Synthesis of new δ2-isoxazoline derivatives and their pharmacological characterization as β-adrenergic receptor antagonists

Article abstract of DOI:10.1016/S0968-0896(97)10051-7

A series of Δ²-isoxazoline derivatives structurally related to Broxaterol 1 and Falintolol 3 has been prepared and evaluated for their binding affinity to β₁- and β₂-adrenergic receptors. Among the tested compounds only the 3-isopropenyl anti derivative 4d is as active as the reference compounds. An electron-releasing group, probably operating through a π- π interaction, in the 3-position of the isoxazoline nucleus greatly enhances the affinity of the compounds. Conversely, the closest analogs of Broxaterol (3-bromo Δ²-isoxazolines 4a and 5a) are at least one order of magnitude less active than the model compound 1. Throughout the series of derivatives the anti stereoisomers are invariably more active than their syn counterparts.

Full text of DOI:10.1016/S0968-0896(97)10051-7

BIOORGANIC &
MEDICINAL
CHEMISTRY
Bioorganic & Medicinal Chemistry 6 (1998) 401±408
Synthesis of New Á2-Isoxazoline Derivatives and
their Pharmacological Characterization as ꢀ-Adrenergic
Receptor Antagonists
Paola Conti,a Clelia Dallanoce,a Marco De Amici,a
Carlo De Michelia,* and Karl-Norbert Klotzb
a
Á
Istituto di Chimica Farmaceutica e Tossicologica, Universita di Milano, v.le Abruzzi, 42-20131 Milano, Italy
È
bInstitute of Pharmacology and Toxicology, Julius-Maximilians Universitat, Versbacher Str. 9 - D-97078 Wurzburg, Germany
È
Received 1 May 1997; accepted 28 October 1997
AbstractÐA series of Á2-isoxazoline derivatives structurally related to Broxaterol 1 and Falintolol 3 has been prepared
and evaluated for their binding anity to b1- and b2-adrenergic receptors. Among the tested compounds only the
3-isopropenyl anti derivative 4d is as active as the reference compounds. An electron-releasing group, probably oper-
ating through a ꢀ ꢀ interaction, in the 3-position of the isoxazoline nucleus greatly enhances the anity of the
compounds. Conversely, the closest analogs of Broxaterol (3-bromo Á2-isoxazolines 4a and 5a) are at least one order
of magnitude less active than the model compound 1. Throughout the series of derivatives the anti stereoisomers are
invariably more active than their syn counterparts. # 1998 Elsevier Science Ltd. All rights reserved.
Introduction
and evaluated as semi-rigid analogs of oxime ethers,
though their structure reproduces an unfavored con-
formation of the parent open derivatives.7±9 Since
Broxaterol may in turn be regarded as the cyclic form of
an oxime ether, we chose 1 as the model compound to
plan the synthesis of stereoisomeric Á2-isoxazolines 4a±e
and 5a±e (Figure 1). The substituent on the heterocyclic
ring was chosen by taking into account the structure of
Broxaterol (i.e., 4a/5a, 4b/5b, and 4e/5e) as well as the
structure of Falintolol (i.e., 4c/5c and 4d/5d). This paper
deals with the synthesis of derivatives 4a±e and 5a±e and
their binding anities to membrane preparations con-
taining b1- and b2-adrenergic receptors. We did not
evaluate the anity of our compounds for the b3-adre-
nergic receptor because, according to the examples
reported in the literature,10 the selectivity for such a
receptor can be achieved by appending a polar group on
the substituent at nitrogen.
Broxaterol (‹)-1, a potent and selective b2-adrenergic
agonist,1,2 has been developed as a potential broncho-
dilatory agent in asthma therapy.3 Inspection of the
structure of 1 suggests a bioisosterism between its het-
erocyclic moiety and the catechol group of Iso-
proterenol 2, the reference drug (Figure 1). This
hypothesis is further supported by the stereochemical
requirements of the two compounds. Despite the di€er-
ent notation, the spatial arrangement of the groups
around the chiral center is in fact the same in both ago-
nists,4 along with common high values of the eudismic
ratio (ERꢀ100).1 Figure 1 reports the absolute con®g-
uration of the eutomer of Broxaterol and Isoproterenol.
It is well known that aliphatic oxime ethers exhibit
interesting b-adrenergic blocking activities.5,6 Among a
series of derivatives, Falintolol 3 (Figure 1) emerged as a
b1/b2-adrenergic antagonist.6 Based on these results, a
number of isoxazolinylethanolamines has been designed
Chemistry
The synthesis of compounds 4a±e and 5a±e was
achieved following a general strategy based on the 1,3-
dipolar cycloaddition of halogenoximes 6a±c to an
Key words: Synthesis; binding anity; adrenergic antagonists;
Broxaterol analogs; Falintolol analogs.
*Corresponding author.
0968-0896/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved
PII: S0968-0896(97)10051-7

402
P. Conti et al./Bioorg. Med. Chem. 6 (1998) 401±408
excess of butadiene in the presence of sodium bicarbo-
nate (Scheme 1). Cycloadducts 7a±c were obtained in
fairly good yields. Ole®nic compounds 7a±c were oxi-
dized with 3-chloroperoxybenzoic acid (mCPBA) to
produce an equimolar mixture of anti 8a±c and syn 9a±c
epoxides, which were separated by column chromato-
graphy. A Wittig methylenation carried out on epoxides
8c and 9c yielded isopropenyl derivatives 8d and 9d,
respectively. Aminolysis of epoxides 8a±d and 9a±d
(Scheme 1) with excess t-butylamine in methanol pro-
duced an almost quantitative yield of aminoalcohols 4a±
d and 5a±d. Derivatives 4e and 5e were obtained, in
excellent yield, by re¯uxing a methanol solution of 4a
and 5a, respectively, in the presence of potassium car-
bonate. Since the transformation of the epoxides into
aminoalcohols did not a€ect the relative con®gurations
of the two chiral centers, anti epoxides 8a±d yielded anti
aminoalcohols 4a±d exclusively, in analogy to the syn
epoxides 9a±d which produced syn derivatives 5a±d
solely.
their 1H NMR pattern with that of the corresponding
analogs 4b and 5b, respectively. In particular, the value
0
of the coupling constant J5,2 of the anti derivatives
(J=5.0±6.5 Hz) was always higher than that of the cor-
responding syn stereoisomer (J=3.2±3.5 Hz).
Results and discussion
Broxaterol 1 and isoxazoline derivatives 4a±e and 5a±e
were assessed for binding anity to membranes con-
taining b1 and b2 adrenergic receptors. Table 1 reports
their dissociation constants (Ki-values) obtained in
competition experiments with 125I-cyanopindolol (125I-
CYP) as a radioligand. Membranes were obtained from
CHO cells transfected with the human b2-adrenergic
receptors and from rat C6 glioma cells containing b1-
adrenergic receptors.
The ®rst point to be addressed is the lack of selectivity
observed with Broxaterol, the reference compound. In
contrast with in vitro functional data where 1 shows
remarkable b2-selectivity as an agonist,1 binding data
(Table 1) give a very low (1.6) b1/b2 ratio. This dichotomy
Compounds 4b and 5b are known from literature.9 The
structure of the other anti (4) and syn (5) aminoalcohol
derivatives was assigned on the basis of the analogy of
Figure 1.

P. Conti et al./Bioorg. Med. Chem. 6 (1998) 401±408
403
Scheme 1. (a) NaHCO3/CH3COOEt; (b) mCPBA/CH2Cl2; (c) Ph3P=CH2/Et2O; (d) t-BuNH2/MeOH; (e) MeOH/K2CO3.
Table 1. Binding anities of compounds 4a±e and 5a±e to rat C6 glioma cells (b1) and to CHO cells transfected with b2-adrenergic
receptors
Compound
Kia(mM) for b2
Kia (mM) for b1
b1/b2 ratiob
1
0.20 (0.13Ð0.31)c
1.44 (0.96Ð2.14)
36.71 (26.8Ð50.3)
6.85 (5.55Ð8.45)
0.32 (0.27Ð0.37)c
2.64 (1.63Ð4.27)
29.70 (13.7Ð64.5)
8.64 (6.02Ð12.4)
1.6
1.8
0.8
1.3
4a
5a
4b
5b
4c
5c
4d
5d
4e
5e
3d
>100
>100
12.60 (7.09Ð22.5)
>100
4.53 (3.52Ð5.82)
28.19 (19.6Ð40.5)
0.31 (0.26Ð0.38)
1.53 (0.87Ð2.71)
2.06 (1.68Ð2.54)
8.79 (4.89Ð15.8)
0.35‹0.042
2.8
>3.5
2.7
0.83 (0.66Ð1.06)
3.66 (2.26Ð5.91)
4.16 (3.21Ð5.38)
14.90 (9.16Ð24.1)
0.64‹0.18
2.4
2.0
1.7
1.8
aKi values are geometric means of 3±4 experiments.
b1/Ki (b2)/1/Ki (b1).
cThe 95% con®dence limits are reported in parentheses.
dData taken from Ref. 11; binding anity to rat lung (b2) and rabbit heart (b1).

404
P. Conti et al./Bioorg. Med. Chem. 6 (1998) 401±408
was examined in detail via competition experiments with
the radioligand antagonist 125I-CYP, both in the pre-
sence and in the absence of GTP, and by measurement
of the activation of adenylate cyclase. In competition
experiments with 125I-CYP the binding curves of Brox-
aterol to b2-receptors evidenced two anity states (high
and low anity states), as expected for agonists (Figure
2).
membrane protein/min). Marginal activation of the
enzyme was observed via b1-receptors (5.5 pmol versus
4.3 pmol) (Figure 3). The EC50 values of Broxaterol for
b1- and b2-adrenergic receptors are 1.3 and 0.17 mM,
respectively. Thus, Broxaterol appears to be a full ago-
nist at b2-receptors and a partial agonist, with a very
low intrinsic activity, at b1-receptors.
The high anity binding of Broxaterol to b2-receptors,
detected in the absence of GTP, makes it roughly 100-fold
more potent at this receptor than at the b1-subtype.
These data provide an explanation for the b2-selectivity
observed in functional experiments, where Broxaterol
relaxes spontaneously contracted guinea pig tracheal
chain at low doses (ED50=6 nM) and marginally
a€ects the contraction of the guinea pig atria
(ED25=327 nM),1 which could not be con®rmed in
binding experiments in the presence of GTP.
The addition of GTP (100 mM) provoked a shift of the
competition curves to the right, abolishing the high a-
nity state. Similar experiments carried out with b1-
receptors demonstrated a single anity state both in the
presence and the absence of GTP (not shown), suggest-
ing that Broxaterol behaves as an antagonist at this
adrenergic receptor subtype. Broxaterol caused about a
fourfold stimulation of the adenylate cyclase (Figure 3)
via b2-receptors (10.6 pmol/mg membrane protein/min)
when compared to the basal activity (2.9 pmol/mg
Inspection of the data reported in Table 1 shows that
derivatives 4a and 5a, the closest analogs of Broxaterol,
are at least one order of magnitude less active than the
reference compound. Furthermore, at variance with 1,
which behaves as an agonist at b2-receptors, compounds
4a and 5a appear to be antagonists as suggested by the
shape of their binding curve (not shown) in the absence
of GTP.
Since the same trend is observed for the other deriva-
tives of the series, it appears that this set of isoxazoline
derivatives behaves as antagonists at both b1- and b2-
receptors and seems to be referred to Falintolol 3 or
related linear oxime ethers rather than to Broxaterol. As
a consequence, the heterocyclic moiety of Broxaterol
can be viewed as the link between catechol, i.e. Iso-
proterenol, and oxime derivatives. A further analogy
among the present isoxazoline derivatives (4a±e and 5a±e)
and linear oxime ethers is the in¯uence of the sub-
stituent on adrenergic binding potency. In both series
the presence of electron-donating groups, probably
operating through a p±p interaction, increases the adre-
nergic binding potency. The bulkiness of the group also
plays an important role in determining the anity of the
ligand for the receptors under investigation. As a matter
of fact, Falintolol emerged as the most promising adre-
nergic antagonist of a huge number of linear aliphatic
oxime ethers;5,6 worth noting is the presence of the
cyclopropyl group which is isoelectronic with the vinyl
moiety. Isoxazoline 4d is the most active compound in
the set of derivatives we prepared and evaluated. It dis-
plays a binding anity equal to Falintolol and a curve
with only the low anity binding site typical of antago-
nists. Furthermore, its structure contains the iso-
propenyl group which can be considered isoelectronic
with the cyclopropyl moiety of the reference compound.
Falintolol is the racemic form of a mixture of the anti 3a
Figure 2. Competition curve of Broxaterol for 125I-CYP bind-
ing to b2-receptors without (*) and in the presence (&) of
100 mM GTP.
Figure 3. Activity of adenylate cyclase mediated by the stimu-
lation of b1- and b2-adrenergic receptors.

P. Conti et al./Bioorg. Med. Chem. 6 (1998) 401±408
405
and syn 3b stereoisomers (Figure 4). The individual
stereoisomers have been prepared and evaluated in
vitro for adrenergic activity on guinea pig atria (b1)
and trachea (b2).12 It emerged that the syn and anti
isomers of 3 display comparable biological e€ects.
Similar potencies have also been observed among the
enantiomers of 3.12
heterocyclic nucleus greatly a€ected the anity for the
adrenergic receptors. It remains to be ascertained if the
activity of cyclic derivatives 4 and 5 is also a€ected by
the absolute con®guration of the chiral centers.
Conclusion
An accurate investigation of the conformational pro®le
of linear aliphatic oxime ethers has been carried out.13 It
turned out that there is a certain degree of freedom
around the O±CH2 bond; when ꢁ2 ranges from 180ꢁ to
90ꢁ the energy di€erences vary within 0.12 Kcal/mol.
On the other hand, the conformation in which ꢁ1 is 0ꢁ
(3c) was found to be unfavored with respect to the pre-
ferred conformations by ca. 30±40 Kcal/mol. As evident
from Figure 4, isoxazoline derivatives 4 and 5 represent
the cyclic form of the unfavorable conformation 3c.
Since derivative 4d is as active as Falintolol 3, we can
deduce that the adrenergic binding potency is not
strictly dependent upon the favorable conformations
around the N±O bond, but a large degree of conforma-
tional freedom is tolerated for the accommodation of
the pharmacophoric groups. At this stage, a require-
ment for the cyclic derivatives is the relative con®gura-
tion of the two chiral centers, since the ethanolamine
side chain needs to be located in a spatial arrangement
suitable for the interaction with the complementary
receptor subsite. In all the pairs of derivatives 4/5 the
anti stereoisomer 4 was always more active than 5, the
syn counterpart. The electronic character as well the
bulkiness of the group located in the 3 position of the
In conclusion, the present results evidence that isox-
azolines derivatives 4 and 5 are not related biologically
to Broxaterol 1 but more appropriately to aliphatic
oxime ethers such as 3. The in¯uence of the substituent
on the biological activity among the two series of deri-
vatives is similar, electron-releasing groups highly
enhancing the anity for the adrenergic receptors. A
detailed investigation of the conformational pro®le of
derivatives 4 and 5 is underway and will be reported in
due course along with the synthesis and pharmacologi-
cal characterization of the chiral forms of 4d and 5d,
and their cyclopropyl analogs.
Experimental
Materials and methods
1H and 13C NMR spectra were recorded with a Bruker
AC-E 200 (200 MHz) spectrometer in CDCl3 solution;
chemical shifts (d) are expressed in ppm and coupling
constants (J) in Hertz. TLC were performed on com-
mercial silica gel GF254 plates; spots were further evi-
denced by spraying with a dilute alkaline potassium
Figure 4.

406
P. Conti et al./Bioorg. Med. Chem. 6 (1998) 401±408
permanganate solution. Liquid compounds were char-
acterized by the oven temperature for Kugelrohr dis-
tillations. Melting points were determined on a Buchi
apparatus and are uncorrected. Microanalyses of new
compounds agreed with theoretical value to within ‹
0.3%. Halogenoximes 6a±c were prepared following
standard procedures.14±16 Broxaterol was prepared fol-
lowing a published methodology.4 Aminoalcohols 4b
and 5b have been previously described.9
distilled at 100±105 ꢁC/1.5 mm Hg; yield 35%; 1H NMR
d 2.84 (m, 2), 3.14 (m, 1), 3.18 (dd, 1, J=7.4 and 17.6),
3.36 (dd, 1, J=10.6 and 17.6), 4.72 (ddd, 1, J=4.1, 7.4
and 10.6). Stereoisomeric epoxides 8b and 9b have been
previously described.9 Compound 8c distilled at 75±
80 ꢁC/0.7 mm Hg; yield 41%; 1H NMR d 2.49 (s, 3),
2.60 (dd, 1, J=2.5 and 4.3), 2.86 (dd, 1, J=4.1 and 4.3),
2.95 (dd, 1, J=7.8 and 17.4), 3.09 (dd, 1, J=6.1 and
17.4), 3.21 (m, 1), 4.86 (ddd, 1, J=4.2, 6.1 and 7.8).
Compound 9c distilled at 100±105 ꢁC/1.5 mm Hg; yield
37%; 1H NMR d 2.47 (s, 3), 2.83 (m, 2), 3.07 (dd, 1,
J=7.8 and 17.4), 3.10 (m, 1), 3.25 (dd, 1, J=11.4 and
17.4), 4.77 (ddd, 1, J=3.8, 7.8 and 11.4).
Synthesis of isoxazoline derivatives 7a±c. To an ethyl
acetate solution (70 mL) of halogenoxime 6 (30 mmol),
cooled to 0 ꢁC, were added butadiene (10 mL, 0.12 mol)
and solid sodium bicarbonate (8.4 g, 0.1 mol). The mix-
ture was stirred at room temperature until dis-
appearance of the halogenoxime; the progress of the
reaction was monitored by TLC. (eluant: 20% ethyl
acetate/cyclohexane). The slurry was poured in water,
the organic layer separated and the aqueous phase was
further extracted with ethyl ether (2Â20 mL). The
organic extracts were dried over anhydrous sodium
sulfate and the solvents removed under vacuum. The
residue was Kugelrohr distilled to give cycloadducts 7a±
c. Compound 7a distilled at 95±100 ꢁC/9 mm Hg. Yield:
82%; 1H NMR d 3.02 (dd, 1, J=9.6 and 16.8), 3.36
(dd, 1, J=10.8 and 16.8), 5.09 (ddd, 1, J=6.8, 9.6 and
10.8), 5.33 (m, 2), 5.94 (m, 1). Compound 7b distilled at
100±105 ꢁC/20 mm Hg.Yield: 65%;9 Compound 7c dis-
tilled at 70±75 ꢁC/1 mm Hg.Yield: 75%; 1H NMR d 2.51
(s, 3), 2.91 (dd, 1, J=9.6 and 17.0), 3.23 (dd, 1, J=10.4
and 17.0), 5.15 (ddd, 1, J=6.6, 9.6 and 10.4), 5.41 (m,
2), 5.96 (m, 1).
Synthesis of epoxides 8d and 9d. To a stirred suspension
of methylenetriphenylphosphorane, generated from
methyltriphenyl-phosphonium bromide (1.8 g, 5 mmol)
and potassium t-butoxide (0.6 g, 4.8 mmol) in anhy-
drous tetrahydrofuran (50 mL), a solution of epoxide 8c
(or 9c) in THF (0.5 g, 3.3 mmol±15 mL) was added
dropwise and stirring was continued at room tempera-
ture overnight. After the usual workup, 8d (or 9d) was
puri®ed through column chromatography on silica gel
(eluant: 20% ethyl acetate/cyclohexane). Compound 8d
distilled at 80±85 ꢁC/1.5 mm Hg. Yield 53%; 1H NMR d
2.05 (s, 3), 2.65 (dd, 1, J=2.8 and 4.7), 2.87 (dd, 1,
J=4.2 and 4.4), 3.04 (dd, 1, J=7.5 and 16.3), 3.12 (m,
1), 3.18 (dd, 1, J=10.6 and 16.3), 4.59 (ddd, 1, J=4.2,
7.5 and 10.6), 5.22 (bs, 1), 5.35 (bs, 1). Compound 9d
distilled at 80±85 ꢁC/1.5 mm Hg. Yield 52%; 1H NMR d
2.05 (s, 3), 2.83 (m, 2), 3.09 (dd, 1, J=7.9 and 16.6), 3.17
(m, 1), 3.25 (dd, 1, J=11.0 and 16.6), 4.67 (ddd, 1,
J=4.6, 7.9 and 11.0), 5.21 (bs, 1), 5.35 (bs, 1).
General procedure for the synthesis of epoxides 8a±c and
9a±c. To a magnetically stirred solution of cyclo-
adducts 7 (24 mmol) in anhydrous dichloromethane
(50 mL) a dichloromethane solution (75 mL) of 55%
3-chloroperoxybenzoic acid (16.7 g, 48 mmol) was added
dropwise. The mixture was stirred at room temperature
until TLC (eluant: 20% ethyl acetate/cyclohexane)
showed the disappearance of the starting material. The
mixture was sequentially treated with 20% aqueous
solutions of sodium iodide and sodium thiosulfate. The
organic layer was separated and the aqueous phase
extracted with dichloromethane (3Â30 mL). The pooled
organic extracts were washed with a 20% potassium
carbonate solution, dried over anhydrous sodium sul-
fate and evaporated to dryness. The crude residue was
puri®ed by ¯ash-chromatography (eluant: 20% ethyl
acetate/cyclohexane) to yield comparable amounts of
anti epoxides 8a±c, as the ®rst fraction, followed by syn
epoxides 9a±c. The anti/syn ratio was in all the exam-
ined reactions near to 1. Compound 8a distilled at 100±
General procedure for the synthesis of aminoalcohols 4a±d
and 5a±d. A stirred solution of epoxides 8a±d or 9a±d
(10 mmol) and t-butylamine (60 mmol) in methanol
(50 mL) was re¯uxed until TLC (eluant: 20% ethyl
acetate/cyclohexane) evidenced the disappearance of the
starting material. The solvent and excess t-butylamine
were removed under vacuum; the oily residue was dis-
solved in 6 N HCl (15 mL) and washed with ethyl ether
(2Â10 mL). The aqueous layer was alkalinized with
solid sodium carbonate and extracted with dichlor-
omethane (3Â15 mL). The organic layer was separated,
dried over anhydrous sodium sulfate, and the solvent
evaporated under vacuum. The residue was dissolved in
methanol and treated with a threefold excess anhydrous
oxalic acid. The salt, separated by addition of anhy-
drous ethyl ether, was crystallized from the appropriate
solvent. 1H and 13C NMR spectra were recorded on the
corresponding free bases. Compound 4a.1/2C2H2O4:
colorless prisms (from abs ethanol), mp 218±219 ꢁC
dec.; 1H NMR d 1.10 (s, 9), 2.52 (dd, 1, J=7.7 and
12.1), 2.80 (dd, 1, J=3.8 and 12.1), 3.24 (dd, 1, J=10.4
and 16.8), 3.32 (dd, 1, J=8.1 and 16.8), 3.62 (m, 1), 4.51
105 ꢁC/1.5 mm Hg; yield 37%; H NMR d 2.61 (dd, 1,
1
J=1.9 and 3.7), 2.87 (dd, 1, J=1.4 and 3.7), 3.00±3.30
(m, 3), 4.68 (ddd, 1, J=4.6, 7.8 and 11.2). Compound 9a

P. Conti et al./Bioorg. Med. Chem. 6 (1998) 401±408
407
(ddd, 1, J=6.0, 8.1 and 10.4); 13C NMR d 30.0 (CMe3),
44.2 (CH2N), 44.9 (C-4), 51.3 (NHC), 70.8 (CHOH),
84.2 (C-5), 139.0 (C-3). Anal C10H18BrN2O4 (C, H, N).
Compound 5a.C2H2O4: colorless needles (from 2-pro-
anhydrous sodium sulfate, and the solvent removed
under vacuum. The residue was dissolved in methanol
(15 mL) and treated with a threefold excess of anhy-
drous oxalic acid to yield crude 4e.C2H2O4 (or
5e.C2H2O4). Compound 4e.C2H2O4: colorless needles
panol/ethyl ether), mp 188±190 ꢁC; H NMR d 1.08 (s,
1
(from 2-propanol), mp 178±180 ꢁC; H NMR d 1.03 (s,
1
9), 2.71 (m, 2), 3.21 (dd, 1, J=4.9 and 11.4), 3.29 (dd, 1,
J=3.1 and 11.4), 3.60 (m, 1), 4.67 (ddd, 1, J=3.4, 8.8
and 10.4); 13C NMR d 29.9 (CMe3), 44.2 (CH2N), 44.9
(C-4), 51.3 (NHC), 71.1 (CHOH), 84.5 (C-5), 138.8 (C-
3). Anal C11H19BrN2O6 (C, H, N). Compound 4b.1/
2C2H2O4: colorless prisms (from abs ethanol), mp 233±
235 ꢁC dec.; 13C NMR d 14.0 (Me), 29.9 (CMe3), 41.4
(CH2N), 45.4 (C-4), 51.2 (NHC), 72.1 (CHOH), 82.8
(C-5), 156.8 (C-3). Anal C11H21N2O4 (C, H, N). Com-
pound 5b.C2H2O4: colorless prisms (from 2-propanol/
ethyl ether), mp 129±131 ꢁC; 13C NMR d 14.0 (Me), 29.9
(CMe3), 41.7 (CH2N), 45.6 (C-4), 51.0 (NHC), 72.1
(CHOH), 82.8 (C-5), 156.8 (C-3). Anal C12H22N2O6 (C,
H, N). Compound 4c.C2H2O4: colorless needles (from
2-propanol), mp 149±151 ꢁC; 1H NMR d 1.10 (s, 9), 2.47
(dd, 1, J=8.5 and 12.2), 2.79 (dd, 1, J=3.9 and 12.2),
3.09 (dd, 1, J=11.3 and 18.0), 3.19 (dd, 1, J=8.1 and
9), 2.44 (dd, 1, J=8.2 and 11.8), 2.68 (dd, 1, J=3.7 and
11.8), 2.89 (dd, 1, J=10.0 and 16.5), 3.01 (dd, 1, J=8.0
and 16.5), 3.64 (m, 1), 3.77 (s, 3), 4.43 (ddd, 1, J=6.0,
8.0 and 10.0); 13C NMR d 29.9 (CMe3), 35.1 (C-4), 45.3
(CH2N), 51.2 (NHC), 58.2 (OMe), 71.3 (CHOH), 84.1
(C-5), 169.0 (C-3). Anal C12H22N2O7 (C, H, N). Com-
pound 5e. C2H2O4: colorless prisms (from 2-propanol),
mp 138±143 ꢁC; 1H NMR d 1.08 (s, 9), 2.69 (d, 2,
J=5.8), 2.94 (dd, 1, J=3.7 and 16.3), 3.04 (dd, 1, J=8.7
and 16.3), 3.59 (m, 1), 3.82 (s, 3), 4.60 (ddd, 1, J=3.4,
5.8 and 5.8); 13C NMR d 29.9 (CMe3), 35.3 (C-4), 45.2
(CH2N), 51.1 (NHC), 58.2 (OMe), 71.5 (CHOH), 84.1
(C-5), 168.9 (C-3). Anal C12H22N2O7 (C, H, N).
Pharmacology
18.0), 3.67 (m, 1), 4.68 (ddd, 1, J=5.0, 8.1 and 11.3); 13
C
Materials. 125I-Cyanopindolol (125I-CYP) was obtained
from Du Pont NEN, Dreieich, Germany. All other
materials were from sources as described earlier.17
DMSO or water 10 mM stock solutions of the sub-
stances under investigation (4a±e and 5a±e) were pre-
pared and then further diluted with binding bu€er to the
appropriate concentration.
NMR d 27.7 (Me), 30.0 (CMe3), 34.1 (C-4), 44.8
(CH2N), 51.3 (NHC), 71.0 (CHOH), 86.7 (C-5), 159.5
(C-3), 194.0 (C=O). Anal C13H22N2O7 (C, H, N).
Compound 5c.C2H2O4: colorless prisms (from 2-propa-
nol/ethyl ether), mp 178±179 ꢁC dec.; 1H NMR d 1.08 (s,
9), 2.47 (s, 3), 2.71 (m, 2), 3.14 (d, 2, J=9.7), 3.58 (m, 1),
4.74 (ddd, 1, J=3.2, 9.9 and 9.9); 13C NMR d 30.0 (Me
and CMe3), 34.9 (C-4), 45.1 (CH2N), 51.2 (NHC), 71.7
(CHOH), 86.9 (C-5), 159.6 (C-3), 194.1 (C=O). Anal
C13H22N2O7 (C, H, N). Compound 4d.1/2C2H2O4: col-
orless prisms (from abs ethanol), mp 260±265 ꢁC dec.;
1H NMR d 1.09 (s, 9), 2.03 (s, 3), 2.54 (dd, 1, J=7.9 and
12.1), 2.82 (dd, 1, J=3.7 and 12.1), 3.13 (dd, 1, J=10.0
and 16.9), 3.20 (dd, 1, J=8.2 and 16.9), 3.56 (m, 1), 4.50
(ddd, 1, J=6.5, 8.2 and 10.0), 5.24 (bs, 1), 5.33 (bs, 1);
13C NMR d 19.9 (Me), 29.9 (CMe3), 37.5 (C-4), 45.1
(CH2N), 51.6 (NHC), 71.3 (CHOH), 83.9 (C-5), 120.8
(CH2=C), 136.4 (C=CH2), 160.0 (C-3). Anal
C13H23N2O4 (C, H, N). Compound 5d.C2H2O4: color-
less prisms (from 2-propanol/ethyl ether), mp 184±
Cells and membranes. CHO cells which do not express
endogenous b-adrenergic receptor were stably trans-
fected with the human b2-adrenergic receptor as descri-
bed.17 As a model for b1-adrenergic receptors rat
C6 glioma cells were used. In these cells we could not
detect measurable amounts of b2-receptors as con®rmed
by competition experiments with the b1-selective
antagonist ICI 118551 for 125I-CYP binding. For bind-
ing studies cells were grown on Petri dishes ( 140 mm) to
con¯uency. Then the medium was removed and cells
were washed and frozen in the dishes. The Petri dishes
were kept at 25 ꢁC until membranes were prepared.
For the preparation of crude membranes, frozen cells
were thawed and then scraped o€ the Petri dishes in the
hypotonic bu€er (5 mM TRIS/HCl, 2 mM EDTA, pH
7.4). The cell suspension was homogenized with an
Ultra-Turrax (IKA) for 2Â15 s at full speed and the
homogenate was spun for 10 min at 1,000 g. The super-
natant was then centrifuged for 40 min at 50,000 g. The
membrane pellet was resuspended in 50 mM TRIS/HCl
bu€er pH 7.4, frozen in liquid nitrogen at a protein
concentration of 1±3 mg/mL and stored at 80 ꢁC.
185 ꢁC dec.; H NMR d 1.09 (s, 9), 2.03 (s, 3), 2.72 (m,
1
2), 3.41 (d, 1, J=9.6), 3.59 (m, 1), 4.63 (ddd, 1, J=3.5,
9.6 and 9.6), 5.22 (bs, 1), 5.32 (bs, 1); 13C NMR d 19.9
(Me), 30.0 (CMe3), 37.2 (C-4), 45.4 (CH2N), 51.1
(NHC), 71.7 (CHOH), 84.3 (C-5), 120.6 (CH2=C), 136.4
(C=CH2), 160.0 (C-3). Anal C14H24N2O6 (C, H, N).
Synthesis of 4e and 5e. A suspension of 4a (or 5a) (0.5 g,
2.3 mmol) and anhydrous potassium carbonate (1.6 g,
11.6 mmol) in methanol (25 mL) was re¯uxed overnight.
Methanol was removed under vacuum and the residue
was taken up with water (20 mL) and dichloromethane
(25 mL). The organic layer was separated, dried over
Radioligand binding. Dissociation constants (Ki-values)
of Broxaterol 1 and its analogues 4a±e and 5a±e were
determined in radioligand competition experiments. The

408
P. Conti et al./Bioorg. Med. Chem. 6 (1998) 401±408
nonselective antagonist 125I-cyanopindolol was used as
the radioligand at a concentration of 15±25 pM. Binding
was carried out in a total volume of 200 mL with 20±
25 mg of membrane protein in 50 mM TRIS/HCl pH 7.4
in the presence of 50 mM GTP. Samples were incubated
for 1 h at 30 ꢁC, ®ltered with a Scatron cell harvester and
washed three times with 4 mL of ice-cold binding bu€er.
Data were analyzed by nonlinear curve-®tting with the
the program SCTFIT.18
5. Leclerc, G.; Bieth, N.; Schwartz, J. J. Med. Chem. 1980, 23,
620.
6. Bouzoubaa, M.; Leclerc, G.; Decker, N.; Schwartz, J.;
Andermann, G. J. Med. Chem. 1984, 27, 1291.
7. Balsamo, A.; Breschi, M.C.; Chini, M.; Domiano, P.; Gian-
naccini, G.; Lucacchini, A.; Macchia, B.; Macchia, M.; Man-
era, C.; Martinelli, A.; Martini, C.; Martinotti, E.; Nieri, P.;
Rossello, A. Eur. J. Med. Chem. 1992, 27, 751.
8. Balsamo, A.; Breschi, M. C.; Chiellini, G.; Lucacchini, A.;
Macchia, M.; Martinelli, A.; Martini, C.; Nardini, C.; Orlan-
dini, E.; Romagnoli, F.; Rossello, A. Eur. J. Med. Chem. 1994,
29, 855.
Adenylate cyclase activity. Adenylate cyclase activity
was determined as described by Klotz et al.19 with
minor modi®cations. The incubation mixture contained
50 mg of membrane protein and about 100,000 cpm of
[a-32P]ATP whereas EGTA and NaCl were omitted.
The samples were incubated for 20 min at 37 ꢁC.
9. Breschi, M.C.; Macchia, M.; Manera, C.; Micali, E.; Nar-
dini, C.; Nencetti, S.; Rossello, A.; Scatizzi, R. Eur. J. Med.
Chem. 1996, 31, 159.
10. Fisher, L. G.; Sher, P. M.; Skwish, S.; Michel, I. M.; Seiler,
S. M.; Dickinson, K. E. J. Bioorg. Med. Chem. Lett. 1996, 6
2253 and references cited therein.
11. Himber, J.; Velly, J.; Kintz, P.; de Burlet, G.; Andermann,
G.; Leclerc, G. Meth. and Find. Exp. Clin. Pharmacol. 1989, 11,
315.
Acknowledgements
This work was supported by grants from M.U.R.S.T.
(Ministero dell'Universita e della Ricerca Scienti®ca e
Tecnologica), Rome. We are greatly indebted to Pro-
fessor Lucio Toma for helpful discussions. The expert
technical assistance of Ms Sonja Kachler is gratefully
acknowledged.
12. Bouzoubaa, M.; Leclerc, G.; Rakhit, S.; Andremann, G. J.
Med. Chem. 1985, 28, 896.
13. Macchia, B.; Balsamo, A.; Lapucci, A.; Martinelli, A.;
Macchia, F.; Breschi, M. C.; Fantoni, B.; Martinotti, E. J.
Med. Chem. 1985, 28, 153.
14. Vyas, D. M.; Chiang, Y.; Doyle, T. W. Tetrahedron Lett.
1984, 25, 490.
References
15. Larsen, K. E.; Torsell, K. B. G. Tetrahedron 1984, 40, 2985.
1. Chiarino, D.; Fantucci, M.; Carenzi, A.; Della Bella, D.;
Frigeni, V.; Sala, R. Il Farmaco Ed. Sci. 1986, 41, 440.
16. Armand, J.; Souchay, P.; Valentini, F. Bull. Soc. Chem.
France 1968, 4592.
17. Lohse, M. J. Naunyn-Schmiedeberg's Arch. Pharmacol.
1992, 345, 444.
2. Sala, R.; Moriggi, E.; Della Bella, D.; Carenzi, A. Eur. J.
Pharmacol. 1991, 203, 17.
3. Broxaterol. Drugs of the Future 1991, 16, 163; 1992, 17, 136;
1993, 18, 162.
18. De Lean, A.; Hancock, A. A.; Lefkowitz, R. J. Mol. Phar-
macol. 1982, 21, 5.
4. De Amici, M.; De Micheli, C.; Carrea, G.; Spezia, S. J. Org.
Chem. 1989, 54, 2646.
19. Klotz, K-N.; Cristalli, G.; Grifantini, M.; Vittori, S.;
Lohse, M. J. J. Biol. Chem. 1985, 260, 14659.