Structure-activity relationships in 1,4-benzodioxan-related compounds. 7.1 Selectivity of 4-phenylchroman analogues for α1-adrenoreceptor subtypes

Article abstract of DOI:10.1021/jm011066n

WB4101 (1)-related compounds 5-10 were synthesized, and their biological profile at α₁-adrenoreceptor (AR) subtypes and 5-HT_1A serotoninergic receptors was assessed by binding assays in Chinese hamster ovary and HeLa cell membranes e

Full text of DOI:10.1021/jm011066n

J . Med. Chem. 2002, 45, 1633-1643
1633
Str u ctu r e-Activity Rela tion sh ip s in 1,4-Ben zod ioxa n -Rela ted Com p ou n d s. 7.1
Selectivity of 4-P h en ylch r om a n An a logu es for r1-Ad r en or ecep tor Su btyp es
Wilma Quaglia,† Maria Pigini,† Alessandro Piergentili,† Mario Giannella,† Francesco Gentili,† Gabriella Marucci,†
Antonio Carrieri,‡ Angelo Carotti,‡ Elena Poggesi,§ Amedeo Leonardi,§ and Carlo Melchiorre*,|
Department of Chemical Sciences, University of Camerino, Via S. Agostino 1, 62032 Camerino (MC), Italy,
Pharmaco-Chemistry Department, University of Bari, Via E. Orabona 4, 70126 Bari, Italy, Research and Development
Division, Recordati S.p.A., Via Civitali 1, 20148 Milano, Italy, and Department of Pharmaceutical Sciences,
University of Bologna, Via Belmeloro 6, 40126 Bologna, Italy
Received October 19, 2001; Revised Manuscript Received J anuary 17, 2002
WB4101 (1)-related compounds 5-10 were synthesized, and their biological profile at
R1-adrenoreceptor (AR) subtypes and 5-HT1A serotoninergic receptors was assessed by binding
assays in Chinese hamster ovary and HeLa cell membranes expressing the human cloned
receptors. Moreover, their receptor selectivity was further determined in functional experiments
in isolated rat prostate (R1A), vas deferens (R1A), aorta (R1D), and spleen (R1B). In functional
assays, compound 5 was the most potent at R1D-ARs with a reversed selectivity profile (R1D
>
R1A > R1B) relative to both prototype 1 and phendioxan (2) (R1A > R1D > R1B), whereas compound
8, bearing a carbonyl moiety at position 1, was the most potent at R1A-ARs with a selectivity
profile similar to that of prototypes. The least potent of the series was the trans isomer 6,
suggesting that optimum R1-AR blocking activity in this series is associated with a cis
relationship between the 2-side chain and the 4-phenyl ring rather than a trans relationship
as previously observed for the 2-side chain and the 3-phenyl ring in 2 and related compounds.
Binding affinity results were not in complete agreement with the selectivity profiles deriving
from functional experiments. Although a firm explanation was not available, neutral and
negative antagonism and receptor dimerization were considered as two possibilities to account
for the difference between binding and functional affinities. Finally, compound 5 was selected
for a modeling study in comparison with 1, mephendioxan (3), and open phendioxan (4) to
achieve information on the physicochemical interactions that account for its high affinity toward
R
1d/D-ARs.
In tr od u ction
cently, it has been advanced that the R1B-AR subtype
may be involved in the regulation of blood pressure as
R1b knockout mice displayed a significantly reduced
responsiveness to phenylephrine-induced increases in
blood pressure.8 A potential therapeutic use for R1D
subtype antagonists has not been defined yet. However,
it has been shown that the R1d subtype is dominant in
the human bladder detrusor.9 Furthermore, it seems
that the R1D receptor blockade may be of benefit for the
irritative symptoms of BPH that result from involuntary
contractions of the bladder smooth muscle.10 It derives
that a valuable drug for the treatment of BPH symp-
toms would be one that would inhibit R1-ARs in the low
urinary tract without blocking R1-ARs responsible for
maintaining the vascular tone. Thus, a useful pharma-
cological agent might be an R1-AR antagonist, which is
selective for the R1A over the R1B subtype and endowed
also with activity at R1D-ARs such as to provide some
potential additional therapeutic benefit.11
2-4
All three subtypes (R1a/A, R1b/B, R1d/D
)
of the R1-
adrenoreceptor (R1-AR) are members of the G protein-
coupled receptor (GPCR) superfamily and are respon-
sible for mediating the actions in the body of the
neurotransmitter noradrenaline. Despite the wide-
spread tissue distribution of these receptors and their
involvement in a variety of physiological processes, the
extremely high sequence homology in the classic ligand-
binding domain across R1-AR subtypes has made it
difficult to design therapeutic agents that can effectively
target one receptor subtype while not affecting all
others.5 Thus, there is an ongoing need for effective
structure-activity studies based on selective ligands for
R1-AR subtypes.
It has been claimed that R1A-AR antagonists can be
useful in the treatment of benign prostatic hyperplasia
(BPH).6 As a matter of fact, in the human prostate,
although mRNA levels are not necessarily in close
correlation with the corresponding protein levels, mRNA
for all three subtypes has been found with that for the
R1a subtype present in the greatest abundance.7 Re-
In continued efforts to identify potent and selective
R1-AR antagonists, we have previously published on the
structure-activity relationships leading to the design
and synthesis of potent R1-antagonists whose chemical
structure incorporated a 1,4-benzodioxan-2-yl or a closely
related moiety as the main feature.12 WB 4101 {N-[2-
(2,6-dimethoxyphenoxy)ethyl]-2,3-dihydro-1,4-benzo-
dioxin-2-methanamine, 1, Figure 1} is the prototype of
R1-AR antagonists bearing a benzodioxan moiety.13 Data
* To whom correspondence should be addressed. Tel.: +39-051-
2099706. Fax: +39-051-2099734. E-mail: camelch@kaiser.alma.unibo.it.
† University of Camerino.
‡ University of Bari.
§ Recordati S.p.A..
| University of Bologna.
10.1021/jm011066n CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/12/2002

1634 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8
Quaglia et al.
Sch em e 2
Sch em e 3
F igu r e 1. Design strategy for the synthesis of compounds
5-10 by changing the place of the phenyl ring from position
3 to 4 and replacing the oxygen atom at position 4 of 2 with a
methylene unit and subsequent ring opening of structure I.
Sch em e 1a
analogue of 2, namely, 4 (Table 4), retained high affinity
for R1-AR subtypes. However, contrary to 2, compound
4 was more potent at R1d/D-ARs than at both R1a/A and
R1b/B subtypes.12
The present article expands on the study of another
aspect of structure-activity relationships of 1, namely,
the effect of exchanging the oxygen atom at position 4
for a phenylmethine group, affording 5. In turn, com-
pound 5 has been modified by replacing the oxygen atom
at position 1 with a sulfur atom, a carbonyl group, or a
methylene unit, giving 6 and 7, 8, and 9, respectively.
Finally, the effect of opening the dihydropyran ring of
5 has been investigated by synthesizing compound 10.
The design strategy for these compounds is shown in
Figure 1.
We describe here the synthesis of compounds 5-10
(Schemes 1-4) and the results of their screening at
human cloned R1-ARs, expressed in Chinese hamster
ovary (CHO) cells, and native R1-ARs in rat prostate,
prostatic vas deferens, thoracic aorta, and spleen.
Furthermore, to improve our knowledge on the main
physicochemical interactions governing the ligand bind-
ing to the R1d-ARs and to interpret at the 3D level the
structure-affinity relationships of the adrenergic ligands
reported in this paper, a series of docking experiments
were undertaken with compounds 3-5 and 1, the latter
being taken as the reference ligand. Compounds 1, 3,
and 5 differ in the presence and position of a phenyl
a
Bn ) benzyl.
showed that the two oxygen atoms at positions 1 and 4
might have a different role in receptor binding.14,15 The
oxygen atom at position 4 of 1 may play only a structural
role as it could be replaced by a sulfur atom or a
methylene unit without affecting potency toward rat vas
deferens R1-ARs. On the contrary, the oxygen atom at
position 1 of 1 appears to contribute to receptor binding
because its replacement with a sulfur atom or a meth-
ylene group resulted in a significant decrease in potency,
whereas its replacement by a carbonyl function did not
modify the biological profile at rat vas deferens R1-ARs.
Among the many other structural modifications per-
formed on 1, the insertion of a phenyl ring or a p-tolyl
moiety at the 3-position having a trans relationship with
the 2-side chain afforded phendioxan {trans-N-[2-(2,6-
dimethoxyphenoxy)ethyl]-2,3-dihydro-3-phenyl-1,4-ben-
zodioxin-2-methanamine, 2, Figure 1} or its p-tolyl
analogue mephendioxan (3, Table 4), respectively, that
displayed the highest affinity for native R1A-ARs relative
to both R1B and R1D subtypes.12,16 Interestingly, an open

1,4-Benzodioxan-Related Compounds
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8 1635
Sch em e 4
compound 6, the C4-H at δ 4.33 ppm is equatorial (J )
2.3 Hz). The irradiation of C4-H caused no NOE effect
at C2-H. However, because of the C3-H multiplicity,
it can be deduced that C2-H is axial and therefore trans
to C4-H. In compound 7, the irradiation of C4-H caused
NOE effect at C2-H. Moreover, C3-Ha at δ 2.02 ppm
coupled with three protons with high coupling constants
(J ) 10.5, 12.3, 13.3 Hz), indicating that C2-H and
C4-H are axial and therefore cis. Consequently, the
stereochemical relationship between the 2-side chain
and the 4-substituent is trans in 6 and cis in 7.
Compound 8 was synthesized by a Mannich reaction
of 4-phenyl-1-tetralone22 and 2-(2,6-dimethoxy-phenoxy)-
ethylamine21 and paraformaldehyde (Scheme 3). This
reaction afforded only one of the two possible isomers.
The stereochemical relationship between the 2-side
chain and the 4-substituent in 8 was determined by 2D
NOE (nuclear Overhauser enhancement sprectroscopy)
measurements. Compound 8 showed cross-peaks be-
tween C4-H and C2-H; moreover, both C4-H and
C2-H showed cross-peaks with the same C3-Ha, indi-
cating a cis relationship between C2-H and C4-H.
Open analogue 10 of 5 was synthesized as shown in
Scheme 4. Alkylation of 2-benzyl-phenol with chloro-
acetaldehyde dimethyl acetal gave 19, which, in turn,
was transformed into the aldehyde 20 by acidic hydroly-
sis. Reductive amination of 20 with 2-(2,6-dimethoxy-
phenoxy)ethylamine21 gave 10.
substituent on the benzo-fused heterocyclic ring, whereas
4 can be considered an open analogue of 3. The diverse
activity profiles of these ligands might therefore be
attributed either to a different binding mode or, most
likely, to some favorable/unfavorable hydrophobic/steric
interactions driven by the presence and spatial arrange-
ment of the aryl group on the benzo-fused heterocyclic
ring.
Ch em istr y
Biology
The compounds used in the present investigation were
synthesized by standard procedures and characterized
by 1H nuclear magnetic resonance (NMR) and elemental
analysis. Compound 5 was synthesized starting from
1117 as shown in Scheme 1. Although amine 12 has been
already reported,17 it was synthesized by following a
different procedure through the catalytic hydrogenation
of 11. This reaction afforded only one of the two possible
stereoisomers. (2,6-Dimethoxy-phenoxy)acetic acid,18 in
chloroform, was amidated in the presence of Et3N and
EtOCOCl with amine 12 to give the corresponding
amide 13. The stereochemical relationship between the
2-side chain and the 4-substituent in 13 was determined
by 1D nuclear Overhauser effect (1D NOE) measure-
ments. A cis relationship between C2-H and C4-H of
13 was deduced by the observation that irradiation of
C4-H caused NOE at C2-H (2%) and C3-Ha (3%) and
irradiation of C2-H gave the same effect at C3-Ha
(2.5%). Reduction of 13 with borane-methyl sulfide
complex in dry diglyme gave the corresponding amine
5.
Isomers 6 and 7 and the carbon analogue 9 were
synthesized as shown in Scheme 2. Basic hydrolysis of
a mixture of cis- and trans-4-phenyl-thiochroman-2-
carboxylic acid ethyl ester19 afforded a cis/trans mixture
(ratio 1:1) of the corresponding acid 14. Acids 14 and
15,20 in chloroform, were amidated in the presence of
Et3N and EtOCOCl with 2-(2,6-dimethoxy-phenoxy)-
ethylamine21 to the corresponding amides 16-18. Iso-
mers 16 and 17 were separated by column chromatog-
raphy. Reduction of amides 16-18 with borane-methyl
sulfide complex in dry diglyme gave the corresponding
amines 6, 7, and 9. The stereochemical relationship
between the amino side chain and the phenyl ring in 6
and 7 was determined by 1D NOE measurements. In
Bin d in g Exp er im en ts. The pharmacological profile
of compounds 5-10 was evaluated by radio-receptor
binding assays using 1 and 8-[2-[4-(2-methoxyphenyl)-
1-piperazinyl]ethyl]-8-azaspiro[4.5]decane-7,9-dione
(BMY-7378) as standard compounds. [3H]Prazosin was
used to label cloned human R1-ARs expressed in CHO
cells.23 Furthermore, [3H]8-hydroxy-2-(di-n-propylami-
no)tetralin was used to label cloned human 5-HT1A
receptors expressed in HeLa cells.24,25
F u n ction a l Stu d ies. Receptor subtype selectivity of
compounds 5-10 was further determined at R1-ARs on
different isolated tissues using 1, 2, and BMY-7378 as
standard compounds. R1-AR subtypes blocking activity
was assessed by antagonism of (-)-noradrenaline-
induced contraction of prostate26 and prostatic vas
27
28a
deferens (R1A
)
or thoracic aorta (R1D
)
and by an-
tagonism of (-)-phenylephrine-induced contraction of
rat spleen (R1B).28b
Resu lts a n d Discu ssion
The functional activity, expressed as pA2 or pKB
values, at R1-AR subtypes of compounds used in the
present study is shown in Table 1 in comparison with
that of prototypes 1 and 2 and the reference compound
BMY-7378, a selective R1d-AR antagonist.
Analysis of the results reveals that all of the com-
pounds were effective antagonists at R1-ARs. Further-
more, all of the compounds were competitive antagonists
at R1-AR subtypes with the exception of 6, 9, and 10 at
rat vas deferens R1A-ARs and 6 and 10 at R1B-ARs for
which a Schild analysis could not be performed due to
a depression of the maximal response to the agonist by
antagonist concentrations higher than 1 µM. Interest-
ingly, the antagonist affinities calculated in rat prostate

1636 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8
Quaglia et al.
Ta ble 1. Antagonist Affinities, Expressed as pA2 Values, of 1,
2, 5-10, and BMY-7378 at R1-ARs on Isolated Rat Prostate,
Ta ble 2. Affinity Constants, Expressed as pKi, of Compounds
1, 5-10, and BMY-7378 for Human Recombinant R1-AR
Subtypes and 5-HT1A Receptorsa
Prostatic Vas Deferens (R1A), Thoracic Aorta (R1D), and Spleen
a
(R1B
)
pKi, human cloned receptors
compd
R1a
R1b
R1d
5-HT1A
1
5
6
7
8
9
9.37
10.30
8.63
10.05
9.60
9.29
9.40
6.42
8.0
10.40
7.56
8.94
8.91
7.91
8.55
6.15
9.29
10.05
7.64
9.36
9.74
7.77
8.44
8.89
8.68
8.19
7.99
7.82
7.96
8.20
7.97
9.43
10
BMY-7378
a
Equilibrium dissociation constants (Ki) were derived from IC50
values using the Cheng-Prusoff equation.53 The affinity estimates
were derived from displacement of [3H]prazosin and [3H]-8-
hydroxy-2-(di-n-propylamino)tetralin binding for R1-ARs and 5-HT1A
receptors, respectively. Each experiment was performed in trip-
licate. Ki values were from two to three experiments, which agreed
within (20%.
the previous observation that the oxygen atom at
position 1 of 1 may have a role in receptor binding.14,15
Interestingly enough, the oxa analogue 5 was more
potent than prototype 2 at all R1-AR subtypes. Further-
more, it displayed a reversed selectivity profile relative
to 2 (R1D > R1A > R1B vs R1A > R1D > R1B). Compound 5
was even more selective than BMY-7378 for R1D-ARs
relative to R1B-ARs, while it was, however, less selective
relative to R1A-ARs. On the other hand, the sulfur
analogue 7 and the oxo analogue 8 behaved like
prototype 2 as they were more potent at R1A-ARs and
less potent at R1B-ARs with a similar selectivity profile,
that is, R1A > R1D > R1B. It derives that from a functional
point of view these new compounds may have a potential
in the treatment of BPH. Opening the dihydropyran
ring of 5 gave 10, which turned out to be a much weaker
antagonist than 5 at all R1-ARs.
pA2
vas deferens
(R1A
prostate
(R1A
spleen
(R1B
thoracic
aorta (R1D)
compd
)
)
)
1
2
5
6
7
8
9
9.27 ( 0.07 9.51 ( 0.06 8.16 ( 0.09 8.80 ( 0.12
8.23 ( 0.09c 8.24 ( 0.07 5.60 ( 0.08 7.49 ( 0.06
8.95 ( 0.07c 8.88 ( 0.02 7.54 ( 0.01 9.56 ( 0.09
6.97 ( 0.05c 6.68 ( 0.10b 6.52 ( 0.14b 7.21 ( 0.09
8.31 ( 0.22c 8.21 ( 0.06 6.81 ( 0.03 7.76 ( 0.02
9.46 ( 0.05c 9.35 ( 0.08 7.58 ( 0.19 8.70 ( 0.09
6.54 ( 0.10 6.75 ( 0.10b 6.96 ( 0.06b 7.04 ( 0.01
7.47 ( 0.20 7.23 ( 0.09b 7.17 ( 0.01 7.78 ( 0.08
7.01 ( 0.08 7.48 ( 0.09 8.40 ( 0.09
10
BMY-7378
a
pA2 values ( SE were calculated according to Arunlakshana
and Schild51a unless otherwise specified, constraining the slope
to -1.51b pA2 is defined as the negative logarithm to the base 10
of that dose of antagonist that requires a doubling of the agonist
b
dose to compensate for the action of the antagonist. The antago-
The binding affinities, expressed as pKi values, in
CHO cells expressing human cloned R1-ARs and HeLa
cells expressing human 5-HT1A receptors of compounds
5-10 are shown in Table 2 in comparison with those of
1 and BMY-7378. Compounds 5-10 were investigated
at 5-HT1A serotoninergic receptors because it is known
that 1 and related compounds are effective ligands also
for the 5-HT1A receptor.12 It turned out that all of the
compounds were potent ligands toward 5-HT1A receptors
but their affinity was much lower than that of proto-
types 1 and BMY-7378. Analysis of the results reveals
that R1 binding affinities of compounds 5-10 did not
correlate with potencies observed in functional assays.
It can be seen that while binding affinities of reference
compound 1 are qualitatively and quantitatively com-
parable with pA2 or pKB values derived from functional
experiments, those observed for 5-10 are not in agree-
ment from both a qualitative and a quantitative point
of view with functional affinities as shown in Table 3.
It would suffice to note that the carbon analogue 9 was
nist affinity was calculated from only one concentration because
there was a depression of the maximum response to the agonist,
which increased with increasing the antagonist concentration.c pKB
values were calculated according to van Rossum51c in the range
of 0.01-1 µM.
and prostatic vas deferens did not differ to each other,
confirming that each of these preparations can be used
to study the R1A-AR subtype.26
It is clear that potency at R1-AR subtypes was
definitely dependent on the stereochemistry and sub-
stitution pattern of the compounds. It turned out that
contrary to the trans stereochemical relationship be-
tween the 2-side chain and the 3-phenyl ring observed
for prototype 2 and related compounds,16 a trans
relationship between the 2-side chain and the 4-phenyl
group is not optimal for binding at R1-ARs as revealed
by a comparison between the trans and the cis isomers
6 and 7. Isomer 7 was significantly more potent (22-
34-fold) than the trans isomer 6 at R1A-ARs, while it
was, however, only slightly more potent at both R1B- and
a markedly weaker antagonist than 5, 7, and 8 at R1A
-
R1D-ARs.
ARs in functional experiments, whereas it displayed
high affinity in binding assays, resulting in a selective
The role of the sulfur atom at position 1 in 7 was
investigated by its replacement with an oxygen atom
(5), a carbonyl group (8), and a methylene unit (9), while
a cis relationship between the side chain and the phenyl
ring was kept constant. It turned out that 5 and 8 were
more potent than 7 at all R1-ARs, whereas the carbon
analogue 9 was the least potent. This result confirms
R1A antagonist (R1A > R1B ) R1D). Furthermore, the open
analogue 10, which was a weak antagonist in functional
experiments, was potent in binding assays with an
affinity almost comparable to that of prototype 1.
However, one aspect was confirmed by binding experi-
ments, namely, that a cis stereochemical relationship

1,4-Benzodioxan-Related Compounds
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8 1637
Ta ble 3. Differences in Antagonist Potency between
changes in the pharmacology of GPCRs, which may be
due to dimerized receptors. These new receptor entities
may not always signal as either a monomer or as
dimers. It derives that the presence of homodimeric or
heterodimeric receptors may certainly have profound
ramifications with regard to the interpretation of bio-
logical data and, as a consequence, to the validity of
structure-activity relationships. Thus, the discrepancy
often observed, like in present case, between functional
and binding affinities may not represent an anomaly
because in screening procedures a homogeneous popula-
tion of cloned receptors is used, which can be organized
differently than native receptors in functional tissues
and, consequently, their biological behavior may not be
coincident. However, the discrepancy between functional
and binding affinities observed in the present study may
simply be accounted for by a different bioavailability of
the compounds at the receptor level.
Functional and Radioligand Binding Assaysa
b
compd
R
1a/R1A
R
1b/R1B
R
1d/R1D
1
5
6
7
8
9
1
24
64
62
1.5
1
3
3
3
40
11
5
724
11
135
21
9
142
112
10
24
5
3
BMY-7378
0.25
0.05
a
Differences in potency were calculated as the antilog of the
difference between pKi values at cloned receptors and pA2 values
at the corresponding native receptors from data reported in Tables
b
1 and 2. pA2 values for native R1A receptors were the mean
between the values obtained in prostate and the values obtained
in vas deferens preparations.
between the 2-side chain and the 4-phenyl ring is
optimal for interaction with R1-ARs. As a matter of fact,
trans isomer 6 was consistently less potent than cis
isomer 7 at all R1-AR subtypes.
Notwithstanding, the discrepancy between functional
and binding affinities, a molecular modeling study, was
undertaken with compound 5 to get information on its
mode of interaction with the R1d-AR, using compounds
1, 3, and 4 for comparison. The choice of 5 was dictated
Recently,29 we discussed the possibility that if an
antagonist does not adhere perfectly to the concept of
neutral antagonism in the interaction with the receptor
but behaves as an inverse agonist, then a discrepancy
between affinity values estimated in functional assays
and affinity values estimated in binding experiments
may not represent a surprise, because the estimated
affinities of inverse agonists are system-dependent.
Interestingly, the prototype of the present study, WB
4101 (1), was shown to be an inverse agonist in a
vascular model containing R1D-ARs and a native and
constitutively activated mutant of the R1a subtype.30
Thus, the 3-fold difference observed for 1 between
binding and functional affinity at R1d/D-ARs might be
explained by the fact that 1 is an inverse agonist at this
subtype, and as a consequence, its affinity is not system-
independent. It is tempting to assume that a similar
explanation might apply for the discrepancy observed
between binding and functional affinities for the other
compounds of the present investigation. However, other
possibilities may offer themselves to rationalize the
discrepancy between binding and functional affinities.
It has been generally assumed that GPCRs exist as
monomers and couple to G proteins in a 1:1 stoichio-
metric ratio. However, increasing evidence indicates
that such classical models of coupling between receptors
and G proteins may be oversimplified.31 An increasing
number of reports have described that receptor systems
form dimers. It has been shown that dimerization of the
receptor can occur when receptors are expressed in
artificial systems and studies with chimeras have
elucidated what part of the receptors participate in this
interaction.32 However, it is clear now that dimerization
and heterodimerization also occur in natural systems.
It has been demonstrated that the GABABR1 and
GABABR2 receptor subtypes form heterodimers in vivo
that are required for proper cell surface receptor local-
ization and function.33 Furthermore, in another study,34
it has been shown that κ and δ opioid receptors can form
heterodimers with distinct ligand binding and functional
properties, raising the possibility that heterodimeriza-
tion may represent a more general mechanism to
modulate GPCRs function. These observations of dimer-
ization and, what may be even more intriguing, hetero-
dimerization open a whole vista of possibilities for subtle
by its very high affinity for both native and cloned R1d/D
ARs.
-
The rhodopsin projection map35,36 has been exten-
sively used as a major template to building 3D theoreti-
cal models of the seven transmembrane helices of
GPCRs, by means of homology building approaches.37
Very recently, a 2.8 Å resolution structure of the
rhodopsin receptor, comprising extra- and intracellular
loops, has been published.38,39 These structural inves-
tigations, along with site-directed mutagenesis studies,40
have been extremely important to elucidate the complex
binding and activation processes of diverse classes of
GPCRs.41 A significant contribution to this field came
just this year with the publication from some of us of a
modeling study of the rat R1d-AR.42 Such a study led to
the identification of two distinct structural arrange-
ments, corresponding to the resting and activated state
of the receptor, and to the determination of the main
structural requirements for an efficient binding to R1d-
ARs of both agonist and antagonist ligands.42
Although the number of analyzed ligands is too low
to derive any significant regression equation between
measured binding affinities and estimated binding
energies, relevant insights from this study might be
gained, at a more qualitative level, by carefully analyz-
ing and comparing the two data sets. The binding
energies calculated for the docked ligands are reported
in Table 4 along with their pKi values. As it can be seen
from Table 4, we were able to rank satisfactorily the
pKi data of most of the docked compounds in terms of
estimated binding energies. The most active compound
4 shows a binding energy nearly three times lower than
the least active compound 3 (-11.14 vs -3.35 kcal/mol),
and also, compound 1 was adequately scored with
respect to its pKi values (-7.63 kcal/mol). In contrast,
compound 5, which has an activity close to that of
compound 4, showed a higher than expected binding
energy (-7.44 kcal/mol). It is worth noting that the QXP
package uses, “by definition”, a quick way to calculate
interaction energy; therefore, it suffers from some
limitations such as an insufficient parametrization of

1638 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8
Quaglia et al.
Ta ble 4. Affinity Constants (pKi) and Binding Energies
(kcal/mol) of Compounds 1 and 3-5
pKi, human
cloned R1d
binding energy
(QXP)
binding energy
(GRID)
compd
1
3
4
5
9.29
7.93a
-7.63
-3.35
-11.14
-7.44
-11.82
-6.39
-12.40
-13.78
10.17b
10.05
a
b
Data from ref 16c. Data from ref 12.
the hydrophobicity for which no entropic effect is taken
into account. This limitation can be overcome by esti-
mating the binding energy with the latest version of the
well-known GRID software.43 Besides a well-param-
etrized force field for the calculation of the enthalpic
contribution, in this method, the entropic component is
also taken into account by considering the role of water
in the solvation-desolvation phases of the binding
process. Moreover, the calculation of the binding energy
is carried out not only as the summation of the energetic
contribution of the different atoms in the molecule
considered as probe atoms but also considering the
whole molecule as a probe.44 So, we decided to apply
this approach to recalculate the binding energies on the
same receptor-ligand complexes retrieved by the QXP
docking runs. As expected, the new values, reported in
the last column of Table 4, show a good linear relation-
ship with the pKi values.
F igu r e 2. Compounds 1 (upper left red), 3 (upper right violet),
5 (lower left green), and 4 (lower right cyan) docked into the
rat R1d-AR binding site.
amine function of Lys and the oxygen atom of one
methoxy group. At the same time, an edge-to-face π-π
interaction between the dimethoxyphenoxy group and
the Trp229 seems to take place.
As can be seen from the molecular complexes in
Figure 2, all of the compounds dock into the receptor
binding site with a similar topology. In particular, the
benzo-fused heterocyclic systems as well as the di-
methoxyphenoxy group bind in a similar fashion. Some
differences may be noted in the docked receptor-ligand
4 complex (see Figure 2D) where the benzyloxy moiety
is much more buried into the core of the receptor helices
bundle than that observed in complexes with ligands 3
(Figure 2B) and 5 (Figure 2C). Stronger favorable
hydrophobic interactions can be therefore expected for
compound 4, and that was the case. It must be pointed
out that this binding conformation can be attained only
by ligand 4 because of its higher flexibility, as compared
to the other ligands, all containing a benzo-fused, more
rigid heterocyclic ring. The least active compound 3
places its p-tolyl substituent in a completely different,
but easily accessible, receptor binding site (see Figure
2B). Its low pKi value can be therefore ascribed to the
lack of efficient hydrophobic contacts with the recepto-
rial counterpart. Moreover, with respect to the other
ligands, compound 3 gives a poorer π-π aromatic
stacking of the benzodioxan moiety with the phenyl ring
of Phe358, which has been shown to play an important
role in the binding of adrenergic ligands.45
Con clu sion s
A most intriguing finding of the present investigation
was the observation that changing place to the phenyl
ring from position 3, as in 2 and 3, to position 4 caused,
as far as R1-blocking activity is concerned, an inversion
of the stereochemical relationship between the side
chain and the phenyl ring. In fact, cis isomer 7 was
significantly more potent than trans isomer 6 at all R1-
AR subtypes, suggesting that optimal activity is likely
associated with a cis relationship rather than a trans
relationship as observed for 2 and analogues. Replacing
the sulfur atom at position 1 in 7 with an oxygen atom
(5) or a carbonyl function (8) improved markedly the
affinity toward all R1-AR subtypes, whereas the replace-
ment with a methylene unit (9) gave a significant
decrease in affinity. These results parallel those ob-
tained following the same structural modifications on
2.16 In functional experiments, compound 5 was a potent
and selective antagonist at R1D-ARs relative to R1B-ARs.
Although the same selectivity profile was not observed
in binding assays, compound 5 may be a useful lead for
the design of more selective ligands for R1D-ARs.
Exp er im en ta l Section
Finally, it is worth noting that in all of the four docked
ligands of Figure 2, the common dimethoxyphenoxy
moiety is always located closely to the charged nitrogen
of Lys379 making possible the formation of a strong
N+H-O hydrogen bond between the positively charged
Ch em istr y. Melting points were taken in glass capillary
tubes on a Bu¨chi SMP-20 apparatus and are uncorrected. IR
1
and H NMR spectra were recorded on Perkin-Elmer 297 and
Varian VXR 300 instruments, respectively. Chemical shifts are
reported in parts per million (ppm) relative to tetramethyl-

1,4-Benzodioxan-Related Compounds
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8 1639
silane (TMS), and spin multiplicities are given as s (singlet),
br s (broad singlet), d (doublet), dd (double doublet), ddd
(doublet of double doublet), t (triplet), br t (broad triplet), dt
(double triplet), q (quartet), or m (multiplet). Although the IR
spectra data are not included (because of the lack of unusual
features), they were obtained for all compounds reported and
were consistent with the assigned structures. The elemental
compositions of the compounds agreed to within (0.4% of the
calculated value. When the elemental analysis was not in-
cluded, crude compounds were used in the next step without
further purification. Chromatographic separations were per-
formed on silica gel columns (Kieselgel 40, 0.040-0.063 mm,
Merck) by flash chromatography. The term “dried” refers to
the use of anhydrous sodium sulfate. Compounds were named
following IUPAC rules as applied by Beilstein-Institut
AutoNom (version 2.1), a software for systematic names in
organic chemistry.
C-(4-P h en yl-ch r om a n -2-yl)m eth yla m in e (12). A solution
of 1117 (2.0 g, 4.79 mmol) in 4.4% HCOOH-MeOH (75 mL)
was added dropwise to a mixture of 10% Pd/C (4.0 g) in 4.4%
HCOOH-MeOH (150 mL). The mixture was stirred overnight
under a nitrogen atmosphere. After the catalyst was filtered
off and washed with MeOH, the solvent was evaporated and
the residue was taken up with 2 N NaOH and extracted with
CHCl3. The dried solvents were evaporated to give an oil; 0.6
g (52% yield). 1H NMR (CDCl3): δ 1.72 (br s, 2, NH2,
exchangeable with D2O), 1.8-2.29 (m, 2, 3-H), 2.7-3.02 (m,
2, CH2N), 4.03-4.41 (m, 2, 2-H and 4-H), 6.59-7.32 (m, 9,
ArH).
extracted with CHCl3, and the aqueous layer was acidified with
concentrated HCl. Extraction with CHCl3, followed by wash-
ing, drying, and evaporation of the extracts, gave 14 as a cis/
trans mixture (ratio 1:1); 0.58 g (92% yield); mp 145-162 °C.
1H NMR (CDCl3): δ 2.3-2.78 (m, 4, 3-H; cis and trans), 3.9
(dd, 1, 4-H; trans), 4.03 (dd, 1, 4-H; cis), 4.27 (dd, 1, 2-H; trans),
4.38 (br t, 1, 2-H; cis), 6.36 (br s, 2, COOH, exchangeable with
D2O; cis and trans), 6.62-7.41 (m, 18, ArH; cis and trans).
cis- an d tr a n s-4-P h en yl-th ioch r om an -2-car boxylic Acid
[2-(2,6-Dim eth oxy-p h en oxy)eth yl]a m id e (16 a n d 17). Eth-
yl chlorocarbonate (0.83 g, 7.40 mmol) was added dropwise to
a stirred and cooled (0 °C) solution of 14 (2.0 g, 7.40 mmol)
and Et3N (0.76 g, 7.40 mmol) in CHCl3 (85 mL), followed after
30 min by the addition of a solution of 2-(2,6-dimethoxy-
phenoxy)ethylamine21 (1.46 g, 7.40 mmol) in CHCl3 (50 mL).
The resulting reaction mixture was stirred at room tempera-
ture for 3 h and then washed with 2 N HCl, 2 N NaOH, and
finally water. Removal of dried solvents gave an oil, which was
purified by column chromatography using petroleum ether-
Et2O (4:6) as the eluent. The trans isomer 16 eluted first; 1.5
g (45% yield). 1H NMR (CDCl3): δ 2.54 (m, 2, 3-H), 3.52 (m, 2,
NCH2), 3.76 (s, 6, OCH3), 3.81 (m, 1, 4-H), 4.09 (m, 2, CH2O),
4.26 (t, 1, 2-H), 6.52-7.33 (m, 12, ArH), 7.56 (br t, 1, NH,
exchangeable with D2O).
The second fraction was the cis isomer 17 as an oil; 0.86 g
(26% yield). 1H NMR (CDCl3): δ 2.41 (dt, J ) 11.0, 13.4 Hz, 1,
3-H), 2.68 (dt, J ) 4.0, 13.4 Hz, 1, 3-H), 3.50 (m, 2, NCH2),
3.85 (s, 6, OCH3), 4.09 (m, 3, CH2O and 4-H), 4.21 (dd, J )
4.0, 11.0 Hz, 1, 2-H), 6.55-7.40 (m, 12, ArH), 7.51 (br t, 1,
NH, exchangeable with D2O).
cis-2-(2,6-Dim eth oxy-p h en oxy)-N-(4-p h en yl-ch r om a n -
2-ylm eth yl)a ceta m id e (13). Ethyl chlorocarbonate (0.13 g,
1.13 mmol) was added dropwise to a stirred and cooled (0 °C)
solution of (2,6-dimethoxy-phenoxy)acetic acid18 (0.24 g, 1.13
mmol) and Et3N (0.12 g, 1.13 mmol) in CHCl3 (10 mL), followed
after 30 min by the addition of a solution of 12 (0.27 g, 1.13
mmol) in CHCl3 (6 mL). The resulting reaction mixture was
stirred at room temperature overnight and then washed with
2 N HCl, 2 N NaOH, and finally water. Removal of dried
solvents gave an oil, which was purified by column chroma-
tography. Eluting with cyclohexanes-Et2O (5:5) gave 13 as a
solid; 0.35 g (71% yield); mp 44-52 °C. 1H NMR (CDCl3): δ
1.95 (dt, J ) 11.9, 13.4 Hz, 1, 3-H), 2.21 (ddd, J ) 1.5, 5.8,
13.4 Hz, 1, 3-H), 3.49-3.92 (m, 2, CH2N), 3.86 (s, 6, OCH3),
4.19 (dd, J ) 11.9, 5.8 Hz, 1, 4-H), 4.34 (m, 1, 2-H), 4.61 (s, 2,
CH2O), 6.54-7.38 (m, 12, ArH), 8.32 (br t, 1, NH, exchangeable
with D2O).
tr a n s-[2-(2,6-Dim eth oxy-ph en oxy)eth yl]-(4-ph en yl-th io-
ch r om a n -2-ylm eth yl)a m in e Oxa la te (6). This was synthe-
sized from 16 (1.0 g, 2.22 mmol) following the procedure
described for 5. Eluting with cyclohexanes-EtOAc-EtOH (8:
1.5:0.5) afforded 6 as the free base; 0.6 g (62% yield). 1H NMR
(CDCl3): δ 1.93 (br s, 1, NH, exchangeable with D2O), 2.13
(ddd, J ) 4.4, 11.1, 13.5 Hz, 1, 3-H), 2.40 (ddd, J ) 3.3, 4.8,
13.5 Hz, 1, 3-H), 2.78-2.98 (m, 4, CH2NCH2), 3.29 (m, 1, 2-H),
3.75 (s, 6, OCH3), 4.10 (t, 2, CH2O), 4.33 (t, 1, 4-H), 6.52-7.33
(m, 12, ArH). The free base was transformed into the oxalate
salt and crystallized from EtOH-MeOH; mp 198-199 °C.
Anal. (C26H29NO3S‚0.5H2C2O4‚0.25H2O) C, H, N, S.
cis-[2-(2,6-Dim et h oxy-p h en oxy)et h yl]-(4-p h en yl-t h io-
ch r om a n -2-ylm eth yl)a m in e Oxa la te (7). This was synthe-
sized from 17 (0.57 g, 1.27 mmol) following the procedure
described for 5. Eluting with cyclohexanes-EtOAc-EtOH (8:
1.5:0.5) afforded 7 as the free base; 0.43 g (78% yield). 1H NMR
(CDCl3): δ 1.92 (br s, 1, NH, exchangeable with D2O), 2.02
(ddd, J ) 10.5, 12.3, 13.3 Hz, 1, 3-H), 2.51 (dt, J ) 3.8, 13.3
Hz, 1, 3-H), 2.77-3.03 (m, 4, CH2NCH2), 3.70 (m, 1, 2-H), 3.82
(s, 6, OCH3), 4.07 (m, 1, 4-H), 4.13 (t, 2, CH2O), 6.53-7.41 (m,
12, ArH). The free base was transformed into the oxalate salt
and crystallized from EtOH-Et2O; mp 114-115 °C. Anal.
(C26H29NO3S‚H2C2O4‚0.75H2O) C, H, N, S.
cis-4-P h en yl-1,2,3,4-tetr a h yd r o-n a p h th a len e-2-ca r box-
ylic Acid [2-(2,6-Dim et h oxy-p h en oxy)et h yl]a m id e (18).
Ethyl chlorocarbonate (0.24 g, 2.14 mmol) was added dropwise
to a stirred and cooled (0 °C) solution of cis-4-phenyl-1,2,3,4-
tetrahydro-naphthalen-2-carboxylic acid20 (0.54 g, 2.14 mmol)
and Et3N (0.22 g, 2.14 mmol) in CHCl3 (25 mL), followed after
30 min by the addition of a solution of 2-(2,6-dimethoxy-
phenoxy)ethylamine21 (0.42 g, 2.14 mmol) in CHCl3 (15 mL).
The resulting reaction mixture was stirred at room tempera-
ture for 3 h and then washed with 2 N HCl, 2 N NaOH, and
finally water. Removal of dried solvent gave an oil, which was
purified by column chromatography. Eluting with petroleum
ether-cyclohexanes-EtOAc (2.5:2.5:5) gave 18 as a solid; 0.85
g (92% yield); mp 115-117 °C. 1H NMR (CDCl3): δ 2.03 (dt, J
) 11.1, 12.5 Hz, 1, 3-H), 2.36 (ddd, J ) 3.5, 4.4, 13.8 Hz, 1,
3-H), 2.72 (m, 1, 2-H), 2.92-3.32 (m, 2, 1-H), 3.57 (q, 2, NCH2),
3.80 (s, 6, OCH3), 4.12 (m, 3, CH2O and 4-H), 6.53-7.36 (m,
12, ArH), 6.94 (br t, 1, NH, exchangeable with D2O).
cis-[2-(2,6-Dim eth oxy-p h en oxy)eth yl]-(4-p h en yl-ch r o-
m a n -2-ylm eth yl)a m in e Oxa la te (5). A solution of 10 M
BH3‚CH3SCH3 (0.09 mL) in dry diglyme (1.5 mL) was added
dropwise at room temperature to a solution of 13 (0.31 g, 0.72
mmol) in dry diglyme (16 mL) with stirring under a stream of
dry nitrogen with exclusion of moisture. When the addition
was completed, the reaction mixture was heated at 120 °C for
6 h. After the mixture was cooled at 0 °C, excess borane was
destroyed by cautious dropwise addition of MeOH (7 mL). The
resulting mixture was left to stand overnight at room tem-
perature, cooled at 0 °C, treated with HCl gas for 10 min, and
then heated at 120 °C for 4 h. Removal of the solvent under
reduced pressure gave a residue, which was dissolved in water.
The aqueous solution was basified with NaOH pellets and
extracted with CHCl3. Removal of dried solvents gave a
residue, which was purified by column chromatography. Elut-
ing with cyclohexanes-EtOAc-EtOH (8:1.5:0.5) afforded 5 as
the free base; 0.25 g (83% yield). 1H NMR (CDCl3): δ 2.01 (dt,
J ) 13.1, 11.8 Hz, 1, 3-H), 2.28 (ddd, J ) 1.3, 5.7, 13.1 Hz, 1,
3-H), 3.03 (m, 4, CH2NCH2), 3.52 (br s, 1, NH, exchangeable
with D2O), 3.85 (s, 6, OCH3), 4.21 (m, 3, 4-H and CH2O), 4.49
(m, 1, 2-H), 6.54-7.39 (m, 12, ArH). The free base was
transformed into the oxalate salt and crystallized from EtOH;
.
mp 160-161 °C. Anal. (C26H29NO4 H2C2O4) C, H, N.
cis- an d tr a n s-4-P h en yl-th ioch r om an -2-car boxylic Acid
(14). A mixture of cis- and trans-4-phenyl-thiochroman-2-
carboxylic acid ethyl ester19 (0.7 g, 2.35 mmol) and 2 N NaOH
(9.5 mL) was heated to 80 °C for 6 h. The mixture was
cis-[2-(2,6-Dim eth oxy-p h en oxy)eth yl]-(4-p h en yl-1,2,3,4-
tetr a h yd r o-n a p h th a len -2-ylm eth yl)a m in e Oxa la te (9).

1640 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8
Quaglia et al.
This was synthesized from 18 (0.85 g, 1.97 mmol) following
the procedure described for 5. Eluting with cyclohexanes-
EtOAc-EtOH (8:1.5:0.5) afforded 9 as the free base; 0.58 g
of 1, retrieved from the Cambridge Crystallographic Database
(refcode BAXBEH),46 by using the fragment library imple-
mented in the QXP software.47 For compounds 3 and 5, the
(S)-equatorial configuration was adopted according to the
X-ray data of 1.48
1
(71% yield). H NMR (CDCl3): δ 1.56 (dt, J ) 12.45, 12.1 Hz,
1, 3-H), 1.86 (br s, 1, NH, exchangeable with D2O), 2.04-2.36
(m, 2, 2-H and 3-H), 2.58-3.10 (m, 6, 1-H and CH2NCH2), 3.80
(s, 6, OCH3), 4.10 (m, 1, 4-H), 4.17 (t, 2, CH2O), 6.52-7.36 (m,
12, ArH). The free base was transformed into the oxalate salt
and crystallized from MeOH; mp 200 °C. Anal. (C27H31NO3‚
0.5H2C2O4‚0.25H2O) C, H, N.
For the 3D receptor model, the coordinates of the rat R1d-
AR subtype in its activated state were recovered from our
recent paper.42 Because site-directed mutagenesis studies49
strongly suggested that Asp170 binds to the charged nitrogen
of both agonist and antagonist ligands, 1 was initially placed
into the cavity of the receptor delimited by transmembrane
domains 3-7 with the charged amino group at a proper
distance from the carboxylate anion of Asp170 to make a
hydrogen bond with it and with the dimethoxyphenoxy group
pointing toward the extracellular loops. This initial complex
was then minimized with the Pollak-Ribiere conjugate gradi-
ent method until the root mean square gradient was lower
than 0.001 kcal/mol Å2. A distance-dependent dielectric con-
stant of 4r was used throughout the calculations. At this step,
the CR trace of the receptor backbone was kept fixed. Com-
pounds 3-5 were then overlayed onto the structure of 1,
extracted from the previous minimization step, with the
module TFIT of QXP. TFIT allows a flexible superposition of
one or more molecules on a selected template using a super-
position force field, which automatically assigns short-range
attractive forces to similar atoms in different molecules. A first
flexible docking run was performed on 4 using the MCDOCK
module of QXP. All of the residues in a region within 5.0 Å
from all of the ligand atoms were allowed to move whereas
the rest of the protein was kept fixed. A total of 1000 different
random positions of the ligand in the receptor were generated
and subjected to a Monte Carlo simulation, followed again by
a Pollak-Ribiere conjugate gradient minimization. During the
energy minimization, constrained atom pairs were defined as
the ligand-protonated nitrogen atom and the Oδ1 atom of
Asp170. A quadratic energy potential of 1.2 kcal/mol Å2 was
applied to them. The receptor structure obtained at the end
of this run was then used as the starting model for a new series
of ligand docking runs. During the simulation, the receptor
was kept fixed, the ligands were allowed to freely move, and
no constrained atom pairs were used. The binding energy was
calculated as the sum of the receptor-ligand nonbonded
energy, the relative internal ligand energy, and the energy,
relative to the local minima, of the binding site. QXP uses a
mixed AMBER/MM2 force field to estimate the binding energy
of the receptor-ligand complex. The lowest energy receptor-
ligand complexes were retrieved and reported in Figure 2. For
binding energy calculation within GRID, the directive FOBE
and TORS were set to 4.0 and 1.0 kcal/mol, respectively. The
modeling and docking calculations were carried out on an SGI-
O2 workstation R10000. The graphical representations of the
protein models in Figure 2 are drawn with the MOLSCRIPT
software package.50
cis-2-{[2-(2,6-Dim eth oxy-p h en oxy)eth yla m in o]m eth yl}-
4-p h en yl-3,4-d ih yd r o-2H-n a p h th a len -1-on e Hyd r och lo-
r id e (8). A solution of 4-phenyl-1-tetralone22 (0.95 g, 4.28
mmol), 2-(2,6-dimethoxy-phenoxy)ethylamine‚HCl21 (1.0 g, 4.28
mmol), paraformaldehyde (0.39 g, 13.0 mmol), and concen-
trated HCl (0.1 mL) was heated at reflux for 8 h. After
evaporation of the solvent, the residue was dissolved in 2 N
NaOH and extracted with CHCl3. Removal of dried solvents
gave an oil, which was purified by column chromatography.
Eluting with cyclohexane-acetone (5:5) gave a mixture of cis
and trans isomers; 1.31 g (71% yield). The mixture of free bases
was transformed into the hydrochloride salt. Following crys-
tallization from 2-PrOH/Et2O, only the cis isomer was isolated;
1
0.3 g; mp 161-163 °C. H NMR (CDCl3 + D2O): δ 2.10 (dd, J
) 13.5, 12.5 Hz, 1, 3-H), 2.40 (dt, J ) 13.5, 4.2 Hz, 1, 3-H),
3.08-3.50 (m, 4, CH2NCH2), 3.87 (s, 6, OCH3), 3.96 (m, 1, 2-H),
4.2-4.5 (m, 3, 4-H and CH2O), 6.55-8.1 (m, 12, ArH). Anal.
(C27H29NO4‚HCl‚0.25H2O) C, H, N.
(2-Ben zyl-ph en oxy)acetaldeh yde Dim eth yl Acetal (19).
A 60% NaH dispersion in mineral oil (0.24 g, 6.0 mmol) was
washed with hexane under nitrogen, suspended in dimethyl-
formamide (DMF; 0.54 mL), and then added over 15 min of a
solution of 2-benzyl-phenol (1.0 g, 5.43 mmol) in dry DMF (0.76
mL). The mixture was stirred at room temperature for 1.5 h
and cooled to 0 °C. Chloroacetaldehyde dimethyl acetal (0.77
g, 6.18 mmol) was added, and the reaction mixture was heated
at 120 °C for 8 h. After it was cooled, the mixture was poured
in 2 N NaOH (10 mL) and extracted with Et2O. Removal of
dried solvents gave a residue, which was purified by column
chromatography. Eluting with cyclohexanes-EtOAc (98:2)
1
gave an oil; 0.53 g (36% yield). H NMR (CDCl3): δ 3.42 (s, 6,
OCH3), 3.99 (d, 2, OCH2), 4.01 (s, 2, CH2Ar), 4.68 (t, 1, CH),
6.81-7.30 (m, 9, ArH).
(2-Ben zyl-p h en oxy)a ceta ld eh yd e (20). Compound 19
(0.53 g, 1.95 mmol) was added to a solution of 2 N HCl (3.2
mL) in acetone (5.4 mL), and the resulting solution was heated
to 70 °C for 1.5 h with stirring. After it was cooled, Et2O (22
mL) and H2O (9.0 mL) were added. The ether layer was
separated and washed with 5% Na2CO3 (1 × 25 mL) and H2O
(2 × 25 mL). Removal of dried solvents gave a residue, which
was purified by column chromatography. Eluting with cyclo-
hexanes-EtOAc (8:2) gave a solid; 0.3 g (68% yield); mp 42-
43 °C. 1H NMR (CDCl3): δ 4.06 (s, 2, CH2Ar), 4.51 (d, 2, OCH2),
6.68-7.32 (m, 9, ArH), 9.76 (d, 1, CHO).
Biology. F u n ction a l An ta gon ism in Isola ted Tissu es.
Male Wistar rats (275-300 g) were killed by cervical disloca-
tion, and the organs required were isolated, freed from
adhering connective tissue, and set up rapidly under a suitable
resting tension in 20 mL organ baths containing physiological
salt solution kept at 37 °C and aerated with 5% CO2:95% O2
at pH 7.4. Concentration-response curves were constructed
by cumulative addition of agonist. The concentration of agonist
in the organ bath was increased approximately 3-fold at each
step, with each addition being made only after the response
to the previous addition had attained a maximal level and
remained steady. Contractions were recorded by means of a
force displacement transducer connected to the MacLab system
PowerLab/800. In addition, parallel experiments in which
tissues did not receive any antagonist were run in order to
check any variation in sensitivity.
[2-(2-Ben zyl-ph en oxy)eth yl]-[2-(2,6-dim eth oxy-ph en oxy)-
eth yl]a m in e Oxa la te (10). A 2.03 M solution of HCl gas in
EtOH (1.31 mL) was added to a solution of 2-(2,6-dimethoxy-
phenoxy)ethylamine21 (1.56 g, 7.91 mmol) and 20 (0.3 g, 1.33
mmol) in EtOH (10 mL), followed by the addition of NaBH3CN
(0.077 g, 1.16 mmol) and molecular sieves (4 Å). The mixture
was stirred at room temperature for 4 h, then acidified at pH
1 with 2 N HCl, filtered, and evaporated. The residue was
taken up with water and basified with 6 N KOH, and the
mixture was extracted with Et2O. After it was dried, the
solvent was evaporated and the residue was purified by column
chromatography. Eluting with EtOAc-EtOH (9:1) gave 10 as
1
the free base; 0.3 g (56% yield). H NMR (CDCl3): δ 3.03 and
3.1 (two t, 4, CH2NCH2), 3.8 (s, 6, OCH3), 3.98 (s, 2, CH2Ar),
4.17 (m, 4, OCH2 and CH2O), 6.55-7.28 (m, 12, ArH). The free
base was transformed into the oxalate salt and crystallized
from EtOH; mp 170-171 °C. Anal. (C25H29NO4‚H2C2O4) C, H,
N.
Ra t P r osta te. This tissue was used to assess R1A-adrenergic
antagonism.26 Prostate strips measuring 8-10 mm in length
and 1-2 mm in width were mounted under 2 g of tension in
modified Krebs’ solution of the following composition (mM):
NaCl, 118; KCl, 4.7; KH2PO4, 1.18; NaHCO3, 25; CaCl2, 2.5;
Molecu la r Mod elin g. Molecular models of ligands 3-5
were built starting from the X-ray crystallographic coordinates

1,4-Benzodioxan-Related Compounds
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8 1641
MgSO4‚7H2O, 1.18; glucose, 5.55. The preparations were
equilibrated for 60 min with washing every 20 min. Before
the concentration-response curves were started, tissues were
exposed to (-)-noradrenaline at a concentration of 1.0 µM. A
minimum response of 0.5 g of tension was required for the
tissue to be used for concentration-response curves. After a
90 min wash period, a cumulative concentration-response
curve to (-)-noradrenaline was constructed. After it was
washed for 90 min, the antagonist was allowed to equilibrate
with the tissue for 30 min, and then, a new concentration-
response curve to the agonist was obtained. (-)-Noradrenaline
solutions contained 0.05% Na2S2O5 to prevent oxidation.
Ra t Va s Defer en s P r osta tic P or tion . This tissue was
used to assess R1A-adrenergic antagonism.27 Prostatic portions
of 2 cm length were mounted under 0.5 g of tension in Tyrode
solution of the following composition (mM): NaCl, 130.0; KCl,
2.0; CaCl2, 1.8; MgCl2, 0.89; NaHCO3, 25.0; NaH2PO4, 0.42;
glucose, 5.6. Cocaine hydrochloride (0.1 µM) was added to the
Tyrode to prevent the neuronal uptake of (-)-noradrenaline.
The preparations were equilibrated for 60 min with washing
every 15 min. After the equilibration period, tissues were
primed two times by addition of 10 µM (-)-noradrenaline.
After another washing and equilibration period of 60 min, a
(-)-noradrenaline concentration-response curve was con-
structed (basal response). The antagonist was allowed to
equilibrate with the tissue for 30 min, followed by 30 min of
washing. Then, a new concentration-response curve to the
agonist was obtained. (-)-Noradrenaline solutions contained
0.05% Na2S2O5 to prevent oxidation.
Ra t Sp leen . This tissue was used to assess R1B-adrenergic
antagonism.28b The spleen was removed and bisected longitu-
dinally into two strips, which were suspended in tissue baths
containing Krebs solution of the following composition (mM):
NaCl, 120.0; KCl, 4.7; CaCl2, 2.5; MgSO4, 1.5; NaHCO3, 20.0;
KH2PO4, 1.2; glucose, 11.0; K2ethylenediaminetetraacetic acid
(EDTA), 0.01. (()-Propranolol hydrochloride (4.0 µM) was
added to block â-ARs. The spleen strips were placed under 1
g of resting tension and equilibrated for 2 h. The cumulative
concentration-response curves to (-)-phenylephrine were
measured isometrically and obtained at 30 min intervals; the
first one was discarded, and the second one was taken as the
control. The antagonist was allowed to equilibrate with the
tissue for 30 min, followed by 30 min of washing. Then, a new
concentration-response curve to the agonist was constructed.
Ra t Aor ta . This tissue was used to assess R1D-adrenergic
antagonism.28a Thoracic aorta was cleaned from extraneous
connective tissue and placed in Krebs solution of the following
composition (mM): NaCl, 118.4; KCl, 4.7; CaCl2, 1.9; MgSO4,
1.2; NaHCO3, 25.0; NaH2PO4, 1.2; glucose, 11.7. Cocaine
hydrochloride (0.1 µM) and (()-propranolol hydrochloride (4.0
µM) were added to prevent the neuronal uptake of (-)-
noradrenaline and to block â-ARs, respectively. Two helicoidal
strips (15 mm × 3 mm) were cut from each aorta beginning
from the end most proximal to the heart. The endothelium was
removed by rubbing with filter paper; the absence of acetyl-
choline (100 µM)-induced relaxation to preparations contracted
with (-)-noradrenaline (1 µM) was taken as an indicator that
the vessel was denuded successfully. Vascular strips were then
tied with surgical thread and suspended in a jacketed tissue
bath containing Tyrode solution. Strips were secured at one
end to Plexiglas hooks and connected to a transducer for
monitoring changes in isometric contraction. After at least 2
encoding each R1-AR subtype. Cloning and stable expression
of the human R1-AR gene was performed as previously
described.23 CHO cell membranes (30 µg proteins) were
incubated in 50 mM Tris-HCl, pH 7.4, with 0.1-0.4 nM
[3H]prazosin, in a final volume of 1.02 mL for 30 min at 25
°C, in absence or presence of competing drugs (1 pM-10 µM).
Nonspecific binding was determined in the presence of 10 µM
phentolamine. The incubation was stopped by addition of ice-
cold Tris-HCl buffer and rapid filtration through 0.2% poly-
ethyleneimine pretreated Whatman GF/B or Schleicher &
Schuell GF52 filters. Genomic clone G-21 coding for the human
5-HT1A receptor was stably transfected in a human cell line
(HeLa).24 HeLa cells were grown as monolayers in Dulbecco’s
modified Eagle’s medium (DMEM), supplemented with 10%
fetal calf serum and gentamicin (100 µg/mL), and 5% CO2 at
37 °C. Cells were detached from the growth flask at 95%
confluence by a cell scraper and were lysed in ice-cold Tris 5
mM and EDTA 5 mM buffer (pH 7.4). Homogenates were
centrifuged at 40 000g for 20 min, and pellets were resus-
pended in a small volume of ice-cold Tris 5 mM and EDTA 5
mM buffer (pH 7.4) and immediately frozen and stored at -70
°C until use. On the experimental day, cell membranes were
resuspended in binding buffer: 50 mM Tris-HCl (pH 7.4), 2.5
mM MgCl2, and 10 µM pargiline.25 Membranes were incubated
in a final volume of 1 mL for 30 min at 30 °C with 0.7-1.4
nM [3H]8-OH-DPAT, in the absence or presence of competing
drugs. Nonspecific binding was determined in the presence of
10 µM 5-HT. The incubation was stopped by the addition of
ice-cold Tris-HCl buffer and rapid filtration through 0.2%
polyethyleneimine-pretreated Whatman GF/B or Schleicher &
Schuell GF52 filters.
Da ta An a lysis. In functional studies, responses were
expressed as percentage of the maximal contraction observed
in the agonist concentration-response curve taken as a
control. Pharmacological computer programs analyzed the
agonist concentration-response curves. pA2 values were cal-
culated according to Arunlakshana and Schild51a from the dose
ratios at the EC50 values of the agonists calculated at three
antagonist concentrations. Each concentration was tested five
times, and Schild plots were constrained to slope -1, as
required by theory.51b When this method was applied, it was
always verified that the experimental data generated a line
whose derived slope was not significantly different from unity
(p > 0.05). In a number of cases, Schild analysis could not be
performed due to depression of the maximal response by high
antagonist concentrations and the resulting nonparallel slopes
of the concentration-response curves. Consequently, pKB
values were calculated from only one concentration. Com-
pounds 2-10 were tested in isolated rat prostate at only one
concentration, in the range of 0.01-1 µM, when determining
R
1A-AR blocking activity. In these cases, pKB values were
calculated according to van Rossum.51c
Binding data were analyzed using the nonlinear curve-
fitting program Allfit.52 Scatchard plots were linear in all
preparations. All pseudo-Hill coefficients (nH) were not sig-
nificantly different from unity (p > 0.05). Equilibrium inhibi-
tion constants (Ki) were derived from the Cheng-Prusoff
equation,53 Ki ) IC50/(1 + L/Kd), where L and Kd are the
concentration and the equilibrium dissociation constant of the
radioligand. pKi values are the mean ( SE of 2-3 separate
experiments performed in triplicate.
h
equilibration period under an optimal tension of 2 g,
Ack n ow led gm en t. This paper was supported by
grants from the Universities of Camerino, Bari, and
Bologna and MURST.
cumulative (-)-noradrenaline concentration-response curves
were recorded at 1 h intervals; the first two were discarded,
and the third one was taken as the control. The antagonist
was allowed to equilibrate with the tissue for 30 min before
the generation of the fourth cumulative concentration-
response curve to (-)-noradrenaline. (-)-Noradrenaline solu-
tions contained 0.05% K2EDTA and 0.9% NaCl to prevent
oxidation.
Refer en ces
(1) For part 6, see ref 12.
(2) (a) Bylund, D. B.; Eikenberg, D. C.; Hieble, J . P.; Langer, S. Z.;
Lefkowitz, R. J .; Minneman, K. P.; Molinoff, P. B.; Ruffolo, R.
R., J r.; Trendelenburg, U. IV. International Union of Pharmacol-
ogy Nomenclature of Adrenoceptors. Pharmacol. Rev. 1994, 46,
121-136. (b) Hieble, J . P.; Bylund, D. B.; Clarke, D. E.;
Ra d ioliga n d Bin d in g Assa ys. Binding to cloned human
R1-AR subtypes was performed in membranes from CHO cells
transfected by electroporation with DNA expressing the gene

1642 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8
Quaglia et al.
Eikenburg, D. C.; Langer, S. Z.; Lefkowitz, R. J .; Minneman, K.
P.; Ruffolo, R. R. International Union of Pharmacology X.
Recommendation for Nomenclature of R1-Adrenoceptors: Con-
sensus Update. Pharmacol. Rev. 1995, 47, 267-270.
omers of trans-[2-(2,6-Dimethoxyphenoxy)ethyl][3-p-tolyl-2,3-
dihydro-1,4-benzodioxin-2-yl)methyl]amine (Mephendioxan), a
Potent Competitive R1A-Adrenoreceptor Antagonist. J . Med.
Chem. 1996, 39, 2253-2258.
(3) Faure, C.; Pimoule, C.; Arbilla, S.; Langer, S. Z.; Graham, D.
Expression of R1-Adrenoceptor Subtypes in Rat Tissues: Impli-
cations for R1-Adrenoceptor Classification. Eur. J . Pharmacol.
Mol. Pharmacol. 1994, 268, 141-149.
(4) Ford, A. P. D. W.; Williams, T. J .; Blue, D. R.; Clarke, D. E. R1-
Adrenoceptor Classification: Sharpening Occam’s Razor. Trends
Pharmacol. Sci. 1994, 15, 167-170.
(5) (a) Kenny, B.; Ballard, S.; Blagg, J .; Fox, D. Pharmacological
Options in the Treatment of Benign Prostatic Hyperplasia. J .
Med. Chem. 1997, 40, 1293-1315. (b) 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: Amsterdam, 1996; pp
135-152. (c) Ruffolo, R. R., J r.; Bondinell, W.; Hieble, J . P. R-
and â-Adrenoceptors. From the Gene to the Clinic. 2. Structure-
Activity Relationships and Therapeutic Applications. J . Med.
Chem. 1995, 38, 3681-3716.
(17) Sime, J . T.; Ainsworth, A. T. Chroman Derivatives and Phar-
maceutical Compositions Containing Them. Eur. Pat. Appl.
20,018, 1980; Chem. Abstr. 1981, 94, 174883w.
(18) Berhenke, L. F.; Bein, L. E. Aryloxy Aliphatic Carboxylates. U.S.
Patent 2,516,611, 1950; Chem. Abstr. 1951, 45, 661b.
(19) Ishibashi, H.; Okada, M.; Sato, K.; Ikeda, M.; Ishiyama, K.;
Tamura, Y. Cationic Polar Cycloadditions of Phenylthionium
Ions with Olefins: A New Route to Thiochromans. Chem. Pharm.
Bull. 1985, 33, 90-95.
(20) Holava, H. M.; Partyka, R. A. 1-Phenyl-3-(aminoalkyl)-1,2,3,4-
tetrahydronaphthalenes and Their Salts as Anorectic Agents or
Central Nervous System Stimulants. U.S. Patent 3,663,608,
1972; Chem. Abstr. 1972, 77, 61675k and references therein.
(21) Woolley, D. W. Probable Evolutionary Relations of Serotonin and
Indole-Acetic Acid, and Some Practical Consequences Therefrom.
Nature 1957, 10, 630-633.
(22) Miyano, S.; Tatsuoka, T.; Suzuki, K.; Imao, K.; Satoh, F.;
Ishihara, T.; Hirotsu, I.; Kihara, T.; Hatta, M.; Horikawa, Y.;
Sumoto, K. The Synthesis and Antilipidperoxidation Activity of
4,4-Diarylbutylamines and 4,4-Diarylbutanamides. Chem. Pharm.
Bull. 1990, 38, 1570-1574.
(6) Matyus, P.; Horvath, K. R-Adrenergic Approach in the Medical
Management of Benign Prostatic Hyperplasia. Med. Res. Rev.
1997, 17, 523-535.
(7) (a) Nasu, K.; Moriyama, N.; Kawabe, K.; Tsujimoto, G.; Murai,
M.; Tanaka, T.; Yano, J . Quantification and Distribution of R1-
Adrenoceptor Subtype mRNAs in Human Prostate: Comparison
of Benign Hypertrophied Tissue and Non-Hypertrophied Tissue.
Br. J . Pharmacol. 1996, 119, 797-803. (b) Price, D. T.; Scwhinn,
D. A.; Lomasney, J . W.; Allen, L. F.; Caron, M. G.; Lefkowitz, R.
J . Identification, Quantitation, and Localization of mRNA for
Three Distinct Alpha-1 Adrenergic Subtypes in Human Prostate.
J . Urol. 1993, 150, 546-551.
(8) Cavalli, A.; Lattion, A.-L.; Hummler, E.; Nenniger, M.;
Pedrazzini, T.; Aubert, J .-F.; Michel, M. C.; Yang, M.; Lembo,
G.; Vecchine, C.; Mostardini, M.; Schmidt, A.; Beermann, F.;
Cotecchia, S. Decreased Blood Pressure Response in Mice
Deficient of the R1b-Adrenergic Receptor. Proc. Natl. Acad. Sci.
U.S.A. 1997, 94, 11589-11594.
(9) Malloy, B. J .; Price, D. T.; Price, R. R.; Bienstock, A. M.; Dole,
M. K.; Funk, B. L.; Rudner, X. L.; Richardson, C. D.; Donatucci,
C. F.; Schwinn, D. A. R1-Adrenergic Receptor Subtypes in
Human Detrusor. J . Urol. 1998, 160, 937-943.
(10) Broten, T.; Scott, A.; Siegl, P. K. S.; Forray, C.; Lagu, B.;
Nagarathnam, W. C.; Marzabadi, W. M.; Murali Dhar, T. G.;
Gluckowski, C. Alpha-1 Adrenoceptor Blockade Inhibits Detru-
sor Instability in Rats With Bladder Outlet Obstruction. FASEB
J . Abstr. 1998, 12, A445.
(23) Testa, R.; Taddei, C.; Poggesi, E.; Destefani, C.; Cotecchia, S.;
Hieble, J . P.; Sulpizio, A. C.; Naselsky, D.; Bergsma, D.; Ellis,
S.; Swift, A.; Ganguly, S.; Ruffolo, R. R.; Leonardi, A. Rec 15/
2739 (SB 216469): a Novel Prostate Selective R1-Adrenoceptor
Antagonist. Pharmacol. Commun. 1995, 6, 79-86.
(24) Fargin, A.; Raymond, J . R.; Regan, J . W.; Cotecchia, S.; Lefkow-
itz, R. J .; Caron, M. G. Effector Coupling Mechanisms of the
Cloned 5-HT1A Receptor. J . Biol. Chem. 1989, 284, 14848-14852.
(25) Fargin, A.; Raymond, J . R.; Lohse, M. J .; Kobilka, B. K.; Caron,
M. G.; Lefkowitz, R. J . The Genomic Clone G-21 Which Re-
sembles a â-Adrenergic Receptor Sequence Encodes the 5-HT1A
Receptor. Nature 1988, 335, 358-360.
(26) Pulito, V. L.; Li, X.; Varga, S. S.; Mulcahy, L. S.; Clark, K. S.;
Halbert, S. A.; Reitz, A. B.; Murray, W. V.; J oliffe, L. K. An
Investigation of the Uroselective Properties of Four Novel R1a
-
Adrenergic Receptor Subtype-Selective Antagonists. J . Phar-
macol. Exp. Ther. 2000, 294, 224-229.
(27) Eltze, M.; Boer, R.; Sanders, K. H.; Kolossa, N. Vasodilatation
Elicited by 5-HT1A Receptor Agonists in Constant-Pressure-
Perfused Rat Kidney Is Mediated by Blockade of R1A-Adreno-
receptors. Eur. J . Pharmacol. 1991, 202, 33-44.
(28) (a) Ko, F. N.; Guh, J . H.; Yu, S. M.; Hou, Y. S.; Wu, Y. C.; Teng,
C. M. (-)-Discretamine, a Selective R1D-Adrenoreceptor Antago-
nist, Isolated from Fissistigma glaucescens. Br. J . Pharmacol.
1994, 112, 1174-1180. (b) Buckner, S. A.; Oheim, K. W.; Morse,
P. A.; Knepper, S. M.; Hancock, A. A. R1-Adrenoceptor-Induced
Contractility in Rat Aorta Is Mediated by the R1D Subtype. Eur.
J . Pharmacol. 1996, 297, 241-248.
(29) Melchiorre, C.; Bolognesi, M. L.; Budriesi, R.; Chiarini, A.;
Giardina`, D.; Minarini, A.; Quaglia, W.; Leonardi, A. Search for
selective antagonists at R1-adrenoreceptors: neutral or negative
antagonism? Farmaco 1998, 53, 278-286.
(30) (a) Noguera, M. A.; Ivorra, I. D.; D’Ocon, P. Functional evidence
of inverse agonism in vascular smooth muscle. Br. J . Pharmacol.
1996, 119, 158-164. (b) Rossier, O.; Abuin, L.; Fanelli, F.;
Leonardi, A.; Cotecchia, S. Inverse Agonism and Neutral An-
tagonism at R1a- and R1b-Adrenergic Receptor Subtypes. Mol.
Pharmacol. 1999, 56, 858-866.
(11) Meyer, M. D.; Altenbach, R. J .; Basha, F. Z.; Carrol, W. A.;
Condon, S.; Elmore, S. W.; Kerwin, J . F., J r.; Sippy, K. B.; Tietje,
K.; Wendt, M. D.; Hancock, A. A.; Brune, M. E.; Buckner, S. A.;
Drizin, I. Structure-Activity Studies for
a Novel Series of
Tricyclic Substituted Hexahydrobenz[e]isoindole R1A Adreno-
ceptor Antagonists as Potential Agent for the Symptomatic
Treatment of Benign Prostatic Hyperplasia (BPH). J . Med.
Chem. 2000, 43, 1586-1603.
(12) Quaglia, W.; Pigini, M.; Piergentili, A.; Giannella, M.; Marucci,
G.; Poggesi, E.; Leonardi, A.; Melchiorre, C. Structure-Activity
Relationships in 1,4-Benzodioxan-Related Compounds. 6. Role
of the Dioxane Unit on Selectivity for R1-Adrenoreceptor Sub-
types. J . Med. Chem. 1999, 42, 2961-2968 and references
therein.
(13) For example, see Melchiorre, C.; Belleau, B. Adrenoceptors and
Catecholamine Action, Part A; Kunos, G., Ed.; Wiley: New York,
1981; pp 131-179.
(14) Melchiorre, C.; Brasili, L.; Giardina`, D.; Pigini, M.; Strappa-
ghetti, G. 2-[[[2-(2,6-Dimethoxyphenoxy)ethyl]amino]methyl]-
1,4-benzoxathian: a New Antagonist with High Potency and
Selectivity Toward R1-Adrenoreceptors. J . Med. Chem. 1984, 27,
1535-1536.
(31) For a review, see Herbert, T. E.; Bouvier, M. Structural and
Functional Aspects of G Protein-Coupled Receptor Oligomeriza-
tion. Biochem. Cell Biol. 1998, 76, 1-11.
(32) For example, see Zeng, F.-Y.; Wess, J . Molecular Aspects of
Muscarinic Receptor Dimerization. Neuropsychopharmacology
2000, 23, S19-S29.
(33) (a) J ones, K. A.; Borowsky, B.; Tamm, J . A.; Craig, D. A.; Durkin,
M. M.; Dai, M.; Yao, W. J .; J ohnson, M.; Gunwaldsen, C.; Huang,
L. Y.; Tang, C.; Shen, Q.; Salon, J . A.; Morse, K.; Laz, T.; Smith,
K. E.; Nagarathnam, D.; Noble, S. A.; Branchek, T. A.; Gerald,
C. GABAB Receptors Function as a Heteromeric Assembly of the
Subunits GABABR1 and GABABR2. Nature 1998, 396, 674-679.
(b) Kaupmann, K.; Malitschek, B.; Schuler, V.; Heid, J .; Froestl,
W.; Beck, P.; Mosbacher, J .; Bischoff, S.; Kulik, A.; Shigemoto,
R.; Karschin, A.; Bettler, B. GABAB-Receptor Subtypes Assemble
into Functional Heteromeric Complexes. Nature 1998, 396, 683-
687. (c) Kuner, R.; Kohr, G.; Grunewald, S.; Eisenhardt, G.;
Bach, A.; Kornau, H. C. Role of Heteromer Formation in GABAB
Receptor Function. Science 1999, 283, 74-77. (d) White, J . H.;
Wise, A.; Main, M. J .; Green, A.; Fraser, N. J .; Disney, G. H.;
Barnes, A. A.; Emson, P.; Foord, S. M.; Marshall, F. H. Hetero-
dimerization Is Required for the Formation of a Functional
GABAB Receptor. Nature 1998, 396, 679-682.
(15) Pigini, M.; Brasili, L.; Giannella, M.; Giardina`, D.; Gulini, U.;
Quaglia, W.; Melchiorre, C. Structure-Activity Relationships in
1,4-Benzodioxan-Related Compounds. Investigation on the Role
of the Dehydrodioxane Ring on R1-Adrenoreceptor Blocking
Activity. J . Med. Chem. 1988, 31, 2300-2304.
(16) (a) Quaglia, W.; Pigini, M.; Giannella, M.; Melchiorre, C.
3-Phenyl Analogues of 2-[[[2-(2,6-Dimethoxyphenoxy)ethyl]ami-
no]methyl]-1,4-benzodioxan (WB 4101) as Highly Selective R1-
Adrenoreceptor Antagonists. J . Med. Chem. 1990, 33, 2946-
2948. (b) Quaglia, W.; Pigini, M.; Tayebati, S. K.; Piergentili,
A.; Giannella, M.; Marucci, G.; Melchiorre, C. Structure-Activity
Relationships in 1,4-Benzodioxan-Related Compounds. 4. Effect
of Aryl and Alkyl Substituents at Position 3 on R-Adrenoreceptor
Blocking Activity. J . Med. Chem. 1993, 36, 1520-1528. (c)
Quaglia, W.; Pigini, M.; Tayebati, S. K.; Piergentili, A.; Gian-
nella, M.; Leonardi, A.; Taddei, C.; Melchiorre, C. Synthesis,
Absolute Configuration, and Biological Profile of the Enanti-

1,4-Benzodioxan-Related Compounds
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 8 1643
(34) J ordan, B. A.; Devi, L. A. G-Protein-Coupled Receptor Hetero-
dimerization Modulates Receptor Function. Nature 1999, 399,
697-700.
Switch Residue Involved in Activation and Catecholamine Ring
Aromatic Bonding. J . Biol. Chem. 1999, 274, 16320-16330.
(46) Allen, F. H.; Kennard, O. 3D Search and Research Using the
Cambridge Structural Database. Chem. Des. Autom. News 1993,
8, 31-37.
(47) McMartin, C.; Bohacek, R. S. QXP: Powerful, Rapid Computer
Algorithms for Structure-Based Drug Design. J . Comput.-Aided
Mol. Des. 1997, 11, 333-344.
(48) Carpy, A.; Colleter, J . C.; Leger, J . M. N-{2-(2,6-Dimethoxy-
phenoxy)ethyl}-1,4-benzodioxane-2-methylamine Hydrochloride
(WB-4101). Cryst. Struct. Commun. 1981, 10, 1391-1395.
(49) Cavalli, A.; Fanelli, F.; Taddei, C.; De Benedetti, P. G.; Cotecchia,
S. Amino Acids of the R1b-Adrenergic Receptor Involved in
Agonist Binding: Differences in Docking Catecholamines to
Receptor Subtypes. FEBS Lett. 1996, 399, 9-13.
(50) Kraulis, P. J . MOLSCRIPT: A Program to Produce Both
Detailed and Schematic Plots of Protein Structures. J . App.
Crystallogr. 1991, 24, 946-950.
(51) (a) Arunlakshana, O.; Schild, H. O. Some Quantitative Uses of
Drug Antagonists. Br. J . Pharmacol. 1959, 14, 48-58. (b)
Tallarida, R. J .; Cowan, A.; Adler, M. W. pA2 and Receptor
Differentiation: A Statistical Analysis of Competitive Antago-
nism. Life Sci. 1979, 25, 637-654. (c) van Rossum, J . M.
Cumulative Dose-Response Curves. II. Techniques for the
Making of Dose-Response Curves in Isolated Organs and the
Evaluation of Drug Parameters. Arch. Int. Pharmacodyn. Ther.
1963, 143, 299-330.
(52) De Lean, A.; Munson, P. J .; Rodbard, D. Simultaneous Analysis
of Families of Sigmoidal Curves: Application to Bioassay,
Radioligand Assay, and Physiological Dose-Response Curves.
Am. J . Physiol. 1978, 235, E97-E102.
(35) Schertler, G. F.; Villa, C.; Henderson, R. Projection Structure
of Rhodopsin. Nature 1993, 362, 770-772.
(36) Unger, V. M.; Hargrave, P. A.; Baldwin, J . M.; Schertler, G. F.
Arrangement of Rhodopsin Transmembrane R-Helices. Nature
1997, 389, 203-206.
(37) Swindells, M. B.; Thornton, J . M. Modelling by Homolog. Curr.
Opin. Struct. Biol. 1991, 1, 219-223.
(38) Palczewski, K.; Kumasaka, T.; Hori, T.; Behnke, C. A.; Mo-
toshima, H.; Fox, A. B.; Le Trong, I.; Teller, D. C.; Okada, T.;
Stenkamp, R. E.; Yamamoto, M.; Miyano, M. Crystal Structure
of Rhodopsin: a G Protein-Coupled Receptor. Science 2000, 289,
739-745
(39) Teller, D. C.; Okada, T.; Behnke, C. A.; Palczewski, K.; Sten-
kamp, R. E. Advances in Determination of a High-Resolution
Three-Dimensional Structure of Rhodopsin,
a Model of G-
Protein-Coupled Receptors (GPCRs). Biochemistry 2001, 40,
7761-7772.
(40) Kristiansen, K.; Dahl, S. G.; Edvardsen, O. A Database of
Mutants and Effects of Site-Directed Mutagenesis Experiments
on G Protein-Coupled Receptors. Proteins 1996, 26, 81-94.
(41) Muller, G. Towards 3D Structures of G Protein-Coupled Recep-
tors: a Multidisciplinary Approach. Curr. Med. Chem. 2000, 7,
861-888.
(42) Carrieri, A.; Centeno, N. B.; Rodrigo, J .; Sanz, F.; Carotti, A.
Theoretical Evidence of a Salt Bridge Disruption As the Initiat-
ing Process for the
Molecular Dynamics and Docking Study. Proteins 2001, 43, 382-
394.
R1d-Adrenergic Receptor Activation: a
(43) Goodford, P. J . A Computational Procedure for Determining
Energetically Favourable Binding Sites on Biologically Impor-
tant Macromolecules. J . Med. Chem. 1985, 28, 849-857.
(44) For a fully detailed description of this methodology, the reader
is referred to the GRID manual available on line from the
Internet at http://www.moldiscovery.com.
(53) Cheng, Y. C.; Prusoff, W. H. Relationship Between the Inhibition
Constant (Ki) and the Concentration of Inhibitor Which Causes
50% Inhibition (I50) of an Enzymatic Reaction. Biochem. Phar-
macol. 1973, 22, 3099-3108.
(45) Chen, S.; Xu, M.; Lin, F.; Lee, D.; Riek, P.; Graham, R. M. Phe310
in Transmembrane VI of the R1b-Adrenergic Receptor Is a Key
J M011066N