New series of aryloxypropanolamines with both human β3-adrenoceptor agonistic activity and free radical scavenging properties

Article abstract of DOI:10.1016/S0960-894X(01)00716-8

A series of 13 novel hybrid molecules designed to possess both free radical scavenging activity and to stimulate the β₃-adrenoceptors in order to improve antidiabetic effect and to restore insulin sensitivity was synthesized and evaluated. Compounds were of quinolyl-, isoquinolyl-, pyridoindolyl- or carbazolyloxypropanolamine structure with a terminal amino group of benzopyranolyl-, di-tert-butylphenolyl- or methoxyindolyl-type. Some of the products possessed both the expected activities.

Full text of DOI:10.1016/S0960-894X(01)00716-8

Bioorganic & Medicinal Chemistry Letters 12 (2002) 209–212
New Series of Aryloxypropanolamines with Both Human
ꢀ3-Adrenoceptor Agonistic Activity and Free Radical
Scavenging Properties
Silvere Aubriot,a Edwige Nicolle,a,* Mireille Lattier,b Cecile Morel,a Wenhong Cao,c
Kiefer W. Daniel,c Sheila Collins,c Gerard Leclerca and Patrice Faureb
a
´ ´ ´
Departement de Pharmacochimie Moleculaire, UMR-CNRS 5063, UFR de Pharmacie, Universite Joseph Fourier, BP 138,
F-38243 Meylan cedex, France
b
´
´
Laboratoire du Stress Cardiovasculaire et des Pathologies Associees, EA 2937 UFR de Pharmacie, Universite Joseph Fourier,
F-38700 La Tronche, France
cDivision of Biological Psychiatry, Duke University Medical Center, Box 3557 Durham, NC 27710, USA
Received 16 July 2001; revised 22 October 2001; accepted 23 October 2001
Abstract—A series of 13 novel hybrid molecules designed to possess both free radical scavenging activity and to stimulate the b3-
adrenoceptors in order to improve antidiabetic effect and to restore insulin sensitivity was synthesized and evaluated. Compounds
were of quinolyl-, isoquinolyl-, pyridoindolyl- or carbazolyloxypropanolamine structure with a terminal amino group of benzo-
pyranolyl-, di-tert-butylphenolyl- or methoxyindolyl-type. Some of the products possessed both the expected activities. # 2002
Elsevier Science Ltd. All rights reserved.
Non insulin-dependent diabetes mellitus (NIDDM) is
an expanding disease throughout industrialized coun-
tries. It is characterized by a resistance to the action of
insulin on peripheral or hepatic tissues, leading to high
levels of circulating glucose. In nearly 50% of the time,
overweight and obesity are present in the same times. In
diabetic patients, free radical overproduction is
observed participating to vascular complications.
Moreover, it has been shown that food supplementation
with antioxidants improves insulin sensitivity in diabetic
patients.1
important differences in the in vivo results between
humans and smaller mammals. It has been also demon-
strated that b3-adrenergic stimulation leads to re-
appearence of brown adipocytes suggesting a possible
antidiabetic effect in humans.7 Clinical investigations
were led on many b3-adrenergic agonists, but it is generally
admitted these products lacked either of selectivity,
potency or biodisponibility.
In order to improve activity of potential b3-adreno-
ceptor agonists, we hypothesized that an original hybrid
molecule possessing both b3-adrenergic and antioxidant
structures might lead to compounds useful for the
treatment of insulin resistance syndrome and its vas-
cular complications. On the basis of this hypothesis, we
designed and synthesized structures corresponding to
the general formula given in Figure 1, bearing various
aromatic rings and different terminal amines derived
The discovery in 1989 of the b3-adrenoceptor present in
brown adipose (BAT) and white adipose tissues (WAT)
where it stimulates thermogenesis and lipolysis, respec-
tively, has open a new way for antidiabetes drug
design.2À4 There is now evidence showing that b3-adre-
nergic stimulation can increase insulin sensitivity and
glucose tolerance, leading to selective loss of adipose
tissue mass in animal models.5,6 In fact, BAT is present
in hibernating animals and in new-borns but disappears
by adulthood; this explains in part why there are
*Corresponding author. Tel./fax: +33-4-76-04-10-22; e-mail: edwige.
nicolle@ujf-grenoble.fr
Figure 1. General structures of aryloxypropanolamines.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00716-8

210
S. Aubriot et al. / Bioorg. Med. Chem. Lett. 12 (2002) 209–212
from known radical scavenging structures such as Tro-
lox (benzopyranol derivative) or di-tert-butyl-4-hydroxy-
toluene (BHT) or 5-methoxytryptamine.
three steps synthesis via phenylhydrazone, Fisher cycliza-
tion and oxidation with carbon on activated palladium.14
Amine 3 derived from BHT, was prepared in four steps
by reduction of 3,5-di-tert-butyl-4-hydroxyphenyl-
propionic acid with LAH, followed by bromination
with PBr3 (Scheme 2B). Alkylation of p-nitrophenol by
the brominated derivative 2 followed by reduction by
hydrazine and Raney nickel, afforded to amine 3. The
precursor of compound 4, an intermediate of the Tro-
glitazone synthesis was obtained according to Yoshioka
et al.15 5-Methoxytryptamine (5) is commercially avail-
able and was used such as. In all the cases, racemic
forms were then biologically evaluated.
The anti-oxidant activity was evaluated compared to
Trolox and also to carvedilol (4-hydroxycarbazole deri-
vative), a b2-adrenergic blocker with potent anti-oxidant
properties.8 b3-Adrenergic activity was compared to L-
755,507, the most potent and selective b3-adrenoceptor
agonist for human b3-receptors actually known
(EC50=0.43 nM).9 In terms of structural features, the
importance of an oxypropanolamines moiety for b3-
adrenergic agonist activity is well documented. In addi-
tion, it should also be noted that in most b3-adrenergic
agonists, an intra- or extracyclic heteroatom in the meta
position of the oxypropanolamine chain is present; this
is true for BRL 37344, CL 316243, carazolol and cya-
nopindolol. Some results obtained in our laboratory
have underscored the importance of one of these two
meta positions for b3-adrenoceptor affinity (unpublished
results). So, we tried to use these observations to inves-
tigate the quinoline, isoquinoline, carbazole and carbo-
line rings (Fig. 1).
Free radical scavenging effects
In diabetes, protein oxidation and lipid peroxidation
lead to vascular degenerative lesions. For these reasons,
the representative derivatives were evaluated for anti-
oxidant activity by determining their ability in vitro to
protect against protein oxidation and plasma lipid per-
oxidation induced by the highly reactive hydroxyl radi-
cals generated by a Fenton reaction (Fe2+/H2O2). The
extent of oxidation was assessed indirectly by measuring
Another important feature appearing now, especially in
L-755,507 is the importance of the length and the
bulkiness of the substituent linked to the nitrogen of the
oxypropanolamine for selectivity toward the b3-adre-
noreceptor subtype. For the present study, we have
chosen three different well known antioxidants: a BHT
derivative, a benzopyranol derivative whose structure is
encountred in tocopherols and, finally, a melatonin deri-
vative (5-methoxytryptamine) which is a highly studied
derivatives for its free radicals scavenging properties.
Table 1. Aryloxypropanolamines 6–18: chemical characteristics
Compd
Ar
R
6
7
8
9
10
11
12
13
14
15
16
17
18
5-Quinoline
5-Quinoline
5-Quinoline
7-Quinoline
7-Quinoline
7-Quinoline
5-Isoquinoline
5-Isoquinoline
7-Isoquinoline
7-Isoquinoline
4-Carbazole
4-Carbazole
5-Carboline
BHT
Trolox
Tryptamine
BHT
Tryptamine
Trolox
BHT
Trolox
BHT
Trolox
Trolox
Tryptamine
Trolox
The general synthetic route used to prepare the target
compounds 6–18 is outlined in Scheme 1.
All the compounds were prepared by alkylation of
hydroxylated rings with epibromohydrin to obtain the
target epoxides which were opened by the different
amines 3–5 in ethanol, either at room temperature or at
reflux.10 The resulting aryloxypropanolamines are
described in Table 1.
The hydroxylated rings were either commercially avail-
able [5-hydroxyisoquinoline, 7-hydroxyquinoline] or
obtained either by known methods or by new synthesis.
5-Hydroxyquinoline was prepared according to
Bucherer synthesis on 5-aminoquinoline.11 7-Hydroxy-
isoquinoline (1) was prepared according to Pomeranz–
Fritsch synthesis.12 5-Hydroxy-a-carboline synthesis
was assessed by Suzuki coupling between 2-fluoro-3-
iodopyridine and N-Boc-m-anisidine boronic acid,
according to the method described by Rocca et al.
(Scheme 2A).13 4-Hydroxycarbazole was obtained in a
Scheme 2. Synthesis of 5-hydroxycarboline (A) and terminal amines
(B).
Scheme 1. General scheme of synthesis.

S. Aubriot et al. / Bioorg. Med. Chem. Lett. 12 (2002) 209–212
Table 2. Aryloxypropanolamines 6–18: b-adrenergic and anti-oxidant activities
211
Compd
cAMP (pmol/min/well)
Ellman’s test
% protectionc
TBARs test
% protectionc
Transfected
cellsa
Transfected Untransfected
cellsb
cells b
500 mM
1 mM
10 mM
100 mM
500 mM
Basal
6
7
8
9
10
11
12
13
14
15
16
17
18
20.6 (Æ0.1)
42.2 (Æ3.5)
28.8 (Æ0.9)
21.6 (Æ1.6) 47.3 (Æ17.9)
26.8 (Æ0.7) 49.2 (Æ16.5)
29.9 (Æ1.1)
120.3 (Æ36.6)*,** 30.3 (Æ0.6)**
33.3 (Æ0.9)*,** 50.4 (Æ2.8)*
102.6 (Æ13.8) 28.6 (Æ1.9)
24.1 (Æ0.2)
82.9 (Æ25.7)*
37.8 (Æ1.5)*,** 64.6 (Æ1.8)*,** 78.2 (Æ0.2)*,**
43.6 (Æ4.7)*,** 52.0 (Æ4.2)*,** 50.7 (Æ2.8)*
97.4 (Æ11.4) 21.4 (Æ0.6)
1.8 (Æ0.1)
23.3 (Æ0.5)
32.5 (Æ1.5)
46.0 (Æ7.7)*
66.3 (Æ3.4)*
26.5 (Æ1.5)**
30.6 (Æ0.7)** 50.8 (Æ2.6)*
129.6 (Æ2.8)
76.3 (Æ6.1)
24.9 (Æ0.6)
43.6 (Æ3.3)*,** 32.8 (Æ4.6)** 49.3 (Æ1.9)*
142.1 (Æ28.1) 41.6 (Æ3.7)
23.9 (Æ1.4)
28.7 (Æ0.6) 54.8 (Æ4.9)*,** 115.2 (Æ36.6)*,** 39.4 (Æ3.9)*,** 49.3 (Æ2.9)*,** 94.4 (Æ9.2)*,**
26.9 (Æ1.6)
33.2 (Æ0.7)
33.0 (Æ2.2)
22.3 (Æ0.6)
31.5 (Æ0.7)
34.4 (Æ3.7)
27.9 (Æ1.6)
44.3 (Æ3.5)*,** 53.6 (Æ3.3)*,** 75.5 (Æ1.8)*,**
Trolox
Carvedilol
31.4 (Æ5.1)
20.0 (Æ9.5)
75.2 (Æ3.3)*
À21.4 (Æ3.7)
1.0 (Æ0.8)
7.6 (Æ2.8)
64.8 (Æ2.8)*
6.2 (Æ2.5)
À88.6 (Æ17.5)
À0.6 (Æ0.6)
L-755,507 166.6 (Æ1.8) 173.6 (Æ17.5) 30.4 (Æ1.1)
Ãp< 0.05 versus BSA or plasma oxidation without tested molecule. ÃÃp< 0.05 versus Trolox.
aWithout propranolol.
bWith propranolol.
cValues are means Æ(SEM) of three different experiments.
the protection of bovine serum albumin (BSA) thiol
protection by using Ellman’s test16 and formation of
thiobarbituric acide reactive substances (TBARs).17
Results are expressed as percentage of protection in
comparison to the benzopyranol structure Trolox and are
presented in Table 2. Statistical method of calculation
was ANOVA test (p<0.05).
Besides the Ellman’s test, it is interesting to observe
that the thiol group protection increased with the
concentration of the tested molecules.
Together, these results showed that bulky polyphenol
derivatives are far more potent to scavenge OH than
indole or carbazole derivatives and that quinonic struc-
tures (BHT and Trolox) are the most potent amongst
them. In addition, it should be noted that, at the tested
concentrations, the novel molecules were generally bet-
ter free radical scavengers against lipid peroxidation
than Trolox which is though the parent molecule for
compounds 7, 13 and 18.
As far as the TBARs test is concerned, the closely rela-
ted molecules derived from benzopyranol, compounds
7, 13, 18 had similar antioxidant activity, between
themselves, and are better (p<0.05) from Trolox at 10
and 100 mM but not 500 mM except for compounds 7,
13 and 18. Moreover, no significant difference was
observed in the antioxidant activity of compounds
bearing different aromatic moities i.e. quinoline, iso-
quinoline or carboline but the same terminal amine
derived from benzopyranol structure (p<0.05). Thus,
the introduction of different aromatic moities seems to
bring no significant changes in the free radical scaveng-
ing. So, only one aromatic structure of each lateral
amine was subsequently tested. On the other side, the
di-tert-butylphenol derivative 12, did not develop the
same scavenging properties compared to benzopyranol
derivatives, which developed dose–response curve
activity. Concerning the tryptamine derivative 8, no
change in anti-oxidant protection was observed between
the three concentration levels. Surprisingly, and in con-
tradiction with literature, carvedilol showed in these
models, weak scavenging properties. Yet, data of litera-
ꢀ-Adrenergic activity
b-Adrenergic activities were calculated by measuring of
cAMP production on human b3-adrenergic transfected
and untransfected COS-7 cells.18 These activities were
compared to L-755,507, one of the most potent and
selective b3 agonists.9 The results are shown in Table 2,
given in pmol of cAMP produced per min and per well
for each compound (10 mM) in the absence or presence
of 0.1 mM propranolol, a b1/b2 antagonist. The levels of
cAMP produced in transfected cells without proprano-
lol is a reflection of the total b-adrenergic agonistic
activity (b1, b2 and b3) while cAMP levels in transfected
cells in the presence of propranolol reflects the selective
b3-adrenergic activity. When untransfected cells were
treated with all the compounds in the presence of pro-
pranolol, all activity levels were equivalent. Together,
these data show that compounds 7, 9, 12 and 13 possess
both b2 and b3 adrenergic agonistic activity. Com-
pounds 7, 8, 9, 14, 15 and 16 were inactive to stimulate
cAMP production in b3-transfected COS-7 cells.19 The
ꢀ
ture referred to a protection against OH exclusively,
generated by a Fenton reaction. The oxidative system
used in the previous works produced less reactive free
ꢀ
radicals (such as O2 À) and it is probable that carvedilol
ꢀ
ꢀ
is a better scavenger for O2 À than OH.

212
S. Aubriot et al. / Bioorg. Med. Chem. Lett. 12 (2002) 209–212
results indicated that compounds 6, 12, 13 and 17
exhibited selective stimulation of cAMP production in
human b3-transfected cells. The compound with the best
b3-adrenergic activity was 12, which was derived from 5-
hydroxyisoquinoline and BHT, giving cAMP produc-
tion in transfected cells of 40% of L-755,507 activity. It
should be noted that one compound (10), derived from
7-hydroxyquinoline and tryptamine, appeared to pos-
sess reverse agonistic activity as demonstrated by a
decrease in cAMP production in cells compared to the
cAMP basal production of untransfected preincubated
(1.8 vs 32.5 pmol/min/well). Evidently, this finding will
require further investigations.
References and Notes
1. Faure, P.; Rossini, E.; Lafond, J. L.; Richard, M. J.;
Favier, A.; Halimi, S. J. Nutr. 1997, 127, 103.
2. Emorine, L. J.; Marullo, S.; Briend-Sutren, M. M.; Patey,
G.; Tate, K.; Delavier-Klutchko, C.; Strosberg, A. D. Science
1989, 245, 1118.
3. Muzzin, P.; Revelli, J. P.; Kuhnet, F. J. Biol. Chem. 1991,
266, 24053.
4. Granneman, J. G.; Lahners, K. N.; Chaudhry, A. Mol.
Pharmacol. 1991, 40, 895.
5. Atef, N.; Lafontan, M.; Double, A.; Helary, C.; Ktorza, A.;
Penicaud, L. Eur. J. Pharmacol. 1996, 298, 287.
6. Green, A.; Caroll, R. M.; Dobias, S. B. Am. J. Physiol.
Endocrinol. Metab. 1996, 34, 271.
7. Ghorbani, M.; Himms-Hagen, J. Int. J. Obes. 1997, 21,
465.
8. Yue, T. L.; Cheng, H. Y.; Lysko, P. G.; McKenna, P. J.;
Feuerstein, R.; Gu, J. L.; Lysko, K. A.; Davis, L. L.; Feuer-
stein, G. J. Pharmacol. Exp. Ther. 1992, 1, 92.
9. Fisher, M. H.; Amend, A. M.; Bach, T. J.; Barker, J. M.;
Brady, E. J.; Candelore, M. R. J. Clin. Invest. 1998, 101, 2387.
10. Tejani-Butt, S. M.; Brunswick, D. J. J. Med. Chem. 1986,
29, 1524.
11. JP Patent 4-95075, 1990.
12. Woodward, R. B.; Doering, W. E. J. Am. Chem. Soc.
1945, 67, 860.
13. Rocca, P.; Marsais, F.; Godard, A.; Queguinier, G. J.
Heterocyclic Chem. 1995, 32, 1171.
14. Rodriguez, J. G.; Temprano, F.; Esteban-Calderon, C.;
Martinez-Ripoll, M. J. Chem. Soc., Perkin Trans. 1 1989,
2117.
15. Yoshioka, T.; Fujita, T.; Kanai, T.; Aizawa, Y.; Kur-
umada, T.; Hasegawa, K.; Horikoshi, H. J. Med. Chem. 1989,
32, 421.
16. Suzuki, Y.; Lyall, V.; Biber, T. U. L.; Ford, G. D. Free
Radical Bio. Med. 1990, 9, 479.
17. Yagi, K. Biochem. Med. Metab. B 1976, 15, 212.
18. cAMP measurement: sCOS-7 cells were cultured in
DMEM with 10% foetal bovine serum (Hyclone, UT, USA),
100 units/mL penicillin, and 100 mg/mL streptomycin at 37 ꢁC
in a humidified 5% CO2 atmosphere. Cells were seeded into 6-
well cluster plates (Falcon) and transfected at 70–80% con-
fluence with 2 mg pBC12-hb3adrenergic receptor plasmid
DNA per well with 5 mL lipofectamine (Life Technologies,
Gaithersburg, USA) in 1 mL of serum-free media. Four hours
after addition of DNA, the cells were gently washed with
serum-free media and incubated in complete growth media for
an additional 24 h. Adenylyl cyclase assay cells were treated in
6-well cluster plates according method of Cotecchia et al.19
Attached cells were incubated for 30 min at 37 ꢁC in serum-
free DMEM/25 mM HEPES, pH 7.5. The medium was
replaced with the same medium containing 0.24 mM IBMX
and the drugs for 20 min. Treatment was terminated by
rapidly aspirating the medium and adding cold 5% TCA (500
mL/well). Adenylyl cyclase was assessed in 20 mL of the TCA
extract by measuring the cAMP formed by RIA20 using a
polyclonal antiserum iodinated to cAMP.21
Concerning the structure–activity relationship, and in
opposition to that suggested in literature, the presence
of an heteroatom in the meta position of the aryl-
oxypropanolamine lateral chain does not seem to have
significance for the activity since the 5-hydroxy-
isoquinoline derivatives (i.e., 12 and 13) developed a
better activity than their quinoline analogues. On the
other hand, in the same manner we previously noted for
carbazole and indole derivatives (unpublished data), the
5-position for the aryloxypropanolamine chain of iso-
quinoline and quinoline seemed to have some sig-
nificance for the activity since no 7-substituted
compound developed any agonistic activity (9, 14, 15).
Moreover, the presence of a bulky amine substituent
appears to be favorable for the b3-adrenergic activity, in
this way, it came to light that BHT derivatives were the
best structures compared to the benzopyranol deriva-
tives which were more potent than their 5-methoxy-
tryptamine analogues. In addition, flexibility of this
chain seems to play an important role since BHT deri-
vatives were better than their Trolox analogues (6>7
and 12>13).
In conclusion, this study brought to light a new series of
aryloxypropanolamines appearing to be unique in that
they possess both antioxidant and b3-adrenergic ago-
nistic activity. The most potent b3 agonist (12) was a
good radical scavenger so, taken on the whole, these
molecules could open the field of a new therapeutic
strategy in diabetic patients, as the improvement of
insulin activity has an indirect beneficial effect on
antioxidant defence system. Further synthesis will be
directed towards improving the activation of cAMP
production in transfected cells, and selectivity toward
b3-adrenoceptors by lengthening the lateral chain in
order to improve its bulkiness.
Acknowledgements
19. Cotecchia, S.; Kobilka, B. K.; Daniel, K. W.; Nolan,
R. D.; Lapetina, E. Y.; Caron, M. G.; Lefkowitz, R. J.;
Regan, J. W. J. Biol. Chem. 1990, 265, 63.
20. Harper, J. F.; Broker, G. J. Cyclic Nucleotide Res. 1975, 1,
207.
We would like to thank Roche Laboratories and Merck
Research Laboratories for providing us with Carvedilol
and L-755,507, respectively. We thank the US NIDDK
National Hormone and Pituitary Program for the anti-
sera to cAMP. We are grateful to Dr. A. Boumendjel
and Pr. C. Ribuot for their profitable advices.
21. Gettys, T. W.; Ramkumar, V.; Uhing, R. J.; Seger, L.;
Taylor, I. L. J. Biol. Chem. 1991, 26.