Analogues of orphan nuclear receptor small heterodimer partner ligand and apoptosis inducer (E)-4-[3-(1-adamantyl)-4-hydroxyphenyl]-3-chlorocinnamic acid. 2. Impact of 3-chloro group replacement on inhibition of proliferation and induction of apoptosis of

Article abstract of DOI:10.1021/jm2011436

The parent phenol of adapalene and its (E)-cinnamic acid analogue were found to induce cancer cell apoptosis but cause adverse systemic effects when administered to mice. In contrast, their respective 5-Cl- and 3-Cl-substituted analogues had their adverse

Full text of DOI:10.1021/jm2011436

Article
pubs.acs.org/jmc
Analogues of Orphan Nuclear Receptor Small Heterodimer Partner
Ligand and Apoptosis Inducer (E)-4-[3-(1-Adamantyl)-4-
hydroxyphenyl]-3-chlorocinnamic Acid. 2. Impact of 3-Chloro Group
Replacement on Inhibition of Proliferation and Induction of
Apoptosis of Leukemia and Cancer Cell Lines
Zebin Xia,† Ricardo G. Correa,† Jayanta K. Das,‡ Lulu Farhana,‡ David J. Castro,† Jinghua Yu,†
Robert G. Oshima,† Joseph A. Fontana,‡ John C. Reed,† and Marcia I. Dawson*,†
†Cancer Center, Sanford-Burnham Medical Research Institute, La Jolla, California 92037, United States
‡Department of Veterans Affairs Medical Center and Wayne State University School of Medicine, Detroit, Michigan 48201,
United States
S
* Supporting Information
ABSTRACT: The parent phenol of adapalene and its (E)-cinnamic acid analogue were
found to induce cancer cell apoptosis but cause adverse systemic effects when
administered to mice. In contrast, their respective 5-Cl- and 3-Cl-substituted analogues
had their adverse effects mitigated without a comparable loss of cancer cell inhibitory
activity. As a result, pharmacologic space in this region of the cinnamic phenyl ring
scaffold was explored. Various substituents were introduced, and their effects on cancer
cell proliferation and viability were evaluated. Cinnamic acids having 3-Br, CN, NO2,
NH2, OMe, and N3 groups had activity comparable to that of 4-[3′-(1-adamantyl)-4′-
hydroxyphenyl]-3-chlorocinnamic acid. A comparative molecular field analysis study
indicated that introduction of an H-bond acceptor at position 3 of the central phenyl
ring would favor inhibition of leukemia cell viability, and docking suggested its hydrogen
bonding with a polar group in a small heterodimer partner homology model. The 3-CN, NO2, NH2, and OH analogues also
inhibited MMTV-Wnt1 murine mammary stem cell viability.
NR box to the AF-2 surface generally represses NR transactiva-
tional activity. While a natural ligand for SHP has not been
reported, our studies showed that small molecules such as 1 and 8
(Figure 1) by interacting with SHP induced SHP to interact with
the mSin3A multiprotein repressor complex,1 thereby modifying
expression of several genes, including c-Fos and c-Jun,8 leading to
activation of JNK and NF-κB signaling pathways and resulting in
apoptosis of cancer and leukemia cell lines.8−10
Without a structure of SHP available, ligand design has relied
on indirect methods such as structure−activity analysis and
docking of candidates to the homology model. The prototype
for SHP ligands was the aromatic retinoid 6-[3′-(1-adamantyl)-
4′-hydroxyphenyl]-2-naphthalenecarboxylic acid (AHPN/
CD437, 1), the parent phenol of the anti-acne drug adapalene.
The retinoid-like adverse effects of 1 were mitigated in the 4-
aryl 3-chlorocinnamic acid 8, which was unable to transcrip-
tionally activate the retinoic acid nuclear receptors (RARs).11
Molecular dynamics of the more rigid 5-chloronaphthalene-
carboxylic acid analogue (3) of 8 suggested why the chloro
group on the central aryl ring ortho to the diaryl bond would
interfere in the docking of 3 (Figure 1; X = Cl in Table 1) to
INTRODUCTION
■
We discovered that 5-Cl-AHPN (3) and 3-Cl-AHPC (8) (Figure 1)
are ligands of the orphan nuclear receptor (NR) small heterodimer
partner (SHP) on the basis of their competitive binding with the
5,5′-tritiated analogue of SHP ligand AHPN (1).1 SHP is one of
two unique orphan nuclear receptors in having only a ligand-
binding domain (LBD) through which it acts to regulate cell
processes such as survival, DNA methylation, and metabolism.2−4
Although a crystal structure of SHP has not been reported, a
homology model was derived from the structure [Protein Data
Bank (PDB) entry 1G2N] of the LBD of ultraspiracle (USP)
protein,5 which is the insect homologue of the vertebrate orphan
NR retinoid X receptor (RXR).6 This model, which has been
partially validated by mutagenesis,7 has the canonical NR LBD
structure of 12 sandwiched α-helices, a short β-sheet between
helices H5 and H6, a H12 containing a ligand-dependent activation
function 2 (AF-2), and a putative ligand-binding pocket (LBP) to
which a USP ligand could be docked.6 This model also revealed
that NR box or coactivator (CoA) motifs (LXXLL, X =
unspecified) were located in helical regions corresponding to NR
LBD H1, H6, and H10. The NR motifs permit a CoA protein to
interact with the NR AF-2 surface generated by NR H3, H4, and
H12, thereby allowing the NR−CoA complex bound to DNA to
recruit the transcriptional complex. In contrast, binding by a SHP
Received: August 26, 2011
Published: December 2, 2011
© 2011 American Chemical Society
233
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
diminish HL-60R cell apoptosis compared to that with 8.
Shifting the 3-Cl group to position 2 of the cinnamic acid ring
in 15 (X2 = Cl) caused a loss of activity against both cell lines
(Table 1), whereas the corresponding 2-F group of 14 (X2 = F)
with a volume of 10.7 Å3 weakened only MDA-MB-231 growth
inhibition but did not affect HL-60R apoptosis induction.
Docking studies described below now suggest reasons for these
differences.
These previous results suggested that further increasing the
volume of the central ring substituent ortho to the diaryl bond
would inversely correlate with levels of growth inhibition and
apoptosis induction. As a corollary, those analogues with
smaller substituents at position 3 would interact with SHP but
those with larger volumes would not. To more thoroughly
explore space in this region, these studies focused on additional
analogues of 6 because of their relative ease of synthesis.
Initially, evaluations employed KG-1 AML cells and then were
extended to OCI-AML2 AML, K562 chronic myeloid leukemia
(CML), and MOLT4 T-cell acute lymphoblastic leukemia
(T-ALL) cell lines for which the analogue dose response was
investigated. Several of the more potent analogues were also
evaluated using the murine MMTV-Wnt1 mammary stem cell
line in which the Wnt1 oncogene had been overexpressed.
RESULTS
■
Figure 1. Structures of the adamantyl-substituted retinoid-related
AHPN series: AHPN/CD437 (1), 5-Me-AHPN (2), 5-Cl-AHPN (3),
and 4. Structures of the similarly substituted AHPC series: 3-A-AHPC
(5), AHPC (6), 3-Me-AHPC (7), 3-Cl-AHPC (8), and 9−15.
Design and Chemical Synthesis. Computational studies
indicated that 8 assumed two poses when docked to the SHP
model.13 Both poses suggested that the carboxylate group of 8
could have ionic interactions with the side chain of R238 in
SHP helix H11, its phenolic hydroxyl could form an H−π bond
with the phenyl ring of F96 in H5, and its 1-adamantyl group
could have van der Waals and/or hydrophobic contacts with
L97, L100, A145, and W148. In contrast, the cinnamyl ring of 8
occupied orthogonal positions so that the 3-Cl group was
directed toward LBP hydrophobic residues or toward the polar
side chains of T55 and C56 in H3. This second orientation
suggested a possible H-bond interaction with the threonine
hydroxyl group and/or cysteine sulfhydryl group. To explore
the importance of the substituent at position 3 with respect to
volume and electrostatic properties and its subsequent impact
on ligand orientation in the SHP LBP, analogues 46−52, 57,
59, and 61 with various substituents at position 3 were
designed, synthesized (Schemes 1 and 2), and evaluated (Table 2).
In addition, the effect on activity of a substituent at cinnamic
acid ring position 6, which is ortho to the cinnamic acid double
bond, was explored in 63 (X1 = Br), and that of restricting the
aryl rings to a planar orientation was investigated in tricyclic
carbazoles 66 and 69, in which a common nitrogen atom and
the diaryl bond combined to restrict ring rotation.
The general route to the 3-substituted analogues of 8
involved constructing a series of (E)-cinnamate intermediates
by a Mizoroki−Heck vinyl−aryl coupling14 of 3,4-disubstituted
phenyl halides with ethyl acrylate or by a Wittig reaction15,16 on
3,4-disubstituted benzaldehydes. After suitable functionalization
followed by the Suzuki−Miyaura diaryl coupling,17 suitably-
functionalized cinnamates with 3-(1-adamantyl)-4-benzyloxy-
phenylboronic acid (34) were used to introduce the central
diaryl bond (Schemes 1 and 2). One-step procedures were used
to prepare four of the five precursors required for constructing
the required 3,4-disubstituted cinnamates (Scheme 1). Thus,
diazotization of aniline 16 at −5 to 0 °C and treatment with
potassium iodide at 22−100 °C afforded 4-bromo-3-trifluor-
omethylphenyl iodide (17) required for intermediate 25.
the agonist conformation of the RARγ LBD (PDB entry
3LBD).12 With the naphthalenecarboxylate group of 3
anchored by ionic and/or H-bonding interactions with the
LBD H5 R278 side chain in the RAR ligand-binding pocket,
the larger 5-Cl group caused the torsion angle between the aryl
rings of 3 to increase so that its 1-adamantyl group (1-Ad)
clashed with side chains in the agonist position of H11, thereby
preventing H12 from forming the AF-2 surface. Fortunately,
the new orientation of the aryl rings did not prevent 3 and 8
from inducing cancer cell apoptosis.11
To explore chemical space about the 5-Cl group of 3 and the
3-Cl group of analogue 8, we had previously synthesized and
evaluated a limited series of analogues. Their abilities to inhibit
the proliferation and induce the apoptosis of all-trans-retinoic
acid (ATRA)-resistant human HL-60R myeloid leukemia and
MDA-MB-231 breast cancer cell lines at 1.0 and 5.0 μM are
summarized in Table 1. The growth inhibitory and apoptotic
activities of several compounds from these initial studies
indicated that introducing particular groups at position 5 of
naphthalene-2-carboxylic acid 1 and position 3 of cinnamic acid
6 negatively impacted efficacy. Their analogues, 2 and 7,
respectively, with a methyl group at these positions were less
active than the parent structures, whereas 3 and 8 with a chloro
group, which was only 3 Å3 smaller (13.5%) in volume than the
methyl group, had these activities partially restored.11 In
contrast, introduction of larger groups such as hydroxymethyl
and ethoxy in 4 and 10, respectively, caused the loss of both
activities. Moreover, the introduction of two methyl and
methoxy substituents at positions 3 and 5 of the central phenyl
ring (11 and 12, respectively), both of which are ortho to the
diaryl bond, further diminished activity against MDA-MB-231
cell proliferation, whereas introducing 3,5-dichloro groups (13)
did not impact MDA-MB-231 growth inhibition but did
234
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
Table 1. Effects of a Substituent on the Central Aryl Ring on Analogue Activity Compared to That of the 5-Chloro Group of 1
and the 3-Chloro Group of 8 in Terms of Human Leukemia and Breast Cancer Cell Growth Inhibition and Apoptosis
a
Induction
a
b
Activities of 1−3, 6−8, and 10 in inducing HL-60R and MDA-MB-231 apoptosis and growth inhibition were previously reported.11 Results
c
represent the average of triplicates ≤10 (standard deviation). Calculated volumes (cubic angstroms) of X: H, 5.77; F, 10.7; Cl, 19.3; Me, 22.3;
CH2OH, 30.6; OMe, 31.3; OEt, 48.1.47 Statistical analysis (probability of difference of the treatment value from that of the nontreated control):
d
e
f
P < 0.001. Statistical analysis (probability of difference of the treatment value from that of the nontreated control): P < 0.01. Statistical analysis
(probability of difference of the treatment value from that of the nontreated control): P < 0.05. Statistical analysis (probability of difference of the
g
treatment value from that of the nontreated control): P > 0.05.
Benzylation of phenols 18 and 20 provided the 3-benzylox-
yphenyl bromides 19 and 21 for preparing the respective
intermediates 26 and 29. Palladium-catalyzed acrylation of 26
and 29, BBr3-mediated debenzylation, and treatment with triflic
anhydride yielded the ethyl 3-cyano- and 3-carbomethoxy-4-
triflyloxycinnamates 28 and 31, respectively. Diisobutylalumi-
num hydride reduction of 4-bromo-3-methoxybenzonitrile (22)
provided benzaldehyde 23. Both 23 and 24 underwent a Wittig
olefination to afford the respective 3-methoxy- and 3-nitro-4-
bromocinnamate ester intermediates 32 and 33.
Debenzylation of 53−55 produced 56, 60, and 62, respectively,
which were hydrolyzed to provide their respective 3-amino-, 3-
bromo-, and 2,5-dibromocinnamic acids 57, 61, and 63. The
amino group of 56 was diazotized (HNO2) and treated with
sodium azide to produce 3-azidocinnamate 58, which after ester
hydrolysis provided 3-azidocinnamic acid 59. Reductive cyclization
of 39 by heating (180 °C) with triphenylphosphine in 1,2-
dichlorobenzene18,19 afforded the regioisomeric carbazoles 64 and
67, which were separated and then deprotected in two steps to give
66 and 69, respectively.
Palladium-catalyzed coupling of aryl bromides 25, 32, and 33
and aryl triflates 28 and 31 with 34 produced the protected 4-
arylcinnamates 35, 38, 39, 36, and 37, respectively, which were
then debenzylated by using BBr3. Another BBr3-mediated
cleavage of the 3-methoxy group of 43 gave 3-OH-substituted
44. Base hydrolysis and acidification yielded target cinnamic
acids 46−48 and 50−52. The hindered 3-carbomethoxy group
of 48 did not hydrolyze at 0−22 °C, which was routinely used
for the ethyl ester hydrolysis, but did at 82 °C to afford 3-
carboxy-substituted 49 (71%).
The 3-nitrocinnamate 39 was also used to prepare analogues
57, 59, 61, 63, 66, and 69 (Scheme 2). Reduction of the 3-nitro
group of 39 using stannous chloride in ethanol provided 3-
aminocinnamate 53. Diazotization of 53 under hydrophobic
conditions using tert-butyl nitrite in the presence of cupric bromide
(2 equiv) at 0−22 °C produced the 2,5-dibromocinnamate 55.
In contrast, treatment of 53 with n-hexyl nitrite in bromoform
(34 equiv) at 100 °C afforded the 3-bromocinnamate 54.
Biological Evaluation. Antiproliferative and Apoptotic
Activities. Analogues 46−52, 57, 59, 61, 63, 66, and 69 were
compared to the 3-chloro congener (8) for their ability to
inhibit proliferation or viability of four human leukemia cell
lines (KG-1 AML, OCI-AML2 AML, K562 CML, and MOLT4
T-ALL) and induce KG-1 cell apoptosis (Table 2). These cell
lines were selected as representative of the acute myeloid, chronic
myeloid, and T-cell acute lymphoblastic types of leukemia having
varied resistances to synthetic retinoid 1 from which these
analogues were originally derived20,21 and to chemotherapeutic
agents and a variety of defects that facilitates their proliferation,22−39
which are described in the Supporting Information. The KG-1
AML line was also used for comparison with the activities of the
first-generation analogues (see Table 1).
The viability of these leukemia lines was determined after
treatment for 72 h by measuring changes in their ATP levels
using assays for luciferase, for which ATP serves as a cofactor
(Table 2). The dose−response curves used to determine their
235
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
a
Scheme 1. Syntheses of 46−52
a
Reagents and conditions: (a) NaNO2, H2SO4, 0 to 5 °C, KI, room temperature to 100 °C; (b) PhCH2Br, K2CO3, acetone, reflux; (c) ethyl acrylate,
Pd(OAc)2, Et3N, tri(o-tolyl)phosphine, reflux; (d) DIBAL, MePh, −78 °C to room temperature; MeOH, 10% H2SO4; (e) (carbethoxymethylene)-
Ph3P, MePh, reflux; (f) BBr3, CH2Cl2, −78 °C; H2O; (g) Tf2O, pyridine, CH2Cl2, 0 °C to room temperature; (h) for compounds 35, 38, and 39,
3-(1-adamantyl)-4-benzyloxyphenylboronic acid (34), Pd(PPh3)4, DME, 2 M Na2CO3, reflux; for compounds 36 and 37, compound 34, Pd(PPh3)4,
LiCl, DME, 2 M Na2CO3, reflux; (i) BBr3, CH2Cl2, −78 °C to room temperature; H2O; (j) for compounds 46, 48, and 51, aqueous NaOH, MeOH,
reflux; H3O+; for compounds 47, 50, and 52, LiOH·H2O, THF/MeOH/H2O, room temperature; H3O+; (k) LiOH·H2O, 5 M NaOH, THF/
MeOH/H2O, room temperature to 82 °C; H3O+.
IC50 values are shown in Figure S1 of the Supporting
Information. Analysis of the effects of the analogues on these
leukemia lines indicated that both the volume and electron
density of the substituent at position 3 on the cinnamic ring
impacted antiproliferative activities. In analogues that induced
the highest levels of KG-1 cell growth inhibition and apoptosis,
a variety of polar substituents were tolerated, including 3-Cl,
CN, OMe, NO2, NH2, N3, and Br. Of these, the 3-OMe and Br
groups of 50 and 61, respectively, provided the most potent
KG-1 growth inhibitory activity at 1.0 and 5.0 μM, which
approximated that of 8 having a 3-Cl group, whereas all six
groups exhibited similar potencies in inducing KG-1 apoptosis.
In contrast, 48 and 51 with 3-CO2Me and 3-OH groups,
respectively, were approximately half as potent at 5.0 μM in
inducing apoptosis. Analogues 46 and 49 with 3-CF3 and
3-CO2H groups, respectively, were essentially inactive. These
substituent effects were observed to track with analogue
potencies in inhibiting K562, OCI-AML2, and MOLT4 cell
viability. In terms of IC50 values, OCI-AML2 AML and
MOLT4 T-ALL cells were more sensitive (approximately 3−4-
fold) to the analogues than K562 CML cells. Analogues that
inhibited viability could be sorted into two groups on the basis
of their IC50 values with those having 3-Cl, CN, OH, NO2, and
NH2 groups being approximately twice as potent as those with
the 3-OMe, N3, and Br groups. An anomaly to these trends was
the 3-CF3 group of 46. At 5.0 μM, 46 was unable to induce KG-
1 cell apoptosis after 48 h but did inhibit OCI-AML2 and
MOLT4 viability by 50% at 3.8 and 3.9 μM, respectively, at 72
h, whereas at 5.0 μM, K562 cell viability was inhibited by only
25%. The second anomaly was the lack of activity of 49 having
the 3-CO2H group. The deprotonation of this group at
physiologic pH may have impeded the cellular uptake of 49.
Introducing a second group at position 5 of the cinnamic ring
led to a further decrease in apoptotic activity in the 3,5-dimethyl,
dimethoxy, and dichloro analogues (11−13, respectively)
compared to their respective monosubstituted analogues (7, 50,
8, and 61). In contrast to the growth inhibitory and apoptotic
activities provided by the 3-amino analogue 57, its planar
analogues 66 and 69 having their diaryl rings linked through an
o-NH group had very low or no activity against the four leukemia
cell lines.
Introducing a second bromo group at position 6 of the
cinnamic ring of the 3-bromocinnamic acid 61 produced 63
with substantially lower KG-1 apoptosis-inducing activity. At
1.0 μM, 63 was inactive and its IC50 values for inhibiting
leukemia cell viability ranged from 1.4 to 4.2 μM. This result is
in agreement with the observation that shifting the chloro
group to position 2 of the cinnamyl ring in 15 led to a loss of
236
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
a
Scheme 2. Syntheses of 57, 59, 61, 63, 66, and 69
a
Reagents and conditions: (a) SnCl2·H2O, EtOH, 80 °C; (b) CuBr2, tert-butyl nitrite, MeCN, 0 °C to room temperature; (c) n-hexyl nitrite, CHBr3,
100 °C; (d) BBr3, CH2Cl2, −78 °C; H2O; (e) LiOH·H2O, THF/MeOH/H2O, room temperature; H3O+; (f) 2 M H2SO4, NaNO2, dioxane, −10 °C;
NaN3, H2O, −10 °C; (g) 25% NaOH, EtOH/Et2O (2:1) (59) or EtOH (69), room temperature; H3O+; (h) PPh3, 1,2-dichlorobenzene, 180 °C.
MDA-MB-231 growth inhibition and HL-60R AML cell
apoptosis-inducing activities, whereas its replacement with the
smaller fluorine in 14 retained HL-60R apoptotic activity but
reduced the level of MDA-MB-231 growth inhibition compared to
that of 6, which had a hydrogen at position 3. In summary, of the
new analogues, those having 3-CN, OH, NO2, and NH2 groups
(47, 51, 52, and 57, respectively) most potently inhibited leukemia
cell viability with IC50 values comparable to those of 3-Cl-
substituted 8. These four groups have unpaired electrons that could
function as H-bond acceptors. Of these, the hydroxyl and amino
groups could also function as H-bond donors. The 3-Cl group of 8
could interact with electron-donating atoms (O of T55 and S of
C56)40 or perhaps have van der Waals interactions with the methyl
group of T55.41
SAR Analysis. The impact of the size of the substituent at
position 3 on apoptotic activity was demonstrated previously by
the inability of the 3-(3-acetamidopropyloxy) analogue (3-A-
AHPC, 5) to induce apoptosis; on the other hand, the 3-Cl
analogue (8) was active.11 A component SAR analysis (Table
S2 and Figure S2 of the Supporting Information) using 11
analogues was undertaken to investigate the impact of the
volume of the substituent at position 3 on K562, OCI-AML2,
and MOLT-4 leukemia cell viability as determined by the
decrease in ATP levels (Table 2). This analysis excluded
inactive 49 (X = CO2H), inactive planar 66 and 69, and 63
with a 3,6-dibrominated cinnamyl ring to provide a series of
active and similar compounds. Despite the fact that a
downstream event (loss of the mitochondrial generation of
ATP) was measured in cells that also were required to take up
the analogues through the cell membrane, correlations were
obtained in the log−log plots of the volume of the substituent
at position 3 versus IC50 values for the three leukemia cell lines.
The less active compounds had substituents at position 3 with
larger volumes (Figure S2B of the Supporting Information).
Thus, 46 with a 3-CF3 group (37 Å3) and 48 with a 3-CO2Me
group (50.3 Å3) were the least active, whereas 52 with a 3-NO2
group (29 Å3) and 59 with a 3-N3 group (31 Å3) were active.
This trend was evident upon comparison of similar
substituents. For example, 8 with a smaller 3-Cl group (19
Å3) was 1.9−3.0-fold more active than 61 with a 3-Br group (24
Å3), and 47 with a 3-CN group (23 Å3) was 2.4−4.5-fold more
active than 59 with a 3-N3 group (30.7 Å3).
Next, the combined impact of the volume of the substituent
at position 3 and H-bond acceptor and donor capacity on the
inhibition of MOLT4 cell viability was examined using
comparative molecular field analysis (CoMFA) (Figure 2).
The resulting partial contour map shows that the favored steric
and positively charged fields were asymmetric and that the
favored negatively charged field was symmetric (Figure 2A).
The large red contour map suggests that an H-bond acceptor
would have a wider role in interacting with a SHP pocket
residue to support the higher activities observed for analogues
47, 51, 52, and 57. Similar partial contour maps were obtained
from analyses of the results from the treatment of K562 and
OCI-AML2 cells. In Figure 2B are shown plots of the actual
and predicted logarithms of the IC50 values for viability
inhibition of the three cell lines obtained from the CoMFA
studies that suggest robust correlations. The data used to
construct the three maps are summarized in Tables S2 and S3
of the Supporting Information. These trends continued when
comparing the physical properties of the substitutents at
position 3 of the 11 analogues with their KG-1 apoptosis-
inducing activities.
The most active analogues (8, 47, 51, 52, and 57) together
with much less active analogue 46 were also evaluated for their
effects on MMTV-Wnt1 murine mammary cancer stem cell
proliferation. Their dose−response curves are shown in Figure
3. On the basis of the IC50 values (Table 2) obtained from
these curves, analogues with 3-Cl and NH2 groups (8 and 57,
respectively) had comparable activity and were more potent
237
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
Table 2. Effects of a Substituent at Position 3 of the AHPC Scaffold on KG-1 AML Cell Growth Inhibition and Apoptosis
Induction and Inhibition of Leukemia Cell (K562, OCI-AML2, and MOLT4) and Mammary Cancer Stem Cell Viability
a
b
Results represent the means of triplicates ≤10 [standard deviation (SD)]. Leukemia cell lines treated with 0, 0.1, 0.25, 0.5, 1.0, 2.5, and 5.0 μM
compound. IC50 values were determined from the dose−response curves shown in Figure S1 of the Supporting Information using data representing
c
means of quadruplicate trials SD (K562, ≤10; OCI-AML2, ≤11; MOLT4, ≤10), as was the determination of viability (percent at 5 μM). Cells
treated with 0, 0.01, 0.03, 0.1, 0.3, 1.0, and 3.0 μM compound. IC50 values determined graphically from each of triplicate dose−response curves are
d
represented as means of triplicates SD. Calculated volume (cubic angstroms) of X: Cl, 19.3; CF3, 37.1; CN, 22.6; CO2Me, 50.3; CO2H, 32.8;
OMe, 31.3; OH, 13.8; NO2, 29.1; NH2, 17.1; N3, 30.7; Br, 23.7.47 Statistical analysis (probability of difference of the treatment value from that of
e
f
the nontreated control): P < 0.01. Statistical analysis (probability of difference of the treatment value from that of the nontreated control): P > 0.05.
g
h
i
Statistical analysis (probability of difference of the treatment value from that of the nontreated control): P < 0.05. Extrapolated value. Statistical
j
analysis (probability of difference of the analogue IC50 value from that of 8): P < 0.001. Statistical analysis (probability of difference of the analogue
IC50 value from that of 8): P < 0.01. Statistical analysis (probability of difference of the analogue IC50 value from that of 8): P < 0.005. Statistical
analysis (probability of difference of the analogue IC50 value from that of 8): P > 0.1. nd, not determined.
k
l
m
than analogues with 3-CN, OH, and NO2 groups (47, 51, and
52, respectively). Compound 46 inhibited MMTV-Wnt1
proliferation with an IC50 value of 2.37 μM, which was more
than 50-fold higher than that determined for 8. The IC50 values
of analogues in the MMTV-Wnt1 murine mammary stem cell
assay were much lower than those obtained using the leukemia
cell lines as would be expected for cells growing in media
without serum.
Binding to SHP. As one of the active analogues listed in
Table 2, 47 (X = CN) was selected to illustrate how interaction
with SHP protein could have roles in leukemia cell apoptosis
induction and inhibition of cell viability. SHP has thus far defied
crystallization and structure determination. In our hands, the
recombinant wild-type human protein proved to be too labile for
determining binding using isothermal titration calorimetry, and
therefore, we resorted to low-temperature, high-field NMR
spectrometry. Binding of 47 to SHP (Figure 4A) was
demonstrated in comparison with 8 (X = Cl) (Figure 4B) as a
positive SHP-binding control. Compound 8 was previously shown
to bind to SHP in a competitive binding assay with
[5,5′-3H2]AHPN.1 The 6.3−7.9 ppm region of the NMR spectra
of the compounds was selected for comparison because ligand
proton signals in that region were not obscured by extraneous
proton signals from water, tris(hydroxymethyl)aminomethane in
238
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
Figure 2. CoMFA of replacement of the 3-chloro group of 8 on leukemia cell viability. (A) Partial CoMFA contour maps (StDev*Coeff) for steric
and electrostatic fields about the substituent at position 3 based on MOLT4 IC50 viability data with analogue 47 displayed as a visual reference.
Favored steric contributions are colored green and disfavored steric contributions yellow; the blue contour indicates regions where positive charges
are favored and red regions where negative charge is favored. (B) Plot of predicted vs actual pIC50 values of inhibition of leukemia cell viability by 8,
46−48, 50−52, 57, 59, 61, and 62 for K562 (green diamonds), OCI-AML2 (blue triangles), and MOLT4 (red dots) leukemia cell lines determined
by CoMFA.
from the structure of the homologous USP LBD by the
Pelliccaria group.6 In the model, SHP helix H12 assumed the
canonical agonist position to form an AF-2 surface with H3 and
H4. This model also has a putative ligand-binding pocket (LBP)
to which the USP phospholipid ligand and analogues 1, 3, and 8
had been successfully docked.6,13 The docking poses for 1, 3, and
8 were highly similar and suggested potential stabilizing
interactions with the following LBP residues: T55 and C56
(H3), F96, L97, and L100 (H5), P139, A145, and W148 (H7),
L231 and L235 (H11), and R238 and I240 (H11−H12 loop).13
Formation of ionic and/or H-bonding interactions between an
analogue carboxylate group and the R238 guanidinium group and
a nonconventional H−π bond between an analogue phenolic
hydroxyl group and the F97 phenyl ring were regarded as the
most important polar interactions for stabilization.
The docked poses of the current analogues substituted at
position 3 were compared with that of 8 (Figure 5A). Active
analogues such as 47 (X = CN) and 52 (X = NO2) assumed
docking poses similar to that of 8. The respective distances
between their carboxylate carbon atoms and the H11 R238
guanidinium carbon were measured as 4.3 and 4.1 Å, respectively,
Figure 3. Effects of 8, 46, 47, 51, 52, and 57 on MMTV-Wnt1
mammary cancer stem cell viability as measured by DAPI staining and
cell counting. Cells were treated with each analogue at the indicated
compared to a value of 3.8 Å for 8. These distances suggest that
concentrations or with vehicle alone (Me2SO control) for 72 h before
their carboxylate groups would contribute H-bond and/or ionic
the viability of treated cells relative to vehicle-treated control was
interactions with the R238 guanidinium group to stabilize binding.
determined. Results represent the means of triplicates the standard
On the basis of position and distance, the 4′-OH groups of 47 and
52, like that of 8, could form H−π bonds with the π-electron
cloud of the H5 F96 phenyl ring to enhance analogue−LBP
interaction.13,42 The respective distances between their OH
oxygens and the F96 phenyl ring centerx were 3.0 and 2.8 Å
compared to 3.0 Å for 8. van der Waals interactions between their
3′-(1-Ad) groups and the surrounding LBP W148, L97, and L100
side chains could also occur to further stabilize binding. The
shortest distances measured between the 3′-(1-Ad) group carbons
of 47 and 52 and the W148 indole ring carbons were 3.6 and
3.3 Å, respectively, compared to 3.9 Å for 8. Similarly, the shortest
distances measured from their 3′-(1-Ad) group carbons to one of
the L97 methyl carbons were 3.7 and 4.1 Å, respectively,
compared to 3.9 Å for 8, and those to one of the L100 methyl
carbons were 3.8 and 4.2 Å, respectively, compared to 3.6 Å for 8.
deviation.
Tris buffer, reduced glutathione, and DMSO, which were present at
higher concentrations. The lower scans in panels A and B of Figure
4 show that the peak heights of the six aromatic and two olefinic
protons in this region of the spectra of both 47 and 8 were
decreased and broadened in the presence of SHP, thereby
indicating their interaction with the protein. In contrast, proton
signals for inactive analogue 46 (X = CF3) were not depressed by
the addition of SHP (Figure 4C), thereby indicating the absence or
only a very weak interaction of 46 with the protein.
Computational Studies. Docking to the SHP Model.
To discern how the substituent at position 3 impacted SHP in
its inhibition of cell viability, docking studies (Figure 5) were
conducted using the three-dimensional model of SHP derived
239
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
Figure 5. Poses assumed by 8, 14, 15, 46, 47, 49, and 52 upon docking
to the ultraspiracle ligand-binding domain-derived model of small
heterodimer partner (SHP). (A) Overlap of docking poses for 8, 47,
and 52 in stick format with LBP residues in line format. C atoms of 8, 47,
and 52 and those of their corresponding pocket residues are colored cyan,
orange, and magenta, respectively, with other atoms colored as follows: Cl,
green; F, light blue; N, blue; O, red; S, yellow. (B) Overlap of docking
poses for 8, 46, and 49 shown in stick format with LBP residues in line
format. C atoms of 8, 46, and 49 and those of their corresponding pocket
residues are colored cyan, magenta, and orange, respectively, with other
atoms colored as in panel A. (C) Overlap of docking poses for 8, 14, and
15 shown in stick format with conformations of SHP ligand-binding
pocket (LBP) residues shown in line format. C atoms of 8, 14, and 15 and
those of their corresponding pocket residues are colored cyan, magenta,
and orange, respectively, with other atoms colored as in panel A. The
residue numbering is that for the human SHP protein sequence. Docking
studies were performed using BioMed Cache version 6.2 with flexible
ligands and protein side chains. 8, 14, 47, and 52 were classified as active
compounds, whereas analogues 15, 46, and 49 were inactive or less active
in terms of growth inhibition, loss of viability, or apoptosis induction.
Figure 4. 1H NMR spectra indicate that apoptosis-inducing 3-
cyanocinnamic acid 47 binds to SHP, whereas inactive 3-
(trifluoromethyl)cinnamic acid 46 does not. The low-field regions of
1
one-dimensional H NMR spectra of 47, 46, and 8 alone and in the
presence of recombinant human SHP are shown. Spectra were
recorded at 11 °C on (A) 47 (100 μM), (B) 8 (100 μM), and (C) 46
(200 μM) in the absence (top spectra) and presence (bottom spectra)
of 0.5 (A and B) and 1.0 μM SHP protein (C). Note that small peaks
at 7.3 and 7.75 ppm in panels A and B represent an impurity in the
D2O (see Figure S4 of the Supporting Information). The spectrum in
panel C was that of a sample containing D2O from another source.
Prior docking using 8 suggested a potential stabilizing inter-
action could occur between its 3-Cl group and the H3 C56 side
chain because of formation of a Cl···H−S bond (4.2 Å Cl−S
distance).13 The measured distances of 3.7 Å from the 3-CN
nitrogen of 47 to the C56 sulfur and 4.3 Å from a 3-NO2 oxygen
of 52 to the C56 sulfur suggest that the 3-CN and 3-NO2 groups
could form similar N···H−S and O···H−S H-bonding inter-
actions.43 The docked poses of 47 and 52 also suggest that their 3-
CN and 3-NO2 groups, respectively, could contribute to ligand
stabilization by forming hydrogen bonds with the T55 OH group,
just as that of 8 suggests its 3-Cl could form a similar bond with
the OH group, while the two aromatic rings of these analogues
could make hydrophobic contacts with the side chains of L231,
L235, and I240.
240
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
Analogues without or with very weak activity such as 46 (X =
CF3) and 49 (X = CO2H) assumed docking poses that differed
from that of 3-Cl-AHPC (Figure 5B). Their carboxylates were
farther from the R238 guanidinium group than that of 8, and
their substituents at position 3 were oriented toward hydro-
phobic LBP residues rather than T55 and C56, which is shown
in the overlap of their docking poses in the SHP LBP. The
measured distances from the carboxylate carbons of 46 and 49
to the R238 guanidinium carbon of 5.8 and 5.1 Å, respectively,
suggest only weak or no H-bond or salt bridge interactions
would occur in contrast to those possible for the carboxylate of
8. Their 3-CF3 and 3-CO2H groups would occupy a side pocket
between the H7 P141 and H11 L235 side chains, where their
respective fluorine and oxygen atoms would be unable to
function as H-bond acceptors. These poses also suggest that
their 4′-OH groups, which are farther from the F96 phenyl ring
(OH O−phenyl center distances of 4.0 and 4.1 Å, respectively)
than that of 8 (3.0 Å), would contribute weaker or no
stabilizing H−π bond interactions. Although the poses do
suggest that the 3′-(1-Ad) groups of 46 and 49 could interact
with the hydrophobic side chains of L97, L100, A145, and
W148 (e.g., the shortest distance to the W148 indole ring was
3.4 Å compared to 3.9 Å for 8), their contributions by being
weaker would contribute far less to binding stabilization,
thereby limiting the residence time of the analogue in the LBP
in agreement with their lack of activity.
Analogues with substituents at position 2 such as 14 (2-F)
and 15 (2-Cl) were also docked and compared with 8 (2-H)
(Figure 5C). Active 14 with the smaller F group at position 2
had a pose similar to that of 8. In contrast, inactive 15 with the
larger 2-Cl group had a pose in which its cinnamic moiety was
farther from H11 and closer to H12 than that of 8. As a result,
the distance (4.7 Å) between the carboxylate carbon of 15 and
the R238 guanidinium carbon, by being longer, suggests that
any H-bond or ionic interactions with the R238 guanidinium
group would be weaker. The short distance (2.46 Å) between
the 2-Cl atom and I240 side chain methyl carbon suggests a
potential clash that would prevent robust interaction with the
pocket. Thus, while the inactive analogues could be docked into
the SHP LBP, their poses suggest that any polar interactions
between the analogue and LBP residues required for stabilizing
binding would be absent or weak.
derived model) or R34 (canonical DAX-1-derived model),
whereas the antagonist 511 assumed a pose in which its
carboxylate interacted with R238 (USP-derived model).
Molecular dynamics using the USP-derived model suggested
that H-bond and/or ionic stabilizing interactions were favored
between the 8 carboxylate and SHP R238 (80%) followed by
Q140 (18%) and R138 (17%) and that SHP NR boxes 1 and 3
were stabilized by 8, whereas NR box 2 and the H12 agonist
position were destabilized.
Mutational analysis indicated that unlike wild-type SHP, the
double mutant L5-6 R138D/H11 R238D, in which reversal of
side chain charge would not support interaction with a ligand
carboxyl group, and the H3 F96A point mutant, which by
lacking the phenyl ring was incapable of stabilizing the ligand
hydoxyl group, were unable to block HNF4α transactivation of
its site on the ApoCIII promoter−Luc reporter construct,
thereby suggesting that R138, R238, and F96 played roles in
mediating SHP repressor activity. Moreover, F96A and
R138D/R238D mutants prevented SHP from inhibiting basal
HNF4α activation by 44%, whereas the W148A, R238A,
R138A, W92A, R34A, and R34Q/Q134R mutants did not have
a similar effect. Although SHP antagonist 5 [X = O-
(CH2)3NHAc] did dock into the LBP “canonical” arm of the
DAX-1-derived model to suggest that its pharmacophoric
elements (carboxylate and hydroxyl) could interact with other
LBP residues, their mutation to alanine did not affect the ability
of SHP to repress HNF4α transactivation and so did not
support this canonical model. These mutational studies support
our use of the USP LBD-derived “allosteric model” for ligand
design. In this model, docking places the carboxylate of 8 at the
pocket entrance, as do the charged termini of phospholipids
bound to USP, SF-1, and LRH-1. Much remains to be learned
about SHP, including the impact of ligands on NR box motif
stabilization and the actual geometry of the LBP in the presence
of a ligand, which highlight the necessity of determining the
actual structure of SHP.
Conclusion. The results indicate that leukemia cell apoptotic
activity and inhibition of cell viability are retained in analogues of 8
as long as their substituents at position 3 had some electron donor
(H-bond-accepting) capacity and restricted volume.
EXPERIMENTAL SECTION
■
Chemistry. General Methods. Chemicals and solvents from com-
mercial sources were used without further purification. 3-(1-Adamantyl)-
4-benzyloxyphenylboronic acid (34) was synthesized using our reported
procedure.44 Abbreviations for solvents and reagents are as follows: BnBr,
benzyl bromide; DIBAL, diisobutylaluminum hydride; DME, 1,2-
dimethoxyethane; Tf2O, trifluoromethanesulfonic anhydride. Experimen-
tal procedures were not optimized. Anhydrous and/or oxygen-sensitive
reactions were conducted under argon gas. Reactions were monitored by
thin-layer chromatography on silica gel (mesh size 60, F254) with
visualization under UV light. Unless specified, the standard workup
involved washing the organic extract with water and brine and drying over
Na2SO4 followed by concentration at reduced pressure. Chromatography
refers to flash column chromatography on silica gel (Merck 60, 230−400
mesh). Melting points of samples were determined in capillaries using a
Mel-Temp II apparatus and are uncorrected. Infrared spectra were
obtained on powdered or liquid samples using an FT-IR Mason satellite
DISCUSSION
■
Our results indicate that the leukemia cell apoptosis-inducing
and viability-inhibiting activities of 3-substituted analogues of 8
would be retained if the substituents had some electron-
donating (H-bond-accepting) capacity and restricted volumes.
Recently, Pelliccaria, Wang, and co-workers conducted an
elegant series of computational and mutagenesis studies to
examine the SHP LBP. Several models of the SHP protein were
constructed on the basis of its homology to the phospholipid-
complexed USP LBD, docosahexaenoic acid and oleic acid-
complexed RXRα LBD, and DAX-1 using their reported crystal
structures.7 These models displayed either the canonical NR
LBP such as those found in the ligand-bound PPARs, RARs,
RXRs, TRs, and VDR or another pocket that they termed
“allosteric” such as that found in NRs bound to charged
phospholipids such as USP, SF-1, and LRH-1. Both allosteric
and canonical models featured an L-shaped LBP; however,
docking studies revealed that 8 occupied different arms of the L
with its adamantyl group residing at the arm juncture so that its
carboxylate could interact with either Gln140 (allosteric USP-
1
spectrophotometer. H NMR spectra were recorded on a 300 MHz
Varian Unity Inova spectrometer or an ECS 400 MHz Jeol spectrometer,
and shift values are expressed in parts per million (δ) relative to CHCl3 as
the internal standard. Unless mentioned, NMR samples were dissolved in
2HCCl3. High-resolution mass spectra were recorded on an Agilent ESI-
TOF mass spectrometer at The Scripps Research Institute (La Jolla, CA).
A Shimadzu HPLC system was used to analyze the purity of target
241
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
molecules (Table S4 of the Supporting Information). General synthetic
methods are described first and designated by letters. Then compound
syntheses and characterizations are listed numerically. The purity of
compounds used in the biological assays was >95% as determined by
reversed-phase HPLC (Table S4 of the Supporting Information lists
actual purities). The purity of 8 was 98% as determined by reversed-phase
HPLC analysis. The SHP residue numbering used in the model and
protein followed that given for human SHP protein (identifier Q15466)
in the UniProtKB protein database; thus, SHP model residue numbers6
were increased by 13.
General Method A: Protection of Phenols 18 and 20 as Benzyl
Ethers 19 and 21, Respectively. To a solution of the phenol
(1 mmol) in acetone (3 mL) were added K2CO3 (1.13 mmol) and
BnBr (1.05 mmol). This mixture was stirred at reflux temperature
under argon. After removal of acetone at reduced pressure, 1 N HCl
was added. This mixture was extracted (EtOAc). The residue obtained
upon concentration of the extract was chromatographed to give the
benzyl ether.
General Method H: Hydrolysis of Ethyl Cinnamates 40, 42, and
44 to Cinnamic Acids 46, 48, and 51, Respectively, Using Aqueous
Sodium Hydroxide. To a stirred solution of the cinnamate (1 mmol)
in MeOH (10 mL) was added 5 M aqueous NaOH (5 mmol). This
mixture was heated at reflux under argon, cooled to room temperature,
acidified with 2 N HCl, and extracted (EtOAc). The extract was
washed (brine) and dried. The yellow solid residue obtained by
concentration was chromatographed or washed (hexane and CH2Cl2
or CHCl3) to afford the cinnamic acid.
General Method I: Hydrolysis of Cinnamates 41, 43, 45, and 65
to Cinnamic Acids 47, 50, 52, and 66, Respectively, Using Aqueous
Lithium Hydroxide. To a stirred solution of the ethyl cinnamate
(1 mmol) in a 50:33:17 THF/MeOH/H2O mixture (16 mL) was
added LiOH·H2O (5 mmol). This mixture was stirred under argon,
the reaction quenched with 2 N HCl to pH 1, and the mixture
extracted (EtOAc). The extract was washed (brine) and dried. The
residue obtained by concentration was chromatographed or washed
(hexane and CH2Cl2) to afford the cinnamic acid.
4-Iodo-2-trifluoromethylphenyl Bromide (17). A reported proce-
dure45 was modified with respect to workup. To a cold solution
prepared by adding 4-bromo-3-(trifluoromethyl)aniline (5.0 g, 20.8
mmol) to prewarmed concentrated sulfuric acid (5 mL) and H2O
(18.5 mL) was added with vigorous stirring and cooling (ice−salt
bath) a solution of NaNO2 (1.56 g, 21.9 mmol) in H2O (10 mL) over
a 30 min period. The mixture was stirred for 28 min before a solution
of KI (4.74 g, 28 mmol) in H2O (3.5 mL) and copper powder
(approximately 3 mg) were added. The reaction mixture was allowed
to warm to room temperature via removal of the ice bath, stirred for
1.3 h, and then heated at 101 °C for 1 h, cooled, and extracted with
EtOAc (280 mL). Extracts were washed (water and brine) and dried.
After solvent removal at reduced pressure, the residue was chromato-
graphed (hexane) to give 4.8 g (66%) of 17 as a white solid: mp 72−
73 °C (lit.45 78−80 °C); 1H NMR δ 7.46 (d, J = 8.4 Hz, 1H, 3-ArH),
7.73 (dd, J = 8.4, 2.1 Hz, 1H, 4-ArH), 8.0 (d, J = 2.1 Hz, 1H, 6-ArH).
2-Benzyloxy-5-bromobenzonitrile (19) (method A). Compound
18 (1.31 g, 6.60 mmol) on reaction for 23 h, workup, and
chromatography (10 to 20% EtOAc/hexane) gave 1.99 g (100%) of
19 as an off-white solid: mp 72−74 °C; IR 2907, 2229, 1486, 1282
General Method B: Coupling of Ethyl Acrylate with Aryl Halides
17, 19, and 21 To Afford Ethyl Cinnamates 25, 26, and 29,
Respectively. A solution of the aryl halide (1.0 mmol), ethyl acrylate
(1.3 mmol), palladium(II) acetate (0.015 mmol), and tri(o-tolyl)-
phosphine (0.04 mmol) in triethylamine (1.1 mL) was stirred while
being heated and then cooled to room temperature, and the reaction
was quenched with 1 N HCl. The resulting suspension was extracted
(EtOAc). The extract was washed (brine) and dried. After solvent
removal at reduced pressure, the residue was purified on silica gel to
give the cinnamate ester.
General Method C: Ethyl Cinnamates 32 and 33 by Wittig
Reaction of (Carbethoxymethylene)triphenylphosphorane with
Aryl Aldehydes 23 and 24, Respectively. A solution of aryl aldehyde
(1.0 mmol) and (carbethoxymethylene)triphenylphosphorane
(1.15 mmol) in PhMe (2.9 mL) was heated at reflux and then cooled
to room temperature. After solvent removal at reduced pressure, the
residue was chromatographed to give the cinnamate.
General Method D: Debenzylation of 26, 29, 35−39, 53−55, 64,
and 67 Using Boron Tribromide. To a stirred solution of the benzyl
ether (1.0 mmol) in CH2Cl2 (8 mL) at −78 °C under argon was
slowly added 1.0 M boron tribromide (3.0 mmol) in CH2Cl2 (3.0
mL). The mixture was stirred at −78 °C for 2 h, the reaction quenched
with water (15 mL), and the mixture extracted (EtOAc). The extract
was washed (brine) and dried. After solvent removal at reduced
pressure, flash chromatography of the residue yielded the phenol.
General Method E: Conversion of Phenols 27 and 30 to Triflates
28 and 31, Respectively. To a stirred solution of the phenol
(1.0 mmol) and pyridine (redistilled from KOH, 2.25 mmol) in
CH2Cl2 (2.5 mL) under argon at 0 °C was slowly added Tf2O (1.35
mmol). The resulting solution was allowed to warm to room
temperature while being stirred, the reaction quenched with 1 N
HCl, and the mixture extracted (CH2Cl2). The extract was washed
(brine), dried, and concentrated at reduced pressure. The residue was
purified by chromatography to afford the triflate.
General Method F: Diaryl Coupling of 3-(1-Adamantyl)-4-
benzyloxyphenylboronic Acid (34) with Aryl Halides 25, 32, and
33. To a solution of 34 (1.2 mmol) and aryl halide (1.0 mmol) in
degassed DME (6.7 mL) were added tetrakis(triphenylphosphine)
palladium (0.12 mmol) and 2 M Na2CO3 (degassed, 1.3 mL). The
reaction mixture was heated at reflux, cooled to room temperature,
diluted with EtOAc, washed (H2O and brine), and dried. After solvent
removal at reduced pressure, flash chromatography on silica gel yielded
the coupling product.
1
cm−1; H NMR δ 5.23 (s, 2H, PhCH2), 6.92 (d, J = 9.0 Hz, 1H, 3-
ArH), 7.33−7.50 (m, 5H, C6H5), 7.61 (dd, J = 9.0, 2.4 Hz, 1H, 4-
ArH), 7.71 (d, J = 2.4 Hz, 1H, 6-ArH); HRMS calcd for C14H10BrNO
[M + Na]+ 309.9838, found 309.8840.
Methyl 2-Benzyloxy-5-bromobenzoate (21) (method A). Methyl
5-bromo-2-hydroxybenzoate (20) (1.50 g, 6.17 mmol) on reaction for
19 h, workup, and chromatography (5 to 9% EtOAc/hexane) gave
1.98 g (100%) of 21 as a light yellow liquid: IR 2909, 1731, 1486, 1242
cm−1; 1H NMR δ 3.94 (s, 3H, CH3), 5.20 (s, 2H, ArCH2), 6.93 (d, J =
9.0 Hz, 1H, 3-ArH), 7.31−7.52 (m, 5H, C6H5), 7.55 (dd, J = 9.0, 2.7
Hz, 1H, 4-ArH), 7.97 (d, J = 2.7 Hz, 1H, 6-ArH); HRMS calcd for
C15H13BrO3 [M + Na]+ 342.9940, found 342.9945.
4-Bromo-3-methoxybenzaldehyde (23). A method different from
that reported in ref 46 was used. To a solution of benzonitrile 22 (1.04
g, 4.90 mmol) in anhydrous PhMe (14 mL) at −78 °C was added 1 M
DIBAL (7.35 mmol) in CH2Cl2 (7.40 mL). The reaction mixture was
stirred at −78 °C for 0.5 h, warmed to room temperature, stirred for
an additional 4.25 h, the reaction quenched with MeOH (4 mL), and
the mixture stirred for 0.5 h. After addition of 10% H2SO4 (45 mL),
the resulting solution was stirred for 1.75 h and then extracted with
EtOAc (100 mL, 50 mL). After solvent removal at reduced pressure,
the residue was chromatographed (10 to 16% EtOAc/hexane) to give
876 mg (83%) of 23 as a white solid: mp 70−72 °C (lit.46 74−76 °C);
1
IR 2960, 1701, 1582, 1042 cm−1; H NMR δ 4.01 (s, 3H, CH3), 7.36
General Method G: Coupling of Aryl Boronic Acid 34 to Aryl
Triflates 28 and 31. To a stirred solution of 34 (1.17 mmol) and the
aryl triflate (1.0 mmol) in degassed DME (8 mL) were added LiCl
(2.6 mmol), tetrakis(triphenylphosphine)palladium (0.12 mmol), and
2 M Na2CO3 (degassed, 1.2 mL). The reaction mixture was heated at
reflux, cooled to room temperature, then diluted with EtOAc, washed
(H2O and brine), and dried. After concentration of the organic layer at
reduced pressure, flash chromatography on silica gel yielded the diaryl
coupling product.
(dd, J = 1.8, 7.8 Hz, 1H, 6-ArH), 7.43 (d, J = 1.8 Hz, 1H, 2-ArH), 7.77
(d, J = 7.8 Hz, 1H, 5-ArH), 9.98 (s, 1H, CHO).
Ethyl (E)-4-Bromo-3-(trifluoromethyl)cinnamate (25) (method
B). 4-Iodo-2-trifluoromethylphenyl bromide (17) (1.0 g, 2.85 mmol)
after coupling at 115 °C for 3 h, workup, and chromatography (10%
EtOAc/hexane) produced 738 mg (80%) of 25 as a yellow solid: mp
1
92−94 °C; IR 2917, 1704, 1362, 1174 cm−1; H NMR δ 1.38 (t, J =
7.5 Hz, 3H, OCH2CH3), 4.32 (q, J = 7.5 Hz, 2H, OCH2CH3), 6.53 (d,
242
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
1
J = 15.9 Hz, 1H, CHCHCO), 7.55 (d, J = 8.1 Hz, 1H, 6-ArH), 7.67
(d, J = 15.9 Hz, 1H, CHCHCO), 7.77 (d, J = 8.1 Hz, 1H, 5-ArH),
7.85 (s, 1H, 2-ArH); HRMS calcd for C12H10BrF3O2 [M + H]+
322.9889, found 322.9897.
mp 58−60 °C; IR 2988, 1709, 1484, 1177 cm−1; H NMR δ 1.36 (t,
J = 7.2 Hz, 3H, OCH2CH3), 3.94 (s, 3H, OCH3), 4.30 (q, J = 7.2 Hz,
2H, OCH2CH3), 6.54 (d, J = 15.9 Hz, 1H, CHCHCO), 7.01 (d, J =
8.1 Hz, 1H, 6-ArH), 7.03 (s, 1H, 2-ArH), 7.56 (d, J = 8.1 Hz, 1H,
5-ArH), 7.64 (d, J = 15.9 Hz, 1H, CHCHCO); HRMS calcd for
C12H13BrO3 [M + H]+ 285.0121, found 285.0128.
Ethyl (E)-4-Benzyloxy-3-cyanocinnamate (26) (method B). 2-Benzyl-
oxy-5-bromobenzonitrile (19) (1.97 g, 6.84 mmol) on coupling at
94 °C for 18 h, workup, and chromatography (5 to 33% EtOAc/
hexane) produced 2.00 g (95%) of 26 as an off-white solid: mp 115−
117 °C; IR 2924, 2230, 1707, 1501, 1268 cm−1; 1H NMR δ 1.38 (t, J =
6.9 Hz, 3H, OCH2CH3), 4.32 (q, J = 6.9 Hz, 2H, OCH2CH3), 5.29 (s,
2H, CH2), 6.36 (d, J = 15.9 Hz, 1H, CHCHCO), 7.05 (d, J = 9.0
Hz, 1H, 5-ArH), 7.33−7.51 (m, 5H, C6H5), 7.59 (d, J = 15.9 Hz, 1H,
CHCHCO), 7.67 (d, J = 9.0 Hz, 1H, 6-ArH), 7.77 (s, 1H, 2-ArH);
HRMS calcd for C19H17NO3 [M + H]+ 308.1281, found 308.1285.
Ethyl (E)-3-Cyano-4-hydroxycinnamate (27) (method D). Ethyl
(E)-4-benzyloxy-3-cyanocinnamate (26) (1.0 g, 3.25 mmol) after
reaction for 2 h, workup, and chromatography (33 to 50% EtOAc/
hexane) gave 0.7 g (99%) of 27 as an off-white solid: mp 198−200 °C;
IR 3302, 2903, 2228, 1711, 1510 cm−1; 1H NMR δ 1.37 (t, J = 7.5 Hz,
3H, OCH2CH3), 4.30 (q, J = 7.5 Hz, 2H, OCH2CH3), 6.34 (s, 1H,
OH), 6.37 (d, J = 16.2 Hz, 1H, CHCHCO), 7.06 (d, J = 8.4 Hz, 1H,
5-ArH), 7.59 (d, J = 16.2 Hz, 1H, CHCHCO), 7.67 (d, J = 8.4 Hz,
1H, 6-ArH), 7.69 (s, 1H, 2-ArH); HRMS calcd for C12H11NO3 [M +
H]+ 218.0812, found 218.0811.
Ethyl (E)-3-Cyano-4-(trifluoromethylsulfonyloxy)cinnamate (28)
(method E). Cinnamate 27 (685 mg, 3.15 mmol) after reaction for
24 h, workup, and chromatography (16.7% EtOAc/hexane) yielded
1.04 g (94%) of 28 as a white solid: mp 80−82 °C; IR 2903, 2226,
1705, 1236 cm−1; 1H NMR δ 1.37 (t, J = 7.2 Hz, 3H, OCH2CH3), 4.31
(q, J = 7.2 Hz, 2H, OCH2CH3), 6.51 (d, J = 16.2 Hz, 1H, CH
CHCO), 7.53 (d, J = 8.4 Hz, 1H, 5-ArH), 7.63 (d, J = 16.2 Hz, 1H,
CHCHCO), 7.85 (dd, J = 8.4, 2.1 Hz, 1H, 6-ArH), 7.90 (d, J = 2.1
Hz, 1H, 2-ArH); HRMS calcd for C13H10F3NO5S [M + H]+ 350.0305,
found 350.0308.
Ethyl (E)-4-Bromo-3-nitrocinnamate (33) (method C). Benzalde-
hyde 24 (3.00 g, 13.03 mmol) after reaction for 18.5 h, workup, and
chromatography (14 to 33% EtOAc/hexane) followed by crystal-
lization (hexane) produced 3.36 g (86%) of 33 as a pale yellow solid:
mp 132−134 °C; IR 2981, 1714, 1530, 1181 cm−1; 1H NMR δ 1.38 (t,
J = 7.2 Hz, 3H, OCH2CH3), 4.31 (q, J = 7.2 Hz, 2H, OCH2CH3), 6.54
(d, J = 16.2 Hz, 1H, CHCHCO), 7.58 (dd, J = 2.1, 8.4 Hz, 1H, 6-
ArH), 7.65 (d, J = 16.2 Hz, 1H, CHCHCO), 7.79 (d, J = 8.4 Hz,
1H, 5-ArH), 8.0 (d, J = 2.1 Hz, 1H, 2-ArH); HRMS calcd for
C11H10BrNO4 [M + H]+ 299.9866, found 299.9865.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-benzyloxyphenyl]-3-(tri-
fluoromethyl)cinnamate (35) (method F). Aryl halide 25 (194 mg,
0.6 mmol) on coupling for 18 h, workup, and chromatography
(6 to 11% EtOAc/hexane) produced 319 mg (94%) of 35 as an off-
white solid: mp 107−109 °C; IR 2903, 1713, 1124, 1176 cm−1; H
1
NMR δ 1.39 (t, J = 7.2 Hz, 3H, OCH2CH3), 1.75 (bs, 6H, AdCH2),
2.06 (bs, 3H, AdCH), 2.18 (bs, 6H, AdCH2), 4.32 (q, J = 7.2 Hz, 2H,
OCH2CH3), 5.19 (s, 2H, ArCH2), 6.55 (d, J = 16.2 Hz, 1H, CH
CHCO), 7.01 (d, J = 8.1 Hz, 1H, 5′-ArH), 7.17 (dd, J = 8.1, 2.1 Hz,
1H, 6′-ArH), 7.25 (d, J = 2.1 Hz, 1H, 2′-ArH), 7.33−7.49 (m, 4H, 3H
from C6H5, 5-ArH), 7.54−7.57 (m, 2H, ArH), 7.71 (d, J = 7.8 Hz, 1H,
6-ArH), 7.77 (d, J = 16.2 Hz, 1H, CHCHCO), 7.90 (s, 1H, 2-ArH);
HRMS calcd for C35H35F3O3 [M + H]+ 561.2611, found 561.2635.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-benzyloxyphenyl]-3-cyanocin-
namate (36) (method G). Aryl triflate 28 (349 mg, 1.00 mmol) on
coupling for 45 h, workup, and chromatography (9% EtOAc/hexane)
produced 310 mg (60%) of 36 as an off-white solid: mp 65−67 °C; IR
2904, 2230, 1711, 1483, 1230 cm−1; H NMR δ 1.39 (t, J = 7.2 Hz,
1
3H, OCH2CH3), 1.77 (bs, 6H, AdCH2), 2.08 (bs, 3H, AdCH), 2.23
(bs, 6H, AdCH2), 4.33 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.22 (s, 2H,
CH2), 6.53 (d, J = 16.2 Hz, 1H, CHCHCO), 7.08 (d, J = 8.4 Hz,
1H, 5′-ArH), 7.35−7.58 (m, 5H, C6H5), 7.47 (dd, J = 8.4, 2.1 Hz, 1H,
6′-ArH), 7.52 (d, J = 2.1 Hz, 1H, 2′-ArH), 7.58 (d, J = 8.4 Hz, 1H, 5-
ArH), 7.71 (d, J = 16.2 Hz, 1H, CHCHCO), 7.78 (d, J = 8.4 Hz,
1H, 6-ArH), 7.90 (s, 1H, 2-ArH); HRMS calcd for C35H35NO3 [M +
H]+ 518.2690, found 518.2686.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-benzyloxyphenyl]-3-carbome-
thoxycinnamate (37) (method G). Aryl triflate 31 (382 g, 1.0
mmol) on coupling for 22 h, workup, and chromatography (10 to 23%
EtOAc/hexane) produced 546 mg (99%) of 37 as a white solid: mp
55−57 °C; IR 2905, 1712, 1479 cm−1; 1H NMR δ 1.38 (t, J = 7.2 Hz,
3H, OCH2CH3), 1.75 (bs, 6H, AdCH2), 2.08 (bs, 3H, AdCH), 2.18
(bs, 6H, AdCH2), 3.71 (s, 3H, CH3), 4.31 (q, J = 7.2 Hz, 2H,
OCH2CH3), 5.21 (s, 2H, ArCH2), 6.53 (d, J = 15.9 Hz, 1H, CH
CHCO), 7.02 (d, J = 8.1 Hz, 1H, 5′-ArH), 7.20 (d, J = 8.1 Hz, 1H, 6′-
ArH), 7.22 (s, 1H, 2′-ArH), 7.46 (d, J = 7.8 Hz, 1H, 5-ArH), 7.34−
7.57 (m, 5H, C6H5), 7.67 (d, J = 7.8 Hz, 1H, 6-ArH), 7.74 (d, J = 15.9
Hz, 1H, CHCHCO), 7.93 (s, 1H, 2-ArH); HRMS calcd for
C36H38O5 [M + H]+ 551.2792, found 551.2789.
Ethyl (E)-4-Benzyloxy-3-carbomethoxycinnamate (29) (method
B). 4-Bromobenzoate 21 (1.10 g, 3.43 mmol) on coupling at 105 °C
for 17 h, workup, and chromatography (5 to 22% EtOAc/hexane)
produced 1.15 g (98%) of 29 as a pale yellow solid: mp 84−85 °C; IR
1
2902, 1729, 1689, 1501, 1269 cm−1; H NMR δ 1.36 (t, J = 7.2 Hz,
3H, OCH2CH3), 3.95 (s, 3H, CH3), 4.29 (q, J = 7.2 Hz, 2H,
OCH2CH3), 5.26 (s, 2H, ArCH2), 6.39 (d, J = 15.9 Hz, 1H, CH
CHCO), 7.05 (d, J = 8.4 Hz, 1H, 5-ArH), 7.31−7.52 (m, 5H, C6H5),
7.60 (dd, J = 8.4, 2.4 Hz, 1H, 6-ArH), 7.67 (d, J = 15.9 Hz, 1H, CH
CHCO), 8.05 (d, J = 2.4 Hz, 1H, 2-ArH); HRMS calcd for C20H2oO5
[M + H]+ 341.1383, found 341.1383.
Ethyl (E)-3-Carbomethoxy-4-hydroxycinnamate (30) (method
D). Cinnamate 29 (623 mg, 1.83 mmol) after reaction for 2 h,
workup, and chromatography (10 to 33% EtOAc/hexane) gave 456
mg (99%) of 30 as a pale yellow solid: mp 62−64 °C; IR 3247, 2941,
1707, 1677 cm−1; 1H NMR δ 1.36 (t, J = 7.2 Hz, 3H, OCH2CH3), 4.01
(s, 3H, CH3), 4.29 (q, J = 7.2 Hz, 2H, OCH2CH3), 6.36 (d, J = 16.2
Hz, 1H, CHCHCO), 7.03 (d, J = 8.7 Hz, 1H, 5-ArH), 7.64 (d, J =
16.2 Hz, 1H, CHCHCO), 7.67 (d, J = 8.7 Hz, 1H, 6-ArH), 8.04 (s,
1H, 2-ArH), 11.0 (s, 1H, OH); HRMS calcd for C13H15O5 [M + H]+
251.0914, found 251.0918.
Ethyl (E)-3-Carbomethoxy-4-trifluoromethylsulfonyloxycinna-
mate (31) (method E). Cinnamate 30 (434 mg, 1.74 mmol) after
reaction for 23 h, workup, and chromatography (10 to 16% EtOAc/
hexane) yielded 641 mg (96%) of 31 as a white solid: mp 51−52 °C;
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-benzyloxyphenyl]-3-methoxy-
cinnamate (38) (method F). Aryl halide 32 (119 mg, 0.41 mmol)
on coupling for 22.6 h, workup, and chromatography (11 to 12%
EtOAc/hexane) produced 192 mg (88%) of 38 as a cream solid: mp
1
IR 2913, 1731, 1431, 1213 cm−1; H NMR δ 1.39 (t, J = 7.2 Hz, 3H,
1
62−64 °C; IR 2903, 1706, 1487, 1163 cm−1; H NMR δ 1.39 (t, J =
OCH2CH3), 4.03 (s, 3H, CH3), 4.33 (q, J = 7.2 Hz, 2H, OCH2CH3),
6.54 (d, J = 16.2 Hz, 1H, CHCHCO), 7.36 (d, J = 8.4 Hz, 1H,
5-ArH), 7.69 (d, J = 16.2 Hz, 1H, CHCHCO), 7.67 (dd, J = 8.4, 2.4
Hz, 1H, 6-ArH), 8.27 (d, J = 2.4 Hz, 1H, 2-ArH); HRMS calcd for
C14H13F3O7S [M + H]+ 383.0407, found 383.0417.
Ethyl (E)-4-Bromo-3-methoxycinnamate (32) (method C). Benzal-
dehyde 23 (876 mg, 4.07 mmol) after reaction for 19 h, workup, and
chromatography (10 to 16% EtOAc/hexane) produced 1.09 g (94%)
of a cream solid (90% E isomer), which after crystallization (hexane)
yielded 712 mg (61%) of the pure E isomer of 32 as a white solid:
7.2 Hz, 3H, OCH2CH3), 1.76 (bs, 6H, AdCH2), 2.08 (bs, 3H, AdCH),
2.21 (bs, 6H, AdCH2), 3.89 (s, 3H, OCH3), 4.31 (q, J = 7.2 Hz, 2H,
OCH2CH3), 5.19 (s, 2H, ArCH2), 6.49 (d, J = 15.9 Hz, 1H, CH
CHCO), 7.01 (d, J = 8.4 Hz, 1H, 5′-ArH), 7.14 (s, 1H, 2′-ArH), 7.22
(dd, J = 8.4, 2.1 Hz, 1H, 6′-ArH), 7.31−7.58 (m, 8H, C6H5, 2,5,6-
ArH), 7.74 (d, J = 15.9 Hz, 1H, CHCHCO); HRMS calcd for
C35H38O4 [M + H]+ 523.2843, found 523.2842.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-benzyloxyphenyl]-3-nitrocinna-
mate (39) (method F). Aryl halide 33 (1.00 g, 3.33 mmol) on coupling
for 16 h, workup, and chromatography (9 to 33% EtOAc/hexane)
243
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
produced 1.66 g (93%) of 39 as a yellow solid: mp 170−171 °C; IR 2902,
1719, 1532, 1179 cm−1; 1H NMR δ 1.39 (t, J = 7.2 Hz, 3H, OCH2CH3),
1.75 (bs, 6H, AdCH2), 2.07 (bs, 3H, AdCH), 2.17 (bs, 6H, AdCH2), 4.33
(q, J = 7.2 Hz, 2H, OCH2CH3), 5.19 (s, 2H, ArCH2), 6.55 (d, J = 16.2
Hz, 1H, CHCHCO), 7.03 (d, J = 8.4 Hz, 1H, 5′-ArH), 7.19 (dd, J =
8.4, 2.1 Hz, 1H, 6′-ArH), 7.24 (d, J = 2.1 Hz, 1H, 2′-ArH), 7.31−7.57 (m,
6H, C6H5, 5-ArH), 7.72 (d, J = 16.2 Hz, 1H, CHCHCO), 7.74 (dd, J =
7.5, 1.5 Hz, 1H, 6-ArH), 7.94 (d, J = 1.5 Hz, 1H, 2-ArH); HRMS calcd for
C34H35NO5 [M + H]+ 538.2588, found 538.2584.
(m, 3H, 2′,6′-ArH, 5-ArH), 7.27 (dd, J = 2.1, 8.7 Hz, 1H, 6-ArH), 7.32
(d, J = 2.1 Hz, 1H, 2-ArH), 7.70 (d, J = 15.9 Hz, 1H, CHCHCO);
HRMS calcd for C27H30O4 [M + H]+ 419.2217, found 419.2216.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-nitrocinna-
mate (45) (method D). Cinnamate 39 (52 mg, 0.097 mmol) after
treatment for 2 h, workup, and chromatography (16 to 20% EtOAc/
hexane) gave 43 mg (100%) of 45 as a yellow solid: mp 236−238 °C;
IR 3271, 2903, 1688, 1533, 1203 cm−1; 1H NMR δ 1.39 (t, J = 7.2 Hz,
3H, OCH2CH3), 1.81 (bs, 6H, AdCH2), 2.12 (bs, 3H, AdCH), 2.15
(bs, 6H, AdCH2), 4.32 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.10 (br s, 1H,
OH), 6.55 (d, J = 15.9 Hz, 1H, CHCHCO), 6.73 (d, J = 8.1 Hz, 1H,
5′-ArH), 7.07 (dd, J = 1.8, 8.1 Hz, 1H, 6′-ArH), 7.2 (d, J = 2.1 Hz, 1H,
2′-ArH), 7.51 (d, J = 7.8 Hz, 1H, 5-ArH), 7.73 (d, J = 15.9 Hz, 1H,
CHCHCO), 7.73 (dd, J = 7.8, 1.5 Hz, 1H, 6-ArH), 7.93 (d, J = 1.5
Hz, 1H, 2-ArH); HRMS calcd for C27H29NO5 [M + H]+ 448.2118,
found 448.2117.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-(tri-
fluoromethyl)cinnamate (40) (method D). Cinnamate 35 (306 mg,
0.54 mmol) after reaction for 2 h, workup, and chromatography
(9 to 14% EtOAc/hexane) gave 240 mg (93%) of 40 as a pale yellow
1
solid: mp 186−187 °C; IR 3408, 2899, 1695, 1324, 1175 cm−1; H
NMR δ 1.39 (t, J = 7.2 Hz, 3H, OCH2CH3), 1.81 (bs, 6H, AdCH2),
2.11 (bs, 3H, AdCH), 2.16 (bs, 6H, AdCH2), 4.32 (q, J = 7.2 Hz, 2H,
OCH2CH3), 4.99 (s, 1H, OH), 6.55 (d, J = 16.2 Hz, 1H, CH
CHCO), 6.72 (d, J = 8.1 Hz, 1H, 5′-ArH), 7.06 (dd, J = 8.1, 1.8 Hz,
1H, 6′-ArH), 7.20 (d, J = 1.8 Hz, 1H, 2′-ArH), 7.40 (d, J = 8.1 Hz, 1H,
5-ArH), 7.71 (d, J = 8.1 Hz, 1H, 6-ArH), 7.76 (d, J = 16.2 Hz, 1H,
CHCHCO), 7.89 (s, 1H, 2-ArH); HRMS calcd for C28H29F3O3 [M
+ H]+ 471.2141, found 471.2151.
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-(trifluoromethyl)-
cinnamic Acid (46) (method H). Cinnamate 40 (207 mg, 0.44 mmol)
after hydrolysis for 35 min, workup, and chromatography (1:5:0 to
5:5:2 EtOAc/hexane/MeOH) afforded 164 mg (84%) of 46 as a pale
yellow solid: mp 180−182 °C; IR 3239, 2900, 1686, 1317, 1140 cm−1;
1H NMR (CD3OD) δ 1.84 (bs, 6H, AdCH2), 2.07 (bs, 3H, AdCH),
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-cyanocinna-
mate (41) (method D). Cinnamate 36 (147 mg, 0.28 mmol) after
treatment for 2 h, workup, and chromatography (8 to 20% EtOAc/
hexane) gave 102 mg (84%) of 41 as a cream solid: mp 220−222 °C;
2.20 (m, 6H, AdCH2), 6.63 (d, J = 15.9 Hz, 1H, CHCHCO), 6.79
(d, J = 8.4 Hz, 1H, 5′-ArH), 6.99 (dd, J = 8.1, 1.6 Hz, 1H, 6′-ArH),
7.12 (d, J = 1.6 Hz, 1H, 2′-ArH), 7.43 (d, J = 7.8 Hz, 1H, 5-ArH), 7.76
(d, J = 15.9 Hz, 1H, CHCHCO), 7.88 (d, J = 7.8 Hz, 1H, 6-ArH),
7.96 (s, 1H, 2-ArH); HRMS calcd for C26H25F3O3 [M + H]+
443.1828, found 443.1840.
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-cyanocinnamic
Acid (47) (method I). Cinnamate 41 (98 mg, 0.23 mmol) after
hydrolysis for 5 h, workup, and washing (hexane and CH2Cl2)
afforded 79 mg (86%) of 47 as a white solid: mp 294−296 °C; IR
3362, 2904, 2230, 1694, 1417, 1253 cm−1; 1H NMR (CD3OD) δ 1.86
(bs, 6H, AdCH2), 2.09 (bs, 3H, AdCH), 2.25 (bs, 6H, AdCH2), 6.64
(d, J = 16.2 Hz, 1H, CHCHCO), 6.88 (d, J = 8.1 Hz, 1H, 5′-ArH),
7.30 (d, J = 8.1 Hz, 1H, 6′-ArH), 7.46 (s, 1H, 2′-ArH), 7.62 (d, J = 8.4
Hz, 1H, 5-ArH), 7.72 (d, J = 16.2 Hz, 1H, CHCHCO), 7.94 (d, J =
8.4 Hz, 1H, 6-ArH), 8.06 (s, 1H, 2-ArH); HRMS calcd for C26H25NO3
[M + H]+ 400.1907, found 400.1903.
1
IR 3373, 2902, 2233, 1693, 1256 cm−1; H NMR (acetone-d6) δ 1.33
(t, J = 7.2 Hz, 3H, OCH2CH3), 1.84 (bs, 6H, AdCH2), 2.09 (bs, 3H,
AdCH), 2.27 (bs, 6H, AdCH2), 4.27 (q, J = 7.2 Hz, 2H, OCH2CH3),
6.75 (d, J = 16.2 Hz, 1H, CHCHCO), 7.0 (d, J = 8.4 Hz, 1H, 5′-
ArH), 7.38 (dd, J = 8.4, 2.1 Hz, 1H, 6′-ArH), 7.53 (d, J = 2.1 Hz, 1H,
2′-ArH), 7.69 (d, J = 8.1 Hz, 1H, 5-ArH), 7.76 (d, J = 16.2 Hz, 1H,
CHCHCO), 8.06 (dd, J = 8.1, 1.8 Hz, 1H, 6-ArH), 8.20 (d, J = 1.8
Hz, 1H, 2-ArH), 8.83 (s, 1H, OH); HRMS calcd for C28H29NO3 [M +
H]+ 428.2220, found 428.2214.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-carbome-
thoxycinnamate (42) (method D). Cinnamate 37 (325 mg, 0.59
mmol) after reaction for 2 h, workup, and chromatography (10 to 33%
EtOAc/hexane) gave 271 mg (100%) of 42 as a pale yellow solid: mp
1
71−73 °C; IR 3340, 2904, 1709, 1255 cm−1; H NMR δ 1.39 (t, J =
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-(methoxycarbonyl)-
cinnamic Acid (48) (method H). Cinnamate 42 (231 mg, 0.48 mmol)
after hydrolysis for 1.5 h, workup, and washing (hexane and CH2Cl2)
afforded 190 mg (91%) of 48 as a white solid: mp 259−262 °C; IR
7.2 Hz, 3H, OCH2CH3), 1.81 (bs, 6H, AdCH2), 2.12 (bs, 3H, AdCH),
2.16 (bs, 6H, AdCH2), 3.72 (s, 3H, CH3), 4.31 (q, J = 7.2 Hz, 2H,
OCH2CH3), 4.97 (s, 1H, OH), 6.53 (d, J = 15.9 Hz, 1H, CH
CHCO), 6.72 (d, J = 8.1 Hz, 1H, 5′-ArH), 7.10 (dd, J = 8.1, 2.1 Hz,
1H, 6′-ArH), 7.18 (d, J = 2.1 Hz, 1H, 2′-ArH), 7.45 (d, J = 7.8 Hz, 1H,
5-ArH), 7.67 (dd, J = 7.8, 1.8 Hz, 1H, 6-ArH), 7.74 (d, J = 15.9 Hz,
1H, CHCHCO), 7.93 (d, J = 1.8 Hz, 1H, 2-ArH); HRMS calcd for
C29H33O5 [M + H]+ 461.2328, found 461.2321.
1
3347, 2906, 1712, 1695, 1253 cm−1; H NMR (CD3OD) δ 1.84 (bs,
6H, AdCH2), 2.08 (bs, 3H, AdCH), 2.19 (bs, 6H, AdCH2), 3.69 (s,
3H, CH3), 6.55 (d, J = 15.9 Hz, 1H, CHCHCO), 6.79 (d, J = 8.1
Hz, 1H, 5′-ArH), 7.04 (dd, J = 8.1, 2.4 Hz, 1H, 6′-ArH), 7.08 (d, J =
2.4 Hz, 1H, 2′-ArH), 7.48 (d, J = 8.1 Hz, 1H, 5-ArH), 7.74 (d, J = 15.9
Hz, 1H, CHCHCO), 7.80 (dd, J = 8.1, 1.4 Hz, 1H, 6-ArH), 7.86 (d,
J = 1.4 Hz, 1H, 2-ArH); HRMS calcd for C27H28O5 [M + H]+
433.2009, found 433.2011.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-methoxycin-
namate (43) (method D). Cinnamate 38 (80 mg, 0.15 mmol) after
reaction for 2 h, workup, and chromatography (16 to 33% EtOAc/
hexane) gave 66 mg (100%) of 43 as a cream solid: mp 197−199 °C;
IR 3332, 2902, 1686, 1493, 1183 cm−1; 1H NMR δ 1.39 (t, J = 7.5 Hz,
3H, OCH2CH3), 1.81 (bs, 6H, AdCH2), 2.11 (bs, 3H, AdCH), 2.19
(bs, 6H, AdCH2), 3.87 (s, 3H, OCH3), 4.32 (q, J = 7.5 Hz, 2H,
OCH2CH3), 5.37 (br s, 1H, OH), 6.49 (d, J = 15.9 Hz, 1H, CH
CHCO), 6.74 (d, J = 7.8 Hz, 1H, 5′-ArH), 7.12 (d, J = 1.5 Hz, 1H, 2′-
ArH), 7.21 (dd, J = 1.5, 7.8 Hz, 1H, 6′-ArH), 7.31 (dd, J = 2.1, 8.1 Hz,
1H, 6-ArH), 7.35 (d, J = 8.1 Hz, 1H, 5-ArH), 7.42 (d, J = 2.1 Hz, 1H,
2-ArH), 7.74 (d, J = 15.9 Hz, 1H, CHCHCO); HRMS calcd for
C28H32O4 [M + H]+ 433.2373, found 433.2370.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-hydroxycin-
namate (44) (method D). Cinnamate 38 (85 mg, 0.16 mmol) after
reaction at −78 °C for 0.33 h and at room temperature for 2.5 h,
workup, and chromatography (16 to 25% EtOAc/hexane) gave 43 mg
(63%) of 44 as a yellow solid: mp 193−194 °C; IR 3295, 2838, 1693,
1463, 1216 cm−1; 1H NMR δ 1.39 (t, J = 7.2 Hz, 3H, OCH2CH3), 1.81
(bs, 6H, AdCH2), 2.12 (bs, 3H, AdCH), 2.18 (bs, 6H, AdCH2), 4.30
(q, J = 7.2 Hz, 2H, OCH2CH3), 5.49 (br s, 2H, OH), 6.47 (d, J = 15.9
Hz, 1H, CHCHCO), 6.83 (d, J = 8.1 Hz, 1H, 5′-ArH), 7.12−7.22
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-carboxycinnamic
Acid (49). To a solution of 48 (61 mg, 0.14 mmol) in a THF/
MeOH/H2O mixture (3:2:1, 1.8 mL) was added lithium hydroxide
monohydrate (30 mg, 0.71 mmol). This mixture was stirred under
argon for 21 h at which time TLC indicated some starting material 48.
Therefore, the mixture was treated with MeOH (2 mL) and 5 N
aqueous NaOH (0.4 mL) at 82 °C for 24 h, the reaction quenched
with 1 N HCl (6 mL), and the mixture extracted with EtOAc (40 and
20 mL). The extract was washed (brine), dried, and concentrated at
reduced pressure. The residue was washed with hexane, CH2Cl2 (three
times), and pentane and dried to give 42 mg (71%) of 49 as a cream
1
solid: mp 274−277 °C; IR 3281, 2895, 1712, 1691, 1250 cm−1; H
NMR (CD3OD) δ 1.83 (bs, 6H, AdCH2), 2.07 (bs, 3H, AdCH), 2.20
(bs, 6H, AdCH2), 6.56 (d, J = 15.9 Hz, 1H, CHCHCO), 6.78 (d,
J = 8.1 Hz, 1H, 5′-ArH), 7.10 (d, J = 8.1 Hz, 1H, 6′-ArH), 7.20 (s, 1H,
2′-ArH), 7.45 (d, J = 7.5 Hz, 1H, 5-ArH), 7.73 (d, J = 15.9 Hz, 1H,
CHCHCO), 7.75 (d, J = 7.5 Hz, 1H, 6-ArH), 7.86 (s, 1H, 2-ArH);
HRMS calcd for C26H26O5 [M + H]+ 419.1853, found 419.1855.
244
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-methoxycinnamic
Acid (50) (method I). Cinnamate 43 (66 mg, 0.15 mmol) on
hydrolysis for 6.3 h, workup, and washing (hexane and CH2Cl2)
afforded 43 mg (70%) of 50 as a cream solid: mp 218−220 °C; IR
Hz, 1H, 2-ArH); HRMS calcd for C34H35BrO3 [M + H]+ 571.1842,
found 571.1838.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-benzyloxyphenyl]-3,6-dibromo-
cinnamate (55). To a solution of 53 (228 mg, 0.45 mmol) in
anhydrous MeCN (2 mL) at 0 °C were added, under argon, tert-butyl
nitrite (107 μL, 0.9 mmol) and CuBr2 (201 mg, 0.9 mmol). The
mixture was stirred at 0 °C for 45 min, warmed to room temperature,
stirred for 4 h, the reaction quenched with 1 N HCl (20 mL), and the
mixture extracted with EtOAc (60 and 40 mL). The residue obtained
on workup was chromatographed (3 to 5% EtOAc/hexane) to give
135 mg (46%) of 55 as a pale yellow solid: mp 74−77 °C; IR 2907,
1718, 1457, 1183 cm−1; 1H NMR δ 1.38 (t, J = 7.2 Hz, 3H,
OCH2CH3), 1.76 (bs, 6H, AdCH2), 2.08 (bs, 3H, AdCH), 2.20 (bs,
6H, AdCH2), 4.32 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.20 (s, 2H, CH2),
6.45 (d, J = 15.9 Hz, 1H, CHCHCO), 7.02 (d, J = 8.4 Hz, 1H, 5′-
ArH), 7.24 (dd, J = 8.4, 2.1 Hz, 1H, 6′-ArH), 7.34 (d, J = 2.1 Hz, 1H,
2′-ArH), 7.32−7.58 (m, 5H, ArH), 7.62 (s, 1H, 5-ArH), 7.92 (s, 1H, 2-
ArH), 8.01 (d, J = 15.9 Hz, 1H, CHCHCO); HRMS calcd for
C34H34Br2O3 [M + H]+ 649.0947, found 649.0943.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-aminocinna-
mate (56) (method D). A solution of 53 (45 mg, 0.089 mmol)
after reaction for 2.5 h, workup, and chromatography (16 to 20%
EtOAc/hexane) gave 30 mg (81%) of 56 as a light yellow solid: mp
214−216 °C; IR 3356, 2905, 1692, 1178 cm−1; 1H NMR δ 1.38 (t, J =
7.2 Hz, 3H, OCH2CH3), 1.81 (bs, 6H, AdCH2), 2.12 (bs, 3H, AdCH),
2.17 (bs, 6H, AdCH2), 3.80 (bs, 2H, NH2), 4.29 (q, J = 7.2 Hz, 2H,
OCH2CH3), 5.03 (bs, 1H, OH), 6.43 (d, J = 16.2 Hz, 1H, CH
CHCO), 6.76 (d, J = 8.1 Hz, 1H, 5′-ArH), 6.94 (d, J = 1.5 Hz, 1H, 2′-
ArH), 7.01 (dd, J = 8.1, 1.5 Hz, 1H, 6′-ArH), 7.16 (d, J = 7.5 Hz, 1H,
5-ArH), 7.19 (dd, J = 2.1, 7.5 Hz, 1H, 6-ArH), 7.32 (d, J = 2.1 Hz, 1H,
2-ArH), 7.65 (d, J = 16.2 Hz, 1H, CHCHCO); HRMS calcd for
C27H31NO3 [M + H]+ 418.2377, found 418.2372.
1
3281, 2905, 1686, 1421, 1252 cm−1; H NMR (CD3OD) δ 1.79 (bs,
6H, AdCH2), 2.03 (bs, 3H, AdCH), 2.16 (bs, 6H, AdCH2), 3.82 (s,
3H, CH3O), 6.47 (d, J = 15.9 Hz, 1H, CHCHCO), 6.70 (d, J = 8.5
Hz, 1H, 5′-ArH), 7.13 (dd, J = 8.5, 1.8 Hz, 1H, 6′-ArH), 7.20 (dd, J =
8.7, 2.4 Hz, 1H, 6-ArH), 7.22 (d, J = 1.8 Hz, 1H, 2′-ArH), 7.25 (d, J =
8.7 Hz, 1H, 5-ArH), 7.27 (d, J = 2.4 Hz, 1H, 2-ArH), 7.66 (d, J = 15.9
Hz, 1H, CHCHCO); HRMS calcd for C26H28O4 [M + H]+
405.2060, found 405.2056.
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-hydroxycinnamic
Acid (51) (method H). Cinnamate 44 (35 mg, 0.08 mmol) after
hydrolysis for 1.5 h, workup, and washing (hexane and CH2Cl2)
afforded 28 mg (85%) of 51 as a light yellow solid: mp >260 °C dec;
IR 3276, 2907, 1689, 1429, 1177 cm−1; 1H NMR (CDCl3/CD3OD) δ
1.54 (bs, 6H, AdCH2), 1.82 (bs, 3H, AdCH), 1.94 (bs, 6H, AdCH2),
6.14 (d, J = 15.9 Hz, 1H, CHCHCO), 6.83 (d, J = 7.5 Hz, 1H, 5′-
ArH), 6.84 (dd, J = 7.5, 1.8 Hz, 1H, 6′-ArH), 6.86 (d, J = 1.8 Hz, 1H,
2′-ArH), 7.03 (dd, J = 8.5, 1.8 Hz, 1H, 6-ArH), 7.04 (d, J = 8.5 Hz, 1H,
5-ArH), 7.13 (d, J = 1.8 Hz, 1H, 2-ArH), 7.38 (d, J = 15.9 Hz, 1H,
CHCHCO); HRMS calcd for C25H26O4 [M + H]+ 391.1904, found
391.1906.
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-nitrocinnamic Acid
(52) (method I). Cinnamate 45 (36 mg, 0.08 mmol) after hydrolysis
for 3.5 h, workup, and washing (hexane and CH2Cl2) afforded 32 mg
(95%) of 52 as a yellow solid: mp 285−287 °C; IR 3278, 2904, 1696,
1
1529, 1254 cm−1; H NMR (CD3OD) δ 1.84 (bs, 6H, AdCH2), 2.08
(bs, 3H, AdCH), 2.18 (m, 6H, AdCH2), 6.65 (d, J = 16.2 Hz, 1H,
CHCHCO), 6.80 (d, J = 8.1 Hz, 1H, 5′-ArH), 7.04 (dd, J = 8.1, 2.1
Hz, 1H, 6′-ArH), 7.10 (d, J = 2.1 Hz, 1H, 2′-ArH), 7.54 (d, J = 7.8 Hz,
1H, 5-ArH), 7.74 (d, J = 16.2 Hz, 1H, CHCHCO), 7.88 (dd, J =
7.8, 1.8 Hz, 1H, 6-ArH), 8.01 (d, J = 1.8 Hz, 1H, 2-ArH); HRMS calcd
for C25H25NO5 [M + H]+ 420.1805, found 420.1800.
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-aminocinnamic
Acid (57) (method I). Cinnamate 56 (27 mg, 0.065 mmol) after
hydrolysis for 4 h, workup, and washing (hexane, CH2Cl2, and pentane)
afforded 25 mg (100%) of 57 as a yellow solid: mp 240−242 °C;
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-benzyloxyphenyl]-3-aminocin-
namate (53). A suspension of 39 (538 mg, 1.00 mmol) and stannous
chloride dihydrate (1.13 g, 5.00 mmol) in EtOH (4.5 mL) was heated
at 80 °C (oil bath) for 2 h. Next, most of the solvent was removed at
reduced pressure; the mixture was cooled to room temperature, and its
pH was adjusted to 7−8 by addition of 2 N NaOH (4 mL) and 5%
NaHCO3 (10 mL). The resulting mixture was stirred for 10 min and
then extracted with EtOAc (80 mL and 2 × 50 mL). The residue
obtained after workup was purified by chromatography on silica gel
(14 to 50% EtOAc/hexane) to give 506 mg (99%) of 53 as a pale
1
IR 3360, 2902, 1687, 1253 cm−1; H NMR (CD3OD) δ 1.84 (bs,
6H, AdCH2), 2.08 (bs, 3H, AdCH), 2.21 (bs, 6H, AdCH2), 6.43 (d,
J = 15.9 Hz, 1H, CHCHCO), 6.82 (d, J = 8.4 Hz, 1H, 5′-ArH), 7.01
(dd, J = 8.4, 1.5 Hz, 1H, 6′-ArH), 7.06 (d, J = 1.5 Hz, 1H, 2′-ArH), 7.08
(d, J = 8.1 Hz, 1H, 5-ArH), 7.10 (dd, J = 8.1, 2.1 Hz, 1H, 6-ArH),
7.21 (d, J = 2.1 Hz, 1H, 2-ArH), 7.62 (d, J = 15.9 Hz, 1H, CH
CHCO); HRMS calcd for C25H27NO3 [M + H]+ 390.2064, found
390.2065.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-azidocinna-
mate (58). To a suspension of 56 (100 mg, 0.197 mmol) in 1,4-
dioxane (1.2 mL) and 2 M H2SO4 (2.5 mL) at −10 to −5 °C under
argon was added a solution of NaNO2 (30 mg, 0.43 mmol) in H2O
(140 μL). The mixture was stirred at −10 to −5 °C for 20 min before
a solution of NaN3 (43 mg, 0.65 mmol) in H2O (220 μL) was added,
and stirring was continued for 30 min. The mixture was warmed to
room temperature and diluted with EtOAc (70 mL). The organic layer
was washed (saturated NaHCO3 and brine), dried, and concentrated
at reduced pressure. The residue was chromatographed (10 to 14%
EtOAc/hexane) to give 76 mg (87%, two steps from 53) of 58 as a
light yellow solid: mp 88−91 °C; IR 3416, 2901, 2111, 1691, 1180
1
yellow solid: mp 150−151 °C; IR 3370, 2905, 1703, 1177 cm−1; H
NMR δ 1.38 (t, J = 7.2 Hz, 3H, OCH2CH3), 1.76 (bs, 6H, AdCH2),
2.08 (bs, 3H, AdCH), 2.20 (bs, 6H, AdCH2), 3.89 (bs, 2H, NH2), 4.30
(q, J = 7.2 Hz, 2H, OCH2CH3), 5.20 (s, 2H, CH2), 6.43 (d, J = 15.9
Hz, 1H, CHCHCO), 6.94 (s, 1H, 2-ArH), 7.02 (d, J = 8.1 Hz, 1H,
6-ArH), 7.05 (d, J = 8.1 Hz, 1H, 5-ArH), 7.18 (d, J = 7.8 Hz, 1H, 5′-
ArH), 7.27 (dd, J = 8.1, 2.1 Hz, 1H, 6′-ArH), 7.36 (d, J = 2.1 Hz, 1H,
2′-ArH), 7.32−7.58 (m, 5H, ArH), 7.65 (d, J = 15.9 Hz, 1H, CH
CHCO); HRMS calcd for C34H37NO3 [M + H]+ 508.2846, found
508.2847.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-benzyloxyphenyl]-3-bromocin-
namate (54). To a solution of n-hexyl nitrite (32 mg, 0.24 mmol)
in CHBr3 (0.6 mL) at 95 °C was added 53 (1.408 g, 4.64 mL) in
CHBr3 (4.5 mL) over a period of 5 min. The mixture was stirred at
100 °C for 5 h and cooled to room temperature. After most of the
CHBr3 had been removed under high vacuum, the residue was purified
on silica gel (3 to 5% EtOAc/hexane) to give 67 mg (58%) of 54 as a
1
cm−1; H NMR (CDCl3) δ 1.35 (t, J = 7.2 Hz, 3H, OCH2CH3), 1.78
(bs, 6H, AdCH2), 2.09 (bs, 3H, AdCH), 2.15 (bs, 6H, AdCH2), 3.80
(bs, 2H, NH2), 4.28 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.14 (bs, 1H,
OH), 6.46 (d, J = 16.2 Hz, 1H, CHCHCO), 6.76 (d, J = 8.4 Hz, 1H,
5′-ArH), 7.18 (dd, J = 8.4, 2.1 Hz, 1H, 6′-ArH), 7.28 (d, J = 2.1 Hz,
1H, 2′-ArH), 7.33 (bs, 3H, 2,5,6-ArH), 7.66 (d, J = 16.2 Hz, 1H, CH
CHCO); HRMS calcd for C27H29N3O3 [M + H]+ 444.2282, found
444.2281.
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-azidocinnamic
Acid (59). To a suspension of 58 (53 mg, 0.12 mmol) in Et2O
(0.1 mL) and EtOH (0.2 mL) was added 25% aqueous NaOH (50 μL,
0.42 mmol). The mixture was stirred for 6 h, acidified with 1 N HCl
(10 mL), and extracted with EtOAc (50 and 30 mL). The extract was
washed (brine), dried, and concentrated at reduced pressure. The
1
white solid: mp 159−161 °C; IR 2903, 1712, 1230 cm−1; H NMR
(CDCl3) δ 1.39 (t, J = 7.2 Hz, 3H, OCH2CH3), 1.76 (bs, 6H, AdCH2),
2.08 (bs, 3H, AdCH), 2.21 (bs, 6H, AdCH2), 4.31 (q, J = 7.2 Hz, 2H,
OCH2CH3), 5.20 (s, 2H, ArCH2), 6.48 (d, J = 15.9 Hz, 1H, CH
CHCO), 7.03 (d, J = 8.4 Hz, 1H, 5′-ArH), 7.28 (dd, J = 8.4, 2.4 Hz,
1H, 6′-ArH), 7.38 (d, J = 2.4 Hz, 1H, 2′-ArH), 7.39 (d, J = 7.8 Hz, 1H,
5-ArH), 7.51 (dd, J = 7.8, 1.5 Hz, 1H, 6-ArH), 7.32−7.57 (m, 6H,
ArH, 5-ArH), 7.67 (d, J = 15.9 Hz, 1H, CHCHCO), 7.86 (d, J = 1.5
245
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
residue was chromatographed (0:50:50 to 14:43:43 MeOH/EtOAc/
hexane) to give 33 mg (66%) of 59 as a tan solid: mp >260 °C; IR
3374, 2904, 2113, 1685, 1251 cm−1; 1H NMR (DMSO-d6) δ 1.76 (bs,
6H, AdCH2), 2.07 (bs, 3H, AdCH), 2.13 (bs, 6H, AdCH2), 6.65 (d,
J = 16.2 Hz, 1H, CHCHCO), 6.85 (d, J = 8.7 Hz, 1H, 5′-ArH), 7.19
(m, 2H, 2′,6′-ArH), 7.38 (d, J = 7.8 Hz, 1H, 5-ArH), 7.57 (d, J =
7.8 Hz, 1H, 6-ArH), 7.66 (d, J = 16.2 Hz, 1H, CHCHCO), 7.67 (s,
1H, 2-ArH), 9.58 (s, 1H, OH), 12.43 (bs, 1H, COOH); HRMS calcd
for C25H25N3O3 [M + H]+ 416.1974, found 416.1988.
6H, AdCH2), 4.31 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.25 (s, 2H,
ArCH2), 6.51 (d, J = 16.2 Hz, 1H, CHCHCO), 6.99 (s, 1H, 8H-
carbazole), 7.36−7.60 (m, 7H, C5H5, 1H, 3H-carbazole), 7.86 (d, J =
16.2 Hz, 1H, CHCHCO), 7.94 (s, 1H, 5H-carbazole), 7.97 (bs, 1H,
9H-carbazole), 7.99 (d, J = 7.8 Hz, 1H, 4H-carbazole); HRMS calcd
for C34H35NO3 [M + H]+ 506.2690, found 506.2683. 67: IR 3407,
2911, 1702, 1272 cm−1; 1H NMR δ 1.39 (t, J = 7.2 Hz, 3H,
OCH2CH3), 1.80 (bs, 6H, AdCH2), 2.12 (bs, 3H, AdCH), 2.61 (bs,
6H, AdCH2), 4.32 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.24 (s, 2H,
ArCH2), 6.53 (d, J = 15.9 Hz, 1H, CHCHCO), 7.03 (d, J = 8.4 Hz,
1H, 6H-carbazole), 7.35−7.59 (m, 7H, C5H5, 1H, 3H-carbazole), 7.85
(d, J = 16.2 Hz, 1H, CHCHCO), 7.88 (d, J = 8.4 Hz, 1H, 5H-
carbazole), 7.95 (d, J = 8.1 Hz, 1H, 4H-carbazole), 8.91 (bs, 1H, 9H-
carbazole); HRMS calcd for C34H35NO3 [M + H]+ 506.2690, found
506.2688.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-bromocinna-
mate (60) (method D). Cinnamate 54 (67 mg, 0.12 mmol) after
reaction for 2 h, workup, and chromatography (9 to 11% EtOAc/
hexane) gave 51 mg (90%) of 60 as a cream solid: mp 213−215 °C; IR
1
3351, 2903, 1691, 1197 cm−1; H NMR δ 1.39 (t, J = 7.2 Hz, 3H,
OCH2CH3), 1.82 (bs, 6H, AdCH2), 2.13 (bs, 3H, AdCH), 2.19 (bs,
6H, AdCH2), 4.31 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.01 (s, 1H, OH),
6.48 (d, J = 15.9 Hz, 1H, CHCHCO), 6.74 (d, J = 8.1 Hz, 1H, 5′-
ArH), 7.18 (dd, J = 8.1, 2.1 Hz, 1H, 6′-ArH), 7.33 (d, J = 2.1 Hz, 1H,
2′-ArH), 7.38 (d, J = 8.1 Hz, 1H, 5-ArH), 7.51 (dd, J = 8.1, 1.8 Hz, 1H,
6-ArH), 7.67 (d, J = 15.9 Hz, 1H, CHCHCO), 7.86 (d, J = 1.8 Hz,
1H, 2-ArH); HRMS calcd for C27H29BrO3 [M + H]+ 481.1373, found
481.1374.
Ethyl (E)-3-[6-(1-Adamantyl)-7-hydroxy-9H-carbazol-2-yl]-
propenoate (65) (method D). Ester 64 (100 mg, 0.20 mmol) after
reaction for 2.3 h, workup, and chromatography (16 to 28% EtOAc/
hexane) gave 60 mg (73%) of 65 as a yellow solid: mp 244−246 °C;
1
IR 3311, 2915, 1670, 1249 cm−1; H NMR δ 1.39 (t, J = 7.2 Hz, 3H,
OCH2CH3), 1.86 (bs, 6H, AdCH2), 2.17 (bs, 3H, AdCH), 2.27 (bs,
6H, AdCH2), 4.32 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.25 (s, 1H, OH),
6.50 (d, J = 15.9 Hz, 1H, CHCHCO), 6.73 (s, 1H, 8H-carbazole),
7.43 (dd, J = 8.1, 1.2 Hz, 1H, 3H-carbazole), 7.46 (s, 1H, 1H-
carbazole), 7.84 (d, J = 15.9 Hz, 1H, CHCHCO), 7.86 (bs, 1H, 9H-
carbazole), 7.91 (s, 1H, 5H-carbazole), 7.97 (d, J = 8.1 Hz, 1H, 4H-
carbazole); HRMS calcd for C27H29NO3 [M + H]+ 416.2226, found
416.2224.
(E)-3-[8-(1-Adamantyl)-7-hydroxy-9H-carbazol-2-yl]propenoic
Acid (66) (method I). Ester 65 (42 mg, 0.10 mmol) upon hydrolysis
for 23 h, workup, and chromatography (0:67:33 to 6.5:93.5:0 MeOH/
EtOAc/hexane) afforded 15 mg (38%) of 66 as a yellow solid: mp
256−258 °C; IR 3334, 2905, 1685, 1280 cm−1; 1H NMR (acetone-d6)
δ 1.87 (bs, 6H, AdCH2), 2.13 (bs, 3H, AdCH), 2.33 (bs, 6H, AdCH2),
6.55 (d, J = 15.9 Hz, 1H, CHCHCO), 7.00 (s, 1H, 8H-carbazole),
7.48 (dd, J = 8.1, 1.5 Hz, 1H, 3H-carbazole), 7.71 (d, J = 1.5 Hz, 1H,
1H-carbazole), 7.84 (d, J = 15.9 Hz, 1H, CHCHCO), 7.93 (s, 1H,
5H-carbazole), 8.07 (d, J = 8.1 Hz, 1H, 4H-carbazole), 10.15 (bs, 1H,
9H-carbazole); HRMS calcd for C25H25NO3 [M + H]+ 388.1913,
found 388.1912.
Ethyl (E)-3-[8-(1-Adamantyl)-7-hydroxy-9H-carbazol-2-yl]-
propenoate (68) (method D). Ester 67 (100 mg, 0.20 mmol) after
reaction for 2 h, workup, and chromatography (16 to 28% EtOAc/
hexane) gave 32 mg (39%) of 68 as a yellow solid: mp 163−166 °C;
IR 3392, 2920, 1680, 1242 cm−1; 1H NMR (acetone-d6) δ 1.39 (t, J =
7.2 Hz, 3H, OCH2CH3), 1.80−2.00 (m, 6H, AdCH2), 2.13 (bs, 3H,
AdCH), 2.68 (bs, 6H, AdCH2), 4.25 (q, J = 7.2 Hz, 2H, OCH2CH3),
5.17 (bs, 1H, OH), 6.51 (d, J = 15.9 Hz, 1H, CHCHCO), 6.86 (d,
J = 8.4 Hz, 1H, 6H-carbazole), 7.47 (dd, J = 8.1, 1.2 Hz, 1H, 3H-
carbazole), 7.55 (d, J = 1.2 Hz, 1H, 1H-carbazole), 7.77 (d, J = 8.4 Hz,
1H, 5H-carbazole), 7.86 (d, J = 15.9 Hz, 1H, CHCHCO), 7.92 (d,
J = 8.1 Hz, 1H, 4H-carbazole), 8.86 (bs, 1H, 9H-carbazole); HRMS
calcd for C27H29NO3 [M + H]+ 416.2220, found 416.2220.
(E)-3-[6-(1-Adamantyl)-7-hydroxy-9H-carbazol-2-yl]propenoic
Acid (69). To a suspension of 68 (30 mg, 0.07 mmol) in EtOH (0.2
mL) and THF (0.1 mL) was added 25% aqueous NaOH (50 μL, 0.42
mmol). This mixture was stirred for 24 h, acidified to pH 1 with 1 N
HCl (10 mL), and extracted with EtOAc (40 mL, 20 mL). The extract
was washed (brine), dried, and concentrated at reduced pressure. The
residue was chromatographed (0:50:50 to 9:91:0 MeOH/EtOAc/
hexane and then 9:91 to 17:83 MeOH/CH2Cl2). The product was
washed (pentane) to give 20 mg (74%) of 69 as a yellow solid: mp
184−187 °C; IR 3342, 2907, 1679, 1270 cm−1; 1H NMR (CD3OD) δ
1.82−2.02 (m, 6H, AdCH2), 2.15 (bs, 3H, AdCH), 2.65 (bs, 6H,
AdCH2), 6.56 (d, J = 16.2 Hz, 1H, CHCHCO), 6.72 (d, J = 8.1 Hz,
1H, 6H-carbazole), 7.31 (d, J = 8.1 Hz, 1H, 3H-carbazole), 7.65 (d,
J = 16.2 Hz, 1H, CHCHCO), 7.70 (s, 1H, 1H-carbazole), 7.71 (d,
J = 8.1 Hz, 1H, 5H-carbazole), 7.85 (d, J = 8.1 Hz, 1H, 4H-carbazole);
HRMS calcd for C25H25NO3 [M + H]+ 388.1907, found 388.1908.
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3-bromocinnamic
Acid (61) (method I). Cinnamate 60 (30 mg, 0.06 mmol) after
hydrolysis for 5 h, workup, and washing (hexane and CH2Cl2)
afforded 25 mg (89%) of 61 as a cream solid: mp 244−247 °C; IR
1
3313, 2915, 1702, 1166 cm−1; H NMR (CD3OD) δ 1.84 (bs, 6H,
AdCH2), 2.08 (bs, 3H, AdCH), 2.21 (bs, 6H, AdCH2), 6.54 (d, J =
16.2 Hz, 1H, CHCHCO), 6.79 (d, J = 8.1 Hz, 1H, 5′-ArH), 7.08
(dd, J = 8.1, 2.1 Hz, 1H, 6′-ArH), 7.22 (d, J = 2.1 Hz, 1H, 2′-ArH),
7.38 (d, J = 8.1 Hz, 1H, 5-ArH), 7.62 (dd, J = 8.1, 1.5 Hz, 1H, 6-ArH),
7.67 (d, J = 16.2 Hz, 1H, CHCHCO), 7.92 (d, J = 1.5 Hz, 1H, 2-
ArH); HRMS calcd for C25H25BrO3 [M + H]+ 453.1060, found
453.1062.
Ethyl (E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3,6-dibromo-
cinnamate (62) (method D). Cinnamate 55 (128 mg, 0.20 mmol)
after reaction for 2 h, workup, and chromatography (4 to 12% EtOAc/
hexane) gave 97 mg (88%) of 62 as a pale yellow solid: mp 211−213
1
°C; IR 3382, 2905, 1694, 1194 cm−1; H NMR δ 1.39 (t, J = 7.2 Hz,
3H, OCH2CH3), 1.82 (bs, 6H, AdCH2), 2.13 (bs, 3H, AdCH), 2.18
(bs, 6H, AdCH2), 4.33 (q, J = 7.2 Hz, 2H, OCH2CH3), 5.03 (s, 1H,
OH), 6.45 (d, J = 15.9 Hz, 1H, CHCHCO), 6.74 (d, J = 8.4 Hz, 1H,
5′-ArH), 7.15 (dd, J = 8.4, 2.1 Hz, 1H, 6′-ArH), 7.30 (d, J = 2.1 Hz,
1H, 2′-ArH), 7.61 (s, 1H, 5-ArH), 7.91 (s, 1H, 2-ArH), 8.01 (d, J =
15.9 Hz, 1H, CHCHCO); HRMS calcd for C27H28Br2O3 [M + H]+
559.0478, found 559.0468.
(E)-4-[3′-(1-Adamantyl)-4′-hydroxyphenyl]-3,6-dibromocinnamic
Acid (63) (method I). Cinnamate 62 (58 mg, 0.104 mmol) upon
hydrolysis for 3.5 h, workup, and washing (hexane, CH2Cl2, and
CHCl3) afforded 53 mg (96%) of 63 as a yellow solid: mp 262−264
°C; IR 3299, 2904, 1701, 1249 cm−1; 1H NMR (CD3OD) δ 1.84 (bs,
6H, AdCH2), 2.08 (bs, 3H, AdCH), 2.21 (bs, 6H, AdCH2), 6.55 (d,
J = 15.9 Hz, 1H, CHCHCO), 6.80 (d, J = 8.4 Hz, 1H, 5′-ArH), 7.08
(dd, J = 8.4, 2.1 Hz, 1H, 6′-ArH), 7.21 (d, J = 2.1 Hz, 1H, 2′-ArH),
7.61 (s, 1H, 5-ArH), 7.97 (d, J = 15.9 Hz, 1H, CHCHCO), 8.07 (s,
1H, 2-ArH); HRMS calcd for C25H24Br2O3 [M + H]+ 531.0165, found
531.0171.
Ethyl (E)-3-[6-(1-Adamantyl)-7-benzyloxy-9H-carbazol-2-yl]-
propenoate (64) and Ethyl (E)-3-[8-(1-Adamantyl)-7-benzyloxy-
9H-carbazol-2-yl]propenoate (67). A reported method was used to
synthesize carbazoles 64 and 67.18,19 To a solution of 39 (0.80 g, 1.49
mmol) in 1,2-dichlorobenzene was added PPh3 (0.97 g, 3.72 mmol).
The mixture was heated at 180 °C while being stirred under argon for
19 h. After cooling, the mixture was chromatographed (9 to 25%
EtOAc/hexane) to give fractions containing 64 or 67, as the major
isomer, each of which was separately crystallized (EtOH, 6 mL) to
afford 390 mg (52%) of 64 as a yellow solid: mp 203−205 °C, and
330 mg (44%) of 67 as a yellow solid: mp 144−145 °C. 64: IR 3334,
2905, 1704, 1291 cm−1; 1H NMR δ 1.38 (t, J = 7.2 Hz, 3H,
OCH2CH3), 1.79 (bs, 6H, AdCH2), 2.12 (bs, 3H, AdCH), 2.28 (bs,
246
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
Computational Studies. Docking Experiments. Docking of
analogues to the SHP homology model6 derived from the USP LBD−
phospholipid complex structure (PDB entry 1G2N) using BioMed
Cache version 6.2 (Fujitsu Ltd., Beaverton, OR) was assessed as we
previously described.15 The binding pocket site was derived by
selecting all neighboring residues within 4 Å (radius) of the USP
ligand 1-stearoyl-2-palmitoylglycero-3-phosphoethanolamine, which
had been docked to the SHP model.6 Both the ligand and side chains
of pocket residues were allowed to be flexible during the docking
process. Final docking poses in Figure 5 were analyzed by measuring
interatom distances after overlaying the docked structures by
superposing the backbone of the SHP model structure.
CoMFA Experiments. Comparative molecular field analysis
(CoMFA) of analogues of 8 substituted at position 3 was performed
using SYBYL version 8.0 (Tripos Inc., St. Louis, MO). Three partial
CoMFA models were derived from 11 analogues and their IC50 values
for inhibiting the viability of three leukemia cell lines (K562, OCI-
AML2, and MOLT4). Each initial structure was minimized by the
conjugate gradient method and convergence criterion of 0.005 kcal/
mol using the Tripos force field with Gasteiger−Huckel charges added.
The asterisk-labeled atoms of analogue 47, which had the lowest IC50
values for inhibiting OCI-AML2 and MOLT4 cell viability, comprised
the common substructure (Figure S3A of the Supporting Information)
used for the alignment. The other 10 analogues were aligned to this
template using the database alignment method in SYBYL version 8.0
as shown in Figure S3B of the Supporting Information and placed in a
three-dimensional lattice. The steric and electrostatic CoMFA fields were
calculated at each lattice intersection of a 2.0 Å spaced grid. An sp3 carbon
atom with a formal charge of +1 and a van der Waals radius of 1.52 Å was
used as a probe atom to generate Lennard-Jones (steric energies) and
Coulombic (electrostatic energies) potential. The van der Waals and
Coulombic potentials representing the steric and electrostatic fields,
respectively, were calculated using standard Tripos force fields with a
distance-dependent dielectric constant of 1.00. Cutoff default values of 30.0
kcal/mol for both steric and electrostatic fields were used.
Biologic Studies. NMR Protein Binding Studies. GST-bound
SHP was expressed and purified as described previously.15 One-
dimensional 1H NMR spectra were recorded with 128 transients and a
sweep width of 12376 Hz using a Bruker Avance 600 MHz NMR
spectrometer equipped with a TCI cryoprobe so that the sample
temperature was maintained at 11 °C. The repetition time was 3 s.
Spectra were recorded for solutions of each compound (100 μM 8 and
47 and 200 μM 46) in the bead elution buffer [50 mM Tris (pH 7.2)
containing 20 mM reduced glutathione] prepared using 92.5% D2O
and 2% DMSO-d6 alone or with GST-bound SHP (0.5 μM for 8 and
47 and 1.0 μM for 46).
Cell Culture. KG-1 AML cells from H. P. Koeffler (UCLA, Cedars-
Sinai Medical Center, Los Angeles, CA) were grown in RPMI 1640
medium supplemented with 5% heat-inactivated fetal bovine serum
(FBS) and gentamycin (100 μg/mL). ATRA-resistant HL-60R
myeloid leukemia cells from S. Collins (University of Washington,
Seattle, WA) were grown under the same conditions. MDA-MB-231
breast cancer cells (ATCC, Manassas, VA) were seeded in DMEM/F-
12 medium supplemented with 10% FBS and gentamycin (25 μg/
mL). Cells were incubated for 24 h before being treated with
compound (final Me2SO concentration of 0.1%) for 48 or 72 h. K562
AML (ATCC), MOLT4 T-ALL (ATCC), and OCI-AML2 (from S.
Kitada, Sanford-Burnham Medical Research Institute) leukemia cell
lines were cultured in RPMI 1640 medium (CellGro, Mediatech,
Manassas, VA) supplemented with 10% heat-inactivated FBS and a 1%
penicillin/streptomycin solution (Omega Scientific, Tarzana, CA) at
37 °C in a 5% CO2 atmosphere.
Cell Apoptosis. KG-1 leukemia cells were seeded and incubated for
48 h as described for the cell growth inhibition experiments. Apoptotic
cells were identified using acridine orange staining for DNA
fragmentation as described previously.11 Results shown are the
averages of triplicate determinations SDs, which were <10%.
Cancer Stem Cell Assay. The effects of compounds on proliferation
of MMTV-Wnt1 murine mammary cancer stem cells derived from
mammospheres were determined as previously described.15 Briefly,
dissociated cells were plated in MSC medium without serum into
gelatin-coated 384-well plates, incubated for 24 h, and then treated with
the indicated concentrations of compound in vehicle (final Me2SO
concentration of 0.1%) or vehicle (Me2SO) alone for 72 h. Cells were
fixed and stained with DAPI. Cell numbers were determined by analyzing
nine images acquired by the IC100 automatic focusing imaging system and
Cyteseer (Vala Sciences, San Diego, CA). Results are expressed as means
SD from triplicate determinations.
Cell Viability Assay. This assay was conducted as previously
reported with several modifications.15 Briefly, K562, MOLT4, and
OCI-AML2 leukemia cell suspensions in medium containing 10% fetal
bovine serum (FBS) (1800 cells in 50 μL of medium/well) were
added to 384-well white-bottomed Greiner plates and incubated with
compounds at the specified concentrations. Compounds were
dissolved in Me2SO (maximal final concentration of 0.1%, which
had no effect on cell growth) and then diluted into medium containing
10% FBS. After 72 h at 37 °C, plates were cooled to room temperature for
15 min before luciferase assays were conducted to determine ATP levels
using CellTiter Glo Luminescent Cell Viability Reagent (20 μL/well;
Promega, Madison, WI) and light emission was measured after 10 min
according to the manufacturer’s instructions. Experiments were performed
in quadruplicate (mean SD). Cell viability was calculated on the basis of
maximal luminescence intensity observed for each cell line in the presence
of the Me2SO vehicle alone (100% value). Assay results are displayed
graphically in Figure S1 of the Supporting Information.
Statistical Analysis. For the data obtained at analogue concen-
trations of 1.0 and 5.0 μM compared to those of the nontreated
control groups, which are given in Tables 1 and 2, the statistical
significance between treated and nontreated control groups (n = 3
each) was determined by the two-sample t test with a degree of
freedom (k) of 4 on the mean standard deviation (SD) data. IC50
values for effects of analogues on the K562, OCI-AML2, and MOLT4
leukemia cell lines were determined from dose−response curves of the
means of quadruplicate determinations
SD, which are shown in
Figure S1 of the Supporting Information. The means of the IC50 values
SD of the effects of compounds on the mammary stem cells that are
listed in Table 2 were determined by a one-tailed t test on the IC50
values determined from dose−response curves in triplicate experi-
ments. Statistical comparisons between the IC50 value of 8 and those
of five other analogues were made using the two-sample t test with a
degree of freedom (k) of 4 on the means standard deviation (SD) of
their IC50 values. The dose−response curves shown in Figure 3
represent the means of triplicate determinations SDs of the effects
of analogues on these stem cells.
ASSOCIATED CONTENT
■
S
* Supporting Information
Dose−response curves showing effects of the analogues on
ATP levels in K562 AML, OCI-AML2 AML, and MOLT4 T-
ALL leukemia cell lines (Figure S1), relationships between
logarithms of analogue 3-substituent volumes and logarithms of
analogue concentrations that inhibited leukemia cell viability by
50% (Figure S2), physical properties of these substituents at
position 3 used for the relationship studies in Figure S2 (Table
S1), common substructure used to generate the CoMFA model
and the compound alignment (Figure S3), CoMFA model
statistics and data used to generate Figure 4 (Tables S2 and
S3), HPLC analytical data on target compound purity (Table
S4), and the 600 MHz 1H NMR spectrum of the D2O used for
the binding studies shown in Figure 4A,B (Figure S4). This
Inhibition of Proliferation. KG-1 leukemia cells were seeded in six-
well plates containing 3 mL of medium and 5% FBS per well and
incubated overnight. After compounds dissolved in Me2SO were
added to provide the designated concentrations, cells were incubated
for 48 h and harvested. Cell numbers were determined by counting
using a hemocytometer as described previously.11 Results shown are
averages of triplicate determinations
standard deviations (SDs),
which were <10%.
247
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
requires NF-κB noncanonical and canonical pathway activation. Cell
Death Differ. 2011, 18, 164−173.
material is available free of charge via the Internet at http://
pubs.acs.org.
(11) Dawson, M. I.; Harris, D. L.; Liu, G.; Hobbs, P. D.; Lange, C.
W.; Jong, L.; Bruey-Sedano, N.; James, S. Y.; Zhang, X.-K.; Peterson,
V. J.; Leid, M.; Farhana, L.; Rishi, A. K.; Fontana, J. A. Antagonist
analogue of 6-[3′-(1-adamantyl)-4′-hydroxyphenyl]-2-
naphthalenecarboxylic acid (AHPN) family of apoptosis inducers
that effectively blocks AHPN-induced apoptosis but not cell-cycle
arrest. J. Med. Chem. 2004, 47, 3518−3536.
AUTHOR INFORMATION
■
Corresponding Author
*Sanford-Burnham Medical Research Institute, 10901 N. Torrey
Pines Rd., La Jolla, CA 92037. Phone: (858) 646-3165. Fax: (858)
646-3197. E-mail: mdawson@sanfordburnham.org.
(12) Klaholz, B. P.; Renaud, J.-P.; Mitschler, A.; Zusi, C.; Chambon, P.;
Gronemeyer, H.; Moras, D. Conformational adaptation of agonists to the
human nuclear receptor RARγ. Nat. Struct. Biol. 1998, 5, 199−202.
(13) Dawson, M. I.; Xia, Z.; Jiang, T.; Ye, M.; Fontana, J. A.; Farhana,
L.; Patel, B.; Xue, L. P.; Bhuiyan, M.; Pellicciari, R.; Macchiarulo, A.;
Nuti, R.; Zhang, X.-K.; Han, Y.-H.; Tautz, L.; Hobbs, P. D.; Jong, L.;
Waleh, N.; Chao, W.-R.; Feng, G.-S.; Pang, Y.; Su, Y. Adamantyl-
substituted retinoid-derived molecules that interact with the orphan
nuclear receptor small heterodimer partner: Effects of replacing the 1-
adamantyl or hydroxyl group on inhibition of cancer cell growth,
induction of cancer cell apoptosis, and inhibition of SRC homology 2
domain-containing protein tyrosine phosphatase-2 activity. J. Med.
Chem. 2008, 51, 5650−5662.
ACKNOWLEDGMENTS
■
We are grateful for support of this research by U.S. Public
Health Service Grants R01 CA109370 (J.A.F. and M.I.D.) and
CA55164 (J.C.R.) and a VA Merit Award (J.A.F.) and to Dr.
Roberto Pellicciari (University of Perugia, Perugia, Italy) for
use of the USP LBD-derived homology model of SHP.
ABBREVIATIONS
■
1-Ad, 1-adamantyl; 3-Cl-AHPC, (E)-4-[3′-(1-adamantyl)-4′-
hydroxyphenyl]-3-chlorocinnamic acid; AHPN, 6-[3′-(1-ada-
mantyl)-4′-hydroxyphenyl]-2-naphthalenecarboxylic acid; AML,
acute myelocytic leukemia; ATRA, all-trans-retinoic acid; CML,
chronic myeloid leukemia; CoMFA, comparative molecular field
analysis; DAPI, 4′,6-diamidino-2-phenylindole dihydrochloride;
RAR, retinoic acid receptor; RXR, retinoid X receptor; SHP,
small heterodimer partner; T-ALL, T-cell acute lymphoblastic
leukemia
(14) Heck, R. F.; Nolley, J. P. Jr. Palladium-catalyzed vinylic hydrogen
substitution reactions with aryl, benzyl, and styryl halides. J. Org. Chem.
1972, 37, 2320−2322.
(15) Xia, Z.; Farhana, L.; Correa, R. G.; Das, J. K.; Castro, D. J.; Yu,
J.; Oshima, R. G.; Reed, J. C.; Fontana, J. A.; Dawson, M. I.
Heteroatom-substituted analogues of orphan nuclear receptor small
heterodimer partner ligand and apoptosis inducer (E)-4-[3-(1-
adamantyl)-4-hydroxyphenyl]-3-chlorocinnamic acid. J. Med. Chem.
2011, 54, 3793−3816.
̈
REFERENCES
(16) Wittig, G.; Schollkopf, U. Uber Triphenyl-phosphin-methylene
̈
■
als olefinbildende Reagenzien I. Chem. Ber. 1954, 87, 1318.
(17) Miyaura, N.; Suzuki, A. Palladium-catalyzed cross-coupling
reactions of organoboron compounds. Chem. Rev. 1995, 95, 2457−
2483.
(18) Freeman, A. W.; Urvoy, M.; Criswell, M. E. Triphenylphosphine-
mediated reductive cyclization of 2-nitrobiphenyls: A practical and
convenient synthesis of carbazoles. J. Org. Chem. 2005, 70, 5014−5019.
(19) Naffziger, M. R.; Ashburn, B. O.; Perkins, J. R.; Carter, R. G.
Diels-Alder approach for the construction of halogenated, o-nitro
biaryl templates and application to the total synthesis of the anti-HIV
agent siamenol. J. Org. Chem. 2007, 72, 9857−9865.
(20) Shao, Z. M.; Dawson, M. I.; Li, X. S.; Rishi, A. K.; Sheikh, M. S.;
Han, Q. X.; Ordonez, J. V.; Shroot, B.; Fontana, J. A. p53 independent
G0/G1 arrest and apoptosis induced by a novel retinoid in human
breast cancer cells. Oncogene 1995, 11, 493−504.
(21) Tang, X. H.; Gudas, L. J. Retinoids, retinoic acid receptors, and
cancer. Annu. Rev. Pathol. 2011, 6, 345−364.
(22) Gieseler, F.; Bauer, E.; Nuessler, V.; Clark, M.; Valsamas, S.
Molecular effects of topoisomerase II inhibitors in AML cell lines:
Correlation of apoptosis with topoisomerase II activity but not with
DNA damage. Leukemia 1999, 13, 1859−1863.
(23) Qi, H.; Ratnam, M. Synergistic induction of folate receptor β by
all-trans retinoic acid and histone deacetylase inhibitors in acute
myelogenous leukemia cells: Mechanism and utility in enhancing
selective growth inhibition by antifolates. Cancer Res. 2006, 66, 5875−
5882.
(24) Drach, J.; Lopez-Berestein, G.; McQueen, T.; Andreeff, M.;
Mehta, K. Induction of differentiation in myeloid leukemia cell lines
and acute promyelocytic leukemia cells by liposomal all-trans-retinoic
acid. Cancer Res. 1993, 53, 2100−2104.
(1) Farhana, L.; Dawson, M. I.; Leid, M.; Wang, L.; Moore, D. D.;
Liu, G.; Xia, Z.; Fontana, J. A. Adamantyl-substituted retinoid-related
molecules bind small heterodimer partner and modulate the Sin3A
repressor. Cancer Res. 2007, 67, 318−325.
(2) Zhang, Y.; Soto, J.; Park, K.; Viswanath, G.; Kuwada, S.; Abel, E.
D.; Wang, L. Nuclear receptor SHP, a death receptor that targets
mitochondria, induces apoptosis and inhibits tumor growth. Mol. Cell.
Biol. 2010, 30, 1341−1356.
(3) Zhang, Y.; Wang, L. Nuclear receptor SHP inhibition of Dnmt1
expression via ERRγ. FEBS Lett. 2011, 585, 1269−1275.
(4) Kim, M.-K.; Chanda, D.; Lee, I.-K.; Choi, H.-S.; Park, K.-G.
Targeting orphan nuclear receptor SHP in the treatment of metabolic
diseases. Expert Opin. Ther. Targets 2010, 14, 453−466.
(5) Billas, I. M.; Moulinier, L.; Rochel, N.; Moras, D. Crystal
structure of the ligand-binding domain of the ultraspiracle protein
USP, the ortholog of retinoid X receptors in insects. J. Biol. Chem.
2001, 276, 7465−7474.
(6) Macchiarulo, A.; Rizzo, G.; Costantino, G.; Fiorucci, S.;
Pellicciari, R. Unveiling hidden features of orphan nuclear receptors:
The case of the small heterodimer partner (SHP). J. Mol. Graphics
Modell. 2006, 24, 362−372.
(7) Cellanetti, M.; Gunda, V.; Wang, L.; Macchiarulo, A.; Pellicciari, R .
Insights into the binding mode and mechanism of action of some atypical
retinoids as ligands of the small heterodimer partner (SHP). J. Comput.-
Aided Mol. Des. 2010, 24, 943−956.
(8) Farhana, L.; Dawson, M. I.; Dannenberg, J.-H.; Xu, L.; Fontana, J.
A. SHP and Sin3A expression are essential for adamantyl-substituted
retinoid-related molecule-mediated nuclear factor-κB activation, c-Fos/
c-Jun expression, and cellular apoptosis. Mol. Cancer Ther. 2009, 8,
1625−1635.
(25) Zhang, Y.; Dawson, M. I.; Ning, Y.; Polin, L.; Parchment, R. E.;
Corbett, T.; Mohamed, A. N.; Feng, K.-C.; Farhana, L.; Rishi, A. K.;
Hogge, D.; Leid, M.; Peterson, V. J.; Zhang, X.-K.; Mohammad, R.; Lu,
J.-S.; Willman, C.; VanBuren, E.; Biggar, S.; Edelstein, M.; Eilender, D.;
Fontana, J. A. Induction of apoptosis in retinoid-refractory acute
myelogenous leukemia by a novel AHPN analog. Blood 2003, 102,
3743−3752.
(9) Farhana, L.; Dawson, M. I.; Xia, Z.; Aboukameel, A.; Xu, L.; Liu,
G.; Das, J. K.; Hatfield, J.; Levi, E.; Mohammad, R.; Fontana, J. A.
Adamantyl-substituted retinoid-related molecules induce apoptosis in
human acute myelogenous leukemia cells. Mol. Cancer Ther. 2010, 9,
2903−2913.
(10) Farhana, L.; Dawson, M. I.; Murshed, F.; Fontana, J. A. Maximal
adamantyl-substituted retinoid-related molecule-induced apoptosis
248
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249

Journal of Medicinal Chemistry
Article
(26) Schmelz, K.; Sattler, N.; Wagner, M.; Lubbert, M.; Dorken, B.;
Tamm, I. Induction of gene expression by 5-aza-2′-deoxycytidine in
acute myeloid leukemia (AML) and myelodysplastic syndrome
(MDS) but not epithelial cells by DNA-methylation-dependent and
-independent mechanisms. Leukemia 2005, 19, 103−111.
(27) Zhou, H. R.; Shen, J. Z.; Fu, H. Y.; Ye, B. G.; Fan, L. P.; Lin, F.
A. Detection of promoter methylation of p16 gene in hematological
malignant cell lines by nested methylation specific polymerase chain
reaction. Zhongguo Shi Yan Xue Ye Xue Za Zhi 2006, 14, 375−378.
(28) Koeffler, H. P.; Golde, D. W. Human myeloid leukemia cell
lines: A review. Blood 1980, 56, 344−350.
heterodimer partner nuclear receptor: Effects of modifying its carboxylate
group on apoptosis, proliferation, and protein-tyrosine phosphatase
activity. J. Med. Chem. 2007, 50, 2622−2639.
(45) Mcbee, E. T.; Pierce, O. R.; Lowery, R. D.; Rapkin, E. Some dihalo-
(trifluoromethyl)benzenes. J. Am. Chem. Soc. 1951, 73, 3932−3934.
(46) Dabrowski, M.; Kubicka, J.; Lulinski, S.; Serwatowski, J. Halogen−
lithium exchange between substituted dihalobenzenes and butyllithium:
Application to the regioselective synthesis of functionalized
bromobenzaldehydes. Tetrahedron 2005, 61, 6590−6595.
(47) Molinspiration Chemoinformatics, Slovensky Grob, Slovak
Republic. http://www.molinspiration.com/cgi-bin/properties (ac-
cessed February 8, 2011).
(29) Ekiz, H. A.; Can, G.; Gunduz, U.; Baran, Y. Nilotinib
significantly induces apoptosis in imatinib-resistant K562 cells with
wild-type BCR-ABL, as effectively as in parental sensitive counterparts.
Hematology 2010, 15, 33−38.
(30) Tang, X. Y.; Wang, C. H.; Xiao, G. F. The effect of ATRA-
induced leukemic cell differentiation on Brd7 gene expression in
leukemia cell lines. Zhongguo Shi Yan Xue Ye Xue Za Zhi 2010, 18,
593−596.
(31) Jimi, S.; Mashima, K.; Matsumoto, T.; Hara, S.; Suzumiya, J.;
Tamura, K. RARα is a regulatory factor for Am-80-induced cell growth
inhibition of hematologic malignant cells. Int. J. Oncol. 2007, 31, 397−
404.
(32) Lee, H. J.; Thompson, J. E.; Wang, E. S.; Wetzler, M.
Philadelphia chromosome-positive acute lymphoblastic leukemia:
Current treatment and future perspectives. Cancer 2011, 117, 1583−
1594.
(33) Ru, H. Y.; Chen, R. L.; Lu, W. C.; Chen, J. H. hBUB1 defects in
leukemia and lymphoma cells. Oncogene 2002, 21, 4673−4679.
(34) Rodrigues, N. R.; Rowan, A.; Smith, M. E.; Kerr, I. B.; Bodmer,
W. F.; Gannon, J. V.; Lane, D. P. p53 mutations in colorectal cancer.
Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 7555−7559.
(35) Yamaguchi, T.; Katoh, I.; Kurata, S. Azidothymidine causes
functional and structural destruction of mitochondria, glutathione
deficiency and HIV-1 promoter sensitization. Eur. J. Biochem. 2002,
269, 2782−2788.
(36) Hakimelahi, G. H.; Ly, T. W.; Yu, S.-F.; Zakerinia, M.; Khalafi-
Nezhad, A.; Soltani, M. N.; Gorgani, M. N.; Chadegani, A. R.;
Moosavi-Movahedi, A. A. Design and synthesis of a cephalosporin-
retinoic acid prodrug activated by a monoclonal antibody−β-lactamase
conjugate. Bioorg. Med. Chem. 2001, 9, 2139−2147.
(37) Anuchapreeda, S.; Tima, S.; Duangrat, C.; Limtrakul, P. Effect of
pure curcumin, demethoxycurcumin, and bisdemethoxycurcumin on
WT1 gene expression in leukemic cell lines. Cancer Chemother.
Pharmacol. 2008, 62, 585−594.
(38) Sugiyama, H. Cancer immunotherapy targeting Wilms’ tumor
gene WT1 product. Expert Rev. Vaccines 2005, 4, 503−512.
(39) Ochsenreither, S.; Fusi, A.; Busse, A.; Bauer, S.; Scheibenbogen,
C.; Stather, D.; Thiel, E.; Keilholz, U.; Letsch, A. “Wilms Tumor
Protein 1” (WT1) peptide vaccination-induced complete remission in
a patient with acute myeloid leukemia is accompanied by the
emergence of a predominant T-cell clone both in blood and bone
marrow. J. Immunother. 2010, 34, 85−91.
(40) Lu, Y.; Shi, T.; Wang, Y.; Yang, H.; Yan, X.; Luo, X.; Jiang, H.;
Zhu, W. Halogen bonding: A novel interaction for rational drug
design? J. Med. Chem. 2009, 52, 2854−2862.
(41) Bissantz, C.; Kuhn, B.; Stahl, M. A medicinal chemist’s guide to
molecular interactions. J. Med. Chem. 2010, 53, 5061−5084.
(42) Scheiner, S.; Kar, T.; Pattanayak, J. Comparison of various types
of hydrogen bonds involving aromatic amino acids. J. Am. Chem. Soc.
2002, 124, 13257−13264.
(43) Kovacs, A.; Varga, Z. Halogen acceptors in hydrogen bonding.
Coord. Chem. Rev. 2006, 250, 710−727.
(44) Dawson, M. I.; Xia, Z.; Liu, G.; Ye, M.; Fontana, J. A.; Farhana, L.;
Patel, B. B.; Arumugarajah, S.; Bhuiyan, M.; Zhang, X.-K.; Han, Y.-H.;
Stallcup, W. B.; Fukushi, J.; Mustelin, T.; Tautz, L.; Su, Y.; Harris, D. L.;
Waleh, N.; Hobbs, P. D.; Jong, L.; Chao, W. R.; Schiff, L. J.; Sani, B. P. An
adamantyl-substituted retinoid-derived molecule that inhibits cancer cell
growth and angiogenesis by inducing apoptosis and binds to small
249
dx.doi.org/10.1021/jm2011436 | J. Med. Chem. 2012, 55, 233−249