Synthesis of anthranylaldoxime derivatives as estrogen receptor ligands and computational prediction of binding modes

Article abstract of DOI:10.1021/jm060560u

N-Me-anthranylaldoximes possess a hydrogen-bonded pseudocyclic A' ring in place of the typical phenolic A-ring that is characteristic of most estrogen receptor (ER) ligands. We have investigated the role played by substituents introduced into either one o

Full text of DOI:10.1021/jm060560u

J. Med. Chem. 2006, 49, 5001-5012
5001
Synthesis of Anthranylaldoxime Derivatives as Estrogen Receptor Ligands and Computational
Prediction of Binding Modes
Tiziano Tuccinardi,† Simone Bertini,† Adriano Martinelli,† Filippo Minutolo,*,† Gabriella Ortore,† Giorgio Placanica,†
Giovanni Prota,† Simona Rapposelli,† Kathryn E. Carlson,‡ John A. Katzenellenbogen,‡ and Marco Macchia†
Dipartimento di Scienze Farmaceutiche, UniVersita` di Pisa, Via Bonanno 6, 56126 Pisa, Italy, and
Department of Chemistry, UniVersity of Illinois, 600 S. Goodwin AVenue, Urbana, Illinois 61801
ReceiVed May 11, 2006
N-Me-anthranylaldoximes possess a hydrogen-bonded pseudocyclic A′ ring in place of the typical phenolic
A-ring that is characteristic of most estrogen receptor (ER) ligands. We have investigated the role played
by substituents introduced into either one or both of the peripheral 3- and 4-phenyl rings in modulating ER
binding affinity. An efficient synthetic strategy was employed for the preparation of differentially substituted
3- and 4-aryl derivatives that involved exploiting the different reactivity of bromo- versus chloro-aryl groups
in palladium-catalyzed cross-couplings. The binding data showed that ERR affinity could be improved by
a single p-OH group in the 4-phenyl ring, whereas the same substitution on the 3-phenyl ring caused a
dramatic reduction of ERâ affinity. The most ERR-selective compound was the one with two p-OH groups
on both phenyl substituents. To rationalize these results, ligand docking followed by molecular mechanics
Poisson-Boltzmann/surface area (MM-PBSA) studies were carried out. These analyses suggested a molecular
basis for the interaction of these compounds with the ERs and enabled the development of models able to
predict the mode of ligand binding.
Introduction
and/or by the different interactions that the ER-ligand complex
might have with the cellular coregulatory proteins or effector
components that vary from tissue to tissue and gene to gene.11,12
Our search for new ER ligands has initially led us to develop
3,4-diarylsalicylaldoximes, either unsubstituted on the phenyl
rings (1a)13 or containing para hydroxy groups on either one
(1b,c)14 or on both (1d)14 peripheral substituents. These
compounds were particularly interesting because they contained
an unprecedented bioisosteric replacement of the phenolic A
group, which is a common feature of typical steroidal and
nonsteroidal ER ligands.15 In salicylaldoximes 1a-d, there is
a six-membered pseudocycle A′ formed by the intramolecular
hydrogen bond between the phenolic OH and the adjacent oxime
nitrogen atom, which might be functioning as an effective mimic
of the phenolic A-ring.
In binding assays, unsubstituted salicylaldoxime 1a proved
to possess an interesting affinity for both receptor subtypes
(relative binding affinity (RBA) ) 1.13% with ERR and 1.71%
with ERâ; RBA (estradiol) ) 100%). The binding properties
were further improved with the introduction of a single p-OH
group, whose position resulted in only a slight shift in the ERR/
ERâ selectivity profile of 1b and 1c. As a matter of fact, when
the p-OH is introduced in the proximal (position 3) aryl ring
(1b), the affinity for the ERâ is increased (RBA ) 2.21%),
whereas when the same group is placed on the other (distal)
aryl substituent (1c), the binding preference for ERR increases
(RBA ) 2.59%).14 However, the simultaneous presence of two
p-OH groups on both phenyl substituents (1d) caused a dramatic
reduction in the binding affinities, especially in the case of
ERâ.14
The actions of estrogens in many organs and tissues (repro-
ductive, skeletal, cardiovascular, and central nervous) are
mediated by estrogen receptors (ERs), which are members of a
superfamily of nuclear receptors that function as ligand-activated
gene transcription factors.1 There are two subtypes of estrogen
receptors known as ERR and ERâ,2,3 and they have a very high
amino acid sequence identity in their DNA-binding domains
(DBDs) but a more modest (59%) homology in their ligand-
binding domains (LBDs).4 Despite the sequence divergence in
the LBD amino acid sequences, however, X-ray analyses of
receptor-ligand complexes have shown that the binding cavities
of ERR and ERâ are extremely similar. The differences are at
two amino acid residues (ERâ has a methionine (M336) in place
of the leucine at position 384 of ERR and an isoleucine (I373)
instead of the methionine at position 421 of ERR) and in the
pocket volume (ERâ being somewhat smaller than ERR).5,6 The
two receptor subtypes have quite different tissue distributions.
In fact, ERR is more enriched in certain reproductive organs,
such as the uterus and mammary glands, whereas ERâ is mostly
expressed in other tissues, such as the central nervous system,
colon, lung, prostate, and the urinary tract.7
Selective estrogen receptor modulators (SERMs)8 are a special
category of estrogens, mostly of a nonsteroidal nature, that show
an estradiol agonist-like action in some tissues but antagonize
estradiol in others. They are particularly attractive as therapeutic
agents because they are able to block estrogen action at those
sites where stimulation would be undesirable (i.e., breast and
uterus), whereas at the same time, they stimulate estrogen actions
in other tissues where they are desired, (i.e., the bone and
liver).9,10 The tissue-selective pharmacology of SERMs is
probably due both to their different actions on each ER subtype
In a parallel study, we also wanted to further investigate the
structural basis for the phenol mimicry of pseudocycle A′ to
see whether we could improve ER binding affinity by modifying
the stereoelectronic nature of this ring. Therefore, we first
synthesized anthranylaldoxime 2a, in which the oxygen atom
of salycilaldoximes was replaced by an aniline-type nitrogen,
and then, we also prepared anthranyl derivatives possessing
* Corresponding author. Tel: +39 050 2219557. Fax: +39 050 2219605.
E-mail: minutolo@farm.unipi.it.
† Universita` di Pisa.
‡ University of Illinois.
10.1021/jm060560u CCC: $33.50 © 2006 American Chemical Society
Published on Web 07/08/2006

5002 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16
Tuccinardi et al.
Chart 1. Estrogen Ligands Possessing Salicylaldoxime (1a-d)
and Anthranylaldoxime (2a-c) Structures
Scheme 1a
Chart 2. Newly Synthesized N-Methylanthranylaldoximes
(3-10)
small alkyl groups on the nitrogen atom (a methyl in 2b and an
ethyl in 2c) to see whether these groups would affect the binding
properties of this class of molecules.16 The N-unsubstituted
derivative 2a proved to possess better binding properties than
its salicylic analogue 1a, showing a 2-fold improvement in both
receptor subtypes. A further increase was observed with
N-methylated analogue 2b, which showed RBA values three
times higher than that of 1a on both receptor subtypes. On the
contrary, the N-ethylated derivative 2c showed a very significant
drop in binding properties compared to those of its N-methylated
analogue 2b, with a nearly 10-fold reduction on ERR and a
remarkable over a 100-fold reduction on ERâ. These results
could be explained by docking experiments which showed that
the N-methyl group of 2b is found to occupy a small
hydrophobic pocket whose boundaries are formed by three
residues (ERR: Leu346, Leu349, and Ala350). However, the
spatial dimension of this pocket is only able to comfortably host
the N-Me group of 2b so that it needs to rearrange to be able
to do the same with even the slightly bulkier ethyl moiety, as
in 2c; the resulting repulsive van der Waals interactions caused
the decreased affinity.16 Therefore, the best phenolic A-ring
substitute found with our initial investigations was the pseudo-
cycle A′ of N-Me anthranylaldoxime (2b).
On the basis of these considerations, we wanted to combine
the beneficial effects of peripheral substituents of the salicyl-
aldoxime series with the highest affinity pseudocycle A′
represented by the N-Me anthranylaldoxime (2b). Therefore,
in this article, we expanded the series of N-Me anthranyl-
aldoximes by introducing various substituents on the 3- and
4-phenyl rings of 2b. Heterodisubstituted compounds 3 and 4
are characterized by the presence of a para hydroxy group on
either one of the aryl substituents, in parallel to what was
previously designed in the salicylaldoxime series with com-
pounds 1c and 1b, respectively. In the homodisubstituted series
(compounds 5-10), we wanted to verify the effect on binding
affinity caused by the presence of polar (OH, OMe) or
hydrophobic (Me) substituents in meta and para positions of
both peripheral aryl rings.
a Key: (a) NH2NH2‚H2O, FeCl3‚6H2O, activated carbon, MeOH, ∆; (b)
(CF3CO)2O, K2CO3, acetone; (c) MeI, K2CO3, DMF; (d) K2CO3, H2O/
MeOH; (e) allyl bromide, K2CO3, acetonitrile, 80 °C; (f) BF3‚Et2O,
sulfolane, 120 °C; (g) t-BuOK, DMSO, RT; (h) 4-R1-C6H4B(OH)2,
Pd(PPh3)4, Na2CO3, toluene, EtOH, ∆; (i) 4-R2-C6H4B(OH)2, Pd2(dba)3,
Cy3P, Cs2CO3, dioxane, ∆; (j) OsO4, NaIO4, dioxane-H2O; (k) BBr3,
CH2Cl2, -78 f 0 °C; (l) NH2OH‚HCl, MeOH-H2O, 50 °C.
tational models of ERR and ERâ were developed. Previously
reported anthranylaldoximes 2a-c16 and newly synthesized
compounds 3-10 were subjected to an automated docking
molecular dynamics (MD) simulation protocol followed by
a
molecular mechanics Poisson-Boltzmann/surface area
(MM-PBSA) analysis,17 and quantitative relationships between
the experimental and calculated free energy of binding were
evaluated with the aim to develop a predictive model.
Results and Discussion
Synthetic Chemistry. The preparation of heterodisubstituted
compounds 3 and 4 (Scheme 1) included key steps we had
already developed,14 represented by a sequential double Pd-
catalyzed cross-coupling reaction on a bromo-chloro-disubsti-
tuted aryl precursor. This approach exploits the different
reactivities of aryl-bromides and aryl-chlorides toward boronic
acids under different reaction conditions, that is, Suzuki-type
conditions for the bromides18 and Fu-type conditions for the
chlorides,19 and it provides an efficient way to selectively
To investigate the molecular interactions that are important
in determining the affinity of these ligands toward the ERs and
also to clarify the spatial orientations that all of the anthranyl-
aldoxime series assume within the receptors, various compu-

Anthranylaldoximes as ER Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16 5003
Table 1. Relative Binding Affinitiesa of Compounds 2-10 for Estrogen
Receptors R and â
Scheme 2a
ligand
hERR
(100)
hERâ
(100)
estradiol
2bb
3
3.7 ( 1.1b
5.38 ( 1.37
1.17 ( 0.33
1.96 ( 0.54
0.14 ( 0.03
2.33 ( 0.20
0.65 ( 0.19
0.95 ( 0.07
3.91 ( 0.30
5.2 ( 1.2b
1.11 ( 0.01
0.72 ( 0.10
0.14 ( 0.04
0.05 ( 0.01
0.47 ( 0.14
0.14 ( 0.03
1.86 ( 0.26
2.13 ( 0.56
4
5
6
7
8
9
10
a Determined by a competitive radiometric binding assay with [3H]es-
tradiol; preparations of purified, full-length human ERR and ERâ (PanVera)
were used (see Experimental Section). The values are reported as the mean
( SD of three independent experiments; the Kd for estradiol for ERR is
0.2 nM and for ERâ is 0.5 nM. b See ref 16.
(E,Z)-2,3-Dichloro-N-methyl-6-(1-propenyl)aniline (23), ob-
tained as previously reported,16 underwent a double cross-
coupling reaction with the appropriate arylboronic acid, using
the catalytic system containing Pd2(dba)3 and tricyclohexyl-
phosphine, which efficiently promotes cross-coupling reactions
on aryl-chloride bonds. Under these conditions, diaryl-substi-
tuted products 24a-d were obtained. Anthranylaldehydes
25a-d were then obtained by oxidative cleavage of the double
bond of 14a-c, using sodium periodate in the presence of
catalytic amounts of osmium tetroxide. Hydroxy-substituted
aldehydes 24e,f were obtained by the demethylation of 25a,b,
respectively, with BBr3. The final condensation of aldehydes
25a-f with hydroxylamine hydrochloride in a 10:1 methanol-
water mixture at 50 °C yielded N-methylanthranylaldoximes
5-10.
In all cases (3-10), the (E)-form of the oxime was the only
diastereoisomer formed, presumably because the intramolecular
hydrogen bond, which can only form in the (E)-isomer,
contributes to the oxime stability. As seen before for similar
compounds,13,14,16 in these cases, the chemical shift value of
the oxime proton, which is always found below 8 ppm (8.33
for 7; 8.34 for 5; 8.35 for 6; 8.36 for 3, 4, 8, and 9; 8.37 for
10), confirmed the (E)-configuration of oximes 5-10.21
a Key: (a) 2 times: Pd2(dba)3, Cy3P, ArB(OH)2, Cs2CO3, dioxane, 80
°C; (b) OsO4, NaIO4, dioxane-H2O; (c) BBr3, CH2Cl2, -78 f 0 °C; (d)
NH2OH‚HCl, MeOH-H2O, 50 °C.
introduce different substituents in those positions, depending
on the boronic acid used. The complete synthesis of 3 and 4
was accomplished as shown in Scheme 1.
2-Bromo-3-chloronitrobenzene (11)20 was first reduced with
hydrazine hydrate in the presence of catalytic amounts of ferric
chloride and activated carbon to yield the corresponding aniline
12. Treatment with trifluoroacetic anhydride afforded trifluo-
roacetamide 13, which was methylated with MeI to give tertiary
amide 14. Alkaline hydrolysis afforded N-methylamine 15,
which underwent an N-allylation reaction with allyl bromide.
N-Allyl-N-methylaniline (16) was submitted to an aza-Claisen
rearrangement in the presence of boron trifluoride etherate,
which selectively gave ortho-allylated aniline 17. The alkaline
rearrangement of the terminal double bond to the internal
position afforded styrene derivative 18 as an E/Z diastereomeric
mixture.
Estrogen Receptor Binding Assays. The binding affinity
of oximes 3-10 for both ERR and ERâ were determined by a
radiometric competitive binding assay, using methods that have
been described elsewhere in detail.22,23 In Table 1, the RBA
values for the newly reported compounds, determined with
purified full-length human alpha (hERR) and beta (hERâ)
receptor subtypes, together with those previously obtained for
2b are reported.16 Binding affinity (RBA) values are reported
relative to that of estradiol (E2), which is set at 100%.
At this point, the bromo-chloro substituted intermediate 18
was first submitted to a classical Suzuki-type cross-coupling
reaction, using Pd(Ph3P)4 as the catalyst, to introduce the first
phenyl or p-methoxyphenyl group in place of the bromine atom,
affording compounds 19a (R1 ) H) or 19b (R1 ) OCH3),
respectively. This step did not affect the chlorine atom of 18,
which was then substituted in the second cross-coupling reaction
with p-methoxyphenyl or unsubstituted phenyl boronic acid,
using the more reactive catalytic system comprised of Pd2(dba)3,
together with tricyclohexylphosphine as the catalyst ligand.
Under these conditions, it was possible to introduce a second
aryl substituent (R2 ) OCH3 for 20a; R2 ) H for 20b) in place
of the aryl-chloride groups of 19a and 19b.
Salicylaldehydes 21a,b were then obtained by oxidative
cleavage of the double bond of 20a,b, using sodium periodate
in the presence of catalytic amounts of osmium tetroxide. The
transformation of the methoxy groups of 21a,b into hydroxyls
was achieved by a BBr3 demethylation, and the resulting
hydroxyaryl-substituted salicylaldehydes 22a,b were then con-
densed with hydroxylamine hydrochloride in a methanol-water
mixture at 50 °C to give final oximes 3 and 4.
The results shown in Table 1 inidicate that the anthranyl-
aldoxime bearing a single para hydroxy group in the distal aryl
ring, that is, compound 3, experiences an increase in its binding
affinity for ERR (RBA ) 5.38%) compared with unsubstituted
2b, whereas its affinity for ERâ is significantly reduced (RBA
) 1.11%). The affinity for ERR found with 3 is the highest
ever found with our pseudocycle-phenol mimics deriva-
tives,13,14,16 corresponding to a Kd of 3.7 nM. A similar increased
ERR/ERâ selectivity ratio caused by the introduction of a p-OH
into the distal phenyl substituent had already been observed in
the salicylaldoxime series (1c),14 although to a lesser extent.
However, when the same para hydroxy group is placed on the
other aryl ring (proximal) as in 4, the affinity decreases with
both receptor subtypes. This time, no parallelism with the
salicylaldoxime series was found because in that series the same
The synthesis of homodisubstituted oximes 5-10 is shown
in Scheme 2.

5004 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16
Tuccinardi et al.
structural modification (p-OH on the proximal ring, as in 1b)
had caused a significant shift in affinity toward ERâ selectivity.
Among the homodisubstituted compounds, the 4-OH disub-
stituted oxime 5 shows a dramatic drop in its affinity for ERâ,
although still preserving a good binding ability on ERR, with
an R/â selectivity Kd ratio of 36. Oxime 7, possessing two para-
methoxy substituents, confirms the general trend observed for
its nonmethylated counterpart 5, with a comparable affinity for
ERR but a lower degree of R/â selectivity due to a slight raise
of ERâ affinity. The binding properties of oximes 6 and 8, which
have the oxygenated substituents (OH and OMe, respectively)
in the meta positions of both peripheral aromatic groups, are
low on both receptor subtypes. These results indicate that the
introduction of two oxygenated substituents, such as OMe and
OH, generally result in a reduction in ERâ affinity. ERR affinity
is only slightly affected when these substituents are placed in
the two para positions, whereas their introduction into the meta
positions is not tolerated by either of the two receptor subtypes.
Compound 9, having two para-methyl groups, has an affinity
for ERâ, which is twice as high as for ERR. In contrast,
analogue 10, where the methyl groups are placed on the meta
positions, has a preference for ERR (RBA ) 3.91%) rather than
ERâ. An interesting profile is found with the introduction of
this small lipophilic substituent (Me) in the para- (10) and meta
positions (9) of the two aryl rings. In fact, in one case (para-
methyl, 9), the RBA ERâ/ERR ratio is about 2, whereas when
the methyl is shifted to the meta position (10), the RBA alpha/
beta ratio is completely reversed (ERR/ERâ ≈2).
Docking Analysis. A series of docking simulations have been
carried out to better rationalize the experimental binding
affinities of the whole series of anthranylaldoximes, including
the previously reported (2a-c)16 and the newly synthesized (3-
10) compounds. To build the receptor models, X-ray structures
of ERR (1G5024) and ERâ (1X7J25) were complexed with
estradiol, and the complexes were subjected to 900 ps of
molecular dynamics (MD) simulations (see Experimental Sec-
tion for details). As shown in Figure 1A, after about 100 ps of
MD, both systems reached an equilibrium because the total
energy for the last 800 ps remained constant. Analyzing the
root-mean-square deviation (RMSD) from the X-ray structures
of all of the heavy atoms of the proteins, we observed that after
an initial increase during the last 800 ps, the RMSD remained
approximately constant, around the range of 0.9-1.0 Å for ERR
and 0.7-0.8 Å for ERâ (Figure 1B).
Figure 1. Analysis of the MD simulation of estradiol complexed with
ERR (black) and ERâ (red). (A) Total energy of the system (kcal/mol)
vs time is reported; (B) RMSD between all of the heavy atoms of the
receptors and the two X-ray starting structures is reported.
Figure 2 shows the minimized average structures of the
binding site of the two receptors complexed with estradiol. As
also confirmed by the H-bond analysis of the MD simulations
(Table 2), estradiol presented a very similar interaction scheme
into both receptors: the 17-hydroxy group interacted with H524-
(475), and the 3-hydroxy substituent formed a H bond with
E353(305) (residue numbers are given for both ERR and ERâ,
the latter in parentheses). Furthermore, as already suggested by
Brzozowski and Pike,26,27 a water molecule is also present at
the core of a H-bond network system between E353(305), R394-
(346), and the 3-hydroxy substituent of estradiol.
Using the two minimized average structures reported
above as receptors, the compounds showed in Table 1 were
docked using an automated docking procedure by means of
AUTODOCK 3.028 (see Experimental Section).
A conformational search of these ligands revealed the
presence of an intramolecular H bond with the formation of a
pseudocycle involving the oxime nitrogen and the ortho disposed
N-H group. The quantum mechanics optimization of 2b
suggested that this interaction has a good strength, about 4 kcal/
Figure 2. Hydrogen-bonding network of the estradiol in the ERR
(panel A) and ERâ (panel B).
mol (see Experimental Section for details), and consequently,
this H bond was maintained during interaction in the binding
site. For this reason, during the AUTODOCK protocol, we

Anthranylaldoximes as ER Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16 5005
Table 2. Hydrogen Bonds Analysis of the
E353(305)-R394(346)-Water-Estradiol-H524(475) System during the
MD Simulation into Both Receptor Subtypes
Table 3. Preferred Binding Orientation (pref.) of the Ligands into ERR
and ERâa
ERR
ERâ
distance
(Å)
compd
pref.
∆EA-B
pref.
∆EA-B
donor
acceptorh
acceptor
% occupied
2a
2b
2c
3
4
5
6
7
8
9
A
A
A
A
A
A
A
A
A
A
A
--
-0.39
--
-2.00
-1.44
--
--
-0.99
--
--
-0.66
A
B
B
A
A
A
B
B
B
A
B
-0.26
+0.37
+0.48
-0.44
-0.37
-0.47
+0.15
+0.49
+0.91
-0.85
+0.31
ERR
E353@OE2
H524@ND1 EST@OH1
WAT@O
E353@OE1
EST@O3
WAT@O
EST@HO3
EST@O3
2.55
3.03
2.98
2.70
2.95
3.24
100
97
94.95
91.20
65.75
63.10
EST@O17
R394@HH12 R394@NH1
WAT@O
WAT@O
R394@HH22 R394@NH2
WAT@H1
WAT@H2
ERâ
EST@O3
EST@O17
E305@OE2
H475@ND1 EST@OH1
WAT@O
E305@OE1
EST@O3
WAT@O
EST@HO3
2.60
3.08
2.99
2.79
3.1
100
94.20
95.20
84.25
59.21
70.80
10
R346@HH12 R346@NH1
a The free energy difference between orientation A and B, calculated by
means of the AUTODOCK scoring function is also reported (∆EA-B). When
AUTODOCK found only one of the two orientations, the free energy
difference is indicated by “--”.
WAT@H1
WAT@H2
WAT@O
WAT@O
R346@HH22 R346@NH2
3.18
As reported by Yang et al. for oxime derivatives,30 there could
be another possible orientation for aldoxime 3 (180° horizontal),
in which the pseudocycle of the ligands interacted with H524-
(475), whereas the two aromatic rings were oriented toward the
E353(305)-R394(346)-H2O system. This orientation was not
taken into account because it was showed only by compound 9
inside the ERR and compound 2b inside the ERâ. Furthermore,
this alternative orientation was found to be energetically less
stable than A and B orientations in both cases.
As shown in Table 3, in ERR, all of the ligands preferred
orientation A, whereas in ERâ, six ligands preferred binding
orientation B. However, for compound 2b in ERR and com-
pounds 2a-c, 3-7, and 10 in ERâ, the difference between the
estimated free energy of binding for the two orientations was
lower than 0.5 kcal/mol. Thus, although the AUTODOCK
results could be considered reliable for indicating the preferred
orientation for ERR, this was not the case for ERâ.
Six compounds out of 11 (2a, 2c, 5, 6, 8, and 9) did not
show the B orientation inside the ERR, and this fact was
somewhat surprising because the ERR binding site was very
similar to that of ERâ. However, the docking analysis for these
compounds revealed that orientation A inside the ERR binding
site was very stable because the corresponding cluster was highly
populated (more than 50% of the total runs).
As a further analysis, we also docked these six compounds
inside the ERR by means of 250 runs instead of 100 runs.
Nevertheless, even in this case, the B orientation was not found.
Therefore, to make a more precise analysis of the ligand-
receptor interaction and to find more reliable results about the
preferred orientation of the ligand into ERR and ERâ, all of
the ligand-receptor complexes were used as starting structures
for 500 ps of MD simulation.
Successively, the MD trajectories were further analyzed
through the MM-PBSA method,17 which has shown to ac-
curately estimate the ligand-receptor energy interaction.31-33
This approach averages contributions of gas-phase energies,
solvation free energies, and solute entropies calculated for
snapshots of the complex molecule as well as the unbound
components extracted from MD trajectories according to the
procedure fully described in the Experimental Section. In
particular, for each ligand, we took into account as starting points
both A and B orientations found by AUTODOCK in both ERs
for a total of 44 calculations; in the case of the ligands for which
AUTODOCK did not find any B orientation into the ERR, these
structures were built on the basis of the B orientation found for
unsubstituted compound 2b.
Figure 3. Docking of compound 2b into ERR. (A) Binding mode A;
(B) binding mode B.
blocked the torsions involved in this intramolecular bond to
prevent the loss of this interaction.
The docking studies revealed that as already hypothesized
for other ER ligands29 these compounds could interact with ERs
through two different binding orientations as shown in Figure
3; these are designated as binding orientations A and B. In both
cases, the hydrogen bonded pseudocycle present in 2b mimicked
the phenolic ring of estradiol, interacting with the E353(305)-
R394(346)-H2O system. The same behavior was found for all
of the other docked N-Me-anthranylaldoximes.
Concerning the other parts of the ligands in binding orienta-
tion A, the distal aromatic ring was directed toward H524(475)
and interacted with L/M384(336), M388(340), M/I421(373),
I424(376), and L525(476), whereas the proximal aromatic ring
interacted with L346(298), F404(356), and M343(295). How-
ever, in binding orientation B, the distal aryl substituent
interacted with L346(298), F404(356), and M343(295), whereas
the proximal ring interacted with M388(340), L/M384(336),
I424(376), and L525(476).

5006 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16
Tuccinardi et al.
In the Table reported in the Supporting Information, the MM-
PBSA results calculated for each ligand (A and B conformation)
into both receptors are reported. The analysis of the energies
differences between the two orientations for each ligand
highlighted that all of the compounds showed the A orientation
as the preferred one in both receptors; only the 5-ERâ complex
showed a better interaction energy when compound 5 was
positioned into the B orientation. Interestingly, this compound
was the one with the highest ERR/ERâ selectivity.
After the MD simulations, the ligand interactions observed
in the starting structures (obtained by means of AUTODOCK)
were generally maintained. In both orientations, the oxime
hydroxy group interacted with the E353(305)-R394(346)-H2O
system, whereas the two aromatic groups interacted with M343-
(295), L346(298), L/M384(336), M388(340), F404(356), M/I421-
(373), I424(376), and L525(476), as described above.
All of the compounds tested showed an affinity selectivity
for ERR, and as already mentioned above, the ERR and ERâ
binding sites are composed of identical residues except for the
presence of L384 and M421 in ERR, which are substituted by
M336 and I373 in ERâ. Moreover, the ERâ binding cavity is
smaller than that of ERR, and this reduction is primarily due to
the replacement of leucine at position 336 in ERR with the
slightly bulkier methionine in ERâ.
In our receptor models, the compounds showed a lipophilic
interaction with the methyl group of M421 in ERR and with
the side chain of I373 in ERâ; these interactions seemed to be
stronger in ERR than in ERâ. As a matter of fact, for all of the
tested ligands, the distance between the aromatic ring and the
methyl group of M421 in ERR was about 3.4 Å; however, in
ERâ, the distance between I373 and the same aromatic ring
was always greater than 4 Å.
As for residue L/M384(336), it was far away from all of the
ligands in both receptor subtypes (the distance was greater than
4 Å), and thus, the nature of the residue at this site appears to
only slightly affect the ligand-receptor interaction.
In conclusion, it appears that the interaction of these ligands
with the non conserved residue M421 in the ERR could be one
of the reasons for their ERR selectivity. As examples of high-
and low-affinity ligands, a binding interaction analysis of
compounds 3 and 6 has been reported below.
Figure 4A shows the superposition of the binding sites of
ERR and ERâ complexed with compound 3, with the two
nonconserved residues L/M384(336) and M/I421(373) repre-
sented as sticks. The compound formed H bonds with the E353-
(305)-R394(346)-H2O system through the oxime OH into both
receptors; furthermore, the para-OH substituent formed a H
bond with H524(475), simulating the interaction of the 17-
hydroxy group of estradiol.
The H-bond interaction between the para-OH substituent and
H475 in ERâ seems to be in contrast to the lower affinity of
this compound, compared with that of the unsubstituted 2b. The
qualitative analysis is also confirmed by the quantitative one
because the MM-PBSA procedure for ERâ calculated a greater
affinity for compound 3 with respect to that for compound 2b
(Table 4). Thus, the computational procedures herein used,
which are based on automated docking, the MD simulation, and
the MM-PBSA analysis, seem to be unable to properly rational-
ize the affinity of compound 3 inside the ERâ.
The substitution of the meta position of each one of the two
aromatic rings with a hydroxy group, as in 6, resulted in a
decrease in affinity with both receptors; Figure 4B shows the
superposition of the binding sites of ERR and ERâ complexed
with 6, which was the lowest affinity compound with both
Figure 4. Superposition of the binding sites of ERR and ERâ
complexed with compound 3 (panel A) and compound 6 (panel B).
The ligand-ERR complex was colored by atom, whereas ligand-ERâ
is in blue. The two nonconserved residues L/M384(336) and M/I421-
(373) (colored red in the ERâ) are shown in stick representation.
Table 4. Energy Contributions to the Free Energy of Binding (∆Gcal
Expressed as kcal/mol) for the Best Ligand Orientation of Each
Compounda
,
ERR
compd
∆EMM
∆GPBSA
-T∆S
∆Gcal
∆Gexp
∆∆G
2a
2b
2c
3
4
5
6
7
8
9
-61.10
-58.35
-62.18
-65.85
-61.62
-70.53
-67.49
-69.84
-67.61
-63.85
-65.77
32.33
28.22
29.22
35.35
33.40
37.31
39.08
36.80
32.13
28.22
30.76
18.50
19.39
24.27
19.26
19.37
22.74
21.5
22.44
22.85
24.91
24.26
-10.27
-10.74
-8.69
-11.11
-11.42
-10.19
-11.64
-10.73
-11.04
-9.48
-11.14
-10.39
-10.61
-11.45
0.84
0.68
1.5
-11.24
-8.85
0.40
1.88
0.56
2.57
0.54
-2.24
-0.11
0.70
-10.48
-6.91
-10.60
-12.63
-10.72
-10.75
10
ERâ
compd
∆EMM
∆GPBSA
-T∆S
∆Gcal
∆Gexp
∆∆G
2a
2b
2c
3
4
5
6
7
8
9
-58.72
-57.55
-57.84
-64.40
-58.89
-83.21
-67.63
-67.79
-66.07
-60.06
-58.67
30.50
26.86
26.37
33.00
30.46
53.66
40.97
33.15
30.07
26.60
27.51
17.5
-10.72
-10.75
-8.32
-10.61
-10.98
-8.20
-10.06
-9.81
-8.85
-8.18
-9.56
-8.85
-10.37
-10.45
-0.11
0.23
-0.12
-1.15
0.30
-0.12
1.48
-0.58
-4.62
-0.71
-0.40
19.94
23.15
20.19
18.92
20.58
19.96
24.50
22.53
22.38
20.31
-11.21
-9.51
-8.97
-6.70
-10.14
-13.47
-11.08
-10.85
10
a The energy difference (∆∆G) between the calculated (∆Gcal) and the
experimental free energy of binding (∆Gexp) is also reported.
receptors among all of the compounds tested. The Figure
highlights that both meta-hydroxy substituents were unable to
interact with any residue, confirming the binding affinity results.

Anthranylaldoximes as ER Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16 5007
the preparation of compounds bearing different 3- and 4-aryl
substituents (3 and 4) is particularly attractive because it exploits
the differential reactivity of bromo- and chloro-aryl function-
alities toward Pd-catalyzed cross-coupling reactions. The assays
of binding affinities revealed that compound 3 has the highest
affinity for ERR of our oxime series, with an RBA value of
5.38% (Kd of 3.7 nM) compared to that of E2 (Kd of 0.2 nM),
together with an ERR/ERâ selectivity Kd ratio of 12. Moreover,
compound 5, possessing two p-hydroxyphenyl substituents in
the 3 and 4 positions, was the most ERR selective compound
with an ERR/ERâ selectivity Kd-ratio of 36. Docking studies
confirmed that in the whole series of anthranylaldoximes the
oxime hydroxy group formed H bonds with the E353(305)-
R394(346)-H2O system, mimicking the phenolic OH group of
estradiol, and all of the compounds tested interacted in the ERR
with the nonconserved residue M421, consistent with their ERR/
ERâ selectivity. Furthermore, we reported calculations of
binding free energies between these ligands and both ERs using
the recently developed MM-PBSA method. This approach
proved to be attractive for rationalizing at a quantitative manner
the interaction of the anthranyaldoxime derivatives with ERs.
Experimental Section
Chemistry. The melting points were determined on a Kofler
hot-stage apparatus and are uncorrected. NMR spectra were obtained
with a Varian Gemini 200 MHz spectrometer. Chemical shifts (δ)
are reported in parts per million downfield from tetramethylsilane
and referenced from solvent references. Electron impact (EI, 70
eV) mass spectra were obtained on an HP-5988A mass spectrom-
eter. Where indicated, the elemental compositions of the compounds
agreed to within (0.4% of the calculated value. Chromatographic
separations were performed on silica gel columns by flash (Kieselgel
40, 0.040-0.063 mm; Merck) or gravity column (Kieselgel 60,
0.063-0.200 mm; Merck) chromatography. The reactions were
followed by thin-layer chromatography (TLC) on Merck aluminum
silica gel (60 F254) sheets that were visualized under a UV lamp.
Evaporation was performed in vacuo (rotating evaporator). Sodium
sulfate was always used as the drying agent.
2-Bromo-3-chloroaniline (12). A solution of 2-bromo-3-chlo-
ronitrobenzene (11)20 (3.94 g, 16.7 mmol) in methanol (170 mL)
was treated with 0.013 g (0.048 mmol) of FeCl3‚6H2O and 0.884
g of activated charcoal. The mixture was heated to 80 °C; hydrazine
monohydrate was then added dropwise (13 mL, 260 mmol), and
the resulting suspension was stirred at the same temperature
overnight. The reaction mixture was then cooled to RT, filtered,
and concentrated under vacuum. The crude residue was dissolved
in chloroform, washed with brine, dried, and concentrated to give
12 as a yellow solid (2.96, 86% yield); 1H NMR (CDCl3) δ (ppm):
4.26 (bs, 2H), 6.65 (dd, 1H, J ) 8.1, 1.5 Hz), 6.84 (dd, 1H, J )
8.1, 1.5 Hz), 7.02 (t, 1H, J ) 8.0 Hz). MS m/z 205 (M+).
Figure 5. Experimental vs predicted free energy of binding for the
eight ligands with ERR (upper) and ERâ (lower). The values are
expressed as kcal/mol.
To verify whether the MM-PBSA method could be used to
extrapolate a quantitative correlation between the calculated free
energy and the experimental affinity of the ligands for ERs, we
used the AMBER module nmode (see Experimental Section for
details) to estimate the entropic contribution for the best 11
ligand-receptor complexes. As shown in Figure 5, a plot of
the experimental versus the calculated free energy of binding
(Table 4), considering compound 8 as an outlier, shows that
there is a good quadratic correlation for both receptor subtypes
(R2 ) 0.83 for ERR and R2 ) 0.81 for ERâ).
With regard to compound 8, the MM-PBSA method predicted
a high affinity into both receptors, and as shown in the Table
reported in the Supporting Information, these affinity values
were mainly due to the van der Waals contribution, which was
the highest one for both receptors. Although it was not possible
to exclude the failure of the MM-PBSA approach for this outlier,
it should be considered that the meta-methoxy group on both
aromatic rings determined an increase of the hindrance of the
ligand, and therefore, the low affinity of compound 8 could be
due to a higher difficulty in reaching the binding site of both
receptors.
2-Bromo-3-chloro-N-trifluoroacetylaniline (13). A solution of
12 (2.96 g, 14.4 mmol) in acetone (35 mL) was treated under
nitrogen with potassium carbonate (2.97 g, 21.5 mmol) and 4.0
mL of trifluoroacetic anhydride (28 mmol), and it was stirred at
RT overnight. An additional amount of trifluoroacetic anhydride
(2 mL, 14 mmol) was added, and stirring was continued for 6 h,
and then the mixture was diluted with water and extracted with
Et2O. The organic phase was dried and concentrated to give 13
(4.33 g, 99% yield) as a yellow solid that was used in the next step
1
without further purification; H NMR (CDCl3) δ (ppm): 7.34-
7.39 (m, 2H), 8.21-8.30 (m, 1H), 8.57 (bs, 1H). MS m/z 301 (M+).
2-Bromo-3-chloro-N-methyl-N-trifluoroacetylaniline (14). A
solution of 13 (4.33 g, 14.3 mmol) in DMF (36 mL) was treated
under nitrogen with potassium carbonate (3.97 g, 28.7 mmol) and
1.8 mL of iodomethane (28 mmol), and it was stirred at RT for 3
h. Two additional amounts of iodomethane (each 0.5 mL, 8 mmol)
were added, and stirring was continued for 3 h, and then the mixture
was diluted with water and extracted with ethyl acetate. The organic
Conclusions
In conclusion, we expanded the series of N-Me anthranyl-
aldoximes by introducing various substituents on the 3- and
4-phenyl rings of 2b, combining the beneficial effects of the
peripheral substituents of the salicylaldoxime series (1a-d), with
the highest affinity pseudocycle A′ represented by the N-Me
anthranylaldoxime nucleus. The synthetic strategy followed for

5008 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16
Tuccinardi et al.
phase was dried and concentrated to give 14 (4.07 g, 90% yield)
1H), 7.02 (d, 1H, J ) 8.4 Hz), 7.23-7.30 (m, 2H), 7.37-7.53 (m,
4H). MS m/z 257 (M+).
as an orange solid that was used in the next step without further
1
purification; H NMR (CDCl3) δ (ppm): 3.31 (s, 3H), 7.21-7.40
(E/Z)-3-Chloro-2-(4-methoxyphenyl)-N-methyl-6-(prop-1-
enyl)aniline (19b). Compound 18 (910 mg, 3.51 mmol) was
submitted to the same procedure described above for the preparation
of 19a by reacting with 4-methoxyphenylboronic acid (0.64 g, 4.2
mmol). Purification of the crude product by flash chromatography
(n-hexane/dichloromethane 9:1) yielded pure 19b (0.77 g, 76%
(m, 2H), 7.54 (dd, 1H, J ) 7.9, 1.6 Hz). MS m/z 315 (M+).
2-Bromo-3-chloro-N-methylaniline (15). A solution of 14 (4.07
g, 12.9 mmol) in methanol (135 mL) and water (55 mL) was treated
with potassium carbonate (11.2 g, 81.2 mmol), and it was stirred
at RT overnight. The reaction mixture was then diluted with water
and extracted with Et2O. The organic phase was dried and
concentrated to give 15 (2.81 g, 99% yield) as a yellowish solid
that was used in the next step without further purification; 1H NMR
(CDCl3) δ (ppm): 2.90 (d, 1H, J ) 5.1 Hz), 4.57 (bs, 1H), 6.49
(dd, 1H, J ) 8.2, 1.3 Hz), 6.80 (dd, 1H, J ) 8.0, 1.4 Hz), 7.13 (t,
1H, J ) 8.1 Hz). MS m/z 219 (M+).
1
yield) as a light yellow solid; H NMR (CDCl3, 9:1 E/Z mixture,
asterisk denotes minor isomer peaks) δ (ppm): 1.84* (dd, 3H, J )
7.0, 1.8 Hz), 1.92 (dd, 3H, J ) 6.6, 1.6 Hz), 2.60 (s, 3H), 3.87 (s,
3H), 6.13 (dq, 1H, J ) 15.7, 6.6 Hz), 6.55 (dq, 1H, J ) 15.9, 1.6
Hz), 7.01 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.3 Hz), 7.03 (d,
1H, J ) 8.2 Hz), 7.18 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.4
Hz), 7.26 (d, 1H, J ) 8.4 Hz); MS m/z 287 (M+).
N-Allyl-2-bromo-3-chloro-N-methylaniline (16). A solution of
15 (2.81 g, 12.8 mmol) in acetonitrile (13 mL) was treated with
potassium carbonate (3.2 g, 23 mmol), and it was heated at 50 °C.
A solution of allyl bromide (6.5 mL, 77 mmol) in acetonitrile (16
mL) was then added dropwise through a dropping funnel, and after
the addition was completed, the reaction mixture was heated to
100 °C for 4 h. TLC analysis still showed the presence of starting
material; therefore, more allyl bromide (6.5 mL) in acetonitrile (16
mL) was added, and the mixture was kept under stirring at 100 °C
overnight. The reaction mixture was then cooled to RT, and the
resulting suspension was filtered under vacuum. The residue was
extracted with Et2O, and the combined filtrates were concentrated
under vacuum to give 16 (3.33 g, 99% yield) as a yellow oil that
(E/Z)-3-(4-Methoxyphenyl)-N-methyl-2-phenyl-6-(prop-1-
enyl)aniline (20a). A solution of 19a (305 mg, 1.19 mmol) in
dioxane (3 mL) was treated with cesium carbonate (658 mg, 2.02
mmol), 4-methoxyphenylboronic acid (270 mg, 1.80 mmol), Pd2-
(dba)3 (35 mg, 0.038 mmol), and 0.15 mL of a 20% solution of
tricyclohexylphosphine (0.10 mmol) in toluene, and the resulting
mixture was heated to 80 °C in a sealed vial for 24 h. The reaction
mixture was then cooled to RT, diluted with Et2O, and filtered
through a Celite pad. The organic filtrate was concentrated under
vacuum, and the crude product was purified by flash chromatog-
raphy (n-hexane/dichloromethane 7:3) to yield pure 20a (331 mg,
1
84% yield) as an oil; H NMR (CDCl3) δ (ppm): 1.94 (dd, 3H, J
) 6.5, 1.6 Hz), 2.62 (s, 3H), 3.72 (s, 3H), 6.20 (dq, 1H, J ) 15.8,
6.6 Hz), 6.64 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.4 Hz),
1
was used in the next step without further purification; H NMR
(CDCl3) δ (ppm): 2.73 (s, 3H), 3.60 (d, 2H, J ) 6.0 Hz), 5.16-
5.23 (m, 1H), 5.29 (dq, 1H, J ) 17.0, 1.6 Hz), 5.92 (ddt, 1H, J )
17.2, 10.2, 6.0 Hz), 6.94-7.01 (m, 1H), 7.12-7.19 (m, 2H). MS
m/z 259 (M+).
6.67-6.71 (m, 1H), 6.94 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′
)
2.4 Hz), 6.97-7.00 (m, 1H), 7.06-7.28 (m, 5H), 7.42 (d, 1H, J )
7.6 Hz). MS m/z 329 (M+).
(E/Z)-2-(4-Methoxyphenyl)-N-methyl-3-phenyl-6-(prop-1-
enyl)aniline (20b). Compound 19b (0.77 g, 2.7 mmol) was
submitted to the same procedure described above for the preparation
of 20a, by reacting with phenylboronic acid (0.77 g, 4.0 mmol).
Purification of the crude product by flash chromatography (n-
hexane/dichloromethane 1:1) yielded pure 20b (0.82 g, 93% yield)
6-Allyl-2-bromo-3-chloro-N-methylaniline (17). A solution of
16 (3.33 g, 12.8 mmol) in sulfolane (4 mL) was treated under
nitrogen with BF3‚Et2O (5.7 mL, 45 mmol), and it was heated at
120 °C under stirring overnight. After cooling to RT, the mixture
was diluted with water, neutralized with a 1 N aqueous NaOH
solution, and extracted with with Et2O. The organic phase was dried
and concentrated under vacuum to give a crude product, which was
purified by flash chromatography (n-hexane/dichloromethane 9:1),
1
as a yellow solid; H NMR (CDCl3) δ (ppm): 1.95 (dd, 3H, J )
6.6, 1.6 Hz), 2.64 (s, 3H), 3.78 (s, 3H), 6.21 (dq, 1H, J ) 15.7, 6.4
Hz), 6.64-6.73 (m, 1H), 6.80 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′
) 2.4 Hz), 6.98-7.14 (m, 8H), 7.42 (d, 1H, J ) 7.9 Hz). MS m/z
329 (M+).
1
yielding pure 17 (2.95 g, 89% yield) as a yellow oil; H NMR
(CDCl3) δ (ppm): 2.82 (s, 3H), 3.46 (dt, 2H, J ) 6.2, 1.6 Hz),
5.04 (dq, 1H, J ) 16.8, 1.8 Hz), 5.14 (dq, 1H, J ) 10.1, 1.6 Hz),
5.96 (ddt, 1H, J ) 16.8, 10.2, 6.4 Hz), 7.02 (d, 1H, J ) 8.2 Hz),
7.07 (d, 1H, J ) 8.2 Hz). MS m/z 259 (M+).
4-(4-Methoxyphenyl)-N-methyl-3-phenylanthranylaldehyde
(21a). A solution of 20a (315 mg, 0.958 mmol) in dioxane (7 mL)
was treated dropwise with water (4 mL), then with sodium periodate
(1.0 g) and 0.05 mL of a 2.5% solution of osmium tetroxide in
tert-butyl alcohol (0.05 mmol). The mixture was stirred at RT for
7 h, and then most of the solvent was removed under vacuum, and
the residue was diluted with water and repeatedly extracted with
chloroform. The combined organic phase was washed with aqueous
Na2S2O3 and brine and then dried and concentrated under vacuum.
The resulting crude residue was purified by flash chromatography
(n-hexane/ethyl acetate 9:1) to yield pure 21a (88 mg, 29% yield)
as a yellow solid: mp ) 120 °C; 1H NMR (CDCl3) δ (ppm): 2.26
(E/Z)-2-Bromo-3-chloro-N-methyl-6-(prop-1-enyl)aniline (18).
A solution of 17 (2.95 g, 11.3 mmol) in DMSO (30 mL) was treated
under nitrogen with t-BuOK (3.18 g, 28.4 mmol) and stirred at RT
for 1 h. The reaction mixture was then quenched with saturated
aqueous NH4Cl and extracted with Et2O. The organic phase was
washed with brine, dried, and concentrated under vacuum, and the
crude product was purified by flash chromatography (n-hexane/
dichloromethane 9:1) to yield pure 18 (1.82 g, 62% yield) as an
oil; 1H NMR (CDCl3, 9:1 E/Z mixture, asterisk denotes minor
isomer peaks) δ (ppm): 1.78* (dd, 3H, J ) 7.0, 1.8 Hz), 1.91 (dd,
3H, J ) 6.6, 1.8 Hz), 2.81 (s, 3H), 2.84* (s, 3H), 6.11 (dq, 1H, J
) 15.8, 6.6 Hz), 6.49-6.58 (m, 1H), 7.00 (d, 1H, J ) 8.2 Hz),
7.20 (d, 1H, J ) 8.4 Hz). MS m/z 259 (M+).
(s, 3H), 3.74 (s, 3H), 6.66 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′
)
2.5 Hz), 6.76 (d, 1H, J ) 8.0 Hz), 6.88 (AA′XX′, 2H, JAX ) 8.8
Hz, JAA′/XX′ ) 2.5 Hz), 7.06-7.22 (m, 5H), 7.49 (d, 1H, J ) 7.9
Hz), 8.25 (bs, 1H), 9.88 (s, 1H). MS m/z 317 (M+).
(E/Z)-3-Chloro-N-methyl-2-phenyl-6-(prop-1-enyl)aniline (19a).
A solution of 18 (453 mg, 1.75 mmol) in toluene (4.5 mL), ethanol
(4.5 mL), and 2 M aqueous Na2CO3 (4.5 mL) was treated under
nitrogen with phenylboronic acid (464 mg, 3.81 mmol) and Pd-
(PPh3)4 (0.088 mmol), and the mixture was heated to reflux for 20
h. The reaction mixture was then cooled to RT, diluted with water,
and extracted with Et2O. The organic phase was washed with brine,
dried, and concentrated under vacuum, and the crude product was
purified by flash chromatography (n-hexane/dichloromethane 9:1
f 8:2) to yield pure 19a (180 mg, 40% yield) as a yellow solid:
mp 62 °C; 1H NMR (CDCl3) δ (ppm): 1.90 (dd, 3H, J ) 6.5, 1.7
Hz), 2.59 (s, 3H), 6.14 (dq, 1H, J ) 15.8, 6.4 Hz), 6.50-6.59 (m,
3-(4-Methoxyphenyl)-N-methyl-4-phenylanthranylaldehyde
(21b). Compound 20b (0.21 g, 0.64 mmol) was submitted to the
same procedure described above for the preparation of 21a.
Purification of the crude product by flash chromatography (n-
hexane/ethyl acetate 9:1) yielded pure 21b (75 mg, 37% yield) as
1
a yellow solid; H NMR (CDCl3) δ (ppm): 2.31 (s, 3H), 3.76 (s,
3H), 6.71 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.5 Hz), 6.77 (d,
1H, J ) 7.9 Hz), 6.95-7.02 (m, 4H), 7.11-7.16 (m, 3H), 7.49 (d,
1H, J ) 8.1 Hz), 9.89 (s, 1H). MS m/z 317 (M+).
4-(4-Hydroxyphenyl)-N-methyl-3-phenylanthranylaldehyde
(22a). A solution of 21a (137 mg, 0.430 mmol) in dichloromethane
(9 mL) was cooled under nitrogen to -78 °C and treated dropwise

Anthranylaldoximes as ER Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16 5009
with a 1 M solution of BBr3 in dichloromethane (1.1 mL, 1.1 mmol),
and the resulting solution was stirred at the same temperature for
5 min and at RT for 3 h. The mixture was then diluted with water
and extracted with ethyl acetate. The organic phase was dried and
concentrated. The crude product was purified by flash chromatog-
raphy (n-hexane/ethyl acetate 7:3) to yield pure 22a (79 mg, 61%
Hz), 2.66 (s, 3H), 3.58 (s, 3H), 3.66 (s, 3H), 6.20 (dq, 1H, J )
15.6, 6.6 Hz), 6.60-6.86 (m, 6H), 7.0-7.24 (m, 4H), 7.44 (d, 1H,
J ) 8.1 Hz). MS m/z 359 (M+).
(E/Z)-2,3-Bis(4-methylphenyl)-N-methyl-6-(prop-1-enyl)-
aniline (24c). Eluent: n-hexane/dichloromethane 7:3; yield 62%;
1
off-white solid: mp 39-41 °C; H NMR (CDCl3) δ (ppm): 2.02
1
yield) as a yellow solid: mp 105 °C; H NMR (CDCl3) δ (ppm):
(dd, 3H, J ) 6.6, 1.6 Hz), 2.26 (s, 3H), 2.29 (s, 3H), 2.35 (s, 3H),
6.28-6.39 (m, 1H), 6.70-6.73 (m, 1H), 6.87-7.05 (m, 8H), 7.70
(d, 1H, J ) 8.1 Hz), 7.59 (d, 1H, J ) 8.1 Hz). MS m/z 327 (M+).
(E/Z)-2,3-Bis(3-methylphenyl)-N-methyl-6-(prop-1-enyl)-
aniline (24d). Eluent: n-hexane/dichloromethane 7:3; yield 35%;
yellow oil; 1H NMR (CDCl3) δ (ppm): 1.92 (dd, 3H, J ) 6.5, 1.6
Hz), 2.18 (s, 3H), 2.41 (s, 3H), 2.94 (s, 3H), 6.08 (dq, 1H, J )
15.4, 6.4 Hz), 6.42 (dq, 1H, J ) 15.4, 1.6 Hz), 6.80-7.41 (m, 10H).
MS m/z 327 (M+).
2.27 (s, 3H), 6.59 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.1 Hz),
6.77 (d, 1H, J ) 8.0 Hz), 6.84 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′
) 2.1 Hz), 7.07-7.20 (m, 5H), 7.49 (d, 1H, J ) 7.9 Hz), 9.88 (s,
1H). MS m/z 303 (M+).
3-(4-Hydroxyphenyl)-N-methyl-4-phenylanthranylaldehyde
(22b). Compound 21b (0.060 g, 0.19 mmol) was submitted to the
same procedure described above for the preparation of 22a.
Purification of the crude product by flash chromatography (n-
hexane/ethyl acetate 8:2) yielded pure 22b (31 mg, 53% yield) as
a yellow solid; 1H NMR (CDCl3) δ (ppm): 2.33 (s, 3H), 4.80 (bs,
1H), 6.65 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.4 Hz), 6.78 (d,
1H, J ) 7.9 Hz), 6.92-7.00 (m, 4H), 7.11-7.16 (m, 3H), 7.50 (d,
1H, J ) 8.1 Hz), 9.89 (s, 1H). MS m/z 303 (M+).
Synthesis of 2,3-Diaryl-N-methylanthranylaldehydes (25a-
d). General Procedure. A solution of the appropriate olefin
precursor (24a-d) (1.25 mmol) in dioxane (15 mL) was treated
dropwise with water (7 mL), then with sodium periodate (1.2 g),
and 0.13 mL of a 2.5% solution of osmium tetroxide in tert-butyl
alcohol. The mixture was stirred at RT for 5 h, then most of the
solvent was removed under vacuum, and the residue was diluted
with water and repeatedly extracted with chloroform. The combined
organic phase was washed with aqueous Na2S2O3 and brine, and
then dried and concentrated under vacuum. The resulting crude
residue was purified by flash chromatography, eluting with the
indicated eluent to obtain compounds 25a-d.
2,3-Bis(4-methoxyphenyl)-N-methylanthranylaldehyde (25a).
Eluent: n-hexane/EtOAc 8:2; yield 34%; off-white solid: mp 105
°C; 1H NMR (CDCl3) δ (ppm): 2.33 (s, 3H), 3.75 (s, 3H), 3.78 (s,
3H), 6.69 (AA′XX′, 2H, JAX ) 9.0 Hz, JAA′/XX′ ) 2.3 Hz), 6.75
(AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.4 Hz), 6.85 (d, 1H, J )
7.9 Hz), 6.91 (AA′XX′, 2H, JAX ) 9.0 Hz, JAA′/XX′ ) 2.4 Hz),
7.03 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.5 Hz), 7.51 (d, 1H,
J ) 7.9 Hz), 9.89 (s, 1H). MS m/z 347 (M+).
4-(4-Hydroxyphenyl)-N-methyl-3-phenylanthranylaldoxime
(3). A solution of 22a (25 mg, 0.082 mmol) in methanol (2 mL)
was treated with hydroxylamine hydrochloride (12 mg, 0.17 mmol),
and the mixture was heated to 50 °C for 1 h. After being cooled to
RT, the solvent was removed under vacuum, and the crude product
was purified by column chromatography (n-hexane/ethyl acetate
7:3) to yield pure 3 (18 mg, 70% yield) as an off-white solid: mp
1
42 °C; H NMR (CDCl3) δ (ppm): 2.39 (s, 3H), 6.58 (AA′XX′,
2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.5 Hz), 6.85 (AA′XX′, 2H, JAX
)
8.8 Hz, JAA′/XX′ ) 2.5 Hz), 6.90 (d, 1H, J ) 8.0 Hz), 7.09-7.24
(m, 5H), 7.37 (d, 1H, J ) 7.9 Hz), 8.36 (s, 1H). MS m/z 318 (M+).
Anal. (C20H18N2O2) C, H, N.
3-(4-Hydroxyphenyl)-N-methyl-4-phenylanthranylaldoxime
(4). Compound 22b (21 mg, 0.068 mmol) was submitted to the
same procedure described above for the preparation of 3. Purifica-
tion of the crude product by column chromatography (n-hexane/
ethyl acetate 7:3) yielded pure 4 (18 mg, 82% yield) as a white
2,3-Bis(3-methoxyphenyl)-N-methylanthranylaldehyde (25b).
Eluent: n-hexane/EtOAc 8:2; yield 32%; oil; 1H NMR (CDCl3) δ
(ppm): 2.49 (s, 3H), 3.59 (s, 3H), 3.72 (s, 3H), 6.61-8.04 (m,
10H), 9.98 (s, 1H). MS m/z 347 (M+).
1
solid: mp 172 °C; H NMR (CDCl3) δ (ppm): 2.42 (s, 3H), 6.68
(AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.3 Hz), 6.89 (d, 1H, J )
7.9 Hz), 6.97 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.3 Hz),
7.00-7.02 (m, 2H), 7.12-7.15 (m, 3H), 7.36 (d, 1H, J ) 7.9 Hz),
8.36 (s, 1H). MS m/z: 318 (M+, 20), 301 (M+ -OH, 100), 300
(M+ -H2O, 29), 286 (M+ -NHOH, 21). Anal. (C20H18N2O2) C,
H, N.
2,3-Bis(4-tolyl)-N-methylanthranylaldehyde (25c). Eluent: n-
1
hexane/Et2O 8:2; yield 38%; yellow solid: mp 110-112 °C; H
NMR (CDCl3) δ (ppm): 2.28 (s, 3H), 2.32 (s, 3H), 2.50 (s, 3H),
6.87-7.34 (m, 9H), 7.63 (d, 1H, J ) 7.9 Hz), 7.98 (d, 1H, J ) 8.0
Hz), 9.90 (s, 1H). MS m/z 315 (M+).
Synthesis of (E/Z)-2,3-Diaryl-N-methyl-6-(prop-1-enyl)-
anilines (24a-d). General Procedure. A solution of 2316 (2 mmol)
in dioxane (3 mL) was treated under nitrogen with cesium carbonate
(4 mmol), the appropriate arylboronic acid (2.8 mmol), Pd2(dba)3
(0.060 mmol), and a 20% solution of tricyclohexylphosphine (0.14
mmol) in toluene. The resulting mixture was heated to 80 °C in a
sealed vial for 16 h. The reaction mixture was then cooled to RT,
diluted with Et2O, and filtered through a Celite pad. The solvent
was removed under vacuum, and the resulting crude mixture was
submitted again to the same treatment described above. The crude
product derived from the second step was purified by column
chromatography, eluting with the indicated eluent to obtain
compounds 24a-d as E/Z mixtures.
2,3-Bis(3-tolyl)-N-methylanthranylaldehyde (25d). Eluent: n-
hexane/ EtOAc 9:1; yield 42%; oil; H NMR (CDCl3) δ (ppm):
2.19 (s, 3H), 2.22 (s, 3H), 2.30 (s, 3H), 6.72-6.87 (m, 3H), 6.76
(d, 1H, J ) 8.0 Hz), 6.94-7.03 (m, 5H), 7.48 (d, 1H, J ) 7.9 Hz),
9.87 (s, 1H). MS m/z 315 (M+).
Synthesis of Hydroxylated 2,3-Diaryl-N-methylanthranylal-
dehydes (25e,f). General Procedure. A solution of the appropriate
methyl-ether precursor (25a,b) (0.27 mmol) in anhydrous dichlo-
romethane (9 mL) was cooled under nitrogen to -78 °C and treated
dropwise with a 1 M solution of BBr3 in dichloromethane (1.2 mL,
1.2 mmol), and the resulting solution was stirred at the same
temperature for 5 min and at RT for 2 h. The mixture was then
diluted with water and extracted with ethyl acetate. The organic
phase was dried and concentrated. The resulting crude residue was
purified by flash chromatography, eluting with the indicated eluent
to obtain compounds 25e and f.
1
(E/Z)-2,3-Bis(4-methoxyphenyl)-N-methyl-6-(prop-1-enyl)-
aniline (24a). Eluent: n-hexane/dichloromethane 1:1; yield 67%;
off-white solid: mp 126-130 °C; 1H NMR (CDCl3, 9:1 E/Z
mixture, asterisk denotes minor isomer peaks) δ (ppm): 1.95 (dd,
3H, J ) 6.6, 1.6 Hz), 2.62* (s, 3H), 2.64 (s, 3H), 3.74 (s, 3H),
3.76* (s, 3H), 3.79 (s, 3H), 3.85* (s, 3H), 6.20 (dq, 1H, J ) 15.5,
6.6 Hz), 6.67 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.4 Hz),
2,3-Bis(4-hydroxyphenyl)-N-methylanthranylaldehyde (25e).
Eluent: n-hexane/EtOAc 1:1; yield 53%; white solid: mp 95 °C;
1H NMR (acetone-d6) δ (ppm): 2.33 (d, 3H, J ) 5.3 Hz), 6.63
(AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.4 Hz), 6.69-6.77 (m,
3H), 6.86 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.5 Hz), 6.93
(AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.5 Hz), 7.58 (d, 1H, J )
8.1 Hz), 8.27 (s, 1H), 8.32 (s, 1H), 9.91 (s, 1H). MS m/z 319 (M+).
2,3-Bis(3-hydroxyphenyl)-N-methylanthranylaldehyde (25f).
Eluent: n-hexane/EtOAc 7:3; yield 41%; light yellow solid: mp
6.73-6.78 (m, 1H), 6.82 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′
)
2.4 Hz), 6.95 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.4 Hz),
7.01 (AA′XX′, 2H, JAX ) 8.7 Hz, JAA′/XX′ ) 2.4 Hz), 7.41 (d, 1H,
J ) 7.9 Hz). MS m/z 359 (M+).
(E/Z)-2,3-Bis(3-methoxyphenyl)-N-methyl-6-(prop-1-enyl)-
aniline (24b). Eluent: n-hexane/dichloromethane 1:1; yield 38%;
yellow oil; 1H NMR (CDCl3) δ (ppm): 1.95 (dd, 3H, J ) 6.6, 1.6
1
95 °C; H NMR (acetone-d6) δ (ppm): 2.35 (d, 3H, J ) 4.8 Hz),
6.46-6.70 (m, 4H), 6.72 (d, 1H, J ) 7.9 Hz), 6.92-7.07 (m, 4H),

5010 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16
Tuccinardi et al.
Table 5. Energy Difference between the Two Optimized Conformations
of Compound 2b
7.60 (d, 1H, J ) 8.0 Hz), 8.34 (s, 1H), 8.36 (s, 1H), 9.92 (s, 1H).
MS m/z 319 (M+).
Synthesis of 2,3-Diaryl-N-methylanthranylaldoximes (5-10).
General Procedure. A solution of the appropriate aldehyde
precursor (25a-f) (0.17 mmol) in methanol (5 mL) was treated
with hydroxylamine hydrochloride (24 mg, 0.35 mmol), and the
mixture was heated to 50 °C for 1 h. After being cooled to RT, the
solvent was removed under vacuum, and the resulting crude residue
was purified by flash chromatography, eluting with the indicated
eluent to obtain compounds 5-10.
2,3-Bis(4-hydroxyphenyl)-N-methylanthranylaldoxime (5). Elu-
ent: n-hexane/EtOAc 1:1; yield 38%; white solid: mp 98-100
1
°C; H NMR (acetone-d6) δ (ppm): 2.34 (s, 3H), 6.61 (AA′XX′,
affinities are expressed as relative binding affinity (RBA) values,
where the RBA of estradiol is 100%; under these conditions, the
Kd of estradiol for ERR is ca. 0.2 nM, and for ERâ 0.5 nM. The
determination of these RBA values is reproducible in separate
experiments with a CV of 0.3, and the values shown represent the
average ( range or SD of two, three, or more separate determina-
tions, respectively.
2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.5 Hz), 6.72 (AA′XX′, 2H, JAX
8.7 Hz, JAA′/XX′ ) 2.5 Hz), 6.80 (d, 1H, J ) 8.1 Hz), 6.84 (AA′XX′,
2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.4 Hz), 6.93 (AA′XX′, 2H, JAX
)
)
8.6 Hz, JAA′/XX′ ) 2.4 Hz), 7.37 (d, 1H, J ) 8.1 Hz), 8.19 (bs, 1H),
8.28 (bs, 1H), 8.34 (s, 1H), 10.21 (bs, 1H). MS m/z 334 (M+).
Anal. (C20H18N2O3) C, H, N.
2,3-Bis(3-hydroxyphenyl)-N-methylanthranylaldoxime (6). Elu-
ent: n-hexane/EtOAc 1:1; yield 43%; glass; 1H NMR (acetone-d6)
δ (ppm): 2.36 (s, 3H), 6.46-6.69 (m, 4H), 6.78 (d, 1H, J ) 7.9
Hz), 6.90-7.09 (m, 4H), 7.39 (d, 1H, J ) 8.0 Hz), 8.17 (s, 1H),
8.23 (s, 1H), 8.35 (s, 1H), 10.27 (s, 1H). MS m/z 334 (M+). Anal.
(C20H18N2O3) C, H, N.
Computational Details. Published coordinates for ERR (1G5024)
and ERâ (1X7J25) were obtained from the Protein Data Bank.36 In
ERâ, the missed loop M410-N421 was constructed through the loop
optimization method of the Modeller program, using 1QKN5 as
the template.
For ERR and ERâ, the 1G50 and 1X7J reference numbering,
was used, respectively; when the amino acid was referred to both
receptors, the ERâ numbering was indicated in parentheses.
Analysis of the Compound’s Geometry. A conformational
search revealed the presence of two low energetic conformations
(Table 5), and conformation a showed an intramolecular H bond
and the formation of a pseudocycle.
To measure the different stabilities of the two conformations and
the strength of the intramolecular H bond, both conformations were
optimized using the B3LYP/6-31G+* method. The energy differ-
ence between the two optimized systems was about 4 kcal/mol,
and this value gave us an idea of the strength of the intramolecular
H bond and the better energetic stability of conformation a.
Automated Docking Procedure. The ligands were submitted
to a conformational search of 1000 steps with an energy window
for saving the structure of 2 kcal/mol by means of the MACRO-
MODEL program.37 The algorithm used was the Monte Carlo
method with MMFFs as the force field and a distance-dependent
dielectric constant of 1.0. The first conformation that showed the
intramolecular H bond was then minimized using the conjugated
gradient method until a convergence value of 0.05 kcal/Å‚mol, using
the same force field and dielectric constant used for the confor-
mational search.
Automated docking was carried out by means of the program
AUTODOCK 3.0;28 AUTODOCK TOOLS38 was used to identify
the torsion angles in the ligands, add the solvent model, and assign
the Kollman and the Gasteiger partial atomic charges to the protein
and the ligands, respectively. The regions of interest used by
AUTODOCK were defined by considering estradiol into both
receptors as the central group; in particular, a grid of 46, 44, and
44 points in the x, y, and z directions was constructed centered on
the center of mass of this compound. A grid spacing of 0.375 Å
and a distance-dependent function of the dielectric constant were
used for the energetic map calculations.
Using the Lamarckian Genetic Algorithm, the docked compounds
were subjected to 100 runs of the AUTODOCK search in which
the default values of the other parameters were used. Cluster
analysis was performed on the results using an RMS tolerance of
1.0 Å.
2,3-Bis(4-methoxyphenyl)-N-methylanthranylaldoxime (7).
Eluent: n-hexane/EtOAc 7:3; yield 70%; off-white solid: mp 175-
177 °C; H NMR (CDCl3) δ (ppm): 2.48 (s, 3H), 3.75 (s, 3H),
1
3.80 (s, 3H), 6.69 (AA′XX′, 2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.4 Hz),
6.82 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.2 Hz), 6.92 (AA′XX′,
2H, JAX ) 8.8 Hz, JAA′/XX′ ) 2.4 Hz), 7.06 (d, 1H, J ) 7.9 Hz),
7.08 (AA′XX′, 2H, JAX ) 8.6 Hz, JAA′/XX′ ) 2.1 Hz), 7.37 (d, 1H,
J ) 8.0 Hz), 8.33 (s, 1H). MS m/z 362 (M+). Anal. (C22H22N2O3)
C, H, N.
2,3-Bis(3-methoxyphenyl)-N-methylanthranylaldoxime (8).
Eluent: n-hexane/EtOAc 7:3; yield 41%; light yellow solid: mp
45-48 °C; 1H NMR (CDCl3) δ (ppm): 2.48 (s, 3H), 3.58 (s, 3H),
3.66 (s, 3H), 6.52-6.78 (m, 4H), 6.98 (d, 1H, J ) 7.9 Hz), 7.03-
7.20 (m, 4H), 7.39 (d, 1H, J ) 7.9 Hz), 8.36 (s, 1H). MS m/z 362
(M+). Anal. (C22H22N2O3) C, H, N.
2,3-Bis(4-tolyl)-N-methylanthranylaldoxime (9). Eluent: n-
hexane/EtOAc 8:2; yield 52%; light yellow glass; 1H NMR (CDCl3)
δ (ppm): 2.26 (s, 3H), 2.31 (s, 3H), 2.43 (s, 3H), 6.87-7.10 (m,
9H), 7.40 (d, 1H, J ) 8.1 Hz), 8.36 (s, 1H). MS m/z 330 (M+).
Anal. (C22H22N2O) C, H, N.
2,3-Bis(3-tolyl)-N-methylanthranylaldoxime (10). Eluent: n-
hexane/EtOAc 9:1; yield 68%; light yellow glass; 1H NMR (CDCl3)
δ (ppm): 2.19 (s, 3H), 2.24 (s, 3H), 2.44 (s, 3H), 6.74-6.78 (m,
1H), 6.84-6.99 (m, 7H), 7.03-7.13 (m, 1H), 7.39 (d, 1H, J ) 8.0
Hz), 8.37 (s, 1H). MS m/z 330 (M+). Anal. (C22H22N2O) C, H, N.
Biological Methods. Purified human full-length ERR and ERâ
were obtained from PanVera (Madison, WI). Cell culture media
were purchased from Gibco BRL (Grand Island, NY). Calf serum
was obtained from Hyclone Laboratories, Inc. (Logan, UT), and
fetal calf serum was purchased from Atlanta Biologicals (Atlanta,
GA). The luciferase assay system was from Promega (Madison,
WI). The expression vectors for human ERR (pCMV5-hERR) and
human ERâ (pCMV5-ERâ) were constructed previously as de-
scribed.34,35 The estrogen responsive reporter plasmid was (ERE)2-
pS2-Luc, was constructed by inserting the (ERE)2-pS2 fragment
from (ERE)2-pS2-CAT into the MluI/BglII sites of the pGL3-Basic
vector (Promega, Madison, WI). The plasmid pCMVâ (Clontech,
Palo Alto, CA), which contains the â-galactosidase gene, was used
as an internal control for transfection efficiency.
MD Simulations. All simulations were performed using AMBER
8.0.39 MD simulations were carried out using the parm94 force field
at 300 K. An explicit solvent model TIP3P water was used, and
the complexes were solvated with a 10 Å water cap. Sodium ions
were added as counterions to neutralize the system. Prior to MD
simulations, two steps of minimization were carried out; in the first
stage, we kept the protein fixed with a constraint of 500 kcal/mol,
and we minimized the positions of the water molecules; then in
Estrogen Receptor Binding Assays. Relative binding affinities
were determined by competitive radiometric binding assays using
10 nM [3H]E2 as a tracer, using previously described methods.22,23
The source of ERs was purified full-length human ERR and ERâ
purchased from Pan Vera (Madison, WI). Incubations were done
at 0 °C for 18-24 h, and hydroxylapatite was used to absorb the
purified receptor-ligand complexes (human ERs).23 The binding

Anthranylaldoximes as ER Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16 5011
the second stage, we minimized the entire system, applying a
constraint of 20 kcal/mol on the R carbon. The two minimization
stages consisted of 5000 steps in which the first 1000 were steepest
descent (SD) and the last 4000 conjugate gradient (CG). Molecular
dynamics trajectories were run using the minimized structure as a
starting input. The time step of the simulations was 2.0 fs with a
cutoff of 12 Å for the nonbonded interaction, and SHAKE was
employed to keep all bonds involving hydrogen atoms rigid.
Constant volume was carried out for 100 ps, during which the
temperature was raised from 0 to 300 K (using the Langevin
dynamics method); then 500 ps (800 ps for the ERR and Erâ
complexed with estradiol) of constant-pressure MD were carried
out at 300 K. All of the R carbons were blocked with a harmonic
force constant of 10 kcal/mol‚Å. The final structure of the
complexes was obtained as the average of the last 400 ps (700 ps
for the ErR and Erâ complexed with estradiol) of MD minimized
with the CG method until a convergence of 0.05 kcal/mol‚Å.
For the ligands, the force fields parameters were taken from the
general Amber force field (GAFF), whereas the atomic partial
charges were derived by semiempirical AM1 geometry optimization
and subsequent single point Hartree-Fock (HF)/6-31G* calculation
of the electrostatic potential, to which the charges were fitted using
the RESP procedure.40
To verify whether using these parameters the ligands were able
to maintain the intramolecular H bond and consequently the
pseudocycle, the simulation protocol applied to the receptor-ligand
complex as described above was applied to compound 2b immersed
in an explicit solvent model TIP3P water with a 20 Å water cap.
The minimized average of the last 400 ps of MD showed the
presence of the intramolecular H bond and the pseudocycle;
furthermore, the superimposition of this structure with the one
optimized with the QM method (conformation a of Table 5) showed
a value of 0.40 of RMSD (evaluated on all the atoms).
(ERR and ERâ). This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(1) Katzenellenbogen, J. A.; Katzenellenbogen, B. S. Nuclear hormone
receptors: ligand-activated regulators of transcription and diverse
cell responses. Chem. Biol. 1996, 3, 529-536.
(2) 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.
(3) Kuiper, G. G. J. M.; Carlsson, B.; Grandien, K.; Enmark, E.;
Haggblad, J.; Nilsson, S.; Gustafsson, J.-A. Comparison of the ligand
binding specificity and transcript tissue distribution of estrogen
receptors alpha and beta. Endocrinology 1997, 138, 863-870.
(4) Mosselman, S.; Polman, J.; Dijekema, R. ERâ: Identification and
characterization of a novel human estrogen receptor. FEBS Lett. 1996,
392, 49-53.
(5) Pike, A. C. W.; Brzozowski, A. M.; Hubbard, R. E.; Bonn, T.;
Thorsell, A. G.; Engstrom, O.; Ljunggren, J.; Gustafsson, J.-A.;
Carlquist, M. Structure of the ligand-binding domain of oestrogen
receptor beta in the presence of a partial agonist and a full antagonist.
EMBO J. 1999, 18, 4608-4618.
(6) Brzozowski, A. M.; Pike, A. C. W.; 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.
(7) Weihua, Z.; Andersson, S.; Cheng, G.; Simpson, E. R.; Warner, M.;
Gustafsson, J.-A. Update on estrogen signaling. FEBS Lett. 2003,
546, 17-24.
(8) Katzenellenbogen, B. S.; Katzenellebogen, J. K. Defining the ,S.
in SERMs. Science 2002, 295, 2380-2381.
(9) Grese, T. A.; Dodge, J. A. Selective estrogen receptor modulators
(SERMs). Curr. Pharm. Des. 1998, 4, 71-92.
(10) Lin, X.; Huebner, V. Non-steroidal ligands for steroid hormone
receptors. Curr. Opin. Drug DiscoVery DeV. 2000, 3, 383-398.
(11) Katzenellenbogen, J. A.; O’Malley, B. W.; Katzenellenbogen, B. S.
Tripartite steroid hormone receptor pharmacology: Interaction with
multiple effector sites as a basis for the cell- and promoter-specific
action of these hormones. Mol. Endocrinol. 1996, 10, 119-131.
(12) McKenna, N. J.; O’Malley, B. W. Combinatorial control of gene
expression by nuclear receptors and coregulators. Cell 2002, 108,
465-474.
Energy Evaluation. We extracted 200 snapshots (at time
intervals of 2 ps) for each species (complex, receptor and ligand)
from the last 400 ps of MD of the 44 ligand-ER complexes. The
various MM-PBSA energy terms in eq 1 were computed as follows:
(13) Minutolo, F.; Bertini, S.; Papi, C.; Carlson, K. E.; Katzenellenbogen,
J. A.; Macchia, M. Salicylaldoxime moiety as a phenolic “A-ring”
substitute in estrogen receptor ligands. J. Med. Chem. 2001, 44,
4288-4291.
G ) Gpolar + Gnonpolar + Emm - TS
(1)
(14) Minutolo, F.; Antonello, M.; Bertini, S.; Rapposelli, S.; Rossello,
A.; Sheng, S.; Carlson, K. E.; Katzenellenbogen, J. A.; Macchia, M.
Synthesis, binding affinity, and transcriptional activity of hydroxy-
and methoxy-substituted 3,4-diarylsalicylaldoximes on estrogen
receptors R and â. Bioorg. Med. Chem. 2003, 11, 1247-1257.
(15) 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.
(16) Minutolo, F.; Antonello, M.; Bertini, S.; Ortore, G.; Placanica, G.;
Rapposelli, S.; Sheng, S.; Carlson, K. E.; Katzenellenbogen, B. S.;
Katzenellenbogen, J. A.; Macchia, M. Novel estrogen receptor ligands
based on an anthranylaldoxime structure: Role of the phenol-type
pseudocycle in the binding process. J. Med. Chem. 2003, 46, 4032-
4042.
(17) Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S.; Chong,
L.; Lee, M.; Lee, T.; Duan, Y.; Wang, W.; Donini, O.; Cieplak, P.;
Srinivasan, J.; Case, D. A.; Cheatham, T. E. 3rd Calculating structures
and free energies of complex molecules: combining molecular
mechanics and continuum models. Acc. Chem. Res. 2000, 33, 889-
897.
(18) Miyaura, N.; Suzuki, A. Palladium-catalyzed cross-coupling reactions
of organoboron compounds. Chem. ReV. 1995, 95, 2457-2483.
(19) Littke, A. F.; Fu, G. C. A convenient and general method for Pd-
catalyzed Suzuki cross-couplings of aryl chlorides and arylboronic
acids. Angew. Chem., Int. Ed. 1998, 37, 3387-3388.
Electrostatic, van der Waals, and internal energies (Emm) were
obtained using the SANDER module in AMBER 8.0.39 Polar
energies (Gpolar) were obtained from the PBSA module of AMBER
8.0 program (using the Poisson-Boltzmann method), applying
dielectric constants of 1 and 80 to represent gas and water phases,
respectively. Nonpolar energies (Gnonpolar) were determined using
the MOLSURF program.
To compare the energetic interactions of the two orientations
(A and B in Figure 3) of the same ligand into the same receptor,
we took into account only the first three terms of eq 1, considering
the entropic value to be approximately constant.
∆Gbind ) Gcomplex - (Gprotein + Gligand
)
(2)
Solute entropy was evaluate only for the best orientation of each
ligand into the two receptors to correlate the predicted free energy
of binding (calculated as in eq 2) with the experimental one. It
was estimated using the NMODE module of AMBER 8.0 on a
total of 10 snapshots. Prior to the normal mode calculations, each
species (complex, receptor, or ligand) was subjected to a CG energy
minimization using a distance dependent dielectric, until a conver-
gence of 0.00001 kcal/mol‚Å.
(20) Lamm, B.; Liedholm, B. Nucleophilic aromatic substitution reactions.
III. Replacement of bromine by chloride ion in 2-bromo-3-nitro- and
2-bromo-5-methyl-3-nitrobenzenediazonium ions. Acta Chem. Scand.
1967, 21, 2679-2688.
(21) Karabatsos, G. J.; Hsi, N. Structural studies by nuclear magnetic
resonance-XI: Conformations and configurations of oxime O-methyl
ethers. Tetrahedron 1967, 23, 1079-1095.
(22) Katzenellenbogen, J. A.; Johnson, H. J., Jr.; Myers, H. N. Photo-
affinity labels for estrogen binding proteins of rat uterus. Biochemistry
1973, 12, 4085-4092.
Acknowledgment. We are grateful for the financial support
provided by grants from the National Institutes of Health (R37
DK15556, R37 CA25836, and P01AG024387 to J.A.K.).
Supporting Information Available: Tables reporting the
combustion analysis data and the MM-PBSA results calculated for
each ligand (A and B conformations) in both receptor subtypes

5012 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 16
Tuccinardi et al.
(23) Carlson, K. E.; Choi, I.; Gee, A.; Katzenellenbogen, B. S.; Katzenel-
lenbogen, 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.
affinities by a combination of molecular mechanics and continuum
solvent models. J. Med. Chem. 2000, 43, 3786-3791.
(32) Donini, O. A.; Kollman, P. A. Calculation and prediction of binding
free energies for the matrix metalloproteinases. J. Med. Chem. 2000,
43, 4180-4188.
(33) Huo, S.; Wang, J.; Cieplak, P.; Kollman, P. A.; Kuntz, I. D. Molecular
dynamics and free energy analyses of cathepsin D-inhibitor interac-
tions: Insight into structure-based ligand design. J. Med. Chem. 2002,
45, 1412-1419.
(34) Wrenn, C. K.; Katzenellenbogen, B. S. Structure-function analysis
of the hormone binding domain of the human estrogen receptor by
region-specific mutagenesis and phenotypic screening in yeast. J.
Biol. Chem. 1993, 268, 24089-24098.
(35) McInerney, E. M.; Weis, K. E.; Sun, J.; Mosselman, S.; Katzenel-
lenbogen, B. S. Transcription activation by the human estrogen
receptor subtype â (ERâ) studied with ERâ and ERR receptor
chimeras. Endocrinology 1998, 139, 4513-4522.
(24) Eiler, S.; Gangloff, M.; Duclaud, S.; Moras, D.; Ruff, M. Over-
expression, purification, and crystal structure of native ER alpha LBD.
Protein Expression Purif. 2001, 22, 165-173.
(25) Manas, E. S.; Xu, Z. B.; Unwalla, R. J.; Somers, W. S. Understanding
the selectivity of genistein for human estrogen receptor-beta using
X-ray crystallography and computational methods. Structure 2004,
12, 2197-2207.
(26) Brzozowski, A. M.; Pike, A. C.; Dauter, Z.; Hubbard, R. E.; Bonn,
T.; Engstrom, O.; Ohman, L.; Greene, G. L.; Gustafsson, J. A.;
Carlquist, M. Molecular basis of agonism and antagonism in the
oestrogen receptor. Nature 1997, 389, 753-758.
(27) Pike, A. C.; Brzozowski, A. M.; Hubbard, R. E.; Bonn, T.; Thorsell,
A. G.; Engstrom, O.; Ljunggren, J.; Gustafsson, J. A.; Carlquist, M.
Structure of the ligand-binding domain of oestrogen receptor beta in
the presence of a partial agonist and a full antagonist. EMBO J. 1999,
18, 4608-4618.
(28) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W.
E.; Belew, R. K.; Olson, A. J. Automated docking using a lamarckian
genetic algorithm and empirical binding free energy function. J.
Comput. Chem. 1998, 19, 1639-1662.
(29) Mewshaw, R. E.; Edsall, R. J., Jr.; Yang, C.; Manas, E. S.; Xu, Z.
B.; Henderson, R. A.; Keith, J. C., Jr.; Harris, H. A. ERbeta ligands.
3. Exploiting two binding orientations of the 2-phenylnaphthalene
scaffold to achieve ERbeta selectivity. J. Med. Chem. 2005, 48,
3953-3979.
(30) Yang, C.; Edsall, R., Jr.; Harris, H. A.; Zhang, X.; Manas, E. S.;
Mewshaw, R. E. ERâ Ligands. Part 2: Synthesis and structure-
activity relationships of a series of 4-hydroxy-biphenyl-carbaldehyde
oxime derivatives. Bioorg. Med. Chem. 2004, 12, 2553-2570.
(31) Kuhn, B.; Kollman, P. A. Binding of a diverse set of ligands to avidin
and streptavidin: an accurate quantitative prediction of their relative
(36) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.;
Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The Protein Data Bank.
Nucl. Acids Res. 2000, 28, 235-242.
(37) Macromodel, version 8.5; Schrodinger Inc., Portland, OR, 1999.
(38) http://www.scripps.edu/∼sanner/python/adt.
(39) Case, D. A.; Darden, T. A.; Cheatham, T. E., III; Simmerling, C. L.;
Wang, J.; Duke, R. E.; Luo, R.; Merz, K. M.; Wang, B.; Pearlman,
D. A.; Crowley, M.; Brozell, S.; Tsui, V.; Gohlke, H.; Mongan, J.;
Hornak, V.; Cui, G.; Beroza, P.; Schafmeister, C.; Caldwell, J. W.;
Ross, W. S.; Kollman, P. A. AMBER 8; University of California:
San Francisco, CA, 2004.
(40) Bayly, C. I.; Cieplak, P.; Cornell, W. D.; Kollman, P. A. A well-
behaved electrostatic potential based method using charge restraints
for determining atom-centered charges: The RESP model. J. Phys.
Chem. 1993, 97, 10269-10280.
JM060560U