Novel 1-oxyl-2-substitutedphenyl-4,4,5,5-tetramethylimidazolines: Synthesis, selectively analgesic action, and QSAR analysis

Article abstract of DOI:10.1016/j.bmc.2007.02.023

Based on the knowledge that imidazoline can result in analgesic action due to its selective binding with the prostacyclin receptor, 20 1-oxyl-2-substitutedphenyl-4,4,5,5-tetramethylimidazolines (3a-t) were prepared in moderate yields. At 0.13 mmol/kg dose, their in vivo analgesic activities were evaluated after the mice were administered at 30, 60, 90, and 150 min. Compared with the pain threshold (12.27 ± 9.56-17.71 ± 7.00%) of normal saline (NS) receiving mice, the pain threshold (23.42 ± 8.14% to 102.58 ± 10.66%) of 3a-t receiving mice increases significantly. Considering a prostacyclin receptor targeting analgesic agent usually had bleeding action and to appraise the bleeding risk, the in vivo tail bleeding time of 1.30 mmol/kg 3a-t receiving mice was found to be ranged from 116.3 ± 8.2 s to 120.3 ± 9.2 s, which was substantially equal to that (117.8 ± 8.4 s to 119.0 ± 8.6 s) of NS receiving mice. Based on the possibility of imidazoline acting as vasodilator, the in vitro vasorelaxations of 3a-t were tested using the rat aortic strip model. When the aortic strip contracted by noradrenaline (NE, final concentration 10^-7 mol/l) was treated with 3a-t (final concentration 5 × 10^-4 mol/l), only lower percentage inhibitions (6.55 ± 5.70-37.40 ± 4.07%) were recorded, implying that the vasorelaxation of 3a-t was neglectable. By selecting appropriate molecular descriptors generated from e-dragon server, the QSAR model of the analgesic activities of 3a-t was constructed using the multiple linear regression method. The established QSAR model showed reasonable accuracy and thus it is promising to be used for screening new 1-oxyl-2-substitutedphenyl-4,4,5,5-tetramethylimidazoline derivatives as analgesic agents.

Full text of DOI:10.1016/j.bmc.2007.02.023

Bioorganic & Medicinal Chemistry 15 (2007) 2815–2826
Novel 1-oxyl-2-substitutedphenyl-4,4,5,5-tetramethylimidazolines:
Synthesis, selectively analgesic action, and QSAR analysis
Ming Zhao,a Zheng Li,a Li Peng,a Yu-Rong Tang,c Chao Wang,b
Ziding Zhangc,* and Shiqi Penga,*
aCollege of Pharmaceutical Sciences, Capital Medical University, Beijing 100054, PR China
bCollege of Pharmaceutical Sciences, Peking University, Beijing 100083, PR China
cCollege of Biological Sciences, China Agricultural University, Beijing 100094, PR China
Received 18 January 2007; revised 12 February 2007; accepted 13 February 2007
Available online 15 February 2007
Abstract—Based on the knowledge that imidazoline can result in analgesic action due to its selective binding with the prostacyclin
receptor, 20 1-oxyl-2-substitutedphenyl-4,4,5,5-tetramethylimidazolines (3a–t) were prepared in moderate yields. At 0.13 mmol/kg
dose, their in vivo analgesic activities were evaluated after the mice were administered at 30, 60, 90, and 150 min. Compared with
the pain threshold (12.27 9.56–17.71 7.00%) of normal saline (NS) receiving mice, the pain threshold (23.42 8.14% to
102.58 10.66%) of 3a–t receiving mice increases significantly. Considering a prostacyclin receptor targeting analgesic agent usually
had bleeding action and to appraise the bleeding risk, the in vivo tail bleeding time of 1.30 mmol/kg 3a–t receiving mice was found to
be ranged from 116.3 8.2 s to 120.3 9.2 s, which was substantially equal to that (117.8 8.4 s to 119.0 8.6 s) of NS receiving
mice. Based on the possibility of imidazoline acting as vasodilator, the in vitro vasorelaxations of 3a–t were tested using the rat aor-
tic strip model. When the aortic strip contracted by noradrenaline (NE, final concentration 10ꢀ7 mol/l) was treated with 3a–t (final
concentration 5 · 10ꢀ4 mol/l), only lower percentage inhibitions (6.55 5.70–37.40 4.07%) were recorded, implying that the vaso-
relaxation of 3a–t was neglectable. By selecting appropriate molecular descriptors generated from e-dragon server, the QSAR model
of the analgesic activities of 3a–t was constructed using the multiple linear regression method. The established QSAR model showed
reasonable accuracy and thus it is promising to be used for screening new 1-oxyl-2-substitutedphenyl-4,4,5,5-tetramethylimidazoline
derivatives as analgesic agents.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
hyperglycemia, depression, liver cell proliferation,
hyperaggregability of platelets, erectile dysfunction and
dermal necrosis. These efforts rationally resulted in some
useful relationships between a certain structure and a
specific inhibition of imidazoline, for instance 2-phenyl-
substituted imidazopurinone possesses high affinity for
human A3 Ars,3 idazoxan selectively blocks a2-adreno-
receptor,7 phentolamine mesylate improves male erectile
dysfunction,8 and 2-phenylaminoimidazoline alleviates
pain,9 which could be used as the clue for developing
more selective drug of imidazoline. By understanding
these relationships, the connection of the structure of
2-phenylaminoimidazoline and its selective inhibition
to prostacyclin receptor was efficaciously used for dis-
covering analgesic agent.9,10 Following this connection,
using radioligand binding technology the prostacyclin
receptor affinities of 460 2-phenylaminoimidazolines
were determined. As a result [(4,5-dihydro-1H-imi-
dazol-2-yl)-[4-(4-isopropoxybenzyl)phenyl]amine] was
found to have analgesic activity in the rat. The results
During the past years intense efforts have been mounted
to develop the structural diversity of imidazolines cho-
sen for the discovery of antagonists of human adenosine
receptors (A3 Ars),1–3 a-adrenergic receptor subtypes,4–7
P2X7 receptor,8 and prostacyclin receptor.9 Activation
of these receptors could result in a series of diseases
including inflammatory, glaucoma, stroke, rheumatoid
arthritis, Alzheimer’s disease, nasal decongestion, hyper-
tension, hypotension, hyperglycemia, liver cell prolifera-
tion, hyperaggregability of platelets, and benign
prostatic hyperplasia, nasal decongestion, hypertension,
Keywords: Tetramethylimidazoline; Analgesic; Vasorelaxation; QSAR.
*
Corresponding authors. Tel.: +86 10 6273 4376; fax: +86 10 6273
1332 (Z.Z.); tel.: +86 10 8391 1528; fax: +86 10 8391 1528 (S.P.);
e-mail addresses: ZidingZhang@cau.edu.cn; sqpeng@mail.bjum.
edu.cn
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.02.023

2816
M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
told us that the 2-position modification of imidazoline
may result in desirable lead compound.
2.2. In vivo analgesic activities of 3a–t
Considering the structural likeness between 3a–t and
analgesic imidazoline reported in the literature,9 the
in vivo analgesic effects of 3a–t were first evaluated by
the tail-flick test.13,14 After the administration of 3a–t
at a level of 0.13 mmol/kg, the pain thresholds of the
mice were measured at 30 min intervals and total mea-
surement was carried out for 150 min. Based on the
tested basic pain threshold (BPT), the tested pain thresh-
old after administration (PTAD), and the AAPT value,
namely PTAD minus BPT, the pain threshold variation
(PTV) was calculated according to the equation
PTV = AAPT/BPT. All values of the PTV for each ani-
mal were averaged and constituted one sample. The data
are listed in Table 1, and the statistical analysis was car-
ried out by use of ANOVA test, p < 0.05 is considered
significant. Under the testing condition, the pain thresh-
old of the mice was increased by nearly all 3a–t after
administering them for 30 min and these analgesic
effects may continue for 150 min.
With this understanding as
a clue for further
structural modification of analgesic imidazoline, the
preparation, in vivo evaluation, and QSAR analysis
of 20 analgesic 1-oxyl-2-substituted-phenyl-4,4,5,5-tet-
ramethyl-imidazolines were performed in the present
study. In order to demonstrate that the analgesic ac-
tion of these 20 1-oxyl-2-substitutedphenyl-4,4,5,5-
tetramethylimidazolines is selective one, their in vivo
tail bleeding time of mouse and in vitro vasorelaxa-
tion activities were also reported.
2. Results and discussion
2.1. Preparations of 1-oxyl-2-substitutedphenyl-4,4,5,5-
tetramethylimidazolines
1-Oxyl-2-substitutedphenyl-4,4,5,5-tetramethylimidazo-
lines 3a–t were synthesized according to the synthetic
route shown in Scheme 1.11,12 After the bromination
and reduction, 2-nitropropane was smoothly converted
into 2,3-bis(hydroxylamino)-2,3-dimethylbutane which
was condensed with substitutedphenyl aldehydes to
give 1,3-dihydroxy-2-substitutedphenyl-4,4,5,5-tetrame-
thylimidazolidines (1a–t) in 31–85% yield. The conden-
sation yield significantly depended on the lipophilicity
and space hindrance of the substituents in the phenyl,
hydrophilic and multiple substitutions usually led
to lower yield (31–85%). The oxidation of 1a–t
with PbO2 provided 1-oxyl-2-(substitutedphen-10-yl)-3-
oxide- 4,4,5,5-tetramethylimidazoline (2a–t) in 52–97%
yield. The oxidation yield also depended on the lipo-
philicity and space hindrance of the substituents in
the phenyl, hydrophilic and multiple substitutions
again led to lower yield (52–68%). The reduction of
2a–t with NaNO2 provided 1-oxyl-2-(substitutedphen-
10-yl)-4,4,5,5-tetramethylimidazoline (3a–t) in 40–98%
yield. The reduction yield depended on the substituent
in the phenyl very much, when the phenyl was substi-
tuted by electron donating polar group, such as
N(CH3)2 and OH, the yield was less than 30%, when
the phenyl was substituted by electron donating nonpo-
lar group, such as CH3 and OCH3, the yield was more
than 50%, and when the phenyl was substituted by
electrondrawing group, such as F, Cl, Br, and NO2,
the yield reached 70%.
As shown in Table 1, all 3a–t exhibited obvious analge-
sic activity at the dosage of 0.13 mmol/kg body weight.
Though after the administration of 30 min the enhanc-
ing pain thresholds in mice were observed for all 3a–t,
the maximal values appeared at different time points.
The enhancing actions of 3e,p–t reached their peaks
after the administration of 60 min, of 3d,f,k reached
their peaks after the administration of 90 min, of
3b,c,g,h reached their peaks after the administration of
120 min, and of 3m,n,o reached their peaks after the
administration of 150 min. The data listed in Table 2
demonstrated that the analgesic activity of 3b–d,f,m,n
was significantly higher than that of the others. From
these data the effect of the substitution in the phenyl
on the analgesic potency can be simply deduced. The
OH, F, Cl, and multiple hydrophobic substitutions usu-
ally led to higher activities, and di-Cl and NO2 substitu-
tions usually led to lower activities. Comparing the
effects of the substituents in the phenyl on the in vitro
vasorelaxation and the in vivo analgesic actions, it was
found that the two kinds of effects exhibited substan-
tially opposite trends. Thus, the in vivo analgesic actions
of 3a–t are independent of their in vitro vasorelaxation.
Based on the relative higher analgesic activity at relative
lower dosage and the relative weaker vasorelaxation
activity at relative higher concentration, it could be
O
O
N
N
HO
iii
N
N OH
iv
O
N
N
NO2
i
ii
v
H3C CH CH3
O2N NO2
HO NHHN OH
R
R
2a-t
1a-t
R
3a-t
Scheme 1. Synthetic route of 1-oxyl-2-(substitutedphen-10-yl)-4,4,5,5-tetramethylimidazolines. (i) Br2 and NaOH (6 mol/l), (ii) Zn and NH4Cl, (iii)
R-C6H4CHO, (iv) PbO2. (v) NaNO2 and hydrochloric acid (pH 5–6). In (1–3)a R = H, (1–3)b R = 2-OH, (1–3)c R = 3-OH, (1–3)d R = 4-OH, (1–3)e
R = 4-CH3, (1–3)f R = 4-OCH3, (1–3)g R = 2,4-dimethoxyl, (1–3)h R = 3,4-dimethoxyl, (1–3)i R = 3,4-methylenedioxy, (1–3)j R = 3-OH-4-OCH3,
(1–3)k R = 3-OCH3-4-OH, (1–3)l R = 4-N(CH3)2, (1–3)m R = 2-F, (1–3)n R = 4-Cl, (1–3)o R = 4-Br, (1–3)p R = 2,4-di-Cl, (1–3)q R = 3,4- di-Cl,
(1–3)r R = 2-NO2, (1–3)s R = 3-NO2, (1–3)t R = 4-NO2.

M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
Table 1. The effect of 3a–t on the pain threshold in mice
2817
Compound
Pain threshold variation ðꢀx ꢁ SD%Þ
90 min
30 min
60 min
120 min
150 min
Control
3a
3b
3c
12.45 7.21
13.40 8.47
36.99 11.60a
45.39 11.42b
50.30 11.75a
70.80 11.37a
53.56 11.30a
102.58 10.66a
59.73 10.45a
60.22 12.60a
64.96 11.80a
48.18 11.88a
42.86 10.38a
35.49 10.30a
67.78 11.18a
66.29 10.42a
35.56 9.23a
63.34 10.83a
72.76 12.63a
66.89 12.23a
66.37 11.72a
60.55 11.07a
17.71 7.00
12.27 9.56
17.43 9.18
40.60 10.71a
48.57 10.59a
49.60 11.78a
55.66 10.38a
40.16 9.27a
35.34 9.35a
42.00 10.84a
44.92 10.26b
52.83 11.27b
33.08 10.03a
54.18 11.17a
30.37 9.00a
62.19 10.40a
59.33 10.25a
41.32 9.62a
39.64 10.61a
50.57 11.04a
38.36 9.38a
49.81 10.75a
56.49 11.11a
34.00 10.11a
64.20 12.46a
53.99 10.74a
75.88 12.60a
43.72 10.04a
99.30 11.70a
62.49 11.09a
59.50 11.97a
74.02 12.43a
48.38 10.23a
66.34 11.61a
36.19 10.00a
63.27 10.21a
61.68 10.29a
59.84 10.60a
48.37 9.22a
55.72 9.45a
57.25 11.54a
47.41 10.19a
57.95 11.55a
32.14 11.55a
70.17 12.00a
72.72 12.07a
55.54 12.40a
35.17 11.14a
68.70 11.97a
71.19 12.22a
78.44 12.50a
58.34 11.85a
56.42 10.72a
43.25 10.64a
34.30 10.61a
61.95 10.04a
50.78 10.22a
59.97 11.22a
30.51 9.48a
28.78 9.34a
23.42 8.14b
43.45 10.08a
35.90 10.37a
33.48 9.19a
62.22 10.50a
70.80 10.04a
29.92 9.46b
25.47 10.41
53.05 9.34a
32.39 9.43a
55.57 11.99a
25.44 10.60b
59.76 11.34a
40.92 10.88a
27.87 9.09b
86.78 12.10a
84.48 11.33a
76.45 11.99a
14.75 9.57
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
3q
3r
26.96 9.33b
19.69 9.91
41.84 10.97a
52.27 11.59a
3s
3t
n = 10.
a Compare to control p < 0.01.
b Compare to control p < 0.05.
Table 2. Effect of 3a–t on the tail bleeding time ðꢀx ꢁ SD%Þ of mice
Compound
Before administration
After administration
5 min
15 min
30 min
60 min
NS
3a
3b
3c
3d
3e
3f
119.0 8.6
117.6 8.1
116.8 8.6
118.4 9.0
120.1 9.1
119.4 8.9
117.4 8.0
120.0 9.3
119.3 8.1
116.5 8.3
118.5 8.2
117.3 8.0
119.5 9.6
118.4 8.5
116.8 9.1
116.3 8.2
118.6 9.1
119.7 9.3
117.7 8.4
118.8 9.0
120.3 9.2
117.8 8.4
118.2 7.6
116.5 9.1
118.8 8.2
118.0 8.5
117.6 8.3
117.7 8.1
119.3 8.3
118.5 8.7
117.4 7.9
117.3 7.6
117.8 9.0
119.2 9.0
119.3 8.8
119.0 8.8
116.7 8.4
118.3 8.9
120.2 8.8
117.5 8.5
118.4 8.9
119.4 8.8
118.3 7.7
117.4 7.9
118.3 7.6
117.2 8.0
116.9 7.6
116.8 8.1
116.5 7.9
116.8 8.2
118.8 9.1
116.8 8.8
116.9 9.1
116.4 9.2
118.7 8.9
118.4 8.3
117.3 9.2
117.5 8.6
118.9 8.9
119.3 9.2
117.8 8.1
116.5 8.6
119.3 8.9
118.2 8.3
119.3 8.1
117.1 7.8
119.3 9.2
117.3 7.7
118.2 9.1
119.3 8.9
117.7 7.9
119.5 9.4
117.9 9.2
118.0 7.9
117.8 8.0
118.5 8.3
117.9 8.6
118.0 8.8
116.6 8.4
117.3 9.3
119.1 8.9
117.2 8.0
117.6 8.5
118.6 8.3
117.9 8.1
118.2 7.6
116.6 8.2
117.3 8.0
117.8 8.2
117.9 7.6
119.1 9.1
118.1 8.4
120.5 9.0
116.8 7.5
117.3 7.6
118.5 8.3
117.7 9.1
117.3 8.5
117.5 8.0
117.4 8.3
117.7 9.2
118.7 9.0
116.7 8.3
118.2 9.0
118.9 8.7
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
3q
3r
3s
3t
n = 10, dose, 1.3 mmol/kg; NS (normal saline), vehicle.
postulated that 3a–t may be selective inhibitor of the
prostacyclin receptor.
(5–60 min) fell into a range of 116–120 s which were sub-
stantially equal to that before administration (Table 2).
In each case the dose was 1.30 mmol/kg body weight.
This observation implied that at ten fold analgesic dose
3a–t exhibited no bleeding risk.
2.3. In vivo tail bleeding activities of 3a–t
As the prostacyclin receptor targeting analgesic agents
usually have bleeding action, to appraise the bleeding
risk of 3a–t, their tail bleeding time assays were per-
formed on mice using a literature method.15 The
observed bleeding time in mice at all time points tested
2.4. In vitro vasorelaxation activities of 3a–t
Since the structure of 3a–t is similar to that of the
vasodilator imidazoline,11,16,17 the conversions of 3a–t

2818
M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
during vasoconstriction induced by noradrenaline (NE)
were examined using the rat aortic strip model. Based on
the model, the aortic strip contracted by NE
(final concentration 10ꢀ7 mol/l) was treated with 3a–t
(final concentration 5 · 10ꢀ4 mol/l, 1 · 10ꢀ4 mol/l, and
2 · 10ꢀ5 mol/l). It was noticed that even at 5 ·
10ꢀ4 mol/l the NE contracted aortic strip can be
relaxed less than 40% by 3a–t (Table 3). This observa-
tion demonstrates 3a–t are weak vasodilators.
2.6. QSAR analysis of 3a–t
Selecting the molecular descriptors generated from e-dra-
gon webserver,18 the multiple linear regression method
was employed to construct the QSAR equation of 3a–t.
The resubstitution and Leave-One-Out (LOO) tests were
carried out to validate the established equations. For 3a–t
Eq. 1 could give the highest predictive accuracy.
PTV90 ¼ 927:3ðꢁ220:5Þ ꢀ 43:5ðꢁ14:7Þ ꢂ GATS4e
ꢀ610:7ðꢁ172:7Þ ꢂ BELe1 ꢀ 2706:5
ðꢁ824:2Þ ꢂ JGI1 þ 92:5ðꢁ36:8Þ ꢂ R2u
N ¼ 20; R ¼ 0:79; S ¼ 8:7; ꢀe ¼ 7:0; RLOO ¼ 0:76;
SLOO ¼ 11:1; ꢀeLOO ¼ 9:3; F ¼ 6:23; p ¼ 0:004:
ð1Þ
2.5. Dose-dependent in vivo analgesic activities of 3c,f,m
Compounds with significant activity (3c,f,m) were
administered using three dosages to study detailed phar-
macological activity profile (Table 4). Statistical analysis
was carried out using one way ANOVA test with
p < 0.05 as significant cut-off. The data indicated that
the in vivo anti-thrombolysis activities of intravenous
3c,f,m were dose-dependent.
In the current QSAR analysis, the PTV90 (i.e., the PTV
value measured at 90 min) is considered as the analgesic
Table 3. Conversion of 3a–t during vasoconstriction induced by NE (n = 6)
Compound
Vasorelaxation ðꢀx ꢁ SD%Þ at following concentrations
1 · 10ꢀ4 mol/l
5 · 10ꢀ4 mol/l
2 · 10ꢀ5 mol/l
3a
3b
3c
3d
3e
3f
17.25 4.79
20.32 4.32
35.70 4.96
8.71 3.18
31.77 4.18
35.08 4.10
37.08 4.03
33.73 4.33
37.40 4.07
16.65 3.52
6.55 5.70
39.98 2.63
23.40 3.92
36.59 5.17
34.78 2.98
34.82 5.03
34.37 3.45
35.58 2.37
34.27 3.91
36.53 4.66
11.04 3.02
13.38 2.99
22.98 3.11
7.66 4.03
19.96 3.21
23.15 2.79
24.03 3.13
22.33 3.05
24.51 2.99
11.04 2.86
6.09 3.33
26.00 3.36
14.94 2.89
23.78 3.03
23.01 3.22
22.99 3.30
22.71 2.98
20.39 3.07
22.07 2.99
24.44 3.32
5.44 2.95
6.01 3.05
7.06 2.78
5.36 4.00
7.00 2.47
6.88 2.80
7.33 2.74
6.66 2.95
8.10 3.30
5.36 3.00
5.98 6.02
8.02 3.46
5.00 2.97
7.19 3.04
6.89 3.01
6.86 3.00
6.74 2.89
6.98 2.76
6.72 3.34
7.36 3.11
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
3q
3r
3s
3t
Table 4. Effect of 3c,f,m at different dosages on the pain threshold in mice
Compound, dose (mmol/kg)
Pain threshold variation ðꢀx ꢁ SD%Þ
30 min
60 min
90 min
120 min
150 min
Control
12.45 7.21
49.60 11.78a
35.69 9.30b
24.16 9.07c
35.34 9.35a
25.00 7.84b
14.92 7.26
62.19 10.40a
49.33 10.25b
31.32 9.32c
13.40 8.47
50.30 11.75a
37.80 10.30b
25.56 10.00c
102.58 10.66a
89.73 10.15b
55.22 10.60c
67.78 11.18a
51.29 10.12b
33.52 9.33c
17.71 7.00
12.27 9.56
17.43 9.18
3c, 0.1300
3c, 0.0130
3c, 0.0013
3f, 0.1300
3f, 0.0130
3f, 0.0013
3m, 0.1300
3m, 0.0130
3m, 0.0013
53.99 10.74a
42.80 10.60b
30.72 10.04c
99.30 11.70a
82.49 10.09b
50.50 10.97c
63.27 10.21a
50.68 10.09b
31.84 9.60c
72.72 12.07a
59.94 10.41b
35.17 11.14c
68.70 11.97a
50.19 10.22b
38.44 10.50c
61.95 10.04a
49.78 9.22b
30.37 10.02c
70.80 10.04a
52.12 9.06b
33.47 10.41c
53.05 9.34a
34.39 9.23b
22.57 7.99c
86.78 12.10a
64.48 10.33b
44.45 9.99c
n = 10.
a Compare to group of 0.0130 mmol/kg dose, p < 0.01.
b Compare to group of 0.0013 mmol/kg dose, p < 0.05.
c Compare to control p < 0.01.

M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
2819
activity of a compound. The resubstitution analysis gave
an absolute average error ðꢀeÞ of 7.0, whereas the abso-
lute average error for the LOO test ðꢀeLOOÞ is 9.3. The
detailed description of the other statistical measure-
ments in Eq. 1 is available in Section 3. Grouped into
the 2D autocorrelation descriptors, the GATS2e is re-
lated to the atomic electronegativities of a molecule.19
The BELe1 is a molecular descriptor obtained from
the lowest eigenvalue of the adjacency matrix, weighting
the diagonal elements with atomic Sanderson electro-
negativities.20 The JGI1 belongs to the descriptors of
charge indices, which evaluate the global charge trans-
ferred between pairs of atoms inside the molecule.21
The R2u is classified into the GETAWAY (GEometry,
Topology, and Atom-Weights AssemblY) molecular
descriptors, which try to match 3D-molecular geometry
provided by the molecular influence matrix and atom
relatedness by molecular topology with chemical infor-
mation by using different atomic weightings.22
the optimized accuracy, Eq. 1 is unable to provide a
more detailed insight into the molecular mechanism of
3a–t to exert their analgesic effects. This is mainly due
to the fact that physicochemical information hidden in
these selected descriptors is complicated. It can be envi-
sioned that the analgesic effects of 3a–t are caused by
their interactions with the same receptor, therefore a de-
tailed CoMFA26 may be able to provide new insight into
the SAR of these compounds. Moreover, molecular
docking studies between the receptor and these com-
pounds will be very helpful in understanding the molec-
ular mechanisms of their analgesic effects provided that
the 3D structure of the receptor is known or can be
confidently predicted.
3. Experimental
3.1. General
Considering the standard deviation in measuring PTV90
(cf. Table 1), the current predictive accuracy of 3a–t is
quite reasonable (Fig. 1). Therefore, Eq. 1 can be prac-
tically used to forecast the analgesic activities of new
1-oxyl-2-substitutedphenyl-4,4,5,5-tetramethylimidazoline
derivatives. In Eq. 1, three descriptors (GATS2e,
BELe1, and JGI1) are significantly related to a mole-
cule’s electrostatic property, while the remaining
descriptor R2u is more correlated with the steric charac-
ter of a molecule. Therefore, the overall analgesic activ-
ities of 3a–t can be reasonably explained by their
electrostatic and steric effects.
The purity of all the compounds was confirmed by TLC
(Merck silica gel plates of type 60 F254, 0.25 mm
layer thickness) and HPLC (waters, C18 column
3.9 · 150 mm). Melting points were measured on a
XT5 hot stage microscope (Beijing keyi electro-optic
factory). The infrared spectra were recorded with a Per-
kin–Elmer 983 instrument. The EI-MS was determined
by Trace MS System instrument (American Thermo
1
Finnigan Company). The H NMR was determined by
Varian INOVA-300 MHz spectrometer. The ESR spec-
tra were obtained from 10ꢀ5 mol/l phosphate-buffered
saline, using a BRUKER 300-E spectrometer. The con-
ditions of measurements are as follows: center field:
3440 G, sweep width: 100 G, sweep time: 60 s, modula-
tion amplitude: 1.1 G, time constant 8.2 · 10ꢀ2 s, modu-
lation frequency 100 kHz, microwave frequency
9.68 GHz, and microwave power: 20 MW.
The QSAR studies on the analgesic activities of different
compounds have been widely reported in the litera-
ture.23–25 In addition to obtain the predictable QSAR
models, the molecular mechanisms of analgesic effects
were also derived to some extent. Focused on obtaining
3.2. Synthesis of the compounds
3.2.1. 2,3-Dimethyl-2,3-dinitrobutane. At ꢀ5 ꢁC, 10 ml
(0.19 mol) of Br2 was added dropwise within 1 h to the
solution of 34.5 g (0.39 mol) of 2-nitropropane in
65 ml (6.0 mol/l) aqueous NaOH. One hundred and
twenty-eight milliliters of ethanol was added into the
solution via stirring. The reaction mixture was stirred
at 84 ꢁC for 3 h. The hot reaction mixture was trans-
ferred into 400 ml of ice water. The formed colorless
crystals were collected by filtration to yield 25 g (73%)
of the title compound, mp 110–112 ꢁC.
100
90
80
70
60
50
40
30
20
3.2.2. 2,3-Bis(hydroxylamino)-2,3-dimethylbutane. 2,3-
Dimethyl-2,3-dinitrobutane (17.6 g, 0.1 mol) was dis-
solved in a mixture of THF (300 ml) and water
(50 ml). Zn powder (27 g) was added all at once to this
solution cooled to 8–10 ꢁC in an ice bath. A solution
of NH4Cl (43 g, 0.8 mol) in H2O (150 ml) was added
dropwise to this slurry within 2 h, with continued stir-
ring for 1 h at 10 ꢁC, and the flask was stored in a fridge
(4–6 ꢁC) for 16 h. The slurry was filtered, and the precip-
itate was carefully washed with THF (4 · 100 ml). The
precipitate was then dried by three washings with diethyl
ether and carefully collected (59 g). The solution was
20
30
40
50
60
70
80
90
100
Tested PTV
Figure 1. Comparison of the predicted analgesic activities of com-
pounds 3a–t in the LOO test with the tested activities. The data points
situated at the zone between the two diagonal-dash lines correspond to
the compounds, in which the absolute values of the prediction errors
are less than 20.0.

2820
M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
evaporated under vacuum until THF ceased to distill
off. Then the solution was protected from air, and
sodium carbonate (50 g) and sodium chloride (30 g)
were added with cooling. Continuous extraction with
chloroform (400 ml) was performed over 18 h. A white
powder was obtained (9.4 g, 63% yield). Mp 182 ꢁC.
d = 1.14 (s, 6H), 1.19 (s, 6H), 1.37 (s, 3H), 4.90 (s,
1H), 7.68 (d, J = 9.0 Hz, 2H), 8.22 (d, J = 8.4 Hz, 2H),
8.35 (s, 2H).
3.2.3.6. 1,3-Dihydroxy-2-(40-methoxyphen-10-yl)-4,4,
5,5-tetramethylimidazolidine (1f). Using the general pro-
cedure mentioned above 372 mg (70%) of the title com-
pound was obtained. Mp 179–181 ꢁC. Rf = 0.52
(CHCl3/CH3OH, 10:1). EI-MS (m/z) 266[M]+. IR (KBr)
3.2.3. General procedure for the preparation of 1,3-
dihydroxy-2-substituted-phenyl-4,4,5,5-tetramethylimidaz-
olidine (1a–t). A solution of 296 mg (2 mmol) of 2,3-
bis(hydroxylamino)-2,3-dimethylbutane and 2 mmol of
various aldehydes in 3 ml of methanol was stirred at
room temperature for 16 h. The title compound was col-
lected by filtration. The filtration was evaporated under
reduced pressure to obtain the second crop of the title
compound. The title compound was directly used for
the next reaction without further purification.
3340, 2835, 1500, 825 cmꢀ1 1H NMR (DMSO-d6)
.
d = 0.99 (s, 6H), 1.03 (s, 6H), 3.73 (s, 3H), 4.56 (s, 1H),
6.88 (d, J = 4.2 Hz, 2H), 7.38 (d, J = 5.7 Hz, 2H), 7.77
(s, 2H).
3.2.3.7. 1,3-Dihydroxyl-2-(20,40-dimethoxylphenyl)-4,
4,5,5-tetramethylimidazolidine (1g). Using the general
procedure mentioned above 349 mg (59%) of the title
compound was obtained. Mp 140–141 ꢁC. Rf = 0.49
(CHCl3/CH3OH, 10:1). EI-MS (m/z) 296[M]+. IR
3.2.3.1. 1,3-Dihydroxy-2-phenyl-4,4,5,5-tetramethyl-
imidazolidine (1a). Using the general procedure men-
tioned above 401 mg (85%) of the title compound was
obtained. Mp 168–169 ꢁC. IR (KBr): 3340, 1606, 1500,
(KBr) 3295, 2825, 1500, 865, 810 cmꢀ1 1H NMR
.
(DMSO-d6) d = 1.03 (s, 12H), 3.72 (s, 6H), 4.99 (s,
1H), 7.02 (s, 1H), 7.42 (d, J = 6.7 Hz, 1H), 7.45 (d,
J = 6.5 Hz, 1H), 7.79 (s, 2H).
1
1450, 1381 cmꢀ1; H NMR (CDCl3) d = 1.14 (s, 12H),
4.78 (s, 1H), 7.31–7.51 (m, J = 6.9 Hz, 5H), 7.71 (s,
2H); ESI-MS (m/e) = 236[M]+.
3.2.3.8. 1,3-Dihydroxy-2-(30,40-dimethoxyphen-10-yl)-
4,4,5,5-tetramethylimidazolidine (1h). Using the general
procedure mentioned above 390 mg (66%) of the title
compound was obtained. Mp 147–149ꢁC. Rf = 0.52
(CHCl3/CH3OH, 6:1). EI-MS (m/z) 296[M]+. IR (KBr)
3.2.3.2. 1,3-Dihydroxy-2-(20-hydroxylphen-10-yl)-4,4,
5,5-tetramethylimidazolidine (1b). Using the general pro-
cedure mentioned above 286 mg (57%) of the title com-
pound was obtained. Mp 139–141ꢁC. Rf = 0.73 (CHCl3/
CH3OH, 6:1). EI-MS (m/z) 252[M]+. IR (KBr) 3325,
1
3310, 2825, 1610, 865, 815 cmꢀ1. HNMR (DMSO-d6)
d = 1.05 (s, 12H), 3.74 (s, 6H), 4.53 (s, 1H), 6.85 (s,
1H), 6.87 (d, J = 6.6 Hz, 1H), 6.92 (d, J = 6.4 Hz, 1H),
7.85 (s, 2H).
1
1600, 1500, 765 cmꢀ1. H NMR (DMSO-d6) d=1.08 (s,
12H), 4.63 (s, 1H), 6.70 (d, J = 6.3 Hz, 1H), 6.74 (t,
J = 6.5 Hz, 1H), 7.14 (t, J = 6.6 Hz, 1H), 7.20 (d,
J = 6.4 Hz, 1H), 8.12 (s, 1H), 8.35 (s, 2H).
3.2.3.9. 1,3-Dihydroxyl-2-(30,40-methylendioxyphen-
10-yl)-4,4,5,5-tetramethylimidazolidine (1i). Using the
general procedure mentioned above 240 mg (43%) of
the title compound was obtained. Mp 188–190ꢁC,
Rf = 0.47 (CHCl3/CH3OH, 10:1), EI-MS (m/z)
280[M]+, IR (KBr) 3315, 1600, 915, 1235,
3.2.3.3. 1,3-Dihydroxy-2-(30-hydroxyphen-10-yl)-4,4,
5,5-tetramethylimidazolidine (1c). Using the general pro-
cedure mentioned above 304 mg (60%) of the title com-
pound was obtained. Mp 244–246ꢁC. Rf = 0.56 (CHCl3/
CH3OH, 6:1). EI-MS (m/z) 252[M]+. IR (KBr) 3320,
1105 cmꢀ1 1H NMR (DMSO-d6) d = 1.05 (s, 12H),
.
3.74 (s, 6H), 4.53 (s, 1H), 6.82 (d, J = 7.7 Hz, 1H),
6.84 (d, J = 8.0 Hz, 1H), 6.87 (s, 1H) 7.85 (s, 2H).
1500, 880, 795, 695 cmꢀ1
.
1H NMR (DMSO-d6)
d=1.03 (s, 6H), 1.05 (s, 6H), 4.51 (s, 1H), 6.89 (dd,
J = 8.4 Hz, J = 8.0 Hz, 1H), 6.92 (d, J = 6.2 Hz, 1H),
7.01 (d, J = 7.3 Hz, 1H), 7.71 (s, 1H), 7.86 (s, 1H),
7.95 (s, 2H).
3.2.3.10. 1,3-Dihydroxyl-2-(30-hydroxyl-40-methoxyl)-
phenyl-4,4,5,5-tetramethylimidazolidine (1j). Using the
general procedure mentioned above 276 mg (49%) of
the title compound was obtained. Mp 245–247ꢁC.
Rf = 0.52 (CHCl3/CH3OH, 6:1). EI-MS (m/z) 282[M]+.
3.2.3.4. 1,3-Dihydroxy-2-(40-hydroxyphen-10-yl)-4,4,5,5-
tetramethylimidazolidine (1d). Using the general proce-
dure mentioned above 275 mg (51%) of the title
compound was obtained. Mp 240–242 ꢁC. Rf 0.67
(CHCl3/CH3OH, 6:1). EI-MS (m/z) 252[M]+. IR (KBr)
1
IR (KBr) 3340, 2825, 1500, 1450, 825 cmꢀ1. H NMR
(DMSO-d6) d = 1.03 (s, 6H), 1.04 (s, 6H) 3.72 (s, 3H),
4.53 (s, 1H), 6.82 (s, 1H), 6.89 (d, J = 6.6 Hz, 1H),
6.94 (d, J = 6.4 Hz, 1H), 7.65 (s, 1H), 8.04 (s, 2H).
3310, 1610, 1500, 1450, 794, 693 cmꢀ1
.
1H NMR
(DMSO-d6) d = 1.03 (s, 6H), 1.05 (s, 6H), 4.39 (s, 1H),
6.70 (d, J = 6.9 Hz, 2H), 7.23 (d, J = 6.9 Hz, 2H), 7.63
(s, 1H), 7.85 (s, 2H).
3.2.3.11. 1,3-Dihydroxyl-2-(30-methoxyl-40-hydroxy-
phen-10-yl)-4,4,5,5-tetramethylimidazolidine (1k). Using
the general procedure mentioned above 186 mg
(33%) of the title compound was obtained. Mp 235–
237ꢁC, Rf = 0.47 (CHCl3/pCH3OH, 6:1), EI-MS (m/z)
3.2.3.5. 1,3-Dihydroxy-2-(40-methylphen-10-yl)-4,4,5,
5-tetramethylimidazolidine (1e). Using the general proce-
dure mentioned above 350 mg (70%) of the title com-
pound was obtained. Mp 199–201 ꢁC. Rf = 0.61
(CHCl3/CH3OH, 10:1). EI-MS (m/z) 250[M]+. IR
282[M]+, IR (KBr) 3310, 2840, 1600, 870, 815 cmꢀ1
.
1H NMR (DMSO-d6) d = 1.00 (s, 6H), 1.03 (s, 6H),
3.74 (s, 3H), 4.41 (s, 1H), 6.70 (d, J = 7.8 Hz, 1H),
1
(KBr) 3335, 2985, 1600, 815 cmꢀ1. H NMR (CDCl3)

M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
2821
6.85 (d, J = 8.1 Hz, 1H), 7.02 (s, 1H), 7.68 (s, 1H), 8.05
(s, 2H).
(KBr) 3325, 1600, 1535, 1365, 750 cmꢀ1
.
1H NMR
(DMSO-d6) d = 1.00 (s, 6H), 1.04 (s, 6H), 5.37 (s, 1H),
7.52 (d, J = 6.6 Hz, 1H), 7.60 (t, J = 7.8 Hz, 1H), 7.92
(t, J = 7.6 Hz, 1H), 8.05 (d, J = 7.2 Hz, 1H), 8.23 (s,
2H).
3.2.3.12. 1,3-Dihydroxyl-2-(40-N,N-dimethylphen-10-
yl)-4,4,5,5-tetramethylimidazolidine (1l). Using the
general procedure mentioned above 173 mg (31%) of
the title compound was obtained. Mp 168–169ꢁC.
Rf = 0.53 (CHCl3/CH3OH, 6:1). EI-MS (m/z) 279[M]+.
3.2.3.19. 1,3-Dihydroxy-2-(30-nitrophen-10-yl)-4,4,5,5-
tetramethylimidazolidine (1s). Using the general proce-
dure mentioned above 421 mg (75%) of the title
compound was obtained. Mp 179–181 ꢁC. Rf = 0.51
(CHCl3/CH3OH, 10:1). EI-MS (m/z) 281[M]+. IR
1
IR (KBr) 3345, 2840, 1595, 1450, 840 cmꢀ1. H NMR
(DMSO-d6) d = 1.03 (s, 6H), 1.05 (s, 6H), 2.85 (s, 6H),
4.01 (s, 1H), 6.68 (d, J = 8.7 Hz, 2H), 7.25 (d,
J = 8.7 Hz, 2H), 7.60 (s, 2H).
(KBr) 3315, 1530, 1360, 1600, 875, 790, 685 cmꢀ1. H
1
NMR (DMSO-d6) d = 0.52 (s, 12H), 4.19 (s, 1 H), 7.04
(dd, J = 8.7 Hz, J = 8.1 Hz, 1H), 7.34 (d, J = 6.0 Hz,
1H), 7.57 (d, J = 7.5 Hz, 1H), 7.77 (s, 1H), 8.03 (s, 2H).
3.2.3.13. 1,3-Dihydroxy-2-(20-fluorophen-10-yl)-4,4,
5,5-tetramethylimidazolidine (1m). Using the general
procedure mentioned above 366 mg (72%) of the title
compound was obtained. Mp 172–173ꢁC, Rf = 0.69
(CHCl3/CH3OH, 6:1), EI-MS (m/z) 254[M]+, IR
3.2.3.20. 1,3-Dihydroxy-2-(40-nitrophen-10-yl)-4,4,5,5-
tetramethylimidazolidine (1t). Using the general proce-
dure mentioned above 335 mg (63%) of the title
compound was obtained. Mp 172–174 ꢁC. Rf = 0.59
(CHCl3/CH3OH, 10:1). EI-MS (m/z) 281[M]+. IR
(KBr) 3325, 1532, 1590, 1500, 1450, 1365, 835, 790,
(KBr) 3310, 1600, 1130, 1235, 775 cmꢀ1 1H NMR
.
(DMSO-d6) d = 1.07 (s, 12H), 4.95 (s, 1H), 7.18 (t,
J = 8.4 Hz, 1 H), 7.30 (d, J = 8.1 Hz, 1H), 7.68 (d,
J = 8.7 Hz, 2H), 8.14 (s, 2H).
1
685 cmꢀ1. H NMR (DMSO-d6) d = 1.00 (s, 6H), 1.05
3.2.3.14.
1,3-Dihydroxy-2-(40-chlorophen-10-yl)-4,4,
(s, 6H), 4.71 (s, 1H), 7.70 (d, J = 9.0 Hz, 2H), 8.21 (d,
J = 7.2 Hz, 2H), 8.40 (s, 2H).
5,5-tetramethylimidazolidine (1n). Using the general pro-
cedure mentioned above 330 mg (61%) of the title com-
pound was obtained. Mp 213–215ꢁC. Rf = 0.73 (CHCl3/
CH3OH, 10:1). EI-MS (m/z) 270[M]+. IR (KBr) 3325,
3.2.4. General procedure for the preparation of 1-oxyl-2-
substitutedphenyl-3-oxide-4,4,5,5-tetramethyl-imidazoli-
dines (2a–t). The mixture of 250 mg (0.8 mmol) of 1,3-
dihydroxy-2-substitutedphenyl-4,4,5,5-tetramethylimi-
dazolidine and 0.5 g of lead dioxide in 80 ml methanol
was stirred at room temperature for 20 min and TLC
(CHCl3/CH3OH, 20:1) indicated the complete disap-
pearance of 1,3-dihydroxy-2-substitutedphenyl-4,4,5,5-
tetramethylimidazolidine. The reaction mixture was
filtered and the filtrate was evaporated under vacuum
to provide the title compound as dark blue crystals.
1600, 1500, 1085, 825 cmꢀ1
.
1H NMR (CDCl3)
d = 1.03 (s, 6H), 1.07 (s, 6H), 4.50 (s, 1H), 7.38 (d,
J = 8.7 Hz, 2H), 7.48 (d, J = 8.4 Hz, 2H), 7.80 (s, 2H).
3.2.3.15. 1,3-Dihydroxy-2-(40-bromophen-10-yl)-4,4,
5,5-tetramethylimidazolidine (1o). Using the general pro-
cedure mentioned above 320 mg (51%) of the title com-
pound was obtained. Mp 205–207 ꢁC. Rf = 0.69 (CHCl3/
CH3OH, 10:1). EI-MS (m/z) 314[M]+. IR (KBr) 3310,
1590, 1450, 1075, 830 cmꢀ1
.
1H NMR (CDCl3)
d = 1.14 (s, 12H), 4.77 (s,1H), 7.37 (d, J = 8.4 Hz, 2H),
7.47 (d, J = 7.8 Hz, 2H), 7.73 (s, 2H).
3.2.4.1. 1-Oxyl-2-(phen-10-yl)-3-oxide-4,4,5,5-tetrame-
thylimidazoline (2a). Using the general procedure men-
tioned above 74 mg (80%) of the title compound was
obtained. Mp 89–91 ꢁC. Rf = 0.85 (CHCl3/CH3OH,
20:1). EI-MS (m/z) 311[M]+. IR (KBr) 1600, 1450,
1357, 1070, 825 cmꢀ1. ESR: a five-line pattern with
intensity ratios of 1:2:3:2:1, aN = 8.16G, g = 2.00997.
3.2.3.16. 1,3-Dihydroxyl-2-(20,40-dichlorophenyl)-4,4,
5,5-tetramethylimidazolidine (1p). Using the general pro-
cedure mentioned above 427 mg (70%) of the title com-
pound was obtained. Mp 201–203ꢁC. Rf = 0.69 (CHCl3/
CH3OH, 10:1). EI-MS (m/z) 305[M]+. IR (KBr) 3315,
1
1590, 1450, 995, 820, 865 cmꢀ1. H NMR (DMSO-d6)
3.2.4.2. 1-Oxyl-2-(20-hydroxylphen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2b). Using the general proce-
dure mentioned above 111 mg (64%) of the title com-
pound was obtained. Mp 83–85 ꢁC. Rf = 0.75 (CHCl3/
CH3OH, 20:1). EI-MS (m/z) 217[M]+. IR (KBr) 3345,
1600, 1500, 1450, 760 cmꢀ1. ESR: a five-line pattern
with intensity ratios of 1:2:3:2:1, aN = 8.16G,
g = 2.00997.
d = 1.11 (s, 12H), 5.01 (s, 1H), 7.50 (d, J = 7.6 Hz,
1H), 7.52 (d, J = 8.1 Hz, 1H), 7.58 (s, 1H), 7.79 (s, 2H).
3.2.3.17. 1,3-Dihydroxyl-2-(30,40-dichlorophenyl)-4,4,
5,5-tetramethylimidazolidine (1q). Using the general pro-
cedure mentioned above 427 mg (70%) of the title com-
pound was obtained. Mp 212–214ꢁC. Rf = 0.69 (CHCl3/
CH3OH, 10:1). EI-MS (m/z) 305[M]+. IR (KBr) 3313,
1
1592, 1453, 992, 863, 822 cmꢀ1. H NMR (DMSO-d6)
3.2.4.3. 1-Oxyl-2-(30-hydroxylphen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2c). Using the general proce-
dure mentioned above 151 mg (76%) of the title com-
pound was obtained. Mp 137–139 ꢁC. Rf = 0.33
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 249[M]+. IR
d = 1.06 (s, 12H), 4.99 (s, 1H), 7.66 (m, 3H), 7.92 (s, 2H).
3.2.3.18. 1,3-Dihydroxy-2-(20-nitrophen-10-yl)-4,4,5,5-
tetramethylimidazolidine (1r). Using the general
procedure mentioned above 410 mg (73%) of the title
compound was obtained. Mp 189–190ꢁC. Rf = 0.54
(CHCl3/CH3OH, 10:1). EI-MS (m/z) 281[M]+. IR
(KBr) 3340, 1590, 1450, 1380, 880, 800, 690 cmꢀ1
.
ESR: a five-line pattern with intensity ratios of
1:2:3:2:1, aN = 8.18 G, g = 2.00994.

2822
M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
3.2.4.4. 1-Oxyl-2-(40-hydroxylphen-10-yl)-3-oxide-4,4,
3.2.4.11. 1-Oxyl-2-(30-methoxy-40-hydroxyphen-10-yl)-
3-oxide-4,4,5,5-tetramethylimidazoline (2k). Using the
general procedure mentioned above 158 mg (71%) of
the title compound was obtained. Mp 83–85 ꢁC.
Rf = 0.38 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
279[M]+. IR (KBr) 3340, 2830, 1590, 1345, 875,
820 cmꢀ1. ESR: a five-line pattern with intensity ratios
of 1:2:3:2:1, an = 8.29 G, g = 2.00990.
5,5-tetramethylimidazoline (2d). Using the general proce-
dure mentioned above 103 mg (52%) of the title com-
pound was obtained. Mp 134–135 ꢁC. Rf = 0.13
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 249[M]+. IR
(KBr) 3250, 1500, 1490, 1385, 882, 843, 690 cmꢀ1
.
ESR: a five-line pattern with intensity ratios of
1:2:3:2:1, aN = 8.18 G, g = 2.00994.
3.2.4.12. 1-Oxyl-2-(40-N,N-dimethylphen-10-yl)-3-
oxide-4,4,5,5-tetramethylimidazoline (2l). Using the
general procedure mentioned above 147 mg (67%) of
the title compound was obtained. Mp 151–154 ꢁC.
Rf = 0.51 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
276[M]+. IR (KBr) 2850, 1595, 1500, 1360,
845 cmꢀ1. ESR: a five-line pattern with intensity
ratios of 1:2:3:2:1, aN = 8.27 G, g = 2.00992.
3.2.4.5. 1-Oxyl-2-(40-methylphen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2e). Using the general
procedure mentioned above 157 mg (80%) of the title
compound was obtained. Mp 86–88 ꢁC. Rf = 0.88
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 247[M]+. IR
(KBr) 2980, 1610, 1500, 1365, 810 cmꢀ1. ESR: a
five-line pattern with intensity ratios of 1:2:3:2:1,
aN = 8.22 G, g = 2.00993.
3.2.4.13. 1-Oxyl-2-(20-fluorophen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2m). Using the general
procedure mentioned above 147 mg (67%) of the title
compound was obtained. Mp 112–114 ꢁC. Rf = 0.52
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 251[M]+. IR
(KBr) 1610, 1450, 1370, 1135, 770 cmꢀ1. Anal. Calcd
for C13H16N2O4 C 62.14, H 6.42, N 11.15. Found C
62.23, H 6.58, N 11.09. ESR: a five-line pattern with
intensity ratios of 1:2:3:2:1, aN = 8.15 G, g = 2.00996.
3.2.4.6. 1-Oxyl-2-(40-methoxyphen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2f). Using the general proce-
dure mentioned above 203 mg (97%) of the title com-
pound was obtained. Mp 89–92 ꢁC. Rf = 0.88 (CHCl3/
CH3OH, 20:1). EI-MS (m/z) 263[M]+. IR (KBr) 2830,
1600, 1355, 835 cmꢀ1. ESR: a five-line pattern with
intensity ratios of 1:2:3:2:1, aN = 8.28 G, g = 2.00991.
3.2.4.7. 1-Oxyl-2-(20,40-dimethoxyphen-10-yl)-3-oxide-
4,4,5,5-tetramethylimidazoline (2g). Using the general
procedure mentioned above 199 mg (85%) of the title
compound was obtained. Mp 75–77 ꢁC. Rf = 0.37
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 293[M]+. IR
(KBr) 2830, 1610, 1365, 805, 870 cmꢀ1. ESR: a five-line
pattern with intensity ratios of 1:2:3:2:1, aN = 8.32 G,
g = 2.00989.
3.2.4.14. 1-Oxyl-2-(40-chlorophen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2n). Using the general proce-
dure mentioned above 240 mg (97%) of the title com-
pound was obtained. Mp 103–105 ꢁC. Rf = 0.67
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 267[M]+. IR
(KBr) 1590, 1500, 1365, 1090, 825 cmꢀ1. ESR: a five-line
pattern with intensity ratios of 1:2:3:2:1, aN = 8.12 G,
g = 2.00997.
3.2.4.8. 1-Oxyl-2-(30,40-dimethoxyphen-10-yl)-3-oxide-
4,4,5,5-tetramethylimidazoline (2h). Using the general
procedure mentioned above 194 mg (83%) of the title
compound was obtained. Mp 95–97 ꢁC. Rf = 0.34
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 293[M]+. IR
(KBr) 2830, 1595, 1355, 870, 820 cmꢀ1. ESR: a five-line
pattern with intensity ratios of 1:2:3:2:1, aN = 8.28 G,
g = 2.00991.
3.2.4.15. 1-Oxyl-2-(40-bromophen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2o). Using the general proce-
dure mentioned above 231 mg (93%) of the title com-
pound was obtained. Mp 89–91 ꢁC. Rf = 0.85 (CHCl3/
CH3OH, 10:1). EI-MS (m/z) 311[M]+. IR (KBr) 3310,
1600, 1450, 1070, 825 cmꢀ1. ESR: a five-line pattern
with intensity ratios of 1:2:3:2:1, aN = 8.12 G,
g = 2.00997.
3.2.4.9. 1-Oxyl-2-(30,40-methylendioxyphen-10-yl)-3-
oxide-4,4,5,5-tetramethylimidazoline (2i). Using the gen-
eral procedure mentioned above 185 mg (84%) of the ti-
tle compound was obtained. Mp 97–99 ꢁC. Rf = 0.75
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 277[M]+. IR
(KBr) 1600, 1365, 910, 1240, 1100 cmꢀ1. ESR: a five-line
pattern with intensity ratios of 1:2:3:2:1, an = 8.28 G,
g = 2.00990.
3.2.4.16. 1-Oxyl-2-(20,40-dichlorophen-10-yl)-3-oxide-
4,4,5,5-tetramethylimidazoline (2p). Using the general
procedure mentioned above 212 mg (88%) of the title
compound was obtained. Mp 123–125 ꢁC. Rf = 0.54
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 302[M]+. IR
(KBr) 1590, 1450, 1365, 1000, 870, 825 cmꢀ1. ESR: a
five-line pattern with intensity ratios of 1:2:3:2:1,
aN = 8.15 G, g = 2.00997.
3.2.4.10. 1-Oxyl-2-(30-hydroxy-40-methoxyphen-10-yl)-
3-oxide-4,4,5,5-tetramethylimidazoline (2j). Using the
general procedure mentioned above 151 mg (68%) of
the title compound was obtained. Mp 117–119 ꢁC.
Rf = 0.34 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
279[M]+. IR (KBr) 3320, 2830, 1595, 1500, 1350,
820 cmꢀ1. ESR: a five-line pattern with intensity ratios
of 1:2:3:2:1, an = 8.25 G, g = 2.00991.
3.2.4.17. 1-Oxyl-2-(30,40-dichlorophen-10-yl)-3-oxide-
4,4,5,5-tetramethylimidazoline (2q). Using the general
procedure mentioned above 214 mg (89%) of the title
compound was obtained. Mp 120–122 ꢁC. Rf = 0.54
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 302[M]+. IR
(KBr) 1593, 1452, 1363, 1004, 871, 822 cmꢀ1. ESR: a
five-line pattern with intensity ratios of 1:2:3:2:1,
aN = 8.15 G, g = 2.00997.

M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
2823
3.2.4.18. 1-Oxyl-2-(20-nitrophen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2r). Using the general
procedure mentioned above 161 mg (73%) of the title
compound was obtained. Mp 145–147 ꢁC. Rf = 0.49
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 278[M]+. IR
(KBr) 1530, 1360, 1450, 740 cmꢀ1. ESR: a five-line
pattern with intensity ratios of 1:2:3:2:1,
aN = 8.14 G, g = 2.00997.
dure mentioned above 182 mg (78%) of the title com-
pound was obtained as red syrup. Rf = 0.52 (CHCl3/
CH3OH, 20:1). EI-MS (m/z) 233[M]+. IR (KBr)
3338, 1588, 1451, 1380, 880, 790 cmꢀ1. Anal. Calcd
for C13H17N2O2: C, 66.93; H, 7.35; N, 12.01. Found:
C, 66.76; H 7.52; N, 12.24.
3.2.5.4. 1-Oxyl-2-(40-hydroxylphen-10-yl)-4,4,5,5-
tetramethylimidazoline (3d). Using the general proce-
dure mentioned above 150 mg (65%) of the title com-
pound was obtained as yellow powder. Mp 98–100 ꢁC.
Rf = 0.36 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
233[M]+. IR (KBr) 3259, 1595, 1492, 1382, 881 845,
693 cmꢀ1. Anal. Calcd for C13H17N2O2: C, 66.93; H,
7.35; N, 12.01. Found: C, 66.77; H 7.54; N, 12.25.
3.2.4.19. 1-Oxyl-2-(30-nitrophen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2s). Using the general
procedure mentioned above 173 mg (78%) of the title
compound was obtained. Mp 161–162 ꢁC. Rf = 0.69
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 278[M]+. IR
(KBr) 1530, 1350, 1600, 1450, 880, 790, 680 cmꢀ1
.
ESR: a five-line pattern with intensity ratios of
1:2:3:2:1, aN = 8.10 G, g = 2.00998.
3.2.5.5. 1-Oxyl-2-(40-methylphen-10-yl)-4,4,5,5-tetra-
methylimidazoline (3e). Using the general procedure men-
tioned above 200 mg (86%) of the title compound was ob-
tained as yellow powder. Mp 156–157 ꢁC. Rf = 0.93
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 231[M]+. IR (KBr)
2985, 1606, 1503, 1368, 804 cmꢀ1. Anal. Calcd for
C14H19N2O: C, 72.69; H, 8.28; N, 12.11. Found: C,
72.80; H 8.42; N, 12.32.
3.2.4.20. 1-Oxyl-2-(40-nitrophen-10-yl)-3-oxide-4,4,
5,5-tetramethylimidazoline (2t). Using the general
procedure mentioned above 162 mg (73%) of the title
compound was obtained. Mp 175–176 ꢁC. Rf = 0.65
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 278[M]+. IR
(KBr) 1530, 1360, 1603, 1500, 1450, 830, 790,
680 cmꢀ1. ESR: a five-line pattern with intensity
ratios of 1:2:3:2:1, aN = 8.10 G, g = 2.00998.
3.2.5.6. 1-Oxyl-2-(40-methoxyphen-10-yl)-4,4,5,5-tet-
ra-methylimidazoline (3f). Using the general procedure
mentioned above 212 mg (86%) of the title compound
was obtained as red powder. Mp 42–43 ꢁC. Rf = 0.63
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 247[M]+. IR
(KBr) 2830, 1610, 1352, 840 cmꢀ1. Anal. Calcd for
C14H19N2O2: C, 67.99; H, 7.74; N, 11.33. Found: C,
67.80; H 7.55; N, 11.55.
3.2.5. General procedure for the preparation of 1-oxyl-2-
substitutedphenyl-4,4,5,5-tetramethyl-imidazolines (3a–t).
The solution of 1.0 mmol of 2-substitutedphenyl-4,4,5,5-
tetramethylimidazoline-3-oxide-1-oxyl and 500 mg
NaNO2 in 15 ml of methanol was adjusted to pH 5 with
hydrochloric acid. The reaction mixture was stirred at
room temperature for 1 h and TLC (CHCl3/CH3OH,
20:1) indicated the complete disappearance of 2-substi-
tutedphenyl-4,4,5,5-tetramethylimidazoline-3-oxide-1-
oxyl. The reaction mixture was evaporated under
vacuum to remove most of the solvent, adjusted to pH
6–7 with aqueous sodium bicarbonate, and extracted
with chloroform. The combined chloroform phase was
dried by anhydrous sodium sulfate and filtered. The
filtrate was evaporated under vacuum to provide the
title compound.
3.2.5.7. 1-Oxyl-2-(20,40-dimethoxylphen-10-yl)-4,4,5,5-
tetramethylimidazoline (3g). Using the general procedure
mentioned above 177 mg (64%) of the title compound
was obtained as yellow powder. Mp 77–79 ꢁC. Rf = 0.41
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 277[M]+. IR (KBr)
2850, 1610, 1590, 1500, 805, 830 cmꢀ1. Anal. Calcd for
C15H21N2O3: C, 64.96; H, 7.63; N, 10.10. Found: C,
64.78; H 7.49; N, 10.22.
3.2.5.8. 1-Oxyl-2-(30,40-dimethoxylphen-10-yl)-4,4,5,5-
tetramethylimidazoline (3h). Using the general procedure
mentioned above 138 mg (50%) of the title compound
was obtained as yellow powder. Mp 74–75 ꢁC. Rf = 0.43
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 277[M]+. IR (KBr)
2833, 1592, 1351, 874, 822 cmꢀ1. Anal. Calcd for
C15H21N2O3: C, 64.96; H, 7.63; N, 10.10. Found: C,
64.80; H 7.50; N, 10.41.
3.2.5.1. 1-Oxyl-2-(phen-10-yl)-4,4,5,5-tetramethylimi-
dazoline (3a). Using the general procedure mentioned
above 171 mg (79%) of the title compound was obtained
as red syrup. Rf = 0.61 (CHCl3/CH3OH, 20:1). EI-MS
(m/z) 217[M]+. IR (KBr) 1602, 1455, 1353, 1072,
822 cmꢀ1. Anal. Calcd for C13H17N2O3: C, 71.86; H,
7.89; N, 14.89. Found: C, 71.69; H, 7.75; N, 12.99.
3.2.5.2. 1-Oxyl-2-(20-hydroxylphen-10-yl)-4,4,5,5-
tetramethylimidazoline (3b). Using the general proce-
dure mentioned above 177 mg (76%) of the title
compound was obtained as red syrup. Rf = 0.83
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 233[M]+. IR
(KBr) 1610, 1503, 1450, 760 cmꢀ1. Anal. Calcd for
C13H17N2O2: C, 66.93; H, 7.35; N, 12.01. Found: C,
66.77; H 7.11; N, 12.33.
3.2.5.9. 1-Oxyl-2-(30,40-methylendioxyphen-10-yl)-
4,4,5,5-tetramethylimidazoline (3i). Using the general
procedure mentioned above 230 mg (81%) of the title
compound was obtained as red powder. Mp 46–
47 ꢁC. Rf = 0.82 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
261[M]+. IR (KBr) 3420, 1600, 1450, 1365, 1260,
935 cmꢀ1. Anal. Calcd for C14H17N2O3: C, 64.35; H,
6.56; N, 10.72. Found: C, 64.17; H 6.44; N, 10.91.
3.2.5.3. 1-Oxyl-2-(30-hydroxylphen-10-yl)-4,4,5,5-
tetramethylimidazoline (3c). Using the general proce-
3.2.5.10. 1-Oxyl-2-(30-hydroxy-40-methoxyphen-10-yl)-
4,4,5,5-tetramethylimidazoline (3j). Using the general

2824
M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
procedure mentioned above 118 mg (45%) of the title
compound was obtained as yellow powder. Mp 70–
71 ꢁC. Rf = 0.45 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
263[M]+. IR (KBr) 3327, 2831, 1605, 1502, 1351,
818 cmꢀ1. Anal. Calcd for C14H19N2O3: C, 63.86; H,
7.27; N, 10.64. Found: C, 63.69; H 7.15; N, 10.81.
mentioned above 215 mg (75%) of the title compound
was obtained as red powder. Mp 66–67 ꢁC. Rf = 0.71
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 286[M]+. IR
(KBr) 3348, 1597, 1456, 1369, 1005, 873, 820 cmꢀ1
.
Anal. Calcd for C13H16N2Ocl2 C 54.56, H 5.28, N
9.79. Found C 54.43, H 5.20, N 9.68.
3.2.5.11. 1-Oxyl-2-(30-methoxyl-40-hydroxyphen-10-
yl)-4,4,5,5-tetramethylimidazoline (3k). Using the gen-
eral procedure mentioned above 106 mg (40%) of the
title compound was obtained as yellow syrup.
Rf = 0.50 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
263[M]+. IR (KBr) 3337, 2835, 1596, 1342, 878,
817 cmꢀ1. Anal. Calcd for C14H19N2O3: C, 63.86; H,
7.27; N, 10.64. Found: C, 63.71; H 7.35; N, 10.82.
3.2.5.18. 1-Oxyl-2-(20-nitrophen-10-yl)-4,4,5,5-tetra-
methylimidazoline (3r). Using the general procedure
mentioned above 189 mg (72%) of the title compound
was obtained as yellow powder. Mp 74–75 ꢁC.
Rf = 0.52 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
262[M]+. IR (KBr) 3343, 1604, 1500, 1452, 1361,
740 cmꢀ1. Anal. Calcd for C13H16N3O3 C 59.53, H
6.15, N 16.02. Found C 59.69, H 6.30, N 16.18.
3.2.5.12. 1-Oxyl-2-(40-N,N-dimethylphen-10-yl)-4,4,
5,5-tetramethylimidazoline (3l). Using the general pro-
cedure mentioned above 123 mg (47%) of the title
compound was obtained as yellow powder. Mp 85–
86 ꢁC. Rf = 0.67 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
260[M]+. IR (KBr) 3341, 2855, 1598, 1502, 1367,
841 cmꢀ1. Anal. Calcd for C15H22N3O: C, 69.20; H,
8.52; N, 16.14. Found: C, 69.38; H 8.45; N, 16.22.
3.2.5.19. 1-Oxyl-2-(30-nitrophen-10-yl)-4,4,5,5-tetra-
methylimidazoline (3s). Using the general procedure
mentioned above 191 mg (73%) of the title compound
was obtained as yellow powder. Mp 71–72 ꢁC.
Rf = 0.78 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
262[M]+. IR (KBr) 3348, 1604, 1533, 1452, 1359, 810,
791, 684 cmꢀ1. Anal. Calcd for C13H16N3O3 C 59.53,
H 6.15, N 16.02. Found C 59.66, H 6.28, N 16.16.
3.2.5.13. 1-Oxyl-2-(20-fluorophen-10-yl)-4,4,5,5-tetra-
methylimidazoline (3m). Using the general procedure
mentioned above 147 mg (62%) of the title compound
was obtained as red syrup. Rf = 0.69 (CHCl3/CH3OH,
20:1). EI-MS (m/z) 235[M]+. IR (KBr) 3345, 2850,
1611, 1455, 1376, 1130, 772 cmꢀ1. Anal. Calcd for
C13H16N2OF C 66.36, H 6.85, N 11.91. Found C
66.25, H 6.68, N 11.74.
3.2.5.20. 1-Oxyl-2-(40-nitrophen-10-yl)-4,4,5,5-tetra-
methylimidazoline (3t). Using the general procedure
mentioned above 186 mg (71%) of the title compound
was obtained as red powder. Mp 119–120 ꢁC.
Rf = 0.74 (CHCl3/CH3OH, 20:1). EI-MS (m/z)
262[M]+. IR (KBr) 3340, 1604, 1520, 1500, 1452, 860,
793, 681 cmꢀ1. Anal. Calcd for C13H16N3O3 C 59.53,
H 6.15, N 16.02. Found C 59.40, H 6.08, N 15.97.
3.2.5.14. 1-Oxyl-2-(40-chlorophen-10-yl)-4,4,5,5-tetra-
methylimidazoline (3n). Using the general procedure
mentioned above 210 mg (83%) of the title compound
was obtained as red syrup. Rf = 0.73 (CHCl3/CH3OH,
20:1). EI-MS (m/z) 251[M]+. IR (KBr) 3350, 1598,
3.3. Pain threshold in vivo assay
Male ICR mice weighing 25 2 g were housed in a 12/
12 light/dark cycle at a room temperature of 21 2 ꢁC
for one day before being used. Each of them in drug
receiving groups was given a single intraperitoneal (ip)
injection of one of 3a–t at a level of 0.13 mmol/kg in
0.2 ml of saline mixed with 0.01 ml DMSO and
0.02 ml Tween 80. The mouse in control group was
given the same mixed solvent as the drug receiving
groups received. Analgesic effects of 3a–t were evaluated
by the tail flick test. The basic pain threshold value of
each mouse was measured for 4 times.
1501, 1369, 1092, 821 cmꢀ1
.
Anal. Calcd for
C13H16N2OCl C 62.03, H 6.41, N 11.13. Found C
62.14, H 6.60, N 11.24.
3.2.5.15. 1-Oxyl-2-(40-bromophen-10-yl)-4,4,5,5-tetra-
methylimidazoline (3o). Using the general procedure
mentioned above 290 mg (98%) of the title compound
was obtained as red powder. Mp 91–92 ꢁC. Rf = 0.95
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 296[M]+. IR
(KBr) 3351, 1590, 1452, 1480, 1369, 840 cmꢀ1. Anal.
Calcd for C15H16N2OBr C 52.72, H 5.44, N 9.46. Found
C 52.52, H 5.67, N 9.67.
After 3a–t were administered, the pain thresholds were
measured at 30 min intervals. This measurement was car-
ried out for 150 min. The potency of analgesia was indi-
cated by the pain threshold variation. The values were
calculated according to equation PTV = AAPT/BPT
wherein PTV = pain threshold variation, BPT = basic
pain threshold andAAPT = pain threshold after adminis-
tration-basic pain threshold. All values of the pain thresh-
old variation for each animal were averaged and
constituted one sample.
3.2.5.16. 1-Oxyl-2-(20,40-dichlorophen-10-yl)-4,4,5,5-
tetramethylimidazoline (3p). Using the general procedure
mentioned above 255 mg (90%) of the title compound
was obtained as red powder. Mp 96–97 ꢁC. Rf = 0.66
(CHCl3/CH3OH, 20:1). EI-MS (m/z) 286[M]+. IR
(KBr) 3351, 1598, 1459, 1362, 873, 824 cmꢀ1. Anal.
Calcd for C13H16N2Ocl2 C 54.56, H 5.28, N 9.79. Found
C 54.45, H 5.22, N 9.64.
3.4. Tail bleeding time in vivo assay
3.2.5.17. 1-Oxyl-2-(30,40-dichlorophen-10-yl)-4,4,5,5-
tetramethylimidazoline (3q). Using the general procedure
Male ICR mice weighing 25 2 g were housed in a 12/
12 light/dark cycle at a room temperature of 21 2 ꢁC

M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
2825
for one day before being used. Each of them in drug
receiving groups was given a single intraperitoneal (ip)
injection of one of 3a–t at a level of 1.30 mmol/kg in
0.2 ml of saline mixed with 0.01 ml DMSO and
0.02 ml Tween 80. The mouse in control group was gi-
ven the same mixed solvent as the drug receiving groups
received. At 5, 15, 30, and 60 min, a mouse was placed in
a tube holder with its tail protruding, and a 2 mm cut
was made on the tail. Flowing blood until it stopped
was gently wiped away with a filter paper every 30 s,
yielding the bleeding time (BT). The assay reported by
Dejana et al.15 was employed.
sion equation, two-step descriptor selection procedure
was applied. The first step consisted in the elimination
of the descriptors that are not able to provide useful sta-
tistical information. For instance, the absolute value of
the correlation coefficient between the selected descrip-
tor (Xi) and the activity (PTV90) under investigation
(jcor(Xi,PTV90)j) should be larger than 0.3. At the sec-
ond step, the descriptor subset was optimized by means
of a genetic algorithm.27 Finally, the optimal QSAR
equation describing the analgesic effects of 3a–t was
determined. The definitions of the selected molecular
descriptors in the final QSAR equation are summarized
in Table 5.
3.5. Vasorelaxation in vitro assay of 3a–t
To systematically assess a QSAR model, a reliable vali-
dation is required. Usually, a QSAR model is evaluated
by the predictive results for a training dataset (resubsti-
tution test) and a testing dataset (cross-validation),
respectively. By the test of resubstitution, the PTV90 va-
lue of each compound in a training data set is predicted
using the rules derived from the same set. Although this
test gives somewhat optimistic error estimate because
the same compounds are used to derive the QSAR equa-
tion and to predict themselves, the resubstitution test is
absolutely necessary due to its ability to reflect the self-
consistency of a given method. On the other hand, a
cross-validation test for an independent testing data
set is also needed because it can reflect the real effective-
ness of a QSAR model. Of the various cross-validation
tests, the LOO test is thought to be a reliable one. By
the LOO test, the PTV90 value of each compound in
the data set is predicted by the rules derived using all
other compounds except the one that is being predicted.
Both tests of resubstitution and LOO were used to eval-
uate the QSAR models in the current work.
The conversion of 3a–t during vasoconstriction induced
by NE was examined using the rat aortic strip according
to a published method.13,14 In brief, immediately after
decapitation, rat aortic strips were obtained and put in
a perfusion bath with 5 ml of warmed (37 ꢁC), oxygen-
ated (95%O2, 5%CO2) Krebs solution (pH 7.4). The aor-
tic strips were mounted to the tension transducers and
the relaxation-contraction curves were recorded. Nor-
adrenaline (NE, final concentration 10ꢀ9 mol/l) solution
was added to induce contraction. When the hypertonic
contraction reached to the maximum level, NE was
washed and the vessel strips were stabilized for 30 min.
After the renewal of the solution, NE (final concentra-
tion 10ꢀ9 mol/l), was added. When the hypertonic con-
traction value of aortic strips reached the peak, 15 ll
of alcohol (vehicle), or the solution of 3a–t in 15 ll of
alcohol (final concentration 5 · 10ꢀ5 mol/l) were added,
respectively, to observe the relaxant response of the
aortic strip to them. The percentage inhibitions of
NE-induced vasoconstriction by 3a–t were recorded.
3.6. QSAR analysis
In the resubstitution test, three statistical parameters (R,
S, and ꢀe) were used to evaluate the performance. The
regression coefficient R is defined as follows:
To perform a QSAR analysis, first the SMILES formats
of compounds 3a–t were generated using JME server
(http://www.molinspiration.com/cgi-bin/properties).
Moreover, the obtained SMILES files were submitted to
e-dragon server (http://www.vcclab.org/lab/edragon/) to
calculate molecular descriptors. Finally, the 1664 differ-
ent molecular parameters for each compound were
obtained.
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
X
N
N
2
2
R ¼ 1 ꢀ
ðyi ꢀ ^yiÞ =
ðyi ꢀ ꢀyÞ
ð2Þ
i¼1
i¼1
where yi, ^yi, and ꢀy are predicted, actual, and mean values
of the activity (i.e., PTV90), respectively. The N is the
number of compounds in the resubstitution test. The
root mean square of errors S was calculated with the
following equation:
The multiple linear regression method was employed to
derive the QSAR equation of 3a–t. Concerning the anal-
gesic effect of a molecule, the PTV value measured at
90 min (i.e., PTV90) was considered in the current QSAR
analysis. As the number of compounds is only 20, the
optimum number of molecular descriptors used in the
QSAR analysis is around 4. To obtain the best regres-
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
To illustrate the predictive accuracy more explicitly, the
N
i¼1
2
ðyi ꢀ ^yiÞ
S ¼
ð3Þ
N
ꢀ
e (absolute average error) is defined as follows:
Table 5. The four molecular descriptors used in the current QSAR analysisa
Abbreviation
Type
Definition
GATS4e
BELe1
JGI1
2D autocorrelation indices
Burden eigenvalue descriptors
Topological charge indices
GETAWAY
Geary autocorrelation ꢀ lag 4/weighted by atomic Sanderson electronegativities
Lowest eigenvalue n. 1 of Burden matrix/weighted by atomic Sanderson electronegativities
Mean topological charge index of order1
R2u
R autocorrelation of lag 2/unweighted
a The definitions of these four descriptors were downloaded from the website of e-dragon (http://www.vcclab.org/lab/edragon/).

2826
M. Zhao et al. / Bioorg. Med. Chem. 15 (2007) 2815–2826
10. Murata, T.; Ushikubi, F.; Matsuoka, T.; Hirata, M.;
P
In the LOO test, the above three statistical parameters,
N
i¼1
jyi ꢀ ^yij
ꢀe ¼
ð4Þ
Yamasaki, A.; Sugimoto, Y.; Ichikawa, A.; Aze, Y.;
Tanaka, T.; Yoshida, N.; Ueno, A.; Oh-ishi, S.; Nar-
umiya, S. Nature 1997, 388, 678.
N
11. Tsunoda, R.; Okumura, K.; Ishizaka, H.; Matsunaga, T.;
Tabuchi, T.; Yasue, H.; Akaike, T.; Sato, K.; Maeda, H.
Eur. J. Pharmacol. 1994, 262, 55.
that is, RLOO,SLOO, and ꢀeLOO, were also calculated.
12. Akaike, T.; Yoshida, M.; Miyamoto, Y.; Sato, K.; Kohno,
M.; Sasamoto, K.; Miyazaki, K.; Ueda, S.; Maeda, H.
Biochemistry 1993, 32, 827.
13. Vocci, F. J.; Petty, S. K.; Dewey, W. L. J. Pharmacol. Exp.
Therap. 1978, 207, 892.
14. Medakovic, M.; Banic, B. J. Pharm. Pharmacol. 1964, 16,
198.
15. Dejana, E.; Villa, S.; De Gaetano, G. Thromb. Haemost.
1982, 48, 108.
Acknowledgments
This work was supported by Beijing area major labora-
tory of peptide and small molecular drugs, National
Natural Scientific Foundation of China (30472071),
973 Project of China (2006CB708501) and Natural Sci-
entific Foundation of Beijing (7052010). Z. Z. was sup-
ported by the talent people’s foundation of China
Agricultural University.
16. Cowart, M.; Kowaluk, E. A.; Daanen, J. F.; Kohlhaas, K.
L.; Alexander, K. M.; Wagenaar, F. L.; Kerwin, J. F.
J. Med. Chem. 1998, 41, 2636.
17. Wu, Y.; Bi, L.; Bi, W.; Li, Z.; Zhao, M.; Wang, C.; Ju, J.;
Peng, S. Bioorg. Med. Chem. 2006, 14, 5711.
18. Tetko, I. V.; Gasteiger, J.; Todeschini, R.; Mauri, A.;
Livingstone, D.; Ertl, P.; Palyulin, V. A.; Radchenko, E.
V.; Zefirov, N. S.; Makarenko, A. S.; Tanchuk, V. Y.;
Prokopenko, V. V. J. Comput. Aided Mol. Des. 2005, 19,
453.
19. Geary, R. C. Incorp. Stat. 1954, 5, 115.
20. Burden, F. R. J. Chem. Inf. Comput. Sci. 1989, 29,
225.
References and notes
1. Ralevic, V.; Burnstock, G. Pharmacol. Rev. 1998, 50, 413.
2. Muller, C. E. Mini-Rev. Med. Chem. 2001, 1, 417.
3. Ozola, V.; Thorand, M.; Diekmann, M.; Qurishi, R.;
Schumacher, B.; Jacobsonb, A. K.; Muller, C. E. Bioorg.
Med. Chem. 2003, 11, 347.
4. Ruffolo, R. R.; Bondinell, W. E.; Hieble, J. P. J. Med.
Chem. 1995, 38, 3681.
5. Ruffolo, R. R.; Waddell, J. E. J. Pharmacol. Exp. Ther.
1983, 224, 559.
6. Pigini, M.; Quaglia, W.; Gentili, F.; Marucci, G.; Cant-
alamessa, F.; Franchini, S.; Sorbic, C.; Brasili, L. Bioorg.
Med. Chem. 2000, 8, 883.
21. Galvez, J.; Garcia-Domenech, R.; De Julian-Ortiz,
V.; Soler, R. J. Chem. Inf. Comput. Sci. 1994, 34,
1198.
22. Consonni, V.; Todeschini, R.; Pavan, M. J. Chem. Inf.
Comp. Sci. 2002, 42, 682.
7. Hong, S.-S.; Bavadekar, S. A.; Lee, S.-I.; Patil, P. N.;
Lalchandani, S. G.; Feller, D. R.; Miller, D. D. Bioorg.
Med. Chem. Lett. 2005, 15, 4691.
23. Doytchinova, I.; Natcheva, R. Acta Pharm. 1997, 47, 189.
24. Wang, X.-H.; Tang, Y.; Xie, Q.; Qiu, Z.-B. Eur. J. Med.
Chem. 2006, 41, 226.
8. Merriman, G. H.; Ma, L.; Shum, P.; McGarry, D.; Volz,
F.; Sabol, J. S.; Gross, A.; Zhao, Z.; Rampe, D.; Wang,
L.; Wirtz-Brugger, F.; Harris, B. A.; Macdonald, D.
Bioorg. Med. Chem. Lett. 2005, 15, 435.
25. Mathur, K. C.; Gupta, S.; Khadikar, P. V. Bioorg. Med.
Chem. 2003, 11, 1915.
26. Cramer, R. D., III; Patterson, D. E.; Bunce, J. D. Prog.
Clin. Biol. Res. 1989, 291, 161.
9. Clark, R. D.; Jahangir, A.; Severance, D.; Salazar, R.;
Chang, T.; Chang, D.; Jett, M. F.; Smith, S.; Bley, K.
Bioorg. Med. Chem. Lett. 2004, 14, 1053.
27. Zhao, M.; Li, Z.; Wu, Y.; Tang, Y. -R.; Wang, C.; Zhang,
Z.; Peng, S. Eur. J. Med. Chem. 2007, doi:10.1016/
j.ejmech.2006.12.027.