Synthesis and activity of substituted 4-(indazol-3-yl)phenols as pathway-selective estrogen receptor ligands useful in the treatment of rheumatoid arthritis

Article abstract of DOI:10.1021/jm049194+

Pathway-selective ligands for the estrogen receptor (ER) inhibit NF-κB-mediated inflammatory gene expression causing a reduction of cytokines, chemokines, adhesion molecules, and inflammatory enzymes. SAR development of a series of 4-(indazol-3-yl)phenols

Full text of DOI:10.1021/jm049194+

© Copyright 2004 by the American Chemical Society
Volume 47, Number 26
December 16, 2004
Letters
production of inflammatory mediators in synoviocytes
and macrophages, likely involving the transcription
factor nuclear factor κB (NF-κB). NF-κB is composed of
a homodimeric and heterodimeric family of the Rel
family of proteins which resides in the cytoplasm in a
nonactive form bound to the cytoplasmic inhibitory
protein-κB (IκB). Cell activation by a variety of extra-
cellular inflammatory cytokines induces a cascade of
events that leads to degredation of the IκB/NFκB com-
plex. Activated NF-κB then translocates to the nucleus
where it binds to DNA elements resulting in the
induction of pro-inflammatory genes such as TNFR,
IL-6, etc.1 Two recent discoveries indicate that inhibition
of this cellular pathway provides a novel intervention
point in the control of the inflammatory process. First,
it was found that overproduction of IL-6 is associated
with states of chronic inflammation.2 Second, in endot-
helial cells it was shown that 17â-estradiol (E2) inhibits
IL-1â-induced NF-κB reporter activity and IL-6 expres-
sion in an estrogen receptor (ER) dependent fashion.3
The latter correlates with the antiinflammatory action
of E2 in vivo which is confirmed in different animal
models of inflammation.4 For a review of recent ad-
vances in ER ligands as a drug class, see ref 5.
A drug discovery project was initiated to confirm that
the ER/ligand complex mimics the antiinflammatory
action of E2 that occurs in chronic conditions such as
RA. The ultimate goal of the project was to identify a
small molecule nonsteroidal ligand for ER which will
mediate the antiinflammatory effect without eliciting
the classical ER responses.
Compounds were screened in vitro using an assay
developed in HAECT-1 cells (immortalized human aortic
endothelial cells).1 HAECT-1 cells are transfected with
human ERR and a reporter gene NFκB-luciferase. These
cells are then treated for 16-18 h with IL-1â and the
test compound. The level of transcription of NF-κB is
directly proportional to the amount of luciferase present.
The ideal profile for the test compound would be an IC50
of NF-κB luciferace reporters of 100 nM or less with an
efficacy (% E2) near 100% coupled with a lack of
classical estrogenic receptor activity. Classical estro-
Synthesis and Activity of Substituted
4-(Indazol-3-yl)phenols as
Pathway-Selective Estrogen Receptor
Ligands Useful in the Treatment of
Rheumatoid Arthritis
Robert J. Steffan,*,† Edward Matelan,†
Mark A. Ashwell,‡,X William J. Moore,†
William R. Solvibile,† Eugene Trybulski,†
Christopher C. Chadwick,†, Susan Chippari,‡
Thomas Kenney,‡ Amy Eckert,‡
Lisa Borges-Marcucci,‡ James C. Keith,§ Zhang Xu,|
Lydia Mosyak,| and Douglas C. Harnish‡
Chemical and Screening Sciences and Cardiovascular/
Metabolic Diseases, Wyeth Research, 500 Arcola Road,
Collegeville, Pennsylvania 19426, and Chemical
and Screening Sciences, Wyeth Research,
Cambridge, Massachusetts 02140
Received October 6, 2004
Abstract: Pathway-selective ligands for the estrogen receptor
(ER) inhibit NF-κB-mediated inflammatory gene expression
causing a reduction of cytokines, chemokines, adhesion mol-
ecules, and inflammatory enzymes. SAR development of a
series of 4-(indazol-3-yl)phenols has led to the identification
of WAY-169916 an orally active nonsteroidal ligand with the
potential use in the treatment of rheumatoid arthritis without
the classical proliferative effects associated with estrogens.
Rheumatoid arthritis (RA) is characterized by chronic
joint inflammation mediated by inflammatory cell in-
filtration into synovial tissues as well as joint destruc-
tion through matrix metalloproteinase (MMP) overex-
pression in articular synoviocytes and chondrocytes. The
pathologic lesions of RA are driven, in part, by the
* Corresponding author: phone (484)865-4648, fax (484) 865-9398,
e-mail steffar@wyeth.com.
† Chemical and Screening Sciences, Wyeth Research, Collegeville.
‡ Cardiovascular/Metabolic Diseases, Wyeth Research, Collegeville.
§ Cardiovascular/Metabolic Diseases, Wyeth Research, Cambridge.
| Chemical and Screening Sciences, Wyeth Research, Cambridge.
Current address: Life Diagnostics Inc., West Chester PA 19380.
X Current address: ArQule, 19 Presidential Way, Woborn, MA
01741.
10.1021/jm049194+ CCC: $27.50 © 2004 American Chemical Society
Published on Web 11/12/2004

6436 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 26
Letters
Table 1. Effects of 1-Substituted Analogues of Templates 4
genic activity is characterized by stimulation of uterine
proliferation, ER-mediated gene expression, and for our
in vitro model, a lack of creatine kinase (CK) activity.
4-(Indazol-3-yl)phenols were identified as potent non-
steroidal estrogen receptor (ER) ligands that selectively
inhibit NF-κB transcriptional activity but are devoid of
conventional estrogenic activity. To confirm that these
ligands for ER were devoid of the proliferative effects
on uterine and breast tissue associated with estrogen,
the compounds were tested in an uterine wet weight
model in female mice in vivo. No increase in uterine wet
weight after dosing for 5 days would confirm a lack of
estrogen receptor activity.
and 8 vs E2 on NF-κB, and CK Expression in
Ad5-wt-ER-Infected HAECT-1 Cellsa
ER/NFκB-luc
IC50, nM
CKEC50
,
compd
R1
R
%E2
nM
%E2
1
6-OH
H
48
622
305
165
62
82
64
71
71
71
49
63
75
72
73
72
87
33
60
73
298
3715
3070
4790
1217
549
iab
149
140
ia
43
50
97
66
33
27
4a 6-OH
4b 6-OH
methyl
The initial indazole lead 1 was identified by high
throughput screening of selected corporate libraries of
compounds.
ethyl
4c
6-OH
CF3CH2
4d 6-OH
n-propyl
4e
4f
6-OH
6-OH
n-butyl
112
443
40
cyclohexyl
4-CH3-phenyl
3-F-phenyl
3-Cl-phenyl
2,5-diCl-phenyl
2,5-diF-phenyl
4g 6-OH
4h 6-OH
69
47
55
39
4i
4j
6-OH
6-OH
50
15
ia
ia
4k 6-OH
31
8a 5-F, 6-Cl benzyl
sdc 19%
105
196
ia
8b 7-Cl
benzyl
390
Compound 17 was confirmed to have strong inhibition
of the NF-κB transcription [IC50 ) 48 nM (82% E2)] in
the ER/NF-κB-luc assay and weakly active in the CK
stimulation activity. Selected libraries of compounds
were prepared via Schemes 1 and 2 to develop the initial
SAR. The acceptable compound purity for the arrays
was >85% by HPLC and exact mass. Confirmation of
activity was conducted using analytically pure dry
powder.
8c
7-CF3
benzyl
8d 6-CF3
benzyl
8e
8f
7-CF3
2-CH2CH2OH
96
ia
69
5-F,6-Cl 2-CH2CH2OH
8g 7-Cl
8h 7-Cl
2-CH2CH2OH
butyl
410
25
71
90
63
a All compounds were ER dependent (only active when ER is
coexpressed with NFkB-luciferase in the HAECT cells). b ia )
inactive. c sd ) % inhibition at single dose of 10 uM %E2 )
efficacy (relative inhibition of test compound at 10 uM vs E2 at
0.1 nM).
Commercially available 2,2′,4,4′-tetrahydroxyben-
zophenone 2 was combined with a variety of 1-substi-
tuted hydrazines in methanol to form the intermediate
hydrazone 3. Heating the intermediate to 200 °C af-
forded the appropriate 1-substituted indazole 4.
Alternatively, the Weinreb amide 5 was prepared
using standard conditions from commercially available
appropriately substituted 2-fluorobenzoic acids.8 The
Weinreb amide was reacted with the methoxy-bearing
phenylmagnesium bromides. The resulting benzophe-
nones 6 were reacted with the 1-substituted hydrazines
in the same manner as in Scheme 1 to give the indazole
7 (R2 ) OCH3). The free hydroxyl derivative 8 (R2 )
OH) was liberated by removal of the methyl group with
excess boron tribromide in methylene chloride.9 Shown
in Table 1 are some of the most active analogues.
Scheme 3a
a (a) NH2NH2‚H2O, DMAP, pyridine, 100 °C; (b) NaH, RX, DMF,
60 °C; (c) BBr3.
Within the trihydroxy series 4a-k the most potent
analogues 4g, 4i, 4j, and 4k were evaluated in vivo for
their ability to inhibit the mRNA of NF-κB target genes
MHC, VCAM-1, RANTES, and TNF-R in C57BL/6 mice
subjected to a high fat diet for 5 weeks.10 All of the
analogues tested in the model produced no detectable
plasma levels, were inactive in the model, and were not
pursued.
Scheme 1a
a (a) RNHNH2, NaOAc, MeOH; (b) 200 °C, Neat.
The compounds lacking a hydroxyl group at R1 from
Scheme 2 were used to guide future analogue develop-
ment. A representative group of compounds 8a-h
illustrate the important positions for SAR development.
Compounds 8b and 8c illustrate the importance of the
7 position. The only compounds from the array of
3-indazolylphenolic and nonphenolic analogues to show
activity had the hydroxyl in the para-postion shown in
template 8.
Scheme 2a
a (a) R2PhMgBr; (b) RNHNH2 then BBr3.

Letters
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 26 6437
Table 2. Effects of R, R1, and R2 Indazole Substitution vs E2
on NF-κB, and CK Expression in Ad5-wt-ER-Infected HAECT-1
Cellsa
ER/NFκB-luc
IC50, nM
CK EC50
nM
,
compd R2
R1
R
%E2
%E2
11a
11b
11c
11d
11e
11f
H
H
H
H
H
H
CH3 propyl
40
3
89
91
303
11
138
iab
98
330
ia
61
44
34
CF3 isopropyl
Figure 1. Dose-response of orally administered WAY-169916
Cl
Cl
Cl
F
propyl
52
37
18
18
93
59
20
77
in the Lewis rat model of adjuvant-induced arthritis.11
cyclohexyl
cyclopentyl
cyclopentyl
105
94
36
34
80
11g OH CF3 allyl
11h CH3 CF3 allyl
11i
81
86
ia
ia
CH3 CF3 isopropyl
116
a All compounds were ER dependent (only active when ER is
coexpressed with NFkB-luciferase in the HAECT cells). b ia )
inactive, %E2 ) efficacy (relative inhibition of test compound at
10 uM vs E2 at 0.1 nM.
Table 3. Effects of Ethynyl Estradiol (EE) (0.01 mg/kg/day)
and Test Compound on NF-κB Target Gene Expression in
C57BL/6 Mice Fed a High Fat Diet
RANTES, VCAM-1,
TNFa,
%inhib
(%EE)
MHC,
%inhib
(%EE)
concn,
compd mg/kg/day
%inhib
(%EE)
%inhib
(%EE)
11a
10
5
10
10
5
0
0
0
66 (45%
29 (83%)
0
0
49 (74%) 53 (46%)
11c
11f
24 (55%) 42 (111%)
0
45 (120%) 22 (92%) 69 (100%) 29 (110%)
66 (71%) 32 (71%) 72 (100%) 39 (61%)
Figure 2. Comparison between ERR/WAY-169916 and ERR/
estradiol binding motifs.
2.5
80 (120%) 49 (94%) 56 (76%
38 (86%) 44 (115%)
31 (93%)
0
11h
11g
10
10
5
0
41 (100%) 28 (100%) 42 (93%) 38 (84%)
58 (100%) 38 (94%) 42 (96%) 38 (103%)
levels of the NF-κB target genes MHC, VCAM-1,
RANTES, and TNF-R in C57BL/6 mice subjected to a
high fat diet for 5 weeks.10 Inhibition of three out of four
target genes was considered orally active. A summary
of the compounds is shown in Table 3.
2.5
10
0
0
44 (92%)
0
11i
41 (100%) 59 (100%) 85 (100%) 21 (88%)
a The test compounds were administered po daily for 5 weeks.
For signifcant inhibition, p < 0.05.
Compound 11g WAY-169916 demonstrated no in-
crease in uterine wet weight versus estradiol after 5 day
dosing at 10 mg/kg in C57BL/6 mice. Estradiol typically
induces a 5-fold increase in uterine weight vs vehicle,
indicative of classical estrogenic effects.
Thus, a more general synthesis (Scheme 3) was
developed to provide a wider variety of substitution at
positions R or R2 while focusing on the 7- position (R1)
of the indazole. Thus the 2-fluorobenzophenone 9 was
reacted with hydrazine hydrate to give the substituted
1-H indazole core compound 10. The R-substitutions
were obtained by initial treatment with sodium hydride
in DMF followed by an alkyl halide. The reactions were
typically heated to 60 °C to complete the alkylation to
give 11 and 12 with ratios typically ranging from 4 to
1, respectively, which were easily separated using
normal phase chromatography. A summary of the in
vitro SAR from these optimized analogues are shown
in Table 2.
In a disease model for adjuvant induced arthritis
(AIA),11,12 WAY-169916 was found to be orally active
at 1 mg/kg. Lewis rats were injected with Freund’s
Adjuvant. After the joints were inflamed, typically after
8 days, the animals were dosed each day with test
compound. The activity was confirmed in the histologi-
cal scoring shown in Table 4. The activity was also
measured at 0.1 and 0.3 mg/kg doses and found to be
dose dependent.11
WAY-169916 displaced [3H]E2 from ERR-ligand bind-
ing domain (LBD) protein with an IC50 value of 1300
nM and from ERâ-LBD with an IC50 value 106 nM. To
Compounds that showed in vitro selectivity were
evaluated in vivo for their ability to inhibit the mRNA
Table 4. Histological Scoring of Synovitis in the Tarsal Joints from Animals with Adjuvant-Induced Arthritis Treated Orally for 2
Weeks with WAY-169916 Beginning on Day 8 after CFA Injection. Means ( SD
dose, mg/kg po
synovial struct 0-3
fibroplasia 0-3
inflammatory cells 0-3
pannus 0-2
total scorea
vehicle
1
2.92 ( 0.21
2.08 ( 0.20b
2.67 ( 0.41
1.58 ( 0.38b
2.92 ( 0.21
1.33 ( 0.41b
2.00 ( 0
10.50 ( 0.63
5.83 ( 1.78b
0.83 ( 0.98b
a Total score is the summation of four individual scores. b p < 0.05, statistically significant improvement in the scores over the vehicle-
treated group.

6438 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 26
Letters
References
(1) Chadwick, C. C.; Chippari S.; Matelan, E.; Borges-Marcucci, L.;
Eckert, A. M.;. Keith Jr., J. C.;. Albert, L. M.; Leathurby, Y.;
Harris, H. A.; Bhat, R. A.; Ashwell, M.; Trybulski, G.;.Winneker,
R. C.; Adelman, S. J.; Steffan, R. J.; & Harnish, D. C. Identifica-
tion of Pathway Selective Estrogen Receptor Ligands That
Inhibit NF-κB Transcriptional Activity. Proc. Nat..Acad.. Sci. U.
S. A., submitted
(2) Bauer J.; Herrmann F., Interleukin-6 in clinical medicine. Ann.
Hematol. 1991, 62, 203-210.
(3) Kurebayashi, S.; Miyashita, Y.; Hirose, T.; Kasayama, S.; Akira,
S.; Kishimoto, T. Characterization of mechanisms of interleu-
kin-6 gene repression by estrogen receptor. J. Steroid Biochem.
Mol. Biol. 1997, 60, 11-17.
(4) Thomas, T. N.; Rhodin, J. A.; Clark, L.; Garces, A.; Bryant, M.
A comparison of the antiinflammatory activities of conjugated
estrogens and 17â-estradiol. Inflam. Res. 2003, 52, 452-460.
(5) Jordan VC. Antiestrogens and selective estrogen receptor modu-
lators as multifunctional medicines. 1. Receptor interactions. J
Med Chem 2003, 46, 883-908. Jordan, V. C. Antiestrogens and
selective estrogen receptor modulators as multifunctional medi-
cines. 2. Clinical considerations and new agents. J. Med. Chem.
2003, 46, 1081-1111.
Figure 3. Comparison between WAY-169916/ERR and estra-
diol/ERR complex on the position of helix-12.
(6) Harris, H. A.; Katzenellenbogen, J. A.; Katzenellenbogen, B. S.;
Characterization of the biological roles of the estrogen receptors,
ERalpha and ERbeta, in estrogen target tissues in vivo through
the use of an ERalpha-selective ligand. Endocrinology 2002, 143,
4172-4177
(7) Krishnan, R.; Lang, S. A., Jr.; Lin, Y.; Wilkinson, R. G. Reactions
of hydroxybenzophenone with hydrazines. J. Heterocycl. Chem.
1988, 25, 447-452.
substantiate the interaction of WAY-169916 (11g) with
the ERR, a cocrystallization was performed with the
LBD of ERR. The compound binds in the same binding
pocket as estradiol as shown in Figure 2.
The phenol of WAY-169916 interacts with the
Glu-353 and Arg-394 salt bridge. The trifluoromethyl
group is pointing directly to the His-524, the second
hydroxyl binding position of estradiol. Of particular
interest in the ERR/WAY-169916 complex is that helix
12 adopts an antagonist position (Figure 3). This sup-
ports the lack of an estrogenic agonist effect associated
with estradiol. WAY-169916, however, does not have the
traditional antagonist effects found in SERMS.
In summary, optimization of the SAR around the
indazole ring led to the identification of WAY-169916
as a selective ligand for ER which is devoid of the
traditional effects of estrogens. WAY-169916 is a potent
pathway-selective inhibitor of NF-κB-induced inflam-
matory events. Its utility has been demonstrated in the
AIA disease model for arthritis. A near complete rever-
sal in hindpaw scores was observed as well as dramatic
improvements in histological scores of synovitis and
cartilage lesions in the tarsal joints.
(8) Nahm, S.; Weinreb, S. M. N-Methoxy-N-methylamides as effec-
tive acylating agents. Tetrahedron Lett. 1981, 22, 3815-3818.
(9) McOmie, J, F. W.; West, D. E. 3,3′-Dihydroxybiphenyl. Organic
Syntheses; Wiley: New York, 1973; Collect. Vol. V, p 412.
(10) Ovariectomized C57BL/6 mice (16-20 g) (Taconic) were sepa-
rated into groups of eight. The mice were fed a chow diet or an
atherogenic diet (15.75% fat, 1.25% cholesterol, and 0.5% sodium
cholate). EE or test compound was administered once daily by
gavage in a methylcellulose/Tween vehicle (0.1 mL per mouse)
for 5 weeks. At the end of the experimental period, the liver was
collected and uterine wet weight was recorded.
(11) Liver total RNA was prepared by using Trizol reagent (BRL).
Estrogen and compound regulation of NF-κB target genes were
verified by real time RT-PCR using an ABI PRISM 7700
Sequence Detection System according to the manufacturer’s
protocol (Applied Biosystems). The data were analyzed using the
Sequence Detector v1.7 software (Applied Biosystems) and
normalized to GAPDH using the Applied Biosystems primer set.
(12) Keith, J. C.; Albert, L.; Leatherby, Y, Follettie, M.; Wang, L.;
Borges-Marcucci, L.; Chadwick, C. C.; Steffan, R. J.; Harnish,
D. C. The Identification of Pathway Selective Estrogen Receptor
Ligands Which Inhibit NF-kB Transcriptional Activity and Their
Utility in RA. Arthritis Res. Ther., submitted.
Acknowledgment. The authors thank Matora Fieo-
ry and Gang Yao (current address Biogen Inc., Cam-
bridge, MA 02142) for their early work with ArQule on
the indazole arrays which guided the SAR.
(13) AIA model: 12 week old ovariectomized female Lewis rats were
injected with Freund’s Adjuvant. After the joints were inflamed,
typically after 8 days, the animals were dosed each day with
test compound and monitored to determine hindpaw joint score
(erythema and swelling; 0-3 score) max. ) 12. Tissue was
collected for histology which included synovial hyperplasia,
inflammatory cell infiltration, pannus formation, and articular
cartilage destruction.
Supporting Information Available: Experimental pro-
cedures for library preparation and characterization data. This
material is available free of charge via the Internet at
http://pubs.acs.org.
JM049194+