Synthesis of some substituted pyrazinopyridoindoles and 3D QSAR studies along with related compounds: Piperazines, piperidines, pyrazinoisoquinolines, and diphenhydramine, and its semi-rigid analogs as antihistamines (H1)

Article abstract of DOI:10.1016/j.bmc.2006.09.018

3D QSAR studies on the title compounds led to the development of a model with three biophoric sites and six secondary sites viz. H-acceptor (ACC), H-donor (DON), heteroatom (presence), hydrophobic (hydrophobicity), steric (refractivity), and a ring (prese

Full text of DOI:10.1016/j.bmc.2006.09.018

Bioorganic & Medicinal Chemistry 14 (2006) 8249–8258
Synthesis of some substituted pyrazinopyridoindoles and 3D QSAR
studies along with related compounds: Piperazines, piperidines,
pyrazinoisoquinolines, and diphenhydramine, and its semi-rigid
analogs as antihistamines (H1)
Mridula Saxena, Stuti Gaur, Philip Prathipati and Anil K. Saxena*
Medicinal and Process Chemistry Division, Central Drug Research Institute, Lucknow 226001, India
Received 20 May 2006; revised 8 September 2006; accepted 9 September 2006
Available online 28 September 2006
Abstract—3D QSAR studies on the title compounds led to the development of a model with three biophoric sites and six secondary
sites viz. H-acceptor (ACC), H-donor (DON), heteroatom (presence), hydrophobic (hydrophobicity), steric (refractivity), and a ring
(presence) along with total hydrophobicity and total refractivity as global properties. The model predicted the test set of compounds
reasonably well. Three of the five newly synthesized 2-substituted octahydropyrazinopyridoindoles have shown potent antihistamin-
ic H1 activity with less toxicity and sedation potential.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
The QSAR studies in 2-b-aroylaminoethyl-1,2,3,4,6,
7,12,12a-octahydropyrazino(20,10:6,1)pyrido(3,4-b)indoles4
emphasized the importance of hydrophobicity of the
substituent at ortho and para-position and bulk at the
ortho position of the aromatic ring of the arylaminoethyl
side-chain, which contribute positively to increase in
activity. Further, in view of the similarity in terms of po-
sitive steric effect of substitution at the phenyl ring of
these molecules as of diphenhydramine, it was suggested
that these molecules bind to the H1 receptor in a folded
conformation whereby the phenyl and indole rings of
these molecules occupy similar positions as the two
phenyl rings of diphenhydramine.6
Histamine1 is an intercellular chemical messenger and
plays a critical role in several diverse physiological pro-
cesses. Four human G-protein coupled histamine recep-
tor subtypes (H1–4)2 are currently recognized to mediate
various actions of the monoamine histamine. These in-
clude smooth muscle contraction, inflammatory re-
sponse, gastric acid secretion, and mediation of
neurotransmitter release in central nervous system.
There has been a tremendous increase in knowledge
about the role of histamine through specific activation
or blockade of these receptor subtypes both in physiol-
ogy and pathology. Among the four subtypes, the hista-
mine H1 receptor has been an attractive target for drug
discovery for several years and H1 receptor antagonists3
have proved to be effective therapeutic agents for respi-
ratory distress, thus contributing to an important class
of drugs today.
Keeping in view that the same structural variations in a
common substructure present in different prototypes
should show similar change in activity (because of its
association with the complementary subsites at the
receptor) the above model was further explored. Thus
different types of compounds viz. 1-[(aroylamino)eth-
yl]-4-benzyl-piperazines and -piperidines, 1-{2-[(arylami-
no)-carbonyl]ethyl}-4-benzyl-piperazines and -piperidines,
2-[(arylamino)carbonyl]ethyl-1,2,3,4,6,7,12,12a-octahydro-
pyrazino[20,10:6,1]pyrido[3,4-b]indoles, and {2-[(aryl-
amino)carbonyl]ethyl}-1,2,3,4,6,11,11a-hexahydro-2H-
pyrazino[1,2-b]isoquinolines were studied to see the
effect of substitution on the phenyl ring of the aryl part
on antihistaminic (H1) activity.5 The similar slope value
In our earlier ligand-based QSAR and pharmacophore
study, models were developed by classical techniques
like Hansch’s physicochemical methods and HASL.4–7
Keywords: QSAR; H1-antihistamine; Apex-3D; Catalyst.
*
Corresponding author. Tel.:+91 522 2612411 18; fax: +91 522
2623405; e-mail: anilsak@gmail.com
0968-0896/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2006.09.018

8250
M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
R
O
a
NH
N
N
H
N
N
N
H
N
R=
5 F
H
8. NO2
1
6. Cl
9. OCH3
7. C2H5
R
O
H2N
O
b
Br
Cl
Br
N
H
R
2
4a-e
3a-e
4d.NO2
4e.OCH3
3d.NO2
3e.OCH3
3a.F
3b.Cl
4a.F
4b.Cl
3c.C2H5
4c.C2H5
Scheme 1. Reagents and conditions: (a) b-bromo-N-arylpropionamide, DMF, 60 ꢁC, K2CO3, NaI (corresponding para-substituted b-bromo-N-
arylpropionamide); (b) THF, TEA, 0 ꢁC.
(0.304( 0.060)) associated with p in QSARs in different
prototypes suggested that all the compounds appear to
act on a common receptor and the essential structural
requirements for the molecules to exhibit H1 antihista-
minic activity are the presence of substructures which
can interact at the proposed four subsites.6 These studies
further indicated that the application and predictive val-
ue of classical QSAR (physicochemical approach) are
not limited to same prototypes. The classical QSAR
can be used for the mapping of receptor sites if the com-
mon substructure competing for the same receptor sites
has been identified in different prototypes.
b]indoles (5–9) were synthesized by the condensation
of the intermediate 1,2,3,4,6,7,-8,12,12a-octahydropy-
razino-[20,10:6,1]pyrido[3,4-b]-indole17 with appropriate
b-bromo-N-arylpropionamides in presence of triethyl-
amine, essentially according to the method described
in the literature for similar compounds6 (Scheme 1).
The required b-bromo-N-arylpropionamides (4a–e)
were synthesized by the condensation of b-bromopropi-
onyl chloride with an appropriate arylamine in presence
of triethylamine (Scheme 1).
3. Pharmacology
In order to further validate these results through ad-
vanced molecular modeling techniques, the HASL ap-
proach has been used to identify the pharmacophore
for H1 receptor antagonists using the above compounds
and some semi-rigid analogs of diphenhydramine, ben-
zylhydrylamine, and phenbenzamine.7–9 These studies
were among the first applications of the HASL ap-
proach, which reinforced the importance of major sites
for the interaction of the tertiary nitrogen and aromatic
rings.
The antihistamine (H1) activity of compounds 5–9 was
measured on the isolated terminal part of the guinea
pig ileum (5.0 cm long) suspended in an organ bath con-
taining aerated Tyrode solution (20 ml) at 35 ꢁC and
spasm of the ileum was induced by 3 · 10ꢀ8 g/mL of his-
tamines. The percentage of inhibition was plotted
against different concentrations of the compound and
the concentration causing 50% inhibition (IC50) was cal-
culated (Table 1).
The above knowledge has been used for the validation
and improvement of our model using advanced soft-
wares like Catalyst10 and APEX-3D.11–16 The APEX-
3D expert system has a limitation in fast generation of
conformations for identifying the suitable conformation
for activity while the CATALYST software is relatively
poor in generating good predictive 3D QSAR models
than APEX-3D. Hence an integrated approach using
both expert systems has been applied to identify
the 3D pharmacophore for the above-mentioned
antihistamines H1 and some newly synthesized 2-[b-
(N-aryl)- proionamido]-1,2,3,4,6,7,8,12,12a-octahy-
dro-pyrazino[20,10:6,1]pyrido[3,4-b]indoles (5–9).
The compounds 5, 6, and 7 showed promising activity
among the newly synthesized compounds. Hence these
compounds were also evaluated for other pharmacolog-
ical effects including acute toxicity, gross observational
effects, antagonism to amphetamine hyperactivity, and
toxicity in aggregated mice. Electroshock seizures were
studied in male mice by the standard method.17–19 The
effect on blood pressure was also studied in anesthetized
cats by administering 2.5 lmol/kg iv.20
4. Molecular modeling
A 3D biophoric model was developed for H1 receptor
antagonists using the newly synthesized 2-[b-(N-ar-
yl)propionamido]-1,2,3,4,6,7,8,12,12a-octahydropyrazi-
no-[20,10:6,1]pyrido[3,4-b]indoles (5–9) (Table 1) and the
compounds 10–49 (Table 2) reported in our previous
papers.5–7 All the 45 molecules were aligned using
2. Chemistry
The 2-[b-(N-4-substitutedphenyl)propionamido]-1,2,3,4,
6,7,8,12,12a-octahydro-pyrazino[20,10:6,1] pyrido[3,4-

M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
Table 1. Antihistaminic (H1) activities for compounds 5–9
8251
Compound
Antihistaminic activity
ALD50 mg/kg Gross effect at 0.2 LD50
dose mg/kg ip (mice)a
ip (mice)
IC50 lmol/L
ꢀlogIC50 observed ꢀlogIC50 calculated ꢀlogIC50 predicted
5
6
7
8
9
0.35 0.06
0.2 0.03
0.19 0.02
2.6 0.4
0.46
0.7
0.69
0.62
0.72
0.59
1000
1000
1000
>1000
>1000
62
Stimulation
Stimulation
Depression
Depression
Depression
0.72
ꢀ0.41
0.57
0.38
ꢀ0.34
0.28
0.32
ꢀ0.33
0.25
2.27 0.07
Mepyramine 0.006 (1.6 lg/ml)
a Depressant implies reduction in spontaneous motoractivity, ataxia, and loss of writhing reflex; Stimulation implies increased active straup
phenomenon, preconvulsiveness, and convulsions.
CATALYST10 molecular modeling software and the
resulting alignment was used in generating the 3D bio-
phoric model by Apex 3D software.11
ture. The second group of two hypotheses is charac-
terized by two ring aromatic (RR) features. The ring
aromatic feature may represent the aromatic rings in-
volved in p–p interactions and the positive ionizable
group represents the charged nitrogen, which may be
involved in electrostatic interaction with an acidic res-
idue on the receptor. The latter is known to be an
important interaction between a ligand and the recep-
tors of G-protein coupled receptor (GPCR) family to
which histamine H1 receptor belongs.
The 3D structures of the training set of the 45 molecules
(Tables 1 and 2) with antihistaminic activities distribut-
ed over a range of 3 orders of magnitude were built and
optimized for their geometry using the CHARMm
forcefield21 in the usual way and the conformations were
generated using the maximum limit of 250 conforma-
tions within a 20 kcal cut-off by applying the poling
algorithm22 as implemented in the Catalyst molecular
modeling software version 4.5 running on SGI O2 work-
station. The common feature hypothesis23 was devel-
oped using the default parameters and taking
diphenhydramine (compound 45, one of the most active
molecules) as a template onto which the rest of the 43 H1
antagonists were superimposed using the Catalyst Hip-
Hop module. The common feature hypothesis was pro-
duced by generating alignments of common features,
which included aromatic rings and positive ionizable
groups.
The ranking of hypotheses is based on the portion of
training set members that fit the proposed pharmaco-
phore and the rarity of the pharmacophore. The high-
er the ranking, it is less likely that the molecules fit the
hypothesis by a chance correlation. The highest rank-
ing hypothesis of each group is shown in (Table 3)
and the active molecule is shown to map well to the
highest ranking hypothesis 1 (Fig. 1) representing a
pharmacophore which is an essential three-dimensional
arrangement of functional groups that a molecule
must possess for its recognition at the active site. This
pharmacophore model has two ring aromatic features
and a positive ionizable group as proposed in our ear-
lier studies and in the pharmacophore model proposed
by Ter Laak et al.24 However, it differs from the
earlier proposed model in that it has no feature to ac-
count for the anionic site and in terms of the inter-fea-
ture distances proposed in the model of Ter Laak
et al. (inter-feature distances in this model: R1-P is
5.247, R1-R2 is 4.971, and R2-P is 6.88, while in the
model proposed by Ter Laak et al.25 R1-P is 8.73,
R1-R2 is 4.79, and R2-P is 9.10). This inconsistency
in the inter-feature distance may be because of the dif-
ferent chemical classes of molecules analyzed in each
of these models. Further the works of Wieland
et al.26 on the active antagonistic site region of hista-
mine H1 receptor prove that, one of the aromatic rings
of the antagonists forms favorable aromatic p–p stak-
ing interactions with Phe 433 and Phe 436, the other
ring establishes aromatic p–p stacking with Trp 167,
in addition the nitrogen establishes a salt bridge inter-
action with ASP 116. While aromatic interactions are
accounted for the ring aromatic features of the
hypotheses, the ionic salt bridge interaction is account-
ed for the positive ionizable feature.
The best alignments generated from common feature
alignment (HipHop) were subjected to different com-
putational chemistry programs including MOPAC
6.0 version (MNDO Hamiltonian)23 for the calcula-
tion of different physicochemical and quantum
chemical parameters: atomic charge, p-population,
H-donor and acceptor index, HOMO, LUMO,
hydrophobicity, and molar refractivity based on
atomic contributions, which were then used by the
APEX-3D expert system running on a SGI Indy
workstation, for the automated identification of
pharmacophore and 3D QSAR model building. The
training set compounds with antihistaminic activities
were classified into the following three classes: (i)
very active (P2.40), (ii) active (<2.40 and
Pꢀ0.85), and (iii) less active (>ꢀ0.85).
5. Results and discussion
HipHop provides feature-based alignment of a collec-
tion of compounds without considering activity.
Among the eight generated hypotheses, six hypotheses
contain two-three features with the ranking scores
ranging from 80.3518 to 25.2954. These six hypotheses
consist of the same common-feature functions of two
ring aromatic and one positive ionizable (RRP) fea-
Among several 3D biophoric models developed by the
APEX-3D expert system using the alignments of com-
mon feature hypothesis 1 for all the molecules of the

8252
M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
Table 2. Molecules of the training set (activity IC50 lmol/L)
O
H
N
N
N
H
N
R1
R2
X
X
O
O
H
N
N
N
H
N
R3
R4
N
N
O
N
H
N
H
O
N
N
H
R5
N
Compound
X
R1
R2
R3
R4
R5
IC50 lmol/La
ꢀlogIC50
Obs.b
Calc.c
Pred.d
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
N
2-I
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
H
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
0.345
0.421
0.696
1.318
7.12
0.46
0.38
0.12
0.35
0.1
0.35
N
2-Br
2-Cl
N
—
—
0.16
0.12
0.12
0.12
0.13
N
2-NO2
2-CH3
H
ꢀ0.12
ꢀ0.85
ꢀ0.84
ꢀ0.19
0.18
N
—
—
0.12
0.18
N
6.94
ꢀ0.51
ꢀ0.26
ꢀ0.01
ꢀ0.01
ꢀ0.17
ꢀ0.17
0.28
ꢀ0.37
ꢀ0.27
ꢀ0.08
ꢀ0.05
ꢀ0.22
ꢀ0.31
0.29
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
H
—
—
1.548
0.661
0.799
1.071
0.477
0.561
0.832
0.963
1.513
0.778
0.836
1.318
1.701
2.167
1.604
0.634
0.979
0.776
1.738
0.537
0.206
0.282
0.364
0.380
0.930
1.342
0.525
0.677
1.316
2-Br
2-Cl
2-CH3
—
—
—
0.1
ꢀ0.03
0.32
0.25
2-C2H5
2-Cl
2-F
—
—
—
—
—
—
0.08
0.02
ꢀ0.41
0.12
0.6
ꢀ0.54
0.13
0.68
—
—
2-OCH3
H
—
—
ꢀ0.18
0.11
0.08
—
—
2-C2H5
2-Cl
2-F
ꢀ0.18
0.15
0.15
ꢀ0.22
0.16
0.2
N
—
—
N
—
—
ꢀ0.12
ꢀ0.23
ꢀ0.34
ꢀ0.2
0.2
N
2-NO2
H
—
—
—
ꢀ0.18
ꢀ0.34
ꢀ0.38
0.55
ꢀ0.18
ꢀ0.35
ꢀ0.53
0.62
N
—
—
—
—
—
—
—
—
—
—
—
2-C2H5
2-Cl
2-F
—
—
0.01
0.11
ꢀ0.1
0.24
0.52
ꢀ0.11
0.27
0.6
—
—
—
—
—
—
2-NO2
—
H
ꢀ0.24
0.27
0.69
—
—
—
ꢀ0.13
0.52
0.6
ꢀ0.2
0.51
0.61
2-Cl
2-CH3
2-OCH3
2-NO2
2-NH2
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
0.55
0.44
0.52
0.21
0.51
0.15
0.42
0.03
0.21
0.12
0.15
0.13
H
ꢀ0.13
0.28
0.17
—
—
2-C2H5
2-Cl
2-NO2
0.12
0.12
0.13
0.12
—
ꢀ0.12
0.01
0.03
O
N
45
0.0039
2.41
2.04
1.78

M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
8253
Table 2 (continued)
Compound
X
R1
R2
R3
R4
R5
IC50 lmol/La
ꢀlogIC50
Obs.b
2.41
Calc.c
Pred.d
46
47
48
49
0.0039
—
—
O
N
3cis
O
0.0039
0.039
0.513
2.41
2.26
2.09
N
2trans
O
N
1.410
—
—
O
0.290
—
—
N
4trans
aInhibitory concentration for 50% block of histamine.
b Observed.
c Calculated.
d Predicted.
Table 3. Hypothesis obtained through Catalyst
S. No. Hypothesisa Rank
DHb
PHc
1
2
RRP
RR
80.3518 111111111111 000000000000
25.2954 111111111111 000000000000
a R, ring aromatic; P, positive ionizable.
b DH, direct hit, all the features of the hypothesis are mapped. Direct
hit = 1, means yes and direct hit = 0, means no.
c PH, partial hit, partial mapping of the hypothesis.
training set, none of the biophores could map to all the
molecules of the training set. So a model which included
maximum number of compounds (42 out of 45) was
selected based on the following criteria; maximum num-
ber of compounds, correlation coefficient r2 > 0.7, the
difference between RMSA and RMSP <0.03 (a measure
of cross-validation), chance 60.1, no. of variables <7,
and compounds >41 (Table 4). The exclusion of the
three compounds (46, 48, and 49) is due to their ability
to map to the three biophoric sites of the model. This
may be due to the misfitting of the spatial geometry of
essential atom types (biophoric sites) which is associated
Figure 1. Mapping of compound 45 to hypothesis 1.
with the introduced semi-rigidity in the diphenhydra-
mine molecule (45). The model is highly significant
and does not suffer from overfitting as evidenced by a

8254
M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
Table 4. 3D QSAR model describing correlation and statistical reliability for H1 antihistaminic activity
Model no.
1
RMSA
0.348
RMSP
0.376
R2
Chance
0.1
Size
3
Match
0.22
Variable
6
No. of compounds
42
0.73
small difference (0.028) between root mean squared
approximation (RMSA) and leave-one-out prediction
(RMSP), as a big difference in RMSA and RMSP is
caused by influential compounds that are overfitted by
the multiple regression and are poorly predicted when
not included in the training set.
C
D
B
A
N
O
C
This model comprised three biophoric features (Figs. 2–
4), corresponding to the earlier suggested sites ABD.6
The biophoric sites A and D in terms of 6p electrons pro-
vided by either of the aromatic rings of the six prototypes
(1-[(aroylamino)ethyl]-4-benzyl-piperazines and -piperi-
dines, 1-{2-[(arylamino)-carbonyl]ethyl}-4-benzyl-piper-
azines and -piperidines, 2-[(arylamino)carbonyl]ethyl-1,
ss6
ss3
ss
5
ss1
ss4
ss2
Figure 2. Representation of biophoric (s) and secondary sites (/).
Table 5. Secondary site parameters: hydrophobicity (ss 1–4) and
refractivity (ss 5–6) for the training set
Compound
ss1
ss2
ss3
ss4
—
ss5
ss6
5
6
—
—
0
ꢀ0.1
ꢀ0.1
—
—
—
2.5
3.5
—
—
—
—
—
—
2.9
2.9
—
—
2.9
2.9
—
—
—
—
2.9
2.9
2.9
3.5
—
2
3.45
—
—
—
—
—
—
—
—
—
—
0
0.15
7
ꢀ0.85
ꢀ0.85
—
—
3.9
—
8
—
—
0
9
—
—
—
—
—
—
—
—
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
47
—
—
ꢀ0.85
—
—
0.15
—
—
—
—
—
—
—
—
—
0.15
0
3.45
3.45
3.45
3.45
3.45
—
—
—
—
—
—
0
—
0
ꢀ0.85
ꢀ0.85
—
0
—
—
0
0.15
0.15
0.15
Figure 3. Compound 18 showing the distances between the biophoric
sites (white circles). The distances are given in angstrom units.
—
—
ꢀ0.1
—
—
—
0
ꢀ0.1
—
—
—
—
ꢀ0.85
—
—
—
—
—
0
—
—
ꢀ0.85
ꢀ0.85
ꢀ0.85
ꢀ0.85
ꢀ0.85
—
—
—
—
—
—
0
—
—
—
—
0.15
—
—
—
0
—
—
0
ꢀ0.1
—
ꢀ0.3
—
ꢀ0.85
—
—
0.15
0.9
3.45
—
ꢀ0.1
—
—
—
—
—
—
—
—
2.9
—
—
—
—
—
—
—
—
—
—
—
—
2.5
—
—
—
0
—
—
—
3.45
—
—
0
0.15
—
—
—
—
—
—
—
—
0
—
—
—
—
—
—
—
—
—
—
—
0
0.15
0.15
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
ꢀ0.1
—
3.45
—
Figure 4. Superimposition of the 42 molecules for pictorial represen-
tation of biophoric sites (solid spheres) and secondary sites (red
circles).
ꢀ1.1
—
—
—
—

M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
8255
2,3,4,6,7,12,12a-octahydropyrazino[20,10:6,1]pyrido[3,4-
b]indoles and {2-[(arylamino)carbonyl]ethyl}-1,2,3,
4,6,11,11a-hexahydro-2H-pyrazino[1,2-b]isoquinolines
as well as the diphenhydramine and its semi-rigid analogs)
correspond to p–p interactions and site B corresponds to
the protonable tertiary nitrogen provided by tertiary ami-
no function present in all the molecules in terms of p-pop-
ulation (0.164 0.009) for electrostatic interactions with
the receptor with a particular spatial disposition; the
mean inter-atomic distances of three biophoric sites A,
B, D being A–D (5.585 0.398), A–B (6.2181 0.421),
and B–D (5.5013 0.488).
derived using this biophore as a template for super-
imposition and the antihistaminic activity (ꢀlogIC50)
as a dependent variable and biophoric center proper-
ties (p-population, charge, HOMO, LUMO, ACC_01,
Don_01, hydrophobicity, and refractivity), global
properties (total hydrophobicity and total refractivi-
ty), secondary sites [H-acceptor (ACC), H-donor
(DON), heteroatom (presence), hydrophobic (hydro-
phobicity), steric (refractivity), and ring (presence)]
as independent variables with the occupancy set at
12, site radius at 0.60, sensitivity at 1.0, and random-
ization value at 100.
In order to understand the interactions and to iden-
tify the secondary sites for explaining the variation
in anti-histaminic activity, the 3D-QSAR model was
The derived 3D QSAR model Eq. 1 correlated the anti-
histaminic activity with secondary site parameters hydro-
phobicity, at secondary site ss1 [(6.269 0.411),
Table 6. Observed versus predicted activity of test compounds
Compound
Test set compounds
Antihistaminic activity H1
Predicted activitya ꢀlogIC50
b
Observed activity K1,K0.5 (nM)
N
50
1.640
1.070
N
51
0.120
0.120
0.450
0.46
0.16
0.53
N
N
52
O
O
N
N
53
a Predicted using Eq. 1.
b As reported in the literature,27 K0.5 is a competition binding assay endpoint and is calculated from IC50 value using the Cheng–Prussof equation,
K0.5 = IC50/(1 + L*/KD).

8256
M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
Figure 5. (a) Compound 51 mapping to the biophore produced by APEX-3D. (b) Compound 52 mapping to the biophore produced by APEX-3D.
(1.511 0.166), (6.085 0.37)ꢁA from the biophoric sites
ABD, respectively], secondary site ss2 [(1.561 0.581),
(5.464 0.373), (4.014 0.814)ꢁA from the biophoric
sites ABD, respectively], secondary site ss3
[(2.08 1.611), (5.119 0.479), (4.993 1.606)ꢁA from
the biophoric sites ABD, respectively], secondary site
ss4 [(4.797 1.067), (4.717 0.274), (1.526 0.875)ꢁA
from the biophoric sites ABD, respectively], and steric ef-
fect in terms of refractivity, at secondary site ss5
[(6.51 0.915), (4.717 0.274), (4.234 0.927)ꢁA from
the biophoric sites ABD, respectively] and secondary site
ss6 [(1.405 0.006), (2.663 0.231), (6.885 0.370)ꢁA
from the biophoric sites ABD, respectively].
cance >99% (F6,36/0.001 = 5.39; F6,36 = 16.38) and also
shows good leave-one-out cross-validation with Q-value
of 0.794 as well low RMSA and RMSP as discussed ear-
lier. Further this is also evident by the good correspon-
dence between the observed, calculated, and (LOO)
predicted values (Table 1).
For the validation of the above 3D QSAR model an
external test set of four compounds devoid of asymmet-
ric center viz. ( )-trans-1-Phenyl-3-(dimethylamino)-
1,2,3,4-tetrahydronaphthalene (50), triprolidine (51),
doxepin (52), and mepyramine (53) was used from the
reported data.27 The model predicted the activity of
these four compounds as ꢀlogIC50 values correspond-
ing to the training sets’ IC50 values (Table 6, four com-
pounds were in terms of K0.5 (nM)), the values as such
may not be compared, nevertheless a very good correla-
tion (R2 = 0.8904) was observed between the observed
and predicted values and also map well to the biophoric
sites of the model (Fig. 5a and b).
logðIC50Þ ¼ ꢀ0:564ðꢁ0:140Þ
ꢂ ½Hydrophobicity at ss1ꢃ
þ 10:097ðꢁ1:287Þ
ꢂ ðHydrophobicity at ss2Þ
ꢀ 1:638ðꢁ0:752Þ
ꢂ ½Hydrophobicity at ss3ꢃ
þ 2:209ðꢁ0:738Þ
Among the five new compounds (5–9) synthesized, the
compounds 5, 6, and 7 showed relatively high activity
and low toxicity among the 2-[b-(N-4-substituted-phen-
yl)propionamido]-1,2,3,4,6,7,8,12,12a-octahydro-pyraz-
ino[20, 10:6,1]pyrido[3,4-b]indole. Though they were less
active than mepyramine, the compounds 5, and 6 were
stimulant in gross behavior and also did not show signif-
icant hypotensive action in cats (>20% B.P fall for
>5 min), thus indicating that they may have no sedative
potential.
ꢂ ½Hydrophobicity at ss4ꢃ
ꢀ 0:270ðꢁ0:044Þ½Refractivity at ss5ꢃ
þ 0:117ðꢁ0:034Þ½Refractivity at ss6ꢃ
þ 0:12
ð1Þ
n = 42, R = 0.855, F6,36 = 16.376, Q = 0.794, S = 0.335
The secondary sites ss1, ss3, and ss5 in terms of hydro-
phobicity and refractivity contribute negatively while
secondary sites ss2 corresponding to the hydrophobicity
at the ortho position of the side-chain phenyl group, and
ss4 and ss6 in terms of hydrophobicity and refractivity,
respectively, contribute positively to the activity. The
values of hydrophobicity and refractivity parameters
at these sites are shown in Table 5. Equation 1 well de-
scribes the observed antihistaminic activity with good
correlation coefficient value (R = 0.86), low standard
deviation (S = 0.335), and is of high statistical signifi-
6. Conclusion
The above studies substantiate the findings of our earlier
QSAR studies where the major limitation was non-in-
clusion of diverse types of molecules into one model.
The 3D QSAR model described here using an integrated
approach of taking the pharmacophore mapping
through the application of CATALYST software and
3D QSAR model development from Apex modeling
not only well describes the variance in activity of the
training set but also maps and explains the estimated

M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
8257
activity of the test set molecules reasonably well. Most
of the published 3D QSAR models on antihistamines
are CoMFA based, which are relatively less versatile
in terms of predicting the activity of different types of
molecules. The present model may be useful in designing
and optimizing non-classical antihistamines H1, which
may also be non-sedative.
FAB-MS = 409 M+1; Anal. Calcd for C23H24ON4Cl:
N, 13.71. Found: N, 13.66. C23H25ClN4O.
7.4. 2-[b-(N-4-Ethylphenyl)propionamido]-
1,2,3,4,6,7,8,12,12a-octahydro-pyrazino[20,10:6,1]pyri-
do[3,4-b]indole (7)
Mp142 ꢁC; yield: 85%; IR (KBr) cmꢀ1: 3250, 2930, 2823,
1660, 742; 1H NMR (CDCl3) d: 1.15–1.25 (m, 3H); 2.29–
2.81 (m, 13H); 3.09–3.23 (m, 2H); 3.53–3.61 (m, 1H);
3.98–4.05 (m, 1H); 7.06–7.46 (m, 8H); 7.78 (s, 1H);
10.75 (s, 1H); FAB-MS = 403 M+1; Anal. Calcd for
C25H29ON4: N, 13.93. Found: N, 13.82. C25H30N4O.
7. Experimental
Microanalysis was performed on a Carlo Erba Analyzer
and compounds were analyzed for nitrogen. Melting
points were determined on an electrically heated mp
apparatus using a silicon oil bath. The compounds were
routinely checked for purity by TLC on silica gel plates
and their structures were verified by their IR spectra
measured on Perkin-FTIR model PC spectrophotome-
ter, FAB mass spectra were recorded on JEOL SX
102/DA-6000 mass using Argon/Xenon (6 KV, 10
7.5. 2-[b-(N-4-Nitrophenyl)propionamido]-
1,2,3,4,6,7,8,12,12a-octahydropyrazino[20,10:6,1]-pyri-
do[3,4-b]indole (8)
Mp 148–149 ꢁC; yield: 80%; IR (KBr) cmꢀ1: 3397, 2930,
2829, 1665, 747; 1H NMR (CDCl3) d: 2.29–3.23 (m,
13H); 3.56–3.62 (m, 1H); 4.00–4.07 (d, 1H, J = 14);
7.10–7.51 (m, 8H); 7.76 (s, 1H); 10.99 (s, 1H); FAB-
MS = 420 M+1; Anal. Calcd for C23H24O3N5: N,
16.71. Found: N, 16.78. C23H25N5O3.
1
MA) as the FAB gas and H NMR spectra recorded
on a Bruker spectrometer (200 MHz) with a multinucle-
ar inverse probehead with gradient at room temperature
(298 K) using CDCl3 or DMSO-d6 or CD3OD as solvent
and tetramethylsilane (TMS) as internal standard.
7.6. 2-[b-(N-4-Methoxyphenyl)propionamido]-
1,2,3,4,6,7,8,12,12a-octahydro-pyrazino[20,10:6,1]pyri-
do[3,4-b]-indole (9)
7.1. General procedure for the synthesis of 2-[b-(N-4-
substituted-phenyl)propionamido]-1,2,3,4,6,7,8,12,12a-
octahydropyrazino[20,10:6,1]pyrido[3,4-b]indoles
Mp 175 ꢁC; yield: 28.30%; IR (KBr) cmꢀ1: 3412, 3334,
1
The appropriate b-bromo-N-(4-substituted phenyl)pro-
pionamide (0.0051 mol) in dry DMF (2 ml) was added
to a suspension of dl-1,2,3,4,6,7,8,12,12a-octahydropy-
razino[20,10:6,1]pyrido[3,4-b]indole (5) (0.005 mol) and
Na2CO3 (0.0025 mol) in dry DMF (5 ml), and the reac-
tion mixture was stirred for 36 h at 60 ꢁC. It was cooled
to 30 ꢁC, diluted with water (20 ml), and extracted with
chloroform (3· 10 ml). The combined chloroform ex-
tracts were washed with water (2· 5 ml), dried over
Na2SO4, and concentrated under reduced pressure to
give the corresponding title compounds which were
purified by column chromatography over silica gel using
methanol (1–2%) in chloroform as elutant.
2824, 2362, 1660, 742; H NMR (CDCl3) d: 2.29–2.85
(m, 11H); 3.10–3.27 (m, 2H); 3.58–3.62 (m, 1H); 3.77
(s, 3H); 4.01–4.11 (m, 1H); 7.16–7.47 (m, 8H); 7.71 (s,
1H); 10.70 (s, 1H); FAB-MS = 405 M+1; Anal. Calcd
for C24H27O2N4: N, 13.86. Found: N, 13.72.
C24H28N4O2.
Acknowledgments
The authors are thankful to the pharmacology division
for providing pharmacological results, Mr. A. S. Kush-
waha for his technical assistance and one of us (MS) to
the DST, New Delhi, for providing financial support un-
der WOS. This is CDRI communication number 6822.
7.2. 2-[b-(N-4-Fluorophenyl)propionamido]-
1,2,3,4,6,7,8,12,12a-octahydro-pyrazino[20,10:6,1]pyri-
do[3,4-b]indole (5)
References and notes
Mp 102 ꢁC; yield: 93.45%; IR (KBr) cmꢀ1: 3298, 3065,
1
2929, 2825, 1663, 744; H NMR (CDCl3) d: 2.29–3.23
(m,13H); 3.54–3.61 (m, 1H); 3.99–4.06 (d, 1H, J = 14);
1. Barger, G.; Dale, H. H. J. Physiol. 1910, 41, 19.
2. Hofstra, C. L.; Desai, P. J.; Thurmond, R. L.; Fung-
Leung, W. P. J. Pharmacol. Exp. Ther. 2003, 305,
1212.
3. Zhang, M. Q.; Leurs, R.; Timmerman, H., 15th ed.. In
Berger Medicinal Chemistry and Drug Discovery; Wolf,
M. E., Ed.; John Wiley: New York, 1997; vol. 5, p
495.
4. Saxena, A. K.; Dhaon, M. K.; Ram, S.; Saxena, M.; Jain,
P. C.; Patnaik, G. K.; Anand, N. Indian J. Chem. 1983,
22B, 1224.
5. Patnaik, G. K.; Saxena, A. K.; Saxena, M.; Srimal, R. C.
Indian J. Exp. Biol. 1992, 30, 144.
6.94–7.53 (m, 8H); 7.78 (s, 1H); 10.83 (s, 1H); FAB-
; Anal. Calcd for C23H24ON4F: N,
MS = 393M+1
14.29. Found: N, 14.28. C23H25FN4O.
7.3. 2-[b-(N-4-Chlorophenyl)propionamido]-
1,2,3,4,6,7,8,12,12a-octahydro-pyrazino[20,10:6,1]pyri-
do[3,4-b]-indole (6)
Mp100 ꢁC; yield: 98.13%; IR (KBr) cmꢀ1: 3278, 2931,
1
2825, 1662, 743; H NMR (CDCl3) d : 2.29–2.90 (m,
11H); 3.05–3.23 (m, 2H); 3.55–3.62 (m, 1H): 3.99–4.07
(m, 1H); 7.09–7.51 (m, 8H); 7.78 (s, 1H); 10.99 (s, 1H):
6. Saxena, M.; Agarwal, S. K.; Patnaik, G. K.; Saxena, A. K.
J. Med. Chem. 1990, 33, 2970.

8258
M. Saxena et al. / Bioorg. Med. Chem. 14 (2006) 8249–8258
7. Saxena, A. K.; Saxena, M.; Chi, H.; Wiese, M. Med.
18. Jain, P. C.; Kapoor, V.; Anand, N.; Ahmad, A. R.;
Patnaik, G. K. J. Med. Chem. 1967, 10, 812.
19. Srimal, R. C.; Dhwan, B. N. Indian J. Pharmacol. 1972,
34, 172.
20. Singh, G. B.; Srimal, R. C.; Nityanand, S.; Dhawan, B. N.
Arzneim.-Forsch 1978, 28, 1087.
21. Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States, D.
J.; Swaminathan, S.; Karplus, A. M. J. Comp. Chem. 1983,
4, 187.
22. Smellie, A.; Teig, S. L.; Towbin, P. I. J. Comp. Chem.
1995, 16, 171.
Chem. Res. 1993, 3, 201; Gaur, S.; Prathipati, P.; Saxena,
M.; Saxena, A. K. Med. Chem. Res. 2004, 13, 724.
8. Ram, S.; Saxena, A. K.; Jain (Late), P. C.; Patnaik, G. K.
Indian J. Chem. 1984, B, 1261.
9. Patnaik, G. K.; Saxena, A. K.; Dhawan, B. N. In Aspects
of Allergy and Applied Immunol; Agarwal, M. K., Ed.,
1984; Vol. 17, 75.
10. Hypotheses in Catalyst, Biosym-MSI, San Diego, CA,
1992.
11. Apex-3D, BIOSYM Technologies, Inc. San Diego, CA,
USA.
12. Arokia Babu, M.; Shakya, N.; Prathipati, P.; Kaskhedikar,
S. G.; Saxena, A. K. Bioorg. Med. Chem. 2002, 10, 4035.
13. Pandya, T.; Pandey, S. K.; Tiwari, M.; Chaturvedi, S. C.;
Saxena, A. K. Bioorg. Med. Chem. 2001, 2, 291.
14. Pandey, S. K.; Naware, N. B.; Trivedi, P.; Saxena, A. K.
SAR QSAR Environ. Res. 2001, 6, 547.
23. Stewart, J. J. P. J. Comput.-Aided Mol. Des. 1990, 4,
101.
24. Terlaak, A. M.; Benhorst, J.; Donne-Op den Kelder, G.
M.; Timmerman, H. J. Med. Chem. 1995, 38, 3351.
25. Terlaak, A. M.; Timmerman, H.; Leurs, R.; Nederkoorn,
P. H.; Smit, M. J.; Donne-Op den Kelder, G. M. J.
Comput.-Aided Mol. Des. 1995, 9, 319.
15. Kashaw, S. K.; Rathi, L.; Mishra, P.; Saxena, A. K.
Bioorg. Med. Chem. Lett. 2003, 15, 2481.
16. Rathi, L.; Kashaw, S. K.; Dixit, A.; Pandey, G.; Saxena,
A. K. Bioorg. Med. Chem. 2004, 1, 63.
17. Saxena, A. K.; Jain, P. C.; Dua, P. R. J. Med. Chem. 1973,
16, 560.
26. Wieland, K.; Terlaak, A. M.; Smit, M. J.; Kuhne, R.;
Timmerman, H.; Leurs, R. J. Biol. Chem. 1999, 42,
22994.
27. Ehren, C.; Bucholtz, E. C.; Brown, R. L.; Tropsha, A.;
Booth, R. G.; Wyrick, S. D. J. Med. Chem. 1999, 42,
3041.