Structural modifications of benzanilide derivatives, effective potassium channel openers. X.

Article abstract of DOI:10.1016/j.ejmech.2006.07.016

Large-conductance calcium-activated potassium (BK) channels are involved in many fundamental cell functions. Consistently, the ability to activate BK channels by exogenous compounds is considered as a promising pharmacodynamic pattern for the potential treatment of several pathologies. In this perspective, the development of new and selective BK-openers can be considered as an actual field of research. This paper reports the synthesis and pharmacological evaluation of new benzanilides, useful for deepening the comprehension of the structure-activity relationships, emerged in previous studies on this class of BK-activators. From a structural point of view, these benzanilides belong to a general class of BK-activators, showing a common pharmacophoric model, consisting of two aryl groups linked through an appropriate "spacer" and the almost obligatory presence of a phenolic hydroxyl. In particular, a new series of benzanilides, in which the phenyl rings have been widely changed both on the acidic portion and the basic one of the amide spacer, were synthesised. Their vasorelaxing effects, induced through the activation of BK channels, were also evaluated. Although many compounds exhibited effects which could not be attributed to the activation of BK channels, two derivatives showed a clear profile of BK-activators with vasodilator activity comparable to or slightly lower than that recorded for the reference benzimidazolone NS1619. A further molecular modelling approach allowed us to obtain a molecular electrostatic potential feature which suggests a suitable interaction with the receptor site of the BK channel, from a tri-dimensional point of view. This approach seems to represent a further contribution for the development of new BK-activators, designed on the basis of the pharmacophoric model above-mentioned.

Full text of DOI:10.1016/j.ejmech.2006.07.016

European Journal of Medicinal Chemistry 41 (2006) 1421e1429
http://www.elsevier.com/locate/ejmech
Original article
Structural modifications of benzanilide derivatives, effective
potassium channel openers. X.
b
b
b
Vincenzo Calderone a, , Alessio Coi , Francesca Lidia Fiamingo , Irene Giorgi ,
*
Michele Leonardi b, Oreste Livi b, Alma Martelli a, Enrica Martinotti a
a Dipartimento di Psichiatria, Neurobiologia, Farmacologia e Biotecnologie, Universita` di Pisa, via Bonanno 6, I-56126 Pisa, Italy
b Dipartimento di Scienze Farmaceutiche, Universita` di Pisa, via Bonanno 6, I-56126 Pisa, Italy
Received 12 May 2006; received in revised form 29 June 2006; accepted 12 July 2006
Available online 6 October 2006
Abstract
Large-conductance calcium-activated potassium (BK) channels are involved in many fundamental cell functions. Consistently, the ability to
activate BK channels by exogenous compounds is considered as a promising pharmacodynamic pattern for the potential treatment of several
pathologies. In this perspective, the development of new and selective BK-openers can be considered as an actual field of research. This paper
reports the synthesis and pharmacological evaluation of new benzanilides, useful for deepening the comprehension of the structureeactivity
relationships, emerged in previous studies on this class of BK-activators. From a structural point of view, these benzanilides belong to a general
class of BK-activators, showing a common pharmacophoric model, consisting of two aryl groups linked through an appropriate ‘‘spacer’’ and the
almost obligatory presence of a phenolic hydroxyl. In particular, a new series of benzanilides, in which the phenyl rings have been widely
changed both on the acidic portion and the basic one of the amide spacer, were synthesised. Their vasorelaxing effects, induced through the
activation of BK channels, were also evaluated. Although many compounds exhibited effects which could not be attributed to the activation
of BK channels, two derivatives showed a clear profile of BK-activators with vasodilator activity comparable to or slightly lower than that
recorded for the reference benzimidazolone NS1619. A further molecular modelling approach allowed us to obtain a molecular electrostatic
potential feature which suggests a suitable interaction with the receptor site of the BK channel, from a tri-dimensional point of view. This
approach seems to represent a further contribution for the development of new BK-activators, designed on the basis of the pharmacophoric
model above-mentioned.
Ó 2006 Elsevier Masson SAS. All rights reserved.
Keywords: BK channel; Benzanilide; BK-activator; Molecular modelling
1. Introduction
neurotransmitters; also, BK channels play a fundamental
role in the regulation of the tone of smooth muscle cells [1,2].
The physiological activation of BK channels, induced
mainly by two triggering signals, such as the rise of intracel-
lular free calcium ions and membrane depolarisation, ensures
the massive flow of potassium ions (with a single channel con-
ductance of 150e300 pS) to the extracellular side of the plas-
malemma, membrane hyperpolarisation and reduction of the
cellular excitability. Conversely, the availability of exogenous
compounds capable of activating BK channels can guarantee
an innovative pharmacological tool for the clinical manage-
ment of many pathological states, due to cell hyperexcitability,
The large-conductance calcium-activated potassium (BK)
channels are expressed in excitable as well as in non-excitable
cells. They control several cell functions: in the nervous sys-
tem, BK channels contribute to the shaping of action potential
and modulate the neuronal excitability and the release of
* Corresponding author. Tel.: þ39 50 2219589.
E-mail address: calderone@farm.unipi.it (V. Calderone).
0223-5234/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2006.07.016

1422
V. Calderone et al. / European Journal of Medicinal Chemistry 41 (2006) 1421e1429
such as asthma, urge incontinence and bladder spasm, gastric
hypermotility, neurological and psychiatric disorders [1,2].
As concerns the cardiovascular system, it is now widely
accepted that BK channels ensure the predominant component
of the outward Kþ current in vascular smooth muscle cells,
accounting for the fundamental function of such ion channels
in the modulation of the muscular tone of vessels [3,4]. Con-
sequently, the vasorelaxing effects of exogenous BK-openers
can furnish the pharmacological rational basis for the treat-
ment of hypertension and/or other diseases related to an
impaired contractility of vessels (for example, coronary vaso-
spasm) [1,2].
In a previous work [5], we could observe that the synthes-
ised 1-(20-hydroxybenzoyl)-5-methyl-benzotriazole, showing
structural analogies with the reference BK-openers NS004
and NS1619 (Fig. 1A) and exhibiting vasorelaxing effects
probably due to the activation of vascular BK channels, was
able to confer significant protection to the myocardial func-
tion, in isolated rat hearts submitted to ischemia/reperfusion
cycles. This result (originally unexpected) can be now ex-
plained thanks to more recent experimental evidence showing
that the activation of cardiac calcium-activated potassium
channels could be involved in the cardioprotective mecha-
nisms of ‘‘ischemic preconditioning’’ and that the administra-
tion of BK-openers, such as NS1619, could reduce cardiac
injury following an ischemic event [6e8]. Of course, these
reports let us foresee a further potential use of BK-activators
in cardiovascular pharmacotherapy, as anti-ischemic drugs.
Some years ago, we undertook a research program concern-
ing the synthesis and pharmacological experimentation of new
compounds such as BK channel openers. As a consequence
1,2,3-triazole derivatives [9e12], benzimidazoles [13,14] and
benzotriazoles [5,13,14] were tested, providing good and
encouraging results. High pharmacological activity was also
detected in some appropriately substituted benzanilide deriva-
tives [15].
Thus, considering that N-(2-hydroxy-5-chloro-phenyl)-2-
methoxy-5-chloro-benzamide (Fig. 2A) was the most active
compound as BK-opener, we began [17] with a modification
of the acid moiety of the benzanilide, introducing a heterocy-
clic ring (furan, thiophene, pyrrole and pyridine) in place of
the phenyl ring but leaving unaltered the basic anilino substit-
uent (Fig. 2B). The next analogous structural modification
considered the basic moiety of benzanilide, by the introduction
of a heterocyclic ring (pyridine, thiazole, morpholine and pyr-
rolidine) in place of the aniline, leaving unaltered the benzoic
acid substituent. In this case the useful phenol function was in-
troduced by cleavage of the ortho-methoxy substituent on the
acid moiety (Fig. 2C). The pharmacological results indicated
that the presence of nitrogen heterocycles on the acid side of
the amide linker seems to be a negative requirement, while fu-
ran and thiophene rings are well tolerated. On the contrary, the
introduction of unsaturated heterocyclic rings (pyridine and
thiazole) on the basic side of the amide linker led to satisfac-
tory biological activity, while the presence of aliphatic hetero-
cycles lowered the pharmacological effect. The presence of
a phenolic function as a probable H-bond donor was
confirmed.
In this new paper, concerning a further deepening of the
structureeactivity relationships of benzanilide derivatives pre-
viously studied as BK channel activators, some substitutions
on the acid or basic moiety of the reference benzanilides,
showing particular and specific properties from a mesomeric
and/or steric point of view, have been taken into consideration.
2. Chemistry
Scheme 1 reports the preparation of three benzanilides
4aec, corresponding to the compounds previously prepared
[5,15], in whose acid moiety a chlorine atom was introduced
to increase the acid property of the phenolic function, whilst
the substituents present on the basic moiety were maintained.
Thus the 2-methoxy-5-chloro-benzoyl chloride (1) reacted
with 2-nitro-4-methyl- (2a), 2-nitro-4-methoxy- (2b) or
2-methoxy-4-nitro-aniline (2c) in refluxing toluene in the pres-
ence of triethylamine, to give the corresponding benzanilides
3aec in good yield. The next reactions of demethylation of
the methoxy substituent, carried out with excess of boron tri-
bromide in dichloromethane at ꢀ20 ꢁC overnight, caused the
cleavage of the methoxy groups present in the ortho position
of the benzoic moiety of the amides 3aec, because adjacent
On the basis of these results and the suggestions reported in
the literature [16], a simple pharmacophoric model consisting
of two suitably substituted phenyl rings bound to a linker of
various kinds, was hypothesised (Fig. 1B).
In order to support this pharmacophoric model and, in
particular, the effectiveness of the amidic linker, by a larger in-
vestigation of the structureeactivity relationships, we decided
to continue our research program developing the simple struc-
tures of the benzanilide derivatives.
A
B
H
N
F3C
LINKER
OH
O
R1
N
NS619 R = CF3
NS004 R = Cl
OH
EWG
R2
R
Fig. 1. A e Benzimidazolones NS004 and NS1619. B e Generic pharmacophoric model for a BK-opener, consisting of two aryl rings, spaced by an appropriate
heterocyclic or acyclic linker. EWG ¼ electron-withdrawing group; R1 and R2 represent various kinds of possible substituents.

V. Calderone et al. / European Journal of Medicinal Chemistry 41 (2006) 1421e1429
1423
A
B
C
HO
HO
Cl
O
OCH
Cl
3 O
O
Heterocycle
NH
N
Heterocycle
NH
H
Cl
Cl
OH
Fig. 2. A e Previously synthesised compounds. (A) The most active benzanilide as BK-opener; (B) and (C) modified benzanilides by introduction of heterocyclic
substituents.
to a carbonyl function which can stabilise the complex with
boron tribromide [15,18]. The methoxyl in the para position
of the anilino moiety was also demethylated in the same man-
ner to give the dihydroxy derivative 4b, but under these exper-
imental conditions, the methoxyl in the ortho position of the
anilino substituent of 3c was left unaltered.
substituents, to differentiate the mesomeric and steric effects.
The basic moiety of the new benzanilides was kept correspond-
ing to that of the more active derivatives, employing again 2-
hydroxy-5-chloroaniline (6a) and 2-hydroxy-5-methylaniline
(6b). Thus starting from 3-chlorobenzoyl chloride (8a), by reac-
tion with the anilines 6a and 6b, the corresponding amides 9a
and 9b were prepared, while by the reaction of 3-iodobenzoyl
chloride (8b) with the same anilines, the derivatives 9c and
9d, respectively, were obtained.
Finally, in order to evaluate the effect of a further displace-
ment of the substituent on the aromatic ring of the acid moiety
of the benzanilide, from the meta to the para position, in
Scheme 4 the synthesis of the benzanilides 11aed is reported.
As new substituents a fluorine atom, a strong electron-with-
drawing group with a few steric hindrance, and a methoxy
group, with different electronic properties but with a larger
steric hindrance, were utilised. The basic moiety of the amide
was kept unaltered utilising the pharmacologically effective
anilines 6a and 6b again. The usual synthetic route for the
new benzanilides involved the 4-fluorobenzoyl chloride
(10a) and the 4-methoxybenzoyl chloride (10b) which, on re-
action with 6a and 6b, provided the derivatives 11a, b and 11c,
d, respectively.
In Scheme 2 is reported the preparation of a series of ben-
zanilides 7aeh, where the substituent in the ortho position of
the benzoic moiety was changed increasing gradually its steric
hindrance, so that the contribution to the pharmacological ac-
tivity of this structural pattern could be evaluated. In addition,
to better evaluate the possible role of the oxygen atom as a hy-
drogen bond acceptor, in the methoxy and ethoxy derivatives,
the ortho-fluoro derivatives 7i and 7l were considered. The
basic moiety of the anilides was exactly similar to that of
previously synthesised active compounds. Therefore 2-hy-
droxy-5-chloroaniline (6a) and 2-hydroxy-5-methylaniline
(6b) were employed. They reacted in the usual manner with
benzoic (5a), 2-methylbenzoic (5b), 2-methoxybenzoic (5c),
2-ethoxybenzoic (5d) or 2-fuorobenzoic (5e) acyl chlorides
to give the corresponding benzanilides 7ael.
In Scheme 3 a further modification of the acid moiety of the
benzanilide is reported, corresponding to the removal of the
substituent in the ortho position and to the introduction of an
electron-withdrawing group in the meta position, to give com-
pounds 9aed. The chlorine and iodine atoms were chosen as
The structures of all the prepared compounds were easily
assigned on the basis of reaction mechanisms and were con-
firmed by analytical and spectroscopic data (Tables 1 and 2).
R1
O
R1
COCl
OCH3
Cl
Cl
R2
N
H
a
R2
+ H2N
OCH3
2, 3a R1= NO2; R2= CH3
2, 3b R1= NO2; R2= OCH3
2, 3c R1= OCH3; R2= NO2
1
3 a-c
b
R1
O
R2
Cl
4a R1= NO2; R2= CH3
4b R1= NO2; R2= OH
N
H
4c R1= OCH3; R2= NO2
OH
Scheme 1. Synthesis of 3aec and 4aec. (a) D, toluene, NEt3; (b) CH2Cl2, BBr3, ꢀ30 ꢁC.

1424
V. Calderone et al. / European Journal of Medicinal Chemistry 41 (2006) 1421e1429
HO
O
COCl
HO
N
a
+
H
H2N
R2
R2
R1
R1
5a R1= H
5b R1= CH3
5c R1= OCH3
6a R2= Cl
6b R2 = CH3
7a R1= H; R2= Cl
7f R1= OCH3; R2= CH3
7b R1= H; R2= CH3.
7c R1= CH3;R2= Cl
7d R1= CH3; R2= CH3
7e R1= OCH3; R2= Cl
7g R1= OCH2CH3; R2= Cl.
7h R1= OCH2 CH3; R2= CH3
7i R1= F; R2= Cl
5d R1= OCH2CH3
7l R1= F; R2= CH3
Scheme 2. Synthesis of 7ael. (a) D, toluene, NEt3.
In particular, the structure of the monomethoxy derivative 4c,
arising from the demethylation with boron tribromide of the
dimethoxy derivative 3c, was assigned by comparison with
an analogous monomethoxy derivative obtained under the
same reaction conditions [15].
NEt3-sensitive potassium channels (such as the BK one).
The vasorelaxing activity of 4b was significantly antagonised
by NEt3, however, the BK-selective blocker IbTX failed to ex-
ert any significant antagonism. Thus, the effects of 4b are
likely to be due to an activation of NEt3-sensitive potassium
channel types, different from the BK ones.
As concerns compound 7, the mechanism of action was
investigated only for compound 7g, which exhibits the highest
level of efficacy and an appreciable index of potency (compa-
rable with that of NS1619). Both NEt3 and IbTX were able to
produce a significant antagonism against the vasorelaxing
effects of this derivative, indicating that the activation of BK
channels plays a significant role in the mechanism of the vaso-
relaxing action. With regard to some possible structureeactiv-
ity relationships, it is possible to observe that a Cl atom in
para position with respect to the phenolic hydroxyl is prefer-
able to a methyl group. In fact, except for the couple 7eef, in
all the other couples of analogous compounds (sharing the R1
substituent) the Cl atom in R2 determines an improvement in
vasorelaxing efficacy. Analysing the influence of the R1 sub-
stituent in the 7 series, it was evident that a F atom in R1
(compounds 7iel) determines a dramatic reduction in efficacy.
This negative impact cannot be due to a steric hindrance.
Indeed, compounds 7ceh (all bearing a steric hindrance in
R1), as well as compounds 7aeb (bearing a H atom in R1),
exhibited vasorelaxing efficacy higher than that exhibited by
their respective analogues bearing a F atom in R1. Thus, the
3. Results and discussion
3.1. Pharmacology
In order to characterise a potential BK-activator profile, all
the synthesised compounds were submitted to preliminary
screening through the evaluation of their possible vasorelaxing
activity on endothelium-denuded rat aortic rings (Table 3).
Compounds 4aec, 7a, 7g, 9a and 11aeb exhibited a full or
nearly full vasorelaxing efficacy (Emax > 80%), while, com-
pounds 9c and 9d showed a completely ineffective vasorelax-
ing profile. All the remaining compounds (7bef, 7hel, 9b
and 11ced) showed partial vasorelaxing efficacy, with Emax
parameters ranging from 30 to 80%.
As concerns the potency indexes, all the compounds with
an efficacy value > 50% showed potency parameters ranging
around the 10 mM order of magnitude, except for compound
4c which exhibited a potency value in the micromolar range.
The non-selective Kþ-channel blocker NEt3 was not able
to antagonise the vasorelaxing effects of 4a and 4c. There-
fore, these effects cannot be ascribed to an activation of
HO
O
COCl
HO
N
H
a
+
R2
H2N
R2
R1
R1
9a R1= Cl ; R2= Cl
9b R1= Cl ; R2= CH3
9c R1= I ; R2= Cl
9d R1= I ; R2= CH3
6a R2= Cl
8a R1= Cl
8b R1= I
6b R2= CH3
Scheme 3. Synthesis of 9aed. (a) D, toluene, NEt3.

V. Calderone et al. / European Journal of Medicinal Chemistry 41 (2006) 1421e1429
1425
HO
O
COCl
HO
N
a
H
+
R2
R2
H2N
R1
R1
11a R1= F; R2= Cl.
11b R1= F; R2= CH3
11c R1= OCH3; R2= Cl
6a R2= Cl
6b R2= CH3
10a R1= F
10b R1= OCH3
11d R1= OCH3; R2= CH3
Scheme 4. Synthesis of 11aed. (a) D, toluene, NEt3.
negative influence of the F atom might be ascribed to other
reasons: (i) an electronic influence on the aromatic system
(F is the only electron-withdrawing substituent in R1 of this
group of compounds); (ii) the possible establishment of an in-
tramolecular H-bond with one of the available H atoms, thus
forcing the molecule into an unfavourable conformation and/
or subtracting a requisite useful for the interaction with the re-
ceptor; (iii) the presence of a polar feature giving unfavourable
interactions with some residues of the channel.
suggesting that BK channels are not involved in their mech-
anism of vasodilation.
In the 9 series, characterised by the R1 substituent in meta
position, the preference for a Cl atom in R2 position with re-
spect to a methyl group was again confirmed. For example,
compound 9a, showing a Cl atom both in R1 and in R2,
showed a NEt3- and IbTX-sensitive vasorelaxing activity, typ-
ical of a BK-opener profile, with a potency index not much
lower than NS1619.
The observation that compounds 11aeb, bearing
a F atom in para position, exhibit a more satisfactory vaso-
relaxing effect, excludes the hypothesis of the importance of
the electronic influence of F on the aromatic system. How-
ever, it should be taken into account that 11aeb bearing
a F atom in R1, represent the most effective compounds in
the 11 series (characterised by an R1 substituent in para
position) but show either NEt3-insensitive effects (11a)
or NEt3-sensitive but IbTX-insensitive activity (11b),
Therefore, among the molecules observed, it appears evi-
dent that for some of them the vasorelaxing effect was clearly
due to the interaction with the BK channel, in particular com-
pounds 7g and 9a.
3.2. Molecular modelling
On the basis of such results, the structure of compounds 7g
and 9c was investigated by molecular modelling studies. The
Table 1
Physico-chemical properties of compounds 3, 4, 7, 9 and 11
Compounds
Yield (%)
Crystallisation
solvent
m.p. (ꢁC)
Analyses (C, H, N)
IR (cmꢀ1
)
3a
3b
3c
69
59
84
85
62
68
78
72
80
75
83
75
70
80
82
79
78
65
70
85
90
52
58
78
Iso-PrOH
MeOH
AcOEt
190e192
185e187
235e237
191e193
224e226
233e235
232e234
135e137
225e227
125e127
234e236
165e167
170e172
133e135
236e238
195e197
220e222
190e192
235e237
228e230
228e230
189e190
203e205
129e131
C
15H13ClN2O4
1680 (CO); 3284 (NH)
1672 (CO); 3309 (NH)
C15H13ClN2O5
C15H13ClN2O5
1666 (CO); 3292 (NH)
1642 (CO); 3250 (NH, OH)
4a
4b
4c
MeOH
C
14H11ClN2O4
13H9ClN2O5
AcOEt/hexane
AcOEt/hexane
MeOH/H2O
MeOH
C
1646 (CO); 3213 (OH); 3388 (NH)
1654 (CO); 3253 (OH); 3392 (NH)
1648 (CO); 3080 b (OH); 3422 (NH)
1645 (CO); 3410 b (NH, OH)
C14H11ClN2O5
C13H10ClNO2
C14H13NO2
7a
7b
7c
MeOH/H2O
MeOH
MeOH
C14H12ClNO2
C15H15NO2
C14H12ClNO3
1648 (CO); 3100 b (OH); 3406 (NH)
1647 (CO); 3090 b (OH); 3407 (NH)
1644 (CO); 3145 b (OH); 3298 (NH)
1642 (CO); 3180 b (OH); 3320 (NH)
1637 (CO); 3380 b (NH, OH)
7d
7e
7f
MeOH/H2O
MeOH/H2O
MeOH
C15H15NO3
C15H14ClNO3
C16H17NO3
7g
7h
7i
1640 (CO); 3100 b (OH); 3353 (NH)
1650 (CO); 3392 (NH); 3120 b (OH)
1649 (CO); 3389 (NH); 3200 b (OH)
1648 (CO); 3116 (OH); 3415 (NH)
1638 (CO); 3220 b (NH, OH)
MeOH
C13H9ClFNO2
7l
MeOH/H2O
MeOH/H2O
MeOH/H2O
MeOH
C14H12FNO2
C13H9Cl2NO2
C14H12ClNO2
9a
9b
9c
C
13H9Cl INO2
14H12INO2
1637 (CO); 3210 b (NH, OH)
1636 (CO); 3230 b (NH, OH)
9d
11a
11b
11c
11d
MeOH
MeOH
C
C13H9ClFNO2
C14H12FNO2
1644 (CO); 3100 b (OH); 3412 (NH)
1650 (CO); 3070 b (OH); 3414 (NH)
1649 (CO); 3070 b (OH); 3415 (NH)
1639 (CO); 3080 (OH); 3369 (NH)
AcOEt/hexane
MeOH
MeOH/H2O
C14H12ClNO3
C15H15NO3

1426
V. Calderone et al. / European Journal of Medicinal Chemistry 41 (2006) 1421e1429
Table 2
1H NMR spectra in DMSO-d6, d values
Table 3
Pharmacological results
3a
3b
3c
2.38 (s, 3H, CH3); 4.05 (s, 3H, OCH3); 7.30e8.46 (m, 6H, Ar.);
10.1 (NH).
Compounds
Emax
pIC50
4a
4b
94.6 ꢂ 5.0
96.7 ꢂ 4.4
97.4 ꢂ 0.1
96.4 ꢂ 9.4
72.4 ꢂ 7.1
62.2 ꢂ 4.6
48.4 ꢂ 8.2
67.5 ꢂ 8.8
77.3 ꢂ 8.7
100a
5.39 ꢂ 0.07
5.28 ꢂ 0.04
6.16 ꢂ 0.04
4.87 ꢂ 0.03
5.03 ꢂ 0.26
4.52 ꢂ 0.03
N.C.
3.85 (s, 3H, OCH3); 4.04 (s, 3H, OCH3); 7.28e8.40 (m, 6H, Ar.);
11.4 (NH).
4c
7a
4.06 (s, 3H, OCH3); 4.08 (s, 3H, OCH3); 6.97e8.11(m, 6H, Ar.);
10.8 (NH).
7b
7c
4a
4b
4c
7a
7b
7c
7d
2.39 (s, 3H, CH3); 7.03e8.40 (m, 6H, Ar.); 10.9 (NH); 11.9 (OH).
7.02e8.16 (m, 6H, Ar.); 10.3 (NH); 11.4 (OH); 12.1 (OH).
4.04 (s, 3H, OCH3); 7.06e8.73 (m, 6H, Ar.); 11.3 (NH); 12.3 (OH).
6.90e7.97 (m, 8H, Ar.); 9.5 (NH); 10.1 (OH).
2.22 (s, 3H, CH3); 6.82e7.98 (m, 8H, Ar.); 9.5 (NH, OH).
2.51 (s, 3H, CH3); 6.90e7.91 (m, 7H, Ar.); 9.4 (NH); 10.2 (OH).
2.22 (s, 3H, CH3); 2.41 (s, 3H, CH3); 6.81e7.54 (m, 7H, Ar.);
9.4 (NH); 9.5 (OH).
7d
7e
4.73 ꢂ 0.06
4.90 ꢂ 0.12
5.51 ꢂ 0.06
4.95 ꢂ 0.07
4.52 ꢂ 0.03
N.C.
7f
7g
7h
7i
67.9 ꢂ 3.2
52.6 ꢂ 4.2
39.7 ꢂ 10.6
100a
7l
9a
7e
7f
4.04 (s, 3H, OCH3); 6.91e8.46 (m, 7H, Ar.); 10.6 (NH); 10.7 (OH).
2.22 (s, 3H, CH3); 4.03 (s, 3H, OCH3); 6.74e8.21(m, 7H, Ar.);
9.9 (NH); 10.5 (OH).
4.93 ꢂ 0.03
N.C.
N.C.
9b
9c
33.7 ꢂ 10.5
Ineffectiveb
Ineffectiveb
81.4 ꢂ 12.6
84.0 ꢂ 6.1
45.7 ꢂ 8.5
46.0 ꢂ 1.8
100a
7g
7h
1.52, 4.30 (t, 3H, q, 2H, CH3CH2O); 6.87e8.51 (m, 7H, Ar.);
10.5 (NH); 10.6 (OH).
9d
N.C.
11a
11b
11c
11d
NS1619
4.76 ꢂ 0.04
4.95 ꢂ 0.04
N.C.
1.52, 4.29 (t, 3H, q, 2H, CH3CH2O); 2.22 (s, 3H, CH3);
6.70e8.21 (m, 7H, Ar.); 10.0 (NH); 10.4 (OH).
6.88e8.21 (m, 7H, Ar.); 9.5 (NH); 10.5 (OH).
2.23 (s, 3H, CH3); 6.74e7.92 (m, 7H, Ar.); 9.4 (NH); 9.8 (OH).
6.92e8.01 (m, 7H, Ar.); 9.7 (NH); 10.1 (OH).
2.22 (s, 3H, CH3); 6.82e8.01 (m, 7H, Ar.); 9.4 (NH); 9.7 (OH).
6.90e8.30 (m, 7H, Ar.); 9.7 (NH); 9.8 (OH).
7i
N.C.
5.31 ꢂ 0.08
7l
9a
The values of potency (pIC50) could not be calculated (N.C.) for those com-
pounds exhibiting an efficacy (Emax) lower than 50.
9b
9c
a
At the concentration of 30 mM, the compound evoked a 100% vasorelaxing
effect, in all the experiments.
9d
2.22 (s, 3H, CH3); 6.83e8.32 (m, 7H, Ar.); 9.4 (NH); 9.8 (OH).
6.91e8.08 (m, 7H, Ar.); 9.5 (NH); 10.1 (OH).
2.23 (s, 3H, CH3); 6.83e8.08 (m, 7H, Ar.); 9.4(NH); 9.5 (OH).
3.84 (s, 3H, OCH3); 6.94e7.97 (m, 7H, Ar.); 9.4 (NH); 10.2 (OH).
2.22 (s, 3H, CH3); 3.82 (s, 3H, OCH3); 6.82e7.98 (m, 7H, Ar.);
9.4 (NH); 9.5 (OH).
11a
11b
11c
11d
b
At the concentration of 30 mM, the compound evoked a mean vasorelaxing
effect lower than 10% and therefore it was considered almost ineffective.
The ortho substitution in the R1 aromatic ring of the vaso-
relaxing molecule 7g with respect to the ineffective 9c is high-
lighted by the electrostatic potential surface profile in
Fig. 3B,C. This observation allows us to hypothesise that
some favourable interaction between bulky non-polar groups
(such as methoxy or ethoxy) in ortho position of one of the ar-
omatic rings and the hydrophobic residues of the receptor
could account for the observed activity of 7g. On the contrary,
the ineffective compound 9c is lacking in the ethoxy substitu-
ent and bears an I atom as substituent in meta position which
could fill different regions of the receptor unable to accommo-
date it. Thus, the inefficacy of 9c could be attributed both to
the lack of the bulky susbtituent and to the role of I in a differ-
ent position of the molecule, thus inducing an unfavourable
interaction with the receptor. Although the limited number
of compounds studied does not allow us to consider it an
exhaustive modelling and SAR study, the hypothesis about
the importance of a bulky non-polar substituent on one of
the two aromatic ring confirms some data previously obtained
in a work by us [9] and seems to be in accord with the obser-
vations of Ohwada et al.[20].
molecules were chosen as representative structures of an effec-
tive opener (7g) and an almost completely ineffective mole-
cule (9c). A Molecular Mechanic (MM) approach was
carried out on these selected molecules and on NS1619, which
was chosen as a representative moiety to be subjected to an ex-
haustive conformational search. The molecules were built us-
ing the molecular modelling package Insight (Accelrys, San
Diego, CA). Different plausible conformers were generated
˚
for NS1619 and solvated with a 5 A layer of water. The con-
formers were subjected to energy minimisation using the cff91
Force Field implemented in the Discover program (Accelrys,
San Diego, CA). The conjugate gradient algorithm was ex-
ploited for energy minimisation and the convergence criterion
˚
A
for the energy gradient was 0.01 kcal molꢀ1 ꢀ1. All calcula-
tions were performed on an SGI Octane R12000 workstation.
A reasonably good sampling of the potential energy surface
was obtained by selecting the optimised conformer for
NS1619 (Fig. 3A) characterised by the lowest energy value.
The structures of the other selected molecules were then opti-
mised on the template of NS1619 using a restraining force of
100 kcal molꢀ1
ꢀ2. The two aromatic nuclei and the OH
˚
A
function were chosen as critical features for the restrain, since
they constitute the major pharmacophoric points of the hy-
pothesised pharmacophore (Fig. 1B). After that, molecular
electrostatic potentials (MEPs) were calculated on the opti-
mised structures of 7g and 9c (Fig. 3B,C) using the program
VEGA [19].
4. Conclusion
Summing up, the synthesis of the series of substituted benza-
nilides and the structureeactivity relationship (SAR) study
performed in this work confirm the pharmacophoric features
described above. The presence of two aromatic nuclei and

V. Calderone et al. / European Journal of Medicinal Chemistry 41 (2006) 1421e1429
1427
Fig. 3. Three-dimensional structure of NS1619 (A). Molecular electrostatic potential (MEP) surface calculated projecting the atomic charge on the surface for 7g
(B), 9c (C). In blue, the polar influence of the phenolic OH and the carbonyl group in the amide function of the linker. (For interpretation of the references to colour
in this figure legend, the reader is referred to the web version of this article.)
a phenolic OH is essential, even if they cannot be considered the
unique elements accounting for the activity of the compounds.
Furthermore, a Cl atom in para position with respect to the phe-
nolic hydroxyl is preferable to a methyl group, according to the
observation of Biagi et al. [15] and Hewawasam et al.[21].
With regard to the compounds of the series 7, bearing the
OH function in the anilino moiety of the benzanilide, it can
be observed that the presence of an appropriate substituent
in ortho position of the acid moiety of the benzanilide might
be able to give favourable interactions with a specific region
of the channel. Indeed, the vasorelaxing efficacy increases in
compounds 7aeh with the steric hindrance of ortho substitu-
ent R1; whereas it is dramatically reduced by the introduction
of the F atom in the same position.
precipitate was formed which was collected by filtration and
treated with H2O. The insoluble material consisted of the
crude title compounds 3aec in variable amounts. The toluene
filtrate was evaporated in vacuo and the residue was dissolved
in CHCl3. The chloroform solution, after washing with 10%
NaOH and 10% HCl, was dried (MgSO4) and evaporated to
give a further amount of the title compounds. The two solid
fractions of each compound 3a, 3b or 3c were combined
and purified by crystallisation (Table 1).
5.1.2. N-(2,4-Substituted-phenyl)-2-hydroxy-5-chloro-
benzamides (4aec)
A solution of 2.00 mmol of methoxyderivative 3a, 3b or 3c
in 150 mL of anhydrous CH2Cl2 was cooled at ꢀ30 ꢁC and,
under stirring, a solution of BBr3 in anhydrous CH2Cl2 was
added drop by drop [1.5 mL (y 15 mmol) in 5 mL of solvent
for 3a; 3 mL (y 30 mmol) in 10 mL of solvent for 3b and 3c].
After the addition the mixture was stirred at ꢀ30 ꢁC for 1 h,
then kept at ꢀ20 ꢁC overnight. The CH2Cl2 solution was
cooled in an ice-bath and the excess of the reagent was decom-
posed by the addition of MeOH and H2O.
The evaluation of this spatial region of the pharmacophore
and the role of the linker could be worth of further investiga-
tion in future perspectives of synthesis.
5. Experimental protocols
5.1. Chemistry
For the isolation of 4a, 6 mL of MeOH followed by 30 mL
of H2O were added, the organic layer was separated and, after
washing with H2O, was extracted with 10% NaOH. Addition
of the alkaline solution caused precipitation of a red solid
(consisting of 4a sodium salt) in the interface of the two liquid
phases, which was collected by filtration. This solid was dis-
solved in H2O and the solution was acidified with 10% HCl
to give 4a as a yellow precipitate. The two liquid phases
were separated and the alkaline extracts acidified to give a fur-
ther amount of 4a, which was combined with the previous
fraction and purified by crystallisation (Table 1).
Melting points were determined on a Kofler hot-stage and
are uncorrected. IR spectra in nujol mulls were recorded on
a Mattson Genesis series FTIR spectrometer. 1H NMR spectra
were recorded with a Varian Gemini 200 spectrometer in
DMSO-d6 or CDCl3, in d units, using TMS as an internal stan-
dard. Mass spectra were performed with a Trace GC Q plus,
thermo quest Finnigan. Elemental analyses (C, H, N) were
within ꢂ 0.4% of the theoretical values and were performed
on a Carlo Erba Elemental Analyzer Mod. 1106 apparatus.
5.1.1. N-(2,4-Substituted-phenyl)-2-methoxy-5-chloro-
benzamides (3aec)
For the isolation of 4b, 8 mL of MeOH followed by 60 mL
of H2O were added, the organic layer was separated and, after
washing with H2O, was extracted with 10% NaOH. The alka-
line extracts were combined and acidified to precipitate 4b
which was collected by filtration and purified by crystallisa-
tion (Table 1). For the isolation of 4c, 8 mL of MeOH followed
by 40 mL of H2O were added, the organic layer was separated
and, after washing with H2O, was extracted with 10% NaOH.
Addition of the alkaline solution caused precipitation of a solid
(consisting of 4c sodium salt) in the interface of the two liquid
phases, which was collected by filtration. This solid was
A solution of 2-methoxy-5-chloro-benzoic acid (0.930 g,
5.0 mmol) in 5e10 mL of SOCl2 was heated under reflux
for 45 min. The solvent was evaporated in vacuo and the liquid
residue consisting of the corresponding acid chloride (1) was
dissolved in 20 mL of anhydrous toluene. This solution was
added drop by drop to a solution of the appropriate substituted
aniline (5.0 mmol of 2a, 2b or 2c) and triethylamine (NEt3)
(1.0 mL, 7.2 mmol) in 20 mL of anhydrous toluene. The mix-
ture was refluxed overnight and, after cooling, a solid

1428
V. Calderone et al. / European Journal of Medicinal Chemistry 41 (2006) 1421e1429
suspended in H2O then acidified to give 4c which was col-
lected and purified by crystallisation (Table 1).
Acidification of the alkaline layer, after separation from the
organic phase, did not give a further precipitate.
anhydrous toluene. This solution was added drop by drop to
a solution of the substituted aniline 6a or 6b (5.0 mmol) and
NEt3 (1.0 mL, 7.2 mmol) in 20 mL of anhydrous toluene.
The mixture was refluxed overnight, the solvent evaporated
in vacuo and the residue dissolved in CHCl3. The chloroform
solution, after washing with 5% NaHCO3 and 10% HCl, was
dried (MgSO4) and evaporated to give the title compounds as
a solid residue which was purified by crystallisation (Table 1).
5.1.3. 2-Substituted benzoyl chlorides (5aee)
A solution of 3.00 mmol of the suitable benzoic acid [ben-
zoic (5a), 2-methylbenzoic (5b), 2-methoxybenzoic (5c),
2-ethoxybenzoic (5d) or 2-fluorobenzoic (5e)] in 5e10 mL
of SOCl2 was heated under reflux for 45e60 min. The solvent
was evaporated in vacuo and the liquid residue was dissolved
in 15 mL of anhydrous toluene to be employed in the follow-
ing reactions.
5.1.7. N-(2-Hydroxy-5-substituted-phenyl)-3-iodo-
benzamides (9c, d)
A solution of 3-iodobenzoic acid (1.24 g, 5.0 mmol) in 5e
10 mL of SOCl2 was heated under reflux for 45e60 min. The
solvent was evaporated in vacuo and the residue, consisting of
the acyl chloride 8b, was dissolved in 20 mL of anhydrous tol-
uene. This solution was added drop by drop to a solution of the
substituted aniline 6a or 6b (5.0 mmol) and NEt3 (1.0 mL,
7.2 mmol) in 20 mL of anhydrous toluene. The mixture was
refluxed overnight and, after cooling, a solid precipitate was
formed which was collected by filtration and treated with
H2O. The insoluble material consisted of the title compounds
9c or 9d. The toluene filtrate, worked up in the usual manner,
gave only a further moderate amount of 9c. The crude solids
obtained were purified by crystallisation (Table 1).
5.1.4. N-(2-Hydroxy-5-chloro-phenyl)-2-substituted-
benzamides (7a, 7c, 7e, 7g and 7i)
To a solution of 2-hydroxy-5-chloro-aniline (6a) (0.430 g,
3.00 mmol) and NEt3 (0.5 mL, 3.6 mmol) in 20 mL of anhy-
drous toluene, a solution of the suitable acyl chloride (5a,
5b, 5c, 5d or 5e) was added drop by drop and the mixture
was refluxed overnight.
For the isolation of 7a, 7c and 7i, after cooling, a solid pre-
cipitate was formed which was collected by filtration and
treated with H2O to give a main fraction of the title
compounds as insoluble material. The toluene filtrate was
evaporated, the residue dissolved in CHCl3 and the chloroform
solution, after washing with 5% NaHCO3 and 10% HCl, was
dried (MgSO4) and evaporated to give a further fraction of 7a,
7c and 7i (Table 1).
For the isolation of 7e, after cooling, a solid precipitate was
not present, whilst for 7g a solid precipitate was formed but
was completely soluble in H2O. Therefore the toluene solu-
tions were evaporated, the residues were dissolved in CHCl3
and the chloroform solutions, after washing with 5% NaHCO3
and 10% HCl, were dried (MgSO4) and evaporated to provide
the title compounds 7e and 7g in good yield (Table 1).
5.1.8. N-(2-Hydroxy-5-substituted-phenyl)-4-fluoro-
benzamides (11a, b)
A solution of 4-fluorobenzoic acid (0.700 g, 5.0 mmol) in
5e10 mL of SOCl2 was heated under reflux for 45e60 min.
The solvent was evaporated in vacuo and the residue, consisting
of the acyl chloride 10a, was dissolved in 20 mL of anhydrous
toluene. This solution was added drop by drop to a solution of
the substituted aniline 6a or 6b (5.0 mmol) and NEt3 (1.0 mL,
7.2 mmol) in 20 mL of anhydrous toluene. The mixture was re-
fluxed overnight and, after cooling, a solid precipitate was
formed which was collected by filtration and treated with
H2O. The insoluble material consisted of the title compounds
11a or 11b. The toluene filtrate, worked up in the usual manner,
gave a further moderate amount of product. The solid fractions
were combined and purified by crystallisation (Table 1).
5.1.5. N-(2-Hydroxy-5-methyl-phenyl)-2-substituted-
benzamides (7b, 7d, 7f, 7h and 7l)
To a solution of 2-hydroxy-5-methyl-aniline (6b) (0.370 g,
3.00 mmol) and NEt3 (0.5 mL, 3.6 mmol) in 20 mL of anhy-
drous toluene, a solution of the suitable acyl chloride (5a,
5b, 5c, 5d or 5e) was added drop by drop and the mixture
was refluxed overnight.
For the isolation of 7b, 7d, 7h and 7l, after cooling, a solid
precipitate was formed and the reactions were worked up as
described for 7a (Table 1).
5.1.9. N-(2-Hydroxy-5-substituted-phenyl)-4-methoxy-
benzamides (11c, d)
A solution of 4-methoxybenzoic acid (0.761 g, 5.0 mmol)
in 5e10 mL of SOCl2 was heated under reflux for
45e60 min. The solvent was evaporated in vacuo and the
residue, consisting of the acyl chloride 10b, was dissolved in
20 mL of anhydrous toluene. This solution was added drop
by drop to a solution of the substituted aniline 6a or 6b
(5.0 mmol) and NEt3 (1.0 mL, 7.2 mmol) in 20 mL of anhy-
drous toluene. The mixture was refluxed overnight and, after
cooling, a solid precipitate was formed which was collected
by filtration and treated with H2O. The insoluble material
consisted of the title compounds 11c or 11d. The toluene fil-
trate, worked up in the usual manner, gave a further moderate
amount of product. The solid fractions were combined and
purified by crystallisation (Table 1).
For the isolation of 7f, after cooling, a solid precipitate was
not present, therefore the toluene solution was worked up as
described for 7e (Table 1).
5.1.6. N-(2-Hydroxy-5-substituted-phenyl)-3-chloro-
benzamides (9a, b)
A solution of 3-chlorobenzoic acid (0.783 g, 5.0 mmol) in
5e10 mL of SOCl2 was heated under reflux for 45e60 min.
The solvent was evaporated in vacuo and the residue, consist-
ing of the acyl chloride 8a, was dissolved in 20 mL of

V. Calderone et al. / European Journal of Medicinal Chemistry 41 (2006) 1421e1429
1429
5.2. Pharmacology
contractile tone induced by KCl 20 mM. The parameter of ef-
ficacy (Emax) represented the vasorelaxing response, expressed
as a percentage (%) of the contractile tone induced by KCl
20 mM, evoked by the limit concentration 0.1 mM (the highest
concentration, which could be administered). Those com-
pounds exhibiting an Emax lower than 10% were considered
as ineffective. The parameter of potency was expressed as
pIC50, calculated as a negative Logarithm of the molar concen-
tration of the tested compounds, evoking a half-reduction of
the contractile tone induced by KCl 20 mM. The pIC50 could
not be calculated for those compounds showing an efficacy
parameter lower than 50%. The parameters of efficacy and
potency were expressed as mean ꢂ standard error, for 5e10
experiments. Student-t test was selected as statistical analysis,
P < 0.05 was considered representative of a significant
statistical difference. Experimental data were analysed by
a computer fitting procedure (software: GraphPad Prism 4.0).
All the experimental procedures were carried out following
the guidelines of the European Community Council Directive
86-609.
To determine a possible vasodilator mechanism of action,
the compounds were tested on isolated thoracic aortic rings
of male normotensive Wistar rats (250e350 g).
The rats were sacrificed by cervical dislocation under light
ether anaesthesia and bled. The aortae were immediately ex-
cised and freed of extraneous tissues. The endothelial layer
was removed by gently rubbing the intimal surface of the ves-
sels with a hypodermic needle. Five-millimeter wide aortic
rings were suspended, under a preload of 2 g, in 20 mL organ
baths, containing Tyrode solution (composition of saline in
mM: NaCl 136.8; KCl 2.95; CaCl2 1.80; MgSO4$7H2O
1.05; NaH2PO4 0.41; NaHCO3 11.9; glucose 5.5), thermo-
stated at 37 ꢁC and continuously gassed with a mixture of
O2 (95%) and CO2 (5%). Changes in tension were recorded
by means of an isometric transducer (Grass FTO3) connected
to a preamplifier (Buxco Electronics) and to a software of data
acquisition (BIOPAC Systems Inc., MP 100).
After an equilibration period of 60 min, endothelial re-
moval was confirmed by the administration of acetylcholine
(ACh) (10 mM) to KCl (20 mM)-precontracted vascular rings.
A relaxation < 10% of the KCl-induced contraction was con-
sidered representative of an acceptable lack of the endothelial
layer, while the organs, showing a relaxation ꢃ 10% (i.e. sig-
nificant presence of the endothelium), were discarded. From
30 to 40 min after confirmation of the endothelium removal,
the aortic preparations were contracted by treatment with a sin-
gle concentration of KCl (20 mM) and when contraction
reached a stable plateau, 3-fold increasing concentrations of
the tested compounds or of the reference drug NS1619
(a well-known BK-activator) were added cumulatively in the
concentration range 10 nMe30 mM.
References
[1] V. Calderone, Curr. Med. Chem. 9 (2002) 1385e1395.
[2] S.-N. Wu, Curr. Med. Chem. 10 (2003) 649e661.
[3] J.E. Brayden, M.T. Nelson, Science 256 (1992) 532e535.
[4] Y. Tanaka, K. Koike, L. Toro, J. Smooth Muscle Res. 40 (2004)
125e153.
[5] G. Biagi, I. Giorgi, O. Livi, V. Scartoni, P.L. Barili, V. Calderone,
E. Martinotti, Farmaco 56 (2001) 827e834.
[6] Y. Shintani, K. Node, H. Asanuma, S. Sanada, S. Takashima, Y. Asano,
Y. Liao, M. Fujita, A. Hirata, Y. Shinozaki, T. Fukushima, Y. Nagamachi,
H. Okuda, J. Kim, H. Tomoike, M. Hori, M. Kitakaze, J. Mol. Cell. Car-
diol. 37 (2004) 1213e1218.
[7] C.-M. Cao, Q. Xia, Q. Gao, M. Chen, T.-M. Wong, J. Pharmacol. Exp.
Ther. 312 (2005) 644e650.
[8] T. Sato, T. Saito, N. Saegusa, H. Nakaya, Circulation 111 (2005)
198e203.
[9] G. Biagi, V. Calderone, I. Giorgi, O. Livi, V. Scartoni, B. Baragatti,
E. Martinotti, Eur. J. Med. Chem. 35 (2000) 715e720.
[10] G. Biagi, V. Calderone, I. Giorgi, O. Livi, E. Martinotti, A. Martelli,
A. Nardi, Farmaco 59 (2004) 397e404.
Preliminary experiments showed that the KCl (20 mM)-
induced contractions remained in a stable tonic state for at
least 40 min.
[11] V. Calderone, I. Giorgi, O. Livi, E. Martinotti, A. Martelli, A. Nardi,
Farmaco 60 (2005) 367e375.
[12] V. Calderone, I. Giorgi, O. Livi, E. Martinotti, E. Mantuano, A. Martelli,
A. Nardi, Eur. J. Med. Chem. 40 (2005) 521e528.
In other sets of experiments, the non-selective potassium
channel blocker tetraethylammonium chloride (NEt3 10 mM)
or the BK-selective blocker Iberiotoxin (IbTX, 100 nM)
were added, after the KCl (20 mM)-induced contraction, fol-
lowed by the administration of selected compounds.
The reference drug NS1619 (Sigma) was dissolved
(10 mM) in EtOH 95% and further diluted in Tyrode solution.
Acetylcholine chloride (Sigma) was dissolved (100 mM) in
EtOH 95% and further diluted in bidistilled water whereas
KCl and NEt3 were both dissolved in Tyrode solution. All
the synthesised derivatives were dissolved (10 mM) in
DMSO; they were all further diluted in Tyrode solution. All
the solutions were freshly prepared immediately before the
pharmacological experimental procedures. Previous experi-
ments showed a complete ineffectiveness of the administration
of the vehicle.
[13] B. Baragatti, G. Biagi, V. Calderone, I. Giorgi, O. Livi, E. Martinotti,
V. Scartoni, Eur. J. Med. Chem. 35 (2000) 949e955.
[14] G. Biagi, V. Calderone, I. Giorgi, O. Livi, V. Scartoni, B. Baragatti,
E. Martinotti, Farmaco 56 (2001) 841e849.
[15] G. Biagi, I. Giorgi, O. Livi, A. Nardi, V. Calderone, A. Martelli,
E. Martinotti, O. LeRoy Salerni, Eur. J. Med. Chem. 39 (2004) 491e498.
[16] M.J. Coghlan, M. Gopalakrishnan, W.A. Carroll, Annual Reports in Me-
dicinal Chemistry, in: A.M. Doherty (Ed.), Academic Press, San Diego,
2001, pp. 11e20.
[17] V. Calderone, F.L. Fiamingo, I. Giorgi, M. Leonardi, O. Livi, A. Martelli,
E. Martinotti, Eur. J. Med. Chem. 41 (6) (2006) 761e767.
[18] F.M. Dean, J. Goodchild, L.E. Houghton, J.A. Martin, R.B. Morton,
B. Parton, A.W. Price, N. Somvichien, Tetrahedron Lett. 7 (1966)
4153e4159.
[19] A. Pedretti, L. Villa, G. Vistoli, J. Comput. Aided Mol. Des. 18 (2004)
167e173.
[20] T. Ohwada, T. Nonomura, K. Maki, K. Sakamoto, S. Ohya, K. Muraki,
Y. Imaizumi, Bioorg. Med. Chem. Lett. 13 (2003) 3971e3974.
[21] P. Hewawasam, N.A. Meanwell, V.K. Gribkoff, S.I. Dworetzky,
C.G. Boissard, Bioorg. Med. Chem. Lett. 7 (1997) 1255e1260.
Vasorelaxing efficacy was evaluated as a maximal vasore-
laxing response, expressed as a percentage (%) of the