Tetrahydroisoquinolines as subtype selective estrogen agonists/antagonists

Article abstract of DOI:10.1016/j.bmcl.2004.03.077

Two series of 6-hydroxy and 7-hydroxy tetrahydroisoquinolines were prepared. Evaluating a range of C-1, C-4, and N-substituents led to the discovery of ER α and ER β selective analogs.

Full text of DOI:10.1016/j.bmcl.2004.03.077

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2729–2733
Tetrahydroisoquinolines as subtype selective estrogen
agonists/antagonists
Richard Chesworth,* Michael P. Zawistoski, Bruce A. Lefker, Kimberly O. Cameron,
Robert F. Day, F. Michael Mangano, Robert L. Rosati, Stacy Colella, Donna N. Petersen,
Amy Brault, Bihong Lu, Lydia C. Pan, Pia Perry, Oicheng Ng, Tessa A. Castleberry,
Thomas A. Owen, Thomas A. Brown, David D. Thompson and Paul DaSilva-Jardine
Pfizer Global Research and Development, CVMD Groton Laboratories, Eastern Point Road, Groton, CT 06340, USA
Received 19 February 2004; revised 24 March 2004; accepted 25 March 2004
Abstract—Two series of 6-hydroxy and 7-hydroxy tetrahydroisoquinolines were prepared. Evaluating a range of C-1, C-4, and
N-substituents led to the discovery of ER a and ER b selective analogs.
ꢀ 2004 Elsevier Ltd. All rights reserved.
The major endogenous ligands that act at the estrogen
receptors (ER) a and b are 17-b estradiol and estrone.1
These two estrogenreceptors are members of the nuclear
hormone receptor superfamily,2 which mediate a num-
ber of physiological processes such as bone mineral
density and lipid levels as well as female reproductive
systems.3 Post-menopausal women have severely de-
creased levels of endogenous estrogens and as such are
often treated with HRT (hormone replacement ther-
apy).4 HRT though, is not without risk. It is associated,
for example, with an increased incidence of both breast
and uterine cancers.5 Due to the side effects of HRT, the
estrogenreceptor(s) have beena target of considerable
interest for the pharmaceutical industry.6 Compounds
targeted initially at ER a (at the time ER b was un-
osteoporosis) as well as lasofoxifene 1,12 currently in
advanced clinical trials for osteoporosis. Since the dis-
covery of these SERMs, major advances in the under-
standing of estrogens have occurred, not the least of
which is the discovery of a second estrogen receptor, ER
b.7 With the discovery of this second subtype it is hoped
that compounds can be found with new pharmacology
compared to the previously described SERMs. Inour
search for new agents, we chose to investigate the tet-
rahydroisoquinoline (THIQ) scaffold 2. As canbe seen
from Figure 1, the THIQ scaffold is structurally similar
to the tetrahydronaphthalene scaffold of lasofoxifene 1,
but incorporation of the nitrogen atom allows for the
rapid synthesis of analogs. The first report of THIQs
analogs as estrogenic compounds was made by Naga-
13
known),7 have beenfoudn to retainsome of 17-
b
rajanet al., who described their use as anti-implanta-
tion agents. More recently the Novartis group has
estradiol’s positive effects, while overcoming some of the
side effects. Molecules that have this mixed agonist/
antagonist profile are referred to as SERMs (Selective
8
EstrogenReceptor Modulators). Tamoxifenwas the
first clinically approved SERM, for the treatment of
breast cancer.9 Tamoxifenhowever, is associated with
an increased incidence of uterine cancer10 and second
generation SERMs have managed to overcome this in-
creased risk. Examples from this class include raloxi-
fene11 (clinically approved for the treatment of
N
O
R2
X
N
R1
HO
HO
Lasofoxifene
Tetra-hydro
isoquinoline
2
1
Keywords: Tetrahydroisoquinolines; ER a; ER b.
* Corresponding author. Tel.: +1-860-7150094; fax: +1-860-6860001;
e-mail: richard_chesworth@groton.pfizer.com
Figure 1. THIQs derived from lasofoxifene.
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.03.077

2730
R. Chesworth et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2729–2733
R1
i
O
HN
N
BnO
O
MeO
N
R2
MeO
3
4
R1
O
O
R2
R2
i, ii
N
CF3
N
CF3
R1
R1
iii
ii
N
N
MeO
MeO
MeO
R2
HO
12
13
5
6a-g,11a
N
N
O
O
For R1=H
R2
O
v, then vi,
vii, or viii,
iv
then ix, OR
NH
N
CF3
iii, iv
R6
X
ix, x
HO
MeO
MeO
N
N
R1
7a-e
8a-e
R5O
R2
N
O
11b-k
14 R5=Bn, R6= COCF3
15 R5=Bn, R6=H
16 R5=H, R6=H
iii
CF3
HO
9a-e
Scheme 2. Synthesis of compounds 11b–k. Reagents and conditions:
(i) 10% Pd/C, H2 44 psi, EtOH; (ii) NaH, DMF, 1-(2-chloroethyl)
pyrrolidine HCl, 100 ꢁC; (iii) BBr3, CH2Cl2, 0 ꢁC; (iv) NaH, BnBr,
DMF, 100 ꢁC; (v) K2CO3, MeOH, reflux; (vi) RCOCl, Et3N, THF;
(vii) RSO2Cl, Et3N, THF; (viii) RCO2Cl, Et3N, THF; (ix) 20%
Pd(OH)2/C, NH4HCO2, MeOH, reflux; (x) RCHO, NaOAc, NaC-
NBH3, MeOH.
CONR3R4
I
v
N
N
ethoxy) benzoic acid. Compounds 11b–k were synthe-
sized as depicted inScheme 2. Starting with the differ-
entially protected THIQ 12 (synthesized as for 8a–e in
Scheme 1 using 4-benzyloxy carboxylic acid), the benzyl
ether was selectively removed via hydogenolysis and the
resulting phenol was alkylated under basic conditions
with 1-(2-chloroethyl) pyrrolidine hydrochloride to give
13. The 6-methoxy group was removed and replaced as a
benzyl ether to give 14 inorder to facilitate analog
synthesis. The trifluoroacetamide group in 14 was re-
moved via basic hydrolysis with K2CO3 inMeOH to
reveal the secondary amine 15, which was capped with
either anacid chloride, a sulfonyl chloride or a chloro-
formate to produce the corresponding amide, sulfon-
HO
HO
6g
10a-b
Scheme 1. Synthesis of compounds for 6a–g, 9a–e, 10a–b, an d11a.
Reagents and conditions: (i) RCO2H, Et3N, CH2Cl2 1-propane
phosphonic cyclic anhydride, DMAP or RCO2H, Et3N, CH2Cl2,
EDC, HOBt; (ii) POCl3, reflux or PCl5, 1,2 DCE reflux, thenNaBH 4,
MeOH; (iii) BBr3, CH2Cl2, )78 ꢁC to rt; (iv) TFAA, Et3N, CH2Cl2,
0 ꢁC; (v) Pd(PPh3)4, Et3N, H2C@CHCONR3R4, DMF, 100 ꢁC.
released their results describing THIQs as possible
SERMs.14
The THIQs were synthesized using standard chemical
transformations, starting from the appropriate 3-meth-
oxy phenethylamine 3 (Scheme 1). Treatment of 3-
methoxy phenethylamine (R1 ¼ H or Ph)13 with the
appropriate R2 containing carboxylic acid in the pres-
ence of the coupling agents such as 1-propane phos-
phonic acid cyclic anhydride (PPAA) for R1 ¼ Ph and
EDC/HOBt for R1 ¼ H provided the desired amides 4.
The amide 4 underwent a Bischler–Napieralski cyclization
amide,
or
carbamate
followed
by
transfer
hydrogenolysis of the benzyl ether group to give the
analogs 11b–i. The basic amines 11j and 11k were syn-
thesized from 15 via transfer hydrogenolysis followed by
reductive amination using the appropriate aldehyde in
the presence of NaCNBH3.
The 7-hydroxy THIQs analogs were synthesized as
depicted inScheme 3. The bromoketone 17 was con-
verted into the amino-ketone 18 upontreatment with
benzyl amine. The amino-ketone underwent reductive
amination with 3-methoxy benzaldehyde, followed by
reductionof the keto group with NaBH 4, to give 19. The
amino-alcohol 19 underwent an acid mediated cycliza-
tionto give the 7-methoxy THIQ 20, which upon
hydrogenolysis of the N-benzyl group gave the inter-
mediate 21. The amide and sulfonamide analogs were
prepared by treating the secondary amine 21 with the
appropriate acid or sulfonyl chlorides or with TFAA to
inthe presecne of POCl
reductionwith NaBH 4, gave the corresponding 6-
(CAUTION), that after
3
methoxy tetrahydroisoquinoline 5. WhenR ¼ phenyl,
1
the methoxy protecting group was removed using BBr3
to unmask the phenol giving the desired analogs 6a–f
and intermediate 6g. The amines 7a–e were capped with
trifluoroacetic anhydride (TFAA) and the resultant
intermediates 8a–e were thensubjected to BBr 3 to reveal
the desired phenolic analogs 9a–e. Examples 10a–b were
prepared from the THIQ 6g via a Heck reactionwith the
appropriate acrylamide.
give 22, which uponexposure to BBr furnished the
3
Compound 11a was prepared using the chemistry de-
picted in Scheme 1 using 4-phenyl-(2-pyrollidin-1-yl
analogs 23a–d. Compound 23e was prepared directly
from 20 by treatment with BBr3.

R. Chesworth et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2729–2733
2731
Table 1. Estrogen receptor binding affinities (IC50) for compounds
6–1015
MeO
O
MeO
O
MeO
R2
ii, iii
i
X
N
R1
OH
HO
R2
Br
NHBn
NBn
MeO
19
17
Compd R1
X
Binding affinity
(nM)a
18
MeO
MeO
ER a
ER b
6a
6b
6c
Ph
Ph
Ph
Ph
Ph
Ph
––
––
––
––
––
––
Ph
30.7
50.0
29.8
42.7
53.1
6.70
385
48.8
26.3
99.0
180
2-Thienyl
Cyclohexyl
i-Pr
iv
v
6d
6e
NBn
NH
MeO
MeO
CH2–cyclohexyl
CH2Ph
295
21
20
MeO
6f
66.4
63.5
27.4
22.0
8.90
>3200
8.70
HO
9a
9b
9c
9db
9ec
10a
CF3 CO
CF3 CO
CF3 CO
CF3 CO
CF3 CO
2-Thienyl
Ph
235
165
p-C6H4–Br
p-C6H4–OH
p-C6H4–OH
vi, vii
or viii
53.6
>3200
8.00
ix
Ph
––
N
X
R2
N
R2
O
MeO
HO
X
22
23a-e
N
Scheme 3. General synthesis of 7-hydroxy THIQs. Reagents and
conditions: (i) BnNH2, Et3N, THF, rt; (ii) 3-methoxy benzaldehyde,
NaB(OAc)3H, 1,2 DCE; (iii) NaBH4, MeOH, 0 ꢁC; (iv) CH2Cl2, TFA,
reflux; (v) Pd/C, H2 50 psi, EtOH; (vi) PhCOCl, Et3N, CH2Cl2;
(vii) RSO2Cl, Et3N, CH2Cl2; (viii) TFAA, Et3N, CH2Cl2, 0 ꢁC;
(ix) BBr3CH2Cl2, )78 ꢁC to rt.
10b
Ph
––
5.90
4.80
O
N
O
We initially chose to keep R1 ¼ phenyl analogous to
lasofoxifene 1 and explore the space described by R2.
We found that the space defined by R2 was tolerant of a
wide variety of substitutents. As can be seen in Table 1,
the introduction of just a simple phenyl group 6a pro-
a Values are means of at least three experiments.
b Single enantiomer. Tentatively assigned as the R-enantiomer.
c Single enantiomer. Tentatively assigned as the S-enantiomer.
duces
a reasonably potent analog. Interestingly,
a good binder, but upon inserting an amide linker into
the molecule 11b, a considerable loss of potency was
observed. Not all amides were impotent, as the trifluo-
roacetamide derivative 11c exhibited improved potency.
Interestingly, this analog showed a slight preference for
ER a over ER b (ca. 3·) in contrast to analogs 9a–d.
This bias for ER a may result from the increased steric
bulk of R2, which is consistent with ER b having a
smaller ligand binding pocket than ER a.16 Upon
replacing the amide linkage with a sulfonamide 11d or a
methylene linker 11j, potency at both the receptors was
restored. We also found that carbamate linkers were
replacement of the phenyl group with a 2-thienyl group
produced compound 6b, which demonstrated modest
selectivity (2·) for ER b. The additionof the dimethyl
acrylamide 10a or the morpholino acrylamide 10b
groups to the phenyl ring, improved potency at ER a by
up to 10·. Anaromatic group at R was not necessary
to retainpotency, as evidenced by compounds
2
6c–e.
Introduction of a methylene spacer group at R2 pro-
duced analogs 6e–f with modest (up to 10·) selectivity
for ER a. The R1 phenyl group could be effectively re-
placed with a trifluoroacetamide group. The introduc-
tionof this group allowed us to access analogs 9a–e that
displayed modest selectivity for ER b. Compounds 9d
and 9e were separated by chiral HPLC and as can be
seen, the biological activity resides in one isomer only.
Compound 9d was putatively assigned as the R-enan-
tiomer based onthe stereochemistry of lasofoxifene 1.
also well tolerated as seenwith compoudn
11e. We
further investigated the scope of the sulfonamide linker
and found that para-substitution of the phenyl sulfon-
amide led generally to less potent compounds 11f–h than
whencompared to the parent sulfonamide
11d. The 1-
napthyl sulfonamide 11i, however, was almost as potent
as the phenyl sulfonamide 11d indicating that reason-
able steric bulk could be tolerated if appropriately
positioned. Interestingly, the two analogs that contain a
methylene linker, 11j and 11k, both show ca. 10–20·
selectivity for ER a over ER b.
In Table 2 the results of our investigations into different
R1 groups canbe seenwhere R 2 was kept constant as the
phenyl alkoxyamine moiety found in lasofoxifene 1. The
direct THIQ analog 11a of lasofoxifene was found to be
The binding affinities of the 7-hydroxy THIQs 23a–e are
showninTable 3. The 7-hydroxy THIQ scaffold is tol-
erant of the amide 23a, sulfonamide 23b, and methylene
23e linkers. The potency of the amide 23a is particularly
InRef. 14 the selectivity for compound 11a was found to be ca. 8·
for ER a, whereas we found the selectivity to be ca. 2·. This
difference in selectivity may be due to differing assay conditions.

2732
R. Chesworth et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2729–2733
17
Table 2. Estrogen receptor binding affinities (IC50) for compounds 1
and 11a–k
Table 4. MCF-7 cell proliferationdata for selected compounds
Agonist (nM)
% act.a
Antagonist (nM)
Compd
EC50
IC50
% inh.a
N
O
6c
11.1
n/a
n/a
n/a
93
n/a
n/a
83
66
96
11c
11f
11i
n/a
n/a
n/a
86.6
29.8
51.3
X
n/a: Not applicable.
a Percent activation and inhibition are calculated from highest dose
N
R1
HO
(1 lM).
Compd
X
R1
Binding affinity (nM)a
ER a
ER b
6c) were found to be agonists.18 Compounds contain-
ing the phenyl alkoxy-amino sidechain at R2, such as
11c, 11f, an d11i, were found to antagonize the prolif-
erative actionof 17- b estradiol inthis cell line thus
confirming that this series could antagonize the estrogen
receptor.
1
n/a
––
n/a
Ph
0.50
11.5
562
1.21
22.2
2250
93.3
18.5
293
11a
11b
11c
11d
11e
11f
11g
11h
11i
11j
CO
CO
SO2
CO2
SO2
SO2
SO2
SO2
CH2
Ph
CF3
Ph
28.7
5.20
44.3
24.8
29.4
112
Et
p-Tolyl
71.2
133
p-C6H4-nPr
2-Napthyl
1-Napthyl
Ph
Insummary, the THIQ scaffold has proved to be a
versatile scaffold. It has allowed the identification of
novel analogs that can act as ER agonists or antagonists
in MCF-7 cells and also has led to the identification of
ER a and ER b selective ligands. Through subtle
changes of the R1 and R2 substituents, as well as the
locationof the phenolic group of the THIQ scaffold,
compounds that are up to 20· selective for ER a, as well
as compounds that are up to 10· selective for ER b have
beenidentified.
1080
21.3
42.2
5.60
4.70
O
O
11k
CH2
9.20
220
n/a: Not applicable.
a Values are means of at least three experiments.
Table 3. Binding affinities of 7-hydroxy THIQs
References and notes
OH
1. Kuiper, G. G. J. M.; Carlsson, B.; Grandien, K.; Enmark,
ꢀ
E.; Haggblad, J.; Nilsson, J.; Gustafsson, J. A. Endocri-
€
nology 1997, 138, 863–870.
2. Sato, M.; Grese, T. A.; Dodge, J. A.; Bryant, H. U.;
Turner, C. H. J. Med. Chem. 1999, 42, 1–24.
3. Nasr, A.; Breckwoldt, M. Gynecol. Endocrinol. 1998, 12,
43–59.
N
R1
HO
X
Compound
X
R1
Binding affinity (nM)a
ER a
ER b
4. Ulmsten, U. Proc. Soc. Exp. Biol. Med. 1998, 217, 2–5.
5. Cauley, J. A.; Cummings, S. R.; Black, D. M.; Mascioli, S.
R.; Seeley, D. G. D. Am. J. Obstet. Gynecol. 1990, 163,
1438–1444.
6. Blizzard, T. A.; Morgan, J. D., II; Mosley, R. T.; Birzin,
E. T.; Frisch, K.; Rohrer, S. P.; Hammond, M. L. Bioorg.
Med. Chem. Lett. 2003, 479–483.
23a
23b
23c
23d
23e
CO
SO2
SO2
CO
CH2
Ph
Ph
160.0
120.0
28.0
27.9
41.7
10.7
11.4
13.1
1-napthyl
CF3
Ph
122.7
20.1
a Values are means of at least three experiments.
7. (a) Kuiper, G. C.; Carlsson, B.; Enmark, E.; Pelto-
ꢀ
Huikko, M.; Nilsson, S.; Gustafsson, J. A. Proc. Natl.
Acad. Sci. U.S.A. 1996, 93, 5925–5930; (b) Mosselmann,
S.; Polman, J.; Dijkema, R. FEBS Lett. 1996, 392, 49–53.
8. Jordan, V. C. J. Am. Chem. Soc. 2003, 46, 1081.
9. Jordan, V. C. Br. J. Pharmacol. 1993, 3, 495.
10. Robinson, D. C.; Bloss, J. D.; Schiano, M. A. Gynecol.
Oncol. 1995, 59, 186–190.
11. Black, L. J.; Sato, M.; Rowley, E. R.; Magee, D. E.;
Bekele, A.; Williams, D. C.; Cullinan, G. J.; Bendele, R.;
Kauffman, R. F.; Bensch, W. R.; Frolik, C. A.; Termine, J.
D.; Bryant, H. U. J. Clin. Invest. 1994, 93, 63–69.
interesting when contrasted to the 6-hydroxy THIQ
benzamide 11b. As with the 6-THIQs, the 1-napthyl
sulfonamide 23c was also reasonably potent. The tri-
fluoroacetamide analog 23d proved to be the most ER b
selective (ca. 10·) analog found during this investiga-
tion. In general we found the 7-hydroxy THIQ template
to be selective for ER b over ER a.
A few select compounds were advanced into the MCF-7
cell proliferationassay as showninTable 4. The MCF-7
cells are derived from breast tumors that predominantly
express ER a.17 As anticipated, compounds that lack the
phenyl alkoxy-amino sidechain at R2, (e.g., compound
12. Rosati, R. L.; Da Silva-Jardine, P.; Cameron, K. O.;
Thompson, D. D.; Ke, H. Z.; Toler, S.; Brown, T. A.; Pan,
L. C.; Ebbinghaus, C. F.; Reinhold, A. R.; Elliot, N. C.;
Newhouse, B. N.; Tjoa, C. M.; Sweetnam, P. M.; Cole, M.
J.; Arriola, M. W.; Gauthier, J. W.; Crawford, D. T.;

R. Chesworth et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2729–2733
2733
Nickerson, D. F.; Pirie, C. M.; Qi, H.; Simmons, H. A.;
Tkalcevic, G. T. J. Med. Chem. 1998, 41, 2928.
17. DNA Fluorometric Assay in96-well Tissue Culture Plates
Using Hoescht 33258 after Cell Lysis by Freezing in
Distilled Water. Rago, R.; Mitchen, J.; Wilding, G. Anal.
Biochem. 1990, 191, 31–34, Assay for estrogenreceptor- a
dependent proliferation of MCF7 cells. MCF7 human
mammary epithelioid adenocarcinoma cells (American
Type Culture Collection, Rockville, Maryland) were
plated at 6000 cells/well in96-well tissue culture plates in
Iscove’s Modified Dulbecco’s Medium (Invitrogen, Carls-
bad, California) supplemented with 5% fetal bovine serum
and antibiotics. Following overnight attachment, cultures
were incubated with two changes of medium plus 5%
charcoal-dextranstripped fetal bovine serum (‘stripped
serum’) over 48 h to deplete the cells of endogenous
steroids. After stripping, cultures were incubated with
fresh medium containing 5% stripped serum plus 0.1%
dimethyl sulfoxide vehicle or test compounds in a final
concentration of 0.1% dimethyl sulfoxide.ꢀC13ompounds
13. Nagarajan, K.; Talwalker, P. K.; Kulkarni, C. L.; Shah,
R. K.; Shenoy, S. J.; Prabhu, S. S. Indian J. Chem., Sect.
B: Org. Chem. Including Med. Chem. 1985, 24B(1), 83–97.
14. Renaud, J.; Bischoff, S. F.; Buhl, T.; Floersheim, P.;
Fournier, B.; Halleux, C.; Kallen, J.; Keller, H.; Schlaeppi,
J. M.; Stark, W. J. Med. Chem. 2003, 46, 2945.
15. Leake, R. E.; Habib, F. In Steroid Hormone Receptors a
Practical Approach; Green, B., Leake, R. E., Eds.; IRL:
Oxford, 1990; pp 67–92, Competition binding assay. The
ability of various compounds to inhibit [3H]-estradiol
binding was measured by a competition binding assay
using dextran-coated charcoal. 293T cell extracts express-
ing either hERa or hERb were incubated in the presence
of increasing concentrations of competitor and a fixed
concentration of [3H]-estradiol (141 Ci/mmol, New Eng-
land Nuclear, Boston, MA) in 50 mM TrisHCl pH 7.4,
1.5 mM EDTA, 50 mM NaCl, 10% glycerol, 5 mM DTT,
0.5 mg/mL b-lactoglobulinina final volume of 0.2 mL. All
competitors were dissolved indimethylsulfoxide. The final
concentration of receptor was 50 pM with 0.5 nM [3H]-
estradiol. After 16 h at 4 ꢁC, dextran-coated charcoal
(20 lL) was added. After 15 minat room temperature the
charcoal was removed by centrifugation and the radioac-
tive ligand present in the supernatant was measured by
scintillation counting. All reagents were obtained from
Sigma (St. Louis, MO).
were typically tested inlog dilutions from 10
to 10ꢀ6 M
in replicates of 8 wells. To assess proliferation-stimulating
activity, medium and test compounds were replaced after
4 days and the DNA content of cultures was determined
by Hoescht 33258 binding1 after a total of 7 days exposure
to test compounds. Alternatively, test compounds were
added inthe presence of 10 ꢀ10 M 17-b estradiol to assess
antagonism of estrogen-dependent proliferation. Estrogen
antagonist activity was similarly determined by Hoescht
dye binding after only 4 days of compound exposure.
18. Brzozowski, A. M.; Pike, A. C.; Dauter, Z.; Hubbard, R.
E.; Bonn, T.; Engstrom, O.; Ohman, L.; Greene, G. L.;
16. Pike, A. C. W.; Brzozowski, A. M.; Hubbard, R. E.;
Bonn, T.; Thorsell, A. G.; Engstrom, O.; Ljunggren, J.;
ꢀ
Gustafsson, J. A.; Carlquist, M. EMBO 1999, 18, 4608–
4618.
ꢀ
Gustafsson, J. A.; Carlquist, M. Nature 1997, 389, 753–
758.