Substituted 4-hydroxyphenyl sulfonamides as pathway-selective estrogen receptor ligands

Article abstract of DOI:10.1016/j.bmcl.2005.11.015

The transcription factor nuclear factor-κB (NF-κB) is a key component in the onset of inflammation. We describe here a series of 4-hydroxyphenyl sulfonamide estrogen receptor (ER) ligands that selectively inhibit NK-κB transcriptional activity but are devoid of conventional estrogenic activity.

Full text of DOI:10.1016/j.bmcl.2005.11.015

Bioorganic & Medicinal Chemistry Letters 16 (2006) 854–858
Substituted 4-hydroxyphenyl sulfonamides
as pathway-selective estrogen receptor ligands
Joseph P. Sabatucci,a,* Mark A. Ashwell,c Eugene Trybulski,a Mary-Margaret OÕDonnell,d
William J. Moore,a Douglas C. Harnishb and Christopher C. Chadwicke
aChemical and Screening Sciences, Wyeth Research, 500 Arcola Road, Collegeville PA 1942,6, USA
bCardiovascular/Metabolic Diseases, Wyeth Research, 500 Arcola Road, Collegeville PA 1942,6, USA
cArQule, 19 Presidential Way, Woborn, MA 01741, USA
dNeurogen Corp, 35 NE Industrial Road, Branford, CT 06405, USA
eLife Diagnostics Inc, West Chester, PA 19380, USA
Received 25 July 2005; revised 3 November 2005; accepted 4 November 2005
Available online 21 November 2005
Abstract—The transcription factor nuclear factor-jB (NF-jB) is a key component in the onset of inflammation. We describe here a
series of 4-hydroxyphenyl sulfonamide estrogen receptor (ER) ligands that selectively inhibit NK-jB transcriptional activity but are
devoid of conventional estrogenic activity.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Overproduction of the cytokine interleukin (IL-6) has
been associated with states of chronic inflammation.1
Production of the IL-6 gene is initiated by the transcrip-
tion factor nuclear factor jB (NF-jB), 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-jB (IjB).
Although well known for their classic effects on the
reproductive tract and action via estrogen response ele-
ments in gene promoters, non-selective estrogens such as
E2 (17b-estradiol) are also known to have anti-inflam-
matory activity.2 This non-classical anti-inflammatory
activity of estrogen has been attributed to interference
of NF-jB activity through a variety of possible
mechanisms.3
of estrogen therapy on bone density, plasma lipid
profiles, vasomotor symptoms, and colon cancer has
been documented.6 However, estrogen therapy has also
been associated with an increased risk for breast cancer
and uterine bleeding. In light of these potential side ef-
fects, more selective ligands for ERa have been sought.
We thus began an HTS screen for selective ER ligands
using an assay developed in HAECT-1 cells.7 HAECT
cells are transfected with human ERa and a reporter
gene NFjB-luciferase. These cells are then treated with
IL-1b and the test compound. The amount of luciferase
present translates directly into the degree of NF-jB
transcription. The target profile for a test compound is
a 6100 nM IC50 of NF-jB luciferase reporters with an
efficacy (% E2) near 100%, coupled with a lack of classi-
cal estrogen receptor activity. Since E2 strongly stimu-
lates creatine kinase (CK) expression, an in vitro assay
measuring CK concentration serves as a monitor of
unwanted classic estrogenic effects.
In endothelial cells, it has been shown that E2 inhibits
IL-1b-induced NF-jB reporter activity and IL-6 expres-
sion in an estrogen receptor (ER) dependent fashion.4
This suppression of Il-6 expression correlates with the
anti-inflammatory action of E2 in vivo as confirmed in
different models of inflammation.5 The beneficial role
Compounds of general type 4 were identified which met
the above criteria. The synthesis of these compounds
was straightforward as depicted in Scheme 1. The appro-
priately substituted anilines (1) were N-alkylated to
compounds 2 by reductive amination with sodium triacet-
oxyborohydride and an aldehyde.8 These anilines were
coupled with a sulfonyl chloride to give sulfonamides 3
Keywords: Nuclear factor-jB; Estrogen receptor; Sulfonamide.
*
Corresponding author. Tel.: +1 484 865 4410; fax +1 484 865 9398;
e-mail: sabatuj@wyeth.com
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.11.015

J. P. Sabatucci et al. / Bioorg. Med. Chem. Lett. 16 (2006) 854–858
855
OZ
OH
R1
H
R1
a
N
b
NH2
R2
c or d
R
R2
SO2Cl
R3
R3
RCHO
R
SO2
R3
R
SO2
R3
N
N
ZO
R2
R2
2
R1
1
3
Z= Me
Z= CO2Et
4
g
f or
OZ
R1
NH2
R2
SO2Cl
e
+
R3
ZO
SO2
R1
HN
R2
1
9
5
R3
SO2Cl
SO2K
R1
OH
h
KO3S
i
X=H, Me
O
O
R1
R1
O
O
O
8
O
7
6
Scheme 1. Reagents and conditions: (a) AcOH/NaBH(OAc)3/DCM (83–89%); (b) Et3N/DCM (79–87%); (c) BCl3/n-Bu4NI (for Z = Me) (45–68%);
(d) 10% K2CO3/MeOH (for Z = CO2Et) (86–91%); (e) Et3N/DCM/14 h (78–88%); (f) RX/K2CO3/CH3CN (69–81%); (g) DEAD/PPh3/THF/ROH
(48–67%); (h) ClCO2Et/aq NaOH (78%); (i) SO2Cl2/toluene/cat. DMF (81%).
and then deprotected to give the final phenolic target
compounds. When a methoxy group was employed as a
protecting group for the phenol, de-methylation of
sulfonamides 3 was achieved with BCl3 and n-Bu4NI.9
The standard procedure employing BBr3 usually resulted
in bromination ortho to the phenol or in some cases,
loss of the N-alkyl group. Compound 8 containing a
carbonate-protecting group was later employed, allowing
deprotection under mild conditions with no side
products.10,11
Compounds 4e–4m in Table 1 summarize the basic SAR
study of the amide nitrogen R group. The optimal group
was straight chain C3 (compounds 4f and 4h) with little
advantage in potency seen in larger chains or branching.
Branching on the carbon directly attached to the nitro-
gen as in isopropyl (4g), cyclohexyl (4k) or phenyl (4j)
was particularly detrimental to activity. Surprisingly,
addition of a hydroxyl group at the end of the 2 carbon
chain (4n) resulted in loss of activity.
We next turned our attention to substitution on the aro-
matic rings. Substitution of a lipophilic group ortho to
the phenol could produce a beneficial effect, but was
not mandatory for activity. In some cases, an ortho
methyl group improved the selectivity. Addition of a
methoxy group (4o) resulted in a loss of activity. It be-
came apparent that the 3 position might increase selec-
tivity but not potency.
Table 1 summarizes the biological results of target mol-
ecules 4a–4n. The initial lead 4a had weak activity in the
luciferase assay, however, it provided a template to be-
gin SAR development. The effect of substituents on both
rings and the chain length/branching of the R group of
the amide nitrogen were two points with which to begin
the SAR study. Two noteworthy points quickly estab-
lished in this series were the need for a substituent R
on the amide nitrogen (as exemplified by the inactivity
of compound 4b) and a 4-OH on the aromatic ring con-
taining the sulfonyl group. Attempts to replace this crit-
ical OH with Cl, OMe, NO2 or a phenol mimic (CO2H,
CH3CONH) resulted in a total loss of activity. Place-
ment of the phenol group was also crucial. Both the
3-phenol and the 3,4-diphenol analogues resulted in a
substantial loss of activity.
The most dramatic effects were seen by the addition of
substituents on the aniline containing aromatic ring.
Placing a methoxy group at either the ortho, meta or
para position of the aniline containing ring (compounds
4p–4r) demonstrated that electron-donating groups were
not favorable. It was noted in these cases that the ortho
substitution seemed to have the most activity in this ser-
ies albeit weak. Nonetheless, the possibility of an Ôortho
effectÕ was investigated. This hypothesis came to fruition
in the case of the 2-bromo derivative (compound 4s). A
dramatic increase in potency resulted, producing a com-
pound with a roughly 100 nM activity in the luciferase
assay and good selectivity, which fell within project cri-
teria. This significant breakthrough shifted our focus to-
ward the exploration of lipophilic ortho substitution.
We concentrated on scaffold 10, maintaining the
R2
O2
S
N
R3
R
HO
R1
4

856
J. P. Sabatucci et al. / Bioorg. Med. Chem. Lett. 16 (2006) 854–858
Table 1. Effects of n-alkyl analogues of sulfonamides versus E2 on NF-jB, and CK expression in Ad5-wt-ER-infected HAECT-1 cellsa
Compoundb R1
R3
R
ER/NFjB-luc % E2-NF/jB CK EC50 (nM) % E2-CK Fold selectivity % E2 selectivity
IC50 (nM)
4a
4b
4c
4d
4e
4f
H
H
Et
404 81
na
100
1600 525
25
3.9
4
H
H
H
Et
Me
H
H
328 105
272 105
na
111
87
2135 805
1391
36
94
6.5
5
3.0
1.1
4-OH
H
Et
Me
H
H
H
n-Pr
i-Pr
Allyl
n-Bu
Phenyl
303 138
871
235
113
81
1430 282
1985
599
88
31
54
25
45
35
34
35
4.7
2.3
2.5
2.2
3.8
1.6
2.0
1.8
.77
4g
4h
4i
H
H
2.6
2.0
3.9
1.9
2.5
2.3
2.6
H
H
107
98
H
H
317
410
705
4j
H
H
84
86
1543
534
4k
4l
H
H
Cyclohexyl 866
Benzyl 500
neo-Pentyl 1371
H
H
79
92
242
>2500
4m
4n
4o
4p
4q
4r
4s
H
H
H
H
2-OHEt
Et
na
na
OMe
H
H
2-OMe Et
3-OMe Et
4-OMe Et
830
1973 800
na
49
37
2171
830
54
37
2.6
.91
H
.42
H
H
2-Br
Et
90
8
98
378
7
4.2
2.6
na = not active.
% E2 = efficacy (relative inhibition of test compound at 10 lM versus E2 at 0.1 nM).
a All compounds were ER dependent (active when ER is coexpressed with NFjB-luciferase in the HAECT cells).
b R2 = H.
optimized n-propyl group on the amide nitrogen. Table
2 summarizes the biological data for compounds con-
taining an ortho substitution. Compound 10a, combin-
ing the best SAR knowledge to date, produced a lead
with 40 nM activity and 10-fold selectivity over CK.
We attribute the greater selectivity of 10a (CK
EC50 = 480) versus 10b (CK EC50 = 151) to the methyl
group ortho to the phenol.
hydride to afford the alcohol, which was cyclized via a
Mitsunubo reaction to give 15. The phenol was deprotec-
ted with BCl3/n-Bu4NI as before to give the target com-
pound 16. Table 3 summarizes the biological results for
the cyclic compounds.
R3
O2
S
N
R2
All of the compounds in Table 2 were potent in the lucif-
erase assay at 100 nM or less with the exception of o-
phenyl (compound 10i) and o-Me (compound 10d) both
of which had activity on a par with the initial leads. The
ortho effect was particularly striking in those com-
pounds with di-ortho substitution. Compounds 10j
through 10o proved to be the most potent with luciferase
activities <30 nM and selectivities in most cases more
than 10-fold over CK.
HO
R1
10
N
O2S
N
O2S
OH
OH
11
12
With the SAR of the simple chain compounds understood
we then explored the effect of having restricted rotation
analogues. These would have the benefit of having an
N-substituted chain and an ortho substituent as the same
moiety. The simple analogue 11 was devoid of activity;
however, with the addition of a methyl group alpha to
the nitrogen (compound 12), enhancement of luciferase
activity was observed. We also investigated the effect of
restricted rotation analogues as exemplified by the benzo-
thiazine 16. This type of constrained analogue is not easily
accessible via commercially available material as in the
previous case, so a synthetic route was devised (Scheme
2). Ethyl 3-methoxyphenyl acetate (13) was chlorosulfo-
nated para to the methoxy group with chlorosulfonic acid
and then coupled with 2-chloro aniline to give sulfon-
amide 14. The ester was reduced with lithium aluminum
The results clearly show the superior activity of the
straight chain analogues versus the cyclic examples.
In general, the SAR of the straight chain sulfonamides
as exemplified by structure 4 can be summarized in the
following manner: (1) a hydroxyl group para to the sul-
fonyl is mandatory for activity, while the addition of a
meta lipophilic group shows an increase in selectivity,
(2) smaller N-alkyl chain analogues show the best
potency with propyl being the optimal group. Some
tolerance of chain extension can be seen but branching
at the a-carbon is not tolerated, (3) a lipophilic group
ortho to the aniline nitrogen enhancing luciferase
activity.

J. P. Sabatucci et al. / Bioorg. Med. Chem. Lett. 16 (2006) 854–858
857
Table 2. Effects of o-substituted analogues of n-propyl sulfonamides versus E2 on NF-jB, and CK expression in Ad5-wt-ER-infected HAECT-1
cellsa
Compound R1
R2
R3
ER/NFjB-luc IC50 (nM) % E2-NF/jB CK EC50 (nM) % E2-CK Fold-selectivity % E selectivity
10a
10b
10c
10d
10e
10f
10g
10h
10i
Me
H
H
H
H
H
H
H
H
H
H
Br
Br
Br
Me
F
40
26 10
60
97
98
480
151 14
na
45
35
12
5.8
2.1
2.8
CF3
Me
H
120
115
104
108
79
109
2
923 57
161
42
63
53
52
39
8.4
2.1
2.7
1.6
2.0
1.5
3.5
74
42
H
Cl
Cl
2
144 31
650
243
3.4
Me
H
51 11
54 21
253
12.8
4.5
CF3
Ph
138
42
Me
H
>2889
63 23
153 22
213 31
378 84
75 35
12
11.4
2.6
10j
Me Cl
Me Cl
24 15
107
83
70
24
42
35
30
43
1.5
3.5
1.7
2.7
3
10k
10l
Me
Me
Me
Me
Me
26
5
5
4
5
5
5
5.8
Et
Cl
Et
Cl
72
96
42.6
23.6
5.3
10m
10n
10o
16
14
7
Me Me
Me Et
90
53
1.7
1.2
na = not active.
% E2 = efficacy (relative inhibition of test compound at 10 lM versus E2 at 0.1 nM).
a All compounds were ER dependent (active only when ER is coexpressed with NFjB-luciferase in the HAECT cells).
Cl
Acknowledgments
CO2Et
a,b
EtO2C
HN
SO2
c,d
The authors wish to acknowledge the expertise of the
following scientists: James Ward for synthetic technical
assistance, Sue Chippari and Tom Kenney for running
the assays and recording the data, Jim Mattes and col-
leagues for NMR analysis, and Mei-Ye Zhang for mass
spectral data.
MeO
MeO
14
13
HO
e
MeO
Cl
Cl
N
N
S
S
O2
O2
16
References and notes
15
Scheme 2. Reagents and conditions: (a) ClSO3H, neat (41%); (b) 2-Cl
aniline/CH2Cl2/Et3N (68%); (c) LAH/ether/0 ꢁC (55%); (d) DEAD/
PPh3/THF (62%); (e) BCl3/n-Bu4NI (68%).
1. Bauer, J.; Herrmann, F. Ann. Hematol. 1991, 62, 203.
2. (a) Evans, M. J.; Harris, H. M.; Miller, C. P.; Karathan-
asis, S. K.; Adelman, S. J. Endocrinology 2002, 143, 375;
(b) Tyree, C. M.; Zou, A.; Allegretto, E. A. J. Steroid
Biochem. Mol. Biol. 2002, 80, 291; (c) Mukherjee, T. K.;
Nathan, L.; Dinh, H.; Reddy, S. T.; Chaudhuri, G. J. Biol.
Chem. 2003, 278, 117.
3. (a) Stein, B.; Yang, M. X. Mol. Cell. Biol. 1995, 15, 4971;
(b) Ray, A.; Prefontaine, K. E.; Ray, P. J. Biol. Chem.
1994, 269, 12940; (c) Ray, P.; Ghosh, S. K.; Zhang, D. H.;
Ray, A. FEBS Lett. 1997, 409, 79; (d) Sun, W. H.; Keller,
E. T.; Stebler, B. S.; Ershler, W. B. Biochem. Biophys. Res.
Commun. 1998, 244, 691; (e) Harnish, D. C.; Scicchitano,
M. S.; Adelman, S. J.; Lyttle, C. R.; Karathanasis, S. K.
Endocrinology 2000, 141, 3403.
Table 3. Effects of cyclic sulfonamides versus E2 on NF-jB, and CK
expression in Ad5-wt-ER-infected HAECT-1 cellsa
Compound ER/NFjB-luc % E2 ER/NFjB CK EC50 % E2
IC50 (nM)
(nM)
CK
11
12
16
na
479
892
98
80
>1000
na
4. Kurebayashi, S.; Miyashita, Y.; Hirose, T.; Akira, S.;
Kishimoto, T. J. Steroid Biochem. Mol. Biol. 1997, 60, 11.
5. Jansson, L.; Holmdahl, R. Arthritis Rheum. 2001, 44,
2168.
6. Riggs, B. L.; Hartmann, L. C. N. Engl. J. Med. 2003, 348,
618.
na = not active.
% E2 = efficacy (relative inhibition of test compound at 10 lM versus
E2 at 0.1 nM).
a All compounds were ER dependent (active when ER is coexpressed
with NFjB-luciferase in the HAECT cells).
7. Chadwick, C. C.; Chippari, S.; Matelan, E.; Borges-
Marcucci, L.; Eckert, A. M.; Keith, J. C., Jr.; 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. Proc. Natl. Acad. Sci. U.S.A.
2005, 102, 2543.
8. Abdel-Magid, A. F.; Carson, K. G.; Harris, B. D.;
Maryanoff, C. A.; Shah, R. D. J. Org. Chem. 1996, 61,
3849.
In conclusion, we have demonstrated that a known class
of sulfonamides are potent pathway-selective inhibitors
of NF-jB induced inflammatory events through ER that
are devoid of the traditional unwanted effects of
estrogens.12

858
J. P. Sabatucci et al. / Bioorg. Med. Chem. Lett. 16 (2006) 854–858
9. Brooks, P. R.; Wirtz, M. C.; Vetelino, M. G.; Rescek, D.
M.; Woodworth, G. F.; Morgan, B. P.; Coe, J. C. J. Org.
Chem. 1999, 64, 9719.
10. Bailey, J.; Clarke, D.; Crawley, M. W.; Marsden, P. D.;
Sidhu, J. PCT Int. Appl. (1983), 87 pp. CODEN: PIXXD2
WO 8300939 A1 19830317 CAN 99:30710 AN
1983:430710.
11. Typical procedures: To a dichloromethane solution of
aniline 1 were added 1.1 equiv of proprionaldehyde,
1.1 equiv of AcOH, and 1.5 equiv of NaBH(OAc)3. The
reaction mixture was stirred at room temperature over-
night and then quenched with 1 N NaOH. The organic
phase was separated and the aqueous layer was extracted
once more with dichloromethane. The organic layers were
combined, dried (MgSO4), and concentrated to yield N-
substituted anilines 2, which could be purified by filtration
through a plug of silica gel, eluting with 15% ethyl acetate/
hexane. The resulting product was dissolved in dichloro-
methane and 1.2 equiv of triethylamine added, whereupon
a solution of 1.2 equiv of 8 in dichloromethane was added
dropwise. Progress of the reaction was monitored by TLC
until most of the aniline was consumed. The solution was
concentrated and the residue was subjected to silica gel
chromatography eluting with 20% ethyl acetate/hexane to
provide sulfonamides 3. The sulfonamide 3 (Z = CO2Et)
was dissolved in methanol and a solution of 10% aqueous
potassium carbonate added and the reaction monitored by
TLC until all the starting material was consumed. The
solution was concentrated and the residue extracted 2·
with ethyl acetate and the combined extracts dried
(MgSO4) and concentrated. The product phenols 4 are
crystallized from ethyl acetate/hexane. All compounds had
NMR and MS data consistent with the proposed
structures.
12. Representative examples of compounds were also
screened against ERR, FXR, PPARg, PPARd,LXRa,
LXRb, and VDR with no cross-activity observed at
10 uM.