Diaryl-dialkyl-substituted pyrazoles: Regioselective synthesis and binding affinity for the estrogen receptor

Article abstract of DOI:10.1016/S0960-894X(02)00057-4

We have developed two novel series of tetrasubstituted pyrazoles, embodying 1,3-diaryl-4,5-dialkyl or 3,5-diaryl-1,4-dialkyl substitution patterns. The scope of a regioselective method, developed by us earlier, was expanded to allow the synthesis of the first series of these tetrasubstituted pyrazoles directly from α,β-unsaturated ketones. The binding affinity of some of these pyrazoles for the estrogen receptor (ER) subtypes ERα and ERβ is very high, and the overall affinity pattern suggests the importance of three phenol substituents for high affinity, ERα-selective binding.

Full text of DOI:10.1016/S0960-894X(02)00057-4

Bioorganic & Medicinal Chemistry Letters 12 (2002) 947–950
Diaryl-Dialkyl-Substituted Pyrazoles: Regioselective Synthesis
and Binding Affinity for the Estrogen Receptor
Gisele A. Nishiguchi, Alice L. Rodriguez and John A. Katzenellenbogen*
Department of Chemistry, University of Illinois, 600 S. Mathews Avenue, Urbana, IL 61801, USA
Received 19 November 2001; accepted 18 January 2002
Abstract—We have developed two novel series of tetrasubstituted pyrazoles, embodying 1,3-diaryl-4,5-dialkyl or 3,5-diaryl-1,4-
dialkyl substitution patterns. The scope of a regioselective method, developed by us earlier, was expanded to allow the synthesis of
the first series of these tetrasubstituted pyrazoles directly from a,b-unsaturated ketones. The binding affinity of some of these pyr-
azoles for the estrogen receptor (ER) subtypes ERa and ERb is very high, and the overall affinity pattern suggests the importance of
three phenol substituents for high affinity, ERa-selective binding. # 2002 Elsevier Science Ltd. All rights reserved.
The estrogen receptor (ER) has the capacity to bind a
wide variety of non-steroidal ligands with high affinity.1
This feature has stimulated the development of ligands
that act as selective estrogen receptor modulators
(SERMs, i.e., compounds that have a mixed endocrine
profile that affords agonistic or antagonistic activity in a
tissue-specific manner.)2 Some of this selectivity may
derive from differential binding to the two ER subtypes,
ERa and ERb.3 Compounds in this class are of use for
menopausal hormone replacement and the prevention
and treatment of breast cancer.4
these pyrazoles, we envisioned the synthesis of com-
pounds having a 1,3-diaryl-4,5-dialkyl or 3,5-diaryl-1,4-
dialkyl substitution patterns (Fig. 1, Series I and II,
respectively). The motivating rationale behind the
design of these new pyrazole ligand series consisted of
(a) keeping the C(3) phenol as a putative A-ring mimic
of estradiol, (b) while modifying other portions of the
pyrazole encompassing substituents at N(1), C(4), and
C(5), thereby probing the compliance of the region of
the ligand-binding pocket with which these substituents
interact and the degree to which these changes affect
ERa selectivity. Herein, we report the synthesis of
tetrasubstituted pyrazoles having two new substitution
patterns and the evaluation of their ERa and ERb
binding affinity (Fig. 1). In the process, we have expan-
ded the scope of a regioselective synthesis route that
allows the first of these series pyrazoles to be prepared
directly from simple a,b-unsaturated ketone precursors.
The tetrasubstituted pyrazole core has emerged as a
target of particular interest in our efforts to develop
SERMs with high affinity and selectivity for the ER
subtypes.5 We have found that some 1,3,5-triaryl-4-
alkyl substituted pyrazoles, such as the propylpyrazole
triol PPT (Fig. 1), have very high ERa-selective binding
affinity and potency;5 others have found that differently
substituted pyrazoles have ERb selectivity.6 Molecular
modeling and phenol deletion/binding affinity studies
suggest that the C(3) phenol of the 1,3,5-triaryl-4-alkyl
substituted pyrazoles we have studied acts as the mimic
of the estradiol A-ring (Fig. 1).5 A comparable orienta-
tion appears to be adopted by the core isomeric 1,3,4-
triaryl-5-alkyl pyrazoles (Fig. 1).7
In our continuing interest in understanding the struc-
tural determinants of the ERa binding selectivity of
*Corresponding author. Tel.: +1-217-333-6310; fax: +1-217-333-
7325; e-mail: jkatzene@uiuc.edu
Figure 1.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00057-4

948
G. A. Nishiguchi et al. / Bioorg. Med. Chem. Lett. 12 (2002) 947–950
A convenient synthesis of the pyrazole core involves the
condensation of 1,3-diketones with hydrazines. This
method is most effective when the diketone is symmet-
rical, [i.e., when the C(3) and C(5) substituents of the
pyrazole are equivalent]8 However, when the 1,3-dione
is unsymmetrical, two regioisomers are produced, and
these can be very difficult to separate.9 We have devel-
oped a regioselective approach to tetrasubstituted pyr-
azoles that produces a single regioisomer starting from a
chalcone (cf. Scheme 1).9 Recently, Katritzky10 has
reported a related method that utilizes a 1-benzotriazolyl
activating substituent at the a-position of the chalcone.
group at an earlier stage of the synthesis, by starting
from a-alkyl-substituted enones. (As noted above, we
had been unable to effect the hydrazine condensation
with a-alkyl-substituted chalcones.9) There is literature
precedent for the synthesis of a alkyl-substituted chal-
cones from aryl ketones and aryl aldehydes under gas-
eous HCl conditions.11 However, our attempts to adapt
this synthesis to produce the a-substituted-a-alkylidene
acylphenones (e.g., 7a–d) directly by the reaction of
acylphenones 5a,b and cyclohexane carboxaldehyde
with anhydrous HCl were unsuccessful, and condensa-
tion under basic conditions (NaOH, KOH, NaOMe)
gave very poor yields.
In the initial investigations of our method,9 we were
unable to produce the tetrasubstituted pyrazoles
directly, because chalcones substituted at the ꢀ position
with alkyl groups proved to be too unreactive in the
initial hydrazine condensation step. However, hydrazine
condensation with a chalcone unsubstituted at the ꢀ
position proceeded effectively, and the resulting pyrazo-
line intermediate could be alkylated at C(4) and oxi-
dized, giving the tetrasubstituted pyrazole.9 Our
synthesis of the Series I targets (i.e., 1,3-diaryl-4,5-dia-
lkyl pyrazoles) was, therefore, envisioned to proceed by
this route, as shown in Scheme 1.
We found that the desired a-alkylidene butyro- and
valerophenones (7a–d) were synthesized satisfactorily in
two steps, aldol condensation and then dehydration
(Scheme 2). The use of Lewis acids for the asymmetric
synthesis of secondary alcohols from propiophenone
and an aldehyde, in the presence of a base, is widely
known.12 Activation of ketones 5a,b with TiCl4, fol-
lowed by enolate formation and condensation with the
cycloalkane aldehydes, gave the aldol products 6a–d.
The stereochemistry of the aldol products was not
determined, because they were simply dehydrated
(MsCl/DBU or p-TSA) to give the desired enone pro-
ducts 7a–d, together with some retro-aldol byproducts.
Attempts to synthesize the alkylidine acetophenone
from commercially available p-methoxyacetophenone
and cyclohexanecarboxaldehyde under various basic
conditions (NaOH, KOH, LDA) gave enone 1, but in
low yield (e.g., NaOH, 18% yield). Reaction of enone 1
with p-methoxyphenylhydrazine hydrochloride gave
two major products, believed to be pyrazole and pyr-
azoline intermediates. However, attempts to isolate the
pyrazoline and alkylate it using LDA and ethyl or pro-
pyl iodine, gave the non-alkylated, oxidized pyrazoline
as the major product. The difficulties in handling this
pyrazoline compounded by low yields in the enone for-
mation prompted us to search for a more efficient
sequence.
Despite our earlier lack of success in reacting a-sub-
stituted chalcones with substituted hydrazines,9 we
found that the a-alkylidine acylphenones (7a–d) did
undergo reaction with p-methoxyphenylhydrazine
hydrochloride in DMF in a slow reaction affording a
crude mixture of pyrazoline (8a–d) and pyrazoles (9a–d
and 10a–d) products (Scheme 3). This mixture was stir-
red in benzene and DDQ to afford the pyrazoles, from
which the desired isomers (9a–d) were isolated in 30–
50% overall yields. The surprising formation of small
amounts (ca. 10%) of undesired isomeric pyrazole side
products (10a–d) appears to be the result of hydrazine
condensation on an enone precursor that had under-
gone double bond isomerization from the a,b- to the
alternate a,b0-position under the acidic reaction condi-
tions. Although these products (10a–d) are constitu-
tional isomers of the desired pyrazoles 9a–d, they are
easily removed from them by chromatography; this was
not the case with many of the regiosiomers that are
One option to circumvent the problems we encountered
in this approach would be to introduce the C(4)-alkyl
Scheme 1.
Scheme 2.

G. A. Nishiguchi et al. / Bioorg. Med. Chem. Lett. 12 (2002) 947–950
949
obtained by hydrazine condensation with unsymmet-
rical 1,3-diketones. The pyrazoles (9a–d) were then
deprotected to give the final phenol products (11a–d).
with anisoyl chloride to give the corresponding 1-aroyl-
cycloalkenes (12a,b), followed by condensation with
p-methoxyphenylhydrazine hydrochloride salt to give
the pyrazolines (13a,b). Oxidation with DDQ in benzene
and deprotection with BBr3 afforded the desired pyr-
azoles (15a,b).
Another set of Series I pyrazoles consisted of those
having a more rigid core, with a six- or seven-membered
ring fused across the C(4) and C(5) positions. The
synthesis of such compounds (Scheme 4) started with a
Friedel–Crafts acylation of cyclohexene or cycloheptene
The targets for the Series II type compounds began with
the introduction of a non-aromatic substituent at N(1)
by reacting the symmetrical 1,3-diketone 168 with
cyclopentyl or cyclohexyl hydrazine to form the desired
.
pyrazole core directly. Deprotection with BF3 SMe2
gave the desired products (18a,b) (Scheme 5).
The binding affinity for each pyrazole phenol for human
estrogen receptors alpha and beta (ERa and ERb) was
determined using a competitive radiometric assay; affi-
nities were expressed relative to that of estradiol, to give
relative binding affinity (RBA) values (Table 1).13
All pyrazoles that we have studied in the past have
shown ERa-binding selectivity.5 This is also a feature of
most of the new pyrazoles studied here. It is curious,
however, that the pyrazoles having the highest affinity,
the N(1)-cycloalkyl compounds of Series II (18a,b),
were also those that showed only very low ERa/ERb
selectivity. These two pyrazoles, in fact, are among the
highest affinity ER ligands in this heterocyclic class, a
result that is consistent with prior indications that a
variety of substituents, even a t-butyl group, is tolerated
at the N(1) position of these pyrazoles.5,7,8 By contrast,
the very low ERa selectivity of the pyrazoles with these
N(1) cycloalkyl groups was unexpected and indicates
that they must be interacting with the receptor in a
manner that does not discriminate in binding affinity
between ERa and ERb (see below).
Scheme 3.
Lowest affinities were shown for the Series I pyrazoles
with the cycloalkyl system fused at C-4,5 (15a,b). The
other Series I pyrazoles having alkyl groups at C(4) and
C(5) (11a–d) had intermediate affinity, but typically less
than that shown previously by the triaryl pyrazoles of
the 1,3,5- and 1,3,4-substitution patterns (cf. Fig. 1).5,7
Both sets of Series I pyrazoles show substantial pre-
ferential binding to ERa. Together, these data suggest
that the phenolic substituent on C(5) of the pyrazole
does play an important role in binding affinity, poten-
tially as the hydrogen bonding partner of His524 (see
Fig. 2). When this phenolic substituent is replaced with
Scheme 4.
Table 1. ERa and ERb binding affinities and selectivitiesa
Compd
ERa
ERb
a/b
11a
11b
11c
11d
15a
15b
18a
18b
0.59
2.1
1.1
3.6
0.56
0.19
3013
74
0.14
0.23
0.11
0.60
0.042
0.047
4.2
8.9
9.4
6.0
13
4
2.3
71
1.0
aRelative binding affinity (RBA) values, where estradiol=100.
Scheme 5.

950
G. A. Nishiguchi et al. / Bioorg. Med. Chem. Lett. 12 (2002) 947–950
Figure 2. Crossed-stereo view of a model of pyrazole 18b in the ligand binding pocket of ERa and ERb. Atom colors are used for the ligand and
residues of ERa, and orange for those of ERb. Residue names and identities are given as ERa/ERb.
an alkyl group (as in compounds 11a–d and 15a,b),
affinity drops to a large degree.
Acknowledgements
This work was supported by a grant from the National
Institutes of Health (PHS 5R37 DK15556). We are
grateful to Kathryn Carlson for biological assays and
Dorina Kosztin for assistance with molecular modeling
studies.
We have previously used molecular modeling to exam-
ine the orientation of pyrazole binding in the ER ligand
binding pocket and to try to rationalize ERa/ERb
selectivity.5,14 Using the same methods, we addressed
three issues with the following outcomes: (1) The very
low affinity of the C(4,5) fused cycloalkyl pyrazoles
15a,b appears to result from a very poor fit in the bind-
ing pocket (not shown); (2) the lower affinity of the
C(4),C(5)-dialkyl pyrazoles 11a–d compared to PPT
appears to result from the loss of the interaction with
the helix-11 histidine (524 in ERa and 475 in ERb)
when the C(5) phenol is replaced by an alkyl group (not
shown), and (3) the high affinity but low ERa/ERb
appears to result from alternative binding orientations
in the two ER subtypes (Fig. 2). When pyrazole 18b is
minimized in the ligand pocket of ERa and ERb, dif-
ferent ligand orientations are found that are related by a
‘rotation’ about an axis through the two phenols. In
both cases, the ligand fits well in the pocket and enjoys
productive interactions with the A-ring (E353/305;
R394/346) and D-ring (H524/472) hydrogen bonding
partners. By flipping, this ligand appears to avoid
unproductive interactions with the residues in the ligand
pocket that differ between ERa and ERb (L384/M336
and M421/I373).
References and Notes
1. Gao, H.; Katzenellenbogen, J. A.; Garg, R.; Hansch, C.
Chem. Rev. 1999, 99, 723.
2. McDonnell, D. P. Trends Endocrinol. Metab. 1999, 10, 301.
3. Pettersson, K.; Gustafsson, J. A. Annu. Rev. Physiol. 2001,
63, 165.
4. Katzenellenbogen, B. S.; Katzenellenbogen, J. A. Breast
Cancer Res. 2000, 2, 335.
5. Stauffer, S. R.; Coletta, C. J.; Tedesco, R.; Sun, J.; Katze-
nellenbogen, B. S.; Katzenellenbogen, J. A. J. Med. Chem.
2000, 43, 4934.
6. Lin, X.; Huebner, V. Curr. Opin. Drug Disc. 2000, 3, 383.
7. Stauffer, S. R.; Huang, Y. R.; Coletta, C. J.; Tedesco, R.;
Katzenellenbogen, J. A. Bioorg. Med. Chem. 2001, 9, 141.
8. Stauffer, S. R.; Katzenellenbogen, J. A. J. Combinat. Chem.
2000, 2, 318.
9. Huang, Y. R.; Katzenellenbogen, J. A. Org. Lett. 2000, 2,
2833.
10. Katritzky, A. R.; Wang, M.; Zhang, S.; Voronkov, M. V.
J. Org. Chem. 2001, 66, 6787.
11. Sangwan, N. K.; Rastogi, S. N. Ind. J. Chem. 1979, 18B,
65.
12. Evans, D. A.; Rieger, D. L.; Bilodeau, M. T.; Urpi, F. J.
Am. Chem. Soc. 1991, 113, 1047.
13. Carlson, K. E.; Choi, I.; Gee, A.; Katzenellenbogen,
B. S.; Katzenellenbogen, J. A. Biochemistry 1997, 36,
14897.
In conclusion, we have extended a regioselective method
for the synthesis of tetrasubstituted pyrazoles and used
it to prepare pyrazole ligands for the estrogen receptor
that have new patterns of diaryl and dialkyl substitu-
tion, some of which have very high affinity for both
ERa and ERb, in contradistinction with the high affi-
nity ERa-selective pyrazoles in the triaryl series. The
lack of subtype selectivity of some of these pyrazoles
might be the consequence of alternative orientations that
they adopt in the binding pocket of the two ER subtypes.
14. Mortensen, D. S.; Rodriguez, A. L.; Carlson, K. E.; Sun,
J.; Katzenellenbogen, B. S.; Katzenellenbogen, J. A. J. Med.
Chem. 2001, 44, 3838.