in ring size in
the spirocycle series, thus the spirocylobutane 14 was the
leastpoetnt(EC 50=28.7 nM), the spirocyclopentyl 15
was intermediate (EC50=9.8 nM) and the spiro-
cyclohexyl derivative was the most potent (EC50=1.6
nM, 8). In the case of the mono-methyl derivative 16,
NMR evidence suggested that the isomeric 2-mercapto-
indole moiety was the observed tautomer and this may
account for its very weak activity (T47D alkaline phos-
phatase EC50>3000 nM). The 3-methyl-3-benzyl deri-
vative was also poor (EC50=1200 nM).
The compounds were evaluated for PR agonist
activity based on their ability to induce alkaline
phosphatase activity in the human breast cancer cell
line T47D, Table 1.5 As previously mentioned
transposing the amide oxygen of oxindole 7 t o a
sulfur atom dramatically switched the activity of the
molecule from antagonist (IC50=56 nM, 7) to ago-
nist(EC 50=1.6 nM, 8), with excellent potency. We
then surveyed the basic substitution pattern of
compound
template.
8 to optimize the core thio-oxindole
Substituting the N-H of the thioamide moiety for an
N-methyl group 9, caused a dramatic loss of potency
(T47D alkaline phosphatase EC50 >3000 nM).
Scheme 1.
Table 1. T47D cell alkaline phosphatase assay data for compounds 2 and 7–17
Compd
R1
R2
R3
Chlorine substituent
PR Alkaline phosphatase4 EC50 (nM)a
2, mpa
7, oxindole
8
0.5
IC50=56
1.6
>3000
5.3
2000
5
200
28.7
9.8
>3000
1200
–CH2CH2CH2 CH2CH2–
–CH2CH2CH2 CH2CH2–
–CH2CH2CH2 CH2CH2–
–CH2CH2CH2 CH2CH2–
–CH2CH2CH2 CH2CH2–
H
H
Me
H
H
H
30
30
30
20
40
300
30
30
30
30
9
10
11
12
13
14
15
16
17
Me
EHtEt
Me
3
–CH2CH2CH2–
–CH2CH2CH2CH2–
Me
Me
H
H
H
H
H
Bn
a50% effective concentration of tested compounds on alkaline phosphatase activity in the human T47D breast carcinoma cell line. Values represent
the average of at least duplicate determinations. The standard deviations for these assays were typicallyꢀ15% of mean or less.

A. Fensome et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1317–1320
1319
With the spirocyclohexyl thio-oxindole as the template,
we then turned our attention to the 5-aryl group, which
was examined next, Table 2. Replacing the 30-chlorine
atom of thio-oxindole 8 with a fluorine atom to afford
20 had no effecton potency (EC 50=1.6 nM, 8 and 20).
Similarly 3-hydroxyl 21 and 3-cyano 22 analogues were
also very potent (EC50=2.5 nM and 2.0 nM respec-
tively), however the 3-biphenyl derivative 23 was much
weaker (EC50=71 nM).
nM). In the isomeric thiophen-5-yl series, the effect of
substitution at the 40-position was explored. Thus the
parent2 -cyano-5-thiophene 35 had potent activity in
0
the T47D assay (EC50=1.0 nM). Adding a methyl
group in the 4-position 36 slightly increased potency
(EC50=0.5 nM), but going to larger substituents
reduced potency (n-propyl 37 EC50=7.7 nM and n-
butyl 38 EC50=23.5 nM). Replacing the cyano group in
compound 35 with a chloride 39 had little effect on
activity (EC50=1.4 nM). Also the furan 40 was nearly
equipotent with thiophene 35 (EC50=1.9 nM).
Based on analogue 11, compounds having a single sub-
stituent at the 40-position appeared to result in sub-
stantially reduced activity. However, adding a second
substituent to the 40-position of compound 8 (in the case
of 24, a fluorine atom) resulted in a compound with
only slightly reduced activity over 8 (T47D alkaline
phosphatase EC50=4.1 nM). Thus substitution at the 40
position is only deleterious if a suitable 30-substituent is
not present. Moving the fluorine to the 50-carbon gave a
small increase in potency (EC50=0.5 nM, 25). The same
trend is also true for the 30-fluoro 20 and 30-cyano 22
derivatives. Thus adding a 40-F to compound 20 or 22
Some of the more potent molecules were tested in rat
decidualization and uterine C3 models for progesta-
tional activity, Table 4.6 Compounds were dosed
orally, in a 2% Tween 80/0.5% methyl cellulose aque-
ous vehicle.
In the decidualization model, the parent thio-oxindole 8
was inactive at the highest dose tested, 3 mg/kg/day.
However the 30-cyano analogue 22 demonstrated
potency equivalent to MPA (EC50=0.4 mg/kg). The
30-cyano-50-fluorophenyl congener 29 was the most
potent in the aryl series (decidual EC50=0.1 mg/kg). In
the heteroaryl series, the 20-cyano-50-thienyl thio-oxi-
ndole 35 was an order of magnitude more potent than
it’s 40-methyl substituted analogue 36 (EC50=0.1 mg/kg
and 1 mg/kg, respectively). The furan 40 was also 10
fold weaker than the parent thiophene 35 (ratdecidual
EC50=1 mg/kg).
caused only
a minor reduction in potency (26
EC50=5.4 nM and 28 EC50=4.2 nM). The 30,
50-difluorophenyl derivative 27 was a little more potent
than the parent fluoride 20 (EC50=0.8 nM) and the
30-cyano-50-fluorophenyl derivative 297 was the most
potent in the aryl series (EC50=0.36 nM). Reversing
the cyano and fluorine functionality in compound 28 to
afford the 30-fluoro-40-cyanophenyl derivative 31 caused
a drop in activity (EC50=14.8 nM). Replacing the
cyano group in compound 29 with a methoxy, to afford
the congener 32, resulted in nearly a 10 fold drop in
potency (EC50=2.6 nM).
The ratcomplimentcomponentC3 assay measures ht e
ability of a progestin agonist to block the estrogen sti-
mulated expression of C3 mRNA.6 The parenthtio-
oxindole 8 and cyano derivative 22 were an order of
magnitude less potent than MPA (C3 IC50=0.37, 0.3
Incorporating a thiophene ring in the 5-position of the
thio-oxindole platform gave opportunities for different
substitution patterns. The 20-cyano-thiophen-4-yl deri-
vative 338 was as active as the most potent phenyl com-
pound 29 (EC50=0.3 nM), Table 3. Exchanging the
sulfur in compound 33 with an oxygen atom gave the
furan 34 which was about5 fold weaker (EC 50=1.9
Table 3. T47D cell alkaline phosphatase assay data for compounds
33–40a
Compd
X
R1
R2
Alkaline phosphatase4 EC50 (nM)a
33
34
35
36
37
38
39
40
S
O
S
S
S
S
S
O
0.3
1.9
1.0
0.5
7.7
23.5
1.4
1.9
CN
CN
CN
CN
Cl
H
Me
n-Pr
n-Bu
H
Table 2. T47D cell alkaline phosphatase assay data for compounds
20–32
Compd R3
R4 R5 R6 PR Alkaline phosphatase4 EC50 (nM)a
CN
H
20
21
22
23
24
25
26
27
28
29
30
31
32
F
OH
CN
Ph
Cl
Cl
F
F
CN
CN
CN
F
H
H
H
H
F
H
F
H
F
H
H
CN
H
H
H
H
H
H
F
H
F
H
F
H
H
H
H
H
H
H
H
H
H
F
1.6
2.5
2.0
71.4
4.3
0.5
5.4
0.8
4.2
a50% effective concentration of tested compounds on alkaline phos-
phatase activity in the human T47D breast carcinoma cell line. Values
represent the average of at least duplicate determinations. The stan-
dard deviations for these assays were typically ꢀ15% of mean or less.
Table 4. Ratdecidual and C3 daat for MPA and compounds 8, 22,
29, 35, 36 and 40
0.36
H
H
F
1.4
14.8
2.6
H
H
Compd
MPA
8
22 29
35 36 40
0.1
MeO
Decidualisationa ED50 (mg/kg) 0.4
C3b IC50 (mg/kg)
0.026
>3 0.4 0.1
0.37 0.3 0.025 0.03
1
1
a50% effective concentration of tested compounds on alkaline phos-
phatase activity in the human T47D breast carcinoma cell line.Values
represent the average of at least duplicate determinations. The stan-
dard deviations for these assays were typicallyꢀ15% of mean or less.
a50% effective dose of test compounds in the rat decidualization.
bC3 models.

1320
A. Fensome et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1317–1320
Table 5. Cross-reactivity data for MPA and 29
Acknowledgements
Compd
AR EC50
(nM)
AR IC50
(nM)
GR EC50
(nM)
GR IC50
(nM)
The authors gratefully appreciate the work of the Dis-
covery Analytical Chemistry department at Wyeth
Research, Collegeville, PA for spectroscopic data.
MPAa
29b
6.1
>10000
10
>10000
>10000
37
aData from ref 9.
References and Notes
bExperimental values represent the mean of at least duplicate deter-
minations. The standard deviation was typically +/ꢂ15% of mean.
1. Mangelsdorf, D. J.; Thummel, C.; Beato, M.; Herrlich, P.;
Schutz, G.; Umesono, K.; Blumberg, B.; Kastner, P.; Mark,
M.; Chambon, P.; Evans, R. M. Cell 1995, 83, 835.
2. (a) Jeppsson, S. Acta Obstet. Gynecol. Scand., Suppl. 1981,
101, 7. (b) O’Connell, B. J. Current Opinion In Pediatrics 1995,
7, 371.
3. (a) Zhang, P.; Terefenko, E. A.; Fensome, A.; Zhang, Z.;
Zhu, Y.; Cohen, J.; Winneker, R.; Wrobel, J.; Yardley, J.
Bioorg. Med. Chem. Lett. 2002, 12, 787. (b) Zhi, L.; Tegley,
C. M.; Marschke, K. B.; Mais, D. E.; Jones, T. K. J. Med.
Chem. 1999, 42, 1466. (c) Palmer, S.; Campen, C. A.; Allan,
G. F.; Rybczynski, P.; Haynes-Johnson, D.; Hutchins, A.;
Kraft, P.; Kiddoe, M.; Lai, M.-T.; Lombardi, E.; Pedersen, P.;
Hodgen, G.; Combs, D. W. J. Steroid Biochem. Mol. Bio.
2001, 5, 33.
and 0.026 mg/kg respectively). As with the decidualisa-
tion data, the 30-cyano-50-fluorophenyl derivative 29
and the 20-cyano-50-thienyl thio-oxindole 35 were the
most potent in the series (IC50=0.025 and 0.03 mg/kg)
and possessed efficacy and potency similar to MPA.
Compound 29 was evaluated in the PR competition
binding assay using human T47D cell cytosol in the
3
presence of 3 nM H-R5020 as the radioligand.5 In this
4. (a) Fensome, A.; Bender, R.; Cohen, J.; Collins, M. A.;
Mackner, V. A.; Miller, L. L.; John, W.; Ullrich, J. W.; Win-
neker, R. C.; Wrobel, J.; Zhang, P.; Zhang, Z; Zhu, Y.;
Bioorg. Med. Chem. Lett., submitted for publication. (b) Fen-
some, A.; Bender, R.; Collins, M.; Fletcher, H., III; Mackner,
V.; Miller, L. L.; Ullrich, J.; Wrobel, J.; Zhang, P.; Zhang, Z.;
Zhu, Y.; Cohen, J.; Winneker, R.: Abstracts of Papers, 223rd
ACS National Meeting, Orlando, FL, USA, 7–11 April, 2002
(2002), MEDI-068.
assay thio-oxindole 29 had an IC50=15.9 nM. In com-
parison MPA 2 has an IC50=10.8 nM. Compound 29
was also evaluated for its androgen receptor (AR)
(mouse fibroblastL929 cell line) and glucocorictoid
receptor (GR) (human lung A549 cell line) activity.6 Itis
clear from Table 5 that compound 29 has no AR ago-
nist or antagonist activity. It does however have GR
antagonist activity.
5. (a) Zhang, Z.; Lundeen, S. G.; Zhu, Y.; Carver, J. M.;
Winneker, R. C. Steroids 2000, 65, 637. (b) Beck, C. A.; Wei-
gel, N. L.; Moyer, M. L.; Nordeen, S. K.; Edwards, D. Proc.
Natl. Acad. Sci. U.S.A. 1993, 90, 4441.
6. Lundeen, S. G.; Zhang, Z.; Zhu, Y.; Carver, J. M.; Winne-
ker, R. C. J. Steroid Biochem. Mol. Bio. 2001, 78, 137.
7. 3-(1,2-Dihydro-2-thioxospiro[cyclohexane-1,3-[3H]indol]-5-
yl)-5-fluorobenzonitrile 29: mp. 236–250 ꢁC; 1H NMR
(CDCl3) d 10.05 (s, 1H), 7.85 (d, 1H, J=1 Hz), 7.65 (s, 1H),
7.46–7.52 (m, 2H), 7.33–7.36 (m, 1H), 7.16 (d, 1H, J=8.1 Hz),
1.86–2,18 (m, 7H) and 1.54–1.66 (m, 3H); MS ((+)-APCI) m/z
337 [M+H]+.
8. (10,20-Dihydro-20-thioxospiro[cyclohexane-1,30-[3H]indol]-50-
yl)-2-thiophenecarbonitrile 35: mp 230–232 ꢁC; 1H NMR
(DMSO-d6) d 12.82 (s, 1H), 8.0–7.98 (m, 2H), 7.73 (d, 1H,
J=4 Hz), 7.69 (dd, 1H, J=8.2 and 1.5 Hz), 7.10 (d, 1H,
J=8.2 Hz), 1.98–1.77 (m, 7H), 1.43–1.33 (m, 3H); MS (EI)
M+ @ m/z 324.
9. Zhi, L.; Tegley, C. M.; Kallel, E. A.; Marschke, K. B.;
Mais, D. E.; Gottardis, M.; Jones, T. K. J. Med. Chem. 1998,
41, 291.
In conclusion, we have developed a new series of potent,
orally active non-steroidal progesterone receptor ago-
nists, based upon the 5-aryl thio-oxindole template.
Initial SAR studies settled on the spirocyclohexyl sub-
stitution pattern for the 3,3-dialkyl substituent on the
thio-oxindole backbone. Although the template was
relatively insensitive to the nature of the 30-substituent
on the 5-aryl group, an additional fluorine substituent
enhanced both in vitro and in vivo potency. A variety of
heteroaromatic groups were also explored and the
20-cyanothien-5-yl moiety was found to be the most
potent.
The 30-cyano-50-fluorophenyl 29 and 20-cyano-thien-50yl
35 spirocyclohexylthio-oxindoles were found to be
equipotent to the steroidal progestin agonist MPA in
the rat decidualisation and C3 models.