Structure-activity relationship refinement and further assessment of 4-phenylquinazoline-2-carboxamide translocator protein ligands as antiproliferative agents in human glioblastoma tumors

Article abstract of DOI:10.1021/jm401721h

Structure-activity relationships (SARs) within the 4-phenylquinazoline-2- carboxamide series of translocator protein (TSPO) ligands have been explored further by the synthesis and TSPO binding affinity evaluation of N-benzyl-N-ethyl/methyl derivatives var

Full text of DOI:10.1021/jm401721h

Article
pubs.acs.org/jmc
Structure−Activity Relationship Refinement and Further Assessment
of 4‑Phenylquinazoline-2-carboxamide Translocator Protein Ligands
as Antiproliferative Agents in Human Glioblastoma Tumors
Sabrina Castellano,† Sabrina Taliani,*,‡ Monica Viviano,† Ciro Milite,† Eleonora Da Pozzo,‡
Barbara Costa,‡ Elisabetta Barresi,‡ Agostino Bruno,§ Sandro Cosconati,∥ Luciana Marinelli,§
Giovanni Greco,§ Ettore Novellino,§ Gianluca Sbardella,† Federico Da Settimo,‡ and Claudia Martini‡
†Dipartimento di Farmacia, Universita
‡Dipartimento di Farmacia, Universita
§Dipartimento di Farmacia, Universita
́
́
́
di Salerno, Via Giovanni Paolo II 132, 84084 Fisciano, Salerno, Italy
di Pisa, Via Bonanno 6, 56126 Pisa, Italy
di Napoli “Federico II”, Via D. Montesano 49, 80131 Napoli, Italy
∥DiSTABiF, Seconda Universita
́
di Napoli, Via Vivaldi 43, 81100 Caserta, Italy
S
* Supporting Information
ABSTRACT: Structure−activity relationships (SARs) within the
4-phenylquinazoline-2-carboxamide series of translocator protein
(TSPO) ligands have been explored further by the synthesis and
TSPO binding affinity evaluation of N-benzyl-N-ethyl/methyl
derivatives variously decorated at the 6-, 2′-, 4′-, and 4″-positions.
Most of the compounds showed high affinity with Ki values in the
nanomolar/subnanomolar range. A pharmacophore model was
developed and employed to better address SAR data presented by
the new TSPO ligands. A subset of the new compounds (5, 8, 12,
and 19) were tested for their ability to inhibit the viability of
human glioblastoma cell line U343. The observed antiproliferative
effect was demonstrated to be specific for compound 19, endowed
with the best combination of binding affinity and efficacy. Furthermore, the ability of 19 to induce mitochondrial membrane
dissipation (Δψm) substantiated the intracellular pro-apoptotic mechanism activated by the binding of this class of ligands to
TSPO.
factor, and Smac) into the cytosol, osmotic swelling of the
mitochondrial matrix, and deregulation of ATP synthesis and
oxidative phosphorylation, causing apoptotic and necrotic
signaling cascade reactions and subsequent cell death.3,6,7
TSPO tends to be expressed in tissues that produce steroids
and that are mitochondrially enriched such as the myocardium,
skeletal muscle, and renal tissue, whereas tissues such as the
liver and brain exhibit comparatively modest expression.3
Aberrant expression of TSPO has been linked to multiple
diseases, including cancer, brain injury, neurodegeneration, and
ischemia−reperfusion injury.8 Specifically, TSPO is highly
expressed in cancerous tissues of the breast, ovary, colon,
prostate, and brain, compared to normal human tissues,
suggesting a role for TSPO in carcinogenesis.9−13 A positive
correlation between TSPO levels and the metastatic potential
of human breast and brain gliomas, as well as astrocytomas, has
also been shown.8,14 Cancer cells are sensitive to apoptotic cell
death induced by binding of ligands to this protein, thus
suggesting TSPO as a potential anticancer drug target.15
INTRODUCTION
■
The translocator protein (TSPO), originally named the
peripheral-type benzodiazepine receptor (PBR),1,2 is an 18
kDa [169 amino acid (aa)] mitochondrial protein typically
localized on the outer mitochondrial membrane,3 although
other expression locations are possible, including the nuclei,
lysosome, Golgi apparatus, peroxisomes, plasma membrane,
and mature erythrocytes.3−5
TSPO plays a key role in the regulation of numerous cellular
processes, such as steroid biosynthesis, cholesterol metabolism,
apoptosis, and cellular proliferation.3 With its five-α-helical
transmembrane domain, the TSPO is believed to form a
complex with several proteins [e.g., voltage-dependent anion
channel (VDAC) and adenine nucleotide transporter (ANT)]
of the outer and inner mitochondrial membranes collectively
known as the mitochondrial permeability transition pore
(MPTP), an important regulator of apoptotic and necrotic
cell death during injury.3,6 In normal homeostatic conditions,
the MPTP maintains the transmembrane potential (Δψm)
through active pore opening and closing. The prolonged
opening of the MPTP results in the release of apoptotic factors
from the mitochondria (cytochrome c, apoptosis-inducing
Received: November 8, 2013
Published: February 28, 2014
© 2014 American Chemical Society
2413
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
a
Chart 1. Structures of Known TSPO Ligands and Newly Synthesized Quinazoline Derivatives 5−23
a
The N,N-dialkyl(2-phenylindol-3-yl)glyoxylamides I and the N,N-dialkyl-4-phenylquinazoline-2-carboxamides II are represented within the
framework of a previously published pharmacophore/topological model of the ligand−TSPO interaction.19,24,30
Actually, classic TSPO ligands such as the isoquinolinecarbox-
amide derivative 1 (PK11195),16,17 Chart 1, and the
benzodiazepine 2 (Ro5-4864),18 Chart 1, as well as (2-
phenylindol-3-yl)glyoxylamide derivative PIGA, previously
synthesized by us,19,20 are able to produce MPTP opening
that leads to dissipation of the Δψm, one of the early events of
the apoptotic cascade activation.7,21 In this view, TSPO ligands
may offer new clinical options to oncologists dealing with
tumors that fail to respond to conventional chemotherapies,
such as glioblastoma multiforme (GBM).
GBM is the most common and deadliest malignant primary
brain tumor in adults. The standard treatment is surgery,
followed by radiation therapy or combined radiation therapy
and chemotherapy. Unfortunately, traditional treatments are
unlikely to result in a prolonged remission, and GBM is one of
the cancers with the worst prognosis and an average life
expectancy of about 12 months after diagnosis.22,23 The
ineffectiveness of the current standard chemotherapy, consist-
ing in the use of DNA-damaging agents, is due, at least partially,
to effective oncogenic blockade of those mechanisms which
would activate cell death following genotoxic insult. In this
view, the activation of the mitochondrion-mediated cell death
machinery by the use of drugs inducing mitochondrial outer
membrane permeabilization may represent a strategic approach
for the development of an alternative therapeutic treatment of
GBM.
isosteres of 1, the well-known TSPO reference ligand.31
Noticeably, the introduction of a second nitrogen atom in
the isoquinoline nucleus of 1 produced a favorable effect on
TSPO affinity, as a great number of derivatives II bind to the
TSPO with Ki values in the nanomolar/subnanomolar range.30
Structure−activity relationships (SARs) of compounds from
class II were rationalized in light of a previously published
pharmacophore/topological model made up of three lipophilic
pockets (L1, L3, and L4) and a H-bond donor group (H1)
(Chart 1).30,32 The main structural requirement for an optimal
interaction with the target protein resulted in the N,N-
disubstitution on the carboxamide moiety. Molecular mechan-
ics calculations indicated that the N,N-disubstituted derivatives
adopt low-energy conformations characterized by the OC
N< moiety lying out of the plane (torsion angle N1C2
CO about 30°) of the quinazoline ring, making the carbonyl
group correctly oriented within the binding cleft to engage a
strong H-bond with the H1 site.30 This hypothesis was
consistent with the results of the work on conformationally
constrained analogues of 1 featuring the isoquinoline-3-
carboxamide core scaffold.33
Furthermore, among the N,N-dialkyl-4-phenylquinazoline-2-
carboxamides II, the most potent derivatives feature at least one
of the two N-alkyl groups with the number of carbon atoms in
the 4−6 range. This suggests that the optimal ligand−TSPO
interaction requires the full occupancy of the L4 hydrophobic
pocket with at least one bulky N-alkyl group as a second crucial
feature.30
A preliminary in vitro biological characterization performed
on selected compounds, chosen as representative of class II,
showed that these quinazoline ligands possess the ability to
bind the TSPO with high affinity, without significantly affecting
In the past decade, we identified two novel classes of potent
and selective TSPO ligands, the N,N-dialkyl(2-phenylindol-3-
yl)glyoxylamides I (Chart 1),19,24−27 designed as conforma-
tionally constrained analogues of the indoleacetamide 3
(FGIN-1-27), Chart 1,28,29 and, more recently, the N,N-
dialkyl-4-phenylquinazoline-2-carboxamides II (Chart 1),30 aza
2414
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of Quinazoline Derivatives 5−23
a
Reagents and conditions: (a) CHOCOOH, NH4OAc, EtOH, rt, 0.5 h; (b) light irradiation, DMF, rt, 12 h, 62−86% (two steps); (c) SOCl2, reflux,
2 h, then monoalkylamine, NEt3, THF, rt, 36−48 h, 43−88%; (d) NaH, CH3I or CH3CH2I, DMF, 0 °C to rt, 0.5 h, 67−93%; (e) LiOH, THF/H2O
(4:1), rt, 12 h, 74%; (f) SOCl2, reflux, 2 h, then N-(4-nitrobenzyl)ethanamine (52) or N-(4-(tosylhydroxy)benzyl)ethanamine (53), NEt3, THF, rt,
36−48 h, 43−88%; (g) 1 N NaOH, EtOH, reflux, 1 h, 76%.
the steroidogenic TSPO function, but producing a modulation
of cell viability on human glioblastoma−astrocytoma cell line
U87MG.30 In addition, data from these experiments suggested
that even small structural modifications on the 4-phenyl-
quinazoline-2-carboxamide core may affect ligand-mediated
modulation of TSPO functions.30
agreement with the proposed pharmacophore/topological
model.24 Moreover, the decoration at the C4 position of the
phenyl ring with various substituents (CH3, OCH3, OH,
COOCH2CH3, COOH) allowed us to probe the steric and
electronic features of the L4 lipophilic pocket of the TSPO
binding site.
Starting from the observation that the presence of a benzyl
moiety is a common feature of high-affinity TSPO ligands [e.g.,
4 (AC-5216), N-benzyl-N-ethyl(2-phenylindol-3-yl)-
glyoxylamides I, Chart 1] endowed with a particularly enhanced
functional profile,24,34,35 herein we designed and synthesized a
novel series of N-benzyl-substituted 4-phenylquinazoline-2-
carboxamides (5−23).
In particular, it was anticipated that the benzyl group would
meet one of the critical structural requirements of 4-
phenylquinazoline-2-carboxamides for binding to TSPO,
namely, the presence of a substituent with the number of
carbon atoms between 4 and 6 on the carboxamide nitrogen. In
fact, in the indolylglyoxylamide class I, the N-benzyl-N-ethyl
derivatives showed affinities roughly similar to those exhibited
by N,N-di-n-butyl compounds, suggesting that a benzyl group
may establish hydrophobic interactions like an aliphatic moiety
of similar size within the lipophilic pockets L3 and L4, in
In this view, the insertion of a benzyl moiety in the 4-
phenylquinazoline-2-carboxamide scaffold has an added value
in the SAR studies of TSPO ligands. The absence of X-ray
structures, not only for TSPO but also for closely related
proteins, makes the construction of a three-dimensional (3D)
TSPO homology model an arduous attempt. This implies that
the rational design of new TSPO ligands must necessarily rely
on ligand-based methods. Thus, comprehensive SAR studies on
different classes of TSPO ligands are of utmost importance for
the design of new TSPO modulators. Actually, to the best of
our knowledge, the insertion of a benzyl group at the amide
moiety has never been studied in terms of SAR even in the
countless series of 1 analogues reported in the literature.
The synthesis and the biological evaluation of the novel 4-
phenylquinazoline derivatives 5−23, 39, and 40 reported in the
present study allowed a new 3D pharmacophore model to be
generated and validated, outlining its predictive power.
2415
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
Table 1. TSPO Affinity of Quinazoline Derivatives IIa−j, 5−23, 39, and 40
a
N
R1
R2
R3
R4
R5
I (%) (1 μM) or Ki (nM)
b
IIa
IIb
(CH2)3CH3
(CH2)4CH3
(CH2)5CH3
(CH2)5CH3
(CH2)5CH3
(CH2)5CH3
CH(CH3)CH2CH3
CH(CH3)CH2CH3
CH(CH3)CH2CH3
CH(CH3)CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH2CH3
CH3
(CH2)3CH3
H
H
H
Cl
H
H
H
Cl
H
H
H
H
F
H
H
H
H
H
H
Cl
H
H
H
Cl
H
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
3.30 0.30
2.48 0.21
0.690 0.070
65.1 7.0
3.42 0.30
38.5 4.0
3.00 0.30
3.06 0.30
2.67 0.30
22%
b
(CH2)4CH3
H
H
H
Cl
H
H
H
Cl
H
H
H
H
H
F
b
IIc
(CH2)5CH3
b
IId
(CH2)5CH3
b
IIe
(CH2)5CH3
b
IIf
(CH2)5CH3
b
b
IIg
CH3
CH3
CH3
CH3
H
IIh
b
IIi
IIj
5
b
1.13 0.10
24.3 2.43
0.888 0.090
0.235 0.020
1.68 0.17
0.762 0.080
1.49 0.15
2.57 0.26
2.77 0.27
8.76 0.88
4.20 0.42
3.65 0.37
719 72
6
H
7
H
8
H
Cl
H
H
H
H
H
H
H
H
H
F
9
H
10
11
12
13
14
15
16
17
18
19
20
21
22
23
39
40
1
H
Cl
CH3
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
H
CH3
OCH3
OH
NO2
COOC2H5
COOH
CH3
CH3
H
0.600 0.060
0.489 0.050
3.99 0.40
1.11 0.11
0.442 0.040
6.12 0.61
41%
Cl
H
F
CH3
H
CH3
H
Cl
H
H
Cl
CH3
H
H
H
H
H
16%
9.30 0.50
23.0 3.0
2
a
The concentration of test molecules that inhibited [3H]1 binding to rat kidney mitochondrial membranes by 50% (IC50) was determined with six
concentrations of the compounds, each performed in triplicate. Ki values and inhibition percentages at 1 μM are the mean
determinations. Data taken from ref 30.
SEM of three
b
Furthermore, selected compounds, representative of the whole
data set, were tested for their ability to inhibit the viability of
human glioblastoma cell line U343.
with methyl or ethyl iodide in the presence of sodium hydride
to furnish compounds 5−13, 16, and 18−23 (Scheme 1).
Unfortunately, the same strategy could not be exploited to
obtain derivatives 14 and 15. In fact, treatment of the secondary
amide 48 with ethyl iodide in the presence of sodium hydride
gave only degradation products. Conversely, the desired tertiary
amide 15 was obtained by direct coupling of the activated acid
31 and N-(4-nitrobenzyl)ethanamine (52). This method was
also followed to obtain derivative 51, using N-(4-(tosyloxy)-
benzyl)ethanamine (53). The phenol derivative 14 was
obtained by deprotecting the tosylate 51 with NaOH. Finally,
hydrolysis of the ethyl ester 16 with lithium hydroxide at room
temperature gave the carboxylic acid derivative 17.
RESULTS AND DISCUSSION
■
Chemistry. The key intermediates in the synthesis of target
compounds 5−23 were the carboxylic acids 31−37, which were
prepared by condensation of the appropriate 2-aminobenzo-
phenones 24−30 with glyoxylic acid in the presence of
ammonium acetate, followed by light irradiation with a 20 W
halogen tungsten lamp, as previously reported.30 After
activation with thionyl chloride, acids 31−37 were coupled
with the appropriate primary amine to give secondary amides
38−50, which were subsequently N-methylated or N-ethylated
Non commercially available 2-aminobenzophenones 24−30
were synthesized according to literature procedures.36 Secon-
2416
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
Figure 1. (a) Representation of the pharmacophore features of the developed model, H-bond acceptor (red sphere A), aromatic rings (orange circles
R1, R2, and R3), custom features (cyan spheres X1 and X2), and excluded volumes (yellow circles). (b) Representation of the developed
pharmacophore model and some of the aligned ligands employed to build the model.
dary amines 52 and 53 were promptly prepared by reductive
alkylation of the appropriate aldehyde with ethylamine using
sodium borohydride (NaBH4) in EtOH (see the Experimental
Section).
best performing element (IIc) of the previously reported series
of N,N-dialkyl-4-phenylquinazolines.30 The favorable effect on
affinity resulting from the simultaneous occupation of the L4
and L3 pockets by an N-benzyl group and a chlorine-
substituted phenyl ring, respectively, contrasts with the SARs
exhibited by derivatives of series II, for which combining a large
alkyl group into L4 and a chlorine on the phenyl into L3 did
not improve the potency (compare IId and IIe with IIc, and
IIh and IIi with IIg).30
Binding Affinity Assays. The binding affinity of all the
newly synthesized 4-phenylquinazoline derivatives 5−23, 39,
and 40 at the TSPO was determined by competition
experiments against [3H]1,30 carried out in rat kidney
membranes. The results, expressed as Ki values, are given in
Table 1, together with the Ki values of the standard TSPO
ligands 1 and 2 (Chart 1). The binding data of some of the
previously investigated quinazoline derivatives30 are included
for comparison purposes at the top of Table 1 (IIa−j).
From a general point of view, it should be noted that most of
the newly synthesized compounds show high affinity for TSPO
in the nanomolar/subnanomolar range, with a potency higher
or comparable to that of previously reported N,N-dialkyl-
substituted derivatives.30 These data suggest that the benzyl
group is effective in engaging a lipophilic interaction within the
L4 hydrophobic pocket similarly to an N-alkyl group with the
number of carbon atoms comprised between 4 and 6.
Furthermore, in agreement with previous findings,30 the
ligand−TSPO interaction requires double substitution at the
carboxamido nitrogen, as the N-monosubstituted quinazolines
tested (39 and 40) are devoid of appreciable affinity.
Compounds 5−19, featuring an ethyl group as the second
substituent on the carboxamido nitrogen, were first inves-
tigated. The unsubstituted derivative 5 (R2 = R3 = R4 = R5 = H)
shows a Ki value in the low nanomolar range (Ki = 1.13 nM),
that is, an affinity slightly worse than that of the most potent
compound of the previously described series II, the N,N-di-n-
hexyl derivative IIc (Ki = 0.69 nM).
The lower affinity of the 6-chloro derivative 6 compared with
the lead 5 is in agreement with our previous findings in series
II, interpreted in terms of unfavorable steric interactions
between the 6-position of the quinazoline ring and the L1
pocket (compare the low affinities of IIf and IIj with those of
IIc and IIg, respectively).
Taken together, all the binding data so far obtained seem to
confirm, as stated in our previous paper,30 that mixed results are
associated with the presence and chemical nature of a
substituent at the 2′-position of the pendant 4-phenyl, due to
interdependent (nonadditive) effects of the R2, R3, and R4
substituents on the affinity of a series of congeneric ligands.
To probe the L4 lipophilic pocket, a small number of
substituents at the 4′-position of the benzyl moiety (R2 = CH3,
OCH3, OH, NO2, COOCH2CH3, COOH) were inserted into
the structure of 5 to yield compounds 12−17. These
substitutions did not cause any significant improvement in
affinity. In this subset it is worth outlining the dramatic drop in
affinity caused by the ionizable carboxylic group featured by 17.
These data suggest that the L4 pocket is mostly lipophilic and
therefore does not tolerate the negative charge carried by a
carboxylate. Moreover, this pocket is not sterically demanding
as it is capable of accommodating the relatively large
COOC2H5 group featured by 16. Furthermore, the insertion
of an electron-withdrawing substituent at the 2′-position of
compound 12 afforded compounds 18 (R3 = F) and 19 (R3 =
Cl) with improved TSPO binding affinity [Ki = 0.600 nM (18)
and Ki = 0.489 nM (19) vs Ki = 2.57 nM (12)], confirming the
SAR trend displayed by derivatives 7 and 8 with respect to 5.
The N-methyl-N-benzyl derivatives 20−23 showed high
affinity in the low nanomolar range (Ki values from 0.442 to
6.12 nM). Also, within this subset, the presence of a chlorine
atom at 2′-position yielded the most active compound 22 (Ki =
0.442 nM). Nonetheless, a pairwise comparison of derivatives
20−23 with their corresponding ethyl analogues highlighted
that the replacement of the N-ethyl with an N-methyl
substituent does not produce any significant improvement in
affinity. Therefore, this class was not further investigated.
Due to the well-established selectivity of N,N-dialkyl-4-
phenylquinazoline-2-carboxamides for TSPO over the central
benzodiazepine receptor (BzR),30 only a few among the novel
quinazoline derivatives (randomly selected) were evaluated for
their BzR affinity by means of experiments against [3H]-
flumazenil in rat cerebral cortex membranes.30 None of the
compounds tested showed appreciable binding properties in
Insertion of small substituents at the 2′-position (F, Cl) or
4′-position (F, Cl, CH3) of the pendant 4-phenyl ring of 5
yielded 7−11, which are provided with a similar or improved
affinity (Ki values ranging from 0.235 to 1.68 nM). The most
beneficial effect in this subset was produced by the presence of
a chlorine at the 2′-position (8) and, to a minor extent, at the
4′-position (10). Compound 8 (Ki = 0.235 nM) turned out to
be the most potent of the whole series, with a 5-fold gain in
affinity with respect to its unsubstituted derivative 5 (Ki = 1.13
nM). Such a compound is also slightly more potent than the
2417
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
this assay (data not shown), thus confirming their selectivity for
TSPO.
ranked compounds is reported in Table S3 in the Supporting
Information). Test set 2 includes 3979 compounds (18 actives
and 3961 putative inactives, i.e., decoys), with an active content
equal to 0.45%, which mimics real-life screening situations.47,48
In this test, the 3.0−2.0 fitness range is only populated by active
ligands (16 out of 18), while the promiscuous fitness region
(range 2.0−1.0) is made up by 2 actives and 8 decoy ligands
(Figure S1, Supporting Information). Finally, in the 1.0−0.0
range of the fitness score only one decoy is ranked, and all the
other compounds do not match the pharmacophore model (the
complete list of the ranked compounds is reported in Table S4
in the Supporting Information).
Computational Studies. The TSPO is an integral
transmembrane (TM) domain which is organized in a five-
TM bundle.37,38 This protein still represents a challenging
target for crystallographers, and to date, no X-ray crystal
structures for TSPO have been disclosed. Also, the absence of
X-ray structures for closely related proteins makes the
construction of a 3D TSPO homology model an arduous
attempt.39,40 On the contrary, in the literature different classes
of TSPO ligands were reported,19,24,30,33,41,42 and in this
context, ligand-based approaches have proven to be useful in
shedding light on the structural requirements necessary to
obtain strong TSPO binders and in rationalizing the activity
profile of different TSPO ligand classes.43,44
Herein, with the aim of better clarifying the structural
features underlying the activity profile of the newly synthesized
compounds (Table 1), we generated a 3D pharmacophore
model. The pharmacophore hypothesis and subsequent ligand
alignment were carried out using the PHASE suite available
within the Maestro package of Schroedinger 9.1,45 on the basis
of the SARs highlighted in previously reported works for
structurally related TSPO ligands tested under similar
experimental conditions (Table S1, Supporting Informa-
tion).30,33
This validation outlined the predictive power of the
presented pharmacophore model, thus encouraging its employ-
ment to better address SAR data presented by the new TSPO
ligands (Table 1). As depicted in Figure 2, the benzyl group of
As reported in Figure 1a, the resulting model consists of one
H-bond acceptor (red sphere A), three aromatic rings (orange
circles R1, R2, and R3), and two custom features accounting for
the disubstituted amide (cyan spheres X1 and X2). Two
excluded volumes were manually added (Figure 1, yellow
spheres): (i) one excluded volume was added in close proximity
to the 6-position of the quinazoline ring, since the insertion of
substituents in this position is detrimental for TSPO ligand
binding (Figure 1, yellow sphere);30 (ii) the second excluded
volume was added in proximity to the disubstituted
carboxamide to account for the low binding affinity of five-
and six-membered ring amides (Figure 1, yellow sphere).30
For the best pharmacophore variant, 1 is the reference ligand
(Figure 1a; see the subsection “Computational Studies” in the
Experimental Section for a full explanation). Interestingly, 1
adopts a conformation characterized by the OCN< group
lying out of the plane of the quinazoline ring (Figure 1a, torsion
angle ∼30°), in agreement with previous studies.30,46 Moreover,
according to the same works, the N-methyl groups adopts a
cisoid conformation with respect to the OC amide group.
The above-described conformational features are shared by all
the aligned active ligands used to generate the 3D
pharmacophore (Figure 1b). Validation of this 3D pharmaco-
phore model was achieved by screening two different test sets
(Tables S2 and S3, Supporting Information). Both test sets
were screened using the Find Matches tool available in
PHASE.45
Test set 1 is made up of 57 compounds (37 actives and 20
inactives/poorly actives). In Figure S1 (Supporting Informa-
tion) the performance of the 3D pharmacophore model is
reported as percentages of the screened database falling into
three calculated fitness score ranges. Interestingly, of the best
31 scoring ligands (range 3.0−2.0), 30 are true actives, while
the 2.0−1.0 fitness score range represents a promiscuous
scoring region including both actives (7) and inactives (8). No
compounds are ranked in the 1.0−0.0 fitness range, while 11
compounds (all inactive) are not ranked as they do not match
the pharmacophore hypothesis at all (the complete list of the
Figure 2. Alignment of compound 8 (green) to the pharmacophore
model and to the reference ligand 1 (orange). It can be appreciated
that the benzyl group extends beyond the region occupied by the less
bulky sec-butyl chain of 1.
5−23, exemplified by compound 8, extends toward the region
occupied by the less bulky sec-butyl chain of 1, suggesting that
the hosting receptor L4 pocket is roomier than postulated from
previous studies,30,33 thus accounting for the generally high
affinity of this new series. Additionally, such a cleft should
feature a rather lipophilic nature as demonstrated by the poor
affinity of the negatively charged 17 compared to the neutral
compounds (12, 13, and 16).
To probe whether the different nature and position of the
substituents at the pendant 4-phenyl ring might favor a specific
conformation of these compounds, a QM energy scan for the
N3−C4−C1′−C2′ dihedral angle (herein defined as ω) of 5, 7,
8, and 10 (unsubstituted and 2′-F-, 2′-Cl-, and 4′-Cl-
substituted, respectively) was performed. Plotting ω vs the
calculated energy (kcal/mol) allowed demonstration that, as
expected, the 4′-substitution has a minor influence on the
dihedral energy landscape with a calculated ω rotational energy
barrier (occurring at ω = 180°) of 5.29 and 4.99 kcal/mol for 5
and 10, respectively, and a minimum energy conformer
featuring an ω value of 220° for both compounds (see Figure
3a,b). In this respect, the slightly higher affinity of 10 compared
to 5 should be ascribed to additional interactions induced by
the chlorine substituent at the 4′-position rather than to
conformational effects. On the contrary, the same calculations
on 7 and 8 revealed that the 2′-substitution influences the
conformational behavior of these compounds (Figure 3c,d). In
particular, the presence of a 2′-Cl atom in 8 forces the lowest
energy ω angle to have a value of 240°, with an energy
difference from the highest energy angle (180°) of 13.28 kcal/
2418
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
Figure 3. Dihedral energy scan plot for compounds 5 (a), 10 (b), 8 (c), and 7 (d). For each compound the ligand maximum energy conformers (*)
and the minimum energy conformers (1 and/or 2) are depicted as insets.
mol (Figure 3c). In this respect, we are tempted to postulate
that the bulky 2′-Cl atom, at least for this N-benzyl-substituted
4-phenylquinazoline-2-carboxamide series, induces a specific
conformation that is more similar to the putative bioactive one,
thus accounting for the enhanced affinity at TSPO. As reported
in Figure 3d, the influence of the 2′-F atom (7) on the
conformational energy landscape of ω is somehow less evident
than that of 2′-Cl. Most specifically, in this case both the ω
values of 220° and 240° are equally stable (two minima), and
the difference from the maximum energy value (0°) is 9.41
kcal/mol.
described in several cellular systems.50−53 Initially, the optimal
membrane protein content of 30 μg was determined and used
for all subsequent binding experiments. Then the specific [3H]1
binding to U343 cell membranes was detected, and the [3H]1
equilibrium binding parameters (dissociation constant, Kd;
maximum number of binding sites, Bmax) were determined by
Scatchard analysis of saturation binding data, a representative
example of which is shown in Figure 4.
According to this hypothesis, the less marked influence of the
smaller 2′-F atom in 7 in inducing a specific conformation
would explain why this compound features an affinity lower
than that of 8 and slightly higher than that of 5.
On the whole, molecular modeling studies seem to indicate
that the coexistence of the 2′-substitution and the benzyl group
on the carboxamide moiety might have a positive effect in
favoring a specific binding pose in the TSPO binding site,
resulting in an enhanced affinity. These observations should be
further validated by experimental studies (i.e., X-ray) and/or
docking calculations/molecular dynamic simulations, which are
not viable at this time in the absence of a thoroughly validated
theoretical receptor model.
Biological Studies on the U343 Glioblastoma Multi-
forme Cell Line. In our previous paper, compound IIc was
shown to be effective in inhibiting U87MG cell viability.30
However, as glioblastoma multiforme (GBM) tumors are
molecularly heterogeneous and, at present, it is not known
which cell line can be the best model for such a heterogeneity,49
we decided to turn our attention to investigate U343 GBM cells
(U-343 MGa, CLS Cell Lines Service), another human GBM
cell line derived from a cancer of a Caucasian male. This cell
line is a monolayer with epithelial morphology and a doubling
time of 1.5 days.
Figure 4. [3H]1 saturation curve and Scatchard plot in U343 cells.
Equilibrium binding parameters (dissociation constant, Kd; maximum
number of binding sites, Bmax) were determined by Scatchard analysis
of saturation binding data, a representative example of which is shown
in the figure.
Specific [3H]1 binding was found to be saturable, whereas
nonspecific binding increased linearly with the radioligand
concentration and was less than 10% of the total binding (data
not shown). Scatchard analysis yielded a single straight-line plot
indicating the presence of a homogeneous population of
First, the presence of TSPO in U343 GBM cells was
investigated by radioligand binding assays using [3H]1, as
2419
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
binding sites. The mean Kd and Bmax values were 8.57 0.25
nM and 6679 280 fmol/mg of proteins, respectively.
compounds, causing a cell viability reduction of barely 40%
with respect to untreated controls (taken as 100%) at the
maximum tested ligand concentration. In summary, com-
pounds IIc and 19 were able to reduce U343 GBM cell viability
in a dose-dependent manner, supporting the specificity of their
effect (see Figure 6).
Then the nature of cell death induced by these TSPO ligands
in U343 cells was evaluated. Depolarization of the mitochon-
drial membrane being the early intracellular event of apoptosis
activated by TSPO ligands, the efficacy of IIc and 19 to induce
mitochondrial membrane potential (Δψm) dissipation was
estimated in U343 GBM cells by the cell fluo-analyzer Muse
(Merck Millipore). The results evidenced a statistically
significant increase in depolarized cells after 5 h of treatment
with 10 μM IIc or 19 (p < 0.005) with respect to the control
(Figure 7). Cells treated with the uncoupling agent carbonyl
Then a representative subset of TSPO ligands (5, 8, 12, and
19) were examined for the effect on the viability of U343 GBM
cells. Compounds IIc and 1 were also tested, for comparison
purpose and as a reference standard, respectively. An initial
screening was performed by exposition of U343 cells to a single
concentration (50 μM) of each compound for 24 h, and cell
viability was quantitatively determined using the colorimetric
viability MTS assay. After the cell exposure with each
compound, the percentage of viable cells was calculated with
respect to control cells (100%), and the results are depicted in
Figure 5. Among the tested compounds, IIc and 19 were
Figure 5. Drug inhibition of cell survival. The cell survival was
evaluated after 24 h of cell treatments with 50 μM compounds. The
percentages of viable cells were calculated with respect to the control
and reported as the mean
SEM of three determinations. The
Bonferroni multiple comparison test showed a significant inhibition in
cell survival for novel compounds.
endowed with the best combination of binding affinity value
and efficacy in inhibiting the viability of U343 GBM cells, being
much more effective than the reference standard 1; thus, they
were further investigated to exclude no specific cytotoxic
effects.
To ascertain whether the observed effect was dose-depend-
ent, cells were exposed to increasing compound concentrations
(from 10 nM to 100 μM) at a single incubation time (24 h).
The classical TSPO ligand 1 was used as a standard. As shown
in Figure 6, the survival assay evidenced a concentration
dependence for the effect produced by the tested compounds.
In particular, this analysis revealed that the concentrations of
IIc and 19 that gave a half-maximal response in survival
inhibition (IC50) were 17.73 and 15.08 μM, respectively. 1
affected cell viability to a much lower extent than the novel
Figure 7. Mitochondrial potential (Δψm) dissipation analyses. The
Δψm dissipation was assessed after 5 h of cell treatments. CCCP was
used as a positive control. The percentages of live, depolarized, and
dead cells for each sample were provided by the instrument and are
reported in the histograms. The Bonferroni multiple comparison test
showed a statistically significant difference in the number of live and
depolarized cells for novel compounds with respect to the control.
cyanide (3-chlorophenyl)hydrazone (CCCP), used as a positive
control, showed the expected increase in depolarization. In
accordance with survival data, 1 affected mitochondrial
depolarization to a lower extent than the novel compounds.
Figure 6. Concentration-dependent survival analyses. The inhibition of cell survival was evaluated after 24 h of cell treatments with increasing
concentrations of compounds. The percentages of viable cells were calculated with respect to the control and reported as the mean SEM of three
determinations. The Bonferroni multiple comparison test showed a significant inhibition in cell survival for novel compounds.
2420
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
These results indicate the intracellular pro-apoptotic mecha-
nism activated following ligand−TSPO binding.
reported as the most involved in GBM, in U343 cells. These
data strongly substantiated the specificity of the inhibition of
U343 GBM cell growth and the intracellular pro-apoptotic
mechanism activated by ligand binding to TSPO, highlighting
TSPO as a potential strategic target for the development of an
alternative therapeutic treatment of GBM.
The ability of 19 to interact with different receptor proteins
or enzymes was assayed to substantiate that the effects elicited
by the compound on U343 GBM cells were specifically due to
its interaction with TSPO. [35S]GTPγS binding experiments
were performed on U343 GBM cell membranes to assess the
compound ability to positively affect G protein-coupled
receptors (GPCRs). The assay measures the binding of the
radiolabeled nonhydrolyzable analogue of GTP to the activated
Gα subunit.54 The results (Figure S2 in the Supporting
Information) showed that 19, tested between 0.1 nM and 100
μM, was not able to increase basal GTPγS binding, suggesting
that this compound did not activate any GPCRs.
EXPERIMENTAL SECTION
■
Chemistry. All chemicals were purchased from Aldrich Chimica
(Milan, Italy) and were of the highest purity. All solvents were reagent
grade and, when necessary, were purified and dried by standard
methods. All reactions requiring anhydrous conditions were conducted
under a positive atmosphere of nitrogen in oven-dried glassware.
Standard syringe techniques were used for anhydrous addition of
liquids. Reactions were routinely monitored by TLC performed on
aluminum-backed silica gel plates (Merck DC, Alufolien Kieselgel 60
F254) with spots visualized by UV light (λ = 254, 365 nm) or using a
KMnO4 alkaline solution. Solvents were removed using a rotary
evaporator operating at a reduced pressure of ∼10 Torr. Organic
solutions were dried over anhydrous Na2SO4. Chromatographic
separations were performed on silica gel (silica gel 60, 0.015−0.040
mm; Merck DC) columns. Melting points were determined on a
Stuart SMP30 melting point apparatus in open capillary tubes and are
uncorrected. Infrared (IR) spectra were recorded neat on a Shimadzu
IR Affinity-1 FTIR instrument, fitted with a MIRacle 10 single-
reflection ATR accessory at room temperature. 1H NMR spectra were
recorded at 300 MHz on a Bruker Avance 300 spectrometer. Chemical
shifts are reported in δ (ppm) relative to the internal reference
tetramethylsilane (TMS). When the amide nitrogen bears two
Furthermore, the ability of compound 19 to target a set of 13
human kinases [Abl kinase, c-kit kinase, EGFR kinase, FLT-1
kinase (VEGFR1), FLT-3 kinase, HER2/ErbB2 kinase, KDR
kinase (VEGFR2), PDGFRα kinase, PDGFRβ kinase, PKCβ1,
PKCβ2, Ret kinase, mTOR kinase (FRAP1)], which has been
reported to be mainly involved in GBM,55,56 was assessed. No
inhibitory activity was detected for 19, when assayed at 1 and
10 μM concentrations (see Figure S3 and Table S5 in the
Supporting Information), thus excluding any involvement of
the tested kinases in the effects of 19 on the viability of U343
cells.
Taken together, these biological results substantiate that the
cytotoxic effects produced by compound 19 on U343 GBM
cells are mediated by a specific interaction with TSPO. Further
studies could be carried out to demonstrate the ability of these
compounds in inhibiting tumor growth in vivo. Actually,
current literature data report that TSPO compounds showing
micromolar in vitro activity were also able to inhibit tumor
growth in vivo.57−60
1
substituents, the H NMR spectra show the presence of two different
rotamers in equilibrium. Copies of these 1H NMR spectra are included
in the Supporting Information. The presence of two different rotamers
in equilibrium was demonstrated for compound 19, taken as
representative of the whole series. 1H NMR spectra of 19 were
recorded at 20 and 80 °C in DMSO-d6, evidencing, at the higher
temperature, coalescence of peaks arising from the benzyl CH2 (data
not shown). Mass spectra were recorded on a Finnigan LCQ DECA
TermoQuest (San Jose, CA) mass spectrometer in electrospray
positive and negative ionization modes (ESI-MS). The purity of the
tested compounds was established by combustion analysis, confirming
a purity ≥95%. Elemental analyses (C, H, N) were performed on a
Perkin-Elmer 2400 CHN elemental analyzer at the laboratory of
microanalysis of the Department of Chemistry and Biology, University
of Salerno (Italy); the analytical results were within 0.4% of the
theoretical values. When the elemental analysis is not included,
compounds were used in the next step without further purification.
General Procedure for the Synthesis of N-Alkyl-N-benzyl-4-
phenylquinazoline-2-carboxamides 5−13, 16, and 18−23.
Sodium hydride (5.50 mmol) was added portionwise, under a
nitrogen atmosphere, to an ice-cooled solution of the appropriate N-
benzyl-4-phenylquinazoline-2-carboxamide 38−50 (5.00 mmol) in dry
DMF (5 mL). The mixture was stirred for 30 min and treated with an
excess of methyl iodide or ethyl iodide (11.0 mmol). The mixture was
stirred for 1 h at 25 °C, and an ice-cooled solution of 1 N HCl and
chloroform were added. The organic layer was separated, washed with
brine, dried, and concentrated in vacuum. Purification by column
chromatography on silica gel (hexane−EtOAc) provided the title
compounds.
CONCLUSION
■
A novel series of N-benzyl-N-ethyl/methyl-4-phenylquinazo-
line-2-carboxamides (5−23) variously decorated at the 6-, 2′-,
4′-, and 4″-positions were synthesized and biologically
evaluated as selective TSPO ligands. Most of the newly
synthesized compounds showed high TSPO binding affinity
with Ki values in the nanomolar/subnanomolar range. SAR data
were addressed by molecular modeling studies, indicating that
the coexistence of the 2′-substitution and the benzyl group on
the carboxamide moiety might have a positive effect in favoring
a specific pose in the TSPO binding site, resulting in an
enhanced affinity. Moreover, the observation that the L4 pocket
is sterically less demanding compared with the L1 and L3
pockets may pave the way for the future development of novel
4-phenylquinazoline-2-carboxamides bearing in the para-
position of the N-benzyl group relatively lengthy R2
substituents as potential fluorescent probes for TSPO, similarly
to that done in series I of (2-phenylindol-3-yl)-
glyoxylamides.25,26
A subset of selected compounds (5, 8, 12, and 19) were
assayed in vitro to determine their ability to induce death in the
U343 GBM cell line. After an initial screening, ligand 19,
showing the best combination of binding affinity and efficacy in
inhibiting the viability of U343 GBM cells, was further
investigated to exclude nonspecific cytotoxic effects. Noticeably,
19 was able to reduce U343 GBM cell viability in a dose-
dependent manner and to cause the dissipation of the
mitochondrial membrane potential. In addition, 19 was
ineffective in activating GPCRs and a set of 13 kinases,
N-Benzyl-N-ethyl-4-phenylquinazoline-2-carboxamide (5). Yield:
89%. Mp: 118−119 °C. 1H NMR (300 MHz, CDCl3, mixture of
rotamers): δ = 8.20−8.13 (m, 2H), 7.96−7.90 (m, 1H), 7.83−7.81 (m,
1H), 7.73−7.26 (m, 10H), 4.88, 4.53 (2s, 2H), 3.60, 3.27 (2q, J = 7.2
Hz, 2H), 1.25, 1.17 (2t, J = 7.2 Hz, 3H) ppm. IR (neat): ν = 1643,
̅
1560, 1544, 1483, 1390 cm−1. MS (ESI): m/z 368 [M + H]+. Anal.
Calcd for C24H21N3O: C, 74.85; H, 5.76; N, 11.44. Found: C, 74.80;
H, 5.76; N, 11.47.
N-Benzyl-N-ethyl-4-(2-fluorophenyl)quinazoline-2-carboxamide
(6). Yield: 92%. Oil. 1H NMR (300 MHz, CDCl3, mixture of
rotamers): δ = 8.22−8.14 (m, 1H), 8.00−7.92 (m, 1H), 7.90−7.82 (m,
2421
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
1H), 7.70−7.25 (m, 10H), 4.88, 4.54 (2s, 2H), 3.59, 3.27 (2q, J = 7.0
7.1 Hz, 2H), 3.60, 3.29 (2q, J = 6.9 Hz, 2H), 1.40, 1.39 (2t, J = 7.1 Hz,
Hz, 2H), 1.25, 1.16 (2t, J = 7.0 Hz, 3H) ppm. IR (neat): ν = 1645,
3H), 1.26, 1.17 (2t, J = 6.9 Hz, 3H) ppm. IR (neat): ν = 1715, 1647,
̅
̅
1562, 1545, 1485, 1389 cm−1. MS (ESI): m/z 386 [M + H]+. Anal.
Calcd for C24H20FN3O: C, 74.79; H, 5.23; N, 10.90. Found: C, 74.86;
H, 5.23; N, 10.88.
1560, 1541, 1485, 1387 cm−1. MS (ESI): m/z 440 [M + H]+. Anal.
Calcd for C27H25N3O3: C, 73.78; H, 5.73; N, 9.56. Found: C, 74.01;
H, 5.71; N, 9.60.
N-Benzyl-4-(2-chlorophenyl)-N-ethylquinazoline-2-carboxamide
(7). Yield: 87%. Mp: 69−70 °C. 1H NMR (300 MHz, CDCl3, mixture
of rotamers): δ = 8.25−8.15 (m, 1H), 8.05−7.95 (m, 1H), 7.70−7.30
(m, 11H), 4.95−4.78, 4.53 (m, s, 2H), 3.64−3.50, 3.27 (m, q, J = 7.1
N-Ethyl-4-(2-fluorophenyl)-N-(4-methylbenzyl)quinazoline-2-car-
1
boxamide (18). Yield: 94%. Mp: 50−51 °C. H NMR (300 MHz,
CDCl3, mixture of rotamers): δ = 8.22−8.15 (m, 1H), 8.00−7.90 (m,
1H), 7.89−7.80 (m, 1H), 7.70−7.49 (m, 4H), 7.40−7.09 (m, 5H),
4.83, 4.48 (2s, 2H), 3.58, 3.24 (2q, J = 6.8 Hz, 2H), 2.35, 2.31 (2s,
Hz, 2H), 1.23, 1.13 (2t, J = 7.1 Hz, 3H) ppm. IR (neat): ν = 1645,
̅
1560, 1543, 1490, 1389 cm−1. MS (ESI): m/z 402 [M + H]+. Anal.
Calcd for C24H20ClN3O: C, 71.73; H, 5.02; N, 10.46. Found: C, 71.73;
H, 5.00; N, 10.44.
3H), 1.24, 1.15 (2t, J = 6.8 Hz, 3H) ppm. IR (neat): ν = 1647, 1560,
̅
1543, 1485, 1389 cm−1. MS (ESI): m/z 422 [M + Na]+. Anal. Calcd
for C25H22FN3O: C, 75.17; H, 5.55; N, 10.52. Found: C, 75.26; H,
5.56; N, 10.55.
N-Benzyl-N-ethyl-4-(4-fluorophenyl)quinazoline-2-carboxamide
(8). Yield: 85%. Mp: 99−100 °C. 1H NMR (300 MHz, CDCl3, mixture
of rotamers): δ = 8.21−8.07 (m, 2H), 7.99−7.91 (m, 1H), 7.88−7.82
(m, 1H), 7.72−7.62 (m, 2H), 7.48−7.42 (m, 2H), 7.40−7.35 (m, 1H),
7.34−7.27 (m, 3H), 7.24−7.19 (m, 1H), 4.87, 4.52 (2s, 2H), 3.60, 3.27
4-(2-Chlorophenyl)-N-ethyl-N-(4-methylbenzyl)quinazoline-2-
1
carboxamide (19). Yield: 91%. Mp: 58−60 °C. H NMR (300 MHz,
CDCl3, mixture of rotamers): δ = 8.23−8.14 (m, 1H), 8.00−7.89 (m,
1H), 7.71−7.05 (m, 10H), 4.84, 4.47 (q, s, J = 14.2 Hz, 2H), 3.60−
3.45, 3.26 (m, q, J = 6.8 Hz, 2H), 2.35, 2.31 (2s, 3H), 1.22, 1.12 (2t, J
(2q, J = 6.8 Hz, 2H), 1.25, 1.16 (2t, J = 6.8 Hz, 3H) ppm. IR (neat): ν
̅
= 1643, 1560, 1541, 1489, 1387 cm−1. MS (ESI): m/z 408 [M + Na]+.
Anal. Calcd for C24H20FN3O: C, 74.79; H, 5.23; N, 10.90. Found: C,
74.58; H, 5.21; N, 10.94.
= 6.8 Hz, 3H) ppm. IR (neat): ν = 1647, 1560, 1543, 1489, 1389 cm−1.
̅
MS (ESI): m/z 438 [M + Na]+. Anal. Calcd for C25H22ClN3O: C,
72.19; H, 5.33; N, 10.10. Found: C, 71.95; H, 5.31; N, 10.13.
N-Benzyl-N-methyl-4-phenylquinazoline-2-carboxamide (20).
N-Benzyl-4-(4-chlorophenyl)-N-ethylquinazoline-2-carboxamide
(9). Yield: 95%. Mp: 117−118 °C. 1H NMR (300 MHz, CDCl3,
mixture of rotamers): δ = 8.21−8.05 (m, 2H), 7.98−7.91 (m, 1H),
7.81−7.78 (m, 1H), 7.70−7.61 (m, 2H), 7.59−7.28 (m, 7H), 4.87,
4.52 (2s, 2H), 3.60, 3.26 (2q, J = 6.4 Hz, 2H), 1.26, 1.16 (2t, J = 6.4
1
Yield: 88%. Mp: 65−68 °C. H NMR (300 MHz, CDCl3, mixture
of rotamers): δ = 8.22−8.13 (m, 2H), 7.99−7.91 (m, 1H), 7.86−7.80
(m, 1H), 7.76−7.53 (m, 5H), 7.49−7.26 (m, 5H), 4.85, 4.52 (2s, 2H),
3.09, 2.94 (2s, 3H) ppm. IR (neat): ν = 1647, 1558, 1541, 1489, 1389
Hz, 3H) ppm. IR (neat): ν = 1636, 1558, 1541, 1478, 1379 cm−1. MS
̅
cm−1. MS (ESI): m/z 354 [M + H]+. Anal. Calcd for C23H19N3O: C,
78.16; H, 5.42; N, 11.89. Found: C, 78.36; H, 5.42; N, 11.85.
N-Benzyl-4-(2-fluorophenyl)-N-methylquinazoline-2-carboxa-
̅
(ESI): m/z 424 [M + Na]+. Anal. Calcd for C24H20ClN3O: C, 71.73;
H, 5.02; N, 10.46. Found: C, 71.56; H, 5.03; N, 10.49.
N-Benzyl-N-ethyl-4-(4-methylphenyl)quinazoline-2-carboxamide
1
mide (21). Yield: 93%. Mp: 52−54 °C. H NMR (300 MHz, CDCl3,
1
(10). Yield: 88%. Mp: 108−109 °C. H NMR (300 MHz, CDCl3,
mixture of rotamers): δ = 8.25−8.13 (m, 1H), 8.02−7.91 (m, 1H),
mixture of rotamers): δ = 8.22−8.10 (m, 2H), 7.97−7.88 (m, 1H),
7.76−7.72 (m, 1H), 7.67−7.59 (m, 2H), 7.50−7.42 (m, 2H), 7.41−
7.35 (m, 2H), 7.35−7.27 (m, 3H), 4.87, 4.52 (2s, 2H), 3.59, 3.26 (2q,
J = 6.5 Hz, 2H), 2.48, 2.47 (2s, 3H), 1.24, 1.16 (2t, J = 6.5 Hz, 3H)
7.89−7.83 (m, 1H), 7.71−7.25 (m, 10H), 4.85, 4.52 (2s, 2H), 3.08,
2.94 (2s, 3H) ppm. IR (neat): ν = 1647, 1560, 1543, 1489, 1389 cm−1.
̅
MS (ESI): m/z 394 [M + Na]+. Anal. Calcd for C23H18FN3O: C,
74.38; H, 4.88; N, 11.31. Found: C, 74.19; H, 4.87; N, 11.33.
N-Benzyl-4-(2-chlorophenyl)-N-methylquinazoline-2-carboxa-
mide (22). Yield: 89%. Mp: 128−129 °C. 1H NMR (300 MHz,
CDCl3, mixture of rotamers): δ = 8.24−8.15 (m, 1H), 8.01−7.92 (m,
1H), 7.73−7.27 (m, 11H), 4.98−4.71, 4.62−4.44 (2m, 2H), 3.07, 2.93
ppm. IR (neat): ν = 1641, 1560, 1543, 1485, 1389 cm−1. MS (ESI):
̅
m/z 382 [M + H]+. Anal. Calcd for C25H23N3O: C, 78.71; H, 6.08; N,
11.02. Found: C, 78.47; H, 6.06; N, 11.06.
N-Benzyl-6-chloro-N-ethyl-4-phenylquinazoline-2-carboxamide
1
(11). Yield: 96%. Mp: 133−134 °C. H NMR (300 MHz, CDCl3,
(2s, 3H) ppm. IR (neat): ν = 1641, 1560, 1543, 1496, 1393 cm−1. MS
̅
mixture of rotamers): δ = 8.15−8.08 (m, 2H), 7.91−7.78 (m, 2H),
7.70−7.65 (m, 1H), 7.63−7.52 (m, 3H), 7.49−7.27 (m, 5H), 4.87,
4.51 (2s, 2H), 3.59, 3.25 (2q, J = 7.2 Hz, 2H), 1.25, 1.17 (2t, J = 7.2
(ESI): m/z 410 [M + Na]+. Anal. Calcd for C23H18ClN3O: C, 71.22;
H, 4.68; N, 10.83. Found: C, 71.13; H, 4.68; N, 10.81.
Hz, 3H) ppm. IR (neat): ν = 1653, 1558, 1541, 1479, 1383 cm−1. MS
N-Benzyl-4-(4-chlorophenyl)-N-methylquinazoline-2-carboxa-
̅
1
mide (23). Yield: 95%. Mp: 61−63 °C. H NMR (300 MHz, CDCl3,
(ESI): m/z 402 [M + H]+. Anal. Calcd for C24H20ClN3O: C, 71.73; H,
5.02; N, 10.46. Found: C, 71.94; H, 5.02; N, 10.45.
mixture of rotamers): δ = 8.22−8.07 (m, 2H), 8.02−7.93 (m, 1H),
7.82−7.28 (m, 10H), 4.85, 4.51 (2s, 2H), 3.10, 2.94 (2s, 3H) ppm. IR
N-Ethyl-N-(4-methylbenzyl)-4-phenylquinazoline-2-carboxamide
(12). Yield: 96%. Oil. 1H NMR (300 MHz, CDCl3, mixture of
rotamers): δ = 8.20−8.11 (m, 2H), 7.98−7.90 (m, 1H), 7.84−7.80 (m,
1H), 7.72−7.69 (m, 1H), 7.68−7.61 (m, 1H), 7.60−7.51 (m, 3H),
7.35, 7.33 (2d, J = 7.8 Hz, 2H), 7.18, 7.11 (2d, J = 7.8 Hz, 2H), 4.83,
4.48 (2s, 2H), 3.58, 3.24 (2q, J = 7.0 Hz, 2H), 2.35, 2.32 (2s, 3H),
(neat): ν = 1647, 1560, 1541, 1483, 1398 cm−1. MS (ESI): m/z 410
̅
[M + Na]+. Anal. Calcd for C23H18ClN3O: C, 71.22; H, 4.68; N, 10.83.
Found: C, 71.07; H, 4.67; N, 10.83.
N-Ethyl-N-(4-hydroxybenzyl)-4-phenylquinazoline-2-car-
boxamide (14). To a solution of 51 (200 mg, 0.37 mmol) in EtOH
(5 mL) was added an aqueous solution of 1 N NaOH (2.00 mL, 2.00
mmol). The mixture was heated at reflux and stirred for 1 h. The
reaction mixture was then cooled to room temperature, treated with 1
N HCl to pH 9, and concentrated in vacuo. The residue was treated
with water (30 mL) and extracted with EtOAc (3 × 30 mL). The
combined organic phases were washed with brine (10 mL), dried
(Na2SO4), filtered, and concentrated in vacuo to yield the title
1.24, 1.16 (2t, J = 7.0 Hz, 3H) ppm. IR (neat): ν = 1645, 1560, 1541,
̅
1485, 1389 cm−1. MS (ESI): m/z 382 [M + H]+. Anal. Calcd for
C25H23N3O: C, 78.71; H, 6.08; N, 11.02. Found: C, 78.45; H, 6.09; N,
11.05.
N-Ethyl-N-(4-methoxybenzyl)-4-phenylquinazoline-2-carboxa-
1
mide (13). Yield: 95%. Oil. H NMR (300 MHz, CDCl3, mixture of
rotamers): δ = 8.20−8.12 (m, 2H), 7.97−7.91 (m, 1H), 7.83−7.80 (m,
1H), 7.74−7.71 (m, 1H), 7.68−7.61 (m, 1H), 7.60−7.53 (m, 3H),
7.40, 7.37 (2d, J = 8.5 Hz, 2H), 6.90, 6.83 (2d, J = 8.5 Hz, 2H), 4.81,
4.45 (2s, 2H), 3.81, 3.79 (2s, 3H), 3.57, 3.24 (2q, J = 6.9 Hz, 2H),
1
compound. Yield: 96%. Mp: 80−81 °C. H NMR (300 MHz, CDCl3,
mixture of rotamers): δ = 8.22−8.10 (m, 2H), 8.00−7.90 (m, 1H),
7.85−7.77 (m, 1H), 7.76−7.50 (m, 5H), 7.39−7.26 (m, 2H), 6.82,
6.77 (2d, J = 8.2 Hz, 2H), 5.37, 5.22 (2s, 1H, exchangeable with
deuterium oxide), 4.78, 4.43 (2s, 2H), 3.56, 3.26 (2q, J = 7.0 Hz, 2H),
1.23, 1.16 (2t, J = 6.9 Hz, 3H) ppm. IR (neat): ν = 1643, 1562, 1543,
̅
1485, 1389 cm−1. MS (ESI): m/z 398 [M + H]+. Anal. Calcd for
C25H23N3O2: C, 75.54; H, 5.83; N, 10.57. Found: C, 75.60; H, 5.83;
N, 10.58.
1.23, 1.17 (2t, J = 7.0 Hz, 3H) ppm. IR (neat): ν = 3238, 1624, 1560,
̅
1516, 1487, 1389 cm−1. MS (ESI): m/z 384 [M + H]+. Anal. Calcd for
C24H21N3O2: C, 75.18; H, 5.52; N, 10.96. Found: C, 75.00; H, 5.50;
N, 11.01.
Ethyl 4-((N-Ethyl-4-phenylquinazoline-2-carboxamido)methyl)-
benzoate (16). Yield: 89%. Oil. 1H NMR (300 MHz, CDCl3, mixture
of rotamers): δ = 8.20−8.08 (m, 2H), 8.07−7.88 (m, 3H), 7.83−7.78
(m, 1H), 7.67−7.50 (m, 7H), 4.92, 4.59 (2s, 2H), 4.38, 4.39 (2q, J =
N-Ethyl-N-(4-nitrobenzyl)-4-phenylquinazoline-2-carboxa-
mide (15). Reaction of carboxylic acid 31 and secondary amine 52,
2422
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
according to the general procedure for the synthesis of N-benzyl-4-
phenylquinazoline-2-carboxamides 5−13, 16, and 18−23, yielded the
2H), 7.46 (d, J = 7.7 Hz, 2H), 2.46 (s, 3H) ppm. IR (neat): ν = 2928,
̅
1718, 1558, 1489, 1387 cm−1. MS (ESI): m/z 263 [M − H]−.
6-Chloro-4-phenylquinazoline-2-carboxylic Acid (37).30 Yield:
1
title compound. Yield: 65%. Mp: 116−117 °C. H NMR (300 MHz,
1
86%. Mp: 213-215 °C (DCM/cyclohexane). H NMR (300 MHz,
CDCl3, mixture of rotamers): δ = 8.27−8.08 (m, 4H), 8.01−7.90 (m,
1H), 7.87−7.81 (m, 1H), 7.72−7.47 (m, 7H), 4.95, 4.65 (2s, 2H),
3.61, 3.35 (2q, J = 7.1 Hz, 2H), 1.29, 1.20 (2t, J = 7.1 Hz, 3H) ppm. IR
DMSO-d6): δ = 8.27 (d, J = 8.9 Hz, 1H), 8.17−8.14 (m, 1H), 8.09−
8.01 (m, 1H), 7.85−7.80 (m, 2H), 7.70−7.63 (m, 3H) ppm. IR
(neat): ν = 3527, 3404, 1718, 1560, 1483, 1381 cm−1. MS (ESI): m/z
(neat): ν = 1647, 1558, 1548, 1508, 1489, 1394, 1339 cm−1. MS (ESI):
̅
̅
283 [M − H]−.
m/z 413 [M + H]+. Anal. Calcd for C24H20N4O3: C, 69.89; H, 4.89; N,
13.58. Found: C, 70.02; H, 4.89; N, 13.60.
General Procedure for the Synthesis of N-Benzyl-4-phenyl-
quinazoline-2-carboxamides 38−50. A solution of the appropriate
2-carboxylic acid 31−37 (2.00 mmol) in thionyl chloride (15 mL) was
refluxed for 2 h under a nitrogen atmosphere. After the solution was
cooled at room temperature, the excess thionyl chloride was removed
at reduced pressure and the crude material dried under vacuum. To
the residue, dissolved in dry THF (10 mL) and cooled to 0 °C, was
added dropwise a mixture of the appropriate primary amine (2.00
mmol) and triethylamine (2.00 mmol) in dry THF (5 mL). The
mixture was stirred at room temperature for 36−48 h (TLC analysis),
filtered, and evaporated. The crude residue was dissolved in DCM (20
mL), washed with 1 N HCl, saturated NaHCO3, and water, dried, and
concentrated in vacuum. Purification by column chromatography on
silica gel (DCM−EtOAc) provided the title compounds.
4-((N-Ethyl-4-phenylquinazoline-2-carboxamido)methyl)-
benzoic Acid (17). To a solution of 16 (200 mg, 0.46 mmol) in a
THF/H2O (4:1) mixture (8 mL) was added LiOH (44.0 mg, 1.80
mmol). The mixture was stirred at room temperature for 12 h and
then treated with HCl (1 N) to pH 7. The solution was extracted with
EtOAc (3 × 30 mL), and the combined organic phases were washed
with brine (10 mL), dried (Na2SO4), filtered, and concentrated in
vacuo. Purification by column chromatography on silica gel (DCM−
EtOAc) provided the title compound. Yield: 74%. Mp: 144−145 °C.
1H NMR (300 MHz, CDCl3, mixture of rotamers): δ = 8.28−8.07 (m,
2H), 8.05−7.87 (m, 3H), 7.84−7.77 (m, 1H), 7.70−7.42 (m, 7H),
4.92, 4.61 (2s, 2H), 3.60, 3.31 (2q, J = 6.9 Hz, 2H), 1.24, 1.17 (2t, J =
6.9 Hz, 3H) ppm. IR (neat): ν = 3433, 1709, 1643, 1562, 1543, 1485,
̅
N-Benzyl-4-phenylquinazoline-2-carboxamide (38). Yield: 82%.
Mp: 232−233 °C. 1H NMR (300 MHz, CDCl3): δ = 8.66 (br t, J = 5.6
Hz, 1H, exchangeable with deuterium oxide), 8.34 (d, J = 8.5 Hz, 1H),
8.18 (d, J = 8.5 Hz, 1H), 8.01−7.97 (m, 1H), 7.81−7.77 (m, 2H),
7.70−7.69 (m, 1H), 7.62−7.57 (m, 3H), 7.44−7.40 (m, 2H), 7.37−
7.32 (m, 2H), 7.31−7.27 (m, 1H), 4.79 (d, J = 5.6 Hz, 2H) ppm. IR
1389 cm−1. MS (ESI): m/z 410 [M − H]−. Anal. Calcd for
C25H21N3O3: C, 72.98; H, 5.14; N, 10.21. Found: C, 72.98; H, 5.14;
N, 10.19.
General Procedure for the Synthesis of 4-Phenylquinazo-
line-2-carboxylic Acids 31−37. An aqueous solution of glyoxylic
acid (10.0 mmol in 10 mL) was added dropwise to a solution of the
appropriate 2-aminobenzophenone 24−30 (10.0 mmol), commercially
available or prepared as previously reported,30 and ammonium acetate
(30.0 mmol) in 96% ethanol (30 mL). The mixture was stirred at
room temperature for 15−30 min. After addition of 20 mL of water,
the orange solid was collected by filtration and washed with 96%
ethanol and diethyl ether to give the corresponding 1,2-dihydro-4-
phenylquinazoline-2-carboxylic acids 31−37, which were dissolved in
DMF (15 mL) and allowed to stand for 12 h under external light
irradiation with a 20 W halogen tungsten lamp. After addition of 15
mL of water, the mixture was cooled in an ice bath, and the yellow
precipitate formed was collected by filtration and washed with water
and diethyl ether to give the 4-phenylquinazoline-2-carboxylic acids
31−37, which were recrystallized from the appropriate solvent.
4-Phenylquinazoline-2-carboxylic Acid (31).30 Yield: 62%. Mp: 98
°C (DCM/cyclohexane). 1H NMR (300 MHz, DMSO-d6): δ = 8.26−
8.23 (m, 1H), 8.16−8.11 (m, 2H), 7.89−7.81(m, 3H). 7.71−7.61 (m,
(neat): ν = 3273, 2926, 1668, 1558, 1522, 1489, 1389 cm−1. MS (ESI):
̅
m/z 340 [M + H]+. Anal. Calcd for C22H17N3O: C, 77.86; H, 5.05; N,
12.38. Found: C, 77.65; H, 5.07; N, 12.34.
N-Benzyl-4-(2-fluorophenyl)quinazoline-2-carboxamide (39).
1
Yield: 76%. Mp: 151−152 °C. H NMR (300 MHz, CDCl3): δ =
8.60 (br t, J = 6.6 Hz, 1H, exchangeable with deuterium oxide), 8.33
(d, J = 8.5 Hz, 1H), 8.04−7.97 (m, 1H), 7.92−7.86 (m, 1H), 7.75−
7.55 (m, 3H), 7.46−7.28 (m, 7H), 4.80 (d, J = 6.0 Hz, 2H) ppm. IR
(neat): ν = 3300, 2928, 1661, 1560, 1514, 1489, 1393 cm−1. MS (ESI):
̅
m/z 358 [M + H]+. Anal. Calcd for C22H16FN3O: C, 74.94; H, 4.51;
N, 11.76. Found: C, 74.05; H, 4.49; N, 11.78.
N-Benzyl-4-(2-chlorophenyl)quinazoline-2-carboxamide (40).
1
Yield: 72%. Mp: 150−151 °C. H NMR (300 MHz, CDCl3): δ =
8.60 (br t, J = 5.7 Hz, 1H, exchangeable with deuterium oxide), 8.34
(d, J = 8.5 Hz, 1H), 8.10−7.89 (m, 1H), 7.76−7.63 (m, 2H), 7.61−
7.39 (m, 5H), 7.40−7.25 (m, 4H), 4.78 (d, J = 5.7 Hz, 2H) ppm. IR
(neat): ν = 3273, 2922, 1684, 1560, 1522, 1497, 1395 cm−1. MS (ESI):
̅
3H) ppm. IR (neat): ν = 3169, 3057, 1751, 1541, 1490, 1346 cm−1.
m/z 396 [M + Na]+. Anal. Calcd for C22H16ClN3O: C, 70.68; H, 4.31;
N, 11.24. Found: C, 70.59; H, 4.32; N, 11.21.
̅
MS (ESI): m/z 249 [M − H]−.
4-(2-Fluorophenyl)quinazoline-2-carboxylic Acid (32). Yield: 65%.
Mp: 203−204 °C (acetonitrile). 1H NMR (300 MHz, DMSO-d6): δ =
8.28−8.23 (m, 1H), 8.20−8.12 (m, 1H), 7.89−7.80 (m, 2H), 7.77−
N-Benzyl-4-(4-fluorophenyl)quinazoline-2-carboxamide (41).
1
Yield: 69%. Mp: 183−184 °C. H NMR (300 MHz, CDCl3): δ =
8.61 (br t, J = 6.0 Hz, 1H, exchangeable with deuterium oxide), 8.33
(d, J = 8.4 Hz, 1H), 8.15 (d, J = 8.4 Hz, 1H), 8.03−7.99 (m, 1H),
7.84−7.79 (m, 2H), 7.76−7.71 (m, 1H), 7.43 (d, J = 7.1 Hz, 2H),
7.38−7.34 (m, 2H), 7.31−7.26 (m, 3H), 4.80 (d, J = 6.0 Hz, 2H) ppm.
7.68 (m, 2H), 7.57−7.45 (m, 2H) ppm. IR (neat): ν = 2930, 2859,
̅
1728, 1566, 1456, 1395 cm−1. MS (ESI): m/z 267 [M − H]−.
4-(2-Chlorophenyl)quinazoline-2-carboxylic Acid (33).30 Yield:
1
IR (neat): ν = 3302, 2930, 1670, 1560, 1514, 1489, 1391 cm−1. MS
68%. Mp: 168−169 °C (acetonitrile). H NMR (300 MHz, DMSO-
̅
(ESI): m/z 358 [M + H]+. Anal. Calcd for C22H16FN3O: C, 73.94; H,
4.51; N, 11.76. Found: C, 74.09; H, 4.50; N, 11.76.
d6): δ = 8.27−8.25 (m, 1H), 8.18−8.13 (m, 1H), 7.87−7.82 (m, 1H),
7.75−7.60 (m, 5H) ppm. IR (neat): ν = 3169, 3066, 1753, 1560, 1473,
̅
1394 cm−1. MS (ESI): m/z 283 [M − H]−.
N-Benzyl-4-(4-chlorophenyl)quinazoline-2-carboxamide (42).
1
4-(4-Fluorophenyl)quinazoline-2-carboxylic Acid (34). Yield: 73%.
Mp: 186−187 °C (acetonitrile). 1H NMR (300 MHz, DMSO-d6): δ =
8.29−8.09 (m, 3H), 7.99−7.81 (m, 3H), 7.59−7.42 (m, 2H) ppm. IR
Yield: 70%. Mp: 181−182 °C. H NMR (300 MHz, CDCl3): δ =
8.60 (br t, J = 6.1 Hz, 1H, exchangeable with deuterium oxide), 8.33
(d, J = 8.4 Hz, 1H), 8.13 (d, J = 8.4 Hz, 1H), 8.01−7.98 (m, 1H),
7.78−7.69 (m, 3H), 7.56 (d, J = 7.8 Hz, 2H), 7.44−7.40 (m, 2H),
7.38−7.33 (m, 2H), 7.32−7.28 (m, 1H), 4.79 (d, J = 3.8 Hz, 2H) ppm.
(neat): ν = 2926, 2855, 1728, 1564, 1491, 1391 cm−1. MS (ESI): m/z
̅
267 [M − H]−.
4-(4-Chlorophenyl)quinazoline-2-carboxylic Acid (35).30 Yield:
IR (neat): ν = 3277, 2914, 1661, 1558, 1513, 1487, 1391 cm−1. MS
̅
1
(ESI): m/z 374 [M + H]+. Anal. Calcd for C22H16ClN3O: C, 70.68; H,
4.31; N, 11.24. Found: C, 70.86; H, 4.31; N, 11.21.
81%. Mp: 197−198 °C (acetonitrile). H NMR (300 MHz, DMSO-
d6): δ = 8.24−8.22 (m, 1H), 8.17−8.12 (m, 2H), 7.90−7.83 (m, 3H),
7.72 (d, J = 7.8 Hz, 2H) ppm. IR (neat): ν = 3523, 3367, 1708, 1593,
N-Benzyl-4-(4-methylphenyl)quinazoline-2-carboxamide (43).
̅
1487, 1404 cm−1. MS (ESI): m/z 283 [M − H]−.
1
Yield: 60%. Mp: 176−177 °C. H NMR (300 MHz, CDCl3): δ =
4-(4-Methylphenyl)quinazoline-2-carboxylic Acid (36). Yield:
90%. Mp: 98−99 °C (acetonitrile). 1H NMR (300 MHz, DMSO-
d6): δ = 8.26−8.08 (m, 3H), 7.90−7.83 (m, 1H), 7.74 (d, J = 7.7 Hz,
8.67 (br t, J = 6.1 Hz, 1H, exchangeable with deuterium oxide), 8.33
(d, J = 8.4 Hz, 1H), 8.20 (d, J = 8.4 Hz, 1H), 8.06−7.96 (m, 1H),
7.74−7.56 (m, 3H), 7.45−7.28 (m, 7H), 4.79 (d, J = 6.0 Hz, 2H), 2.49
2423
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
(s, 3H) ppm. IR (neat): ν = 3292, 2922, 1655, 1560, 1516, 1489, 1387
carboxamides 5−13, 16, and 18−23, yielded the title compound.
Yield: 95%. Mp: 85−87 °C. 1H NMR (300 MHz, CDCl3): δ = 8.20−
8.10 (m, 2H), 8.00−7.90 (m, 1H), 7.85−7.79 (m, 1H), 7.75−7.53 (m,
7H), 7.44−7.26 (m, 4H), 7.01−6.90 (m, 2H), 4.81, 4.48 (2s, 2H),
3.54, 3.26 (2q, J = 7.4 Hz, 2H), 2.45 (s, 3H), 1.20, 1.15 (2t, J = 7.4 Hz,
̅
cm−1. MS (ESI): m/z 376 [M + Na]+. Anal. Calcd for C23H19N3O: C,
78.16; H, 5.42; N, 11.89. Found: C, 78.16; H, 5.44; N, 11.85.
N-Benzyl-6-chloro-4-phenylquinazoline-2-carboxamide (44).
1
Yield: 82%. Mp: 170−171 °C. H NMR (300 MHz, CDCl3): δ =
3H) ppm. IR (neat): ν = 1647, 1558, 1541, 1487, 1373 cm−1. MS
8.60 (br t, J = 6.1 Hz, 1H, exchangeable with deuterium oxide), 8.29
(d, J = 9.0 Hz, 1H), 8.14 (d, J = 2.3 Hz, 1H), 7.92 (dd, J = 9.0 Hz, 2.3
Hz, 1H), 7.80−7.74 (m, 2H), 7.63−7.58 (m, 3H), 7.45−7.28 (m, 5H),
̅
(ESI): m/z 538 [M + H]+. Anal. Calcd for C31H27N3O4S: C, 69.25; H,
5.06; N, 7.82. Found: C, 69.41; H, 5.07; N, 7.80.
N-(4-Nitrobenzyl)ethanamine (52). To a solution of 4-nitro-
benzaldehyde (500 mg, 3.31 mmol) in dry ethanol (10 mL) was added
ethylamine (2 M solution in THF, 1.80 mL, 3.64 mmol) under a
nitrogen atmosphere. The solution was stirred at room temperature
for 12 h and then refrigerated to 0 °C, and NaBH4 was added
portionwise until disappearance of the intermediate imine (TLC
analysis). The reaction mixture, after addition of water (20 mL), was
concentrated in vacuo and extracted with EtOAc (3 × 40 mL). The
combined organic phases were washed with a saturated solution of
NaHCO3 (3 × 20 mL) and brine (10 mL), dried (Na2SO4), filtered,
and concentrated in vacuo to yield the title compound as a pale-yellow
oil, which was used without further purification. Yield: 93%. 1H NMR
(300 MHz, CDCl3): δ = 8.17 (d, J = 8.5 Hz, 2H), 7.50 (d, J = 8.5 Hz,
2H), 3.89 (s, 2H), 2.67 (q, J = 7.1 Hz, 2H), 1.64 (br s, 1H,
exchangeable with deuterium oxide), 1.14 (2t, J = 7.1 Hz, 3H) ppm.
4-((Ethylamino)methyl)phenyl 4-Methylbenzenesulfonate
(53). Reductive alkylation of 4-formylphenyl 4-methylbenzenesulfo-
nate, according to the procedure for the synthesis of 52, yielded the
4.79 (d, J = 6.1 Hz, 2H) ppm. IR (neat): ν = 3287, 2918, 1693, 1560,
̅
1518, 1479, 1385 cm−1. MS (ESI): m/z 374 [M + H]+. Anal. Calcd for
C22H16ClN3O: C, 70.68; H, 4.31; N, 11.24. Found: C, 70.47; H, 4.30;
N, 11.27.
N-(4-Methylbenzyl)-4-phenylquinazoline-2-carboxamide (45).
1
Yield: 72%. Mp: 212−213 °C. H NMR (300 MHz, CDCl3): δ =
8.61 (br t, J = 6.0 Hz, 1H, exchangeable with deuterium oxide), 8.34
(d, J = 8.5 Hz, 1H), 8.18 (d, J = 8.5 Hz, 1H), 8.00−7.96 (m, 1H),
7.80−7.77 (m, 2H), 7.71−7.68 (m, 1H), 7.60−7.56 (m, 3H), 7.31 (d, J
= 7.8 Hz, 2H), 7.16 (d, J = 7.8 Hz, 2H), 4.75 (d, J = 6.0 Hz, 2H), 2.34
(s, 3H) ppm. IR (neat): ν = 3264, 2926, 1659, 1558, 1516, 1489, 1391
̅
cm−1. MS (ESI): m/z 354 [M + H]+. Anal. Calcd for C23H19N3O: C,
78.16; H, 5.42; N, 11.89. Found: C, 78.31; H, 5.43; N, 11.86.
N-(4-Methoxybenzyl)-4-phenylquinazoline-2-carboxamide (46).
1
Yield: 80%. Mp: 188−189 °C. H NMR (300 MHz, CDCl3): δ =
8.59 (br t, J = 6.0 Hz, 1H, exchangeable with deuterium oxide), 8.34
(d, J = 8.4 Hz, 1H), 8.17 (d, J = 8.4 Hz, 1H), 8.00−7.96 (m, 1H),
7.81−7.78 (m, 2H), 7.71−7.67 (m, 1H), 7.60−7.56 (m, 3H), 7.35 (d, J
= 8.2 Hz, 2H), 6.88 (d, J = 8.2 Hz, 2H), 4.72 (d, J = 5.9 Hz, 2H), 3.80
1
title compound as a pale yellow oil. Yield: 96%. H NMR (300 MHz,
CDCl3): δ = 7.69 (d, J = 8.1 Hz, 2H), 7.30 (d, J = 8.1 Hz, 2H), 7.23
(d, J = 8.4 Hz, 2H), 6.92 (d, J = 8.4 Hz, 2H), 3.74 (s, 2H), 2.65 (q, J =
7.1 Hz, 2H), 2.53 (s, 3H), 1.62 (br s, 1H, exchangeable with deuterium
oxide), 1.12 (t, J = 7.1 Hz, 3H) ppm.
Computational Studies. Training Set. Twenty-two isoquinoline
and quinazoline derivatives tested under similar conditions using rat
kidney membrane were selected as the training set (Table S1,
Supporting Information).30 The ligand set, thus generated, consists of
10 active and 12 inactive TSPO ligands. The compounds were built
using the fragment builder tool available in the Maestro package of
Schroedinger 9.145 and minimized employing Macromodel, MMFFs as
the force field, and water as the implicit solvent until a convergence
value of 0.05 kJ/(mol·Å). The computational protocol applied consists
(s, 3H) ppm. IR (neat): ν = 3283, 2930, 1657, 1560, 1514, 1489, 1393
̅
cm−1. MS (ESI): m/z 370 [M + H]+. Anal. Calcd for C23H19N3O2: C,
74.68; H, 5.18; N, 11.37. Found: C, 74.88; H, 5.18; N, 11.34.
Ethyl 4-((4-Phenylquinazoline-2-carboxamido)methyl)benzoate
1
(47). Yield: 81%. Mp: 114−115 °C. H NMR (300 MHz, CDCl3):
δ = 8.73 (br t, J = 6.0 Hz, 1H, exchangeable with deuterium oxide),
8.35 (d, J = 8.0 Hz, 1H), 8.19 (d, J = 8.0 Hz, 1H), 8.05−7.95 (m, 3H),
7.84−7.76 (m, 2H), 7.75−7.67 (m, 1H), 7.64−7.55 (m, 3H), 7.47 (d, J
= 8.5 Hz, 2H), 4.85 (d, J = 6.0, 2H), 4.37 (q, J = 7.1 Hz, 2H), 1.38 (t, J
= 7.1 Hz, 3H) ppm. IR (neat): ν = 3292, 2926, 1717, 1667, 1560,
̅
1522, 1489, 1391 cm−1. MS (ESI): m/z 412 [M + H]+. Anal. Calcd for
C25H21N3O3: C, 72.98; H, 5.14; N, 10.21. Found: C, 72.86; H, 5.14;
N, 10.18.
of the application of 500 steps of the Polak−Ribier
́
e conjugate gradient
N-(4-Nitrobenzyl)-4-phenylquinazoline-2-carboxamide (48).
1
(PRCG) for structure minimizations.
Yield: 70%. Mp: 154−155 °C. H NMR (300 MHz, CDCl3): δ =
Generation of the Pharmacophore. A relevant step, in generating
a pharmacophore model, is represented by the exhaustive sampling of
the conformational space of the ligands under investigation. This
represents a crucial step in our study, since it has been hypothesized
that both isoquinoline and quinazoline derivatives are characterized by
peculiar bioactive conformations at the carboxamide region.30,46
Particularly relevant features related to the carboxamide moiety are
(i) the OCN< moiety lying out of the plane of the aromatic
system and (ii) the smaller substituent of the disubstituted amidic
nitrogen in a cisoid conformation with respect to the OC amide
group.46 Therefore, the 22 compounds were submitted to an extensive
conformational search employing the advanced search method of the
ConfGen tool available in the Maestro package of Schroedinger 9.1.45
Conformations were generated by the torsional search method, using
implicit water as the solvent and MMFFs as the force field. For each
molecule, the maximum number of conformations returned was set
equal to 1000 and the search method was conducted applying the
thorough method, with the number of rotatable bonds equal to 36 and
sampling various amide bond conformations. The maximum number
of ring conformations was set equal to 32, and the energy window for
saving conformations was fixed equal to 200 kJ/mol. Finally,
redundant conformations were eliminated if the heavy atom RMSD
was lower than 0.8 Å. The conformations, thus obtained, were
imported into the “Develop Pharmacophore Hypothesis” panel of the
PHASE suite, available in the Maestro package of Schroedinger 9.1.45
The pharmacophore hypothesis and subsequent ligand alignment were
carried out using the PHASE suite.45 On the basis of the ligand
structural features, the following pharmacophore features were used:
8.85 (br t, J = 6.0 Hz, 1H, exchangeable with deuterium oxide), 8.35
(d, J = 8.2 Hz, 1H), 8.25−8.15 (m, 3H), 8.05−7.95 (m, 1H), 7.90−
7.70 (m, 3H), 7.65−7.50 (m, 5H), 4.88 (d, J = 6.3 Hz, 2H) ppm. IR
(neat): ν = 3314, 2930, 1672, 1560, 1513, 1489, 1393, 1346 cm−1. MS
̅
(ESI): m/z 385 [M + H]+. Anal. Calcd for C22H16N4O3: C, 68.74; H,
4.20; N, 14.58. Found: C, 68.95; H, 4.21; N, 14.62.
4-(2-Fluorophenyl)-N-(4-methylbenzyl)quinazoline-2-carboxa-
mide (49). Yield: 73%. Mp: 219−220 °C. 1H NMR (300 MHz,
CDCl3): δ = 8.55 (br t, J = 6.0 Hz, 1H, exchangeable with deuterium
oxide), 8.32 (d, J = 8.4 Hz, 1H), 8.02−7.96 (m, 1H), 7.94−7.81 (m,
1H), 7.77−7.52 (m, 3H), 7.40−7.27 (m, 4H), 7.21 (d, J = 7.8 Hz,
2H), 4.74 (d, J = 6.0 Hz, 2H), 2.34 (s, 3H) ppm. IR (neat): ν = 3277,
̅
2926, 1663, 1558, 1516, 1489, 1393 cm−1. MS (ESI): m/z 394 [M +
Na]+. Anal. Calcd for C23H18FN3O: C, 74.38; H, 4.88; N, 11.31.
Found: C, 74.22; H, 4.87; N, 11.30.
4-(2-Chlorophenyl)-N-(4-methylbenzyl)quinazoline-2-carboxa-
mide (50). Yield: 77%. Mp: 150−152 °C. 1H NMR (300 MHz,
CDCl3): δ = 8.55 (br t, J = 5.8 Hz, 1H, exchangeable with deuterium
oxide), 8.33 (d, J = 8.5 Hz, 1H), 8.08−7.86 (m, 1H), 7.82−7.60 (m,
2H), 7.59−7.42 (m, 4H), 7.31 (d, J = 7.9 Hz, 2H), 7.16 (d, J = 7.9 Hz,
2H), 4.74 (d, J = 5.5 Hz, 2H), 2.34 (s, 3H) ppm. IR (neat): ν = 3321,
̅
2924, 1668, 1565, 1516, 1495, 1391 cm−1. MS (ESI): m/z 410 [M +
Na]+. Anal. Calcd for C23H18ClN3O: C, 71.22; H, 4.68; N, 10.83.
Found: C, 71.30; H, 4.68; N, 10.81.
4-((N-Ethyl-4-phenylquinazoline-2-carboxamido)methyl)-
phenyl 4-Methylbenzenesulfonate (51). Reaction of the carbox-
ylic acid 31 and secondary amine 53, according to the general
procedure for the synthesis of N-benzyl-4-phenylquinazoline-2-
2424
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
(i) hydrogen bond acceptor (A), (ii) hydrogen bond donor (D), (iii)
hydrophobic group (H), (iv) aromatic ring (R), (v) a custom feature X
accounting for substituted amides, and (vi) a custom feature Y, which
corresponds to the combination of hydrophobic groups and aromatic
rings. Common pharmacophores were identified by applying a tree-
based partitioning technique, which groups similar pharmacophores
according to their intersite distances (distance between pairs of sites in
the pharmacophore). The minimum intersite distance was set equal to
2 Å, with a maximum tree depth of 5, an initial box size of 32 Å, and a
final box size of 1 Å. PHASE returns a list of pharmacophore variants,
and for each one the representative ligand (the ligand exactly matching
all the pharmacophore features of the 3D pharmacophore variant) is
reported. The 3D pharmacophore models, thus obtained, were
evaluated on the basis of a scoring procedure, which yielded the
best alignment function of the active compounds. In detail, 40
common pharmacophoric variants were obtained, which were
clustered using the complete linkage method. For this clustering
algorithm the distance between clusters is the largest distance between
any pair of objects (one object from each cluster). Finally, the best
pharmacophore variant was chosen on the basis of the (i) post hoc
scores (this score is the result of rescoring and is a weighted
combination of the vector, site, volume, and selectivity scores) and (ii)
size of the clusters (number of pharmacophore variants which belong
to the cluster).
Assessment of the Pharmacophore. The selected pharmacophore
model was evaluated on the basis of its ability to discriminate between
active and inactive compounds using the Find Matches tool available
in PHASE45 and screening two different test sets. Find Matches is a
panel where the users can specify the database to search and the
pharmacophore hypothesis to use in the search. The search is
performed in two steps: finding matches to the hypothesis and the
scoring of the hits. Ligands matching the pharmacophore are ranked
on the basis of the fitness score (0 ≤ F ≤ 3), which represents a
measure of how well the matching pharmacophore site points align to
those of the hypothesis, how well the matching vector features
(acceptors, donors, aromatic rings, etc.) overlie those of the
hypothesis, and how well the matching conformation superimposes,
in an overall sense, with the reference ligand conformation. The fitness
score is defined by
Biological Evaluation. Materials. Cell culture media and growth
supplements were obtained from Lonza (Milan, Italy). Nonessential
amino acids, N-methyl-N-(1-methylpropyl)-1-(2-chlorophenyl)-3-iso-
quinolinecarboxamide (1), CCCP, and protease inhibitors were from
Sigma-Aldrich srl (Milan, Italy). [3H]1 (specific activity 73.6 Ci/
mmol) was purchased from Perkin-Elmer Italia spa (Monza, Italy).
The Cell Titer 96 Aqueous One Solution Cell Proliferation Assay was
obtained from Promega (Milan, Italy). The Muse Mitopotential Assay
Kit was from Millipore spa (Milan, Italy). All other reagents were from
standard commercial sources.
The experimental protocol for the [35S]GTPγS binding experiments
is reported in the Supporting Information.
The screening of compound 19 on a set of 13 human kinases [Abl
kinase, c-kit kinase, EGFR kinase, FLT-1 kinase (VEGFR1), FLT-3
kinase, HER2/ErbB2 kinase, KDR kinase (VEGFR2), PDGFRα
kinase, PDGFRβ kinase, PKCβ1, PKCβ2, Ret kinase, mTOR kinase
(FRAP1)] was performed by Cerep (France), according to the
company’s standard operating procedures.
Compound Affinity Studies: [3H]1 Binding to Rat Kidney
Mitochondrial Membranes. For affinity binding studies, crude
mitochondrial membranes were incubated with 0.6 nM [3H]1 in the
presence of a compound concentration range of 0.05 nM to 20 μM in
50 mM Tris−HCl, pH 7.4, as previously described.19 For the active
compounds, the IC50 values were determined and Ki values were
derived in accordance with the equation of Cheng and Prusoff.62
U343 GBM Cell culture Conditions. The human U343 GBM cell
line was purchased by the National Institute for Cancer Research
(ICLC, Genoa, Italy) and maintained in standard culture conditions
(37 °C, 95% humidity, 5% CO2) in Eagle’s minimum essential
medium (Earle’s BSS) supplemented with 2 mM L-glutamine, 100 U/
mL penicillin, 100 μg/mL streptomycin, 0.1 mM nonessential amino
acids (NEAAs), 1.0 mM sodium pyruvate, and 10% fetal bovine serum
(FBS).
TSPO Characterization in U343 Cells: [3H]1 Binding Assays. As
previously reported,52 for crude membrane preparation, confluent
U343 cells derived from a 175 cm2 Petri dish were harvested using
phosphate-buffered saline (PBS), pH 7.4, supplemented with 0.04%
EDTA. After cell collection by centrifugation (200g for 5 min), the
pellet was suspended in ∼10 mL of 5 mM ice-cold Tris−HCl buffer,
pH 7.4, containing protease inhibitors (160 μg/mL benzamidine, 200
μg/mL bacitracin, and 20 μg/mL trypsin inhibitor) and homogenized
with an Ultraturrax. Then the homogenate was centrifuged at 48000g
for 15 min at 4 °C, and the supernatant was discarded. The obtained
pellet was suspended in ∼10 mL of Tris−HCl, 50 mM, pH 7.4 (assay
buffer), containing the same amounts of protease inhibitors as
described above, and the homogenate was pelleted by centrifugation
(48000g, 15 min, 4 °C). The pellet was washed once with assay buffer,
and an additional centrifugation step followed (48000g, 15 min, 4 °C).
The resulting cell membrane pellet was suspended at a final
concentration of 1 mg of proteins/mL in assay buffer and used for
binding assays. The protein content of 20 μL of membrane suspension
was measured by the Bradford method63 using the Bio-Rad Protein
Assay reagent, according to the manufacturer’s protocol, with bovine
serum albumin (BSA) used as the standard.
⎛
⎝
⎞
⎠
Salign
Calign
⎜
⎟
⎟
S = W 1 −
+ W Svec + W Svol + WivolSivol
site
vec
vol
⎜
where Salign is the alignment score, Calign the alignment cutoff (1), Wsite
the weight of the site score (1), Svec the vector score, Wvec the weight
of the vector score (1), Svol the volume score, Wvol the weight of the
volume score, Sivol the included volume score, and Wivol the weight of
the volume score. Two test sets were screened using the Find Matches
tool and the selected pharmacophore model: test set 1, which is
composed of 57 isoquinoline and quinazoline derivatives (37 actives
and 20 inactives, Table S2 Supporting Information), and test set 2 (18
TSPO active ligands and 3961 compounds, retrieved from the whole
DUD database).61 It is worth mentioning that test set 2 has an active
content equal to 0.45% that mimics real-life screening situations.47,48
The compounds included in test sets 1 and 2 were prepared by
applying the same protocol described for the training set compounds
and were submitted to the same conformational search methods.
QM Geometry Scan. The Jaguar tool within Schroedinger software
suite 9.145 was used to perform a geometry scan for compounds 5, 7,
8, and 10. Geometry scans are a series of jobs run that vary only the
value of one or more variables used to define an internal or Cartesian
coordinate in the input structure. A preliminary relaxed geometry scan
(without constraints on the ligand geometry) was conducted on
compound 5. In this case the DFT/B3LYP 6-31G** level of theory
was used in the gas phase, and the N3−C4−C1′−C2′ dihedral angle
(herein defined as ω) was varied in 19 steps from 0 to 360°. The
obtained geometrical minimum was used as the starting point to
rebuild compounds 7, 8, and 10. Therefore, compounds 5, 7, 8, and 10
were submitted to a rigid geometry scan allowing the rotation of the ω
dihedral angle.
To determine the presence of specific [3H]1 binding to U343 cell
membrane suspensions, equilibrium radioligand binding assays were
performed. Briefly, different aliquots of U343 cell membranes (10−
100 μg of proteins) were incubated with [3H]1 (1.5 nM) in the
presence (nonspecific binding) or in the absence (total binding) of
unlabeled 1 (1 μM), in a final volume of 500 μL of assay buffer for 90
min at 0 °C. For saturation experiments, aliquots of U343 cell
membranes (30 μg of proteins) were incubated in duplicates with
eight increasing [3H]1 concentrations (0.5−30 nM) in the same
above-described conditions. In each assay, the final vehicle
concentration in the incubation buffer was less than 1% and did not
interfere with specific [3H]1 binding.
TSPO Functional Studies: Compound Cell Treatments. For
functional experiments, U343 cells, grown in flasks or Petri dishes,
were detached by mild trypsinization and counted using the Scepter
2.0 hand-held automated cell counter (http://www.millipore.com/
2425
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
catalogue/item/phcc20060, Millipore, Milan, Italy). Then the cells
were seeded in 96-well plates at a density of ∼4.5 × 103 cells per well
for the proliferation assay or in 6-well plates at a density of ∼60 × 103
cells per well for the apoptosis measurement and allowed to proliferate
for 24 h. Then the culture medium was replaced by complete medium
supplemented with different TSPO ligands or by complete medium
added with DMSO or EtOH (control cells). The cells were incubated
for 24 h (proliferation assay) or 5 h (apotosis assay). In each assay, the
vehicle in which compounds were dissolved never exceeded 1% of the
final assay volume, and we verified that this amount did not affect cell
survival.
Notably for the apoptotic measurement, U343 cells were exposed
for 15 min to the uncoupling agent CCCP (3 μM) as a depolarized
positive control.
Proliferation Analyses by MTS Conversion Assay. The number of
living cells in U343 treated for 24 h was measured by using the
quantitative colorimetric MTS conversion assay as previously
described.52 All measurements were performed in duplicate, and the
experiments were repeated at least three times.
Mitochondrial Membrane Potential Assay. The mitochondrial
depolarization state of treated cells was assessed using the Muse cell
analyzer (Merck Millipore), which is a miniaturized cytometry packing
three-parameter analysis using specific fluorescent dyes. We simulta-
neously measured changes in the mitochondrial membrane potential
(using the red Muse MitoPotential dye) and in cellular plasma
membrane permeabilization (using the Muse MitoPotential 7-AAD
reagent). Briefly, both floating and adherent treated cells were
collected, centrifuged at 300g for 5 min, and suspended in cell culture
medium. Then a 100 μL aliquot of cell suspension (about 20 000
cells/mL) was first added to 95 μL of diluted Muse MitoPotential dye,
and after 20 min at rt, it was added to 5 μL of 7-AAD reagent dye.
After 5 min, the cell suspensions were acquired by a Muse cell
analyzer, and the percentage of live, depolarized, and dead cells was
analyzed in accordance with the Millipore guidelines.
Data Analyses. Scatchard analyses of saturation binding data,
displacement curves, graphic presentation, and statistical analyses were
performed using the nonlinear multipurpose curve-fitting Graph-Pad
Prism computer program (Graph Pad Software, version 5.0; San
Diego, CA). In detail, for saturation binding studies one-site binding
curve fitting was used. IC50 values were derived by semilog plots of
ligand displacement experiment data. The Cheng and Prusoff equation
was used to calculate Ki values.55
ACKNOWLEDGMENTS
■
This study was financially supported by MIUR (PRIN 2009,
Futuro in Ricerca 2010, PRIN 2011), the University of Pisa,
and the University of Salerno. We thank Dr Simona Daniele for
the [35S]GTPγS binding assay.
DEDICATION
■
This paper is dedicated to Prof. Giorgio Stefancich on the
occasion of his 70th birthday.
ABBREVIATIONS USED
■
BzR, central benzodiazepine receptor; 3D, three-dimensional;
GBM, glioblastoma multiforme; MPTP, mitochondrial perme-
ability transition pore; QM, quantomechanical; SAR, struc-
ture−activity relationship; TM, transmembrane; TSPO, 18 kDa
translocator protein
REFERENCES
■
(1) Braestrup, C.; Squires, R. F. Specific benzodiazepine receptors in
rat brain characterized by high-affinity (3H)diazepam binding. Proc.
Natl. Acad. Sci. U.S.A. 1977, 74, 3805−3809.
(2) Braestrup, C.; Albrechtsen, R.; Squires, R. F. High densities of
benzodiazepine receptors in human cortical areas. Nature 1977, 269,
702−704.
(3) Papadopoulos, V.; Baraldi, M.; Guilarte, T. R.; Knudsen, T. B.;
Lacapere, J. J.; Lindemann, P.; Norenberg, M. D.; Nutt, D.; Weizman,
A.; Zhang, M. R.; Gavish, M. Translocator protein (18 kDa): new
nomenclature for the peripheral-type benzodiazepine receptor based
on its structure and molecular function. Trends Pharmacol. Sci. 2006,
27, 402−409.
(4) Bouyer, G.; Cueff, A.; Egee, S.; Kmiecik, J.; Maksimova, Y.;
Glogowska, E.; Gallagher, P. G.; Thomas, S. L. Erythrocyte peripheral
type benzodiazepine receptor/voltage-dependent anion channels are
upregulated by Plasmodium falciparum. Blood 2011, 118, 2305−2312.
(5) Olson, J. M.; Ciliax, B. J.; Mancini, W. R.; Young, A. B. Presence
of peripheral-type benzodiazepine binding sites on human erythrocyte
membranes. Eur. J. Pharmacol. 1988, 152, 47−53.
Statistical analyses were performed by one-way ANOVA with the
Bonferroni post-test, as appropriately specified in the relevant figure
legends. A p value of <0.05 was considered statistically significant. All
data are presented as the mean SEM derived from at least three
independent experiments done in duplicate.
(6) Veenman, L.; Gavish, M. The role of 18 kDa mitochondrial
translocator protein (TSPO) in programmed cell death, and effects of
steroids on TSPO expression. Curr. Mol. Med. 2012, 12, 398−412.
(7) Veenman, L.; Papadopoulos, V.; Gavish, M. Channel-like
functions of the 18-kDa translocator protein (TSPO): regulation of
apoptosis and steroidogenesis as part of the host-defense response.
Curr. Pharm. Des. 2007, 13, 2385−2405.
ASSOCIATED CONTENT
■
(8) Batarseh, A.; Papadopoulos, V. Regulation of translocator protein
18 kDa (TSPO) expression in health and disease states. Mol. Cell.
Endocrinol. 2010, 327, 1−12.
S
* Supporting Information
Tables listing the set of compounds used to generate the 3D
pharmacophore model, test set 1, the ranking for test sets 1 and
2, and the enzyme assay for compound 19, figures showing the
general performance of the 3D pharmacophore model,
[35S]GTPγS binding assays on U343 GBM cells, and the
enzyme assay for compound 19, biological evaluation via the
[35S]GTPγS binding assay, and 1H NMR spectra of compounds
5−23. This material is available free of charge via the Internet
at http://pubs.acs.org.
(9) Galiegue, S.; Casellas, P.; Kramar, A.; Tinel, N.; Simony-
Lafontaine, J. Immunohistochemical assessment of the peripheral
benzodiazepine receptor in breast cancer and its relationship with
survival. Clin. Cancer Res. 2004, 10, 2058−2064.
(10) Han, Z. Q.; Slack, R. S.; Li, W. P.; Papadopoulos, V. Expression
of peripheral benzodiazepine receptor (PBR) in human tumors:
relationship to breast, colorectal, and prostate tumor progression. J.
Recept. Signal Transduction Res. 2003, 23, 225−238.
(11) Hardwick, M.; Fertikh, D.; Culty, M.; Li, H.; Vidic, B.;
Papadopoulos, V. Peripheral-type benzodiazepine receptor (PBR) in
human breast cancer: correlation of breast cancer cell aggressive
phenotype with PBR expression, nuclear localization, and PBR-
mediated cell proliferation and nuclear transport of cholesterol. Cancer
Res. 1999, 59, 831−842.
(12) Katz, Y.; Ben-Baruch, G.; Kloog, Y.; Menczer, J.; Gavish, M.
Increased density of peripheral benzodiazepine-binding sites in ovarian
carcinomas as compared with benign ovarian tumours and normal
ovaries. Clin. Sci. (London) 1990, 78, 155−158.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: +39-0502219547. Fax: +39-0502219605. E-mail:
sabrina.taliani@farm.unipi.it.
Notes
The authors declare no competing financial interest.
2426
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
(13) Maaser, K.; Grabowski, P.; Sutter, A. P.; Hopfner, M.; Foss, H.
D.; Stein, H.; Berger, G.; Gavish, M.; Zeitz, M.; Scherubl, H.
Overexpression of the peripheral benzodiazepine receptor is a relevant
prognostic factor in stage III colorectal cancer. Clin. Cancer Res. 2002,
8, 3205−3209.
(14) Bode, J.; Veenman, L.; Caballero, B.; Lakomek, M.; Kugler, W.;
Gavish, M. The 18 kDa translocator protein influences angiogenesis, as
well as aggressiveness, adhesion, migration, and proliferation of
glioblastoma cells. Pharmacogenet. Genomics 2012, 22, 538−550.
(15) Austin, C. J.; Kahlert, J.; Kassiou, M.; Rendina, L. M. The
translocator protein (TSPO): a novel target for cancer chemotherapy.
Int. J. Biochem. Cell Biol. 2013, 45, 1212−1216.
(16) Le Fur, G.; Vaucher, N.; Perrier, M. L.; Flamier, A.; Benavides,
J.; Renault, C.; Dubroeucq, M. C.; Gueremy, C.; Uzan, A.
Differentiation between two ligands for peripheral benzodiazepine
binding sites, [3H]Ro5-4864 and [3H]PK11195, by thermodynamic
studies. Life Sci. 1983, 33, 449−457.
(17) Takai, N.; Kira, N.; Ishii, T.; Yoshida, T.; Nishida, M.; Nishida,
Y.; Nasu, K.; Takano, M.; Midori, H.; Koga, S.; Narahara, H. A
translocator protein ligand PK11195 shows antigrowth activity in
human choriocarcinoma cells. Tumour Biol. 2012, 33, 1505−1510.
(18) Marangos, P. J.; Patel, J.; Boulenger, J. P.; Clark-Rosenberg, R.
Characterization of peripheral-type benzodiazepine binding sites in
brain using [3H]Ro 5-4864. Mol. Pharmacol. 1982, 22, 26−32.
(19) Primofiore, G.; Da Settimo, F.; Taliani, S.; Simorini, F.; Patrizi,
M. P.; Novellino, E.; Greco, G.; Abignente, E.; Costa, B.; Chelli, B.;
Martini, C. N,N-dialkyl-2-phenylindol-3-ylglyoxylamides. A new class
of potent and selective ligands at the peripheral benzodiazepine
receptor. J. Med. Chem. 2004, 47, 1852−1855.
(20) Chelli, B.; Rossi, L.; Da Pozzo, E.; Costa, B.; Spinetti, F.;
Rechichi, M.; Salvetti, A.; Lena, A.; Simorini, F.; Vanacore, R.; Scatena,
F.; Da Settimo, F.; Gremigni, V.; Martini, C. PIGA (N,N-Di-n-butyl-5-
chloro-2-(4-chlorophenyl)indol-3-ylglyoxylamide), a new mitochon-
drial benzodiazepine-receptor ligand, induces apoptosis in C6 glioma
cells. ChemBioChem 2005, 6, 1082−1088.
(21) Pilkington, G. J.; Parker, K.; Murray, S. A. Approaches to
mitochondrially mediated cancer therapy. Semin. Cancer Biol. 2008, 18,
226−235.
(22) Hess, K. R.; Broglio, K. R.; Bondy, M. L. Adult glioma incidence
trends in the United States, 1977−2000. Cancer 2004, 101, 2293−
2299.
(23) Johannesen, T. B.; Angell-Andersen, E.; Tretli, S.; Langmark, F.;
Lote, K. Trends in incidence of brain and central nervous system
tumors in Norway, 1970−1999. Neuroepidemiology 2004, 23, 101−
109.
(24) Da Settimo, F.; Simorini, F.; Taliani, S.; La Motta, C.; Marini, A.
M.; Salerno, S.; Bellandi, M.; Novellino, E.; Greco, G.; Cosimelli, B.;
Da Pozzo, E.; Costa, B.; Simola, N.; Morelli, M.; Martini, C.
Anxiolytic-like effects of N,N-dialkyl-2-phenylindol-3-ylglyoxylamides
by modulation of translocator protein promoting neurosteroid
biosynthesis. J. Med. Chem. 2008, 51, 5798−5806.
(25) Taliani, S.; Simorini, F.; Sergianni, V.; La Motta, C.; Da Settimo,
F.; Cosimelli, B.; Abignente, E.; Greco, G.; Novellino, E.; Rossi, L.;
Gremigni, V.; Spinetti, F.; Chelli, B.; Martini, C. New fluorescent 2-
phenylindolglyoxylamide derivatives as probes targeting the peripheral-
type benzodiazepine receptor: design, synthesis, and biological
evaluation. J. Med. Chem. 2007, 50, 404−407.
(26) Taliani, S.; Da Pozzo, E.; Bellandi, M.; Bendinelli, S.; Pugliesi, I.;
Simorini, F.; La Motta, C.; Salerno, S.; Marini, A. M.; Da Settimo, F.;
Cosimelli, B.; Greco, G.; Novellino, E.; Martini, C. Novel irreversible
fluorescent probes targeting the 18 kDa translocator protein: synthesis
and biological characterization. J. Med. Chem. 2010, 53, 4085−4093.
(27) Pike, V. W.; Taliani, S.; Lohith, T. G.; Owen, D. R.; Pugliesi, I.;
Da Pozzo, E.; Hong, J.; Zoghbi, S. S.; Gunn, R. N.; Parker, C. A.;
Rabiner, E. A.; Fujita, M.; Innis, R. B.; Martini, C.; Da Settimo, F.
Evaluation of novel N(1)-Methyl-2-phenylindol-3-ylglyoxylamides as a
new chemotype of 18 kDa translocator protein-selective ligand suitable
for the development of positron emission tomography radioligands. J.
Med. Chem. 2010, 54, 366−373.
(28) Kozikowski, A. P.; Ma, D.; Brewer, J.; Sun, S.; Costa, E.; Romeo,
E.; Guidotti, A. Chemistry, binding affinities, and behavioral properties
of a new class of “antineophobic” mitochondrial DBI receptor complex
(mDRC) ligands. J. Med. Chem. 1993, 36, 2908−2920.
(29) Romeo, E.; Auta, J.; Kozikowski, A. P.; Ma, D.; Papadopoulos,
V.; Puia, G.; Costa, E.; Guidotti, A. 2-Aryl-3-indoleacetamides (FGIN-
1): a new class of potent and specific ligands for the mitochondrial
DBI receptor (MDR). J. Pharmacol. Exp. Ther. 1992, 262, 971−978.
(30) Castellano, S.; Taliani, S.; Milite, C.; Pugliesi, I.; Da Pozzo, E.;
Rizzetto, E.; Bendinelli, S.; Costa, B.; Cosconati, S.; Greco, G.;
Novellino, E.; Sbardella, G.; Stefancich, G.; Martini, C.; Da Settimo, F.
Synthesis and biological evaluation of 4-phenylquinazoline-2-carbox-
amides designed as a novel class of potent ligands of the translocator
protein. J. Med. Chem. 2012, 55, 4506−4510.
(31) Matarrese, M.; Moresco, R. M.; Cappelli, A.; Anzini, M.;
Vomero, S.; Simonelli, P.; Verza, E.; Magni, F.; Sudati, F.; Soloviev, D.;
Todde, S.; Carpinelli, A.; Kienle, M. G.; Fazio, F. Labeling and
evaluation of N-[11C]methylated quinoline-2-carboxamides as poten-
tial radioligands for visualization of peripheral benzodiazepine
receptors. J. Med. Chem. 2001, 44, 579−585.
(32) Campiani, G.; Nacci, V.; Fiorini, I.; De Filippis, M. P.; Garofalo,
A.; Ciani, S. M.; Greco, G.; Novellino, E.; Williams, D. C.; Zisterer, D.
M.; Woods, M. J.; Mihai, C.; Manzoni, C.; Mennini, T. Synthesis,
biological activity, and SARs of pyrrolobenzoxazepine derivatives, a
new class of specific “peripheral-type” benzodiazepine receptor ligands.
J. Med. Chem. 1996, 39, 3435−3450.
(33) Cappelli, A.; Anzini, M.; Vomero, S.; De Benedetti, P. G.;
Menziani, M. C.; Giorgi, G.; Manzoni, C. Mapping the peripheral
benzodiazepine receptor binding site by conformationally restrained
derivatives of 1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl)-3-
isoquinolinecarboxamide (PK11195). J. Med. Chem. 1997, 40, 2910−
2921.
(34) Kita, A.; Kohayakawa, H.; Kinoshita, T.; Ochi, Y.; Nakamichi,
K.; Kurumiya, S.; Furukawa, K.; Oka, M. Antianxiety and
antidepressant-like effects of AC-5216, a novel mitochondrial
benzodiazepine receptor ligand. Br. J. Pharmacol. 2004, 142, 1059−
1072.
(35) Kita, A.; Furukawa, K. Involvement of neurosteroids in the
anxiolytic-like effects of AC-5216 in mice. Pharmacol., Biochem. Behav.
2008, 89, 171−178.
(36) Sugasawa, T.; Toyoda, T.; Adachi, M.; Sasakura, K. Amino-
haloborane in organic synthesis. 1. Specific ortho substitution reaction
of anilines. J. Am. Chem. Soc. 1978, 100, 4842−4852.
(37) Korkhov, V. M.; Sachse, C.; Short, J. M.; Tate, C. G. Three-
dimensional structure of TSPO by electron cryomicroscopy of helical
crystals. Structure 2010, 18, 677−687.
(38) Murail, S.; Robert, J. C.; Coic, Y. M.; Neumann, J. M.; Ostuni,
M. A.; Yao, Z. X.; Papadopoulos, V.; Jamin, N.; Lacapere, J. J.
Secondary and tertiary structures of the transmembrane domains of
the translocator protein TSPO determined by NMR. Stabilization of
the TSPO tertiary fold upon ligand binding. Biochim. Biophys. Acta,
Biomembr. 2008, 1778, 1375−1381.
(39) Anzini, M.; Cappelli, A.; Vomero, S.; Seeber, M.; Menziani, M.
C.; Langer, T.; Hagen, B.; Manzoni, C.; Bourguignon, J. J. Mapping
and fitting the peripheral benzodiazepine receptor binding site by
carboxamide derivatives. Comparison of different approaches to
quantitative ligand-receptor interaction modeling. J. Med. Chem.
2001, 44, 1134−1150.
(40) Bernassau, J. M.; Reversat, J. L.; Ferrara, P.; Caput, D.; Lefur, G.
A 3D model of the peripheral benzodiazepine receptor and its
implication in intra mitochondrial cholesterol transport. J. Mol.
Graphics 1993, 11, 236−244.
(41) Taliani, S.; Pugliesi, I.; Da Settimo, F. Structural requirements to
obtain highly potent and selective 18 kDa translocator protein
(TSPO) ligands. Curr. Top. Med. Chem. 2011, 11, 860−886.
(42) Simorini, F.; Marini, A. M.; Taliani, S.; La Motta, C.; Salerno, S.;
Pugliesi, I.; Da Settimo, F. Medicinal chemistry of indolylglyoxylamide
TSPO high affinity ligands with anxiolytic-like effects. Curr. Top. Med.
Chem. 2012, 12, 333−351.
2427
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428

Journal of Medicinal Chemistry
Article
(43) Ortore, G.; Tuccinardi, T.; Martinelli, A. Computational studies
on translocator protein (TSPO) and its ligands. Curr. Top. Med. Chem.
2012, 12, 352−359.
(61) Huang, N.; Shoichet, B. K.; Irwin, J. J. Benchmarking sets for
molecular docking. J. Med. Chem. 2006, 49, 6789−6801.
(62) Cheng, Y.; Prusoff, W. H. Relationship between the inhibition
constant (K1) and the concentration of inhibitor which causes 50%
inhibition (IC50) of an enzymatic reaction. Biochem. Pharmacol. 1973,
22, 3099−3108.
(63) Bradford, M. M. A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the principle
of protein-dye binding. Anal. Biochem. 1976, 72, 248−254.
(44) Tuccinardi, T.; Taliani, S.; Bellandi, M.; Da Settimo, F.; Da
Pozzo, E.; Martini, C.; Martinelli, A. A virtual screening study of the 18
kDa translocator protein using pharmacophore models combined with
3D-QSAR studies. ChemMedChem 2009, 4, 1686−1694.
(45) Schrodinger. Maestro, version 9.1; Schrodinger, LLC: New York,
̈
̈
2009; http://www.schrodinger.com/.
(46) Lee, Y. S.; Simeon, F. G.; Briard, E.; Pike, V. W. Solution
structures of the prototypical 18 kDa translocator protein ligand,
1
PK11195, elucidated with H/13C NMR spectroscopy and quantum
chemistry. ACS Chem. Neurosci. 2012, 3, 325−335.
(47) Polgar, T.; Keseru, G. M. Virtual screening for beta-secretase
(BACE1) inhibitors reveals the importance of protonation states at
Asp32 and Asp228. J. Med. Chem. 2005, 48, 3749−3755.
(48) Truchon, J. F.; Bayly, C. I. Evaluating virtual screening methods:
good and bad metrics for the “early recognition” problem. J. Chem. Inf.
Model. 2007, 47, 488−508.
(49) Shankavaram, U. T.; Bredel, M.; Burgan, W. E.; Carter, D.;
Tofilon, P.; Camphausen, K. Molecular profiling indicates orthotopic
xenograft of glioma cell lines simulate a subclass of human
glioblastoma. J. Cell. Mol. Med. 2012, 16, 545−554.
(50) Gavish, M.; Fares, F. Solubilization of peripheral benzodiaze-
pine-binding sites from rat kidney. J. Neurosci. 1985, 5, 2889−2893.
(51) Costa, B.; Salvetti, A.; Rossi, L.; Spinetti, F.; Lena, A.; Chelli, B.;
Rechichi, M.; Da Pozzo, E.; Gremigni, V.; Martini, C. Peripheral
benzodiazepine receptor: characterization in human T-lymphoma
Jurkat cells. Mol. Pharmacol. 2006, 69, 37−44.
(52) Chelli, B.; Salvetti, A.; Da Pozzo, E.; Rechichi, M.; Spinetti, F.;
Rossi, L.; Costa, B.; Lena, A.; Rainaldi, G.; Scatena, F.; Vanacore, R.;
Gremigni, V.; Martini, C. PK11195 differentially affects cell survival in
human wild-type and 18 kDa translocator protein-silenced ADF
astrocytoma cells. J. Cell. Biochem. 2008, 105, 712−723.
(53) Dell’Osso, L.; Da Pozzo, E.; Carmassi, C.; Trincavelli, M. L.;
Ciapparelli, A.; Martini, C. Lifetime manic-hypomanic symptoms in
post-traumatic stress disorder: relationship with the 18 kDa
mitochondrial translocator protein density. Psychiatry Res. 2010, 177,
139−143.
(54) Harrison, C.; Traynor, J. R. The [35S]GTPγS binding assay:
approaches and applications in pharmacology. Life Sci. 2003, 74, 489−
508.
(55) De Witt Hamer, P. C. Small molecule kinase inhibitors in
glioblastoma: a systematic review of clinical studies. Neuro-Oncology
(Cary, NC, U.S.) 2010, 12, 304−316.
(56) Mellinghoff, I. K.; Schultz, N.; Mischel, P. S.; Cloughesy, T. F.
Will kinase inhibitors make it as glioblastoma drugs? Curr. Top.
Microbiol. Immunol. 2012, 355, 135−169.
(57) Xia, W.; Spector, S.; Hardy, L.; Zhao, S.; Saluk, A.; Alemane, L.;
Spector, N. L. Tumor selective G2/M cell cycle arrest and apoptosis of
epithelial and hematological malignancies by BBL22, a benzazepine.
Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 7494−7499.
(58) Decaudin, D.; Castedo, M.; Nemati, F.; Beurdeley-Thomas, A.;
De Pinieux, G.; Caron, A.; Pouillart, P.; Wijdenes, J.; Rouillard, D.;
Kroemer, G.; Poupon, M. F. Peripheral benzodiazepine receptor
ligands reverse apoptosis resistance of cancer cells in vitro and in vivo.
Cancer Res. 2002, 62, 1388−1393.
(59) Walter, R. B.; Raden, B. W.; Cronk, M. R.; Bernstein, I. D.;
Appelbaum, F. R.; Banker, D. E. The peripheral benzodiazepine
receptor ligand PK11195 overcomes different resistance mechanisms
to sensitize AML cells to gemtuzumab ozogamicin. Blood 2004, 103,
4276−4284.
(60) Shoukrun, R.; Veenman, L.; Shandalov, Y.; Leschiner, S.;
Spanier, I.; Karry, R.; Katz, Y.; Weisinger, G.; Weizman, A.; Gavish, M.
The 18-kDa translocator protein, formerly known as the peripheral-
type benzodiazepine receptor, confers proapoptotic and antineoplastic
effects in a human colorectal cancer cell line. Pharmacogenet. Genomics
2008, 18, 977−988.
2428
dx.doi.org/10.1021/jm401721h | J. Med. Chem. 2014, 57, 2413−2428