Pyrazole ligands: Structure - Affinity/activity relationships and estrogen receptor-α-selective agonists

Article abstract of DOI:10.1021/jm000170m

We have found that certain tetrasubstituted pyrazoles are high-affinity ligands for the estrogen receptor (ER) (Fink et al. Chem. Biol. 1999, 6, 205-219) and that one pyrazole is considerably more potent as an agonist on the ERα than on the ERβ subtype (Sun et al. Endocrinology 1999, 140, 800-804). To investigate what substituent pattern provides optimal ER binding affinity and the greatest enhancement of potency as an ERα-selective agonist, we prepared a number of tetrasubstituted pyrazole analogues with defined variations at certain substituent positions. Analysis of their binding affinity pattern shows that a C(4)-propyl substituent is optimal and that a p-hydroxyl group on the N(1)-phenyl group also enhances affinity and selectivity for ERα. The best compound in this series, a propylpyrazole triol (PPT, compound 4g), binds to ERα with high affinity (ca. 50% that of estradiol), and it has a 410-fold binding affinity preference for ERα. It also activates gene transcription only through ERα. Thus, this compound represents the first ERα-specific agonist. We investigated the molecular basis for the exceptional ERα binding affinity and potency selectivity of pyrazole 4g by a further study of structure-affinity relationships in this series and by molecular modeling. These investigations suggest that the pyrazole triols prefer to bind to ERα with their C(3)-phenol in the estradiol A-ring binding pocket and that binding selectivity results from differences in the interaction of the pyrazole core and C(4)-propyl group with portions of the receptor where ERα has a smaller residue than ERβ. These ER subtype-specific interactions and the ER subtype-selective ligands that can be derived from them should prove useful in defining those biological activities in estrogen target cells that can be selectively activated through ERα.

Full text of DOI:10.1021/jm000170m

4934
J . Med. Chem. 2000, 43, 4934-4947
Articles
P yr a zole Liga n d s: Str u ctu r e-Affin ity/Activity Rela tion sh ip s a n d Estr ogen
Recep tor -r-Selective Agon ists
Shaun R. Stauffer,† Christopher J . Coletta,† Rosanna Tedesco,† Gisele Nishiguchi,† Kathryn Carlson,† J un Sun,‡
Benita S. Katzenellenbogen,‡,# and J ohn A. Katzenellenbogen*,†
Departments of Chemistry, Physiology, and Cell and Structural Biology, University of Illinois and
University of Illinois College of Medicine, Urbana, Illinois 61801
Received J une 9, 2000
We have found that certain tetrasubstituted pyrazoles are high-affinity ligands for the estrogen
receptor (ER) (Fink et al. Chem. Biol. 1999, 6, 205-219) and that one pyrazole is considerably
more potent as an agonist on the ERR than on the ERâ subtype (Sun et al. Endocrinology
1999, 140, 800-804). To investigate what substituent pattern provides optimal ER binding
affinity and the greatest enhancement of potency as an ERR-selective agonist, we prepared a
number of tetrasubstituted pyrazole analogues with defined variations at certain substituent
positions. Analysis of their binding affinity pattern shows that a C(4)-propyl substituent is
optimal and that a p-hydroxyl group on the N(1)-phenyl group also enhances affinity and
selectivity for ERR. The best compound in this series, a propylpyrazole triol (PPT, compound
4g), binds to ERR with high affinity (ca. 50% that of estradiol), and it has a 410-fold binding
affinity preference for ERR. It also activates gene transcription only through ERR. Thus, this
compound represents the first ERR-specific agonist. We investigated the molecular basis for
the exceptional ERR binding affinity and potency selectivity of pyrazole 4g by a further study
of structure-affinity relationships in this series and by molecular modeling. These investiga-
tions suggest that the pyrazole triols prefer to bind to ERR with their C(3)-phenol in the estradiol
A-ring binding pocket and that binding selectivity results from differences in the interaction
of the pyrazole core and C(4)-propyl group with portions of the receptor where ERR has a smaller
residue than ERâ. These ER subtype-specific interactions and the ER subtype-selective ligands
that can be derived from them should prove useful in defining those biological activities in
estrogen target cells that can be selectively activated through ERR.
In tr od u ction
ally quite easy to imagine the orientation that this
ligand is likely to adopt when it is bound by ER.3,4 ER
mutagenesis studies,5 and the recent X-ray crystal-
lographic structures of ER complexed with both estra-
diol and three nonsteroidal ligands (raloxifene, hydroxy-
tamoxifen, and diethylstilbestrol), provide additional
guidance in the selection of reasonable binding orienta-
tions for ligands of this type.4,6 However, when the
nonsteroidal estrogens have structures that are more
divergent from those of steroidal estrogens, it becomes
a greater challenge to predict ligand-binding orienta-
tion.7
The recent characterization of a second ER gene,
encoding ERâ, places a further premium on our under-
standing of the details of ligand-receptor interaction,8,9
because it would be especially interesting to have
ligands that could activate or inhibit each of the ER
subtypes with high selectivity. Such ligands would be
valuable tools to define the biological effects that are
mediated by ERR and ERâ. So far, however, there have
been only a few reports of ER subtype-selective estro-
gens, and in many cases, the selectivity has been
relatively modest.10
The estrogen receptor (ER) displays a remarkable
capacity for binding nonsteroidal ligands with high
affinity.1 Many of these ligands have been developed
into hormonal agents having mixed agonist-antagonist
and tissue-selective activities that are useful in meno-
pausal hormone replacement, in fertility regulation, and
in the prevention and treatment of breast cancer.
Because of their unusual pharmacology, some of these
agents have been termed selective estrogen r eceptor
m odulators (SERMs).2 To understand the molecular
basis of the tissue selectivity of these SERMs, it would
be helpful to know in detail how they are interacting
with the ER. However, short of performing X-ray
crystallographic analysis of ER complexes with each
ligand, it can be a challenge to obtain such information.
If the nonsteroidal ligand bears a reasonable struc-
tural relationship with steroidal estrogens, it is gener-
* Address correspondence to: J ohn A. Katzenellenbogen, Dept. of
Chemistry, University of Illinois, 600 S. Mathews Ave., Urbana, IL
61801. Tel: 217-333-6310. Fax: 217-333-7325. E-mail: jkatzene@
uiuc.edu.
† Department of Chemistry.
Recently, we investigated various heterocyclic diazole
structures as core elements for nonsteroidal estrogens
‡ Department of Physiology.
#
Department of Cell and Structural Biology.
10.1021/jm000170m CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/28/2000

Pyrazole Ligands: ERR-Selective Agonists
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26 4935
Sch em e 1. Synthesis of C(4)-Alkylpyrazole Analogues
F igu r e 1. Pyrazole core optimized for high-affinity ER binding
(X ) H, OH; R ) alkyl).
of novel design.11,12 Our aim was to identify systems that
would be amenable to combinatorial assembly and
library synthesis. From among the several systems we
studied that included imidazoles, oxazoles, thiazoles,
and isoxazoles, we found that high-affinity ER ligands
could be obtained by appropriate substitution on a
pyrazole core. Subsequently, we used a parallel, solid-
phase synthesis approach to prepare some combinatorial
pyrazole libraries of moderate size.13 In the cases we
investigated, high-affinity binding required tetrasub-
stitution of the pyrazole core and an appropriate display
of aromatic, phenolic, and aliphatic groups.11,12 An
example of this optimal pattern of substitution that we
have thus far delineated for pyrazoles includes three
aromatic groups at the 1-, 3-, and 5-positions, specifi-
cally phenols at the 3- and 5-positions, and an alkyl
group at the 4-position (Figure 1). Initial studies on one
of these compounds (Figure 1; X ) H, R ) Et) showed
that it acted as an agonist on both ER subtypes, but it
was considerably more potent on ERR than on ERâ.14
Thus, this pyrazole was termed an ERR potency-
selective agonist.14
In this report, we have investigated structure-
activity relationships in these tetrasubstituted pyrazoles
to delineate two aspects of their behavior: (1) the C(4)-
alkyl substituent and phenol hydroxyl pattern that
provide optimal ER subtype selectivity and (2) the
extent to which the ERR-selective binding affinity and
potency of these pyrazoles can be understood in the
context of crystal structures of the ERR ligand-binding
domain (LBD) and models for ERâ derived from these
structures. Our studies provide new information on
these issues, and in the process, we have identified a
pyrazole that shows complete ERR selectivity in tran-
scription activation by the receptor.
separate the protected product from traces of the
diketone precursor by chromatography. So, in these
cases the crude material was passed through a short
silica gel column and, without further purification, was
then deprotected using BBr3. Pyrazoles 3d ,e were also
treated in this manner. The final phenolic pyrazoles
4a -i were isolated in 30-98% yield after purification
by recrystallization and/or chromatography.
In addition to the final pyrazole products 4a -i, we
wished to prepare the isopropyl analogue derived from
ketone 1f. However, we encountered difficulties in the
preparation of this hindered dione because of significant
O-acylation that occurred during the Claisen condensa-
tion. Separation of the O-acylated byproduct from the
desired dione proved to be difficult. Moreover, once the
dione was isolated, the pyrazole condensation failed
using the conditions described above, which had been
optimized for solution-phase synthesis. To avoid these
problems, we utilized a solid-phase synthesis according
to the methodology that we had previously developed,13
and in this manner we were able to obtain the isopro-
pylpyrazole 6 starting from the resin-bound ketone 5,
as shown in Scheme 2, although the overall yield was
relatively low.
To prepare a complete series of all of the possible
diphenol and monophenol analogues of the C(4)-ethyl
triphenol 4f, we also synthesized monophenols 7a -c
and the remaining diphenols 8a ,b (all mono- and
diphenols are shown in Scheme 3, top). Monophenol 7a
was prepared from dibenzoylmethane (9) and 4-meth-
oxyphenylhydrazine by the usual sequence11 (Scheme
3, middle). The synthesis of the remaining two mono-
phenols (7b,c) by a regioselective route has been de-
scribed elsewhere (not shown).15 The two remaining
Resu lts a n d Discu ssion
Ch em ica l Syn th eses. The compounds we have stud-
ied have various C(4)-alkyl substituents and have either
a phenyl or 4-hydroxylphenyl substituent on N(1).
Shown in Scheme 1 is the route used for the synthesis
of the pyrazoles 4a -i. The starting alkylphenones 1a -f
were readily prepared in good yields by Friedel-Crafts
acylation of anisole. The requisite â-diketones 2a -e
were then produced in moderate to good yields by acyl-
ation of the corresponding lithium enolates with 4-ni-
trophenyl 4-methoxybenzoate. Because the pyrazoles
are sterically crowded, rather harsh conditions were
required for their formation (>16 h at reflux at 110-
120 °C in DMF/THF solution). Nonetheless, these
conditions served quite well for pyrazole formation in
solution. Several of the pyrazole intermediates protected
as methyl ethers were isolated in good yield. However,
with the trimethoxy pyrazoles 3g-i, it was difficult to

4936 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26
Stauffer et al.
Sch em e 2. Synthesis of C(4)-Isopropylpyrazole 6 by Solid-Phase Methodology13
diphenols (8a ,b ) were also prepared by this regio-
selective method, as shown in Scheme 3 (bottom). The
two chalcones (12a ,b ) are treated with p-methoxy-
phenylhydrazine under anerobic conditions to give the
pyrazolines 13a ,b. The corresponding anions, generated
by treatment with LDA, were ethylated, and the C(4)-
ethylpyrazolines 14a ,b were then oxidized to the pyra-
zoles 15a ,b and deprotected to the desired bisphenols
8a ,b.
ER Bin d in g Affin ity of Tetr a su bstitu ted P yr a -
zoles. The ER binding affinity of pyrazoles 4a -i, 6, 7a -
c, and 8a ,b was determined in a competitive radiometric
binding assay using purified full-length human ERR and
ERâ, as previously described.16,17 The affinities are
expressed as relative binding affinity (RBA) values and
are presented in two tables: Table 1, which is discussed
here, covers the effect of the nature of the C(4)-alkyl
substituent on the binding affinity and ER subtype
selectivity of pyrazoles in the triphenol and the principal
3,5-diphenol series; the data in Table 2, which compares
the affinity of the three possible mono- and diphenols
with the triphenol in the C(4)-ethylpyrazole series,
relates to the issue of ligand orientation and is discussed
in a later section.
Two interesting trends are notable in the binding
affinity data presented in Table 1. In both the diphenol
series (X ) H, 4a -e, 6) and the triphenol series (X )
OH, 4f-i), optimal binding affinity for ERR requires a
C(4)-alkyl substituent that is not too long (n-Bu: 4e,i)
or too short (Me: 4a ), the highest affinity being found
with the intermediate size substituents: Et (4b,f), n-Pr
(4c,g), and i-Bu (4d ,h ). Thus, it appears that the
subpocket that is accommodating this group has a
limited size and relatively narrow shape. A more
significant trend evident in this series has to do with
the effect of the phenolic substituent on the N(1)-phenyl
group. In each case where a comparison can be made
(namely, Et, Pr, i-Bu, or n-Bu), the triphenols (X ) OH,
4f-i) have higher affinity on ERR and/or lower affinity
on ERâ than do the diphenols (X ) H, 4a -e). As a
result, the triphenols consistently have higher ERR
affinity selectivity than do the corresponding 3,5-di-
phenols.18 In fact, some of the members show a remark-
able selectivity in their binding affinity for ERR vs ERâ,
which is 30-40 for some of the pyrazole diols (4c,d ) and
as high as 200-400 for some of the pyrazole triols (4f,g).
The increased ERR binding affinity of the triphenols
is not expected, because additional polar substituents
are generally poorly tolerated in the center of the ligand-
binding pocket of ER, at least in most nonsteroidal
ligand systems that have been examined, such as the
benzo[b]thiophenes19 and 2,3-diarylindans.7 One excep-
tion is found with certain triphenylacrylonitriles, where
addition of a third hydroxyl increases binding affinity.20
We have previously reported binding affinities for one
of these pyrazoles (4b), using receptor preparations
containing only the LBDs of human ERR and ERâ,
rather than full-length human ERR and ERâ.14 The
values obtained previously were higher but less ERR-
selective (60 ( 16 for ERR and 18 ( 4 for ERâ). At this
point, the reasons for these affinity differences between
full-length ER and the ER LBD are not clear, although
they have been seen, as well, in a structurally different
class of ER ligands.17 However, it is of note that the
higher ERR/ERâ affinity ratio that we obtain for pyra-
zole 4b with the full-length ERs is more consistent with
the 120-fold potency ratio we described in transcription
assays than was the 3-fold ERR/ERâ affinity ratio
obtained with ERs containing only the LBDs.14
Tr a n scr ip tion a l Activity a n d P oten cy of Tetr a -

Pyrazole Ligands: ERR-Selective Agonists
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26 4937
Sch em e 3. Synthesis of the Remaining Mono- and Diphenolic C(4)-Ethylpyrazoles
su bstitu ted P yr a zoles. For investigation of transcrip-
tion activation ability, we selected two compounds: the
pyrazole having the highest ERR/ERâ affinity selectivity
(410-fold, the propylpyrazole triol 4g, called PPT for
convenience) and the corresponding pyrazole in the diol
series (4c, called propylpyrazole diol or PPD). These two
pyrazoles were assayed for estrogen agonist activity in
transactivation assays using human endometrial cancer
(HEC-1) cells transfected with expression plasmids for
ERR and ERâ and an estrogen-responsive reporter gene
plasmid. The dose-response curves for these compounds
are shown in Figure 2.
Both compounds are potent in activating gene tran-
scription through ERR, but the PPD (4c) is weak in
transcriptional activation through ERâ and PPT (4g)
is completely inactive in stimulating transcription via
ERâ. Thus, PPT (4g) is an ERR-specific agonist. This
pharmacological profile is reminiscent of the behavior
of the C(4)-ethyl analogue of pyrazole 4c (namely,
pyrazole 4b), that we have previously described as an
agonist that is more potent on ERR than on ERâ. This
compound (4b) had an EC50 of ca. 1 nM on ERR and
showed a 120-fold potency selectivity for this ER sub-
type.14 However, in cell transfection assays both PPT
and PPD are nearly 10-fold more potent than the
original pyrazole 4b on ERR, and they are much more
ERR-selective.
As we had noted in our earlier study on pyrazole 4b,14
and is clearly evident here as well, the ERR selectivity
of these pyrazoles in terms of their potency in transcrip-
tion assays is substantially greater than their affinity
selectivity in binding assays. For example, the ERR
affinity selectivity of PPD (4c) is 32, yet its potency
selectivity is ca. 1000 (cf. Figure 2). Likewise, the ERR
affinity selectivity of PPT (4g) is 410, and its potency
selectivity, which although is difficult to accurately
evaluate (cf. Figure 2), is probably greater than 10000.
Such a discordance between affinity and potency might
be explained in the context of “tripartite receptor
pharmacology”, a concept that we advanced some time

4938 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26
Stauffer et al.
Ta ble 1. Relative Binding Affinity Data for C(4)-Alkyl
These six binding modes are given the designations:
N(1), N(1)′, C(3), C(3)′, C(5), and C(5)′.
Triphenol and 3,5-Diphenol Pyrazole Analoguesa
A classical approach that has been used to try to
ascertain which orientation di- and triphenolic com-
pounds adopt within the ER binding pocket has been
to systematically delete the phenolic hydroxyl group and
examine the effect of this deletion on the binding
affinity.7 Because the hydroxy group on the phenol that
corresponds to the A-ring of estradiol is thought to make
the major contribution to ligand-binding affinity, the
greatest reduction in binding should occur when this
hydroxyl group is the one deleted.3 We have undertaken
such an analysis by comparing the affinities of pyrazoles
in which the three phenols in the triphenol 4f were each
singly deleted (diphenol set: 4b, 8a ,b) or were deleted
in three pairs (monophenol set: 7a -c). The affinities
of these three diphenols and three monophenols, to-
gether with the triphenol parent 4f, are listed in Table
2. Because we had prepared the most pyrazoles in the
C(4)-ethyl series, we chose to do this structure-affinity
study in this series.
ERR/ERâ
compd
4a
4b
6
4c (PPD) n-Pr
4d
4e
R
X
RBA ERRb
RBA ERâb
RBA ratio
3,5-Diphenols
Me
Et
i-Pr
H
0.76 ( 0.18 0.28 ( 0.16
2.7
28
6.5
32
40
19
H
H
H
H
H
31 ( 15
5.6 ( 2
1.1 ( 0.2
0.86 ( 0.11
0.52 ( 0.03
1.4 ( 0
16.8 ( 0.3
56 ( 6
i-Bu
n-Bu
8.7 ( 2.0
0.47 ( 0.1
1,3,5-Triphenols
4f
Et
OH
OH
OH
36 ( 6
49 ( 12
75 ( 6
14 ( 4
0.15 ( 0.014
240
410
84
4g (PPT) n-Pr
4h
4i
0.12 ( 0.04
0.89 ( 0.06
0.18 ( 0.09
i-Bu
n-Bu OH
77
The simplest analysis we can make of the structure-
binding affinity pattern shown in Table 2 is the follow-
ing: In the monophenol set (7a -c), the N(1)- and C(3)-
phenols (7a ,b) have comparable affinities, but the C(5)-
phenol (7c) binds ca. 100-fold less well. Because of the
very low affinity of the C(5)-monophenol (7c), we believe
that the C(5)-phenol cannot function as the mimic of
the A-ring of estradiol, in essence, eliminating orienta-
tions C(5) and C(5)′ as possibilities. (It is of note that
these two “eliminated” binding modes project bulky
groups simultaneously in directions that correspond to
C(7) and C(11) of estradiol; cf. Figure 3.) Elimination of
the C(5)-phenol leaves the N(1)- and C(3)-phenols as
potential estradiol A-ring mimics, but the very similar
affinities of both of these monophenols suggest that each
one may function as the A-ring mimic.
a
Relative binding affinity (RBA), where estradiol is 100%.
Values are the mean of at least 2 and more typically 3 or more
independent determinations ((SD).33 Competitive radiometric
b
binding assays were done with purified full-length human ERR
and ERâ (PanVera Inc.), using 10 nM [3H]E2 as tracer and HAP
for adsorption of the receptor-tracer complex.16,17
ago.21 Binding measured in vitro with purified ERs
involves only the interaction between ligand and recep-
tor, whereas transcription measured in cells involves
the additional interaction of the ligand-receptor com-
plex with coactivators and other cellular components.
Thus, compared to their relative affinities for ERR and
ERâ, the relative potency of two ligands can be modu-
lated by differences in the strength with which their
respective ligand-receptor complexes bind to coactiva-
tors or are modified by other cellular elements, although
other factors, as well, might be involved. In this case, it
appears that the pyrazole complexes with ERR are
better able to bind coactivators than are the ERâ-
pyrazole complexes, as we have documented in a recent
study.22
Analysis of the binding affinity pattern of the bis-
phenols is somewhat more ambiguous, but suggestive,
nevertheless: Deletion of the N(1)-phenol from the
triphenol has only a minor effect on the binding affinity
(4b vs triphenol 4f), which means that in the context of
the triphenol in ERR, the N(1)-phenol is not contributing
significantly to binding affinity. Thus, one would expect
that the C(3)-phenol is likely to be the more important
of the two and thus that the orientations C(3) and C(3)′
are more likely than the orientations N(1) and N(1)′ .
Deletion of either the C(3)-phenol or the C(5)-phenol
results in a 4-5-fold decrease in binding (8a ,b vs 4f).
The fact that the N(1),C(3)-bisphenol (8b) still binds
quite well is not surprising, because the C(3)-phenol is
still available to be the estradiol A-ring mimic. However,
the fact that the N(1),C(5)-phenol (8a ) also binds quite
well might seem surprising, because this diphenol lacks
the important C(3)-hydroxyl.
A Mod el for th e Bin d in g of P yr a zoles w ith th e
ER Su btyp es. The high affinity, and particularly the
high selectivity, with which these pyrazoles bind to ERR
raises the important issue of what molecular features
underlie the differences in their interaction with ERR
vs ERâ. Without crystal structures available for the
comparison of any of these pyrazoles complexed with
both ERR and ERâ, we are currently limited to inves-
tigating this issue by molecular modeling.
1. St r u ct u r e-Affin it y R ela t ion sh ip s a n d t h e
Or ien t a t ion of P yr a zoles in t h e Liga n d -Bin d in g
P ock et of ERR. Because the pyrazoles we have studied
here are rather symmetrical and are polyphenolic, it is
not obvious which orientation these ligands might prefer
to adopt within the ligand-binding pocket of ER. In
principle, the triphenolic pyrazoles could adopt six
orientations (see Figure 4): Each of the three phenolic
rings could play the role of the A-ring of estradiol, and
in each case the remainder of the pyrazole could adopt
two orientations about the bond connecting the A-ring
phenol mimic to the pyrazole core (darkened bond).
We believe that the reasonably good affinity of di-
phenol 8a suggests that this pyrazole can bind with the
N(1) group as the estradiol A-ring mimic, meaning that
the N(1) and N(1)′ ligand orientations are also possible,
though, based on the minimal effect of the N(1)-phenol
deletion, they are probably not significantly populated
when a C(3)-phenol is available. (It seems unlikely that
diphenol 8a would bind with the C(5)-phenol in the
A-ring pocket, because of the very low affinity of the

Pyrazole Ligands: ERR-Selective Agonists
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26 4939
F igu r e 2. Transcription activation by ERR (left) and ERâ (right) in response to pyrazoles 4c (PPD) and 4g (PPT). Human
endometrial cancer (HEC-1) cells were transfected with expression vectors for ERR or ERâ and an (ERE)3-pS2-CAT reporter
gene and were treated with the indicated concentrations of ligand for 24 h. CAT activity was normalized for â-galactosidase
activity from an internal control plasmid. Values are expressed as a percent of the ERR or ERâ response with 1 nM E2, which is
set at 100%.32
F igu r e 3. Potential pyrazole binding relative to estradiol. Both modes A and A′ have the C(3)-phenol oriented to project in to the
estradiol A-ring binding pocket. Modes A and A′ are related to one another by rotation of the pyrazole core around the darkened
bond.
C(5)-monophenol 7c.) Thus, the phenol deletion/binding
affinity approach has only reduced the number of
possible ligand orientations in the binding pocket from
six to four, with a suggested preference for two orienta-
tions, C(3) and C(3)′. As described in the section below,
we have tried to narrow down the possible ligand
orientations even further by molecular modeling.
The conclusions thus far that the C(3)-phenol is the
preferred A-ring mimic but that the N(1)-phenol is a
possible alternative are consistent with two other stud-
ies that we have recently completed. In a related but
more limited investigation of the orientation of ligand
binding in two isomeric pyrazole series, we concluded
that the orientation with the C(3)-phenol as the A-ring
mimic was most likely in the pyrazole series that is the
same as the one studied here and that in the isomeric
series the phenol that occupied topologically congruent
substituent position also functioned as the A-ring
mimic.23 In another investigation in which we worked
to convert these pyrazoles from agonists to antagonists
by appending a basic side chain onto the four substit-
uents on these pyrazoles, we found that high affinity
was retained only when this group was attached to the
C(5)-hydroxyl.24 On the basis of this structure-affinity
pattern, as well as some molecular modeling and
consideration of the structure of the ER LBD complexes
with antagonists, we concluded in that study that the
most likely orientation of these basic side chain-
substituted pyrazoles was N(1).24 In this orientation, the
basic side chain could project outward through the 11â-
channel that forms in the ER LBD complexes with
antagonists.4,6
2. Molecu la r Mod elin g of th e Or ien ta tion of
P yr a zole Tr iols in th e Liga n d -Bin d in g P ock et of
th e ERs. We have previously done modeling of the
pyrazole ligands in the ER LBDs, first in our initial
study on these ligands11 and, more recently, in connec-
tion with a study of the orientation of these ligands in

4940 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26
Stauffer et al.
F igu r e 4. Crossed-stereoview of propylpyrazole triol (PPT, 4g) docked and minimized in ERR LBD pocket according to binding
mode A, showing selected residues close to the ligand. The pyrazole ligand is shown with standard atom colors. The residues in
ERR are identified and shown in yellow. At the two positions where the ERâ sequence differs from ERR, the ERâ residues are
shown in red (Met in ERâ in place of ERR Leu384 and Ile in ERâ in place of ERR Met421).
Ta ble 2. Relative Binding Affinity Data for C(4)-Ethylpyrazole
Modeling). During the course of the ligand docking
routine, the PPT that began in orientation mode C(3)′
Tri-, Di-, and Monophenolsa
became reoriented into mode C(3), whereas the one that
began in mode C(3) remained in mode C(3). Therefore,
mode C(3)′ was not considered further. After the dock-
ing/minimization routine, the PPT in mode C(3) was
nicely accommodated in this orientation, with only
minimal changes in the side chain conformations of
some of the residues that line the ligand-binding pocket.
We also modeled the four other possible orientations of
this pyrazole using the same routine (not shown), but
we found that none of them gave a fit as good as that of
the pyrazole orientation illustrated as mode C(3) in
Figure 3. The fact that the N(1) and N(1)′ modes did
not score well in our modeling, although they seemed
possible on the basis of the structure-affinity consid-
erations above, suggests either limitations in our model-
ing method or the fact that these modes are not
important in the triphenolic pyrazoles and only become
important when certain phenols are missing (see above)
or a basic side chain is appended.24
In Figure 4, we show a crossed-stereoview skeletal
model illustrating PPT (4g) in the ERR structure,
oriented in binding mode C(3), after the ligand docking/
minimization routine. For clarity and for the ERR and
ERâ comparison below, only selected residues from ER
are shown. In this orientation, the N(1)-phenyl group
of PPT (4g) projects into a region of the binding pocket
which represents the C/D-subpocket that is normally
occupied by portions of the C- and D-rings and the 18-
methyl substituent on estradiol (E2). Despite the im-
portance of the N(1)-phenol in enhancing ERR binding
selectivity (see above and Table 1), there are no obvious
sites for its interaction in this subpocket. The N(1)- and
C(5)-aryl groups of PPT project somewhat more deeply
into the C/D-subpocket than do the corresponding
regions of estradiol. However, this region is known to
tolerate a variety of substituents,3 and certain residues
have been repositioned by the docking/minimization
routine to make additional space for the pyrazole ligand.
It is of note that these residue conformational changes,
which allow for PPT binding in this mode, involve
relatively minimal alteration in the total protein energy.
3. An a lysis of In ter a ction s Betw een P yr a zole
Su bstitu en ts a n d Am in o Acid Resid u es in ERR
a n d ERâ. We have used the C(3) orientation mode
X
Y
Z
ERR/ERâ
compd N(1) C(3) C(5) RBA ERRb
RBA ERâb RBA ratio
Monophenols
7a
7b
7c
OH
H
H
H
OH
H
H
H
3.1 ( 0.5
2.6 ( 0.1 0.61 ( 0.2
1.5 ( 0.2
2.1
4.3
0.67
OH 0.04 ( 0.11 0.06 ( 0.01
Diphenols
4b
8a
8b
H
OH
OH OH
OH OH
H
31 ( 15
7.0 ( 0.6 0.80 ( 0.09
8.9 ( 0.6 0.32 ( 0.01
1.1 ( 0.2
28
9
28
OH
H
Triphenols
4f
OH OH OH
36 ( 6
0.15 ( 0.014
240
a
Relative binding affinity (RBA), where estradiol is 100%.
Values are the mean of at least 2 and more typically 3 or more
independent determinations ((SD).33 Competitive radiometric
b
binding assays were done with purified full-length human ERR
and ERâ (PanVera Inc.), using 10 nM [3H]E2 as tracer and HAP
for adsorption of the receptor-tracer complex.16,17
the binding pocket.23 As a result of this latter study,23
we have come to appreciate that there are some ambi-
guities in how polyphenolic ligands such as these
pyrazoles might be accommodated by ER, at least at the
level at which we are currently able to model. Such
ambiguity is echoed in the results of the structure-
binding affinity considerations discussed above. There-
fore, for the purposes of this paper, we have chosen to
illustrate a binding mode for these pyrazoles in which
the C(3)-phenyl group is placed in the ER LBD pocket
that normally binds the A-ring of estradiol (Figure 3,
see C(3) and C(3)′), as this appeared to be the most likely
orientation for these polyphenolic systems (see above
and ref 23). We have, however, also examined the other
four possible orientations.
We conducted a ligand docking/binding minimization
routine by placing PPT (4g) in ERR in both orienta-
tions: mode C(3) and mode C(3)′ (Flexidock routine
within Sybyl; see Experimental Section, Molecular

Pyrazole Ligands: ERR-Selective Agonists
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26 4941
F igu r e 5. Crossed-stereoview of ligand skeletal structures and receptor molecular surfaces for the ERR- and ERâ-propylpyrazole
triol (PPT, 4g) complexes. The PPT in ERR is shown with standard atom colors, and the ERR receptor surface is shown in yellow;
in the ERâ-PPT complex, both the ligand and receptor surface are shown in red. For further details, see text.
model derived above to try to understand the origins of
the high ERR affinity selectivity of PPT (4g), and to do
this, we have constructed a model for ERâ that is
derived from ERR LBD-diethylstilbestrol (DES) struc-
ture (see Experimental Section for details).6 Although
the LBDs of ERR and ERâ have only 56% amino acid
identity, the residues that line the ligand-binding pocket
are nearly identical.25 In fact, of the 22-24 residues that
are considered to be in contact with the ligand (i.e.,
within 4 Å), all but two are identical in ERR and ERâ.25
The only differences are at ERR position 384, where
ERR has a leucine but ERâ has a methionine, and at
ERR position 421, where ERR has a methionine but ERâ
has an isoleucine. It has been speculated that these
sequence differences, in particular, may underlie the
differences in the ERR and ERâ binding affinities of
ligands that have been studied so far.25 For clarity, we
have distinguished these residues in Figure 4 by color:
The residues common to ERR and ERâ are illustrated
in yellow, but at the two positions where the ERâ
residues are different, the ERâ residues, colored in red,
are superimposed over the ERR residues in yellow (see
Figure 4 and legend).
In this model with the most ERR potency-selective
ligand PPT (4g), we can see that the pyrazole core itself
has the closest contact with Leu384 in the case of ERR
and Met384 in the case of ERâ. For the Met/Ile
discriminating residues at ERR position 421, the C(4)-
propyl group as well as the C(5)-aryl group appear to
have close contacts, although this interaction seems less
severe.
A model for PPT (4g) binding to ERâ was generated
by changing the two residues in the ERR ligand-binding
pocket to those found in ERâ, inserting the ligand
according to mode C(3), and then conducting the ligand
docking/minimization routine as was done before with
ERR (see Experimental Section). The comparative fit of
this pyrazole in the ligand-binding pocket of ERR and
ERâ can be appreciated by the composite receptor
surface model shown in Figure 5 in crossed-stereoview.
In this figure, we have displayed the pyrazoles in both
of the receptors as skeletal structures, and the two
pockets as colored surfaces. The ligand in the ERR
model is shown with standard atom colors and the ERR
surface in yellow; both the ligand and receptor in the
ERâ model are shown in red. In this figure, the view
has been rotated, twisting the C(3)-phenol away from
the viewer; this orientation is the best we have found
to display differences between the two models.
The principal difference in the surface of the ligand-
binding pocket is at the site where ERR has a smaller
residue (Leu384) than ERâ (Met384). This is illustrated
by the sharp red interpenetrating surface in the middle
left region of the pocket. The increased steric bulk in
ERâ at this site makes contact with the pyrazole core
of this ligand, displacing it relative to the pyrazole
position in the ERR model. Because the C(3)-phenol
remains relatively fixed in the rather rigid A-ring
binding pocket, this pyrazole core displacement causes
a large shift in the position of the N(1)-phenol and a
significant but somewhat smaller shift of the C(5)-
phenol and the propyl group. The large shift of the N(1)-
phenol is intriguing, because this group contributes
significantly to the ERR affinity and potency selectivity
of these pyrazoles. However, we have been unable to
identify any interaction in the N(1)-phenol subpocket
that would favor the presence of a hydroxyl in ERR and
disfavor it in ERâ.
The shift of the propyl group forces this substituent
to adopt a gauche conformation in ERâ (vs an anti
orientation in ERR); this is a change that would most
likely increase ligand internal energy and reduce bind-
ing affinity. The shift of the pyrazole position in the ERâ
model moves the C(5)-phenol toward the site where the
smaller Ile residue in ERâ replaces the bulkier Met421
in ERR. Thus, the different shape of the ligand-binding
pocket in ERâ, in particular the increased bulk of the
ERâ Met residue where ERR has a Leu, appears to
interfere with the optimal binding of the pyrazole ligand
and is therefore thought to account for its lowered

4942 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26
Stauffer et al.
hydroxylapatite (HAP) to adsorb bound receptor-ligand com-
plex.16,17 HAP was prepared following the recommendations
of Williams and Gorski.31 All incubations were done at 0 °C
for 18-24 h. Binding affinities are expressed relative to
estradiol (RBA ) 100%) and are reproducible with a coefficient
of variation of 0.3.
Tr a n scr ip tion Activa tion Assa ys. Human endometrial
cancer (HEC-1) cells were maintained in culture and trans-
fected as described previously.32 Transfection of HEC-1 cells
in 60-mm dishes used 0.4 mL of a calcium phosphate precipi-
tate containing 2.5 µg of pCMVâGal as internal control, 2 µg
of the reporter gene plasmid, 100 ng of the ER expression
vector, and carrier DNA to a total of 5 µg DNA. CAT activity,
normalized for the internal control â-galactosidase activity,
was assayed as previously described.32
Molecu la r Mod elin g for P yr a zole 4g. The starting
conformation for 4g used for receptor docking studies was
generated from a random conformational search performed
using the MMFF94 force field as implemented in Sybyl 6.6.
The resulting lowest-energy conformer was then used for
docking studies. Charge calculations were determined using
the MMFF94 method and molecular surface properties dis-
played using MOLCAD module in Sybyl 6.6.
Pyrazole 4g, generated as noted above, was prepositioned
in the DES-ERR LBD crystal structure (Protein Data Bank
code 3ERD)6 using a least-squares multifitting of select atoms
within the DES ligand. Once prepositioned, DES was deleted
and ligand 4g optimally docked in the ERR binding pocket
using the Flexidock routine within Sybyl (Tripos). Both
hydrogen-bond donors and acceptors within the pocket sur-
rounding the ligand (Glu353, Arg394, and His524), the ligand
itself, and select torsional bonds were defined. The best docked
receptor-ligand complex from Flexidock then underwent a
three-step minimization: first nonring torsional bonds of the
ligand were minimized in the context of the receptor using the
torsmin command, followed by minimization of the side chain
residues within 8 Å of the ligand while holding the backbone
and residues Glu353 and Arg394 fixed. A final third minimiza-
tion of both the ligand and receptor was conducted using the
Anneal function (hot radius 8 Å, interesting radius 16 Å from
pyrazole 4g) to afford the final model.
binding affinity. It is interesting to note that even after
extensive minimization, there is a substantial energy
difference between the two complexes, with the ERR
complex having greater stabilization by ca. 30 kcal/mol.
Con clu sion s
A series of C(4)-alkyl tetrasubstituted pyrazole ana-
logues have been prepared and their RBA values
determined for the ER. These compounds show an
interesting binding affinity pattern, including high-
affinity selectivity for ERR, which for the propylpyrazole
triol (PPT, 4g) reaches 410-fold. This compound also has
very high potency selectivity for ERR in reporter gene
transcription assays in cells, and it is the first ERR-
specific agonist. Structure-affinity relationships in a
series of mono-, di-, and triphenolic pyrazoles and
molecular modeling suggests that these pyrazoles bind
to the ER with the C(3)-phenol in the A-ring binding
pocket. A model of PPT (4g) docked in the ERR LBD-
DES crystal structure in the most likely binding orien-
tation suggests that the very high ERR binding selec-
tivity of this pyrazole derives from particular interactions
between the pyrazole core and the C(4)-propyl group
with a region on the ligand-binding pocket where ERR
has a smaller residue (Leu384) than ERâ (Met384).
These interactions may serve as the basis for future
structure-based design of ERR- and ERâ-specific ligands.
Exp er im en ta l Meth od s
Gen er a l. All reagents and solvents were obtained from
Aldrich, Fisher, or Mallinckrodt. Tetrahydrofuran was freshly
distilled from sodium/benzophenone. Dimethylformamide was
vacuum distilled prior to use and stored over 4 Å molecular
sieves. Melting points were determined on a Thomas-Hoover
UniMelt capillary apparatus and are uncorrected. All reactions
were performed under a dry N2 atmosphere unless otherwise
specified. Reaction progress was monitored by analytical thin-
layer chromatography using GF silica plates purchased from
Analtech. Visualization was achieved by short-wave UV light
(254 nm) or potassium permanganate stain. Radial prepara-
tive-layer chromatography was performed on a Chromatotron
instrument (Harrison Research, Inc., Palo Alto, CA) using EM
Science silica gel Kieselgel 60 PF254 as adsorbent. Flash column
chromatography was performed using Woelm 32-63 µm silica
gel packing.26 Synthetic procedures for 1a -f and spectral data
for 1d ,f are described below. Compounds 1a ,27 1b,28 1c,29 and
1e30 were spectroscopically identical to the reported com-
pounds. The synthesis of compounds 7b,c has been described
elsewhere.15
The ERâ molecular model was generated within Sybyl by
first modifying residues Leu384 to Met384 and Met421 to
Ile421 in the DES-ΕRR LBD crystal structure6 and conducting
an initial minimization of these two residues while holding
the remaining atoms fixed. This was followed by a final
minimization using the Anneal function (hot radius 4 Å,
interesting radius 8 Å from Met384 and Ile421) using condi-
tions similar to above. The pyrazole 4g was introduced into
this ERâ model and the Flexidock routine implemented as for
the ERR model (see above). All minimizations were done using
the MMFF94 force field with the Powell gradient method (final
rms < 0.02 kcal/mol‚Å).
1H and 13C NMR spectra were recorded on either a Varian
Unity 400 or 500 MHz spectrometers using CDCl3 or MeOD-
d4 as solvent. Chemical shifts are reported as parts per million
downfield from an internal tetramethylsilane standard (δ 0.0
for 1H) or from solvent references. NMR coupling constants
are reported in hertz (Hz). 13C NMR were determined using
either the attached proton test (APT) or standard 13C pulse
sequence parameters. Low- and high-resolution electron im-
pact mass spectra were obtained on Finnigan MAT CH-5 or
70-VSE spectrometers. Low- and high-resolution fast atom
bombardment (FAB) were obtained on a VG ZAB-SE spec-
trometer. Elemental analyses were performed by the Micro-
analytical Service Laboratory of the University of Illinois.
Those final components that did not give satisfactory combus-
tion analysis (i.e., 4b-d , 6, 8a ) gave satisfactory exact mass
determinations and were found to be at least 98% pure by
HPLC analysis under normal and reversed-phase conditions.
RBA Assa ys. Purified ERR and ERâ binding affinities were
determined using a competitive radiometric binding assay
using 10 nM [3H]estradiol as tracer, commercially available
ΕRR and ERâ preparations (PanVera Inc., Madison, WI), and
Ch em ica l Syn th eses. Gen er a l P r oced u r e for P r ep a r a -
tion of Alk ylp h en on es Usin g AlCl3 (1a -c,f): Meth od A.
To a stirred solution of AlCl3 (46.8 mmol) in 1,2-dichloroethane
(10 mL) at 0 °C was added the commercial acid chloride (39.0
mmol) dropwise over 10 min. The resulting solution was
allowed to come to room temperature for 20 min until all of
the AlCl3 had dissolved. The reaction mixture was cooled to 0
°C and a solution of anisole (5.1 mL, 46.8 mmol) in 1,2-
dichloroethane (20 mL) was then added dropwise over 30 min.
Upon completion, the reaction was allowed to reach room
temperature and stir for 8-15 h. The mixture was then
quenched by pouring over 100 g of ice and extracted with
CH2Cl2 (3 × 25 mL). The combined organic layers were washed
with water, NaHCO3 (satd), brine, then dried over MgSO4 and
concentrated under reduced pressure. Excess anisole was
removed in vacuo and then the ketone product distilled.
Gen er a l P r oced u r e for P r ep a r a tion of Alk ylp h en on es
Usin g P P A (1d ,e): Meth od B. A mechanically stirred
mixture consisting of carboxylic acid (1.0 equiv), anisole (1.1
equiv) and polyphosphoric acid (PPA; 6.15 g/mL of anisole) was
heated to 90-100 °C for 1.5 h. Upon cooling to near room

Pyrazole Ligands: ERR-Selective Agonists
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26 4943
temperature, the dark mixture was poured over ice (100 g/0.1
mol) and extracted repeatedly with EtOAc. The organic layers
were then washed with satd NaHCO3 followed by brine. After
drying the extract over Na2SO4 and removal of solvent, a crude
oil was obtained. The final products were purified by bulb-to-
bulb distillation.
1-(4-Meth oxyp h en yl)-4-m eth ylp en ta n -1-on e (1d ). Pre-
pared according to method B outlined above to afford the title
compound as a pale yellow oil (75%): 1H NMR (CDCl3, 500
MHz) δ 0.94 (d, 6H, J ) 6.0), 1.60 (m, 3H, overlapping methine
and â-CH2), 2.90 (t, 2H, J ) 6.8), 3.87 (s, 3H), 6.92 (d, 2H, J
) 8.8), 7.90 (d, 2H, J ) 8.8); 13C NMR (CDCl3, 125 MHz) δ
22.6, 28.5, 33.9, 36.4, 55.2, 113.5, 130.2, 163.6, 198.7; MS (EI,
70 eV) m/z 206.1 (M+). Anal. (C13H18O2) C, H.
1-(4-Met h oxyp h en yl)-3-m et h ylb u t a n -1-on e (1f). Pre-
pared according to general method A outlined above to afford
the title compound as a pale yellow oil (84%): 1H NMR (CDCl3,
500 MHz) δ 0.95 (d, 6H, J ) 6.7), 2.24 (sept, 1H, J ) 6.6), 2.73
(d, 2H, J ) 6.7), 3.81 (s, 3H), 6.88 (d, 2H, J ) 8.9), 7.90 (d,
2H, J ) 9.2); 13C NMR (CDCl3, 125 MHz) δ 22.6, 25.2, 46.2,
55.2, 113.5, 130.2, 130.3, 163.1, 198.7; HRMS (EI, M+) calcd
for C12H16O2 192.1150, found 192.1153.
2-Me t h yl-1,3-b is(4-m e t h oxyp h e n yl)p r op a n e -1,3-d i-
on e (2a ). To a solution of 1a (160 mg, 0.97 mmol) and
4-nitrophenyl 4-methoxybenzoate (prepared from p-nitro-
phenol and 4-methoxybenzoic acid using diisopropylcarbodi-
imide and 4-(dimethyamino)pyridine in THF (35 mL) at 0 °C
was added a 1.0 M solution of lithium hexamethyldisilamide
(3.03 mL, 3.03 mmol) dropwise over 5 min. The reaction was
warmed to room temperature and allowed to stir for 1.5 h. At
this time the reaction was quenched by the addition of H2O
(25 mL). The mixture was then repeatedly extracted with
diethyl ether. The organic layers were combined and washed
with H2O, then dried over Na2SO4 and concentrated under
reduced pressure to afford a crude solid. Unreacted ester was
removed by adding a solution of 40% ethyl acetate/hexanes
and filtering off the insoluble ester. The remaining filtrate was
concentrated and subjected to flash chromatography (40%
ethyl acetate/hexanes) to afford the title compound as a light
yellow oil (266 mg, 92%): 1H NMR (CDCl3, 500 MHz) δ 1.58
(d, 3H, J ) 7.3), 3.85 (s, 3H), 5.13 (q, 1H, J ) 7.2), 6.92 (d, 2H,
J ) 6.8), 7.93 (d, 2H, J ) 7.0).
2-E t h yl-1,3-b is(4-m et h oxyp h en yl)p r op a n e-1,3-d ion e
(2b).11 Prepared according to the procedure outlined above
from 1b and purified by flash chromatography (ethyl acetate/
hexanes) to afford a viscious oil (95%): 1H NMR (CDCl3, 500
MHz) δ 1.01 (t, 3H, J ) 7.4), 2.12 (quint, 2H, J ) 7.1), 3.80 (s,
6H), 4.98 (t, 1H, J ) 6.6), 6.87 (AA′XX′, 2H, J ) 8.8, 2.1), 7.94
(AA′XX′, 2H, J ) 8.5, 2.3); APT 13C NMR (CDCl3, 125 MHz) δ
12.9 (CH3), 23.2 (CH2), 55.6 (CH), 58.8 (CH3O), 114.0 (ArCH),
129.3 (ArC), 131.0 (ArCH), 163.8 (ArC), 195.0 (CdO); MS (EI,
70 eV) m/ z 312.3 (M+).
on e (2e). Prepared according to the procedure outlined above
from 1e and purified by flash chromatography (ethyl acetate/
hexanes) to afford product as a light yellow oil (69%): 1H NMR
(CDCl3, 500 MHz) δ 0.89 (t, 3H, J ) 7.0), 1.37 (m, 4H), 2.11
(q, 2H, J ) 7.1), 3.85 (s, 6H), 6.91 (d, 4H, J ) 9.0), 7.98 (d, 4H,
J ) 9.0); 13C NMR (CDCl3, 125 MHz) δ 14.0 (CH3), 22.9 (CH2),
22.7 (CH2), 30.7 (CH2), 55.7 (CH3O), 57.7 (CH), 114.2 (ArCH),
129.5 (ArC), 131.2 (ArCH), 163.2 (CdO); HRMS (EI, M+) calcd
for C13H18O4 340.1674, found 340.1675. Anal. (C13H18O4) C, H.
Gen er a l P r oced u r e for P yr a zole Syn th esis. To a DMF
(30 mL) and THF (10 mL) solution containing diketone (1.0
mmol) was added phenylhydrazine hydrochloride (3-5 equiv).
The mixture was brought to reflux (oil bath temperature 120
°C) until disappearance of diketone as evident by TLC analysis
(8-20 h). The reaction mixture was then allowed to cool to
room temperature and diluted with H2O (30 mL). The product
was extracted repeatedly with ethyl acetate (3 × 25 mL) and
the organic layers combined and washed sequentially with a
satd LiCl solution (25 mL), satd NaHCO3 (25 mL), and brine
(25 mL). The oganic layer was dried over Na2SO4 and
concentrated under reduced pressure to afford the crude
product in the form of an oil, which was purified by flash
chromatography or by passage through a short silica plug
eluting with an ethyl acetate/hexane solvent system.
3,5-Bis(4-m eth oxyp h en yl)-4-m eth yl-1-p h en yl-1H-p yr a -
zole (3a ). The diketone 2a (250 mg, 0.84 mmol) was reacted
with phenylhydrazine hydrochloride (423 mg, 2.94 mmol)
according to the general procedure above. Upon purification
by flash chromatography (40% ethyl acetate/hexanes) the title
compound was obtained as a tan solid (220 mg, 71%): mp 142-
143 °C; 1H NMR (CDCl3, 500 MHz) δ 2.21 (s, 3H), 3.83 (s, 3H),
3.86 (s, 3H), 6.89 (AA′XX′, 2H, J ) 8.9, 2.5), 6.99 (AA′XX′, 2H,
J ) 8.9, 2.5), 7.15 (AA′XX′, 2H, J ) 8.8, 2.5), 7.19-7.32 (m,
5H), 7.74 (AA′XX′, 2H, J ) 8.9, 2.5); 13C NMR (CDCl3, 125
MHz) δ 10.2, 55.2, 55.3, 113.6, 113.9, 113.9, 123.1, 124.7, 126.5,
126.6, 128.7, 129.1, 131.3, 140.4, 141.2, 150.9, 159.2, 159.4;
MS (EI, 70 eV) m/z 370 (M+). Anal. (C24H22N2O2) C, H, N.
4-E t h yl-3,5-b is(4-m et h oxyp h en yl)-1-p h en yl-1H -p yr a -
zole (3b).11 Diketone 2b and phenylhydrazine hydrochloride
were reacted as outlined above to afford 3b as an orange solid
after flash chromatography purification (87%): 1H NMR
(CDCl3, 400 MHz) δ 1.04 (t, 3H, J ) 7.6), 2.63 (q, 2H, J )
7.6), 3.83 (s, 3H), 3.86 (s, 3H), 6.90 (AA′XX′, 2H, J ) 8.8, 2.4),
6.99 (AA′XX′, 2H, J ) 8.8, 2.6), 7.17 (AA′XX′, 2H, J ) 8.8,
2.4), 7.20 (m, 2H), 7.24 (m, 3H), 7.72 (AA′XX′, 2H, J ) 9.0,
2.4); 13C NMR (CDCl3, 100 MHz) δ 15.8, 17.3, 55.4, 55.5, 114.1,
114.2, 120.7, 123.5, 124.8, 126.8, 127.0, 128.8, 129.3, 131.5,
140.5, 141.2, 150.8, 159.4, 159.6; HRMS (EI, M+) calcd for
C25H24N2O2 384.1835, found 384.1837.
3,5-Bis(4-m eth oxyp h en yl)-1-p h en yl-4-p r op yl-1H-p yr a -
zole (3c). Diketone 2c (200 mg, 0.61 mmol) was reacted with
phenylhydrazine hydrochloride (444 mg, 3.07 mmol) according
to the general procedure to afford 3c as an orange oil (130
mg, 53%) after purification by flash chromatography (40%
2-P r op yl-1,3-bis(4-m eth oxyp h en yl)p r op a n e-1,3-d ion e
(2c). Prepared according to the procedure outlined above from
1c and purified by flash chromatography (ethyl acetate/
hexanes) to afford product as a clear viscous oil (76%): 1H
NMR (CDCl3, 500 MHz) δ 0.95 (t, 3H, J ) 7.4), 1.36-1.47 (m,
2H), 2.06-2.11 (m, 2H), 3.83 (s, 6H), 5.05 (t, 1H, J ) 6.7) 6.90
1
ethyl acetate/hexanes): H NMR (CDCl3, 500 MHz) δ 0.79 (t,
3H, J ) 7.4), 1.42 (sext, 2H, J ) 7.7), 2.57 (t, 2H, J ) 8.0),
3.83 (s, 3H), 3.86 (s, 3H), 6.90 (AA′XX′, 2H, J ) 8.6, 2.5), 6.99
(AA′XX′, 2H, J ) 8.8, 2.5), 7.18-7.32 (m, 5H), 7.15 (AA′XX′,
2H, J ) 8.6, 2.5), 7.72 (AA′XX′, 2H, J ) 8.8, 2.5); 13C NMR
(CDCl3, 125 MHz) δ 14.05, 23.79, 25.86, 55.09, 55.15, 113.74,
113.85, 118.87, 118.86, 122.86, 124.60, 126.52, 129.49, 131.17,
139.66, 141.31, 150.36, 159.14, 159.33; HRMS (EI, M+) calcd
for C26H26N2O2 398.1994, found 398.2000.
(AA′XX′, 2H, J ) 9.0, 2.5), 7.95 (AA′XX′, 2H, J ) 9.0, 2.5); 13
C
NMR (CDCl3, 125 MHz) δ 14.0, 21.5, 31.6, 55.4, 57.1, 113.8,
129.0, 130.8, 163.5, 194.8; HRMS (EI, M+) calcd for C20H22O4
326.1518, found 326.1512.
2-Isob u t yl-1,3-b is(4-m et h oxyp h en yl)p r op a n e-1,3-d i-
on e (2d ). Prepared according to the procedure outlined above
from 1d and purified by flash chromatography (ethyl acetate/
hexanes) to afford product as a clear viscious oil (77%): 1H
NMR (CDCl3, 500 MHz) δ 0.96 (d, 3H, J ) 6.5), 1.70 (m, 1H,
J ) 6.7), 2.00 (br t, 2H, J ) 6.9), 3.82 (s, 6H), 5.18 (t, 1H, J )
6.5), 6.89 (d, 2H, J ) 8.6), 7.97 (d, 2H, J ) 9.0); 13C NMR
(CDCl3, 125 MHz) δ 22.8 (CH3), 27.2 (CH), 38.5 (CH2), 55.7
(CH3O), 55.8 (CH), 114.2 (ArCH), 129.4 (ArC), 131.2 (ArCH),
163.2 (CdO); HRMS (EI, M+) calcd for C13H18O4 340.1674,
found 340.1679. Anal. (C21H24O4) C, H.
4-Isobu tyl-3,5-bis(4-m eth oxyph en yl)-1-ph en yl-1H-pyr a-
zole (3d ). Diketone 2d was reacted with phenylhydrazine
hydrochloride according to the general procedure to afford 3d
as an orange oil (85%) after a short silica plug (20% ethyl
acetate/hexanes). The purified material was then directly used
in the subsequent deprotection step.
4-Bu t yl-3,5-b is(4-m et h oxyp h en yl)-1-p h en yl-1H -p yr a -
zole (3e). Diketone 2e was reacted with phenylhydrazine
hydrochloride according to the general procedure to afford 3e
as a reddish oil (86%) after purification by flash chromatog-
2-n -Bu t yl-1,3-b is(4-m e t h oxyp h e n yl)p r op a n e -1,3-d i-

4944 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26
Stauffer et al.
raphy (20% ethyl acetate/hexanes). This material was then
directly used in the subsequent deprotection step.
ylation procedure. The crude product was purified by flash
chromatography (10% CH3OH/CH2Cl2) to afford 4c as a tan
solid (85 mg, 86%): mp 240-245 °C; 1H NMR (MeOD-d4, 500
MHz) δ 0.72 (t, 3H, J ) 7.4), 1.35 (sext, 2H, J ) 7.5), 2.55 (t,
2H, J ) 7.7), 6.77 (AA′XX′, 2H, J ) 8.7, 2.5), 6.88 (AA′XX′,
2H, J ) 8.5, 2.4), 7.03 (AA′XX′, 2H, J ) 8.5, 2.4), 7.32-7.27
(m, 5H), 7.50 (AA′XX′, 2H, J ) 8.8, 2.5); 13C NMR (MeOD-d4,
400 MHz) δ 12.73, 13.28, 25.28, 114.75, 114.89, 118.26, 121.13,
124.78, 124.82, 126.69, 128.26, 128.99, 131.02, 139.71, 142.12,
151.30, 157.08, 157.49; HRMS (EI, M+) calcd for C24H22N2O2
370.1681, found 370.1676.
4-Isobu tyl-3,5-bis(4-h ydr oxyph en yl)-1-p h en yl-1H-p yr a -
zole (4d ). A stirred CH2Cl2 solution of 3d (100 mg, 0.24 mmol)
was deprotected using BBr3 according to the general demeth-
ylation procedure. The crude product was purified by flash
chromatography (10% CH3OH/CH2Cl2) to afford 4d as a tan
powder (70 mg, 76%): mp 225 °C dec; 1H NMR (MeOD-d4, 500
MHz) δ 0.63 (d, 6H, J ) 6,5), 1.51 (m, 1H, J ) 7.0), 2.51 (d,
2H, J ) 7.5), 6.76 (AA′XX′, 2H, J ) 9.0, 2.3), 6.87 (AA′XX′,
2H, J ) 8.9, 2.3), 7.02 (AA′XX′, 2H, J ) 8.5, 2.4), 7.26 (m, 5H),
7.49 (AA′XX′, 2H, J ) 8.0, 2.3); 13C NMR (MeOD-d4, 125 MHz)
δ 22.8, 29.9, 33.8, 116.4, 116.6, 119.1, 123.0, 126.6, 126.8, 128.4,
129.9, 130.9, 132.9, 141.5, 144.2, 153.4, 158.8, 159.2; FAB-
HRMS (M + 1) calcd for C25H25N2O2 385.1916, found 385.1916.
4-Eth yl-1,3,5-tr is(4-m eth oxyp h en yl)-1H-p yr a zole (3f).
Diketone 2b (400 mg, 1.28 mmol) was reacted with 4-meth-
oxyphenylhydrazine hydrochloride (1.11 g, 6.40 mmol) accord-
ing to the general procedure. The crude product was purified
by flash chromatography (2% acetone/CH2Cl2) to afford 3f as
an oil (351 mg, 66%): 1H NMR (CDCl3, 500 MHz) δ 1.06 (t,
3H, J ) 7.5), 2.65 (q, 2H, J ) 7.5), 3.73 (s, 3H), 3.79 (s, 3H),
3.83 (s, 3H), 6.78 (AA′XX′, 2H, J ) 9.0, 2.8) 6.89 (AA′XX′, 2H,
J ) 8.8, 2.4), 6.69 (AA′XX′, 2H, J ) 8.8, 2.4), 7.16 (AA′XX′,
2H, J ) 8.8, 2.4), 7.21 (AA′XX′, 2H, J ) 9.0, 2.7), 7.74 (AA′XX′,
2H, J ) 8.7, 2.4); 13C NMR (CDCl3, 125 MHz) δ 15.52, 17.09,
55.05, 55.10, 55.22, 113.65, 113.80, 113.83, 119.78, 123.11,
125.96, 126.78, 128.95, 131.18, 133.45, 140.84, 149.94, 158.99,
157.97, 159.22; HRMS (EI, M+) calcd for C26H26N2O3 414.1943,
found 414.1942.
1,3,5-Tr is(4-m eth oxyph en yl)-4-pr opyl-1H-pyr azole (3g).
Diketone 2c (500 mg, 1.52 mmol) was reacted with 4-meth-
oxyphenylhydrazine hydrochloride (800 mg, 4.59 mmol) ac-
cording to the general procedure to afford 362 mg of 3g as a
red oil (67%) after a short silica plug (20% ethyl acetate/
hexanes). This material was then directly used in the subse-
quent deprotection step.
4-Isobu tyl-1,3,5-tr is(4-m eth oxyph en yl)-1H-pyr azole (3h ).
Diketone 2d was reacted with 4-methoxyphenylhydrazine
hydrochloride according to the general procedure to afford 3h
as a light orange oil (85%) after a short silica plug (20% ethyl
acetate/hexanes). The purified material was then directly used
in the subsequent deprotection step.
4-Bu tyl-1,3,5-tr is(4-m eth oxyp h en yl)-1H-p yr a zole (3i).
Diketone 2e was reacted with 4-methoxyphenylhydrazine
hydrochloride according to the general procedure to afford 3i
as a light orange oil (87%) after a short silica plug (20% ethyl
acetate/hexanes). The purified material was then directly used
in the subsequent deprotection step.
Gen er a l Dem eth yla tion P r oced u r e. To a stirred solution
of methyl-protected pyrazole (1 equiv) in CH2Cl2 at -78 °C
was added dropwise a 1 M BBr3 solution in CH2Cl2 (3-5
equiv). Upon complete addition of BBr3, the reaction was
maintained at -78 °C for 1 h and then allowed to reach room
temperature and stir for an additional 16 h. The mixture was
cooled to 0 °C and carefully quenched with H2O (15-25 mL).
The product was then repeatedly extracted with EtOAc and
the organic layers dried over Na2SO4. Upon solvent removal
the crude phenolic products were purified by flash chroma-
tography and/or recrystallization from MeOH/CH2Cl2 mix-
tures.
4-Bu t yl-3,5-b is(4-h yd r oxyp h en yl)-1-p h en yl-1H -p yr a -
zole (4e). A stirred CH2Cl2 solution of 3e (100 mg, 0.24 mmol)
was deprotected using BBr3 according to the general demeth-
ylation procedure. The crude product was purified by flash
chromatography (30% EtOAc/hexanes) to afford a tan solid.
This material was subsequently recrystallized from 5-10%
MeOH/CH2Cl2 to afford the title compound as small off-white
crystals (59 mg, 64%): mp 205.5-207.5 °C; 1H NMR (MeOD-
d4, 400 MHz) δ 0.72 (t, 3H, J ) 7.2), 1.16 (sext, 2H, J ) 7.2),
1.34 (quint, 2H, J ) 7.2), 2.60 (t, 2H, J ) 7.2), 4.95 (br s, 2H
exchange with D2O), 6.78 (d, 2H, J ) 9.0), 6.91 (d, 2H, J )
8.0), 7.02 (d, 2H, J ) 8.8), 7.26 (m, 5H), 7.53 (d, 2H, J ) 8.4);
13C NMR (MeOD-d4, 125 MHz) δ 14.6, 23.9, 24.8, 34.3, 116.8,
116.9, 120.5, 123.2, 126.8, 128.7, 130.3, 131.0, 131.1, 133.1,
141.7, 144.0, 153.3, 159.0, 159.4. Anal. (C25H24N2O2‚0.1H2O)
C, H, N.
4-Eth yl-1,3,5-tr is(4-h yd r oxyp h en yl)-1H-p yr a zole (4f). A
stirred CH2Cl2 solution of 3f (200 mg, 0.48 mmol) was
deprotected using BBr3 according to the general demethylation
procedure. A crude oil was isolated which was triturated with
a 10% CH3OH/CH2Cl2 solution from which the desired product
precipitated. The white powder was collected by filtration and
recrystallized from CH3OH/CH2Cl2 to afford the title compound
1
4f (175 mg, 98%): mp 210-215 °C; H NMR (MeOD-d4, 400
MHz): δ 0.96 (t, 3H, J ) 7.5) 2.58 (q, 2H, J ) 7.5) 6.70 (AA′XX′,
2H, J ) 8.8, 2.6), 6.77 (AA′XX′, 2H, J ) 8.6, 2.3), 6.87 (AA′XX′,
2H, J ) 8.8, 2.3), 7.10 (m, 4H), 7.48 (AA′XX′, 2H, J ) 8.6, 2.4);
13C NMR (MeOD-d4, 100 MHz) δ 14.5, 16.6, 114.8, 114.9, 114.9,
119.2, 121.4, 125.0, 126.8, 129.1, 131.2, 131.9, 142.1, 150.5,
156.7, 157.1, 157.5. Anal. (C23H20N2O3‚0.7H2O) C, H, N.
3,5-Bis(4-h yd r oxyp h en yl)-4-m eth yl-1-p h en yl-1H-p yr a -
zole (4a ). A stirred CH2Cl2 solution of 3a (200 mg, 0.54 mmol)
was deprotected using BBr3 according to the general demeth-
ylation procedure. Purification by flash chromatography (5%
CH3OH/CH2Cl2) afforded the title compound as a tan solid (54
1
mg, 30%): mp 225-230 °C; H NMR (MeOD-d4, 500 MHz) δ
2.13 (s, 3H), 6.76 (AA′XX′, 2H, J ) 8.4, 2.7), 6.93 (AA′XX′, 2H,
J ) 8.7, 2.5), 7.01 (AA′XX′, 2H, J ) 8.8, 2.5), 7.54 (AA′XX′,
2H, J ) 8.6, 2.4), 7.35-7.22 (m, 5H); 13C NMR (MeOD-d4, 125
MHz) 8.7, 112.8, 114.7, 114.8, 120.9, 124.4, 124.8, 126.7, 128.3,
128.9, 130.9, 139.8, 151.4, 157.1, 157.4; MS (EI, 70 eV) m/z
342. Anal. (C22H18N2O2‚H2O) C, H, N.
1,3,5-Tr is(4-h ydr oxyph en yl)-4-pr opyl-1H-pyr azole (4g).
A stirred CH2Cl2 solution of 3g (200 mg, 0.48 mmol) was
deprotected using BBr3 according to the general demethylation
procedure. A crude oil was isolated which was triturated with
a 10% CH3OH/CH2Cl2 solution from which the desired product
precipitated. The white powder was collected by filtration and
recrystallized from CH3OH/CH2Cl2 to afford the title compound
4-E t h yl-3,5-b is(4-h yd r oxyp h en yl)-1-p h en yl-1H -p yr a -
zole (4b).11 A stirred CH2Cl2 solution of 3b (100 mg, 0.26
mmol) was deprotected using BBr3 according to the general
demethylation procedure. Purification by flash chromatogra-
phy (5% CH3OH/CH2Cl2) afforded the title compound as a
1
4g (125 mg, 68%): mp 230 °C dec; H NMR (MeOD-d4, 400
MHz) δ 0.72 (t, 3H, J ) 7.2), 1.33 (sext, 2H, J ) 7.6), 2.54 (t,
2H, J ) 8), 6.70 (AA′XX′, 2H, J ) 8.8, 2.4), 6.76 (AA′XX′, 2H,
J ) 6.8, 2.0), 6.87 (AA′XX′, 2H, J ) 8.8, 2.4), 7.02 (AA′XX′,
2H, J ) 8.8, 2.4), 7.05 (AA′XX′, 2H, J ) 9.2, 2.4), 7.47 (AA′XX′,
2H, J ) 8.8, 2.0); 13C NMR (MeOD-d4, 100 MHz) δ 25.8 (CH3),
36.4 (CH2), 38.4 (CH2), 127.0 (C), 128.0 (C), 128.5 (C), 130.5
(CH), 134.3 (CH), 137.9 (CH), 138.9 (C), 140.3 (C), 141.3 (C),
142.8 (C), 143.3 (C), 144.1 (C),144.7 (CH), 155.3 (CH), 163.6
(CH), 169.5 (CH), 170.3 (CH). Anal. (C24H22N2O3‚0.6H2O) C,
H, N.
1
white solid (50 mg, 54%): mp 247-248 °C; H NMR (MeOD-
d4, 400 MHz) δ 0.98 (t, 3H, J ) 7.4), 2.60 (q, 2H, J ) 7.5), 6.78
(AA′XX′, 2H, J ) 8.8, 2.4), 6.88 (AA′XX′, 2H, J ) 8.7, 2.5),
7.05 (AA′XX′, 2H, J ) 8.8, 2.4), 7.24-7.42 (m, 5H), 7.51
(AA′XX′, 2H, J ) 8.7, 2.5); HRMS (EI, M+) calcd for C23H21N2O2
357.1611, found 357.1603.
3,5-Bis(4-h yd r oxyp h en yl)-1-p h en yl-4-p r op yl-1H-p yr a -
zole (4c). A stirred CH2Cl2 solution of 3c (107 mg, 0.27 mmol)
was deprotected using BBr3 according to the general demeth-
4-Isobu tyl-1,3,5-tr is(4-h ydr oxyph en yl)-1H-pyr azole (4h ).
A stirred CH2Cl2 solution of 3h (250 mg, 0.56 mmol) was

Pyrazole Ligands: ERR-Selective Agonists
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26 4945
deprotected using BBr3 according to the general demethylation
procedure. Recrystallization from CH3OH/CH2Cl2 afforded the
title compound 4h as an off-white powder (80 mg, 37%): mp
226-228 °C; 1H NMR (MeOD-d4, 500 MHz) δ 0.61 (d, 6H, J )
6.5), 1.50 (m, 1H, J ) 7.0), 2.50 (d, 2H, J ) 7.5), 4.87 (s, 3H,
OH), 6.70 (AA′XX′, 2H, J ) 8.5, 2.0), 6.75 (AA′XX′, 2H, J )
8.5, 2.3), 6.85 (AA′XX′, 2H, J ) 9.0, 2.3), 7.01 (AA′XX′, 2H, J
) 9.0, 2.3), 7.05 (AA′XX′, 2H, J ) 9.0, 2.3), 7.45 (AA′XX′, 2H,
J ) 8.5, 2.3); 13C NMR (MeOD-d4, 125 MHz) δ 22.9, 29.9, 33.9,
116.4, 116.5, 116.6, 118.3, 123.2, 126.9, 128.3, 130.9, 132.9,
133.5, 144.3, 152.7, 158.2, 158.6, 158.9. Anal. (C25H24N2O3‚
0.7H2O) C, H, N.
4-Bu tyl-1,3,5-tr is(4-h yd r oxyp h en yl)-1H-p yr a zole (4i). A
stirred CH2Cl2 solution of 3i (200 mg, 0.45 mmol) was
deprotected using BBr3 according to the general demethylation
procedure. Recrystallization from CH3OH/CH2Cl2 afforded the
title compound 4i as an off-white powder (90 mg, 50%): mp
214-230 °C dec; 1H NMR (MeOD-d4, 500 MHz) δ 0.71 (t, 3H,
J ) 7.5), 1.13 (sext, 2H, J ) 7.0), 1.30 (quint, 2H, J ) 8.5),
2.57 (t, 2H, J ) 8.0), 6.70 (AA′XX′, 2H, J ) 9.0, 2.4), 6.76
(AA′XX′, 2H, J ) 8.5, 2.3), 6.86 (AA′XX′, 2H, J ) 9.0, 2.5),
7.02 (AA′XX′, 2H, J ) 8.5, 2.3), 7.05 (AA′XX′, 2H, J ) 8.5,
2.3), 7.46 (AA′BB′, 2H, J ) 8.5, 2.3); 13C NMR (MeOD-d4, 400
MHz) δ 14.1, 23.5, 24.4, 33.9, 116.3, 116.4, 116.5, 119.3, 123.0,
126.7, 128.3, 130.7, 132.8, 133.5, 143.9, 152.4, 158.3, 158.7,
159.0. Anal. (C25H24N2O3‚0.3H2O) C, H, N.
126.6, 127.7, 128.0, 128.2, 128.4, 128.5, 130.0, 130.6, 132.6,
133.8, 141.6, 150.5, 155.2; MS (EI, 70 eV) m/ z 340.2 (M+).
Anal. (C23H20ON2) C, H, N.
1,3-Bis(4-m et h oxyp h en yl)-5-p h en yl-4,5-d ih yd r o-1H -
p yr a zole (13a ). A mixture of commercially available 4′-
methoxychalcone (12a ; 253 mg, 1.1 mmol) and 807 mg of
4-methoxyphenylhydrazine (4.6 mmol) in 10 mL of anhydrous
DMF was heated to 85 °C overnight. The reaction solution was
cooled to room temperature and partitioned with diethyl ether
and water. The organic layer was washed with water, dried
(MgSO4), and concentrated. The crude was then recrystallized
from a mixture of ethyl acetate and hexanes to give 198 mg of
13a as light yellow crystals (48%): 1H NMR (CDCl3, 400 MHz)
δ 3.12 (dd, 1H, J ) 16.7, 8.4), 3.74 (s, 3H), 3.86 (s, 3H), 3.74-
3.88 (m, 1H), 5.14 (dd, 1H, J ) 11.9, 8.6), 6.8 (AA′XX′, 2H, J
) 9.0), 6.94 (AA′XX′, 2H, J ) 9.0), 7.04 (AA′XX′, 2H, J ) 9.0),
7.26-7.42 (m, 5H), 7.68 (d, 2H, J ) 9.0); 13C NMR δ 43.9, 55.3,
55.6, 65.8, 114.0, 114.4, 114.8, 126.1, 127.5, 127.8, 128.0, 129.1,
140.1, 142.9, 146.9, 153.2, 160.0; HRMS (EI, M+) calcd for
C23H22N2O2 358.1688, found 358.1681.
1,5-Bis(4-m et h oxyp h en yl)-3-p h en yl-4,5-d ih yd r o-1H -
p yr a zole (13b). A mixture of commercially available 4-meth-
oxychalcone (12b ; 2 g, 8.4 mmol) and 7.3 g of 4-methoxy-
phenylhydrazine (42 mmol) in 80 mL of anhydrous DMF was
heated to 85 °C overnight. The reaction solution was cooled to
room temperature and partitioned with diethyl ether and
water. A yellow solid precipitated and was collected by
filtration to give 2.6 g of 13b (86%): 1H NMR (CDCl3, 400
MHz) δ 3.1 (dd, 1H, J ) 15.6,7.8), 3.72 (s, 3H), 3.78 (s, 3H),
3.8 (d, 1H, J ) 5.6), 5.13 (dd, 1H, J ) 12.0, 9.0), 6.76 (AA′XX′,
2H, J ) 9.1, 2.3), 6.87 (AA′XX′, 2H, J ) 8.7, 2.1), 7.01 (d, 2H,
J ) 9.0), 7.2-7.4 (m, 5H), 7.7 (d, 2H, J ) 7.3); 13C NMR δ
43.7, 55.2, 55.6, 65.2, 114.4, 114.4, 114.9, 125.6, 127.3, 128.3,
128.5, 132.9, 134.7, 139.7, 146.3, 153.3, 158.9; MS (EI, 70 eV)
m/ z 358.2 (M+).
4-Eth yl-1,3-bis(4-m eth oxyph en yl)-5-ph en yl-4,5-dih ydr o-
1H-p yr a zole (14a ). To a solution of lithium diisopropylamide
in 20 mL of THF [prepared by dropwise addition of 0.88 mL
of n-BuLi (1.41 mmol) to 0.21 mL (1.5 mmol) of diisopropyl-
amine in 18 mL of THF at -78 °C] was added a solution of
317 mg (0.88 mmol) of pyrazoline 13a in 8 mL of THF dropwise
via syringe at -78 °C and stirred for 1 h. To the dark red
solution was added iodoethane (0.11 mL, 1.31 mmol) in one
portion and the resulting yellow solution was warmed to room
temperature overnight. The reaction was quenched with 5 mL
of brine, the aqueous layer was separated and extracted with
CH2Cl2, the organic layer was dried (MgSO4) and concentrated
in vacuo. The residue was purified by flash chromatography
(ether/hexanes, 3:2) to afford 262 mg of 14a as a yellow foam
(77%): 1H NMR (CDCl3, 400 MHz) δ 0.8 (t, 3H, J ) 7.3), 1.45
(tq, 1H, J ) 4, 2), 1.55 (m, 1H), 3.2 (m, 1H), 3.25 (s, 3H), 3.3
(s, 3H), 4.8 (d, 1H, 3.4), 6.77 (AA′XX′, 2H, J ) 8.9), 6.82
(AA′XX′, 2H, J ) 8.5), 7.0 (AA′XX′, 2H, J ) 8.7), 7.13 (AA′XX′,
2H, J ) 8.5), 7.25-7.40 (m, 3H), 7.70 (AA′XX′, 2H, J ) 7.3);
13C NMR δ 10.5, 25.4, 54.8, 55.1, 57.8, 69.7, 114.4, 114.4, 115.0,
125.8, 126.0, 127.5, 127.7, 129.3, 139.3, 142.8, 148.4, 153.6,
160.2; HRMS (EI, M+) calcd for C25H26N2O2 386.1999, found
386.1994.
4-Eth yl-1,5-bis(4-m eth oxyph en yl)-3-ph en yl-4,5-dih ydr o-
1H-p yr a zole (14b). Prepared from lithium diisopropylamide
(4.65 mmol), pyrazoline 13b (1 g, 2.8 mmol), and ethyl iodide
(0.45 mL, 5.6 mmol) by the procedures used to prepare
pyrazoline 14a . The crude product was purified by flash
chromatography (hexanes/ether 3:2) to give 0.33 g (38%) of 14b
a yellow foam: 1H NMR (CDCl3, 400 MHz) δ 1.03 (t, 3H, J )
7.3), 1.69 (tq, 1H, J ) 4, 2), 1.85 (m, 1H), 3.37 (m, 1H), 3.73 (s,
3H), 3.75 (s, 3H), 4.91 (d, 1H, 3.4), 6.77 (AA′XX′, 2H, J ) 8.9),
6.82 (AA′XX′, 2H, J ) 8.5), 7.0 (AA′XX′, 2H, J ) 8.7), 7.13
(AA′XX′, 2H, J ) 8.5), 7.25-7.40 (m, 3H), 7.70 (AA′XX′, 2H, J
) 7.3); 13C NMR δ 10.5, 25.1, 55.2, 55.6, 57.3, 68.9, 113.9,
114.4, 114.5, 125.8, 126.8, 127.9, 128.5, 132.6, 134.1, 138.5,
148.5, 152.8, 158.8; MS (EI, 70 eV) m/ z 386.2 (M+).
4-Isop r op yl-3,5-b is(4-m e t h oxyp h e n yl)-1-p h e n yl-1H -
p yr a zole (6).13 Upon solid support cleavage and solvent
removal, the crude solid was recrystallized from 25% ethyl
acetate/hexane to afford 6 as small cubic crystals (30 mg, 11%
1
over three steps): mp 225-230 °C; H NMR (MeOD-d4, 400
MHz) δ 1.09 (d, 6H, J ) 7.0 Hz) 2.98 (septet, 1H, J ) 7.11
Hz), 6.76 (AA′XX′, 2H, J ) 9.1, 2.6), 6.88 (AA′XX′, 2H, J )
9.0, 2.6), 7.1 (AA′XX′, 2H, J ) 8.7, 2.5), 7.20-7.30 (m, 5H),
7.39 (AA′XX′, 2H, J ) 8.8, 2.40); 13C NMR (MeOD-d4, 100 MHz)
δ 22.5, 24.5, 114.5, 114.6, 121.6, 124.6, 125.0, 126.7, 128.2,
130.2, 131.8, 139.6, 141.3, 151.3, 157.2, 157.6; HRMS (EI, M+)
calcd for C24H22N2O2 370.1681, found 370.1674.
1,3-Bisp h en ylp r op a n e-1,3-d ion e (10). To a stirred solu-
tion of diketone 9 (2g, 8.9 mmol) in freshly distilled THF (50
mL), was added 8.9 mL (8.9 mmol) of tetrabutylammonium
fluoride (1 M in THF) and stirred for 30 min. The solution
was concentrated under reduced pressure and dissolved in 50
mL of CHCl3. Ethyl iodide (1.4 mL, 17.8 mmol) was added in
one portion and stirred at room temperature overnight. The
solution was concentrated and the crude was purified by flash
chromatgraphy (petroleum ether/CH2Cl2, 1:1) to give 920 mg
of 19 as a white solid in 41% yield: 1H NMR (CDCl3, 400 MHz)
δ 1.05 (t, 3H, J ) 7.4), 2.17 (quint, 2H, J ) 7.4), 5.12 (t, 1H,
J ) 6.5) 7.42-7.58 (m, 6H), 7.94-7.98 (m, 4H); 13C NMR δ
12.8, 22.9, 58.7, 128.5, 128.8, 133.4, 136.1, 196.1; MS (EI, 70
eV) m/ z 252.1 (M+).
4-E t h yl-1-(4-m et h oxyp h en yl)-3,5-b isp h en yl-1H -p yr a -
zole (11). The diketone 10 (300 mg, 1.2 mmol) was reacted
with 4-methoxyphenylhydrazine hydrochloride (830 mg, 4.7
mmol) according to the general procedure for pyrazole syn-
thesis. The crude product was purified by flash chromatogra-
phy (petroleum ether/CH2Cl2, 1:1) to afford 276 mg of 11 as a
yellow oil (65%): 1H NMR (CDCl3, 400 MHz) δ 1.04 (t, 3H, J
) 7.5), 2.67 (q, 2H, J ) 7.5), 3.77 (s, 1H), 6.77 (AA′XX′, 2H, J
) 1.0, 2.2), 7.19 (AA′XX′, 2H, J ) 9.1, 2.2), 7.23-7.48 (m, 9H),
7.78 (AA′XX′, 2H, J ) 8.2, 2.5); 13C NMR δ 15.6, 17.1, 55.4,
113.8, 120.4, 126.1, 127.5, 127.9, 128.1, 128.41, 128.45, 130.1,
130.9, 133.4, 134.1, 141.2, 150.4, 158.2; MS (EI, 70 eV) m/ z
354.2 (M+).
4-E t h yl-1-(4-h yd r oxyp h en yl)-3,5-b isp h en yl-1H -p yr a -
zole (7a ). A stirred CH2Cl2 solution of 11 (274 mg, 0.77 mmol)
was deprotected with BBr3 according to the general demeth-
ylation procedure. The crude was purified by flash chroma-
tography (CH2Cl2/acetone, 3:1) to give 84 mg of 7a as a white
1
solid (32%): mp 184-185 °C; H NMR (CD3OD, 400 MHz) δ
1.03 (t, 3H, J ) 7.5), 2.66 (q, 2H, J ) 7.5), 6.65 (AA′XX′, 2H J
) 8.8, 2.2), 7.09 (AA′XX′, 2H, J ) 9.0, 2,1), 7.2-7.5 (m, 8H),
7.77 (AA′XX′, 2H, J ) 8.4); 13C NMR δ 15.5, 17.0, 115.8, 120.2,
4-E t h yl-1,3-b is(4-m et h oxyp h en yl)-5-p h en yl-1H -p yr a -
zole (15a ). To a stirred solution of 14a (27.8 mg, 0.07 mmol)

4946 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26
Stauffer et al.
in 3 mL of benzene was added 78 mg (0.9 mmol) of MnO2. The
solution was heated to 100 °C with a Dean-Stark trap for 2
h, cooled to room temperature, filtered through Celite and
concentrated. The crude was chromatographed (EtOAc/hex-
anes, 1:10) to afford 28 mg of 15a (100%): 1H NMR (CDCl3,
400 MHz) δ 1.12 (t, 3H, J ) 7.5), 2.65 (q, 2H, J ) 7.5), 3.78 (s,
3H), 3.85 (s, 3H), 6.77 (AA′XX′, 2H, J ) 9.0, 2.0), 7.0 (AA′XX′,
2H, J ) 8.8, 2.0), 7.2 (AA′XX′, 2H, J ) 9.0, 2.0), 7.22-7.40 (m,
5H), 7.71 (d, 2H, J ) 8.6, 2.0); 13C NMR δ15.5, 17.1, 55.2, 55.4,
113.8, 113.8, 120.1, 126.1, 126.8, 128.1, 128.4, 129.1, 130.1,
131.1, 133.5, 141.1, 150.2, 158.2, 159.1; HRMS (EI, M+) calcd
for C25H24N2O2 385.1912, found 385.1916.
4-E t h yl-1,5-b is(4-m et h oxyp h en yl)-3-p h en yl-1H -p yr a -
zole (15b). A mixture of 147 mg (0.37 mmol) of pyrazoline
14b and 127 mg (0.56 mmol) of dichlorodicyanoquinone in 10
mL of benzene was heated to reflux for 5 h. The mixture was
cooled to room temperature and filtered through a plug of
Celite with diethyl ether. The filtrate was concentrated in
vacuo and the residue was chromatographed (hexanes/ethyl
acetate, 4:6) to give 130 mg (90%) of 15b as a white solid: 1H
NMR (CDCl3, 400 MHz) δ 1.04 (t, 3H, J ) 7.5), 2.65 (q, 2H, J
) 7.5), 3.78 (s, 3H), 3.83 (s, 3H), 6.79 (AA′XX′, 2H, J ) 8.9,
2.0), 6.89 (AA′XX′, 2H, J ) 8.5, 2.0), 7.16 (AA′XX′, 2H, J )
8.3, 1.8), 7.20 (AA′XX′, 2H, J ) 8.7, 2.1), 7.3-7.5 (m, 3H), 7.77
(AA′XX′, 2H, J ) 8.8, 1.8); 13C NMR δ 15.6, 17.12, 55.2, 55.4,
113.7, 113.9, 120.2, 123.2, 126.1, 127.4, 127.9, 128.4, 131.3,
133.5, 134.27, 141.07, 150.3, 158.1, 159.3; MS (EI, 70 eV) m/ z
384.2 (M+).
4-E t h yl-1,3-b is(4-h yd r oxyp h en yl)-5-p h en yl-1H -p yr a -
zole (8a ). A stirred CH2Cl2 solution of 15a (26 mg, 0.068
mmol) was deprotected with BBr3 according to the general
demethylation procedure. The crude was purified by flash
chromatography (CH2Cl2/MeOH, 10:1) to give 17 mg of 8a as
a tan solid (70%): mp 225-227 °C; 1H NMR (CD3OD, 500
MHz) δ 0.8 (t, 3H, J ) 7.5), 2.5 (q, 2H, J ) 7.5), 6.58 (AA′XX′,
2H, J ) 9.0, 2.2), 6.78 (AA′XX′, 2H, J ) 8.5, 2.0), 6.94 (AA′XX′,
2H, J ) 8.9, 2.0), 7.1-7.3 (m, 5H), 7.39 (AA′XX′, 2H, J ) 8.9,
2.0); 13C NMR δ 15.8, 17.9, 116.3, 116.4, 121.0, 122.6, 128.2,
128.9, 129.3, 129.6, 132.6, 133.2, 135.3, 143.8, 151.7, 158.2,
159.0; HRMS (EI, M+) calcd for C23H20N2O2 356.1598, found
356.1603.
Refer en ces
(1) Gao, H.; Katzenellenbogen, J . A.; Garg, R.; Hansch, C. Compara-
tive QSAR Analysis of Estrogen Receptor Ligands. Chem. Rev.
1999, 99, 723-744.
(2) Grese, T. A.; Dodge, J . A. Selective Estrogen Receptor Modula-
tors (SERMs) [Review]. Curr. Pharm. Des. 1998, 4, 71-92.
(3) Anstead, G. M.; Carlson, K. E.; Katzenellenbogen, J . A. The
estradiol pharmacophore: ligand structure-estrogen receptor
binding affinity relationships and a model for the receptor
binding site. Steroids 1997, 62, 268-303.
(4) Brzozowski, A. M.; Pike, A. C.; Dauter, Z.; Hubbard, R. E.; Bonn,
T.; Engstro¨m, O.; O¨ hman, L.; Greene, G. L.; Gustafsson, J .-A.;
Carlquist, M. Molecular basis of agonism and antagonism in the
oestrogen receptor. Nature 1997, 389, 753-758.
(5) Ekena, K.; Weis, K. E.; Katzenellenbogen, J . A.; Katzenellen-
bogen, B. S. Different residues of the human estrogen receptor
are involved in the recognition of structurally diverse estrogens
and antiestrogens. J . Biol. Chem. 1997, 272, 5069-5075.
(6) Shiau, A. K.; Barstad, D.; Loria, P. M.; Cheng, L.; Kushner, P.
J .; Agard, D. A.; Greene, G. L. The structural basis of estrogen
receptor/coactivator recognition and the antagonism of this
interaction by tamoxifen. Cell 1998, 95, 927-937.
(7) Anstead, G. M.; Peterson, C. S.; Katzenellenbogen, J . A. Hy-
droxylated 2,3-diarylindenes: Synthesis, estrogen receptor bind-
ing affinity, and binding orientation considerations. J . Steroid
Biochem. 1989, 33, 877-887.
(8) Kuiper, G. G. J . M.; Enmark, E.; Pelto-Huikko, M.; Nilsson, S.;
Gustafsson, J . A° . Cloning of a novel receptor expressed in rat
prostate and ovary. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 5925-
5930.
(9) Mosselman, S.; Polman, J .; Dijkema, R. ERâ: identification and
characterization of a novel human estrogen receptor. FEBS Lett.
1996, 392, 49-53.
(10) Kuiper, G. G. J . M.; Carlsson, B.; Grandien, K.; Enmark, E.;
Ha¨ggblad, J .; Nilsson, S.; Gustafsson, J .-A° . Comparison of the
ligand binding specificity and transcript tissue distribution of
estrogen receptor R and â. Endocrinology 1997, 138, 863-870.
(11) Fink, B. E.; Mortensen, D. S.; Stauffer, S. R.; Aron, Z. D.;
Katzenellenbogen, J. A. Novel Structural Templates for Estrogen-
Receptor Ligands and Prospects for Combinatorial Synthesis of
Estrogens. Chem. Biol. 1999, 6, 205-219.
(12) Katzenellenbogen, J . A.; Katzenellenbogen, B. S.; Fink, B. S.;
Stauffer, S. R.; Mortensen, D. S.; Sattigeri, V. J .; Huang, Y.
Estrogen Receptor Ligands. Patent WO 00/19994, Apr 13, 2000.
(13) Stauffer, S. R.; Katzenellenbogen, J . A. Solid-Phase Synthesis
of Tetrasubstituted Pyrazoles, Novel Ligands for the Estrogen
Receptor. J . Comb. Chem. 2000, 2, 318-329.
(14) Sun, J .; Meyers, M. J .; Fink, B. E.; Rajendran, R.; Katzenellen-
bogen, J . A.; Katzenellenbogen, B. S. Novel Ligands that
Function as Selective Estrogens or Antiestrogens for Estrogen
Receptor-R or Estrogen Receptor-â. Endocrinology 1999, 140,
800-804.
(15) Huang, Y.; Katzenellenbogen, J . A. Regioselective synthesis of
1,3,5-triaryl-4-alkylpyrazoles: Novel ligands for the estrogen
receptor. Org Lett. 2000, 2, 2833-2836.
(16) Carlson, K. E.; Choi, I.; Gee, A.; Katzenellenbogen, B. S.;
Katzenellenbogen, J . A. Altered ligand binding properties and
enhanced stability of a constitutively active estrogen receptor:
Evidence that an open pocket conformation is required for ligand
interaction. Biochemistry 1997, 36, 14897-14905.
(17) Meyers, M. J .; Sun, J .; Carlson, K. E.; Katzenellenbogen, B. S.;
Katzenellenbogen, J . A. Estrogen receptor subtype-selective
ligands: Asymmetric synthesis and biological evaluation of cis-
and trans-5,11-dialkyl-5,6,11,12-tetrahydrochrysenes. J . Med.
Chem. 1999, 42, 2456-2468.
(18) We are grateful to a referee for noting this relationship.
(19) Grese, T. A.; Cho, S.; Finley, D. R.; Godfrey, A. G.; J ones, C. D.;
Lugar, C. W., III; Martin, M. J .; Matsumoto, K.; Pennington, L.
D.; Winter, M. A.; Adrian, M. D.; Cole, H. W.; Magee, D. E.;
Phillips, D. L.; Rowley, E. R.; Short, L. L.; Glasebrook, A. L.;
Bryant, H. U. Structure-activity relationships of selective
estrogen receptor modulators: Modifications at the 2-arylbenzo-
thiophene core of raloxifene. J . Med. Chem. 1997, 40, 146-167.
(20) Pons, M.; Michel, F.; Crastes de Paulet, A.; Gilbert, J .; Miquel,
J .-F.; Precgoux, G.; Hospital, M.; Ojasoo, T.; Raynaud, J .-P.
Influence of new hydroxylated triphenylethylene (TPE) deriva-
tives on estradiol binding to uterine cytosol. J . Steroid Biochem.
1984, 20, 137-145.
4-E t h yl-1,5-b is(4-h yd r oxyp h en yl)-3-p h en yl-1H -p yr a -
zole (8b). A stirred CH2Cl2 solution of 15b (130 mg, 0.34
mmol) was deprotected with BBr3 according to the general
demethylation procedure. The crude was purified by flash
chromatography (CH2Cl2/acetone, 3:1) to give 42 mg of 8b as
1
a white solid (35%): mp 225-226 °C; H NMR (CD3OD, 500
MHz) δ 0.96 (t, 3H, J ) 7.5), 2.62 (q, 2H, J ) 7.5), 6.71 (AA′XX′,
2H, J ) 9.1, 2.2), 6.77 (AA′XX′, 2H, J ) 8.5, 2.0), 7.06 (AA′XX′,
2H, J ) 8.9, 2.1), 7.07 (AA′XX′, 2H, J ) 9.1, 2.2), 7.35-7.47
(m, 3H), 7.67 (AA′XX′, 2H, J ) 8.9, 2.0); 13C NMR δ 15.8, 17.9,
116.3, 116.4, 121.0, 122.6, 128.2, 128.9, 129.3, 129.6, 132.6,
133.2, 135.3, 143.8, 151.7, 158.2, 159.0; HRMS (EI, M+) calcd
for C23H20N2O2 356.151939, found 356.152478 Anal. (C23
20N2O2‚0.75H2O) C, H, N.
-
H
Ack n ow led gm en t. We are grateful for support of
this research through grants from the U.S. Army Breast
Cancer Research Program (DAMD17-97-1-7076 to J.A.K.)
and the National Institutes of Health (PHS 5R37
DK15556 to J .A.K. and PHS 5R37 CA18119 to B.S.K.).
We thank Dr. Ying R. Huang for the preparation of some
compounds used in this study. NMR spectra were
obtained at the Varian Oxford Instrument Center for
Excellence in NMR Laboratory. Funding for this instru-
mentation was provided in part from the W. M. Keck
Foundation and the National Science Foundation (NSF
CHE 96-10502). Mass spectra were obtained on instru-
ments supported by grants from the National Institute
of General Medical Sciences (GM 27029), the National
Institute of Health (RR 01575), and the National Science
Foundation (PCM 8121494).
(21) Katzenellenbogen, J . A.; O’Malley, B. W.; Katzenellenbogen, B.
S. Tripartite steroid hormone receptor pharmacology: Interac-
tion with multiple effector sites as a basis for the cell- and
promoter-specific action of these hormones. Mol. Endocrinol.
1996, 10, 119-131.

Pyrazole Ligands: ERR-Selective Agonists
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 26 4947
(22) Kraichely, D. M.; Sun, J .; Katzenellenbogen, J . A.; Katzenellen-
bogen, B. S. Conformational Changes and Coactivator Recruit-
ment by Novel Ligands for Estrogen Receptor-R and Estrogen
Receptor-â: Correlations with Biological Character and Distinct
Differences among SRC Coactivator Family Members. Endo-
crinology 2000, 141, 3534.
(23) Stauffer, S. R.; Huang, Y.; Coletta, C. J .; Katzenellenbogen, J .
A. Estrogen Pyrazoles: Defining the Pyrazole Core Structure
and Orientation of Substituents in the Ligand Binding Pocket
of the Estrogen Receptor. Biorg. Med. Chem. 2000, in press.
(24) Stauffer, S. R.; Huang, Y.; Aron, Z. D.; Coletta, C. J .; Sun, J .;
Katzenellenbogen, B. S.; Katzenellenbogen, J . A. Triarlpyrazoles
with basic side chains: Development of pyrazole-based estrogen
receptor antagonists. Biorg. Med. Chem. 2000, in press.
(25) Pike, A. C.; Brzozowski, A. M.; Hubbard, R. E.; Bonn, T.;
Thorsell, A. G.; Engstrom, O.; Ljunggren, J .; Gustafsson, J .;
Carlquist, M. Structure of the ligand-binding domain of oestro-
gen receptor beta in the presence of a partial agonist and a full
antagonist. EMBO J . 1999, 18, 4608-4618.
(28) Wang, D.; Chen, D.; Haberman, J .; Li, C. J . Ruthenium-
Catalyzed Isomerization of Homoallylic Alcohols in Water.
Tetrahedron 1998, 54, 5129-5142.
(29) Horiuchi, Y.; Oshima, K.; Utimoto, K. Titanium Tetrachloride-
Induced Three-Component Coupling Reaction of R-Haloacyl-
silane, Allysilane, and Carbonyl Compound. J . Org. Chem. 1996,
61, 4483-4486.
(30) Narasaka, K.; Kusama, H. Friedel-Crafts Acylation of Arenes
Catalyzed by Bromopentacarbonylrhenium (I). Bull. Chem. Soc.
J pn. 1995, 68, 2379-2383.
(31) Williams, D.; Gorski, J . Equilibrium binding of estradiol by
uterine cell suspensions and whole uteri in vitro. Biochemistry
1974, 13, 5537-5542.
(32) McInerney, E. M.; Tsai, M. J .; O’Malley, B. W.; Katzenellen-
bogen, B. S. Analysis of estrogen receptor transcriptional
enhancement by a nuclear hormone receptor coactivator. Proc.
Natl. Acad. Sci. U.S.A. 1996, 93, 10069-10073.
(33) Katzenellenbogen, J . A.; J ohnson, H. J ., J r.; Myers, H. N.
Photoaffinity labels for estrogen binding proteins of rat uterus.
Biochemistry 1973, 12, 4085-4092.
(26) Still, W. C.; Kahn, M.; Mitra, A. Rapid chromatographic tech-
nique for preparative separations with moderate resolution. J .
Org. Chem. 1978, 43, 2923-2926.
(27) Hachiya, I.; Moriwaki, M.; Kobayashi, S. Hafnium (IV) Trifluoro-
methanesulfonate, An Efficient Catalyst for the Friedel-Crafts
Acylation and Alkylation Reactions. Bull. Chem. Soc. J pn. 1995,
68, 2053-2060.
J M000170M