Pharmacophore-based design, synthesis, biological evaluation, and 3D-QSAR studies of aryl-piperazines as α1-adrenoceptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2005.05.003

Phenyl-piperazines were designed and synthesized based on pharmacophore for uro-selective α₁-adrenoceptor antagonists and 3D chemical database searching. Within this series, three compounds, 2, 3, and 13, showed similar or better α₁-

Full text of DOI:10.1016/j.bmcl.2005.05.003

Bioorganic & Medicinal Chemistry Letters 15 (2005) 3216–3219
Pharmacophore-based design, synthesis, biological evaluation,
and 3D-QSAR studies of aryl-piperazines as
a1-adrenoceptor antagonists
Min-Yong Li,a Hao Fangb and Lin Xiaa,*
aDepartment of Medicinal Chemistry, China Pharmaceutical University, Nanjing 210009, China
bSchool of Pharmacy, Shandong University, Jinan 250012, China
Received 15 March 2005; revised 27 April 2005; accepted 2 May 2005
Abstract—Phenyl-piperazines were designed and synthesized based on pharmacophore for uro-selective a1-adrenoceptor antago-
nists and 3D chemical database searching. Within this series, three compounds, 2, 3, and 13, showed similar or better a1-AR antag-
onistic activity compared with prazosin. The 3D-QSAR study of these compounds may provide useful information for the
development of novel aryl-piperazines as uro-selective a1-adrenoceptor antagonists, which can be used for the treatment of BPH
with fewer side effects.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Benign prostatic hyperplasia (BPH) is a common disease
in aging male population, and a substantial percentage
of men with BPH develop bladder outlet obstruction
(BOO) as part of age-related urological disorder.1 Re-
cent studies provide that BPH is primarily mediated
by a1-adrenoceptor (a1-AR) in prostatic tissue, and as
a result, some a1-AR antagonists, such as terazosin,
doxazosin, and alfuzosin, are clinically used for their
treatment.2 However, these agents have side effects, such
as hypotension, dizziness, muscle fatigue, and so on,
which are presumably due to an inability of these antag-
onists to adequately discriminate between the a1-AR in
the vascular and lower urinary tracts.3 These observa-
tions suggest that a powerful uro-selective a1-AR antag-
onist could alleviate the symptoms associated with BPH
within minimal cardiovascular side effects.
distances of A–P, A–HBD, and P–HBD are 5.296–
˚
5.477, 5.429–6.823, and 3.000 A, respectively.
In an on-going study, the previously generated pharma-
cophore model was used to search our in-house chemical
database using the 3DFS flexible searching program.5
One hit (1) with an aryl-piperazine skeleton was found.
Compound 1 has an excellent fit of the pharmacophore
model. The mapping picture with distance is shown in
Figure 1.
In our first trial, compound 1 was evaluated for its a1-
AR antagonistic activity to phenylphrine-reduced con-
tractions of isolated Sprague–Dawley rat anococcygeal
muscles. The results showed that 1 had moderate a1-
AR antagonistic activity (pA2 = 7.07). Such preliminary
results suggest that compound 1 can be used as a lead
compound, and aryl-piperazine can be used as scaffold
of a1-AR antagonists for further development.
Recently, we reported a three-point pharmacophore for
some uro-selective a1-AR antagonists, such as tamsulo-
sin, silodosin, indoramin, GG-818, and RS-100975,
using the Apex-3D program.4 This pharmacophore
model contained an aromatic ring (A), a positive ioniz-
able center (P), and a hydrogen bond donor (HBD). The
To improve the a1-AR antagonistic activity for this aryl-
piperazine lead compound, we decided to modify the
structure based on pharmacophore features. Com-
pounds 2–22 were designed by replacing the pyridine
ring with other heterocycles and changing the substitu-
ent on the phenyl ring of the aryl-piperazine structure.
These compounds (2–22) have been synthesized via the
route outlined in Scheme 1. The alkylation of
Keywords: Pharmacophore-based design; a1-Adrenoceptor; Antago-
nists; 3D-QSAR.
*
Corresponding author. Tel.: +86 25 8327 1440; fax: +86 25 8330
1606; e-mail: phenopro@cpu.edu.cn
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.05.003

M.-Y. Li et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3216–3219
3217
Figure 1. Pharmacophore model mapped to 1.
Scheme 1. Reagents and conditions: (i) Et3N, CH3CN, 100 ꢁC; (ii) N2H4 Æ H2O, EtOH, 80 ꢁC; (iii) ArCOCl, K2CO3, CH3CN, 100 ꢁC; (iv) ethanolic
HCl.
Table 1. Chemical and biological data for phenyl-piperazines
phenylpiperazine 23a–f with 2-(2-bromoethyl)phthali-
Compound Ar
R
pA2
mide 24 in the presence of Et3N in acetonitrile yielded
25a–f. Compounds 25a–f were then converted into
26a–f by deprotecting the phthalimide groups with
hydrazine hydrate. Finally, 26a–f were acylated with
aroyl chloride and then converted to water-soluble
hydrochloride salts to give target compounds 2–22.6
1
4-Pyridine
2,4-Dimethyl 7.07
5-Chloro-2-methoxy 8.56
2
2-Furan
2-Furan
2-Furan
3
3-Methoxy
4-Methoxy
8.56
7.85
7.71
4
5
2-Oxo-2H-chromene 2-Methoxy
6
2-Oxo-2H-chromene 5-Chloro-2-methoxy 7.31
7
2-Oxo-2H-chromene 4-Methoxy
7.74
6.53
6.50
6.47
6.92
5.94
9.12
7.72
7.36
5.79
5.95
7.87
7.91
8
3-Pyridine
4-Pyridine
2-Furan
4-Methyl
4-Methyl
The target compounds were evaluated for their a1-AR
antagonistic activities and the biological data (pA2) are
summarized in Table 1. The result showed that three
compounds, 2, 3, and 13, displayed high a1-AR antago-
nistic pA2 of 8.56, 8.56, and 9.12, respectively. All three
compounds have similar or better a1-AR antagonistic
activities compared with those of prazosin, which is a
classical agent used in clinic for the treatment of BPH.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Prazosin
4-Methyl
4-Methyl
2-Benzofuran
3-Pyridine
2-Furan
2,4-Dimethyl
2-Methoxy
2,4-Dimethyl
2,4-Dimethyl
4-Methoxy
4-Methoxy
2-Methoxy
2-Methoxy
2-Furan
2-Benzofuran
3-Pyridine
4-Pyridine
3-Pyridine
4-Pyridine
3-Pyridine
4-Pyridine
2-Benzofuran
On the basis of these preliminary findings, an exten-
sive SAR study of these aryl-piperazines was initiated
using a novel 3D-QSAR technique, Self-Organizing
Molecular Field Analysis (SOMFA).7,8 The structures
of these compounds were built and optimized by
5-Chloro-2-methoxy 5.56
5-Chloro-2-methoxy 7.18
5-Chloro-2-methoxy 5.63
8.82

3218
M.-Y. Li et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3216–3219
Figure 2. Superposition of all 22 compounds.
molecular mechanics methodology with CVFF force
field using InsightII program.9 The optimized struc-
tures were assigned with Gasteiger–Marsili charges
and performed RMS overlapping to fit with the com-
pound 13, which has the best biological activity, as a
reference, according to three-point pharmacophore.
The final overlaid geometries, as represented in Figure
2, were processed with the SOMFA program to devel-
op the 3D-QSAR model.
In this SOMFA study, shape and electrostatic potential
were generated. To integrate the predictive power of
these two properties into one final model, their individ-
ual predictions were combined using a weighted average
of the shape and electrostatic potential based QSAR or
a mixing coefficient (c1). With the highest value of r2, the
SOMFA models were then derived by the partial least-
squares (PLS) regression with cross-validation. From
statistical analysis of the SOMFA model, good cross-
validated correlation coefficient q2 value (0.708), moder-
ate noncross-validated correlation coefficient r2 value
Figure 3. The shape master grid with compound 13.
(0.743), F test value F(57.67), and standard error of
estimate SE value (0.52) were obtained, indicating a
good conventional statistical correlation. The SOMFA
results indicated that the value of mixing coefficient c1
is 0.9, which means that the contribution of shape field
and electrostatic field to QSAR equation is 90% and
10%, respectively. Table 2 summarizes the actual and
estimated pA2 of 22 compounds that were synthesized.
Table 2. Actual and estimated activities pA2 for compound 1–22 by
SOMFA analysisa
Compound
Actual pA2
Estimated pA2
Error
The SOMFA shape master grid for the analysis is shown
in Figure 3. In this figure, we found a high density of red
points around the 2- and 5-positions at the phenyl ring,
which indicates a favorable steric interaction; simulta-
neously we found high density of blue points around
the 3- and 4-positions at the phenyl ring, and around
the Ar substituent, which means an unfavorable shape
interaction.
1
2
7.07
8.56
8.56
7.85
7.71
7.31
7.74
6.53
6.50
6.47
6.92
5.94
9.12
7.72
7.36
5.79
5.95
7.87
7.91
5.56
7.18
5.63
6.39
8.24
8.30
6.79
8.00
7.79
6.39
6.71
6.50
6.99
6.85
6.54
8.42
7.34
6.90
6.13
6.23
7.33
7.60
5.12
7.30
5.38
0.68
0.32
3
0.26
1.06
4
5
À0.29
À0.48
1.35
6
7
8
À0.18
0.00
9
In conclusion, in the current work we successfully de-
signed a series of potent aryl-piperazine a1-AR antago-
nists based on a three-point pharmacophore. The
study of the 3D-QSAR for this series of aryl-piperazines
may provide some useful information for further design
of more aryl-piperazine derivatives as potent and selec-
tive a1-adrenoceptor antagonists.
10
11
12
13
14
15
16
17
18
19
20
21
22
À0.52
0.07
À0.60
0.70
0.38
0.46
À0.34
À0.28
0.54
0.31
0.44
Acknowledgments
À0.12
0.25
We gratefully acknowledge Dr. Ting Wang (European
Molecular Biology Laboratory, Germany) for the use
a Error = Actual pA2 À Estimated pA2.

M.-Y. Li et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3216–3219
3219
6. For example, compound 2: Yield (48%). 1H NMR (CDCl3,
300 MHz) d (ppm): 3.19–3.21(m, N–CH2, 2H), 3.33 (m, N–
CH2, 2H), 3.50–3.65 (m, 3· N–CH2, 6H), 3.86 (s, ArOCH3,
3H), 3.98 (m, CONH–CH2, 2H), 6.49 (m, Ar–H, 1H), 6.80–
6.96 (m, Ar–H, 1H), 7.03 (m, Ar–H, 1H), 7.04–7.07 (m, Ar–
H, 1H), 7.29–7.31 (m, Ar–H, 1H), 7.55 (m, Ar–H, 1H), 8.51
(s, NHCO, 1H), 12.55 (br, N+H, 1H). IR (cmÀ1): 3437,
3248, 1667, 1498. EI-MS: 363 (M+), 239 (base peak), 154,
95. HR-MS: Calcd mass, 363.1350, Meas. mass, 363.1344.
7. Robinson, D. D.; Winn, P. J.; Lyne, P. D.; Richards, W. G.
J. Med. Chem. 1999, 42, 573.
of 3DFS program. This work was supported by the
National High-Tech Research and Development Plan
(Grant No. 2002AA2Z3118).
References and notes
1. Thorpe, A.; Neal, D. Lancet 2003, 361, 1359.
2. Chess-Williams, R. Expert Opin. Pharmacother. 2002, 3,
167.
3. Pool, J. L.; Kirby, R. S. Int. Urol. Nephrol. 2001, 33, 407.
4. Fang, H.; Lu, J. F.; Xia, L. J. Chin. Pharm. Sci. 2003, 12,
188.
5. Wang, T.; Zhou, J. H. J. Chem. Inf. Comput. Sci. 1998, 38,
71.
8. Li, M. Y.; Du, L. P.; Wu, B.; Xia, L. Bioorg. Med. Chem.
2003, 11, 3945.
9. InsightII molecular modeling package, Version 2000.
Accelrys Inc., San Diego, CA.