Synthesis, pharmacological evaluation, and structure-activity relationship and quantitative structure-activity relationship studies on novel derivatives of 2,4-diamino-6,7-dimethoxyquinazoline α1-adrenoceptor antagonists

Article abstract of DOI:10.1021/jm9805337

A new series of novel piperazine and non-piperazine derivatives of 2,4- diamino-6,7-dimethoxyquinazoline was synthesized and evaluated for binding affinity toward α₁-adrenergic and other G-protein-coupled aminergic receptors. The α₁-adrenoceptor (AR) subtype selectivity was also investigated for the most interesting compounds. Only compound 16 showed moderate selectivity toward the α(1b)-AR subtype. Selected compounds were tested in vivo in a dog model indicating activity on blood pressure and on the lower urinary tract. Compound 10 showed in vivo potency close to that of prazosin. Powerful interpretative and predictive theoretical QSAR models have been obtained. The theoretical descriptors employed in the rationalization of the α₁-adrenergic binding affinity depict the key features for receptor binding which can be summarized in an electrostatic interaction between the protonated amine function and a primary nucleophilic site of the receptor, complemented by short-range attractive (polar and dispersive) and repulsive (steric) intermolecular interactions. Moreover, on predictive grounds, the ad hoc derived size and shape QSAR model developed in a previous paper (Rastelli, G.; et al. J. Mol. Struct. 1991, 251, 307-318) proved to be successful in predicting nanomolar α₁-adrenergic binding affinity for compound 28.

Full text of DOI:10.1021/jm9805337

J . Med. Chem. 1999, 42, 427-437
427
Syn th esis, P h a r m a cologica l Eva lu a tion , a n d Str u ctu r e-Activity Rela tion sh ip
a n d Qu a n tita tive Str u ctu r e-Activity Rela tion sh ip Stu d ies on Novel Der iva tives
of 2,4-Dia m in o-6,7-d im eth oxyqu in a zolin e r1-Ad r en ocep tor An ta gon ists
Amedeo Leonardi,† Gianni Motta,*,§ Carlo Boi,§ Rodolfo Testa,‡ Elena Poggesi,‡ Pier G. De Benedetti, and
M. Cristina Menziani
Recordati S.p.A., Via M. Civitali 1, 20148 Milano, Italy, and Dipartimento di Chimica, Universita` di Modena, Via Campi 183,
41100 Modena, Italy
Received September 22, 1998
A new series of novel piperazine and non-piperazine derivatives of 2,4-diamino-6,7-dimethoxy-
quinazoline was synthesized and evaluated for binding affinity toward R1-adrenergic and other
G-protein-coupled aminergic receptors. The R1-adrenoceptor (AR) subtype selectivity was also
investigated for the most interesting compounds. Only compound 16 showed moderate selectivity
toward the R1b-AR subtype. Selected compounds were tested in vivo in a dog model indicating
activity on blood pressure and on the lower urinary tract. Compound 10 showed in vivo potency
close to that of prazosin. Powerful interpretative and predictive theoretical QSAR models have
been obtained. The theoretical descriptors employed in the rationalization of the R1-adrenergic
binding affinity depict the key features for receptor binding which can be summarized in an
electrostatic interaction between the protonated amine function and a primary nucleophilic
site of the receptor, complemented by short-range attractive (polar and dispersive) and repulsive
(steric) intermolecular interactions. Moreover, on predictive grounds, the ad hoc derived size
and shape QSAR model developed in a previous paper (Rastelli, G.; et al. J . Mol. Struct. 1991,
251, 307-318) proved to be successful in predicting nanomolar R1-adrenergic binding affinity
for compound 28.
Ch a r t 1. 2-(Substituted-amino)-4-amino-6,7-dimeth-
oxyquinazoline R1-AR Antagonists
In tr od u ction
Among the compounds classified as antagonists of the
R1-adrenoceptor,1-7 the derivatives of 2,4-diamino-6,7-
dimethoxyquinazoline proved to be the first to show
potent and selective activity.8 This relevant character-
istic granted the synthesis and pharmacological evalu-
ation of a large number of derivatives by different
research groups.9-22 In Chart 1 the structure of the
earliest example, prazosin (1), is shown, together with
those of other very potent R1-adrenoceptor antagonists
(2-4) of this class.
Recent structure-activity relationship (SAR) studies
on the wide-ranging series of 2,4-diamino-6,7-dimeth-
oxyquinazoline8-23 and -quinoline24 derivatives have
qualitatively rationalized the relevant affinity and
selectivity of these compounds for the R1-adrenoceptor.
It has been proposed that the 2,4-diamino-6,7-dimethoxy-
quinazoline nucleus acts as a conformationally re-
stricted bioisosteric replacement of noradrenaline with
the N1-protonated form being particularly suited for
effective charge-reinforced hydrogen bonding with the
carboxylate counterion in the ground-state conformation
of the R1-adrenoceptor.8,9 This is corroborated by the
lack of significant affinity and antihypertensive activity
of the structurally closely related isoquinolines, where
a
b
See ref 11. See ref 15. c See ref 12.
the bioisosteric substitution (C-H instead of N1) abol-
ishes any biological activity.14
Further support comes from a recent report on mo-
lecular modeling and quantitative structure-activity
relationship (QSAR) analysis of a heterogeneous series
of quinazoline, quinoline and isoquinoline derivatives.25
The QSAR models obtained in this paper by correlating
theoretical molecular descriptors with both the experi-
mental acidity constants and the R1-adrenoceptor bind-
ing affinity data values clearly depicted the fundamental
role of the protonated quinazoline nucleus for a produc-
tive interaction with the receptor. Furthermore, subse-
quent studies26 have pointed out that once the electronic
requirements of the common quinazoline moiety are
satisfied, the binding affinities are modulated by the
molecular shape of the quinazoline 2-substituent, through
* To whom correspondence should be addressed. Tel: 0039-02-
48787493.
Fax:
0039-02-48709017.
E-mail:
MOTTA.G@
RECORDATI.IT.
† Division of Medicinal Chemistry and Pharmacology.
§ Medicinal Chemistry Department.
‡ Pharmacology Department.
Chemistry Department, Universita` di Modena.
10.1021/jm9805337 CCC: $18.00 © 1999 American Chemical Society
Published on Web 02/11/1999

428 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3
Leonardi et al.
Sch em e 1
Sch em e 2
the optimization of both dispersive and steric interac-
tions. In this respect, ad hoc derived size and shape
descriptors defined on the ligand bioactive molecular
form have proven to be very successful to derive QSAR
models for several molecular series of G-protein-coupled
receptor ligands.27-29 These indices describe the size-
shape similarity with respect to a reference supermol-
ecule obtained by superposition of the most active and
structurally different compounds, better if conforma-
tionally constrained. According to the ligand pharma-
cophore similarity-target receptor complementarity
paradigm, the supermolecule represents the overall
shape and the conformational flexibility of the receptor
binding site.
presence of dicyclohexylcarbodiimide (DCC) and 4-(di-
methylamino)pyridine (DMAP) at room temperature
(method B).
On these bases, in order to validate the recently
proposed theoretical QSAR models, novel piperazine
(Table 1) and non-piperazine (Table 2) derivatives of 2,4-
diamino-6,7-dimethoxyquinazoline were designed, syn-
thesized, and tested for their binding affinity for the R1-
adrenoceptor in rat cortex. Selectivity toward other
G-protein-coupled receptors, namely, R2, D2, and 5-HT1A
sites, was also investigated. Moreover, the most inter-
esting compounds were submitted to enlarged receptor
binding screening on R1-adrenoceptor subtypes and in
vivo tests to evaluate their effects on the lower urinary
tract and the cardiovascular system. Finally, theoretical
descriptors calculated on a single structure and ad hoc
defined size and shape descriptors30,31 have been em-
ployed in order to capture, on a quantitative ground,
the molecular features responsible for the observed
variation in the R1-adrenoceptor binding affinity.
Compounds 3, 17, 20, 21, and 24-29 were synthe-
sized as shown in Scheme 2. 4-Amino-2-chloro-6,7-
dimethoxyquinazoline was reacted with a slight excess
of the appropriate cyclic or acyclic base in isoamyl
alcohol at reflux, optionally in the presence of potassium
iodide (method C). The benzyl derivative 20 was oxi-
dized with magnesium monoperoxyphthalate in metha-
nol to afford in high yield the N4-oxide derivative 22,
which was submitted to Meisenheimer thermal rear-
rangement conditions and afforded 23 in low yield, as
depicted in Scheme 3.
The 2-isopropyl-6-methoxyphenoxy acids 31, 33, and
35 were obtained by alkaline hydrolysis of the corre-
sponding ethyl esters 30, 32, and 34 (not isolated),
which were prepared by O-alkylation of the sodium salt
of 2-isopropyl-6-methoxyphenol with the suitable R-bro-
mo esters under phase-transfer conditions at 60-65 °C,
as depicted in Scheme 4.
Chloride 36 (Scheme 5) was prepared by chlorination
of the corresponding acid with SOCl2 and used to acylate
piperazine monohydrobromide in EtOH-THF mixture
at reflux temperature to afford the piperazinyl inter-
mediate 37 in moderate yield. In this last step a discrete
amount of the bis-acyl derivative was formed and easily
separated as an insoluble solid by acid treatment of the
reaction mixture. 37 was then used for the synthesis of
17.
Ch em istr y
The synthesis of 2-(1-piperazinyl)-4-amino-6,7-dimeth-
oxyquinazolines 6-16, 18, and 19 is summarized in
Scheme 1. The 2-(1-piperazinyl)-4-amino-6,7-dimethoxy-
quinazoline (5) was treated with a suitable acyl chloride,
in the presence of TEA as proton acceptor in CHCl3 or
DMF at room temperature, to afford derivatives 6, 8,
9, 13, 16, 18, and 19 (method A). Using Ac2O instead of
acetyl chloride, without TEA or solvent, the diacetyl
derivative 7 was obtained. Compounds 10-15 were
prepared by an alternative acylation procedure reacting
5 with the appropriate acids in DMF or CHCl3 in the

Novel Derivatives of R1-Adrenoceptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3 429
Sch em e 3
Sch em e 4
Sch em e 5
Resu lts a n d Discu ssion
Acylp ip er a zin e Der iva tives. Double acylation of pip-
erazine N4 and 4-NH2 (7) groups led to a loss of affinity
for compound 5, itself showing a low affinity. The
absence of affinity in compound 7 can be due to a
reduction in basicity of the anilino nitrogen (as shown
by the values of the nucleophilic superdelocalizability:
SL(NH2) ) 0.018 eV-1 for 5 and 0.011 eV-1 for 7) and/
or to violation of the steric requirements of the receptor
binding site.
When acylation is limited to the piperazine N4, a
consistent increase of affinity is evident and apparently
modulated by the bulkiness of the substituent, the
optimum being obtained when planar rings (1 and 9)
or mono- or disubstituted methylene units are in posi-
tion R to the carbonyl group. When this methylene unit
is trisubstituted, affinity is reduced to a different extent
(8 and 17) or almost lost (19).
The different nature of the substituents present in
compounds with high affinity for this receptor supports
the relative tolerance to bulkiness in this portion of the
molecules, which is challenged when methyl groups are
inserted at the R-carbonyl methylene and an isopropyl
group is simultaneously inserted at position 2 (6) of the
phenoxy ring (18 and 19).
2. Oth er P ip er a zin es. Alkylation of piperazine N4
with bulky groups improves affinity (20 and 21) also
but to a lesser extent than acylation with groups having
close steric requirements (9 and 12). Affinity is slightly
reduced when the benzyl group is replaced by a benz-
yloxy group (23) and is lost when N4 of compound 20 is
transformed into its N-oxide derivative. These results
suggest that changes in basicity of the piperazine N4
may play a relevant role in determining affinity for the
R1-adrenoceptor.
Affinity of all compounds for the R1-adrenoceptor was
determined in rat cortex32 and was found to be in the
nanomolar to micromolar range, as shown in Tables 1
and 2. Affinity for the native R2,33 D2,34 5-HT1A
,
and
35
36
5-HT2A receptors was also checked. All compounds
proved inactive at these sites (IC50 > 1000 nM) except
compound 14, showing IC50 ) 170 nM at the 5-HT1A
receptor.
The compounds endowed with high affinity for the
native R1-adrenoceptor were also tested on the three R1-
adrenoceptor subtypes of animal origin expressed in the
COS-7 cell line (Table 3). In agreement with the
literature, prazosin exhibited high affinity for the three
subtypes, with no selectivity. The absence of selectivity
was confirmed for all tested compounds, except com-
pound 16, which exhibited high affinity and moderate
selectivity for the R1b-adrenoceptor subtype. Despite its
relative absolute value, it is worth noting that selectivity
for the R1b subtype of compound 16 could be due to the
bulkiness of the aromatic group. Bulkiness, even if in a
different position, closer to the protonated N1 atom of
quinazoline, is also a common characteristic of two
selective R1b-adrenergic antagonists recentely re-
ported: (+)-cyclazosin ([4-(4-amino-6,7-dimethoxyquinazo-
lin-2-yl)-cis-octahydroquinoxalin-1-yl]furan-2-ylmetha-
none
hydrochloride34)
and
(S)-4-(4-amino-6,7-
dimethoxyquinazolin-2-yl)-N-(1,1-dimethylethyl)-1-(2-
furoyl)piperazine-2-carboxamide.37
On the basis of the overall lack of high subtype
selectivity and on the fact that affinity for the native
R1-adrenoceptor shown in Table 1 adequately reflects
the data from Table 3, other compounds were not
further tested for subtype selectivity. Thus the following
comments and QSAR analysis refer to the native R1-
adrenoceptor.
In fact, the trend in the electrophilic superdelocaliz-
ability on this atom suggests that the most active
ligands are less susceptible to an electrophilic attack
SAR for Na tive r1-Ad r en ocep tor Affin ity. 1. N-

430 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3
Leonardi et al.
Ta ble 1. Derivatives (5-23) of 2-(1-Piperazinyl)-4-amino-6,7-dimethoxyquinazoline
a
b
Mean of 2-3 different evaluations. SEM (always less than 15%) was omitted for clarity. See refs 8 and 39. c See ref 23.
(protonation) to this site: SH(N4) ) 8.38 × 10-4 and 9.74
× 10-4 eV-1 for the PhCO derivative 9 and the PhCH2
analogue 20, respectively, and 2.61 × 10-3 and 2.65 ×
10-3 eV-1 for 23 (R ) PhCH2O) and 5 (R ) H),
respectively.
Open chain or cyclized amines bearing phenyl groups
showed limited affinity (24, 25, and 29).
On the contrary, compound 28 proved to be a high-
affinity ligand for the R1-AR. It was designed on the
basis of the superposition of compound 3 to compound
2 (Figure 1). Its affinity was calculated in advance to
be in the nanomolar range by making use of a QSAR
model recently published26 and by comparison of its
reactivity determinants (SL(N) and ∆Eprot, Table 4) with
3. Oth er Am in es a t P osition 2 of th e Qu in a zolin e
Rin g. The attempt to realize a “hybrid” structure
bearing the substituted phenylpropylamino chain char-
acteristic of tamsulosin (26) gave an inactive compound.

Novel Derivatives of R1-Adrenoceptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3 431
Ta ble 2. 2-(Substituted-amino)-4-amino-6,7-dimethoxyquinazolines 3 and 24-29
a
b
Mean of 2-3 different evaluations. SEM (always less than 15%) was omitted for clarity. See ref 8. c See ref 15.
Ta ble 3. Affinity of Selected Compounds for R1-Adrenoceptor
The data values of the molecular descriptors used in
the selected correlation equations are listed in Table 4,
together with the cologarithmic form of the R1-AR
binding affinity (pIC50) of the considered ligands. An
exhaustive definition of the descriptor involved in the
selected QSAR models is given in the Computational
Procedures section. A quantitative overview of the
collinearities existing between the theoretical descrip-
tors of Table 4 is shown in Table 5, where the respective
intercorrelation coefficients are given. The descriptors
involved in the rationalization of the R1-adrenoceptor
binding affinity emphasize (a) the role of the quinazoline
N1-nitrogen atom as a hydrogen-bonding donor with
respect to a suitable amino acid residue of the receptor
(eqs 2 and 3), (b) the hydrogen-bonding acceptor pro-
pensity (eqs 4 and 5) of the ligands, and (c) the
dispersive and steric interactions realized by the quinazo-
line substituents (eq 1).
Subtypes
Ki (nM), animal clones
compd
R1a
R1b
R1d
10
11
13
14
15
16
18
1
10.99
5.68
1.48
18.95
36.20
7.5
155.6
0.72
5.43
4.43
1.57
8.56
9.05
0.45
8.17
19.38
1.59
18.95
23.72
10.34
112.8
1.39
14.7
0.46
respect to compounds 3 and 27. The QSAR model
derived in that paper refers to a supermolecule obtained
by superpositioning compounds 1-3 and all the acces-
sible conformations of the flexible compound 4 (Chart
1), chosen within 0.5 kcal mol-1 of its absolute minimum
(i.e., 19 conformers). The regression obtained in ref 26:
The critical role of the quinazoline protonated N1-
nitrogen is well-described by the valency of its hydrogen
atom Vf(H). In fact, the negative value of the regression
coefficient indicates that ligands with lower values for
minimum valency give stronger hydrogen bonds. The
Vf(H) values for compounds 7, 19, and 29 have not been
reported in Table 4 since their minimum valency is not
due to the protonated N1-nitrogen but to a substituent
hydrogen, whereas compound 8 has been omitted from
eq 2 since it is an outlier with overpredicted binding
affinity. The role of a long-range electrostatic interaction
as a preliminary recognition step in the ligand-receptor
complex formation is further emphasized by the nucleo-
philic index computed on the protonated nitrogen atom
(S1(N)) (eq 3). This index, which provides a numerical
differentiation of the capability of the ligand to donate
the proton, explains 60% of the variation in the binding
affinity, by omitting compounds 7, 24, and 25.
pA ) 5.321((1.801)Vdif - 2.426((0.679)
n ) 14, r ) 0.834, s ) 0.302, F ) 27.3
(a)
where pA is the negative logarithm of the binding
affinity of a generic compound normalized with respect
to prazosin (IC50 ) 2 nM), provided a pA value for
compound 28 (Vdif ) 0.4045) of -0.27, which corresponds
to an IC50 of 3.7 nM. The experimental result (IC50
1.6 nM) confirms the soundness of the model used to
predict these values.
Qu a n t it a t ive St r u ct u r e-Affin it y R ela t ion sh ip
An a lysis. Successful correlation equations can be ob-
tained for the R1-AR binding affinity by means of
theoretical descriptors derived on a single structure (i.e.,
molecular orbital indices and charged partial surface
area descriptors) and ad hoc derived descriptors com-
puted with respect to a supermolecule obtained by
superimposing the highest affinity ligands (see Com-
putational Procedures section).
)
Equation 1 involves the ad hoc derived size and shape
descriptor Vdif, which, by taking into account in its

432 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3
Leonardi et al.
F igu r e 1. Structural design of compound 28, based on the superimposition of compound 3 to compound 2.
Ta ble 4. Binding Affinities (pIC50) and Theoretical Descriptors of the Protonated Forms of the Quinazoline-like Ligands
HACA-1
(Å2)
HACA-1/TMSA
SL(N)
∆Eprot
(kcal/mol)
compd
pIC50
Vmol (Å3)
Vin (Å3)
Vout (Å3)
Vdif
Vf(H)
(×10-3
)
(eV-1
)
5
9
27
3
23
1
17
16
18
19
13
11
10
8
6.47
8.34
8.39
9.70
6.95
8.70
7.31
8.40
6.91
5.93
8.85
7.82
8.14
6.47
7.35
5.29
7.49
8.49
7.40
7.07
7.77
8.80
6.59
7.15
5.50
8.22
246.50
330.75
285.13
331.88
337.75
314.00
402.13
417.75
438.63
450.38
429.63
378.13
363.25
327.75
280.13
339.50
329.63
413.38
397.13
375.63
374.50
324.88
377.50
280.25
355.13
452.00
243.63
314.88
285.13
331.88
287.25
314.00
318.75
356.13
335.75
300.88
429.63
339.63
323.00
284.50
279.38
274.13
314.50
365.63
339.25
304.88
338.88
324.88
294.88
260.00
250.75
361.63
2.87
15.87
0.00
0.00
50.50
0.00
0.4528
0.5623
0.5362
0.6241
0.4452
0.5905
0.4426
0.5539
0.4379
0.2847
0.8080
0.5663
0.5317
0.4537
0.5240
0.3926
0.5630
0.5978
0.5291
0.4403
0.5703
0.6110
0.3992
0.4509
0.2753
0.5101
0.9253
0.9236
0.9224
0.9224
0.9241
0.9237
0.9243
0.9233
0.9242
3.4882
2.4363
2.3817
2.1270
2.1909
2.3759
2.4330
2.3125
2.4951
2.4386
2.3756
2.5686
2.3094
2.3792
2.3159
3.5630
2.1852
2.1894
2.1909
2.5053
2.3156
2.5698
2.7671
2.4433
2.5772
2.3156
6.05
3.24
3.74
2.97
2.95
3.30
3.00
2.65
2.82
2.81
2.69
3.04
2.91
3.43
3.57
5.06
3.04
2.56
2.67
3.18
2.75
3.67
3.58
3.83
3.22
2.67
0.0152
0.0164
0.0170
0.0168
0.0165
0.0163
0.0159
0.0165
0.0160
0.0154
0.0164
0.0164
0.0166
0.0161
0.0165
0.0170
0.0158
0.0162
0.0156
0.0143
0.0165
0.0167
0.0148
0.0147
0.0150
0.0162
-191.07
-188.70
-190.50
-189.12
-187.98
-189.13
-189.69
-188.54
-189.51
-192.53
-188.58
-188.85
-187.74
-189.54
-188.38
-187.84
-190.78
-189.25
-191.29
-190.31
-188.64
-190.68
-193.56
-190.24
-187.95
-189.80
83.38
61.62
102.88
149.50
0.00
38.50
40.25
43.25
0.75
65.37
15.13
47.75
57.88
70.75
35.62
0.00
82.62
20.25
104.38
90.37
0.9235
0.9237
0.9235
0.9240
0.9235
6
7
20
12
21
25
14
28
29
24
26
15
0.9242
0.9235
0.9240
0.9243
0.9233
0.9221
0.9247
0.9253
0.9233
formulation both the inner and outer molecular volumes
of the ligands considered with respect to the reference
volume of the supermolecule, indicates that once the
main docking has been accomplished the binding affin-
ity might be modulated by the optimization of the short-
range intermolecular interactions and dispersion con-
tributions of the quinazoline substituents. In this
respect, it is worth noting that the selected supermol-
ecule, being obtained by superimposing the highest
affinity ligands (1, 3, 13, and 28) for the native R1-
adrenoceptor, is a good template for the R1 binding site,
but not for the R1 subtypes. In fact, the selective R1b-

Novel Derivatives of R1-Adrenoceptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3 433
Ta ble 5. Correlation Matrix between Experimental Binding Affinity and Theoretical Descriptors Reported in Table 4
pIC50
Vmol
Vin
Vout
Vdif
Vf(H)
HACA-1
HACA-1/TMSA
SL(N)
∆Eprot
pIC50
Vmol
1.000
0.038
1.000
0.689
0.689
Vin
Vout
Vdif
0.648
-0.596
0.858
1.000
-0.051
0.725
1.000
-0.725
0.416
0.000
1.000
-0.647
-0.365
-0.265
0.520
Vf(H)
-0.838
-0.561
-0.415
0.487
-0.097
-0.296
-0.690
-0.035
-0.060
-0.478
-0.468
-0.667
0.353
1.000
0.466
0.336
-0.791
-0.048
HACA-1
HACA-1/TMSA
SL(N)
∆Eprot
0.061
1.000
0.871
-0.146
-0.104
-0.283
-0.401
-0.227
1.000
-0.113
-0.134
1.000
0.542
0.165
0.144
0.256
1.000
Ta ble 6. Regression Models for the Correlation between Theoretical Molecular Structure Descriptors and R1-Adrenoceptor Binding
Affinity
eq
1
descriptor
intercept
Vdif
intercept
Vf(H)
x ( Dx
t-test
n
R2
R2
s2
F
CV
3.23 ( 0.53
8.47 ( 1.03
905.4 ( 122.1
-971.8 ( 132.2
6.03
8.18
7.42
-7.35
omitted: 8
-3.82
5.76
26
22
0.736
0.730
0.651
0.655
0.324
0.250
66.96
54.05
2
3
intercept
-15.08 ( 3.94
1410.7 ( 244.8
SL(N)
23
0.613
0.536
0.439
33.20
omitted: 7, 24, and 25
4
5
intercept
Vdif
HACA-1
intercept
Vout
5.99 ( 1.06
7.44 ( 0.97
-0.90 ( 0.31
11.39 ( 0.59
-0.02 ( 0.003
-891.1 ( 160.2
5.65
7.66
-2.90
19.21
-6.80
-5.56
26
26
0.807
0.725
0.723
0.643
0.248
0.353
48.04
30.29
HACA-1/TMSA
adrenoceptor derivatives 16 and 18 are well-accomo-
dated by eq 1 (Table 6) for their native R1 binding
affinity, whereas no interpolation can be done for their
R1b-adrenoceptor binding affinity by making use of the
same equation.
A significant improvement in the stability of the
regression involving Vdif is obtained in eq 4 by including
the HACA-1 index, as underlined by the cross-validated
correlation coefficient (RCV2 ) 0.723). It is worth noting
that this descriptor contributes to the regression with
a negative sign; therefore proliferation of hydrogen bond
acceptor centers in the ligands seems to not be essential
for the modulation of the binding affinity. The HACA-1
index is also involved, together with Vout, in eq 5, where
it is normalized for the total molecular surface area of
the ligands. Notwithstanding this correlation shows
worse statistical parameters with respect to eq 4, the
predictive power for the ligands showing higher affinity
is notably improved (compare Figures 2 and 3, where
the experimental and calculated R1 binding affinity for
eq 4 (Table 6) and eq 5 (Table 6) are plotted).
F u n ction a l Selectivity. Compounds 10, 11, 13, and
16 were tested in a dog model38 aiming to evaluate the
effects on stimulated urethral contractility in parallel
to hypotension. Prazosin was tested as a reference
standard, as shown in Table 7. None of the tested
compounds showed selectivity for the lower urinary
tract. Compound 10 was the only derivative showing
in vivo potency close to that of prazosin.
F igu r e 2. Plot of the predicted versus experimental R1 binding
affinity for eq 4.
tive effect given by acylation of the 4-amino group and
the modulating role of the piperazine N4 susceptibility
to protonation. Tolerance to bulky substituents intro-
duced at the piperazine N4 was also confirmed, with
some exceptions. Moreover, the use of theoretical de-
scriptors derived on a single structure and ad hoc
defined size and shape descriptors in deriving QSAR
models provided elucidation of the role of the main
pharmacophoric components for R1-AR recognition and
binding. In fact, the electrostatic interactions between
the protonated amine function and a primary nucleo-
philic site of the receptor necessary for recognition are
described by the nucleophilic superdelocalizability and
the free valency of the protonated N1-nitrogen atom,
while the short-range attractive and repulsive intermo-
lecular interactions mainly responsible for the modula-
tion of the binding affinities are described by charged
partial surface area descriptors computed on the whole
molecules (polar forces) and by ad hoc defined size and
Con clu sion s
In addition to confirmation of the pivotal role of
protonation of N1 in the quinazoline ring, the obtained
results allowed us to define other relevant features for
the affinity of 4-amino-2-(1-piperazinyl)-6,7-dimethoxy-
quinazolines for the R1-adrenoceptor, such as the nega-

434 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3
Leonardi et al.
solid: mp 244-245 °C; 1H NMR (DMSO-d6) δ 2.03 (s, 3H),
3.35-3.55 (m, 4H), 3.60-3.80 (m, 4H), 3.79 (s, 3H), 3.83 (s,
3H), 6.75 (s, 1H) 7.17 (bs, 2H), 7.44 (s, 1H).
1-Acet yl-4-(4-a cet yla m in o-6,7-d im et h oxy-2-q u in a zo-
lin yl)p ip er a zin e (7). A mixture of 5 (4.63 g, 0.016 mol) and
Ac2O (100 mL) was stirred at reflux for 1 h. The reaction was
poured into H2O (300 mL), was alkalinized with 12 N NaOH
(190 mL) in 15 min, and after cooling to room temperature
was extracted with CHCl3 (3 × 140 mL). The organic layer
was washed with H2O (2 × 100 mL) and dried (anhydrous
Na2SO4) and the solvent evaporated to afford a crude that was
purified by flash chromatography (CHCl3-MeOH gradient
100:1 to 100:5) followed by crystallization from DMF-H2O (1:
2) to afford 3.7 g (62%) of a white solid: mp >270 °C; 1H NMR
(DMSO-d6) δ 2.06 (s, 3H), 2.41 (s, 3H), 3.45-3.60 (m, 4H),
3.71-3.81 (m, 4H), 3.84 (s, 3H), 3.89 (s, 3H), 6.90 (s, 1H), 7.47
(s, 1H), 10.46 (s, 1H).
The following compounds were prepared by method A,
above.
1-(4-Am in o-6,7-dim eth oxy-2-qu in azolin yl)-4-pivaloylpip-
er a zin e h yd r och lor id e‚0.75H2O (8): 1H NMR (DMSO-d6) δ
1.25 (s, 9H), 3.30-4.30 (m, 8H), 3.80 (s, 7.5H), 7.45 (s, 1H),
7.60 (s, 1H), 8.20-8.80 (bs, 2H).
F igu r e 3. Plot of the predicted versus experimental R1 binding
affinity for eq 5.
Ta ble 7. In Vivo Effects of Selected Compds after Intravenous
Administration in the Dog Model
1-(4-Am in o-6,7-dim eth oxy-2-qu in azolin yl)-4-(3,3-diph en -
ylp r op ion yl)p ip er a zin e h yd r och lor id e (13): 1H NMR (DM-
SO-d6) δ 3.80 (s, 6H), 3.00-4.10 (m, 10H), 4.20-4.70 (m, 1H),
6.90-7.30 (m, 10H), 7.50-7.80 (m, 2H), 8.10-9.20 (bb, 2H),
12.00-12.60 (bb, 1H).
UP
DBP
(ED25
ratio
(DBP/UP)
compd (ED50
)
)
10
10.6
35
31
65
3.6
(6.8-16.4)
(27.6-44.7)
(25-37)
8.1
132
23
183
6.6
(6.7-9.8)
(73-240)
(18-29)
(96-347)
(4.2-10)
0.7
11
3.77
0.74
2.82
1.83
13a
16a
1a
1-(4-Am in o-6,7-d im eth oxy-2-qu in a zolin yl)-4-[(2-isop r o-
p yl-6-m eth oxyp h en oxy)a cetyl]p ip er a zin e h yd r och lor id e
(51-82)
(3.1-4.2)
1
(16): H NMR (DMSO-d6) δ 1.30 (d, 6H), 3.10-4.90 (m, 9H),
a
These data have already been published, see ref 55.
4.30 (s, 3H), 4.35 (s, 3H), 5.15 (s, 2H), 7.40-7.90 (m, 3H), 8.50-
8.80 (m, 2H), 9.50-10.00 (bb, 2H), 13.70-14.70 (bb, 1H).
1-(4-Am in o-6,7-d im e t h oxy-2-q u in a zolin yl)-4-[(2-iso-
p r op yl-6-m eth oxyp h en oxy)p r op ion yl]p ip er a zin e h yd r o-
shape descriptors (dispersive and steric forces). Finally,
the predictive ability of previously reported QSAR
models obtained by employing ad hoc derived size and
shape descriptors has been corroborated by the experi-
mental binding affinity of compound 28.
1
ch lor id e‚0.4H2O (18): H NMR (DMSO-d6) δ 1.10-1.70 (m,
9H), 3.50-5.00 (m, 10.8H), 4.30-4.35 (m, 9H), 7.60-8.10 (m,
3H), 8.70-9.00 (m, 2H), 9.50-10.60 (bb, 2H), 13.60-15.00 (bb,
1H).
1-(4-Am in o-6,7-d im eth oxy-2-qu in a zolin yl)-4-[(2-isop r o-
p yl-6-m et h oxyp h en oxy)-2-m et h ylp r op ion yl]p ip er a zin e
h yd r och lor id e (19): 1H NMR (DMSO-d6) δ 1.50 (d, 6H), 1.75
(s, 6H), 3.50-5.30 (m, 9H), 4.45 (s, 3H), 4.70 (s, 6H), 8.00-
8.60 (m, 3H), 9.10-9.50 (m, 2H), 9.80-11.20 (bb, 2H), 14.70-
15.30 (bb, 1H).
Exp er im en ta l Section
Ch em istr y. Melting points were determined with a Buchi
535 apparatus and are uncorrected. 1H NMR spectra were
recorded on a Brucker AC 200; chemical shifts are reported
as δ values relative to tetramethylsilane as internal standard.
Splitting patterns are designated as follows: s, singlet; d,
doublet; t, triplet; q, quartet; and m, multiplet. Analytical TLC
was performed on silica 60 F254 plates (Merck), and the spots
were made visible by a UV lamp or iodine vapor or were
sprayed with Dragendorff’s reagent. Flash chromatography
was performed on silica gel (Merck 70-230 mesh). Analyses
indicated by the symbols of the elements were within (0.4%
of the theoretical values and were performed by Redox s.n.c.
Unless stated otherwise, starting materials were used as high-
grade commercial products.
The compound 5 was synthesized by a published procedure.
The physical data are in agreement with those given in ref
39. Intermediates N-methyl-3,3-diphenylpropylamine40 used
for 25, (R)-1-methyl-2-(3-aminosulfonyl-4-methoxyphenyl)eth-
ylamine41 used for 26, 1,2,3,4-tetrahydrobenz[f]isoquinoline42
used for 28, and 4,4-diphenylpiperidine43 used for 29 were
synthesized by published procedures.
Meth od A. 1-Acetyl-4-(4-am in o-6,7-dim eth oxy-2-qu in azo-
lin yl)p ip er a zin e Hyd r och lor id e‚2.5H2O (6). To a suspen-
sion of 5 (5.78 g, 0.02 mol), CHCl3 (120 mL), and triethylamine
(4.86 mL, 0.0347 mol) was dropped in 15 min a solution of
acetyl chloride (2.24 mL, 0.031 mol) in CHCl3 (20 mL). The
mixture was stirred at 22-25 °C for 8 h and then was washed
with H2O (3 × 50 mL); the organic layer was dried (anhydrous
Na2SO4) and evaporated to dryness in vacuo. The residue was
suspended into boiling EtOH (150 mL), and ethanolic HCl was
added to the mixture obtaining a solution followed by precipi-
tation of the solid, which was collected and crystallized from
EtOH affording 5.5 g (66%) of the desired product as a white
Meth od B. 1-(4-Am in o-6,7-d im eth oxy-2-qu in a zolin yl)-
4-(ben zoyla cetyl)p ip er a zin e (10). To a suspension of 97%
DCC (10.5 g, 0.05 mol), DMAP (0.37 g, 0.003 mol), 5 (8.7 g,
0.03 mol), and anhydrous CHCl3 (140 mL) was dropped in 3
min a solution of benzoylacetic acid (8.2 g, 0.05 mol) in
anhydrous CHCl3 (30 mL). The mixture was stirred at 22-25
°C for 6 h and then was diluted with CH2Cl2 (200 mL) and
MeOH (16 mL), and the insoluble DCU was filtered off. The
solvents were evaporated to dryness, and the residue was
purified by flash chromatography (CH2Cl2-MeOH, 100:3)
followed by crystallization from ACN to afford 7.8 g (60%) of
1
a white solid: mp 214-215 °C; H NMR (DMSO-d6) δ 3.00-
4.00 (m, 8H), 2.75 (s, 6H), 4.20 (s, 2H), 6.65 (s, 1H), 6.80-7.10
(bs, 2H), 7.30 (s, 1H), 7.20-7.50 (m, 3H), 7.60-7.90 (m, 2H).
The following compounds were prepared by method B,
above.
1-(4-Am in o-6,7-d im et h oxy-2-q u in a zolin yl)-4-[(3-b en -
zoyl)p r op ion yl]p ip er a zin e h yd r och lor id e m on oh yd r a te
1
(11): H NMR (DMSO-d6) δ 2.50-2.90 (m, 2H), 3.80 (s, 6H),
3.00-4.00 (m, 12H), 7.20-7.90 (m, 7H), 8.20-8.90 (bb, 2H),
11.50-12.50 (bb, 1H).
1-(4-Am in o-6,7-d im e t h oxy-2-q u in a zolin yl)-4-(2,2-d i-
p h en yla cetyl)p ip er a zin e h yd r och lor id e‚0.75H2O (12): 1H
NMR (DMSO-d6) δ 3.20 (s, 1.5H), 3.30 (s, 6H), 3.40-4.00 (m,
8H), 5.50 (s, 1H), 7.10 (s, 10H), 7.50 (s, 1H), 7.60 (s, 1H), 8.20-
8.80 (bb, 2H), 11.20-12.70 (bb, 1H).
1-(4-Am in o-6,7-d im eth oxy-2-qu in a zolin yl)-4-[(2-m eth -
oxyp h en oxy)a cetyl]p ip er a zin e h yd r och lor id e (14): 1H
NMR (DMSO-d6) δ 2.90-4.20 (m, 8H), 3.70 (s, 3H), 3.80 (s,

Novel Derivatives of R1-Adrenoceptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3 435
1
6H), 4.75 (s, 2H), 6.75 (s, 4H), 7.50 (s, 1H), 7.60 (s, 1H), 8.30-
8.70 (bb, 2H), 11.70-12.30 (bb, 1H).
(17.26); H2O, 1.13 (0.79); H NMR (DMSO-d6) δ 2.80 (d, 2H),
3.30 (t, 2H), 3.56 (t, 2H), 3.78 (s, 3H), 3.83 (s, 3H), 4.33 (s,
2H), 4.50 (s, 2H), 6.74 (s, 1H), 7.15-7.35 (bs, 2H), 7.30-7.45
(m, 3H), 7.45 (s, 1H), 7.50-7.65 (m, 2H).
Meth od C. 1-(4-Am in o-6,7-d im eth oxy-2-qu in a zolin yl)-
4-[2-(2-m e t h oxyp h e n oxy)-2-m e t h ylp r op ion yl]p ip e r a -
zin e Hyd r och lor id e (17). A mixture of 37 (2.78 g, 0.01 mol),
4-amino-2-chloro-6,7-dimethoxyquinazoline (2.39 g, 0.01 mol),
and isoamyl alcohol (50 mL) was stirred at reflux temperature
for 5 h. After cooling to 0-5 °C the mixture was kept resting
for 30 min; then the precipitate was collected by filtration,
washed first with isoamyl alcohol and second with acetone.
The crude was suspended in H2O (50 mL) and CHCl3 (50 mL),
alkalinized with 20% aqueous Na2CO3, and stirred at 22-25
°C until two phases became clear. The organic layer was
washed with H2O, dried (Na2SO4), and evaporated to dryness.
The residue was purified by flash chromatography (EtOAc-
MeOH gradient 10:0 to 10:1). The crude base (4.45 g) was
dissolved in boiling EtOH (90 mL), the solution acidified with
ethanolic HCl, and the solid crystallized from 80% EtOH to
1-(4-Am in o-6,7-d im eth oxy-2-qu in a zolin yl)-4-(ben zylox-
y)p ip er a zin e (23). Compound 22 (2.2 g, 0.0056 mol) was
heated at 220-225 °C for 5 min. The brown crude was purified
by flash chromatography (CHCl3-5 N methanolic NH3, 100:
1.5) followed by crystallization from ACN to afford 0.24 g (11%)
of a white solid: mp 179-180 °C; 1H NMR (DMSO-d6) δ 2.48
(d, 2H), 2.97 (t, 2H), 3.29 (t, 2H), 3.78 (s, 3H), 3.82 (s, 3H),
4.57 (d, 2H), 4.71 (s, 2H), 6.73 (s, 1H), 7.13 (bs, 2H), 7.31-
7.37 (m, 5H), 7.42 (s, 1H). For the interpretation of this
spectrum, it was very helpful looking up the source of ref 44.
Eth yl 2-Meth oxy-6-isop r op ylp h en oxya ceta te (30). To
a stirred suspension of 50% NaH (0.6 g, 0.0125 mol) in
anhydrous benzene (50 mL) was dropped in 15 min at 22-25
°C a solution of 2-methoxy-6-isopropylphenol45 (1.7 g, 0.010
mol) in anhydrous benzene (20 mL). After 30 min of stirring
at 22-25 °C and 3 h at reflux, H2O (1.2 mL) was added to the
mixture and the solvents were evaporated in vacuo. The green
residue was dissolved into isobutyl methyl ketone (50 mL), and
a solution of ethyl bromoacetate (2.3 mL, 0.014 mol) in isobutyl
methyl ketone (5 mL) was added. The mixture was stirred at
22-25 °C for 2 h and at 100-105 °C for 14 h. After that period,
ethyl bromoacetate (0.67 mL, 0.006 mol) was added, and
stirring at 100-105 °C was continued for an additional 8 h.
After cooling to 22-25 °C, the mixture was cautiously diluted
with 2 N HCl (30 mL), the organic phase was separated, and
the aqueous phase was extracted with Et2O. The residue
obtained after evaporation of the reunited organic phases was
purified by flash chromatography (n-hexane-Et2O gradient
100:5 to 100:20) to afford 1.78 g (71%) of a yellowish thick oil:
1H NMR (CDCl3) δ 1.25 (d, 6H), 1.40 (t, 3H), 3.20-3.90 (m,
1H), 4.00 (s, 3H), 4.20-4.70 (m, 2H), 4.80 (s, 2H), 6.90-7.40
(m, 3H).
2-Met h oxy-6-isop r op ylp h en oxya cet ic Acid (31). To a
well-stirred mixture of NaOH (20 g, 0.5 mol), TEBA (1.1 g,
0.005 mol), 2-methoxy-6-isopropylphenol (8.4 g, 0.05 mol), H2O
(30 mL), and toluene (40 mL) was dropped in 15 min a solution
of ethyl bromoacetate (11.1 mL, 0.1 mol) in toluene (10 mL).
The mixture was stirred at 22-25 °C for 2 h, at 60-65 °C for
2 h, and at reflux for 1.5 h. After that, a solution of ethyl
bromoacetate (5.5 mL, 0.05 mol) in toluene (10 mL) was added
to the mixture, and refluxing was continued for an additional
5 h. After cooling to room temperature, H2O (250 mL) was
added and the phases were separated. The aqueous phase was
washed with Et2O (3 × 50 mL), acidified with 37% HCl, and
extracted with Et2O (3 × 50 mL). The organic phase was
washed with H2O (3 × 30 mL) and extracted with 20% Na2CO3
solution (40 mL) stirring at room temperature for almost 15
min. The alkaline phase was washed with Et2O (2 × 30 mL),
acidified with 37% HCl, and extracted with Et2O (3 × 40 mL).
The ethereal solution was dried (Na2SO4) and evaporated to
dryness to afford 8 g (72%) of 31 as a yellowish oil. The
analytical sample was obtained by distillation (190 °C/7
mmHg): 1H NMR (CDCl3) δ 1.25 (d, 6H), 3.20-3.70 (m, 1H),
3.90 (s, 3H), 4.65 (s, 2H), 6.70-7.40 (m, 3H), 10.30-10.60 (bs,
1H).
1 -[2 -(2 -M e t h o x y p h e n o x y )-2 -m e t h y l p r o p i o n y l ]-
p ip er a zin e‚HCl‚H2O (37). To a boiling solution of 2-(2-
methoxyphenoxy)-2-methylpropionic acid46 (10.5 g, 0.05 mol)
in anhydrous CHCl3 (50 mL) was dropped within 30 min a
solution of SOCl2 (5.4 mL, 0.074 mol) in anhydrous CHCl3 (20
mL). The mixture was stirred at reflux for 2 h and then was
evaporated to dryness to give 11.5 g (100%) of the 2-(2-
methoxyphenoxy)-2-methylpropionyl chloride as a light-brown
oil: 1H NMR (CDCl3) δ 1.70 (6H, s), 4.00 (3H, s), 7.10-7.50
(4H, m). To a stirred solution of anhydrous piperazine (12.9
g, 0.15 mol), 95% EtOH (50 mL), and H2O (15 mL) was added
at room temperature 48% HBr (25.3 g, 0.15 mol). After cooling
to 10 °C a solution of the above acid chloride (11.5 g, 0.05 mol)
in THF (50 mL) was dropped in 30 min without exceeding 20
°C. The mixture was stirred for 1.5 h at the same temperature
1
afford 2.93 g (57%) of a white solid: mp 288 °C dec; H NMR
(DMSO-d6) δ 1.60 (s, 6H), 3.30-3.70 (m, 2H), 3.98 (s, 3H), 4.02
(s, 6H), 3.70-4.03 (m, 8H), 7.00-7.40 (m, 4H), 7.96 (s, 1H),
8.08 (s, 1H), 8.70-9.30 (bs, 1H).
The following compounds were prepared by method C, as
above.
1-(4-Am in o-6,7-d im eth oxy-2-qu in a zolin yl)-4-ben zylp ip-
er azin e dih ydr och lor ide em ih ydr ate (20): 1H NMR (DMSO-
d6) δ 3.00-3.60 (m, 4H), 3.85 (s, 7H), 4.00-4.80 (m, 6H), 7.40
(s, 5H), 7.50 (s, 1H), 7.60 (s, 1H), 8.40-9.10 (bb, 2H).
1-(4-Am in o-6,7-d im et h oxy-2-q u in a zolin yl)-4-b en zh y-
d r ylp ip er a zin e d ih yd r och lor id e em ih yd r a te (21): 1H
NMR (DMSO-d6) δ 2.20-2.60 (m, 4H), 3.50-4.00 (m, 4H), 3.75
(s, 3H), 3.80 (s, 3H), 4.10 (s, 1H), 5.25-5.45 (bs, 2H), 6.70 (s,
1H), 6.80 (s, 1H), 6.90-7.40 (m, 10H).
4-Am in o-6,7-d im e t h oxy-2-(N -b e n zylm e t h yla m in o)-
qu in a zolin e h yd r och lor id e (24): 1H NMR (DMSO-d6) δ 3.30
(s, 3H), 3.90 (s, 6H), 5.00 (s, 2H), 7.30 (s, 5H), 7.75 (s, 1H),
7.85 (s, 1H), 8.50-8.85 (bs, 2H), 11.50-12.40 (bb, 1H).
4-Am in o-6,7-d im e t h oxy-2-[N -(3,3-d ip h e n ylp r op yl)-
m eth yla m in o]qu in a zolin e h yd r och lor id e (25): 1H NMR
(DMSO-d6) δ 2.30-2.80 (m, 2H), 3.20 (s, 3H), 3.40-4.30 (m,
3H), 3.90 (s, 6H), 6.80-7.40 (m, 10H), 7.70 (s, 2H), 8.00-8.80
(bb, 2H), 11.20-11.70 (bb, 1H).
(R)-4-Am in o-6,7-d im eth oxy-2-[1-m eth yl-2-(3-a m in osu l-
fon yl-4-m eth oxyp h en yl)eth yla m in o]qu in a zolin e h yd r o-
ch lor id e‚0.8H 2O (26): 1H NMR (DMSO-d6) δ 1.25 (d, 3H),
2.80-3.20 (m, 3.6H), 3.90 (s, 9H), 4.10-4.70 (m, 1H), 6.90-
7.20 (m, 4H), 7.30-8.20 (m, 5H), 8.30-9.10 (bb, 2H).
4-Am in o-6,7-dim eth oxy-2-(1,2,3,4-tetr ah ydr oben z[f]iso-
qu in olin -2-yl)qu in a zolin e‚0.25C2H5OH (28): 1H NMR (DM-
SO-d6) δ 3.17 (t, 2H), 3.79 (s, 3H), 3.85 (s, 3H), 4.17 (t, 2H),
5.02 (s, 2H), 6.80 (s, 1H), 7.19 (s, 2H), 7.36 (d, 1H), 7.44 (s,
1H), 7.48-7.55 (m, 2H), 7.75-8.02 (m, 3H).
4-Am in o-6,7-d im eth oxy-2-(4,4-d ip h en yl-1-p ip er id in yl)-
qu in a zolin e h yd r och lor id e‚0.65H2O (29): 1H NMR (DMSO-
d6) δ 2.20-2.70 (m, 4H), 3.30-4.00 (m, 5.3H), 3.80 (s, 6H),
6.80-7.30 (m, 10H), 7.50-7.75 (m, 2H), 8.30-8.75 (bs, 2H),
10.00-11.00 (bb, 1H).
1-(4-Am in o-6,7-d im eth oxy-2-qu in a zolin yl)-4-ben zylpip -
er a zin e N4-Oxid e‚2HCl‚0.75H2O (22). To a stirred solution
of 20 (5.32 g, 0.014 mol) in MeOH (70 mL) was dropped in 30
min at 0-5 °C a solution of 83% Mg monoperoxyphthalate‚
6H2O (4.34 g, 0.0073 mol) in H2O (40 mL). After 1 h of stirring
the mixture was evaporated to dryness, and the residue was
purified by flash chromatography (CHCl3-5 N methanolic
NH3, 100:3). The crude base was dissolved into EtOH (50 mL),
the solution acidified with ethanolic HCl and diluted with Et2O
(100 mL), and the precipitate collected by filtration and dried
to afford 5 g (86%) of a white solid: mp 219-221 °C; 1H NMR
(DMSO-d6) δ 3.80 (s, 6H), 3.00-4.50 (m, 6H), 4.50-5.40 (m,
4H), 7.10-7.60 (m, 7H), 8.30-9.20 (bs, 2H), 11.40-12.90 (bb,
1H). After D2O addition all exchangeable protons were evalu-
ated as HDO peak. A sample of the crude base was crystallized
from EtOAc to give a white solid: mp 188 °C; for C21H25N5O3‚
0.25H2O, calcd (found) C, 63.06 (63.19); H, 6.43 (6.41); N, 17.51

436 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3
Leonardi et al.
and then for 3 h at reflux, diluted with THF (50 mL), and
cooled to 0-5 °C for 30 min. The piperazine salts were
separated by filtration, the filtrate was evaporated to dryness,
and the residue was treated with 1 N HCl (100 mL) and Et2O
(100 mL). The insoluble 1,4-bis[2-(2-methoxyphenoxy)-2-meth-
ylpropionyl]piperazine was separated by filtration; the aqueous
phase was neutralized with NaOH, washed with Et2O (4 ×
50 mL), alkalinized with concentrated NaOH, and extracted
with Et2O (5 × 50 mL). The organic phase was acidified with
HCl in EtOH and evaporated to dryness. The residue was
crystallized from MEK to give 5.6 g (33%) of 37 as a white
ecule constituted by the most structurally different ligands
which show the highest binding affinity for the R1-adrenoceptor
(1, 3, 13, and 28). Vdif is computed according to the following
formula: Vdif ) Vin - Vout/Vsup, where Vsup is the resultant van
der Waals volume of the reference supermolecule (531.75 Å3).
According to the ligand pharmacophore similarity-target
receptor complementarity paradigm, this approach assumes
that the volume obtained by superimposing the most structur-
ally different ligands which show the highest affinities for the
same receptor might reflect the overall shape and the confor-
mational flexibility of the high-affinity receptor binding site.
1
solid: mp 95-98 °C; H NMR (CDCl3) δ 1.60 (s, 6H), 2.90-
3.50 (m, 4H), 3.90 (s, 3H), 3.90-4.60 (m, 4H), 6.80-7.60 (m,
Ack n ow led gm en t. M.C.M. and P.G.D.B. are grate-
ful to Prof. Mati Karelson for the CODESSA program.
The authors would like to thank Mr. Roberto Cappelletti
for obtaining NMR data.
6H).
Com p u ta tion a l P r oced u r es. Geom etr y op tim iza tion :
Conformational analysis of the N1-protonated form of some
quinazoline derivatives was recently performed,26 and the
resulting absolute minimum structure of prazosin (1) was
considered in this study as the starting geometry. The sub-
stituents were constructed within the Editor module of the
Quanta96 program.47 Then, the protonated structures were
fully optimized by means of molecular orbital calculations
(AM1),48 using the MOPAC 6.0 (QCPE 455) program.
Molecu la r su p er im p osition : The most structurally dif-
ferent ligands which show the highest affinity for the R1-
adrenoceptor (1, 3, 13, and 28) were chosen for the construction
of the reference supermolecule and were superimposed, by a
rigid fit procedure on the quinazoline moiety. The Quanta96
molecular modeling software was utilized for molecular com-
parisons, matching, and computation of van der Waals vol-
umes.
Molecu la r d escr ip t or s a n d st a t ist ica l t r ea t m en t of
d a ta : The MOPAC output files were loaded into the CODESSA
program49 along with the R1-adrenoceptor binding affinity data
values. A large number of global and fragment descriptors
(constitutional, topological, electrostatic, geometrical, and
quantum-chemical) were generated for each compound. The
ad hoc defined molecular size and shape parameters, described
above, were added as external descriptors.31,50
Refer en ces
(1) Ruffolo, R. R., J r.; Stadel, J . M.; Hieble, J . P. R-Adrenoceptors:
Recent Developments. Med. Res. Rev. 1994, 14, 229-270.
(2) Ruffolo, R. R., J r. Chirality in R- and â-Adrenoceptor Agonists
and Antagonists. Tetrahedron 1991, 48, 9953-9980.
(3) De Marinis, R. M.; Wise, M.; Hieble, J . P.; Ruffolo, R. R., J r.
Structure-Activity Relationships for Alpha-1 Adrenergic Recep-
tor Agonists and Antagonists. In The R1-Adrenergic Receptor;
Ruffolo, R. R., J r., Ed.; Humana Press: Clifton, NJ , 1987; pp
211-265.
(4) Vizi, E. S. Compounds Acting on Alpha1- and Alpha2-Adreno-
ceptors: Agonists and Antagonists. Med. Res. Rev. 1986, 6, 431-
449.
(5) Timmermans, P. B. M. V. M.; De J onge, A.; Thoolen, M. J . M.
C.; Wilffert, B.; Batink, H.; van Zwieten, P. A. Quantitative
Relationships between R-Adrenergic Activity and Binding Af-
finity of R-Adrenoceptor Agonists and Antagonists. J . Med.
Chem. 1984, 27, 495-503.
(6) Melchiorre, C. Selectivity of R1 and R2 Adrenergic Agonists and
Antagonists. II Farmaco Ed. Sci. 1980, 35, 535-550.
(7) Leonardi, A.; Testa, R.; Motta, G.; De Benedetti, P. G.; Hieble,
P.; Giardina`, D. R1-Adrenoceptors: Subtype- and Organ-
Selectivity of Different Agents. In Perspective in Receptor
Research; Giardina`, D., Piergentili, A., Pigini, M., Eds.; Elsevier
Science B.V.: Amsterdam, 1996; pp 135-152.
The search for the best correlation equation was achieved
by means of the heuristic method, which accomplished a
preselection of descriptors on the basis of their statistical
significance.51,52 Default values for control parameters and
criteria were used: minimum squared correlation coefficient
to consider one-parameter correlation significant, Rmin ) 0.1;
t-test value to consider descriptor significant in one-parameter
correlation, t1 ) 1.5; t-test value to consider descriptor
significant in multiparameter correlation, t2 ) 3.0; highest pair
correlation coefficient of two descriptors scales, rfull ) 0.99;
significant intercorrelation level, rsig ) 0.80. Evaluation of the
best correlation models was carried out by validation of each
model by cross-validation techniques. The QSAR models
reported in this paper were selected on the basis of the best
statistical parameters and the largest diversity of the descrip-
tors involved.
Defin ition of th e d escr ip tor s53 u sed in th e selected
QSAR m od els: SL(N) is the nucleophilic superdelocalizability
on the protonated nitrogen atom. Being computed on the
lowest unoccupied molecular orbital (LUMO), it characterizes
the hydrogen bond donor propensity. Vf(H), the free valency
for the H atom of the protonated nitrogen, characterizes the
hydrogen-bonding donor capability of the ligands. ∆Eprot is the
energy difference between the protonated and neutral forms
of the ligands considered in the vapor phase. HACA-154 is
calculated according to the following formula: HACA-1 )
SAqASA, where qA is the charge on the hydrogen-bonding
acceptor atom and SA is its surface area. The atomic partial
charges used in this formula were calculated using a Zefirov
empirical method. The HACA-1/TMSA parameter represents
the ratio of the surface area of the hydrogen acceptor atoms
to the total surface area of the molecule (TMSA). Vin and Vout
are ad hoc defined size and shape descriptors26 and represent,
respectively, the intersection and the outer van der Waals
volumes of the ligands considered with respect to a supermol-
(8) Campbell, S. F. Structure-activity Relationships for Quinazoline
R-Adrenoceptor Antagonists. In X-ray Crystallography and Drug
Action; Horn, A. S., De Rauter, C. J ., Eds.; Clarendon: Oxford,
1984; pp 347-366.
(9) Campbell, S. F. In Second SCI-RSC Medicinal Chemistry
Symposium; Emmet, J . C., Ed.; Royal Society of Chemistry:
Letchworth, 1984; p 18.
(10) Campbell, S. F.; Davey, M. J .; Hardstone, J . D.; Lewis, B. N.;
Palmer, M. J . 2,4-Diamino-6,7-dimethoxyquinazolines. 1. 2-[4-
(1,4-Benzodioxan-2-ylcarbonyl)piperazin-1-yl] Derivatives as R-A-
drenoceptor Antagonists and Antihypertensive Agents. J . Med.
Chem. 1987, 30, 49-57.
(11) Campbell, S. F.; Plews, R. M. 2,4-Diamino-6,7-dimethoxyquinazo-
lines. 3. 2-(4-Heterocyclylpiperazin-1-yl) Derivatives as R1-
Adrenoceptor Antagonists and Antihypertensive Agents. J . Med.
Chem. 1987, 30, 1794-1798.
(12) Alabaster, V. A.; Campbell, S. F.; Danilewicz, J . C.; Greengrass,
C. W.; Plews, R. M. 2,4-Diamino-6,7-dimethoxyquinazolines. 2.
2-(4-Carbamoylpiperidino) Derivatives as R1-Adrenoceptor An-
tagonists and Antihypertensive Agents. J . Med Chem. 1987, 30,
999-1003.
(13) Campbell, S. F.; Danilewicz, J . C.; Greengrass, C. W.; Plews, R.
M. 2,4-Diamino-6,7-dimethoxyquinazolines. 4. 2-[4-(Substituted
oxyethoxy)piperidino] Derivatives as R1-Adrenoceptor Antago-
nists and Antihypertensive Agents. J . Med. Chem. 1988, 31,
516-520.
(14) Bordner, J .; Campbell, S. F.; Palmer, M. J .; Tute, M. S.
1,3-Diamino-6,7-dimethoxyquinazoline Derivatives as Potential
R1-Adrenoceptor Antagonists. J . Med. Chem. 1988, 31, 1036-
1039.
(15) Danilewicz, J . C.; Kenp, J . E. G.; Wright, J . R. U.K. Patent
Specification 1383409, Sept 9, 1972.
(16) Sekiya, T.; Hiranuma, H.; Hata, S.; Mizogami, S.; Hanazuka,
M.; Yamada, S. Pyrimidine Derivatives. 4. Synthesis and Anti-
hypertensive Activity of 4-Amino-2-(4-cinnamoylpiperazino)-6,7-
dimethoxyquinazoline Derivatives. J . Med. Chem. 1983, 26,
411-416.
(17) Manoury, P. M.; Binet, J . L.; Dumas, A. P.; Le`fevre-Borg, F.;
Cavero, I. Synthesis and Antihypertensive Activity of a Series
of 4-Amino-6,7-dimethoxyquinazoline Derivatives. J . Med. Chem.
1986, 29, 19-25.

Novel Derivatives of R1-Adrenoceptor Antagonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 3 437
(18) Carenzi, A.; Chiarino, D.; Napoletano, M.; Reggiani, A.; Sala,
A.; Sala, R. New Isoxazole Derivatives Provided with Antihy-
pertensive Activity. Arzneim.-Forsch./ Drug Res. 1989, 39, 642-
646.
(19) Rampa, A.; Valenti, P.; Da Re, P.; Carrara, M.; Zampiron, S.;
Cima, L.; Giusti, P. Some New Prazosin Analogues. Arch.
Pharm. (Weinheim) 1989, 322, 359-361.
(20) Rampa, A.; Carrara, M.; Zampiron, S. Some New Prazosin-like
Derivatives. Boll. Chim. Farm. 1989, 128, 129-131.
(21) Giardina`, D.; Brasili, L.; Gregori, M.; Massi, M.; Picchio, M. T.;
Quaglia, W.; Melchiorre, C. Structure-Activity Relationships in
Prazosin-Related Compounds. Effect of Replacing a Piperazine
Ring with an Alkanediamine Moiety on R1-Adrenoceptor Block-
ing Activity. J . Med. Chem. 1989, 32, 50-55.
(22) Giardina`, D.; Gulini, U.; Massi, M.; Piloni, M. G.; Pompei, P.;
Rafaini, G.; Melchiorre, C. Structure-Activity Relationships in
Prazosin-Related Compounds. 2. Role of the Piperazine Ring on
R-Blocking Activity. J . Med. Chem. 1993, 36, 690-698.
(23) Pfizer Inc. GB 1,543,668.
(24) Campbell, S. F.; Hardstone, J . D.; Palmer, M. J . 2,4-Diamino-
6,7-dimethoxyquinoline as R1-Adrenoceptor Antagonists and
Antihypertensive Agents. J . Med. Chem. 1988, 31, 1031-1035.
(25) De Benedetti, P. G.; Menziani, M. C.; Rastelli, G.; Cocchi, M.
Molecular Orbital Study of the Nitrogen Basicity of Prazosin
Analogues in Relation to Their R1-Adrenoceptor Binding Affinity.
J . Mol. Struct. (Theochem) 1991, 233, 343-351.
(26) Rastelli, G.; Fanelli, F.; Menziani, M. C.; Cocchi, M.; De
Benedetti, P. G. Conformational Analysis, Molecular Modeling
and Quantitative Structure-Activity Relationships Studies in
2,4-Diamino-6,7-Dimethoxy-2-Substituted Quinazolines (R1-Adr-
energic Antagonists. J . Mol. Struct. (Theochem) 1991, 251, 307-
318.
(27) De Benedetti, P.G.; Cocchi, M.; Menziani, M.C.; Fanelli, F.
Theoretical Quantitative Structure-Activity Analysis and Phar-
macophore Modeling of Selective Non Congeneric ctla-Adrenergic
Antagonists. J . Mol. Struct. (Theochem), 1993, 280, 283-290.
(28) De Benedetti, P. G.; Cocchi, M.; Menziani, M. C.; Fanelli, F.
Theoretical Quantitative Size and Shape Activity Selectivity
Analyses of 5-HT1A Serotonin and R-Adrenergic Receptor Ligands.
J . Mol. Struct. (Theochem) 1994, 305, 101-110.
(29) Cocchi, M.; Menziani, M. C.; Fanelli, F.; De Benedetti, P. G.
Theoretical Quantitative Structure-activity Relationship Analy-
sis of Congeneric and non Congeneric R1-Adrenoceptor Antago-
nists. A Chemometric Study. J . Mol. Struct. (Theochem) 1995,
331, 79-93.
(30) Karelson, M.; Lobanov, V. S.; Katritzky, A. R. Quantum-
Chemical Descriptors in QSAR/QSPR Studies. Chem. Rev. 1996,
96, 1027-1043 and references therein.
(31) Menziani, M. C.; De Benedetti, P. G.; Karelson, M. Theoretical
Descriptors in Quantitative Structure-Affinity and Selectivity
Relationship Study of Potent N4-Substituted Arylpiperazine
5-HT1A Receptor Antagonists. Bioorg. Med. Chem. 1998, 6, 535-
550.
(32) Morrow, A. L.; Creese, I. Characterization of R1-Adrenergic
Receptor Subtypes in Rat Brain: a Reevaluation of [3H]WB4101
and [3H]Prazosin Binding. Mol. Pharmacol. 1986, 29, 321-330.
(33) Diop, L.; Dausse, J .-P.; Meyer, P. Specific Binding of [3H]-
Rauwolscine to R2-Adrenoceptors in Rat Cerebral Cortex: Com-
parison between Crude and Synaptosomal Plasma Membranes.
J . Neurochem. 1983, 41, 710-715.
(34) Giardina`, D.; Crucianelli, M.; Romanelli, R.; Leonardi, A.;
Poggesi, E.; Melchiorre, C. Synthesis and Biological Profile of
Enantiomers of [4-(4-Amino-6,7-dimethoxyquinazolin-2-yl)-cis-
octahydroquinoxalin-1-yl]furan-2-ylmethanone (Cyclazosin), a
Potent Competitive R1B-Adrenoceptor Antagonist. J . Med. Chem.
1996, 39, 4602-4607.
(35) Hoyer, D.; Engel, G.; Kalkman, H. O. Molecular Pharmacology
of 5-HT1 and 5-HT2 Recognition Sites in Rat and Pig Brain
Membranes: Radioligand Binding Studies with [3H]5-HT, [3H]8-
OH-DPAT, (-)-[125I]Iodocyanopindolol, [3H]Mesulergine and [3H]-
Ketanserin. Eur. J . Pharmacol. 1985, 118, 13-23.
(36) Craig, A. S.; Kenneth, J . K. In Vivo Regulation of the Serotonin-2
Receptor in Rat Brain. Life Sci. 1986, 38, 117-127.
(37) Patane, M. A.; Scott, A. L.; Broten, T. P.; Chang, R. S. L.;
Ransom, R. W.; DiSalvo, J .; Forray, C.; Bock, M. G. 4-Amino-
2-[4-[1-(benzyloxycarbonyl)-2(S)-[[(1,1-dimethylethyl)amino]car-
bonyl]piperazinyl]-6,7-dimethoxyquinazoline (L-765,314): A Po-
tent and Selective R1b Adrenergic Receptor Antagonist. J . Med.
Chem. 1998, 41, 1205-1208.
(38) Imagawa, J .; Akima, M.; Sakai, K. Functional Evaluation of
Sympathetically Mediated Responses in in Vivo Lower Urinary
Tract of Dogs. J . Pharmacol. Methodol. 1989, 22, 103-111.
(39) Althuis, T. H.; Hess, H.-J . Synthesis and Identification of the
Major Metabolites of Prazosin Formed in Dog and Rat. J . Med.
Chem. 1977, 20, 146-149.
(40) Bockmuhl, M.; Stein, L. Ger. Offen. 925,468; Chem. Abstr. 1958,
52, 3869a.
(41) Yamanouchi Pharm. EP257,787.
(42) Kessar, S. V.; Chaud, R.; J it, P.; Azasteroids: Part XII
-
Synthesis of(+-)-13,16-Diaza-15-oxo-18-norequilenin Methyl
Ether. Indian J . Chem. 1974, 12, 113-116.
(43) von Scaefer, H.; Hackmack, G.; Eistetter, K.; Kruger, U.; Menge,
H. G.; Klosa, J . Synthesis, Physico-chemical Properties and
Pharmacological Screening Results of Budipine and Related
1-Alkyl-4,4-diphenylpiperidines. Arzneim.-Forsh. 1984, 34, 233-
240.
(44) Castagnoli, N., J r.; Cymerman, C. J .; Melikian, A. P.; Roy, S. K.
Amine-N-oxide Rearrangements Mechanism and Products of
Thermolysis. Tetrahedron 1970, 26, 4319-4327.
(45) J ohnson, W. S.; Sheuvi, A. B.; Boots, S. G. An Approach to
Taxodione Involving Biomimetic Polyene Cyclization Methodol-
ogy. Tetrahedron 1982, 38, 1397-1404.
(46) Audisio, G.; Ruggieri, R. Derivatives of R-Hydroxyisobutyric Acid.
Gazz. Chim. Ital. 1963, 93, 335-338.
(47) QUANTA/CHARMm, 1996; Molecular Simulations Inc., 16 New
England Executive Park, Burlington, MA 01803-5297.
(48) Dewar, M. J . S.; Zoebisch, E. G.; Healey, E. F.; Stewart, J . J . P.
AM1: A New General Purpose Quantum Mechanical Molecular
Model. J . Am. Chem. Soc. 1985, 107, 3902-3909.
(49) Katritzky, A. R.; Lobanov, V. S.; Karelson, M. QSPR: The
Correlation and Quantitative Prediction of Chemical and Physi-
cal Properties from Structure. Chem. Soc. Rev. 1995, 279-287.
(50) Cappelli, A.; Anzini, M.; Vomero, S.; Mennuni, L.; Makovec, M.;
Douchet, E.; Hamon, M.; Bruni, G.; Romeo, M. R.; Menziani, M.
C.; De Benedetti, P. G.; Langer, T. Novel Potent and Selective
Central 5-HT3 Receptor Ligands Provided with Different Intrin-
sic Efficacy. 1. Mapping the Central 5-HT3 Receptor Binding Site
by Arylpiperazine Derivatives. J . Med. Chem. 1998, 41, 728-
741.
(51) Katritzky, A. R.; Mu, L.; Lobanov, V. S.; Karelson, M. Correlation
of Boiling Points with Molecular Structure. 1. A Training Set of
298 Diverse Organics and a Test Set of 9 Simple Inorganics. J .
Phys. Chem. 1996, 100, 10400-10407.
(52) Katritzky, A. R.; Mu, L.; Karelson, M. QSPR Treatment of
Unified Nonspecific Solvent Polarity Scale. J . Chem. Inf. Com-
put. Sci. 1997, 37, 756-761.
(53) Katritzky, A. R.; Lobanov, V. S.; Karelson, M. CODESSA,
training manual; Gainesville, FL, 1995.
(54) Stanton, D. T.; J urs, P. C. Development and Use of Charged
Partial Surface Area Structural Descriptors in Computer-
Assisted Quantitative Structure-Property Relationship Studies.
Anal. Chem. 1990, 62, 2323-2329.
(55) Testa, R.; Guarneri, L.; Angelico, P.; Poggesi, E.; Taddei, C.;
Sironi, G.; Colombo, D.; Sulpizio, A. C.; Naselsky, D. P.; Hieble,
J . P.; Leonardi, A. Pharmacological Characterization of the
Uroselective Alpha-1 Antagonist Rec 15/2739 (SB 216469): Role
of the Alpha-1L Adrenoceptor in Tissue Selectivity, Part II. J .
Pharmacol. Exp. Ther. 1997, 281, 1284-1293.
J M9805337