Elaborate ligand-based pharmacophore exploration and QSAR analysis guide the synthesis of novel pyridinium-based potent β-secretase inhibitory leads

Article abstract of DOI:10.1016/j.bmc.2010.03.043

β-Secretase (BACE) inhibitors have potential as anti-Alzheimer's disease treatments prompting us to explore the pharmacophoric space of 129 known BACE inhibitors. QSAR analysis was employed to select optimal combination of pharmacophoric models and 2D physicochemical descriptors capable of explaining bioactivity variation (r² = 0.88, F = 60.48, r_LOO² = 0.85, r_PRESS² against 25 external test inhibitors = 0.71). We were obliged to use ligand efficiency as the response variable because the logarithmic transformation of bioactivities failed to access self-consistent QSAR models. Three pharmacophoric models emerged in the successful QSAR equation suggesting at least three binding modes accessible to ligands within BACE binding pocket. QSAR equation and pharmacophoric models were validated through ROC curves and were employed to guide synthesis of novel pyridinium-based BACE inhibitors. The best inhibitor illustrated an IC₅₀ value of 1.0 μM against BACE.

Full text of DOI:10.1016/j.bmc.2010.03.043

Bioorganic & Medicinal Chemistry 18 (2010) 3088–3115
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Elaborate ligand-based pharmacophore exploration and QSAR analysis guide
the synthesis of novel pyridinium-based potent b-secretase inhibitory leads
Afaf Al-Nadaf a, Ghassan Abu Sheikha b, Mutasem O. Taha c,
*
a Department of Pharmaceutical Chemistry, Applied Science University, Amman, Jordan
b Faculty of Pharmacy Al-Zaytoonah Private University of Jordan, Amman, Jordan
c Drug Discovery Unit, Department of Pharmaceutical Sciences, Faculty of Pharmacy, University of Jordan, Amman, Jordan
a r t i c l e i n f o
a b s t r a c t
Article history:
b-Secretase (BACE) inhibitors have potential as anti-Alzheimer’s disease treatments prompting us to
explore the pharmacophoric space of 129 known BACE inhibitors. QSAR analysis was employed to select
optimal combination of pharmacophoric models and 2D physicochemical descriptors capable of explain-
ing bioactivity variation (r2 = 0.88, F = 60.48, r2LOO = 0.85, rP2RESS against 25 external test inhibitors = 0.71).
We were obliged to use ligand efficiency as the response variable because the logarithmic transformation
of bioactivities failed to access self-consistent QSAR models. Three pharmacophoric models emerged in
the successful QSAR equation suggesting at least three binding modes accessible to ligands within BACE
binding pocket. QSAR equation and pharmacophoric models were validated through ROC curves and
were employed to guide synthesis of novel pyridinium-based BACE inhibitors. The best inhibitor illus-
Received 22 January 2010
Revised 16 March 2010
Accepted 17 March 2010
Available online 21 March 2010
Keywords:
b-Secretase inhibitors
Pharmacophore modeling
Ligand-efficiency
Quantitative structure–activity relationship
Receiver-operating characteristic curve
Pyridinium
trated an IC50 value of 1.0 lM against BACE.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
Clearly, the main focus of recent efforts towards the develop-
ment of new BACE inhibitors concentrate on structure-based li-
gand design,15,16 with few ligand-based exceptions.72 To date,
several human BACE X-ray complexes are documented in the Pro-
tein Data Bank (e.g., PDB codes: 3HVG, 3HW1, 2WEZ, 2WF1, 2WF2,
2WF3, 2WF4, 2NVNN, 2VIE, 2VIJ, 2VIY, 2VIZ, 2VJ6, 2VJ7, 2VJ9 res-
olution range: 1.60–2.48 Å).17–23 However, although considered
the most reliable structural information that can be used for drug
design, crystallographic structures are limited by inadequate reso-
lution24 and crystallization-related artifacts of the ligand–protein
complex.25–27 Moreover, crystallographic structures generally
ignore structural heterogeneity related to protein anisotropic mo-
tion and discrete conformational substrates particularly in cases
of pronounced induced-fit protein flexibilities.28
Alzheimer’s disease (AD) is the most common cause of demen-
tia in older people. The progression of AD involves the destruction
of cells that control memory especially in the hippocampus. AD is
thought to be caused by the formation of neuritic plaques of aber-
rantly folded proteins in the brain.1
Fibrils in neuritic plaques consist mainly of amyloid b-protein
(Ab), which is a 40–42 residue-protein generated by cleavage of
the extracellular domain of the transmembrane amyloid precursor
protein (APP) catalysed by the proteases b- and c
-secretase.2
Familial AD shows several mutations at APP secretase cleavage
sites. These seem to affect the pattern of APP processing and lead to
increased production of fibril-forming Ab-peptide.3
It has been shown recently that Ab-peptide increases with age,
The continued interest in designing new BACE inhibitors, com-
bined with the drawbacks of structure-based design and the sig-
nificant induced fit flexibility observed for BACE,29 prompted us
to explore the possibility of developing ligand-based three-dimen-
sional (3D) pharmacophore(s) integrated within self-consistent
QSAR model. This approach avoids the pitfalls of structure-based
techniques; furthermore, the pharmacophore model(s) can be
used as 3D templates to synthesize new BACE inhibitory scaffolds.
We previously reported the use of this innovative approach
towards the discovery of new inhibitory leads against glycogen
synthase kinase 3b,30 hormone sensitive lipase,31 bacterial
MurF,32 protein tyrosine phosphatase 1B33 and influenza neur-
aminidase.34
particularly in AD patients,4,5 prompting significant recent interest
in designing potent b- and
c-secretase inhibitors as potential ther-
apeutic agents against AD.6,7
Several classes of b-secretase (BACE) inhibitors were recently re-
ported, including: peptidomimetic inhibitors,8 imidazolidinone
analogs,9 hydroxymethylcarbonyl isosteres,10 coumarin deriva-
tives,11 flavonoids,12 cyclic ureas13 and acyl guanidines.14 Further-
more,
a recent in silico docking-based screening campaign
identified several diverse new micro molar BACE inhibitors.15
* Corresponding author. Tel.: +962 6 5355000x23305; fax: +962 6 5339649.
E-mail address: mutasem@ju.edu.jo (M.O. Taha).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.03.043

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3089
We employed the HYPOGEN module from the CATALYST software
package35 to construct numerous reasonable binding hypotheses
for BACE inhibitors. Subsequently, genetic function algorithm
(GFA) and multiple linear regression (MLR) analyses were em-
ployed to search for an optimal QSAR that combine high-quality
binding pharmacophores with other molecular descriptors and
capable of explaining bioactivity variation across a collection of di-
verse BACE inhibitors. The optimal pharmacophores were further
validated by evaluating their abilities to successfully classify a list
of compounds as actives or inactives by assessing their receiver-
operating characteristic (ROC) curves. The optimal pharmaco-
phores were employed as guides to synthesize new series of active
BACE inhibitors. The synthesized compounds illustrated low micro
molar potencies.
binding modes within glycogen synthase kinase 3b,30 hormone
sensitive lipase,31 bacterial MurF,32 protein tyrosine phosphatase
1B,33 influenza neuraminidase,34 estrogen receptor b ligands.76
The collected inhibitors were carefully selected in such a way
that they were all bioassayed employing identical assay proce-
dures and conditions to allow proper QSAR modelling. The 2D
structures of the inhibitors were imported into CATALYST and con-
verted automatically into plausible 3D single conformer represen-
tations. The resulting 3D structures were used as starting points for
conformational analysis and in the determination of various
molecular descriptors for QSAR modeling.
The conformational space of each inhibitor was extensively
sampled utilizing the poling algorithm employed within the CON-
49
FIRM module of CATALYST and via the ‘Best’ module to ensure
CATALYST models drug–receptor interaction using information de-
rived only from the drug structure.35–43 HYPOGEN identifies a 3D ar-
ray of a maximum of five chemical features common to active
training molecules, which provides a relative alignment for each
input molecule consistent with their binding to a proposed com-
mon receptor site. The chemical features considered can be hydro-
gen bond donors and acceptors (HBD and HBA), aliphatic and
aromatic hydrophobes (Hbic), positive and negative ionizable (Po-
sIon and NegIon) groups and aromatic planes (RingArom). The con-
formational flexibility of training ligands is modeled by creating
multiple conformers, judiciously prepared to emphasize represen-
tative coverage over a specified energy range. CATALYST pharmaco-
phores have been used as 3D queries for database searching and
in 3D-QSAR studies.44–46,32,47,33
extensive sampling of conformational space. Efficient conforma-
tional coverage guarantees minimum conformation-related noise
during pharmacophore generation and validation stages.49
2.2. Exploration of BACE pharmacophoric space
CATALYST-HYPOGEN enables automatic pharmacophore construction
by using a collection of at least 16 molecules with bioactivities
spanning over 3.5 orders of magnitude.35,39–43 Accordingly, the fact
that we have an informative list of 129 BACE inhibitors of evenly
spread bioactivities over more than 3.5 orders of magnitude
prompted us to employ HYPOGEN algorithm to identify as many phar-
macophoric binding modes assumed by diverse BACE inhibitors as
possible.
HYPOGEN implements an optimization algorithm that evaluates
large number of potential binding models for a particular target
through fine perturbations to hypotheses that survived the con-
structive and subtractive phases of the modeling algorithm (see
Section 4.1.4).39 The number of evaluated pharmacophoric models
is reflected by the configuration (Config.) cost calculated for each
modeling run. It is generally recommended that the Config. cost
of any HYPOGEN run not to exceed 17 (corresponding to 217 hypoth-
eses to be assessed by CATALYST) to guarantee thorough analysis of
all models.40
2. Results and discussion
CATALYST enables automatic pharmacophore construction by
using a collection of molecules with activities ranging over a num-
ber of orders of magnitude. CATALYST pharmacophores (hypotheses)
explain the variability of bioactivity with respect to the geometric
localization of the chemical features present in the molecules used
to build it. Different hypotheses were generated for a series of
BACE inhibitors. A total of 129 compounds were used in this study
(Fig. 1 and Table A in Supplementary data). Six training subsets
were selected from the collection (Table 1). Each subset consisted
of inhibitors of wide structural diversity. The biological activity in
the training subsets spanned from 3.5 to 4.0 orders of magnitude.
Genetic algorithm and multiple linear regression statistical analy-
sis were subsequently employed to select an optimal combination
of complementary pharmacophores capable of explaining bioactiv-
ity variations among all inhibitors.
The size of the investigated pharmacophoric space is a function
of training compounds, selected input chemical features and other
CATALYST control parameters.49 Restricting the extent of explored
pharmacophoric space should improve the efficiency of optimiza-
tion via allowing effective evaluation of limited number of phar-
macophoric models. On the other hand, extensive restrictions
imposed on the pharmacophoric space might reduce the possibility
of discovering optimal pharmacophoric hypotheses, as they might
occur outside the ‘boundaries’ of the pharmacophoric space.
Therefore, we decided to explore the pharmacophoric space of
BACE inhibitors within plausible ‘boundaries’ through 24 HYPOGEN
automatic runs and employing six carefully selected training sub-
sets (i.e., from the collected compounds): subsets A, B, C, D, E,
and F in Table 1. The training compounds in these subsets were se-
lected in such away to guarantee maximal 3D diversity and contin-
uous bioactivity spread over more than 3.5 logarithmic cycles.
Furthermore, the training inhibitors were selected in such a way
that differences in their anti-BACE bioactivities are primarily
attributable to the presence or absence of pharmacophoric features
(e.g., HBA or HBD or Hbic or RingArom) rather than steric shielding
and/or bioactivity-enhancing or -reducing auxiliary groups (e.g.,
electron donating or withdrawing groups). We gave special
emphasis to the 3D diversity of the most-active compounds in each
training subset (Table 1) because of their significant influence on
the extent of the evaluated pharmacophoric space during the con-
structive phase of HYPOGEN algorithm (see Section 4.1.4).
2.1. Data mining and conformational coverage
The literature was surveyed to collect as many structurally di-
verse BACE inhibitors as possible (1–129, see Table A in Supple-
mentary data and Fig. 1).15,48 Although the collected compounds
are quite diverse and hence not expected assume similar binding
modes or bind to the same set of binding pockets, the combination
of pharmacophore modeling and QSAR analysis should allow ac-
cess to all possible binding modes assumed by different inhibitors
within the enzyme. In this approach we employ large list of diverse
inhibitors to generate large number of potential binding hypothe-
ses (240 pharmacophores in this case). Subsequently, the resulting
pharmacophores are allowed to compete within the context of
QSAR analysis. Pharmacophore models that emerge within the
optimal QSAR model(s) (i.e., capable of explaining bioactivity var-
iation across all collected compounds) should cover all binding
modes of diverse training molecules. We previously reported the
use of this innovative approach towards unveiling diverse ligand-
Guided by our reasonably restricted pharmacophoric exploration
concept, we instructed HYPOGEN to explore only 4- and 5-featured

3090
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
R
R
2
1
R
O
1
O
HN
N
N
H
N
N
N
X
N
X
O
O
R
2
R
3
R
1
1-15
16-28
29-58
O
R
R
2
2
S
R
3
N
O
H
N
O
OH
H
N
H
N
H
N
O
O
HN
R
4
O
O
H N
2
R
1
R
1
59-63
64-71
O
N
R
6
O
R
R
O
O
5
O
S
N
HN
O
R
4
1
O
R
3
R
2
O
R
1
HN
Cl
72-76
77-81
S
N
O
NH
O
OH
N
O
O
O
N
N
N
H
HN
N
82
83
O
HN
N
Br
N
N
O
H
N
OH
O
N
H
N
I
I
84
85
H
H
N
N
O
O
F
H
N
O
S
S
NH
F
O
O
O
HN
F
F
S
F
O
F
F
HN
87
86
Figure 1. The chemical scaffolds of training compounds, the corresponding structures and bioactivities are as in Table A in Supplementary data.
pharmacophores, that is, ignore models of lesser number of features
(as shown in Table B in Supplementary data). The later restriction
has the advantage of narrowing the investigated pharmacophoric
space and representing the feature-rich nature of BACE ligands.

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3091
O
S
Br
N
NH
N
O
O
N
S
O
H
NH
N
NH
N
O
Cl
89
88
CH3
H
CH3
CH3
O
H3C
N
N
O
H
N
H
N
N
NH
O
O
S
O
H3C
S
CH3
N
S
H3C
O
90
91
O
C
H
O
N
O
S
HN
N
O
HN
S
S
N
N
H
O
S
O
S
O
NH2
O
NH2
92
93
Cl
H
HN
O
N
Cl
O
3
O
S
F
N
S
O
N
HN
O
HN
CH
O
94
95
N
O
O
O
N
O
CH3
S
NH
NH
O
N
H
O
S
O
HN
N
O
O
96
97
H
H
N
O
N
S
Cl
O
N
O
O
N
S
H3C
HN
O
N
CH3
N
N
N
O
H3C
CH3
CH3
H3C
98
99
Figure 1 (continued)
In each run, the resulting binding hypotheses were automatically
ranked according to their corresponding ‘total cost’ value, which is
defined as the sum of error cost, weight cost and configuration cost
(see Section 4.1.5for details about pharmacophore validation).35,39–43

3092
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
Br
HN
O
N
HN
S
N
N
H
N
N
O
O
Cl
H2N NH
101
100
CH3
N
O
N
O
O
N
HN
HO
O
N
S
CH3
HN
O
Br
N
H
N
O
S
H3C
CH3
H
102
103
O
H
N
O
O
O
N
N
N
O
O
NH
N
H
O
Cl
HN
H3C
CH3
104
105
Cl
HO
N
HN
O
N
O
O
O
N
H2N
N
N
S
O
CH3
O
O
S
N
H
106
H2N
107
OH
O
S
H3C
S
O
H
O
N
O
NH2
S
CH3
O
O
O
O
H3C
108
OH
109
H3C
N
N
NH
H
N
N
Cl
S
O
N
N
O
OH
111
110
Figure 1 (continued)

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3093
OH
CH3
OH
OH
O
Br
O
O
HN
S
O
O
O
N
O
N
Cl
H
112
113
F
H
N
F
F
O
S
Cl
O
O
O
O
N
S
NH
O
O
H3C
NH
N
H
O
S
N
CH3
114
115
O
S
Cl
S
O
O
Br
N
Br
HN
N
O
N
H
N
N
H
116
117
O
H3C
O
O
OH
O
H
N
NH
O
O
N
H
N
O
118
119
HO
O
O
N
H
H
N
HO
O
O
HN
O
O
H C
3
O
CH
3
N
H
O
O
O
H C
3
120
121
O
O
HN
OH
HN
S
NH
O
O
O
HN
O
O
NH
NH
O
O
O
N
S
122
123
S
Figure 1 (continued)
Eventually, 240 pharmacophore models emerged from 24
automatic HYPOGEN runs, out of which only 226 models illustrated
confidence levels P85% (Fisher scrambling criteria, see Section
4.1.5).35,39–43 These successful models were clustered and their best
representatives (45 models, see Section 4.1.6) were used in subse-
quent QSAR modeling. Table C in Supplementary data shows the
statistical criteria of the best representatives, which shared compa-
rable features and acceptable statistical success criteria.
Emergence of several statistically comparable pharmacophore
models suggests the ability of BACE ligands to assume multiple

3094
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
O
N
HO
OH
O
Br
O
H
O
HN
N
H
O
O
H
N
O
NH
N
O
124
125
O
Cl
O
O
NH
O
O
O
Cl
NH
O
NH
O
N
H
O
O
O
O
OH
O
126
127
O
OH
O
O
CH
3
CH
3
N
HO
HN
CH
3
H
N
HN
HO
O
N
H
O
I
O
O
128
129
Figure 1 (continued)
Table 1
Training subsets employed in exploring the pharmacophoric space of BACE inhibitors, numbers correspond to compounds in Table A in Supplementary data and Figure 1
Training sets
Most activea
Moderate active
Least activeb
A
B
C
D
E
F
5, 7, 9, 11, 13
1, 2, 4, 6, 16, 32, 33, 38, 56
53, 58, 98
5, 8, 20, 27, 28
5, 8, 20, 27, 28
7, 8, 14, 19, 27, 28
5, 8, 9, 20, 28
18, 31, 32, 33, 61, 63, 65, 67, 69, 70, 71
31, 32, 61, 63, 65, 69, 67, 70, 71
31, 33, 35, 45, 61, 63, 85, 110
17, 18, 26, 38, 39, 41, 45, 60, 63
1, 3, 17, 26, 30, 31, 43, 53, 55, 56
108, 125, 126
108, 125, 126
123, 129
125, 126
119, 126, 123
7, 8, 13, 14, 27
a
Potency categories as defined by Eqs. 2 and 3.
Potency categories as defined by Eqs. 2 and 3.
b
pharmacophoric binding modes within the binding pocket. There-
fore, it is quite challenging to select any particular pharmacophore
hypothesis as a sole representative of the binding process.
QSAR (GFA-MLR-QSAR) analysis to search for optimal QSAR equa-
tion (s).35
Fit values obtained by mapping the 45 representative hypotheses
against collected BACE inhibitors (1–129) were enrolled together
with a selection of 2D descriptors as independent variables GFA-
MLR-QSAR analysis (see Section 4.1.7).51,52 Unfortunately, all our at-
tempts to achieve self-consistent and predictive QSAR models were
futile prompting us to evaluate an alternative modeling strategy,
namely, to employ ligand efficiency [log (IC50)/MW] as the response
variable instead of activity [log (IC50)]. The fact that many training
compounds exhibited similar potencies despite their wide struc-
tural variations (e.g., 61 training compounds illustrated IC50 values
2.3. QSAR modeling
The predictive value of pharmacophore hypotheses as 3D-QSAR
models is usually limited by steric shielding and bioactivity-
enhancing or reducing auxiliary groups.42 This point combined
with the fact that pharmacophore modeling of BACE inhibitors
furnished numerous binding hypotheses of comparable success
criteria (Table C in Supplementary data) prompted us to employ
classical QSAR analysis to search for the best combination of phar-
macophore (s) and other 2D descriptors capable of explaining
bioactivity variation across the whole list of collected inhibitors
(1–129, Table A in Supplementary data and Fig. 1). We employed
genetic function approximation and multiple linear regression
of 100 lM) prompted us to maximize bioactivity variance through
division(normalization) over molecularweights. This novel strategy
proved successful in achieving self-consistent QSAR models.
To access the predictive power of the resulting QSAR models on
an external set of inhibitors, we randomly selected 25 molecules

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3095
determined for 25 test compounds.50,51 Hypo10/10, Hypo6/18 and
Hypo1/21 represent the fit values of the training compounds (as
calculated from Eq. 5) against the 10th, 6th and 1st pharmacophore
models in 10th, 18th and 21st automatic runs, respectively, as ar-
ranged in Tables B and C under Supplementary data. JursDPSA3 is
the difference in atomic charge-weighted surface areas calculated
by subtracting atomic charge-weighted positive solvent-accessible
surface area minus atomic charge-weighted negative solvent-acces-
sible surface area. Log P is the logarithmic transformation of the oil/
water partition coefficient. AtypeC3 is atom-type-based Alog P
descriptor related to the hydrophobic contribution of carbon atoms
in a particular molecule. SdsCH is the electro-topological state index
of –CH@ atoms. v0 and v2 are the zero and 2nd order molecular
connectivity indices. ShadowXZfrac is a Shadow descriptor related
to the area of molecular shadow in the XZ plane calculated by align-
ing the molecules according to their principal moments of inertia in
the X, Y and Z axes. LUMO is the energy of the lowest unoccupied
molecular orbital calculated by semiempirical quantum mechanical
method (MOPAC).45,46,32,51,53
(marked with asterisks in Table A in Supplementary data) and em-
ployed them as external testing set for validating the QSAR models.
Moreover, all QSAR models were cross-validated automatically
using the leave-one-out cross-validation in CERIUS2.50,51
Eq. 1 shows the details of the optimal QSAR model. Figure 2
shows the corresponding scatter plots of experimental versus esti-
mated bioactivities for the training and testing inhibitors:
Logð1=IC50
Þ
¼ 7:9 ꢀ 10ꢁ5 Hypo1=21 þ 3:0 ꢀ 10ꢁ4 Hypo6=18
MW
þ 1:5 ꢀ 10ꢁ4 Hypo10=10 ꢁ 4:7 ꢀ 10ꢁ5 JursDPSA3
ꢁ 1:4 ꢀ 10ꢁ4 LogP 9:0 ꢀ 10ꢁ4 AtypeC3
þ 3:6 ꢀ 10ꢁ4 SdsCH ꢁ 4:7 ꢀ 10ꢁ4 v0
þ 1:1 ꢀ 10ꢁ3
v
24:5 ꢀ 10ꢁ3 ShadowXZfrac
ꢁ 5:1 ꢀ 10ꢁ4 LUMO ꢁ 3:1 ꢀ 10ꢁ3 r1204 ¼ 0:88;
F ¼ 60:48; n ¼ 104; r2BS ¼ 0:89; rL2OO ¼ 0:85;
r2PRESS ¼ 0:71
ð1Þ
Emergence of three pharmacophoric models in Eq. 1 suggests the
existence of at least three binding modes assumed by inhibitors
within the binding pocket of BACE. Figures 3 and 6 show
Hypo10/10; Hypo6/18 and Hypo1/21 and how they map training
where, r104 is the correlation coefficient against 104 training com-
pounds, r2LOO is the leave-one-out correlation coefficient, rB2S is the
bootstrapping regression coefficient and r2PRESS is the predictive r2
(x 10-3)
10
8
6
4
2
(x 10-3)
0
-10
-8
-6
-4
-2
0
2
4
6
8
10
-2
-4
-6
-8
-10
Experimental Log (IC50/MW)
10 (x 10-3)
8
6
4
2
0
(x 10-3)
10
-10
-8
-6
-4
-2
0
2
4
6
8
-2
-4
-6
-8
-10
Experimental Log (IC50/MW)
Figure 2. Experimental versus fitted (A, 105 compounds, r2LOO= 0.85) and predicted (B, 25 compounds, r2PRESS= 0.71) bioactivities calculated from the best QSAR model (Eq. 1).
The solid lines are the regression lines for the fitted and predicted bioactivities of training and test compounds, respectively, whereas the dotted lines indicate the 2 ꢀ 10ꢁ3
log (IC50)/MW error margins.

3096
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
compound 13 (IC50 = 0.046
l
M) and some synthesized compounds,
2.4. Receiver operating characteristic (ROC) curve analysis
while Table 2 shows the X, Y, and Z coordinates of the three
pharmacophores.
To further validate the resulting models (both QSARs and phar-
macophores), QSAR-selected pharmacophores were subjected to
receiver-operating curve (ROC) analysis. In ROC analysis, the ability
of a particular pharmacophore model to correctly classify a list of
compounds as actives or inactives is indicated by the area under
the curve (AUC) of the corresponding ROC as well as other
parameters: overall accuracy, overall specificity, overall true posi-
tive rate and overall false negative rate (see Section 4.1.8 for more
details). Table 3 and Figure 4 show the ROC performances of our
QSAR-selected pharmacophores.
Emergence of electrotopological, Shadow and connectivity
descriptors in Eq. 1 illustrate certain role played by the ligands’
topology in the binding process. However, despite their predictive
significance, their information content is quite obscure. Emergence
of a hydrophilicity-related descriptor, that is, JursDPSA3, combined
with two hydrophobicity-related descriptors, that is, log P and Aty-
peC3, all in association with negative regression coefficients sug-
gests that optimal ligand/BACE affinity requires certain optimal
hydrophilic/hydrophobic balance.
Finally, emergence of LUMO in Eq. 1 combined with a negative
slope suggests that ligand/BACE affinity favors electrophilic ligands
Clearly from Table 3 and Figure 4 that Hypo10/10 and
Hypo6/18 significantly outperformed Hypo1/21, which is probably
attributed to the absence of a positive ionizable feature from
Hypo1/21 compared to Hypo10/10 and Hypo6/18.
probably due to
p-stacking with certain electron-rich aromatic
centers in the binding pocket.
Figure 3. Pharmacophoric features of QSAR-selected binding models. Light blue spheres represent Hbic features, red spheres represent PosIon features, green-vectored
spheres encode for HBAs, violet vectored spheres encode HBDs, and RingArom as orange vectored spheres. (A) Hypo10/10, (B) Hypo10/10 fitted against 13 (Table A in
Supplementary data and Figure 1, Fit value = 9.9, IC50 = 0.046 lM), (C) Hypo6/18, (D) Hypo6/18, fitted against 13 (Fit value = 9.6), (E) Hypo1/21, and (F) Hypo1/21 fitted
against 13 (Fit value = 10.3).

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3097
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
of the ligands and compares their BACE complexes with the ways
they map Hypo10/10 and Hypo6/18 employing rigid mapping, that
is, fitting the ligands’ bound states against corresponding pharma-
cophores without conformational adjustments.
80
Fitting the hydroxyethyl amine fragment of 2ISO ligand against
HBA and PosIon features in Hypo10/10 ( Fig. 5b) corresponds to
hydrogen-bonding and electrostatic interactions connecting this
fragment with the hydroxy and carboxylate side chains of SER35,
ASP32 and ASP228, respectively, as in Figure 5c. Similarly, mapping
the ligand’s amidic carbonyl with a HBA feature in Hypo10/10 corre-
sponds to hydrogen-bonding interaction connecting this carbonyl
with the hydroxyl side chain of THR232 (Fig. 5b and c). Finally, fitting
the fluorobenzene and benzylic groups of 2ISO ligand against two
Hbic features in Hypo10/10 (Fig. 5b) correlates with stacking the
fluorobenzene and benzylic rings against the aromatic side chains
of TYR14, PHE108 and TRP115, respectively, as in Figure 5c.
Similarly, fitting the benzylic amino of 2IRZ ligand against
PosIon feature in Hypo6/18 (Fig. 5e) corresponds to electrostatic
attraction connecting this group with the carboxylate side chains
of ASP32 and ASP228 (Fig. 5f), while mapping the amidic NH of
the ligand against HBD feature in Hypo6/18 (Fig. 5e) agrees with
hydrogen bonding connecting this group with the peptidic
carbonyl of GLY230 ( Fig. 5f). On the other hand, mapping the
fluorobenzene of 2IRZ ligand against two Hbic features in Hypo6/
18 (Fig. 5e) seems to correspond to hydrophobic interactions tying
this fragment with the hydrophobic side chains of ALA335 and
LEU30, as in Fig. 5f. Finally, the central meta trisubstituted benzene
ring of 2IRZ ligand is shown in Figure 5e to have less-than-optimal
mapping against a central RingArom feature in Hypo6/18, which
seems to correlate with aromatic ring stacking against the adjacent
co-planar amide side chain of GLN73.
0
0.1
0.1
0.1
0.2
0.2
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.8
0.8
0.9
0.9
0.9
1
FALSE POSITIVE RATE
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Clearly from the above discussion, Hypo10/10 and Hypo6/18
represent two valid binding modes assumed by ligands within
BACE. Furthermore, these models point to limited number of criti-
cal interactions required for high ligand-BACE affinity in each of
the binding modes. In contrast, crystallographic complexes reveal
many bonding interactions without highlighting critical ones. Inci-
dentally, Figure 5c and f only show interactions corresponding to
pharmacophoric features while other binding interactions were
hidden for clarity.
0
0.3
0.4
0.5
0.6
0.7
1
FALSE POSITIVE RATE
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
2.6. Synthesis and bioactivities of pharmacophore-guided novel
BACE inhibitors
The fact that pharmacophores Hypo10/10 and Hypo6/18 were
significantly superior to Hypo1/21 vis-à-vis their ROC perfor-
mances and QSAR slopes, prompted us to employ the former mod-
els as templates for building novel BACE inhibitors. We envisaged
that a terminal pyridinium moiety should correspond to the posi-
tive ionizable features in both pharmacophores, furthermore, the
pyridinium ring should provide wide electrophilic surface of pro-
nounced negative LUMO energy to satisfy the requirements of
QSAR Eq. 1.
0
0.3
0.4
0.5
0.6
0.7
1
Incidentally, although positively-charged, and therefore unable
to cross the blood–brain barrier, pyridinium-based drugs can be
devised in the form of uncharged 1,4-dihydro-pyridine surrogate
prodrugs capable of crossing the blood–brain-barrier and get oxi-
dized to pyridinium salts in the brain.70,71
FALSE POSITIVE RATE
Figure 4. Receiver operating characteristic (ROC) curves of QSAR-selected phar-
macophores: (A) Hypo10/10, (B) Hypo6/18 and (C) Hypo1/21.
The spatial arrangement of pharmacophoric features in both
models confined us to meta-disubstituted benzene as linker scaf-
folds in our proposed compounds. Moreover, we proposed mono-
substituted aromatic amide on the other side of the molecules to
correspond with the terminal hydrophobic features in Hypo10/10
and Hypo6/18 (Figs. 3 and 7). We evaluated several electronically
diverse substituents at this side, namely, unsubstituted benzene,
m-methyl, p-methyl, p-chloro, and p-methoxy groups.
2.5. Comparing pharmacophore models with crystallographic
complexes
To further emphasize the validity of our pharmacophore/QSAR
modeling approach, we compared the crystallographic structures
of two BACE/ligand complexes77 (PDB codes: 2ISO and 2IRZ) with
Hypo10/10 and Hypo6/18. Figure 5 shows the chemical structures

3098
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
O
S
O
HN
N
O
F
O
O
NH2
OH
O
HN
S
O
N
F
O
ALA335
GLN73
LEU30
GLY230
N
O
N
ASP32
ASP228
NH2
Figure 5. Co-crystallized ligands of BACE X-ray structures 2ISO (resolution = 2.20 Å) and 2IRZ (resolution = 1.80 Å).77 (A) and (D) the chemical structures of the co-crystallized
ligands of 2ISO and 2IRZ, respectively, (B) and (E) mapping Hypo10/10 and Hypo6/18 against the co-crystallized ligands of 2ISO and 2IRZ (rigid mapping), respectively, (C) and
(F) binding pockets of 2ISO and 2IRZ, respectively, complexed with their corresponding co-crystallized ligands.
Accordingly, we explored the structure-activity profiles of 22
novel structures constructed based on m-aminobenzoic acid or
m-diaminobenzene scaffolds, as shown in Table 4.
the proposed compounds,74 and therefore, provided us with fur-
ther impetus to explore the anti-BACE bioactivities of this novel
class. Tables D and E in the Supplementary data show the differ-
ent descriptor values of collected and synthesized anti-BACE
compounds.
Synthesis commenced by forming the N-phenyl-m-nitrobenza-
mide, N-(m-nitrophenyl)-benzamide and N-(m-nitrophenyl)-benz-
sulfonamide derivatives (Schemes 1–3). Subsequent reduction of
the nitro groups with either sodium dithionite63 or stannous dichlo-
ride64,65 offered the corresponding aromatic amines (Schemes 4 and
5) that were acylated with mono-chloroacetyl chloride (Scheme 6).
Finally, fusion of the resulting chloroacetylated derivatives with
pyridine, nicotinamide or N-phenylethyl nicotinamide offered the
targeted pyridinium derivatives (Scheme 7).
The synthesized compounds were bioassayed employing fluo-
rescent resonance energy transfer (FRET)-based assay.68 The bioas-
say procedure was validated several times employing several
concentrations of the standard BACE inhibitor Lys-Thr-Glu-Glu-
Ile-Ser-Glu-Val-Asn-Statine-Val-Ala-Glu-Phe.75
Although nicotinamide-based pyridinium derivatives (171–178,
Table 4) successfully fitted all 3 pharmacophore models, as clearly
evident fromFigure 7a andTable 4, we were prompted to evaluated
simpler unsubstituted pyridinium derivatives (164–170, Table 4)
to assess the significance of the meta-amide substituents (in nico-
tinamide analogues 171–178) on anti-BACE bioactivity. These
derivatives successfully fit Hypo6/18 and Hypo1/21 and miss a
hydrogen-bond donor feature in Hypo10/10.
Surprisingly, bioassay results show that nicotinamide analogues
(171–178) exhibited generally similar or inferior anti-BACE
profiles compared to their unsubstituted pyridinium analogues,
as seen in Table 4. We believe this odd conduct is related to
hydration-promoting effects produced by the nicotinamide amidic
moieties, which seems to offset their bioactivity gains from fitting
the hydrogen-bond donor feature in Hypo10/10 (Fig. 7a). This point
combined with the observation that sulfonamide derivatives (170,
177, 178, and 185) exhibited generally inferior anti-BACE bioactiv-
ities compared to their amide analogues (167, 174, 176 and 184,
respectively), which seem also to be explainable by the profound
hydration of their sulfonamide linkers, prompted us to tether
hydrophobic phenylethyl side chains to the nicotinamide moieties
in 179–185 to attempt counterbalance hydration. Extensive ligand
hydration can compete, and therefore, hinder ligand-receptor
binding.
Addition of the phenylethyl fragment enhanced the bioactivity
of 184 (% inhibition at 50
unsubstituted analogue 169 (% inhibition at 50
13.0 M) and the nicotinamide analogue 176 (% inhibition at
lM = 100%, IC50 = 2.2
lM) compared to
lM = 68%, IC50
=
l
Clearly from Table 4, our proposed compounds were predicted
to have moderate anti-BACE bioactivities (low micromolar) com-
pared to the nanomolar potencies of some training compounds
(e.g., 13, 11, 12, 5, 14, and 9, Fig. 1 and Table A in Supplemen-
tary data). Still, the fact that pyridinium rings possess LUMO val-
ues outside the corresponding range of training compounds
(Tables D and E in Supplementary data) cast doubt on the ability
of our QSAR equation to accurately predict the bioactivities of
50
50
l
l
M = 24%). A similar trend is seen with 185 (% inhibition at
M = 36%) compared to the nicotinamide analogue 178 (% inhi-
bition at 50 lM = 3.6%). Phenylethyl side chain-related bioactivity
improvements in these cases can also be attributed to extra inter-
actions in the binding site unexplained by the pharmacophore
model and mediated by the phenylethyl side chain (i.e., extra to
their proposed hydration-counterbalance effect mentioned
earlier). In fact, docking studies suggest the existence of mutual

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3099
GLN73, respectively. On the other hand, the electron-rich
methoxybenzene aromatic ring and chloro terminals of 166 and
167 are involved in charge-transfer complex and electrostatic
attraction, respectively, with the quaternary ammonium of
LYS321. These later interactions are not represented in CATALYST
pharmacophoric features, which explain the failure of our models
in predicting these binding modes and the excellent potencies of
166 and 167. The nicotinamide and phenylethyl nicotinamide sub-
stituents in the corresponding analogues 173, 174, 181 and 182
seem to shift the chloro and methoxybenzene away from the qua-
ternary ammonium of LYS321 and therefore disrupt interactions
with LYS321 leading to reduction in anti-BACE bioactivity.
It remains to be mentioned that we validated the selectivity of
our potent derivatives against a closely-related aspartic protease:
renin. Table 5 shows the inhibitory profiles of 166, 167, 169, 174,
179, 181 and 184 against renin. Clearly, the compounds showed
minimal inhibitory profiles against renin albeit with increasing
trend upon association with phenylethyl side chains.
2.7. Comparison of QSAR-selected pharmacophores with
docking into BACE binding site
Pharmacophore features obtained by pharmacophore/QSAR
modeling can be compared with the structure of BACE binding site
to identify probable residues important for ligand binding and
inhibition. Therefore, some of our potent synthesized compounds
were fitted against our QSAR-selected pharmacophores, and the
resulting mapped conformers were compared with docked poses
of these compounds into BACE binding site (PDB code: 2IRZ, reso-
lution 1.8 Å). The docking experiments were performed employing
LIGANDFIT docking engine and through default docking parame-
ters.69 However, we confined the docking simulation to rigid (or
semi-rigid) docking of pharmacophore-fitted conformers to avoid
unnecessary exploration of irrelevant docked conformers, which
should focus the attention on probable binding residues corre-
sponding to features within our QSAR-selected pharmacophores.
The features in Hypo10/10 as well as the alignment of 184
(IC50 = 2.2 lM) as proposed by Hypo10/10 were compared with
the way this compound docks into the binding pocket of BACE,
as in Figure 7a and b. A marked similarity was observed between
the features proposed by the pharmacophore model and the ligand
binding features in the docked structure.
In one of the high-ranking docked poses of 184 (Fig. 7b) the
pyridinium group was placed at close proximity to the carboxylate
of ASP228 in the binding pocket. This interaction corresponds to a
PosIon feature in Hypo10/10 mapping the pyridinium group of
184. Similarly, the docking experiment suggests that the nicotin-
amide NH in 184, interacts with amidic carbonyl of GLY230, which
seems to agree with mapping this group with a HBD feature in
Hypo10/10 (Fig. 7a). The central meta-diaminobenzene ring of
184 is docked adjacent to the aliphatic side chain of THR72, which
correspond nicely to mapping this group against a central Hbic fea-
ture in Hypo10/10. Finally, mapping the terminal toloyl amide
moiety of 184 with HBA and Hbic features in Hypo10/10 corre-
sponds with hydrogen bonding and hydrophobic interactions with
the side chains of LYS321, LEU267 and VAL309, respectively
(Fig. 7a and b).
Figure 6. Optimal docked poses of compounds 166 and 167.
p
-stacking interactions between the phenolic side chain of TYR71
and the phenylethyl side chain (see discussion below, Fig. 6b).
Still, in cases of chloro and methoxy derivatives 181 (% inhibi-
tion at 50
50 M = 44%), respectively, addition of phenylethyl fragments sig-
nificantly reduced their anti-BACE bioactivities compared to their
unsubstituted counterparts 166 (% inhibition at 50 M = 100%,
IC50 = 1.0 M) and 167 (% inhibition at 50 M = 100%, IC50
3.7 M), respectively, and their nicotinamide analogues 173 (%
inhibition at 50 M = 64%, IC50 = 30.2 M) and 174 (% inhibition
at 50 M = 100%, IC50 = 1.9 M). We believe this conduct is because
lM = 78%, IC50 = 12.0 lM) and 182 (% inhibition at
l
l
l
l
=
l
l
l
l
l
Comparably, the docked pose of 174 (IC50 = 1.9 lM) corre-
166 and 167 assume different, and more stable, binding poses from
those predicted by our pharmacophore hypotheses. Figure 6 shows
two optimal binding poses of the two compounds produced
employing LIGANDFIT docking engine via default docking set-
tings.69 Clearly from the figure the pyridinium moieties of 166
and 167 dip in an electron-rich aromatic/anionic pocket composed
of TRP115, PHE108, ASP32, ASP228 and TYR71, while the middle
amidic NHs of 166 and 167 are hydrogen-bonded to GLY230 and
sponds to fitting 174 against binding model Hypo6/18 (Fig. 7c
and d). The pyridinium moiety of 174 apparently electrostatically
interacts with the carboxylates of ASP228 and ASP32, which corre-
sponds to mapping this moiety with the PosIon feature in Hypo6/
18. Similarly, the central m-aminobenzamide linker of 174 is direc-
ted by the docking engine towards the indole of TRP115 suggesting
significant mutual
p-stacking, as in Figure 7d. This agrees with
mapping the m-aminobenzamide of 174 onto a RingArom feature

3100
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
Table 2
Pharmacophoric features and corresponding weights, tolerances and 3D coordinates of Hypo10/10, Hypo 6/18 and Hypo 1/21
Model
Definitions
Chemical features
HBD
HBA
Hbic
Hbic
PosIon
Hypo10/10a
Weights
Tolerances
Coordinates
2.30497
1.60
1.47
2.30497
1.60
2.3049
1.60
ꢁ0.64
ꢁ3.18
2.42
2.30497
1.60
6.81
0.07
ꢁ2.23
2.30497
1.60
2.20
1.09
2.20
X
Y
Z
ꢁ4.28
ꢁ6.60
ꢁ2.59
ꢁ2.04
ꢁ3.83
4.23
4.63
0.96
ꢁ2.63
ꢁ5.01
0.11
ꢁ1.74
ꢁ0.10
HBD
Hbic
Hbic
PosIon
RingArom
Hypo 6/18b
Weights
Tolerances
Coordinates
2.11185
1.60
2.1118
1.60
2.03
0.30
ꢁ1.04
2.11185
1.60
5.70
2.17
ꢁ0.39
2.11185
1.60
2.11185
1.60
2.20
1.60
ꢁ6.06
2.61
0.09
X
Y
Z
ꢁ0.02
ꢁ0.43
ꢁ3.79
ꢁ4.53
3.24
3.32
0.35
5.17
ꢁ1.05
ꢁ4.02
0.30
ꢁ0.22
HBA
Hbic
RingArom
RingArom
Hypo 1/21c
Weights
Tolerances
Coordinates
2.59597
1.60
4.61
2.59597
1.60
3.80
2.22
1.54
2.59597
1.60
7.97
2.82
1.27
2.59597
1.60
1.78
2.20
5.75
ꢁ4.49
2.20
1.60
8.92
0.55
2.97
1.60
4.36
X
Y
Z
ꢁ1.86
ꢁ4.35
ꢁ5.10
1.20
ꢁ1.90
ꢁ3.23
a
b
c
Hypo10/10: the 10th pharmacophore hypothesis generated in the 10th HYPOGEN run (as in Tables B and C in Supplementary data).
Hypo6/18: the 6th pharmacophore hypothesis generated in the 18th HYPOGEN run (as in Tables B and C in Supplementary data).
Hypo 1/21: the 1st pharmacophore hypothesis generated in the 21st HYPOGEN run (as in Tables B and C in Supplementary data).
explored via six diverse sets of inhibitors and using CATALYST-HYPOGEN
Table 3
ROCa performances of QSAR-selected pharmacophores as 3D search queries
to identify high quality binding model(s). Subsequently, genetic
algorithm and multiple linear regression analysis were employed
to access optimal QSAR model capable of explaining anti-BACE
Pharmacophore model
ROCa–AUCb
ACCc
SPCd
TPRe
FNRf
Hypo10/10
Hypo6/18
Hypo1/21
0.982
0.981
0.738
0.961
0.961
0.961
0.988
0.975
0.9611
0.28
0.60
0.96
0.011345
0.024311
0.038898
bioactivity variation across 129 collected BACE inhibitors (r1204
¼
0:879, F = 60.48, n = 104, r2BS = 0.001, rL2OO ¼ 0:846, rP2RESS against 25
external test inhibitors = 0.705). Three pharmacophoric models
emerged in the QSAR equation suggesting the existence of at least
three distinct binding modes accessible to ligands within BACE
binding pocket. The QSAR equation and the associated pharmaco-
phoric models were used to guide synthetic exploration of a new
series of BACE inhibitors that resulted in several novel low micro-
molar BACE inhibitors.
a
b
c
d
e
f
ROC: receiver operating characteristic.
AUC: area under the curve.
ACC: overall accuracy.
SPC: overall specificity.
TPR: overall true positive rate.
FNR: overall false negative rate.
in Hypo6/18. The amidic NH of the terminal p-anisidine of 174
seem to hydrogen-bond with the side chain of THR232 in the
docked pose, which agrees with mapping the same NH by a HBD
feature in model Hypo6/18. Similarly, mapping the methoxyben-
zene terminal of 174 with two Hbic features correlates with hydro-
phobic interactions tying this group with the side chain linkers of
SER325 and ARG235.
Similarly, the docked pose of 169 approximate its mapped pose
against Hypo1/21: The toloylamide ring of 169 is placed by the
docking engine at close proximity to the aromatic rings of TYR71,
PHE108, and TRP115 (Fig. 7f) suggesting the existence of mutual
4. Experimental
4.1. Molecular modeling
4.1.1. Software and hardware
The following software packages were utilized in the present
research:
ꢂ
ꢂ
CATALYST (Version 4.11), Accelrys Inc. (www.accelrys.com), USA.
CERIUS2 (Version 4.10), Accelrys Inc. (www.accelrys.com), USA.
ꢂ CS ChemDraw Ultra 6.0, Cambridge Soft Corp. (http://www.cam-
aromatic p-stacking interactions and apparently corresponding to
mapping the toloylamide with Hbic feature in Hypo1/21 (Fig. 7e).
Furthermore, LigandFit directed the central m-diaminobenzene
169 towards ILE110 suggesting mutual hydrophobic interactions,
which agrees with fitting this group against a Hbic feature in
Hypo1/21. Similarly, mapping the acetamido carbonyl of 169
against a HBA feature in Hypo1/21 corresponds to a hydrogen-
bonding interaction tying this group with the hydroxyl of
THR231 (Fig. 7e and f). Finally, the docking engine directed the ter-
minal pyridinium moiety of 169 perpendicular to the amidic side
bridgesoft.Com), USA.
ꢂ Pharmacophore and QSAR modeling studies were performed using
CATALYST (HYPOGEN module) and CERIUS2 software suites from Accelrys
Inc. (San Diego, California, www.accelrys.com) installed on a Sili-
con Graphics Octane2 desktop workstation equipped with a dual
600 MHz MIPS R14000 processor (1.0 GB RAM) running the Irix
6.5operatingsystem.Structuredrawingwasperformedemploying
ChemDraw Ultra 6.0 which was installed on a Pentium 4 PC.
chain of ASN233 suggesting the existence of mutual
p-stacking
interactions and corresponding to a RingArom feature in Hypo1/
21 mapping this pyridinium moiety.
4.1.2. Data set
The structures of 129 BACE-1 inhibitors (Table A in Supplemen-
tary data) were collected from published literature.15,48 All
collected inhibitors were assayed employing identical assay proce-
dures and conditions in order to allow proper QSAR correlation.
The in vitro bioactivities of the collected inhibitors were expressed
3. Conclusion
BACE inhibitors are currently considered as potential treat-
ments for AD. The pharmacophoric space of BACE inhibitors was

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3101
Figure 7. (A), (C) and (E) Show mapping of Hypo10/10, Hypo6/18 and Hypo1/21 against the synthesized compounds 184 (IC50 = 2.2
l
M), 174 (IC50 = 1.9
lM), and 169
(IC50 = 13 lM) (Table 4), respectively, while (B), (D) and (F) Illustrate corresponding docked poses of the same compounds in BACE (PDB code: 2IRZ, resolution = 1.8 Å).
as the concentration of the test compound that inhibited the activ-
ity of BACE-1 by 50% (IC50). Table A in Supplementary data shows
the structures and IC50 values of the considered inhibitors. The log-
500 lM. (Table A in Supplementary data, Fig. 1). These assumptions
are necessary to allow statistical correlation and QSAR analysis. The
logarithmic transformation of IC50 values should minimize any
potential errors resulting from this assumption.
arithm of measured IC50 (lM) values were used in pharmacophore
modeling and QSAR analysis, thus correlating the data linear to the
free energy change.
The two-dimensional (2D) chemical structures of the inhibitors
were sketched using ChemDraw Ultra, installed on a PC, and saved
in MDL-mol file format. Subsequently, they were imported into
CATALYST, converted into corresponding standard 3D structures and
energy minimized to the closest local minimum using the molecu-
lar mechanics CHARMm force field implemented in CATALYST. The
In cases where IC50 is expressed as being higher than 100
lM
(e.g., 74, 100, 101, 107, 112 and 117), it was assumed it equals
100
500
l
l
M. In cases where IC50 is expressed as being higher than
M (e.g., 118, 120, 122, 125 and 129), it was assumed it equals

Table 4
The synthesized compounds with their fit values against (Hypol0/10; Hypo 6/18; Hypo 1/21), their corresponding QSAR estimates from Eq. 1 and their in vitro anti-BACE bioactivities
No.a
164
165
166
Structure
Fit valuesb
Hypo 6/18
QSAR predictions
% Inhibition at 50
Experiments
% Inhibition at 50
Mc
Hypo 10/10
0.0
Hypo 1/21
6.9
lM
IC50
9.2
(lM)
l
IC50 (l
M)c
CH
3
O
O
H
N
Cl
N
2.1
0.0
0.8
88
42
—
N
H
O
O
H
N
Cl
N
0.0
0.0
6.8
7.2
75
73
13.6
14.7
61
—
N
H
CH
Cl
3
O
O
H
N
Cl
N
100
1.0 (98%)d
N
H
O
O
CH
3
H
N
Cl
N
167
168
169
0.0
0.0
0.0
1.3
0.0
0.0
7.1
7.0
7.3
76
75
76
12.9
13.5
13.3
100
51
3.7 (99%)d
N
H
O
O
H
N
H
N
Cl
N
—
O
CH
O
3
Cl
N
H
N
H
N
68
13.0 (99%)d
O
O
O
O
S
H
N
Cl
N
170
171
0.0
7.8
1.6
3.7
9.2
9.4
56
33.4
1.8
50
69
—
—
N
H
O
CH
3
O
H
N
Cl
N
N
H
100
O
H N
2
O
O
H
N
Cl
N
O
N
H
CH
3
172
9.l
6.8
9.4
100
0.9
44
—
O
H2N

Cl
O
H
N
Cl
N
N
H
173
8.4
2.6
9.0
100
2.8
64
30.2 (88%)d
O
H N
2
O
O
O
CH
3
H
N
Cl
N
O
N
H
174
175
176
6.9
7.3
9.0
5.0
0.0
4.2
8.9
8.5
8.3
100
100
100
2.3
6.6
1.5
100
1.9 (96%)d
44.7 (87%)d
—
O
O
H N
2
H
N
H
N
Cl
N
O
54
O
H N
2
CH3
Cl
H
N
H
N
N
O
24
O
O
H2N
O
O
S
H
N
Cl
N
N
H
O
177
178
8.9
8.9
5.9
3.9
9.6
9.6
100
100
2.8
3.8
34
—
—
O
H N
2
O
H
N
H
N
Cl
O
N
S
O
O
4
CH
3
H N
2
O
Cl
O
N
O
H
HN
N
179
5.7
1.0
8.9
100
2.3
72
4.0 (97%)d
N
H
O
CH3
(continued on next page)

Table 4 (continued)
No.a
Structure
Fit valuesb
Hypo 6/18
QSAR predictions
% Inhibition at 50
Experiments
% Inhibition at 50
Mc
Hypo 10/10
Hypo 1/21
lM
IC50
(lM)
l
IC50 (l
M)c
Cl
Cl
Cl
Cl
Cl
Cl
O
O
O
O
O
O
N
O
H
N
HN
HN
HN
HN
HN
HN
180
6.1
0.0
1.3
0.8
0.0
0.9
3.1
9.4
100
6.1
58
—
N
H
O
O
O
CH
3
N
O
H
N
181
182
183
184
185
4.0
4.3
5.6
6.7
7.1
9.4
9.1
9.1
9.7
8.8
100
3.0
8.9
8.9
4.4
6.4
78
44
55
100
36
12.0 (88%)d
N
H
Cl
N
O
H
N
89
—
N
H
O
CH
3
N
O
O
89
—
N
H
N
H
N
O
O
100
2.2 (99%)d
N
H
N
H
CH3
N
O
O
S
100
—
N
H
N
H
O
CH
3
a
b
c
Compound numbers as in Scheme 7.
Best-fit values calculated by Eq. 5.
Bioactivity values are the average of at least duplicate measurements.
d
Values between brackets represent the correlation coefficients of the corresponding dose–response lines at three concentrations.

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3105
resulting 3D structures were utilized as starting conformers for
conformational analysis.
4.1.5. Assessment of the generated hypotheses
When generating hypotheses, CATALYST attempts to minimize a
cost function consisting of three terms: Weight cost, error cost
and configuration cost.35,39–43 Weight cost is a value that increases
as the feature weight in a model deviates from an ideal value of 2.
The deviation between the estimated activities of the training set
and their experimentally determined values adds to the error cost.
The activity of any compound can be estimated from a particular
hypothesis through Eq. 4:35
4.1.3. Conformational analysis
The molecular flexibilities of the collected compounds were ta-
ken into account by considering each compound as a collection of
conformers representing different areas of the conformational
space accessible to the molecule within a given energy range.
Accordingly, the conformational space of each inhibitor (1–129,
Table A in Supplementary data) was explored adopting the ‘best
conformer generation’ option within CATALYST. Default parameters
were employed in the conformation generation procedure, that
is, a conformational ensemble was generated with an energy
threshold of 20 kcal/ mol from the local minimized structure which
has the lowest energy level and a maximum limit of 250 conform-
ers per molecule.35
LogðEstimated ActivityÞ ¼ I þ Fit
ð4Þ
where I = the intercept of the regression line obtained by plotting
the log of the biological activity of the training set compounds
against the Fit values of the training compounds. The Fit value for
any compound is obtained automatically employing Eq. 5:35
X
X
2
Fit ¼
mapped hypothesis features ꢀ W½1 ꢁ
ðdisp=tolÞ ꢃ
4.1.4. Pharmacophoric hypotheses generation
ð5Þ
All 129 molecules with their associated conformational models
were regrouped into a spreadsheet. The biological data of the inhib-
itors were reported with an ‘Uncertainty’ value of 3, which means
that the actual bioactivity of a particular inhibitor is assumed to be
situated somewhere in an interval ranging from one-third to
three-times the reported bioactivity value of that inhibitor. Typi-
cally, CATALYST requires informative training sets that include at least
16 compounds of evenly spread bioactivities over at least three and a
half logarithmic cycles. Lesser training lists could lead to chance cor-
relation and thus faulty models.39,41,51 Six structurally diverse train-
ing subsets (Table 1) were carefully selected from the collected
compounds for pharmacophore modeling.
Each training subset was utilized to conduct four modeling runs
to explore the pharmacophoric space of BACE inhibitors. Different
hypotheses were generated by altering the interfeature spacing
and the number of allowed features in the resulting pharmaco-
phores (Table B in Supplementary data).
Pharmacophore modeling employing CATALYST proceeds through
three successive phases: the constructive phase, subtractive phase
and optimization phase. During the constructive phase, CATALYST
generates common conformational alignments among the most-
active training compounds. Only molecular alignments based on
a maximum of five chemical features are considered. The program
identifies a particular compound as being within the most active
category if it satisfies Eq. 2:39,41,51
where
R mapped hypothesis features represents the number of
pharmacophore features that successfully superimpose (i.e., map
or overlap with) corresponding chemical moieties within the fitted
compound, W is the weight of the corresponding hypothesis feature
spheres. This value is fixed to 1.0 in CATALYST-generated models. disp
is the distance between the center of a particular pharmacophoric
sphere (feature centroid) and the center of the corresponding
superimposed chemical moiety of the fitted compound; tol is the
radius of the pharmacophoric feature sphere (known as Tolerance,
equals to 1.6 Å by default).
R
(disp/tol)2 is the summation of
(disp/tol)2 values for all pharmacophoric features that successfully
superimpose corresponding chemical functionalities in the fitted
compound.35
The third term, that is, the configuration cost, penalizes the com-
plexityof the hypothesis. This is a fixed cost, which is equal to the en-
tropy of the hypothesis space. The more the numbers of features (a
maximum of five) in a generated hypothesis, the higher is the entro-
py with subsequent increase in this cost. The overall cost (total cost)
of a hypothesis is calculated by summing over the three cost factors.
However, error cost is the main contributor to total cost.
CATALYST also calculates the cost of the null hypothesis, which
presumes that there is no relationship in the data and that exper-
imental activities are normally distributed about their mean.
Accordingly, the greater the difference from the null hypothesis
cost, the more likely that the hypothesis does not reflect a chance
correlation. In a successful automatic modeling run, CATALYST ranks
the generated models according to their total costs.35
ðMAct ꢀ UncMActÞ ꢁ ðAct=UncActÞ > 0:0
ð2Þ
where ‘MAct’ is the activity of the most active compound in the
training set, ‘Unc’ is the uncertainty of the compounds and ‘Act’ is
the activity of the training compounds (Table 1).
In the subsequent subtractive phase, CATALYST eliminates some
hypotheses that fit inactive training compounds. A particular
training compound is defined as being inactive if it satisfies Eq.
3:39,41,51
An additional approach to assess the quality of CATALYST-HYPOGEN
pharmacophores is to cross-validate them using the CAT-SCRAMBLE
program implemented in CATALYST. This validation procedure is
based on Fisher’s randomization test.51 In this validation test, a
95% confidence level was selected, which instruct CATALYST to gener-
ate 19 random spreadsheets by the CAT-SCRAMBLE command. Subse-
quently, CATALYST-HYPOGEN is challenged to use these random
LogðActÞ ꢁ log;ðMActÞ > BS
ð3Þ
spreadsheets to generate hypotheses using exactly the same fea-
tures and parameters used in generating the initial unscrambled
hypotheses. Success in generating pharmacophores of comparable
cost criteria to those produced by the original unscrambled data
reduces the confidence in the training compounds and the
unscrambled original pharmacophore models.
where, ‘BS’ is the bioactivity spread (equals 3.5 by default, Table 1).
However, in the optimization phase, CATALYST applies fine pertur-
bations in the form of vectored feature rotation, adding new fea-
ture and/or removing a feature, to selected hypotheses that
survived the subtractive phase, in an attempt to find new models
of enhanced bioactivity/mapping correlations. CATALYST selects the
highest-ranking models (10 by default) and presents them as the
optimal pharmacophore hypotheses resulting from the particular
automatic modeling run.
Table C in Supplementary data shows the success criteria of rep-
resentative pharmacophores from each run.
4.1.6. Clustering of the generated pharmacophore hypotheses
The successful models (226) were clustered into 45 groups uti-
Eventually, our pharmacophore exploration efforts (24 auto-
matic runs, Table 1 and Table B in Supplementary data) culminated
in 226 pharmacophore models of variable qualities.
lizing the hierarchical average linkage method available in CATALYST
.
Therefore, closely-related pharmacophores were grouped in five-

3106
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
R
1
R
2
O
O
H N
2
R
R
1
2
a
O N
2
O N
2
N
H
Cl
135 R = H, R = CH
3
131 R = H, R = CH
3
1
2
130
1
2
136 R = CH , R = H
132 R = CH , R = H
1
3
2
1
3
2
137 R = H, R = Cl
133 R = H, R = Cl
1
2
1
2
138 R = H, R = OCH
1
2
3
134 R = H, R = OCH
1
2
3
Scheme 1. Synthesis of N-phenyl-m-nitrobenzamide derivatives: (a) triethylamine in dry acetone.
O N
2
R
1
R
H N
2
Y
O N
2
Y
1
Cl
S
a
O
S
NH
2
O
O
NH
R
2
R
2
a
O
148 Y = CONH, R = H, R = CH
135 Y = CONH, R = H, R = CH
3
1
2
3
1
2
149 Y = CONH, R = CH , R = H
O N
2
136 Y = CONH, R = CH , R = H
1
3
2
1 3 2
OCH
150 Y = CONH, R = H, R = Cl
137 Y = CONH, R = H, R = Cl
3
1
2
1
2
OCH
140
3
151 Y = CONH, R = H, R = OCH
138 Y = CONH, R = H, R = OCH
1
2
3
1
2
3
139
138
152 Y = NHCO, R = H, R = H
145 Y = NHCO, R = H, R = H
1
2
1
2
153 Y = NHCO, R = H, R = CH
146 Y = NHCO, R = H, R = CH
1
2
3
1
2
3
Scheme 2. Synthesis of N-(4-methoxyphenyl)-3-nitro-benzenesulfonamide (140):
(a) triethylamine and DMAP in dry acetone.
Scheme 4. Reduction of the nitro-phenylbenzamides to their corresponding amine
derivatives. (a) Na2S2O4/Na2CO3 in water.
membered clusters. Subsequently, the highest-ranking representa-
tives, as judged based on their fit-to-bioactivity correlation F-val-
ues (calculated against collected compounds 1–129, Table A in
Supplementary data and Fig. 1), were selected to represent their
corresponding clusters in subsequent QSAR modeling (Table C in
Supplementary data).
O N
Y
H N
Y
2
2
a
R
1
R
1
140 Y = SO NH, R = CH O
147 Y = NHSO , R = CH
154 Y = SO NH, R = CH O
155 Y = NHSO , R = CH
2
1
3
3
2
1
3
3
2
1
2
1
4.1.7. QSAR modeling
Scheme 5. Reduction of the nitro-phenylbenzsulfonamides to their corresponding
amine derivatives. (a) SnCl2ꢄ2H2O in ethanol.
A subset of 104 compounds from the total list of inhibitors
(1–129) was utilized as a training set for QSAR modeling. However,
since it is essential to access the predictive power of the resulting
QSAR models on an external set of inhibitors, the remaining 25
molecules (ca. 20% of the dataset) were employed as an
external test subset for validating the QSAR models. The test
molecules were selected as follows: the collected inhibitors
(1–129, Table A in Supplementary data and Fig. 1) were ranked
according to their IC50 values, and then every fifth compound
was selected for the test set starting from the high-potency
end. This selection considers the fact that the test molecules
must represent a range of biological activities similar to that
of the training set.
H
R
H N
2
Y
1
N
Y
R1
R2
a
Cl
O
R
2
156 Y = CONH, R1 = H, R2 = CH3
157 Y = CONH, R1 = CH3, R2 = H
158 Y = CONH, R1 = H, R2 = Cl
159 Y = CONH, R1 = H, R2 = OCH3
160 Y = NHCO, R1 = H, R2 = H
161 Y = NHCO, R1 = H, R2 = CH3
162 Y = SO2NH, R1= CH3O
148 Y = CONH, R = H, R = CH
3
1
2
149 Y = CONH, R = CH , R = H
1
3
2
150 Y = CONH, R = H, R = Cl
1
2
151 Y = CONH, R = H, R = OCH
1
2
3
152 Y = NHCO, R = H, R = H
1
2
153 Y = NHCO, R = H, R = CH
1
2
3
154 Y = SO NH, R = CH O
2
1
3
3
163 Y = NHSO2, R1= CH3
155 Y = NHSO , R = CH
2
1
The chemical structures of the inhibitors were imported into
CERIUS2 as standard 3D single conformer representations in SD for-
mat. Subsequently, different descriptor groups were calculated for
each compound employing the C2.DESCRIPTOR module of CERIUS2.
The calculated descriptors included various simple and valence
connectivity indices, electro-topological state indices and other
molecular descriptors (e.g., logarithm of partition coefficient,
polarizability, dipole moment, molecular volume, molecular
Scheme 6. Synthesis of the mono-chloromethyl-acetamido-derivatives (a) mono-
chloroacetylchloride/triethylamine in dry acetone.
weight, molecular surface area, energies of the lowest and highest
occupied molecular orbitals, etc.).51 Furthermore, the training
compounds were fitted (using the Best-fit option in CATALYST
)
against the representative pharmacophores (45 models, Table C
R
1
H
N
X
O N
2
Z
O N
2
NH
2
a or b
R
1
145 Z = CO, R = H
1
141
142 X = COCl, R = H
1
146 Z = CO, R = CH
1
3
143 X = COCl, R = CH
1
3
147 Z = SO , R = CH
2
1
3
144 X = SO Cl, R = CH
2
1
3
Scheme 3. Synthesis N-(m-nitrophenyl)-benzamide and -benzsulfonamide derivatives: (a) triethylamine in dry acetone, (b) triethylamine/DMAP in dry acetone.

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3107
H
Cl
N
Y
R1
R2
H
Cl
a
R3
N
Y
R1
R2
N
O
O
156 Y = CONH, R1 = H, R2 = CH3
157 Y = CONH, R1 = CH3, R2 = H
158 Y = CONH, R1 = H, R2 = Cl
164-185
159 Y = CONH, R1 = H, R2 = OCH3
160 Y = NHCO, R1 = H, R2 = H
161 Y = NHCO, R1 = H, R2 = CH3
162 Y = SO2NH, R1= CH3O
163 Y = NHSO2, R1= CH3
Scheme 7. Synthesis of the pyridinum acetamido-phenylbenzamides and -phenylbenz sulfonamides (see Table 4 for detailed structures). (a) Fusion with pyridine,
nicotinamide or N-phenylethyl nicotinamide (186).
Table 5
Therefore, it was necessary to prepare valid evaluation struc-
tural database (testing set) that contains an appropriate list of de-
Inhibitory activities of potent synthesized BACE inhibitors against renin enzyme
coy compounds in combination with diverse list of known active
compounds. The decoy list was prepared as described by Verdonk
and co-workers.55,56 Briefly, the decoy compounds were selected
based on three basic one-dimensional (1D) properties that allow
the assessment of distance (D) between two molecules (e.g., i
and j), namely: (1) the number of hydrogen-bond donors (Num-
HBD); (2) number of hydrogen-bond acceptors (NumHBA) and
(3) count of nonpolar atoms (NP, defined as the summation of Cl,
F, Br, I, S and C atoms in a particular molecule). For each active
compound in the testing set, the distance to the nearest other ac-
tive compound is assessed using their Euclidean Distance (Eq. 7):
Compound
Percent renin inhibition at 50 lM
166
167
169
174
179
181
184
0
0
0
0
26
26
22
in Supplementary data), and their fit values were added as addi-
tional descriptors. The fit value for any compound is obtained auto-
matically via Eq. 5.35
Dði;jÞ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
2
2
¼
ðNumHBDi ꢁ NumHBDjÞ þ ðNumHBAi ꢁ NumHBAjÞ þ ðNPi ꢁ NPjÞ
Genetic function approximation (GFA) was employed to search
for the best possible QSAR regression equation capable of correlating
the variations in biological activities of the training compounds with
variations in the generated descriptors, that is, multiple linear
regression modeling (MLR). The fitness function employed herein
is based on Friedman’s ‘lack-of-fit’ (LOF).51 However, to avoid over-
whelming GFA-MLR with large number of poor descriptors; we re-
moved 20% of those showing lowest variance prior to QSAR analysis.
We were obliged to normalize the potencies of the training
ð7Þ
The minimum distances are then averaged over all active compounds
(Dmin). Subsequently, for each active compound in the testing set an
average of 25 decoys were randomly chosen from the ZINC data-
base.57 The decoys were selected in such a way that they did not ex-
ceed Dmin distance from their corresponding active compound.
Moreover, to further diversify the actives members, that is, to
avoid close similarity among actives in the testing set, any active
compound having zero distance (D(i, j)) from other active com-
pound (s) in the testing set were excluded. Active testing com-
pounds were defined as those possessing BACE-1 affinities (IC50
compounds via division by their corresponding molecular weights,
Log ð1=IC50
i.e., ligand efficiency ð
ÞÞ,54 to achieve reasonable self-consis-
Mwt
tent QSAR models.
values) ranging from 2 nM to 10 lM. The testing set included 25
active compounds and 617 ZINC compounds.
Our preliminary diagnostic trials suggested the following opti-
mal GFA parameters: explore linear, quadratic and spline equa-
tions at mating and mutation probabilities of 50%; population
size = 500; number of genetic iterations = 30,000 and lack-of-fit
(LOF) smoothness parameter = 1.0. However, to determine the
optimal number of explanatory terms (QSAR descriptors), it was
decided to scan and evaluate all possible QSAR models resulting
from 4 to 20 explanatory terms.
The testing set (642 compounds) was screened by each particu-
lar pharmacophore for ROC analysis employing the ‘Best flexible
search’ option implemented in CATALYST, while the conformational
spaces of the compounds were generated employing the ‘Fast con-
formation generation option’ implemented in CATALYST. Compounds
missing one or more features were discarded from hit lists. The in
silico hits were scored employing their fit values (best fit values) as
calculated by Eq. 5.
ROC curve analysis describes the sensitivity (Se or true positive
rate, Eq. 8) for any possible change in the number of selected com-
pounds (n) as a function of (1 ꢁ Sp). Sp is defined as specificity or
true negative rate (Eq. 9):55,58
All QSAR models were validated employing leave one-out cross-
validation (rL2OO), bootstrapping (rB2S) and predictive r2 (rP2RESS) calcu-
lated from the test subsets. The predictive r2PRESS is defined as:
r2PRESS ¼ SD ꢁ PRESS=SD
ð6Þ
where SD is the sum of the squared deviations between the biolog-
ical activities of the test set and the mean activity of the training set
molecules, PRESS is the squared deviations between predicted and
actual activity values for every molecule in the test set.
Number of Selected Actives
Total Number of Actives
Number of Discarded Inactives
TP
TP þ FN
Se ¼
Sp ¼
¼
ð8Þ
ð9Þ
TN
¼
Total Number of Inactives
TN þ FP
4.1.8. Receiver operating characteristic (ROC) curve analysis
Successful pharmacophore models (i.e., Hypo10/10, Hypo6/18,
and Hypo1/21) were further validated by assessing their abilities
to selectively capture diverse BACE inhibitors from a large list of
decoys employing ROC analysis.
where, TP is the number of active compounds captured by the vir-
tual screening method (true positives), FN is the number of active
compounds discarded by the virtual screening method, TN is the

3108
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
number of discarded decoys (presumably inactives), while FP is the
number of captured decoys (presumably inactives).
360 nm). Elemental analysis was performed using EuroVector ele-
mental analyzer. Chemicals and solvents were used without fur-
ther purification.
A ROC curve is plotted by setting the score of the active mole-
cule as the first threshold. Afterwards, the number of decoys with-
in this cutoff is counted and the corresponding Se and Sp pair is
calculated. This calculation is repeated for the active molecule with
the second highest score and so forth, until the scores of all actives
are considered as selection thresholds.55,58
The ROC curve representing ideal distributions, where no over-
lap between the scores of active molecules and decoys exists, pro-
ceeds from the origin to the upper-left corner until all the actives
are retrieved and Se reaches the value of 1. Thus, the ideal ROC
curve continues as a horizontal straight line to the upper-right cor-
ner where all actives and all decoys are retrieved, which corre-
sponds to Se = 1 and Sp = 0. In contrast to that, the ROC curve for
a set of actives and decoys with randomly distributed scores tends
towards the Se = 1 ꢁ Sp line asymptotically with increasing num-
ber of actives and decoys.55
4.2.1. Preparation of N-(3-nitrophenyl)benzamide derivatives
(135–138)
To a magnetically-stirred, ice-bathed, solution of the substituted
aniline and triethylamine (2 equiv) in dry acetone (50 mL), benzoyl
chloride (2 equiv) in acetone was added. The solution was stirred at
room temperature until completion (as revealed by TLC). The reac-
tion mixture was quenched by slow addition into sufficient 5%
aqueous sodium bicarbonate solution to neutralize all generated
acid. The precipitated crude N-(3-nitrophenyl)benzamide products
were purified by recrystallization from acetone/water (Scheme 1).61
4.2.1.1. 3-Nitro-N-p-tolyl-benzamide (135). Prepared from com-
mercially available 3-nitrobenzoyl chloride 130 (25.9 g, 0.14 mol)
dissolved in acetone (50 mL) and p-tolylamine 131 (10.0 g,
0.09 mol) to yield 135 as white crystalline powder (20.0 g, 87%)
The success of a particular virtual screening workflow can be
judged from the following criteria:
mp: 160–161 °C; mmax (KBr disc) 3298, 1647, 1527 cmꢁ1 1H NMR
;
(300 MHz, DMSO-d6): d 2.26 (s, 3H), 7.15 (d, 2H, J 7.8 Hz), 7.63
(d, 2H, J 7.5 Hz), 7.79 (t, 1H), 8.39 (t, 2H), 8.75 (s, 1H), 10.49 (s,
1H, NH) ppm; 13C NMR (75 MHz, DMSO-d6): d 21.20 (CH3),
121.25 (2 ꢀ CH), 123.05 (CH), 126.76 (CH), 129.78 (2 ꢀ CH),
130.84 (CH), 133.84 (C), 134.81 (CH), 136.82 (C), 137.01 (C),
148.41 (C), 163.76 (C@O) ppm; HRMS-ESI m/z [M+Na]+ calcd for
C14H12N2NaO3: 279.07456, found: 279.07401.
(1) Area under the ROC curve (AUC).58 In an optimal ROC curve
an AUC value of 1 is obtained; however, random distribu-
tions cause an AUC value of 0.5. Virtual screening that per-
forms better than a random discrimination of actives and
decoys retrieve an AUC value between 0.5 and 1.55,58
(2) Overall accuracy (ACC): describes the percentage of correctly
classified molecules by the screening protocol (10). Testing
compounds are assigned a binary score value of zero (com-
pound not captured) or one (compound captured):55,59,60
4.2.1.2. 3-Nitro-N-m-tolyl-benzamide (136). Prepared from 130
(25.9 g, 0.14 mol) and m-tolylamine 132 (10 g, 0.09 mol) to yield
136 as white crystalline powder (20.0 g, 87%), mp 115 °C; mmax
ꢀ
ꢁ
TP þ TN
A
N
A
N
ACC ¼
¼
ꢄ Se þ 1 ꢁ
ꢄ Sp
ð10Þ
(KBr disc) 3294, 1647, 1531 cmꢁ1 1H NMR (300 MHz, DMSO-d6):
;
N
d 2.29 (s, 3H, CH3), 6.93 (d, 1H, J 7.5 Hz), 7.23 (t, 1H), 7.55 (d, 1H,
J 7.8 Hz), 7.58 (s, 1H), 7.80 (t, 1H), 8.37 (d, 1H, J 9.9 Hz), 8.41 (d,
1H, J 6 Hz), 8.75 (s, 1H), 10.49 (s, 1H, NH); 13C NMR (75 MHz,
DMSO-d6): d 21.86 (CH3), 118.42 (CH), 121.77 (CH), 123.06 (CH),
125.54 (CH), 126.23 (CH), 129.23 (CH), 130.87 (CH), 134.82 (CH),
136.69 (C), 138.54 (C), 139.24 (C), 148.40 (C), 163.94 (C@O) ppm;
HRMS-ESI m/z [MꢁH]+ calcd for C14H11N2O3: 255.07697, found:
255.07752.
where N is the total number of compounds in the testing
database, A is the number of true actives in the testing
database.
(3) Overall specificity (SPC): describes the percentage of dis-
carded inactives by the particular virtual screening work-
flow. Inactive test compounds are assigned a binary score
value of zero (compound not captured) or one (compound
captured).55,59,60
(4) Overall true positive rate (TPR or overall sensitivity):
describes the fraction percentage of captured actives from
the total number of actives. Active test compounds are
assigned a binary score value of zero (compound not cap-
tured) or one (compound captured).
(5) Overall false negative rate (FNR or overall percentage of dis-
carded actives): describes the fraction percentage of active
compounds discarded by the virtual screening method. Dis-
carded active test compounds are assigned a binary score
value of zero (compound not captured) or one (compound
captured).
4.2.1.3.N-(4-Chloro-phenyl)-3-nitro-benzamide (137). Prepared
from 130 (28.94 g, 0.16 mol) and 4-chloro-phenylamine 133 (10.0 g,
0.08 mol) to yield 137 as white crystalline powder (20.0 g, 91%) mp:
173–175 °C; mmax (KBr disc) 3290, 1651, 1527 cmꢁ1 1H NMR
;
(300 MHz, DMSO-d6): d 7.37 (d, 2H J 9.3 Hz), 7.79 (d, 2H, J 8.7 Hz),
7.79 (t, 1H), 8.37 (d, 1H, J 6.9 Hz), 8.41 (d, 1H, J 8.1 Hz), 8.75 (s, 1H),
10.68 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 122.76
(2 ꢀ CH), 123.11 (CH), 126.99 (CH), 128.47 (C), 129.31 (2 ꢀ CH),
130.93 (CH), 134.88 (CH), 136.66 (C), 138.33 (C), 148.41 (C), 164.12
(C@O) ppm; HRMS-ESI m/z [M+H]+ calcd for C13H10ClN2O3:
277.03799, found: 277.03745.
4.2. Synthetic procedures
4.2.1.4. N-(4-Methoxy-phenyl)-3-nitro-benzamide (138). Pre-
pared from 130 (14.0 g, 0.08 mol) and 4-methoxy-phenylamine
134 (4.6 g, 0.038 mol) to yeild 138 as yellowish-green crystalline
powder (4.87 g, 47%) mp:176–177 °C; mmax (KBr disc) 3298, 1647,
Melting points were measured using Gallenkampf melting point
apparatus and are uncorrected. 1H NMR and 13C NMR spectrums
were collected on a Varian Oxford NMR300 spectrometer. High res-
olution mass spectrometry was performed using LC Mass Bruker
Apex-IV mass spectrometer utilizing an electrospray interface.
Infrared spectra were recorded using Shimadzu IR Affinity-1 spec-
trophotometer. The samples were dissolved in CHCl3 and analyzed
as thin solid films using NaCl plates or as KBr pellets. Analytical
thin layer chromatography (TLC) was carried out using pre-coated
aluminum plates and visualized by UV light (at 254 and/or
1527 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 3.73 (s, 3H), 6.94 (d,
;
2H, J 9 Hz), 7.67 (d, 2H, J 9.3), 7.811 (t, 1H), 7.37 (d, 1H, J 6.6),
8.40 (d, 1H, J 7.2), 8.76 (s, 1H), 10.46 (s, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6):
d
55.87 (CH3), 114.50 (2 ꢀ CH), 122.85
(2 ꢀ CH), 122.99 (CH), 126.71 (CH), 130.85 (CH), 132.37 (C),
134.76 (CH), 137.04 (C), 148.43 (C), 156.53 (C), 163.53 (C@O)
ppm; HRMS-ESI m/z [MꢁH]+ calcd for C14H11N2O4: 271.07188,
found: 271.07243.

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3109
4.2.2. Preparation of N-(4-methoxyphenyl)-3-nitro-benzene-
sulfonamide and N-(m-nitrophenyl)-benzsulfonamide deriv-
atives (140, 145–147)
and Na2CO3 (7.0 equiv) were added as dry powders. The reaction
mixture was stirred at room temperature for seven days. The reac-
tion mixture was supplemented with Na2S2O4 (2.0 equiv) every
48 h until thin-layer chromatography revealed that all starting
material had reacted. The reaction was terminated by filteration.
The solid residue was rinsed with water (200 mL) over 2–3 h and
dried under vacuum (Scheme 4).63
To a magnetically-stirred, ice-bathed, solution containing the
substituted aniline (1.0 equiv) and triethylamine (2 equiv) in dry
acetone (50 mL), a solution of 3-nitro-benzenesulfonyl chloride
(1.5–2.0 equiv) and DMAP (0.5 equiv) in dry acetone (50 mL) was
added. The reaction was stirred at room temperature overnight un-
til TLC revealed all aniline had reacted. The reaction mixture was
quenched by slow addition onto sufficient 5% aqueous sodium
bicarbonate to neutralize all generated acid. The precipitated crude
products were purified by recrystallization from acetone/water
(Schemes 2 and 3).62
4.2.3.1. 3-Amino-N-p-tolyl-benzamide (148). Prepared from
compound 135 (23.0 g, 0.089 mol) to yield 148 as white powder
(2.82 g, 14%) mp: 165–167 °C; mmax (KBr disc) 3479, 3387, 3286,
1651 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.24 (s, 3H), 5.28 (s,
;
2H), 6.70 (d, 1H, J 7.8), 7.00–7.13 (m, 5H), 7.54 (d, 2H, J 8.1), 9.96
(s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 21.17 (CH3), 113.61
(CH), 115.32 (CH), 117.32 (CH), 120.86 (2 ꢀ CH), 129.4 (CH),
129.60 (2 ꢀ CH), 132.93 (C), 136.71 (C), 137.54 (C), 149.43 (C),
166.85 (C@O) ppm; HRMS-ESI m/z [M+H]+ calcd for C14H15N2O:
228.12179, found: 228.12125.
4.2.2.1. N-(4-Methoxy-phenyl)-3-nitro-benzenesulfonamide (140).
Prepared from commercially available 3-nitro-benzenesulfonyle
chloride 139 (7.0 g, 0.0315 mol) and 4-methoxy-phenylamine
138 (3.88 g, 0.0315 mol) to yield 140 as white powder (3.59 g,
37%) mp: 193–194 °C; mmax (KBr disc) 3448, 1604 cmꢁ1 1H NMR
;
(300 MHz, DMSO-d6): d 3.70 (s, 3H), 6.97 (d, 1H, J 6 Hz), 7.05 (d,
1H, J 9 Hz), 7.98 (t, 1H), 8.23 (d, 1H, J 9 Hz), 8.64 (s, 1H), 8.64 (d,
1H, J 9 Hz) ppm; 13C NMR (75 MHz, DMSO-d6): d 56.30 (CH3),
115.77 (2 ꢀ CH), 123.49 (2 ꢀ CH), 125.39 (CH), 130.16 (CH),
132.55 (CH), 133.42 (C), 134.63 (C), 140.03 (C), 148.71 (C),
161.68 (C@O) ppm; HRMS-ESI m/z [MꢁH]+ calcd for C13H11N2O5S:
308.04222, found: 308.04275.
4.2.3.2. 3-Amino-N-m-tolyl-benzamide (149). Prepared from
compound 136 (20.0 g, 0.078 mol) as white powder (3.97 g, 22%)
mp: 95–96 °C; mmax (KBr disc) 3421, 3336, 3255, 1647 cmꢁ1 1H
;
NMR (300 MHz, DMSO-d6): d 2.27 (s, 3H), 5.29 (s, 2H), 6.71 (d,
1H, J 7.2 Hz), 6.86 (d, 1H, J 7.5 Hz), 7.70–7.20 (m, 4H), 7.51 (d,
1H, J 7.8 Hz), 7.58 (s, 1H), 9.96 (s, 1H) ppm; 13C NMR (75 MHz,
DMSO-d6): d 21.93 (CH3), 113.63 (CH), 115.36 (CH), 117.38 (CH),
118.06 (CH), 121.39 (CH), 124.75 (CH), 129.06 (CH), 129.06 (CH),
136.69 (C), 138.32 (C), 139.97 (C), 149.44 (C), 167.00 (C@O) ppm;
HRMS-ESI m/z [M+H]+ calcd for C14H15N2O: 228.12179, found:
228.12125.
4.2.2.2. N-(3-Nitrophenyl)-benzamide (145). Prepared from com-
mercially available 3-nitroaniline 141 (12.0 g, 0.09 mol) and ben-
zoyl choride 142 (24.0 g, 0.17 mol) to yield 145 as light yellow
crystalline powder (20 g, 92%) mp: 157–160 °C; mmax (thin film)
1651, 1539 cmꢁ1 1H NMR (300 MHz, CDCl3): d 7.2 (s, , 1H),
;
7.47–7.58 (m, 5H), 7.88 (d1H, J 7.2 Hz) 7.99 (d, 1H, J 8.4 Hz),
8.116 (m, 1H), 8.49 ppm (s, 1H); 13C-NMR (CDCl3): d 113.91 (CH),
118.10 (CH), 124.85 (CH), 126.08 (2 ꢀ CH), 127.95 (2 ꢀ CH),
128.93 (CH), 131.45 (CH), 132.96 (C), 138.03 (C), 147.57 (C),
164.92) (C@O); HRMS-ESI m/z [MꢁH]+ calcd for C13H9N2O3:
241.06132, found 241.06187.
4.2.3.3. 3-Amino-N-(4-chloro-phenyl)-benzamide (150). Prepared
from compound 137 (25.0 g, 0.09 mol) to yield 150 as white pow-
der (2.92 g, 13%) mp: 175–177 °C; mmax (KBr disc) 3402, 3321,
1651 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 5.31 (s, 2H), 6.73 (d,
;
1H, J 7.2 Hz), 7.01–7.14 (m, 3H), 7.35 (d, 2H, J 8.6 Hz), 7.77 (d,
2H, J 8.7 Hz), 10.18 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d
113.60 (CH), 115.38 (CH), 117.58 (CH), 122.32 (2 ꢀ CH), 127.62
(C), 129.14 (2 ꢀ CH), 129.50 (CH), 136.33 (C), 139.04 (C), 149.49
(C), 167.17 (C@O) ppm; HRMS-ESI m/z [M+H]+ calcd for
C13H12ClN2O: 247.06381, found: 247.06327.
4.2.2.3. 4-Methyl-N-(3-nitro-phenyl)-benzamide (146). Prepared
from 141 (6.0 g, 0.05 mol) and 4-methyl benzoylchloride 143
(12.0 g, 0.09 mol) to yield 146 as light yellow crystalline powder
(13.3 g, 63%) mp: 175 °C; mmax (thin film) 1654, 1523 cmꢁ1 1H
,
NMR (CDCl3): d 8.78 (s, 1H), 8.776 (s, 1H), 8.16 (dd, 1H, J 7.2,
0.9 Hz), 7.9 (d, 1H, J 1.2 Hz), 7.88 (d, 2H, J 7.8 Hz), 7.60 (m, 1H),
7.32 (dd, 2H, J 8.1 Hz), 2.28 (s, 3H) ppm; 13C NMR (CDCl3): d 21.71
(CH3), 114.93 (CH), 118.64 (CH), 126.76 (CH), 128.49 (2 ꢀ CH),
129.69 (2 ꢀ CH), 130.66 (CH), 131.97 (C), 141.14 (C), 142.85 (C),
148.52 (C), 166.48 (C@O) ppm; HRMS-ESI m/z [M+H]+ calcd for
C14H12N2NaO3: 279.07456, found: 279.07401.
4.2.3.4. 3-Amino-N-(4-methoxy-phenyl)-benzamide (151). Pre-
pared from compound 138 (4.87 g, 0.019 mol) to yield 151 as grey
powder (1.58 g, 34%); mp: 202–204 °C;
mmax (KBr disc) 3387, 3321,
1651 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 3.72 (s, 3H), 5.29
;
(s, 2H), 6.71 (dd, 1H, J 0.9, 7.2 Hz), 6.88 (d, 2H, J 9 Hz), 7.00–7.19
(m, 3H), 7.62 (dd, 2H, J 2.1, 4.8 Hz), 9.92 (s, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6): d 55.82 (CH3), 113.60 (CH), 114.34 (2 ꢀ CH),
115.29 (CH), 117.28 (CH), 122.44 (2 ꢀ CH), 129.40 (CH), 133.14
(C), 136.70 (C), 149.40 (C), 156.00 (C), 166.61 (C@O) ppm;
HRMS-ESI m/z [M+H]+ calcd for C14H15N2O2: 244.11671, found:
244.11616.
4.2.2.4. 4-Methyl-N-(3-nitro-phenyl)-benzenesulfonamide (147).
Prepared from 141 (10.0 g, 50 mmol) and 4-methyl benzenesulfo-
nyl chloride 144 (15.0 g, 78 mmol) to yield 147 as white crystalline
powder (11.1 g, 75%) mp: 182–185 °C; mmax (KBr disc) 3580, 1535,
1381, 1350 cmꢁ1; 1H NMR (CDCl3): d 2.4 (s, 3H), 7.36 (m, 3H), 7.554
(t, 1H), 7.791 (d, 1H, J 8.4 Hz), 7.88 (s, 2H), 8.3 (d, 2H, J 7.2 Hz); 13C
NMR (CDCl3): d 22.00 (CH3), 125.23 (CH), 126.91 (CH), 128.79
(2 ꢀ CH), 130.13 (2 ꢀ CH), 130.16 (CH), 135.86 (C), 136.15 (C),
137.94 (CH), 146.00 (C), 148.66 (C) ppm; HRMS-ESI m/z [MꢁH]+
calcd for C13H11N2O4S: 292.04731, found: 292.04786.
4.2.3.5. N-(3-Amino-phenyl)-benzamide (152). This compound
was prepared from 145 (9.2 g, 0.04 mol) to yield 152 as white pow-
der (5.2 g, 49%) mp: 124 °C; mmax (thin film) 3587, 1651, 1543,
1458 cmꢁ1 1H NMR (CDCl3): d 3.61 (s, 2H), 6.43 (dd, 1H, J 7.8,
;
0.9 Hz), 6.84 (dd, 1H, J 7.8, 0.9 Hz), 7.08 (t, 1H), 7.23–7.49 (m,
5H), 7.80 (d, 2H, J 5.25 Hz); 13C NMR (CDCl3): d 107.31 (CH),
110.56 (CH), 111.60 (CH), 127.35 (2 ꢀ CH), 128.89 (2 ꢀ CH),
129.96 (CH), 131.91 (CH), 135.34 (C), 139.27 (C), 147.56 (C),
166.27 (C@O) ppm; HRMS-ESI m/z [M+H]+ calcd for C13H13N2O:
213.10279, found: 213.10224.
4.2.3. Reduction of nitro-phenylbenzamides derivatives (148–
153)
To a magnetically-stirred aqueous suspension of the particular
nitro-phenylbenzamide (1.0 equiv in 300 mL), Na2S2O4 (5.0 equiv)

3110
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
4.2.3.6. N-(3-Amino-phenyl)-4-methyl-benzamide (153). Pre-
pared from 146 (6.5 g, 0.025 mol) to yield 153 as white powder
(1.9 g, 34%) mp: 138–139 °C; mmax (thin film) 3564, 1650, 1616
(t, 1H), 7.62 (d, 2H, J 7.8 Hz), 7.65 (d, 1H, J 6.6 Hz), 7.81 (d, 1H, J
7.8 Hz), 8.06 (s, 1H), 10.19 (s, 1H), 10.5 (s, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6): d 21.19 (CH3), 44. 23 (CH2), 119.52 (CH),
121.02 (2 ꢀ CH), 122.85 (CH), 123.41 (CH), 129.64 (2 ꢀ CH),
133.33 (C), 136.60 (C), 137.26 (C), 139.29 (C), 165.57 (C@O),
165.84 (C@O) ppm; HRMS-ESI m/z [M+Na]+ calcd for
C16H15ClN2NaO2: 325.07197, found: 325.07143.
cmꢁ1 1H NMR (CDCl3): d 2.35 (s, 3H), 5.06 (s, 2H), 6.27 (d, 1H, J
;
7.2 Hz), 6.83 (d, 1H, J 7.5 Hz), 6.91 (m, 1H), 7.08 (s, 1H), 7.28
(d, 2H, J 7.8 Hz), 7.81(d, 2H, J 7.8 Hz), 9.84 (s, 1H); 13C NMR (CDCl3):
d 21.66 (CH3), 106.75 (CH), 109.02 (CH), 110.31 (CH), 128.32
(2 ꢀ CH), 129.44 (2 ꢀ CH), 129.50 (CH), 133.07 (C), 140.47 (C),
141.95 (C), 149.58 (C), 165.76 (C@O) ppm; HRMS-ESI m/z [M+H]+
calcd for C14H15N2O: 228.12179, found: 228.12124.
4.2.5.2. 3-(2-Chloro-acetylamino)-N-m-tolyl-benzamide (157).
Prepared from compound 149 (3.7 g, 0.016 mol) to yield 157 as
grey crystalline powder (1.65 g, 34%) mp: 208–209 °C; mmax (KBr
disc) 3259, 1670, 1651, 1597 cmꢁ1 1H NMR (300 MHz, DMSO-
;
4.2.4. Reduction nitro-phenylbenzsulfonamides derivatives
(154, 155)
d6): d 2.28 (s, 3H), 4.26 (s, 2H), 6.90 (d, 1H, J 7.2 Hz), 7.2 (t, 1H),
7.45 (d, 1H, J 8.1 Hz), 7.49 (t, 1H), 7.58 (s, 1H), 7.64 (d, 1H, J
6.6 Hz), 7.8 (d, 1H, J 7.8 Hz), 8.06 (s, 1H), 10.19 (s, 1H), 10.5 (s,
1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 21.92 (CH3), 44.23
(CH2), 118.21 (CH), 119.53 (CH), 121.54 (CH), 122.91 (CH),
123.43 (CH), 125.09 (CH), 129.14 (CH), 129.61 (CH), 136.57 (C),
138.44 (C), 139.30 (C), 139.7 (C), 165.57 (C@O), 165.98 (C@O)
ppm; HRMS-ESI m/z [M+Na]+ calcd for C16H15ClN2NaO2:
325.07197, found: 325.07143.
To a magnetically-stirred suspension of the particular nitro-
phenylbenzosulfonamide derivative (1.0 equiv) in anhydrous
ethanol (100 ml), SnCl2ꢄ2H2O (5.0 equiv) was added. The reaction
mixture heated to reflux for 2 h then allowed to cool at room tem-
perature; poured onto ice (100 mL) and made slightly basic (pH
8.5) by addition of saturated NaHCO3 solution. Subsequently, solid
NaCl (ca. 1.0 g) was added and the resulting mixture was extracted
with ethyl acetate (3 ꢀ 100 mL). The organic layers were combined
and washed thoroughly with brine (3 ꢀ 50 mL) and dried over
anhydrous MgSO4. Finally, ethyl acetate was evaporated under re-
duced pressure to yield the free base of the title compounds
(Scheme 5).64,65
4.2.5.3. 3-(2-Chloro-acetylamino)-N-(4-chloro-phenyl)-benzam-
ide (158). Prepared from compound 150 (2.5 g, 0.018 mol) to
yield 158 as light brown powder (4.5 g, 77%) mp: 245–246 °C;
mmax (KBr disc) 3282, 1670, 1651, 1593 cmꢁ1 1H NMR (300 MHz,
;
DMSO-d6): d 4.24 (s, 2H), 7.40 (d, 2H, J 4.98 Hz), 7.48 (t, 1H), 7.64
(d, 1H, J 5.2 Hz), 7.78 (d, 2H, J 4.98 Hz), 7.82 (d, 1H, J 6.4 Hz),
8.09 (s, 1H), 10.4 (s, 1H), 10.54 (s, 1H) ppm; 13C NMR (75 MHz,
DMSO-d6): d 44.23 (CH2), 119.53 (CH), 122.50 (2 ꢀ CH), 123.09
(CH), 123.49 (CH), 127.98 (C), 129.23 (2 ꢀ CH), 129.68 (CH),
136.23 (C), 138.77 (C), 139.36 (C), 165.61 (C@O), 166.16 (C@O)
ppm; [M+Na]+ found 345.01680 C15H12Cl2N2NaO2 requires
345.01735.
4.2.4.1. 3-Amino-N-(4-methoxy-phenyl)-benzenesulfonamide
(154). Prepared from compound 140 (3.11 g, 0.01 mol) to yield
154 as brown powder (1.49 g, 53%) mp: 154–155 °C; mmax (KBr
disc) 3448, 3371, 1635, 1600 cmꢁ1 13C NMR (75 MHz, DMSO-d6):
;
d 56.15 (CH3), 112.60 (CH), 114.86 (2 ꢀ CH), 115.06 (2 ꢀ CH),
119.53 (CH), 126.71 (C), 130.38 (CH), 133.29 (CH), 140.09 (C),
150.18 (C), 160.94 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C13H14N2O3S: 279.07587, found: 279.07532.
4.2.5.4. 3-(2-Chloro-acetylamino)-N-(4-methoxy-phenyl)-benz-
amide (159). Prepared from compound 151 (1.49 g, 6.2 mmol) to
4.2.4.2. N-(3-Amino-phenyl)-4-methyl-benzenesulfonamide (155).
Prepared from 147 (8.8 g, 0.03 mol) to yield 155 as brown powder
(4.0 g, 51%) mp: 230–233 °C; m ;
max (thin film) 3564, 1519 cmꢁ1 1H
yield 159 as white powder (0.74 g, 37%) mp: 242–242 °C; mmax (KBr
disc) 3305, 1674, 1647, 1593 cmꢁ1; 1H NMR (300 MHz, DMSO-d6):
d 3.72 (s, 3H), 4.26 (s, 2H), 6.90 (dd, 2H, J 1.5, 5.1 Hz), 7.45 (t, 1H),
7.63 (d, 3H, J 8.4 Hz), 7.78 (d, 1H, J 7.2 Hz), 8.06 (s, 1H), 10.15 (s,
1H), 10.50 (s, 1H) ppm; HRMS-ESI m/z [M+Na]+ calcd for
C16H15ClN2NaO3: 341.06689, found: 341.06634.
NMR (300 MHz, DMSO-d6): d 2.479 (s, 3H), 5.35 (s, 2H, D2O
exchangable, NH2), 5.99 (dd, 1H, J 7.8, 0.9 Hz), 6.30 (s, 1H), 6.61
(dd, , 1H, J 8.2, 1.5 Hz), 6.97 (t, 1H), 7.45 (d, 2H, J 8.1 Hz), 7.67 (d,
2H,
J 8.25 Hz), 7.70 (s, D
exchangble, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6): d 21.86 (CH3), 116.14 (CH), 116.95 (CH),
118.57 (CH), 128.68 (2 ꢀ CH), 129.98 (2 ꢀ CH), 130.49 (CH),
134.95 (C), 136.85 (C), 145.80 (C), 150.50 (C) ppm; HRMS-ESI m/z
[MꢁH]+ calcd for C13H13N2O2S: 261.06977, found: 261.13107.
4.2.5.5. N-[3-(2-Chloro-acetylamino)-phenyl]-benzamide (160).
Prepared from 152 (5.8 g, 0.027 mol) to yield 160 as white powder
(5.1 g, 69%) mp: 240–243 °C; mmax (thin film) 3286, 1670,
1647 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 4.26 (s, 2H), 7.27–
;
8.16 (m, 9H), 10.33 (s, 1H), 10.38 (s, 1H); 13C-NMR (DMSO): d
44.34 (CH2), 112.19 (CH), 115.49 (CH), 116.69 (CH), 128.41
(2 ꢀ CH), 129.06 (2 ꢀ CH), 129, 63 (CH), 132.29 (CH), 135.60 (C),
139.38 (C), 140.28 (C), 165.33 (C@O), 166.34 (C@O) ppm; HRMS-
ESI m/z [M+Na]+ calcd for C15H13ClN2NaO2: 311.05632, found:
311.05913.
4.2.5. Synthesis of the mono-chloromethyl-acetamido
derivatives (156–163)
To a magnetically-stirred, ice-bathed, solution or suspension of
the particular aniline (148–155, 1.0 equiv) and triethylamine
(2.0 equiv) in dry acetone (75 mL), 2-chloroacetyl chloride
(1.0 equiv) in dry acetone (100 mL) was gradually added over
30 min. The reaction mixture was stirred at room temperature un-
til TLC revealed complete consumption of the starting aniline. Sub-
sequently, the reaction mixture was poured slowly onto sufficient
5% aqueous sodium bicarbonate to neutralize the generated acid.
The precipitated crude products were purified by recrystallization
from acetone/water (Scheme 6).66
4.2.5.6. N-[3-(2-Chloro-acetylamino)-phenyl]-4-methyl-benzam-
ide (161). Prepared from 153 (1.5 g, 6.6 mmol) to yield 161 as
pale yellow powder (1.3 g, 67%) mp: 233–235 °C; mmax (thin film)
3502, 1651, 1539 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.36 (s,
;
3H), 4.23 (s, 2H), 7.26–7.32 (m, 5H), 7.42 (s, 1H), 7.85 (d, 1H, J
7.8 Hz), 8.11 (s, 1H), 10.21 (s, 1H), 10.34 (s, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6): d 21.70 (CH3), 44.31 (CH2), 112.16 (CH),
115.34 (CH), 116.65 (CH), 128.42 (2 ꢀ CH), 129.57 (2 ꢀ CH),
132.65 (C), 139, 31 (C), 140.31 (C), 142.30 (C), 165.28 (C@O),
166.09 (C@O) ppm; HRMS-ESI m/z [MꢁH]+ calcd for C16H14ClN2O2:
301.07438, found: 301.07494.
4.2.5.1. 3-(2-Chloro-acetylamino)-N-p-tolyl-benzamide (156).
Prepared from compound 148 (2.82 g, 0.0125 mol) to yield 156 as
grey crystalline powder (2.02 g, 53%), mp: 228 (decomp.) °C; mmax
(KBr disc) 3448, 1670, 1647, 1597 cmꢁ1 1H NMR (300 MHz,
;
DMSO-d6): d 2.25 (s, 3H), 4.26 (s, 2H), 7.12 (d, 2H, J 7.8 Hz), 7.46

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3111
4.2.5.7. 2-Chloro-N-[3-(4-methoxy-phenylsulfamoyl)-phenyl]-acet-
amide (162). Prepared from compound 154 (1.49 g, 5.3 mmol) to
yield 162 as light brown powder (0.57 g, 30%) mp: 165–169 °C;
4.2.6.3. 1-{[3-(4-Chloro-phenylcarbamoyl)-phenylcarbamoyl]-
methyl}-pyridinium chloride (166). Prepared from 158 (0.3 g,
1.0 mmol) and pyridine to yield 166 as white powder (0.19 g,
55.5%) mp: 266–269 °C (Decomp.); mmax (KBr disc) 3448, 3294,
mmax (KBr disc) 3352, 1670, 1593, 1535 cmꢁ1 1H NMR (300 MHz,
;
DMSO-d6): d 3.77 (s, 2H), 4.29 (s, 3H), 6.95–7.62 (m, 6H), 7.92 (d,
2H, J 7.2 Hz), 8.16 (s, 1H), 10.74 (s, 1H) ppm; 13C NMR (75 MHz,
DMSO-d6): d 44.15 (CH2), 56.23 (CH3), 115.41 (CH), 118.81 (CH),
123.57 (CH), 125.39 (CH), 126.14 (C), 130.89 (2 ꢀ CH), 133.26
(2 ꢀ CH), 139.86 (C), 139.98 (C), 161.27 (C), 165.95 (C@O) ppm;
HRMS-ESI m/z [M]+ calcd for C15H15ClN2O4S: 354.04410, found:
354.26951.
1689, 1651 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 5.76 (s, 2H),
;
7.37 (d, 2H, J 8.7 Hz), 7.48 (t, 1H), 7.72 (dd, 1H, J 0.9, 7.5 Hz),
7.81 (d, 2H, J 6.6 Hz), 7.83 (s, 1H), 8.17 (d, 1H, J 4.8 Hz), 8.22 (d,
2H, J 6.6 Hz), 8.67 (t, 1H), 9.10 (d, 2H, J 6.6 Hz), 10.48 (s, 1H),
11.41 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 62.82 (CH2),
119.46 (CH), 122.56 (2 ꢀ CH), 122.92 (CH), 123.57 (CH), 127.97
(C), 128.22 (2 ꢀ CH), 129.20 (2 ꢀ CH), 129.69 (CH), 136.26 (C),
138.77 (C), 139.26 (C), 146.96 (2 ꢀ CH), 147.15 (CH), 164.28
(C@O), 166.06 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C20H17ClN3O2: 366.10093, found: 366.10038.
4.2.5.8. 2-Chloro-N-[3-(toluene-4-sulfonylamino)-phenyl]-acet-
amide (163). Prepared from 155 (13.0 g, 0.049 mol) to yield 163
as light brown powder (14.57 g, 80%) mp 173–176 °C; mmax (thin
film) 3564, 1720, 1543 cmꢁ1
;
1H NMR (300 MHz, DMSO-d6): d
4.2.6.4. 1-{[3-(4-Methoxy-phenylcarbamoyl)-phenylcarbamoyl]-
methyl}-pyridinium chloride (167). Prepared from compound
159 (0.2 g, 0.6 mmol) and pyridine to yield 167 as Off-white powder
2.06 (s, 3H), 4.23 (s, 2H), 6.61 (d, 1H, J 7.8 Hz), 7.35 (m, 1H), 7.41
(s, 1H), 7.41 (m, 2H), 7.46 (d, 2H, J 8.1 Hz), 7.67 (d, 2H, J 8.1 Hz),
10.49 (s, 1H); 13C NMR (75 MHz, DMSO-d6): d 21.88 (CH3), 44.20
(CH2), 121.48 (CH), 122.50 (CH), 127.03 (CH), 128.71 (2 ꢀ CH),
130.41 (CH), 130.66 (2 ꢀ CH), 134.67 (C), 136.36 (C), 140.12 (C),
146.19 (C), 165.60 (C@O) ppm.
(0.2 g, 85%); mp: 245–246 °C (Decomp.); mmax(KBr disc) 3406, 1685,
1620 cmꢁ1; 1H NMR (300 MHz, DMSO-d6): d 3.71 (s, 2H), 5.74 (s, 2H),
6.90 (dd, 2H, J 3, 9 Hz), 7.47 (t, 1H), 7.64 (dd, 2H, J 3, 6 Hz), 7.70 (d, 1H,
J 8.1 Hz), 7.80 (dd, 1H, J 7.8, 1.2 Hz), 8.14 (s, 1H), 8.21 (dd, 2H, J 6.6,
7.8 Hz), 8.65 (t, 1H), 9.08 (d, 2H, J 5.4 Hz), 10.19 (s, 1H), 11.32 (s,
1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 55.85 (CH3), 62.81 (CH2),
114.40 (2 ꢀ CH), 119.43 (CH), 122.61 (2 ꢀ CH), 122.68 (CH), 123.40
(CH), 128.23 (2 ꢀ CH), 129.62 (CH), 132.80 (C), 136.64 (C), 139.16
(C), 146.96 (CH), 147.15 (2 ꢀ CH), 156.23 (C), 164.23 (C@O), 165.49
(C@O) ppm; HRMS-ESI m/z [M]+ calcd for C21H20N3O3: 363.15382,
found: 363.15328.
4.2.6. Synthesis of the pyridinum acetamido-phenylbenzamides
and -phenylbenz sulfonamides (164–185)
To magnetically-stirred neat pyridine, nicotinamide or N-phen-
ylethyl nicotinamide (178) (3.0 equiv) heated to 120–130 °C, the
particular mono-chloromethyl-acetamido-derivative (156–163)
was added as neat powder (1.0 equiv). The reaction mixture was
stirred at 120–130 °C for 10–15 min then cooled to room temper-
ature. Subsequently, the resulting solid mass was suspended in
dry acetone (50 mL) and stirred at room temperature for 1 h. The
resulting suspensions were filtered and the solid residues were fur-
ther washed with acetone (3 ꢀ 100 mL) over 1 h periods to yield
the pyridinum derivatives (Scheme 7).
4.2.6.5. 1-[ (3-Benzoylamino-phenylcarbamoyl)-methyl]-pyrid-
inium chloride (168). Prepared from 160 (0.286 g, 0.99 mmo-
l)and pyridine to yield 168 (0.3 g, 88%) mp: 238–239 °C
(Decomp.); mmax (thin film) 3444, 1730, 1543 cmꢁ1 1H NMR
;
(300 MHz, DMSO-d6): d 5.75 (s, 2H), 7.25–7.56 (m, 7H), 7.95 (d,
2H, J 7.2 Hz), 8.21 (t, 2H), 8.67 (t, 1H), 9.11 (d, 2H, J 6 Hz), 10.38
(s, 1H), 11.22 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 62.87
(CH2), 112.16 (CH), 115.30 (CH), 116.79 (CH), 128.18 (CH), 128.43
(2 ꢀ CH), 129.03 (2 ꢀ CH), 129.60 (CH), 132.3 (CH), 135.30 (CH),
139.18 (CH), 140.35 (CH), 146.88 (2 ꢀ CH), 147.13 (2 ꢀ CH),
163.94 (C@O), 166.27 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C20H18N3O2: 332.13990, found: 332.13935.
4.2.6.1. 1-[(3-p-Tolylcarbamoyl-phenylcarbamoyl)-methyl]-py-
ridinium chloride (164). Prepared from compound 156 (0.3 g,
1.0 mmol) and pyridine to yield 164 as white powder (0.3 g,
80%) mp: 263–266 °C (Decomp.); mmax (KBr disc) 3448, 1689,
1670 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.24 (s, 3H), 5.7 (s,
;
2H), 7.11 (d, 2H, J 8.7 Hz), 7.46 (t, 1H), 7.62 (d, 2H, J 8.4 Hz), 7.70
(d, 1H, J 7.8 Hz), 7.82 (d, 1H, J 7.8 Hz), 8.18 (m, 3H), 8.70 (t, 1H),
9.12 (d, 2H, J 6 Hz), 10.26 (s, 1H), 11.52 (s, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6): d 21.18 (CH3), 62.82 (CH2), 119.44 (CH),
121.08 (2 ꢀ CH), 122.68 (CH), 123.47 (CH), 128.21 (CH),
129.57 (2 ꢀ CH), 129.65 (2 ꢀ CH), 133.31 (C), 136.63 (C), 137.26
(C), 139.23 (C), 146.93 (CH), 147.14 (2 ꢀ CH), 164.25 (C@O),
165.79 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for C21H20N3O2:
346.15555, found: 346.15500.
4.2.6.6. 1{[3-(4-Methyl-benzoylamino)-phenylcarbamoyl]-me-
thyl}-pyridinium chloride (169). Prepared from 161 (0.283 g,
0.93 mmol) and pyridine to yield 169 (0.31 g, 89%): mp: 262–
265 °C (Decomp.); mmax (thin film) 3614, 1654, 1539 cmꢁ1 1H
;
NMR (300 MHz, DMSO-d6): d 2.37 (s, 3H), 5.70 (s, 2H), 7.28–7.44
(m, 7H), 7.86 (d, 2H, J 7.8 Hz), 8.22 (t, 2H), 8.67 (t, 1H), 9.08 (d,
1H, J 6 Hz), 10.25 (s, 1H), 11.02 (s, 1H) ppm; 13C NMR (75 MHz,
DMSO-d6): d 21.72 (CH3), 62.86 (CH2), 112.04 (CH), 115.12 (CH),
116.72 (CH), 128.19 (2 ꢀ CH), 128.44 (2 ꢀ CH), 129.58 (CH),
129.65 (2 ꢀ CH), 132.60 (C), 139.11 (C), 140.46 (C), 142.33 (C),
146.91 (CH), 147.15 (2 ꢀ CH), 163.92 (1C@O), 166.09 (1C@O)
ppm; HRMS-ESI m/z [M]+ calcd for C21H20N3O2: 346.15555, found:
346.15500.
4.2.6.2. 1-[(3-m-Tolylcarbamoyl-phenylcarbamoyl)-methyl]-py-
ridinium chloride (165). Prepared from compound 157 (0.3 g,
1.0 mmol) and pyridine to yield 165 as white powder (0.28 g,
80%) mp: 271 °C (Decomp.); mmax (KBr disc) 3448, 3352, 1701,
1654, 1539 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.27 (s, 3H),
;
5.74 (s, 2H), 6.90 (d, 1H, J Hz), 7.19 (t, 1H), 77.51 (m, 3H), 7.69
(d, 1H, J 7.8), 7.82 (d, 1H, J 7.8 Hz), 8.14 (s, 1H), 8.21 (t, 2H), 8.68
(t, 1H), 9.09 (d, 2H, J 6 Hz), 10.23 (s, 1H), 11.35 (s, 1H) ppm; 13C
NMR (75 MHz, DMSO-d6): d 21.89 (CH3), 62.84 (CH2), 118.29
(CH), 119.52 (CH), 121.63 (CH), 122.78 (CH), 123.50 (CH), 125.11
(CH), 128.23 (2 ꢀ CH), 129.11 (CH), 129.64 (CH), 136.56 (C),
138.41 (C), 139.14 (C), 139.67 (C), 146.98 (2 ꢀ CH), 147.14 (CH),
164.20 (C@O), 165.86 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C21H20N3O2: 346.15555, found: 346.15500.
4.2.6.7. 1-{[3-(4-Methoxy-phenylsulfamoyl)-phenylcarbamoyl]-
methyl}-pyridinium chloride (170). Prepared from compound
162 (0.1 g, 0.3 mmol) and pyridine to yield 170 as light brown
powder (0.11 g, 98%); mp 209 °C (Decomp.); mmax (KBr disc)
3421, 1701, 1635 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 3.76 (s,
;
3H), 5.77 (s, 2H), 6.93 (s, 1H), 7.50 (d, 2H, J 8.1 Hz), 7.65 (m, 3H),
7.97 (d, 2H, J 8.4 Hz), 8.24 (m, 3H), 8.71 (t, 1H), 9.10 (d, 2H, J
5.4 Hz), 11.61 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 56.20
(CH3), 62.79 (CH2), 115.36 (2 ꢀ CH), 118.64 (2 ꢀ CH), 123.68

3112
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
(CH), 125.23 (CH), 126.04 (C), 128.21 (2 ꢀ CH), 131.01 (CH), 133.18
(CH), 139.75 (C), 139.94 (C), 147.02 (CH), 147.15 (2 ꢀ CH), 161.18
(C), 164.68 (C@O) ppm; HRMS-ESI m/z [MꢁH]+ calcd for
C20H19N3O4S: 397.10963, found: 397.10908.
123.41 (CH), 127.90 (CH), 129.61 (2 ꢀ CH), 132.79 (C), 133.96
(C), 136.66 (C), 139.08 (C), 144.71 (CH), 147.40 (CH), 148.72
(CH), 156.25 (C), 163.43 (C@O), 163.99 (C@O), 165.44 (C@O)
ppm; HRMS-ESI m/z [M]+ calcd for C22H21N4O4: 406.15964, found:
406.15909.
4.2.6.8. 3-Carbamoyl-1-[(3-p-tolylcarbamoyl-phenylcarbamoyl)-
methyl]-pyridinium chloride (171). Prepared from compound
156 (0.3 g, 0.99 mmol) and nicotinamide to yield 171 as white
powder (0.35 g, 83%) mp: 287–290 °C (Decomp.); mmax (KBr disc)
4.2.6.12. 1-[(3-Benzoylamino-phenylcarbamoyl)-methyl]3-car-
bamoyl-pyridiniumchloride (175). Prepared from 160 (0.5 g,
1.48 mmol) and nicotinamide to yield (0.17 g, 49%) mp: 286–
3298, 1689, 1651 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.24 (s,
288 °C (Decomp.); mmax (thin film) 3564, 1732, 1508 cmꢁ1 1H
;
;
3H), 5.81 (s, 2H), 7.11 (d, 2H, J 7.8 Hz), 7.46 (t, 1H), 7.61 (d, 2H, J
7.8 Hz), 7.69 (d, 1H, J 7.8 Hz), 7.80 (d, 1H, J 7.8 Hz), 8.16 (d, 2H, J
12 Hz), 8.32 (t, 1H), 8.77 (s, 1H), 9.09 (d, 1H, J 7.8 Hz), 9.21 (d,
1H, J 5.7 Hz), 9.62 (s, 1H), 10.24 (s, 1H), 11.39 (s, 1H) ppm; 13C
NMR (75 MHz, DMSO-d6): d 21.19 (CH3), 63.06 (CH2), 119.4 7
(CH), 121.09 (2 ꢀ CH), 122.70 (CH), 123.51 (CH), 127.91 (CH),
129.66 (3 ꢀ CH), 133.32 (C), 133.89 (C), 136.64 (C), 137.25 (C),
139.16 (C), 144.80 (CH), 147.40 (CH), 148.74 (CH), 163.43 (C@O),
164.04 (C@O), 165.74 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C22H21N4O3: 390.16472, found: 390.16411.
NMR (300 MHz, DMSO-d6): d 5.77 (s, 2H), 7.29–7.60 (m, 7H),
7.95 (d, 1H, J 8.4 Hz), 8.31 (d, 1H, J 7.5 Hz), 8.33 (t, 1H), 8.76 (s,
2H), 9.10 (d, 1H, J 8.1 Hz), 9.21 (d, 1H, J 6 Hz), 9.61 (s, 1H), 10.36
(s, 1H), 11.07 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 63.09
(CH2), 112.05 (CH), 115.24 (CH), 116.74 (CH), 127.87 (CH), 128.42
(2 ꢀ CH), 129.05 (2 ꢀ CH), 129.68 (CH), 132.32 (CH), 133.87 (C),
135.51 (C), 139.13 (C), 140.38 (C), 144.71 (CH), 147.44 (CH),
148.44 (CH), 163.47 (C@O), 166.30 (C@O), 175.63 (C@O) ppm;
HRMS-ESI m/z [M]+ calcd for C21H19N4O3: 375.14572, found:
375.14517.
4.2.6.9. 3-Carbamoyl-1-[(3-m-tolylcarbamoyl-phenylcarbamo-
yl)-methyl]-pyridinium chloride (172). Prepared from com-
pound 157 (0.3 g, 1.0 mmol) and nicotinamide to yield 172 as
Off-white powder (0.14 g, 33%) mp: 242–244 °C (Decomp.); mmax
4.2.6.13. 3-Carbamoyl-1-{[3-(4-methyl-benzoylamino)-phenylc-
arbamoyl]-methyl}-pyridinium chloride (176). Prepared from
161 (0.25 g, 0.87 mmol) and nicotinamide to yield (0.22 g, 64%)
mp: 289 °C (Decomp.); mmax (thin film) 3730, 1680, 1539 cmꢁ1
;
(KBr disc) 3298, 1685, 1651 cmꢁ1 1H NMR (300 MHz, DMSO-d6):
1H NMR (300 MHz, DMSO-d6): d 2.47 (s, 3H), 5.77 (s, 2H), 7.23–
7.46 (m, 5H), 7.86 (d, 2H, J 8.4 Hz), 8.20 (t, 1H), 8.31 (dd, 1H, J
2.1, 6 Hz), 8.78 (s, 2H), 9.11 (d, 1H, J 8.4 Hz), 9.21 (d, 1H, J 6 Hz),
9.62 (s, 1H), 10.27 (s, 1H), 11.12 (s, 1H) ppm; 13C NMR (75 MHz,
DMSO-d6): d 21.7 (CH3), 63.08 (CH2), 112.12 (CH), 115.16 (CH),
116.76 (CH), 127.85 (CH), 128.46 (2C), 129.54 (3 ꢀ CH), 132.58
(C), 133.83 (C), 139.10 (C), 140.44 (C), 142.30 (C), 144.72 (CH),
147.41 (CH), 148.72 (CH), 163.44 (C@O), 163.71 (C@O), 166.07
(C@O) ppm; HRMS-ESI m/z [M]+ calcd for C22H21N4O3: 390.16472,
found: 390.16411.
;
d 2.27 (s, 3H), 5.79 (s, 2H), 6.88 (d, 1H, J 7.2 Hz), 7.19 (t, 1H),
7.46 (t, 1H), 7.52 (d, 1H, J 9), 7.58 (s, 1H), 7.69 (d, 1H, J 8.1 Hz),
7.81 (dd, 1H, J 1.2, 6.9 Hz), 8.13 (s, 1H), 8.19 (s, 1H), 8.32 (t, 1H),
8.73 (s, 1H), 9.07 (d, 1H, J 8.1 Hz), 9.20 (d, 1H, J 6 Hz), 9.6 (s, 1H),
10.23 (s, 1H), 11.30 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d
21.9 (CH3), 63.03 (CH2), 118.26 (CH), 119.46 (CH), 121.60 (CH),
122.74 (CH), 123.55 (CH), 125.11 (CH), 127.91 (CH), 129.13 (CH),
129.67 (CH), 133.93 (C), 136.63 (C), 138.42 (C), 139.15 (C),
139.68 (C), 144.76 (CH), 147.43 (CH), 148.75 (CH), 163.46 (C@O),
164.07 (C@O), 165.87 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C22H21N4O3: 390.16472, found: 390.16411.
4.2.6.14. 3-Carbamoyl-1-{[3-(4-methoxy-phenylsulfamoyl)-phe-
nylcarbamoyl]-methyl}-pyridinium chloride (177). Prepared
from compound 162 (0.122 g, 3.4 mmol) and nicotinamide to yield
4.2.6.10. 3-Carbamoyl-1-{[3-(4-chloro-phenylcarbamoyl)-phe-
nylcarbamoyl]-methyl}-pyridinium chloride (173). Prepared
from compound 158 (0.3 g, 1.0 mmol) and nicotinamide as white
powder (0.168 g, 44%) mp: 294–296 °C (Decomp.); mmax (KBr disc)
177 (0.10 g, 64%) mp: 210–212 °C (Decomp.);
mmax (KBr disc) 3406,
1689, 1600 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 3.76 (s, 3H),
;
5.82 (s, 2H), 6.94 (s, 1H), 7.51 (d, 2H, J 7.8 Hz), 7.65 (m, 3H), 7.96
(d, 2H, J 8.4 Hz), 8.19 (s, 1H), 8.26 (s, 1H), 8.35 (t, 1H), 8.72 (s,
1H), 9.10 (d, 1H, J 7.8 Hz), 9.22 (d, 1H, J 5.7 Hz), 9.61 (s, 1H),
11.64 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 56.20 (CH3),
63.04 (CH2), 115.39 (2 ꢀ CH), 118.67 (2 ꢀ CH), 123.72 (CH),
125.28 (CH), 126.07 (C), 127.91 (CH), 131.01 (CH), 132.01 (C),
133.95 (CH), 139.75 (C), 139.96 (C), 144.80 (CH), 147.46 (CH),
148.74 (CH), 151.74 (C), 163.43 (C@O), 164.50 (C@O) ppm;
HRMS-ESI m/z [MꢁH]+ calcd for C21H20N4O5S: 441.11880, found:
441.11825.
3282, 1693, 1651 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 5.78 (s,
;
2H), 7.36 (d, 2H, J 8.7 Hz), 7.49 (t, 1H), 7.71 (d, 1H J 7.2 Hz), 7.79
(d, 3H, J 8.7 Hz), 8.16 (d, 2H, J 12 Hz), 8.32 (t, 1H), 8.71 (s, 1H),
9.07 (d, 1H, J 8.4 Hz), 9.19 (d, 1H, J 6.3 Hz), 9.58 (s, 1H), 10.46 (s,
1H), 11.28 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 63.02
(CH2), 119.47 (CH), 122.55 (2 ꢀ CH), 122.92 (CH), 123.61 (CH),
127.91 (CH), 129.20 (2 ꢀ CH), 129.74 (CH), 129.75 (C), 133.91 (C),
136.27 (C), 138.75 (C), 139.19 (C), 144.74 (CH), 147.42 (CH),
148.74 (CH), 163.45 (C@O), 164.08 (C@O), 166.02 (C@O) ppm;
HRMS-ESI m/z [M]+ calcd for C21H18ClN4O3: requires 409.10674,
found: 409.10619.
4.2.6.15. 3-Carbamoyl-1-{[3-(toluene-4-sulfonylamino)-phenylc-
arbamoyl]-methyl}-pyridinium chloride (178). Prepared from
163 (0.5 g, 1.4 mmol) and nicotinamide to yield 178 as white pow-
der (0.46 g, 68%) mp: 260 °C (Decomp.); mmax (thin film) 3614,
4.2.6.11. 3-Carbamoyl-1-{[3-(4-methoxy-phenylcarbamoyl)-phe-
nylcarbamoyl]-methyl}-pyridinium chloride (174). Prepared
from compound 159 (0.2 g, 0.6 mmol) and nicotinamide to yield
174 as light brown powder (0.23 g, 87%) mp: 255–257 °C
1700, 1539, 1219 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.31 (s,
;
3H), 5.76 (s, D2O exchangeable, 2H), 6.63 (d, 1H, J 8.1 Hz), 7.34–
7.65 (m, 10H), 8.21 (s, 1H), 8.34 (t, 1H), 8.67 (s, 1H), 9.1 (dd, 2H,
J 20.1, 7.8 Hz), 9.53 (s, 1H), 11.11 (s, 1H); 13C NMR (75 MHz,
DMSO-d6): d 21.86 (CH3), 63.024 (CH2), 121.39 (CH), 122.577
(CH), 127.09 (CH), 127.91 (CH), 128.72 (2 ꢀ CH), 130.53 (CH),
130.69 (2 ꢀ CH), 133.90 (C), 134.76 (C), 136.33 (C), 139.92 (C),
144.74 (CH), 146.18 (C), 147.511 (CH), 148.78 (CH), 163.46
(C@O), 164.16 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
(Decomp.); mmax (KBr disc) 3298, 1685, 1647 cmꢁ1
;
1H NMR
(300 MHz, DMSO-d6): d 3.74 (s, 3H), 5.79 (s, 2H), 6.89 (d, 2H, J
9 Hz), 7.96 (t, 1H), 7.65 (d, 2H, J 8.7 Hz), 7.71 (d, 2H, J 9), 7.80
(d, 1H, J 9 Hz), 8.16 (d, 2H, J 9 Hz), 8.34 (t, 1H), 8.69 (s, 1H),
9.08 (d, 1H, J 9 Hz), 9.20 (d, 1H, J 6 Hz), 9.59 (s, 1H), 10.17 (s,
1H), 11.19 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 55.85
(CH3), 63.03 (CH2), 114.4 (2 ꢀ CH), 119.47 (CH), 122.68 (2 ꢀ CH),

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3113
C21H21N4O4S: 426.13171, found: 426.13116; Anal. Calcd for
C21H21ClN4O4S: C, 54.72; H, 4.59. Found: C, 54.50; H, 4.43.
(75 MHz, DMSO-d6): d 35.37 (CH3), 41.84 (CH2), 55.85 (CH2),
63.07 (CH2), 114.37 (2 ꢀ CH), 119.45 (CH), 122.68 (3 ꢀ CH),
123.48 (CH), 126.92 (CH), 127.90 (CH), 129.08 (2 ꢀ CH), 129.35
(2 ꢀ CH), 129.58 (CH), 132.84 (C), 133.92 (C), 136.64 (C), 139.18
(C), 139.81 (C), 144.52 (CH), 147.15 (CH), 148.69 (CH), 156.21 (C),
161.74 (C@O), 164.03 (C@O), 165.52 (C@O) ppm; HRMS-ESI m/z
[M]+ calcd for C30H29N4O4: 509.21888, found: 509.21833.
4.2.6.16. 3-Phenethylcarbamoyl-1-[(3-p-tolylcarbamoyl-phenyl-
carbamoyl)-methyl]-pyridinium chloride (179). Prepared from
compound 156 (0.3 g, 0.99 mmol) and 186 as white powder
(0.36 g, 68%) mp: 231–234 °C (Decomp.); mmax (KBr disc) 3290,
1693, 1647 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.25 (s, 3H),
;
2.88 (t, 2H), 3.56 (t, 2H), 5.7 (s, 2H), 7.13 (d, 2H, J 8.4 Hz), 7.2 (dd,
1H, J 1.5, 7.2 Hz), 7.26 (m, 4H), 7.47 (t, 1H), 7.62 (d, 2H, J 8.1 Hz),
7.70 (d, 1H, J 7.8 Hz), 7.80 (d, , 1H, J 8.1 Hz), 8.14 (s, 1H), 8.32 (t,
1H), 9.06 (d, 1H, J 8.4 Hz), 9.20 (d, J 6 Hz, 1H), 9.52 (s, 1H), 9.58 (s,
1H), 10.23 (s, 1H), 11.32 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-
d6): d 21.19 (CH3), 35.38 (CH2), 41.85 (CH2), 63.05 (CH2), 119.46
(CH), 121.07 (2 ꢀ CH), 122.68 (CH), 123.50 (CH), 126.94 (CH),
127.91 (CH), 129.09 (2 ꢀ CH), 129.36 (2 ꢀ CH), 129.67 (2 ꢀ CH),
133.34 (C), 133.96 (C), 136.66 (C), 137.24 (C), 139.15 (C), 139.78
(C), 140.89 (CH), 144.46 (CH), 147.16 (CH), 148.71 (CH), 161.78
(C@O), 164.04 (C@O), 165.72 (C@O) ppm; HRMS-ESI m/z [M]+ calcd
for C30H29N4O3: 493.22397, found: 493.22342; Anal. Calcd for
C30H29ClN4O3: C, 68.11; H, 5.53. Found: C, 67.80; H, 5.13.
4.2.6.20. 1-[(3-Benzoylamino-phenylcarbamoyl)-methyl]-3-phen-
ethylcarbamoyl-pyridinium chloride (183). Prepared from 160
(0.3 g, 1 mmol) and 186 to yield 183 (0.47 g, 92%) mp: 204 °C (De-
comp.); mmax (KBr disc) 3448, 1705, 1662, 1543 cmꢁ1 1H NMR
;
(300 MHz, DMSO-d6): d 2.88 (t, 2H), 3.56 (t, 2H), 5.77 (s, 2H),
7.15–7.58 (m, 12H), 7.94 (d, 2H, J 7.8 Hz), 8.22 (s, 1H), 8.31 (t,
1H), 9.10 (d, 1H, J 8.1 Hz), 9.20 (d, 1H, J 5.7 Hz), 9.5 (s, 1H), 10.36
(s, 1H), 11.11 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 35.38
(CH2), 41.84 (CH2), 63.11 (CH2), 112.12 (CH), 115.28 (CH), 116.78
(CH), 126.92 (CH), 127.86 (CH), 128.41 (2 ꢀ CH), 129.03 (2 ꢀ CH),
129.09 (2 ꢀ CH), 129.35 (2 ꢀ CH), 129.64 (CH), 132.30 (CH),
133.89 (C), 135.50 (C), 139.13 (C), 139.13 (C), 139.81 (C), 140.38
(C), 144.44 (CH), 147.16 (CH), 148.68 (CH), 161.76 (C@O), 163.72
(C@O), 166.28 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C29H27N4O3: 479.20832, found: 479.20777.
4.2.6.17. 3-Phenethylcarbamoyl-1-[(3-m-tolylcarbamoyl-phenyl-
carbamoyl)-methyl]-pyridinium chloride (180). Prepared from
compound 157 (0.3 g, 0.1 mmol) and 186 as white powder (0.37 g,
70%) mp: 206–209 °C (Decomp.); mmax (KBr disc) 3294, 1689,
4.2.6.21. 1-{[3-(4-Methyl-benzoylamino)-phenylcarbamoyl]-
methyl}-3-phenethylcarbamoyl-pyridinium; chloride (184).
This compound was prepared from 161 (0.3 g, 1.0 mmol) and
186. (0.31 g, 59%) mp: 225 °C (Decomp.); mmax (KBr disc) 3309,
1662 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.27 (s, 3H), 2.85 (t,
;
2H), 3.54 (t, 2H), 5.80 (s, 2H), 6.89 (d, 1H, J 7.2 Hz), 7.22 (m, 5H),
7.55 (m, 4H), 7.70 (d, 1H, J 7.2 Hz), 7.81 (d, 1H, J 7.5 Hz), 8.1 (s,
1H), 8.32 (t, 1H), 9.07 (d, 1H, J 6.9 Hz), 9.20 (d, 1H, J 6.3 Hz), 9.54
(t, 1H), 9.58 (s, 1H), 10.23 (s, 1H), 11.35 (s, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6): d 21.91 (CH3), 35.38 (CH2), 41.84 (CH2), 63.08
(CH2), 118.27 (CH), 119.48 (CH), 121.60 (CH), 122.76 (CH), 123.56
(CH), 125.10 (CH), 126.94 (CH), 127.92 (CH), 129.06 (2 ꢀ CH),
129.12 (CH), 129.36 (2 ꢀ CH), 129.67 (CH), 133.95 (C), 136.64 (C),
138.41 (C), 139.17 (C), 139.69 (C), 139.79 (C), 144.49 (CH), 147.18
(CH), 148.71 (CH), 161.75 (C@O), 164.05 (C@O), 165.89 (C@O)
ppm; HRMS-ESI m/z [M]+ calcd for C30H29N4O3: 493.22397, found:
493.22342; Anal. Calcd for C30H29ClN4O3: C, 68.11; H, 5.53. Found:
C, 68.35; H, 5.64.
1689, 1666, 1608, 1562 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d
;
2.35 (s, 3H), 2.88 (t, 2H), 3.53 (t, 2H), 5.78 (2, 2H), 7.18–7.47 (m,
9H), 7.86 (d, 2H, J 7.5 Hz), 8.22 (s, 1H), 8.31 (t, 1H), 9.11 (d, , 1H,
J 7.5 Hz), 9.20 (d, 1H, J 5.4 Hz), 9.62 (t, 2H), 9.65 (s, 1H), 10.27 (s,
1H), 11.14 (s, 1H) ppm; 13C NMR (75 MHz, DMSO-d6): d 21.70
(CH3), 35.38 (CH2), 41.84 (CH2), 63.10 (CH2), 112.16 (CH), 115.19
(CH), 116.8 (CH), 126.92 (CH), 127.85 (CH), 128.46 (2 ꢀ CH),
129.08 (2 ꢀ CH), 129.35 (2 ꢀ CH), 129.54 (3 ꢀ CH), 132.60 (C),
133.88 (C), 139.11 (C), 139.81 (C), 140.44 (C), 142.29 (C), 144.47
(CH), 147.15 (CH), 148.67 (CH), 161.75 (C@O), 163.70 (C@O),
166.07 (C@O) ppm; HRMS-ESI m/z [M+H]+ calcd for C30H30N4O3:
494.23179, found: 494.26384.
4.2.6.18. 1-{[3-(4-Chloro-phenylcarbamoyl)-phenylcarbamoyl]-
methyl}-3-phenethylcarbamoyl-pyridinium chloride (181). Pre-
pared from compound 158 (0.3 g, 1 m mol) and 186 as white pow-
der (0.262 g, 56%) mp: 240–242 °C (Decomp.); mmax (KBr disc)
4.2.6.22. 3-Phenethylcarbamoyl-1-{[3-(toluene-4-sulfonylami-
no)-phenylcarbamoyl]-methyl}-pyridinium chloride (185). Pre-
pared from 163 (0.32 g, 0.9 mmol) and 186 to yield 185 (0.26 g,
50%) mp: 204 °C (Decomp.); mmax (KBr disc) 3421, 1697, 1666,
3421, 1689, 1662, 1593 cmꢁ1
;
1H NMR (300 MHz, DMSO-d6): d
1600, 1543 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.39 (s, 3H),
;
2.87 (t, 2H), 3.55 (t, 2H), 5.81 (s, 2H), 7.18–7.81 (m, 12H), 8.17 (s,
1H), 8.32 (t, 1H), 9.09 (d, 1H, J 7.8 Hz), 9.21 (d, 1H, J 6 Hz), 9.58
(s, 2H), 10.49 (s, 1H), 11.42 (s, 1H) ppm; 13C NMR (75 MHz,
DMSO-d6): d 35.38 (CH2), 41.84 (CH2), 63.07 (CH2), 119.49 (CH),
122.57 (2 ꢀ CH), 122.94 (CH), 123.64 (CH), 126.92 (CH), 127.19
(2 ꢀ CH), 127.96 (C), 129.08 (2 ꢀ CH), 129.19 (2 ꢀ CH), 129.35
(CH), 129.69 (CH), 133.93 (C), 136.25 (C), 138.79 (C), 139.22 (C),
139.79 (C), 144.49 (CH), 147.15 (CH), 148.69 (CH), 161.75 (C@O),
164.07 (C@O), 166.05 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C29H26ClN4O3: 513.16934, found: 513.16879.
2.88 (t, 2H), 3.55 (t, 2H), 5.75 (s, 2H), 6.61 (d, 1H, J 8.1 Hz), 7.19–
7.70 (m, 12H), 8.33 (t, 1H), 9.08 (d, , 2H, J 7.8 Hz), 9.17 (d, 1H, J
5.7 Hz), 9.53 (s, 1H), 9.56 (s, 1H), 11.36 (s, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6): d 21.83 (CH3), 35.40 (CH2), 41.84 (CH2), 63.29
(CH2), 121.36 (CH), 122.59 (CH), 126.92 (CH), 127.09 (CH), 127.88
(CH), 128.68 (2 ꢀ CH), 129.09 (2 ꢀ CH), 129.34 (2 ꢀ CH), 130.49
(CH), 130.64 (2 ꢀ CH), 134.00 (C), 134.77 (C), 136.35 (C), 139.75
(C), 139.82 (C), 144.38 (CH) 146.13 (C), 147.22 (CH), 148.71 (CH),
161.80 (C@O), 164.09 (C@O) ppm; HRMS-ESI m/z [M]+ calcd for
C29H29N4O4S 529.19095, found: 529.19040.
4.2.6.19. 1-{[3-(4-Methoxy-phenylcarbamoyl)-phenylcarbamo-
yl]-methyl}-3-phenethylcarbamoyl-pyridinium chloride (182).
Prepared from compound 159 (0.2 g, 0.6 mmol) and 186 to yield
182 as light yellow powder (0.18 g, 55%) mp: 222–225 °C (De-
4.2.7. Synthesis of N-phenethyl-nicotinamide (186)
Excess oxalyl chloride (99%) was added to nicotinic acid (15 g,
0.12 mol) to form slurry. The mixture was left to dry in the fume
hood. Subsequently, neat phenylethylamine (10 mL) was added
to the resulting powder in ice bath. The reaction was subsequently
warmed to room temperature and stirred for 15 min. Then it was
terminated by quenching with 5% aqueous NaHCO3 solution
(300 ml). The resulting crude precipitate was collected and recrys-
tallized from acetone/water (Scheme 8).
comp.);
m ;
max3282, 1662, 1593, 1512 (KBr disc) cmꢁ1 1H NMR
(300 MHz, DMSO-d6): d 2.88 (t, 2H), 3.52 (t, 2H), 3.69 (s, 3H),
5.75 (s, 2H), 6.87 (d, 2H, J 8.7 Hz), 7.16–7.78 (m, 10H), 8.11 (s,
1H), 8.30 (t, 1H), 9.02 (d, 1H, J 7.5 Hz), 9.16 (d, 1H, J 5.4 Hz), 9.42
(s, 1H), 9.53 (s, 1H), 10.13 (s, 1H), 11.19 (s, 1H) ppm; 13C NMR

3114
A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
Acknowledgments
O
O
a, b
HO
N
N
H
N
This project was partially sponsored by the Faculty of Graduate
Studies (Ph.D. Thesis of Afaf Al-Nadaf). The authors thank the
Deanship of Scientific Research and Hamdi-Mango Center for Sci-
entific Research at the University of Jordan for their generous
funds. The authors also thank Ms. Hiba Zalloum for her technical
assistance.
186
Scheme 8. Synthesis of 186. (a) Oxalyl chloride, (b) quench with 2-phenyleth-
ylamine.
Supplementary data
4.2.7.1. N-Phenethyl-nicotinamide (186). Yellow crystalline
needles (8.12 g, 29%) mp: 58–59 °C; mmax (KBr disc) 3325,
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.03.043.
1635 cmꢁ1 1H NMR (300 MHz, DMSO-d6): d 2.83 (t, 2H), 3.51
;
(t, 2H), 7.15–7.30 (m, 5H), 7.48 (t, 1H), 8.12 (d, 1H, J 7.8 Hz),
8.67 (d, 1H, J 3.6 Hz), 8.76 (s, 1H), 8.94 (s, 1H) ppm; 13C NMR
(75 MHz, DMSO-d6): d 35.67 (CH2), 41.55 (CH2), 124.14 (CH),
126.83 (CH), 129.04 (2 ꢀ CH), 129.36 (2 ꢀ CH), 130.70 (C),
135.54 (CH), 140.06 (C), 148.96 (CH), 152.45 (CH), 165.39
(C@O) ppm; HRMS-ESI m/z [M+H]+ calcd for C14H15N2O:
228.12179, found 228.12124.
References and notes
1. Maria, L. B.; Riccardo, M.; Anna, M.; Michela, R.; Carlo, M. Curr. Opin. Chem. Biol.
2009, 13, 303.
2. Shawn, J. S. Drug Dev. Res. 2009, 70, 101.
3. Jeffrey-Tri, N.; Abdellah, Y.; Yoshiaki, K. Curr. Pharm. Des. 2006, 12, 4295.
4. Li-Bang, Y.; Kristina, L.; Riqiang, Y.; Martin, C.; Weiming, X.; Xiao-Li, Y.;
Thomas, B.; Lucia, S.; Philipwong; Donald, P.; Rena, L.; Yong, S. Nat. Med. 2003,
9.
5. Vinicius, B. S.; Adriana, M. N.; Daniela, G. P.; Tainá, F. P.; Carlos, H. T. Curr.
Bioactive Compd. 2009, 5, 119.
4.3. Quantification of BACE-1 activity in a fluorometric assay
The BACE-1 fluorescence resonance energy transfer (FRET) as-
say was performed as described by the BACE kit manufacturer (Sig-
ma, product number CS0010)67,68 Principle of the assay: The
substrate is linked to a fluorescent dye on one end and to a quench-
ing group on its other end. The fluorescence of the substrate is sig-
nificantly reduced due to intra molecular resonance energy
transfer to the quenching group. Upon substrate cleavage by the
enzyme, there is a disturbance of the energy transfer resulting in
the enhancement of the fluorescent signal. The assay procedure
can be described briefly as follows: The BACE substrate is prepared
6. Cooper, J. B. Curr. Drug Targets 2002, 3, 155.
7. Varghese, J. Curr. Top. Med. Chem. 2006, 6, 569; Schmidt, B.; Baumann, S.;
Hannes, A. B.; Larbig, G. Curr. Top. Med. Chem. 2006, 6, 377.
8. Wen-Hai, H.; Rong, S.; Yong-Zhou, H. Curr. Med. Chem. 2009, 16, 1806.
9. Cumming, J. N.; Le, T. X.; Babu, S.; Carroll, X.; Chen, L.; Favreau, P.; Gaspari, T.;
Guo, D. W.; Hobbs, Y.; Huang, U.; Iserloh, M. E., et al Bioorg. Med. Chem. Lett.
2008, 18, 3236.
10. Yoshio, H.; Hiroko, O.; Naoko, M.; Ryoji, Y.; Abdellah, Y.; Koushi, H.; Tooru, K.;
Kazuki, S.; Yoshio, H.; Shoichi, I.; Yoshiaki, K. Bioorg. Med. Chem. Lett. 2008, 18,
1654.
11. Lorna, P.; Andrea, C.; Francesco, C.; Federica, B.; Manuela, B.; Francesca, M.;
Maurizio, R.; Vincenza, A.; Angela, R. Bioorg. Med. Chem. Lett. 2008, 18, 423.
12. Yoshiari, S.; Takeshi, K.; Akinori, A.; Tetsuhiro, N.; Hachiro, S. Biochim. Biophys.
Acta 2008, 1780, 819.
13. Heuisul, P.; Kyeongsik, M.; Hyo-Shin, K.; Ki, D. K.; Dongchul, L.; Sang-Won, S.;
Jae-Ung, C.; Bettina, P.; Deog-Young, C. Bioorg. Med. Chem. Lett. 2008, 18, 2900.
14. Derek, C. C.; Joseph, R. S.; Rajiv, C.; Rebecca, C.; John, W. E.; Kristi, Y. F.; Boyd, L.
H.; Yun, H.; Steve, J.; Lee, D. J.; Guixian, J.; Peter, A. L.; Michael, S. M.; Eric, S. M.;
William, J. M.; Mary-Margaret, O.; Andrea, M. O.; Albert, J. R.; Kristine, S.;
JunJun, W.; Eric, W.; Jonathan, B. Bioorg. Med. Chem. Lett. 2008, 18, 1063.
15. Huang, D.; Urs, L.; Peter, K.; Marco, C.; Alcide, B.; Amedeo, C. J. Am. Chem. Soc.
2006, 128, 5436.
16. Arun, K. G.; Geoffrey, B.; Cynthia, H.; Reiko, K.; Dongwoo, S.; Khaja, A.; Lin, H.;
Jeffrey, A. L.; Chan, N.; Gerald, K.; Jacques, E.; Jordan, T. J. Med. Chem. 2001, 44,
2865.
17. Godemann, R.; Madden, J.; Kramer, J.; Smith, M. A.; Fritz, U.; Hesterkamp, T.;
Barker, J.; Hoeppner, S.; Hallett, D.; Cesura, A.; Ebneth, A.; Kemp, J.
Biochemistry-Us 2009, 48, 10743.
in the buffer to a concentration of 50
in the same buffer to a concentration of approximately 0.3 units/
l
M. BACE enzyme is prepared
l.
l
Stock solutions of test samples are prepared in DMSO, and then
serially diluted in 50 mM sodium acetate, pH 4.5 to give the de-
sired working concentrations. Triton X-100 was added to each well
to a final concentration of 160
BACE enzyme, substrate, standard, test samples and buffer are
then added to the wells for a total volume of 100 l, with the
lM.
l
BACE-1 enzyme being added last, just prior to reading. Baseline
fluorescence is recorded immediately after the addition of the
BACE enzyme on a fluorometer set at excitation 320 nm, emission
405 nm. The reaction rate was monitored for 2 h at 37 °C using
FLX800TBI Microplate Fluorimeter (BioTek Instruments, Winooski,
USA) and the linear time-relative fluorescence units (RFU) sections
were taken for rate calculation.68
18. Charrier, N.; Clarke, B.; Cutler, L.; Demont, E.; Dingwall, C.; Dunsdon, R.;
Hawkins, J.; Howes, C.; Hubbard, J.; Hussain, I.; Maile, G.; Matico, R.; Mosley, J.;
Naylor, A.; O‘Brien, A.; Redshaw, S.; Rowland, P.; Soleil, V.; Smith, K. J.;
Sweitzer, S.; Theobald, P.; Vesey, D.; Walter, D. S.; Wayne, G. Bioorg. Med. Chem.
Lett. 2009, 19, 3664.
19. Charrier, N.; Clarke, B.; Demont, E.; Dingwall, C.; Dunsdon, R.; Hawkins, J.;
Hubbard, J.; Hussain, I.; Maile, G.; Matico, R.; Mosley, J.; Naylor, A.; O‘Brien, A.;
Redshaw, S.; Rowland, P.; Soleil, V.; Smith, K. J.; Sweitzer, S.; Theobald, P.;
Vesey, D.; Walter, D. S.; Wayne, G. Bioorg. Med. Chem. Lett. 2009, 19, 3669.
20. Charrier, N.; Clarke, B.; Cutler, L.; Demont, E.; Dingwall, C.; Dunsdon, R.;
Hawkins, J.; Howes, C.; Hubbard, J.; Hussain, I.; Maile, G.; Matico, R.; Mosley, J.;
Naylor, A.; O‘Brien, A.; Redshaw, S.; Rowland, P.; Soleil, V.; Smith, K. J.;
Sweitzer, S.; Theobald, P.; Vesey, D.; Walter, D. S.; Wayne, G. Bioorg. Med. Chem.
Lett. 2009, 19, 3674.
21. Clarke, B.; Demont, E.; Dingwall, C.; Dunsdon, R.; Faller, A.; Hawkins, J.;
Hussain, I.; Macpherson, D.; Maile, G.; Matico, R.; Milner, P.; Mosley, J.; Naylor,
A.; O‘Brien, A.; Redshaw, S.; Riddell, D.; Rowland, P.; Soleil, V.; Smith, K.;
Stanway, S.; Stemp, G.; Sweitzer, S.; Theobald, P.; Vesey, D.; Walter, D. S.;
Ward, J.; Wayne, G. Bioorg. Med. Chem. Lett. 2008, 18, 1017.
22. Beswick, P.; Charrier, N.; Clarke, B.; Demont, E.; Dingwall, C.; Dunsdon, R.;
Faller, A.; Gleave, R.; Hawkins, J.; Hussain, I.; Johnson, C. N.; Macpherson, D.;
Maile, G.; Matico, R.; Milner, P.; Mosley, J.; Naylor, A.; O‘Brien, A.; Redshaw, S.;
Riddell, D.; Rowland, P.; Skidmore, J.; Soleil, V.; Smith, K. J.; Stanway, S.; Stemp,
G.; Stuart, A.; Sweitzer, S.; Theobald, P.; Vesey, D.; Walter, D. S.; Ward, J.;
Wayne, G. Bioorg. Med. Chem. Lett. 2008, 18, 1022.
4.4. Quantification of renin activity in a fluorometric assay
SensoLyte™ 520 Renin Assay Kit was used for the assay using
a Mc-Ala/Dnp FRET peptide. The sequence of this peptide is de-
rived from the cleavage site of renin. In the FRET peptide the
fluorescence of Mc-Ala is quenched by Dnp. Upon cleavage into
two separate fragments by renin, the fluorescence of Mc-Ala is
recovered, and can be monitored at excitation/emission = 490/
520 nm.
Test compounds and renin solutions were added into the micro-
plate wells and incubated at 37 °C for 30 min. Subsequently, 50 lL
renin substrate solution were added into each well. The reagents
were subsequently mixed thoroughly by shaking the plate gently
for 30 s. The fluorescence intensity was immediately measured
continuously and recorded every 5 min for 15 min at 37 °C. Appro-
priate positive and negative controls were prepared.73
23. Clarke, B.; Demont, E.; Dingwall, C.; Dunsdon, R.; Faller, A.; Hawkins, J.;
Hussain, I.; Macpherson, D.; Maile, G.; Matico, R.; Milner, P.; Mosley, J.; Naylor,

A. Al-Nadaf et al. / Bioorg. Med. Chem. 18 (2010) 3088–3115
3115
A.; O‘Brien, A.; Redshaw, S.; Riddell, D.; Rowland, P.; Soleil, V.; Smith, K.;
Stanway, S.; Stemp, G.; Sweitzer, S.; Theobald, P.; Vesey, D.; Walter, D. S.;
Ward, J.; Wayne, G. Bioorg. Med. Chem. Lett. 2008, 18, 1011.
24. Beeley, N. R. A.; Sage, C. Targets 2003, 2, 19.
52. Ramsey, L. F.; Schafer, W. D. The Statistical Sleuth, 1st ed.; Wadsworth
Publishing Company: Belmont, CA, 1997.
53. Kier, L. B.; Hall, L. H. In Topological Indices and Related Descriptors in QSAR and
QSPR; Devillers, J., Balaban, A. T., Eds.; Gordon and Breach: London, 1999; pp
491–562.
25. Klebe, G. Drug Discovery Today 2006, 11, 580.
26. Steuber, H.; Zentgraf, M.; Gerlach, C.; Sotriffer, C. A.; Heine, A.; Klebe, G. J. Mol.
Biol. 2006, 363, 174.
27. Stubbs, M. T.; Reyda, S.; Dullweber, F.; Moller, M.; Klebe, G.; Dorsch, D.;
Mederski, W.; Wurziger, H. ChemBioChem 2002, 3, 246.
28. DePristo, M. A.; de Bakker, P. I. W.; Blundell, T. L. Structure 2004, 12, 831.
29. Georgia, B. M.; Dennis, C.; Samuel, L. G.; Ming-Tain, L.; Sanjeev, K. M.; Philippe,
G. N.; Beth, P.; Hemaka, A. R.; Harold, G. S.; Shaun, R. S.; Katharine, M. H. Bioorg.
Med. Chem. Lett. 2007, 17, 1117.
54. Andrew, L. H.; Colin, R. G.; Alexander, A. Drug Discovery Today 2004, 9, 430.
55. Kirchmair, J.; Markt, P.; Distinto, S.; Wolber, G.; Langer, T. J. Comput. Aided Mol.
Des. 2008, 22, 213.
56. Verdonk, M. L.; Marcel, L.; Berdini, V.; Hartshorn, M. J.; Mooij, W. T. M.; Murray,
C. W.; Taylor, R. D.; Watson, P. J. Chem. Inf. Comput. Sci. 2004, 44, 793.
57. Irwin, J. J.; Shoichet, B. K. J. Chem. Inf. Comput. Sci. 2005, 45, 177.
58. Triballeau, N.; Acher, F.; Brabet, I.; Pin, J.-P.; Bertrand, H.-O. J. Med. Chem. 2005,
48, 2534.
30. Taha, M. O.; Bustanji, Y.; Al-Ghussein, M.; Mohammad, M.; Zalloum, H.; Al-
Masri, I. M.; Atallah, N. J. Med. Chem. 2008, 51, 2062.
59. Gao, H.; Williams, C.; Labute, P.; Bajorath, J. J. Chem. Inf. Comput. Sci. 1999, 39,
164.
31. Taha, M. O.; Dahabiyeh, L. A.; Bustanji, Y.; Zalloum, H.; Saleh, S. J. Med. Chem.
2008, 51, 6478.
60. Jacobsson, M.; Liden, P.; Stjernschantz, E.; Bostroem, H.; Norinder, U. J. Med.
Chem. 2003, 46, 5781.
32. Taha, M. O.; Atallah, N.; Al-Bakri, A. G.; Paradis-Bleau, C.; Zalloum, H.; Younis,
K.; Levesque, R. C. Bioorg. Med. Chem. 2008, 16, 1218.
33. Taha, M. O.; Bustanji, Y.; Al-Bakri, A. G.; Yousef, M.; Zalloum, W. A.; Al-Masri, I.
M.; Atallah, N. J. Mol. Graphics Modell. 2007, 25, 870.
34. Abu Hammad, A. M.; Taha, M. O. J. Chem. Inf. Model. 2009, 49, 978.
35. CATALYST 4.11 Users’ Manual; Accelrys Software Inc.: San Diego, CA, 2005.
36. Sprague, P. W.; Hoffmann, R.; Van de Waterbeemd, H.; Testa, B.; Folkers, G.
Curr. Tools Med. Chem. 1997, 230.
37. Barnum, D.; Greene, J.; Smellie, A.; Sprague, P. J. Chem. Inf. Comput. Sci. 1996, 36,
563.
38. Smellie, A.; Teig, S.; Towbin, P. J. Comput. Chem. 1995, 16, 171.
39. Li, H.; Sutter, J.; Hoffmann, R. In Pharmacophore Perception, Development, and
Use in Drug Design; Güner, O. F., Ed.; International University Line: La Jolla, CA,
2000; pp 173–189.
40. Sutter, J.; Güner, O.; Hoffmann, R.; Li, H.; Waldman, M. In Pharmacophore
Perception, Development, and Use in Drug Design; Güner, O. F., Ed.; International
University Line: La Jolla, CA, 2000; pp 501–511.
41. Kurogi, Y.; Guner, O. F. Curr. Med. Chem. 2001, 8, 1035.
42. Bersuker, I. B.; Bahçeci, S.; Boggs, J. E. In Pharmacophore Perception,
Development, and Use in Drug Design; Güner, O. F., Ed.; International
University Line: La Jolla, CA, 2000; pp 457–473.
61. Darvesh, S.; McDonald, R. S.; Darvesh, K. V.; Mataija, D.; Mothana, S.; Cook, H.;
Carneiro, K. M.; Richard, N.; Walsh, R.; Martin, E. Bioorg. Med. Chem. 2006, 14,
4586.
62. Shunqi, Y.; Todd, A.; Esmir, G.; Jae Hoon, S.; Tania, T.; Hongwoo, K.; Frank, R.;
Huaming, C.; Robert, H.; Wu, J. Z.; Zhi, H.; Nanhua, Y. Bioorg. Med. Chem. Lett.
2007, 17, 28.
63. Chuansheng, N.; Jun, L.; Terrence, W. D.; Shu-Hui, C. Tetrahedron 1998, 54, 6311.
64. Brett, D. A.; Victor, K. P.; Laura, C. M.; Mark, R.; Magda, M.; Clodagh
Prendergast, ; Terry, B.; Guy, L.; Jamie, F.; Lina, L.; Xiaodong, W.; Hariharan,
V.; Marna, P.; Craig, W.; Miche‘le, C. R.; Michael, H.; Kenway, H.; Xiaohu, D.;
Christopher, K.; Nigel, P. S.; Michael, H. R. J. Med. Chem. 2006, 49, 6371.
65. Laixing, H.; Zhuo-rong, L.; Yan, L.; Jinrong, Q.; Yi-He, L.; Jian-dong, J.; David, W.
B. J. Med. Chem. 2006, 49, 6273.
66. Sante, M.; Maria, D.; Barbara, S.; Maria, B.; Edward, B. J. Med. Chem. 1988, 31,
1956.
67. BACE1 Activity Detection Kit (Fluorescent), CS0010.
68. Francesca, M.; Marina, N.; Vanni, C.; Vincenza, A. Anal. Bioanal. Chem. 2007, 388,
1175.
69. CERIUS
2 4.10 LigandFit User Manual; Accelrys Inc.: San Diego, CA, 2000.;
Venkatachalam, C. M.; Jiang, X.; Oldfield, T.; Waldman, M. J. Mol. Graphics
Modell. 2003, 21, 289.
43. Poptodorov, K.; Luu, T.; Langer, T.; Hoffmann, R.. In Methods and Principles in
Medicinal Chemistry. Pharmacophores and Pharmacophores Searches; Hoffmann,
R. D., Ed.; Wiley-VCH: Weinheim, Germany, 2006; 2, pp 17–47.
44. Singh, J.; Chuaqui, C. E.; Boriack-Sjodin, P. A.; Lee, W. C.; Pontz, T.; Corbley, M.
J.; Cheung, H. K.; Arduini, R. M.; Mead, J. N.; Newman, M. N.; Papadatos, J. L.;
Bowes, S.; Josiah, S.; Ling, L. E. Bioorg. Med. Chem. Lett. 2003, 13, 4355.
45. Taha, M. O.; Qandil, A. M.; Zaki, D. D.; AlDamen, M. A. Eur. J. Med. Chem. 2005,
40, 701.
46. Taha, M. O.; Bustanji, Y.; Al-Ghussein, M. A. S.; Mohammad, M.; Zalloum, H.; Al-
Masri, I. M.; Atallah, N. J. Med. Chem. 2008, 51, 2062.
47. Taha, M. O.; Al-Bakri, A. G.; Zalloum, W. A. Bioorg. Med. Chem. Lett. 2006, 16,
5902.
70. Perioli, L.; Ambrogi, V.; Bernardini, C.; Grandolini, G.; Ricci, M.; Giovagnoli, S.;
Rossi, C. Eur. J. Med. Chem. 2004, 39, 715.
71. Laszlo, P.; Katalin, P.; Alevtina, D. Z.; Vien, N.; Pal, P.; Stanley, M. S. J. Med. Chem.
2004, 47, 6025.
72. Huang, W.; Yu, H.; Sheng, R.; Li, J.; Hu, Y. Bioorg. Med. Chem. 2008, 16, 10190.
73. SensoLyte™ 390 Renin Assay Kit.
74. Cronin, M. T. D.; Schultz, T. W. J. Mol. Struct. (Theochem.) 2003, 622, 39.
75. Sinha, S.; Anderson, J. P.; Barbour, R.; Basi, G. S.; Caccavello, R.; Davis, D.; Doan,
M.; Dovey, H. F.; Frigon, N.; Hong, J.; Jacobson-Croak, K.; Jewett, N.; Keim, P.;
Knops, J.; Lieberburg, I.; Power, M.; Tan, H.; Tatsuno, G.; Tung, J.; Schenk, D.;
Seubert, P.; Suomensaari, S. M.; Wang, S.; Walker, D.; Zhao, J.; McConlogue, L.;
John, V. Nature 1999, 402, 537.
48. Garino, C.; Tomita, T.; Pietrancosta, N.; Laras, Y.; Rosas, R.; Herbette, G.;
Maigret, B.; Quéléver, G.; Iwatsubo, T.; Kraus, J. J. Med. Chem. 2006, 49, 4275.
49. Sheridan, R. P.; Kearsley, S. K. Drug Discovery Today 2002, 7, 903.
76. Taha, M. O.; Tarairah, M.; Zalloum, H.; Abu-Sheikha, G. J. Mol. Graphics Modell.
2010, 28, 383.
77. Rajapakse, H. A.; Nantermet, P. G.; Selnick, H. G.; Munshi, S.; Mcgaughey, G. B.;
Lindsley, S. R.; Young, M. B.; Laia, M. T.; Espeseth, S.; Shi, X. P.; Colussi, D.;
Pietrakm, B.; Crouthamel, C.; Tugusheva, K.; Huang, Q.; Xu, M.; Simon, A. J.;
Kuo, L.; Hazuda, D. J.; Graham, S.; Vacca, J. P. J. Med. Chem. 2006, 49, 7270.
50. Fischer, R. The Principle of Experimentation Illustrated by
a Psycho-Physical
ExpeHafner Publishing Co., 8th ed.; Hafner Publishing.
51. CERIUS2, QSAR Users’ Manual, version 4.10; Accelrys Inc.: San Diego, CA, 2005, 43–
88, 221–235.