Discovery of 7-hydroxyaporphines as conformationally restricted ligands for beta-1 and beta-2 adrenergic receptors

Article abstract of DOI:10.1039/c7md00656j

A series of (-)-nornuciferidine derivatives was synthesized and the non-natural enantiomer of the aporphine alkaloid was discovered to be a potent β₁- and β₂-adrenergic receptor ligand that antagonized isoproterenol and procaterol induced cyclic AMP increases from adenylyl cyclase, respectively. Progressive deconstruction of the tetracyclic scaffold to less complex cyclic and acyclic analogues revealed that the conformationally restricted (6a-R,7-R)-7-hydroxyaporphine 2 (AK-2-202) was necessary for efficient receptor binding and antagonism.

Full text of DOI:10.1039/c7md00656j

MedChemComm
View Article Online
View Journal
RESEARCH ARTICLE
Discovery of 7-hydroxyaporphines as
conformationally restricted ligands for beta-1 and
Cite this: DOI: 10.1039/c7md00656j
beta-2 adrenergic receptors†
*
Angela F. Ku
and Gregory D. Cuny
A series of (−)-nornuciferidine derivatives was synthesized and the non-natural enantiomer of the
aporphine alkaloid was discovered to be a potent β1- and β2-adrenergic receptor ligand that antagonized
isoproterenol and procaterol induced cyclic AMP increases from adenylyl cyclase, respectively. Progressive
deconstruction of the tetracyclic scaffold to less complex cyclic and acyclic analogues revealed that the
conformationally restricted (6a-R,7-R)-7-hydroxyaporphine 2 (AK-2-202) was necessary for efficient recep-
tor binding and antagonism.
Received 31st December 2017,
Accepted 5th January 2018
DOI: 10.1039/c7md00656j
rsc.li/medchemcomm
G protein-coupled receptors (GPCRs), the largest protein
superfamily in the human genome,1 participate in a wide
range of physiologic and pathophysiologic processes and have
emerged as targets for chemotherapeutic development.2 The
β2-adrenergic receptor (β2-AR),2a,d which is involved in pain
signaling3 and the regulation of the respiratory,4 cardiovascu-
lar5 and neurological6 systems, is one of these important
GPCRs. Many reported ligands for the β2-AR have a common
pharmacophore consisting of an aromatic ring connected to a
ethanolamine moiety (Fig. 1),7–10 including the antagonist
(R)-pronethalol,8 the agonist (R)-salbutamol9 and the endoge-
nous agonist epinephrine.7d,10
eral related compounds, employing acid catalysed reductive
cyclization followed by Pd-mediated ortho-arylation as two
key steps.13 As shown in Table 1, (−)-nornuciferidine has an
absolute configuration of (S,S) at (C-a, C-b). However, several
β2-AR ligands (e.g., salbutamol) with a R-configuration at the
corresponding C-b position were reported to be more active
and/or selective than their racemates or stereoisomers.8,9
Therefore, nornuciferidine diastereomer 1, enantiomer 2, 6a-
alkyl substituted derivative 3, two tricyclic derivatives 4 and
5, and the acylic derivative 6 were selected for conducting the
structure–activity relationship (SAR) analysis for β2-AR bind-
ing functional activity.
7-Hydroxyaporphine alkaloids also possess the β2-AR
pharmacophore raising the possibility that this class of natu-
ral products might possess β2-AR binding affinity. However,
conformational restrictions imposed by the tetracyclic system
could impede optimal spatial orientation of the functional
groups necessary for efficient binding.11 Herein we report the
exploration of 7-hydroxyaporphine derivatives as β2-AR li-
gands. Furthermore, progressive dismantling of the tetra-
cyclic scaffold to less conformationally restricted derivatives
was conducted to evaluate the impact on receptor binding
affinity.
The preparation of test compounds 1–6 is summarized in
Schemes 1 and 2. The synthesis was initiated with the conver-
sion of (R)-(−)-2-chloromendelic acid (7) into ester 9. Then,
coupling of alcohol 9 and isocyanate 10, derived from amine
8, furnished carbamate 11. Subsequent one-pot reductive cy-
clization with sequential addition of DIBAL-H, methanol and
BF3·OEt2, followed by silyl deprotection with TBAF, gave oxa-
zolidinone 12 as a mixture favouring the anti-isomer (dr
67 : 33). The diastereomers could be differentiated based on
their 1H NMR coupling constants. The assignment of absolute
(−)-Nornuciferidine is a 7-oxygenated aporphine alkaloid
isolated from the bark of Bolivian Duguetia spixiana
(Annonaceae) by Rasamizafy and coworkers12 and was se-
lected as the lead molecule for this study. We recently
reported the synthesis of this natural product, as well as sev-
Department of Pharmacological and Pharmaceutical Sciences, University of
Houston, Science and Research Building 2, Houston, Texas 77204, USA.
E-mail: gdcuny@central.uh.edu
Fig.
1 Examples of β2-AR ligands and the 7-hydroxyaporphine
† Electronic supplementary information (ESI) available: Experimental procedures
and characterization data for new compounds. See DOI: 10.1039/c7md00656j
scaffold; β2-AR pharmacophore elements highlighted in red for
ethanolamine and blue for the aromatic ring.
This journal is © The Royal Society of Chemistry 2018
Med. Chem. Commun.

View Article Online
Research Article
MedChemComm
Table 1 Human β2-AR binding inhibition
intermediate previously reported by our group, followed by
hydrolysis of the carbamate.
Compounds 4 and 5 (Table 1, type B) possess the same ab-
solute configuration as compound 2, but the linkage between
the biaryl ring systems was disconnected, imparting increased
freedom of rotation. These two compounds differ from one
another in the ortho-substituent (R2). Compound 4 (R2 = Cl)
was synthesized by methylation and hydrolysis of intermedi-
ate 12-anti. Compound 5 (R2 = H) was derived from two com-
mercially available starting materials, mandelic ester 16 and
amine 17. After formation of carbamate 19, one-pot reductive
cyclization provided oxazolidinone 20 exclusively as the
anti-isomer. Subsequent hydrolysis furnished compound 5.
Compound 6 (Table 1, type C) is an acyclic R-enantiomer
with a fully rotatable ethanolamine scaffold. As shown in
Scheme 2, hydrolysis of mandelic ester 16, followed by EDC
coupling with amine 17, generated amide 21 in 69% yield.
Then, reduction of 21 with borane, produced in situ by addi-
tion of NaBH4 and BF3·OEt2, generated compound 6 in 82%
yield.
The specific binding affinities of (−)-nornuciferidine and
compounds 1–6 for human β2-AR were evaluated in a
radioligand binding assay using [3H]-CGP-12177.15 The half
maximal inhibitory concentration (IC50) values are summa-
rized in Table 1. The natural product (−)-nornuciferidine was
found to be a moderately potent ligand for β2-AR (IC50 = 1.6
0.5 μM). When the chirality at C-b was switched from S to R
(e.g., compound 1), the binding affinity increased by 7.6-fold
Compound
Type R1
R2 (C-a, C-b) IC50 (μM)
(−)-Nornuciferidine
A
A
A
A
B
B
C
H
H
H
Me
—
—
—
—
—
—
—
Cl
H
(S,S)
(S,R)
1.6 0.5
1
2
3
4
5
6
0.21 0.035
0.0014 0.0005
0.020 0.0025
>10
(R,R)
rel-(R,R)
(R,R)
(R,R)
(—,R)
>10
—
8.4 0.5
configurations was previously discussed.13 Oxazolidinone 12
was then carried forward directly to the ortho-phenol arylation
step. After heating at 110 °C for 40 minutes in the presence of
10 mol% XPhos precatalyst and Cs2CO3, arylated products 13-
anti and 13-syn were obtained in 70% and 24% yields, respec-
tively. Subsequent methylation and removal of the carbamate
converted 13-anti and 13-syn into the desired products 2 and
1, respectively.
Compound 3 (Table 1, type A) has the same structure as
( )-nornuciferidine except for a methyl group at C-a. This
compound was generated by methylation of phenol 14,14 an
Scheme 1 Summarized synthetic routes for compounds 1–5.
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2018

View Article Online
MedChemComm
Research Article
and σ2), δ-opioid (DOR), κ-opioid (KOR), μ-opioid (MOR),
peripheral-type benzodiazepine (PBR) receptors, as well as
dopamine (DAT), norepinephrine (NET), serotonin (SERT)
transporters and benzodiazepine (BZP) rat brain site, which
is an allosteric benzodiazepine binding site of the GABAA
receptor.
The compound was initially accessed at 10 μM and for
those that demonstrated >50% inhibition determinations of
inhibition constant (Ki) values were conducted.20 The profil-
ing results are shown in Table 3. Overall, compound 2 was
found to be quite selective, demonstrating moderate to weak
activities (Ki values 0.12–1.2 μM) against seven of the molecu-
lar targets and potent activity (Ki < 0.10 μM) against only one
other GPCR (e.g., 5-HT1A) besides β1- and β2-ARs.
Scheme 2 Synthetic route for compound 6.
(IC50 = 0.21 0.035 μM). Surprisingly, when the absolute con-
figurations at C-a and C-b were switched from (S,S) in (−)-
nornuciferidine to (R,R) in compound 2, the binding affinity
dramatically increased by almost a thousand fold (IC50
0.0014 0.0005 μM). However, addition of a substituent at
C-a as in compound 3 reduced potency by 14.3-fold (IC50
=
=
In conclusion, modification of the absolute stereo-
chemistry in (−)-nornuciferidine resulted in the discovery of 2
with the configuration of (R,R) at (C-a, C-b) as a potent ligand
for human β1- and β2-ARs. The compound was selective (>10-
fold) verses β3-AR and 40 other molecular targets and blocked
isoproterenol and procaterol induced cyclic AMP increases
from adenylyl cyclase in human CHO cells at β1- and β2-AR,
respectively. Due to the conformationally restricted structure,
compound 2 (AK-2-202) may help reveal the preferred bind-
ing conformations for β1- and β2-AR ligands. Additional struc-
ture–activity relationship studies using related aporphine de-
rivatives could further refine the pharmacophore for β1- and
β2-AR antagonism, as well as for generating biased agonists.21
Finally, compound 2 defines very specific chemical space that
may be used for in silico screening or design to achieve struc-
turally distinct compounds that elicit similar pharmacology.
This work was financially supported by the Welch Founda-
tion (E-1916-20170325) and the University of Houston. The Ki
determinations for profile screening were generously pro-
vided by the National Institute of Mental Health's Psychoac-
tive Drug Screening Program, Contract # HHSN-271-2013-
00017-C (NIMH PDSP), directed by Bryan L. Roth MD, PhD at
0.020 0.0025 μM) compared to compound 2. Interestingly,
releasing part of the conformational restriction of the
aporphine ring system by disconnecting the biaryl linkage
resulted in complete loss of activity with or without an ortho-
substituent (IC50 > 10 μM for 4 and 5) likely due to rotation
of the C-a–C-b bond changing the relative orientation of the
amino alcohol. Further dismantling of the tetra-
hydroisoquinoline (e.g., the bond between C-b and the phenyl
residue) restored weak binding activity (IC50 = 8.4 0.5 μM
for 6).
Based on the β2-AR binding assay results, compound 2
was selected for several additional studies. First, the com-
pound was evaluated for selectivity among human β-AR sub-
types.16 As displayed in Table 2, compound 2 demonstrated
significant binding activity at the β1-AR (IC50 = 0.0017
0.0003 μM). However, it was about 150-fold less potent for β3-
AR (IC50 = 0.24 0.064 μM). Next, the function of 2 was exam-
ined in cellular assays.17,18 The compound failed to agonize
either β1- or β2-AR at 10 μM, whereas, it blocked isoprotere-
nol17 and procaterol18 induced cyclic AMP increases from
adenylyl cyclase in human CHO cells (Eurofins Panlabs Dis-
covery Services Taiwan Ltd.) at β1 (IC50 = 0.017 0.014 μM)
and β2-AR (IC50 = 0.0069 0.0027 μM), respectively.
In order to examine the broader selectivity of 2, it was
screened against a panel of 45 GPCRs and monoamine trans-
Table
transporters
3
Profiling results of
2
against 45 GPCRs and monoamine
porters,19 including serotonin (e.g., 5-HT1A, 5-HT1B, 5-HT1D
,
Target
Kia (μM)
Target
Kia (μM)
Target
Kia (μM)
5-HT1E, 5-HT2A, 5-HT2B, 5-HT2C, 5-HT3, 5-HT5A, 5-HT6 and
5-HT7), adrenergic (e.g., α1A, α1B, α1D, α2A, α2B, α2C, β1, β2
and β3), dopamine (e.g., D1, D2, D3, D4 and D5), histamine
(e.g., H1, H2, H3 and H4), muscarinic (e.g., M1, M2, M3, M4
and M5), γ-aminobutyric acid type A (GABAA), sigma (e.g., σ1
5-HT1A
5-HT1B
5-HT1D
5-HT1E
5-HT2A
5-HT2B
5-HT2C
5-HT3
5-HT5A
5-HT6
5-HT7
α1A
0.071
1.209
0.857
—
—
0.246
—
—
—
0.862
0.120
—
α2B
α2C
β1
β2
β3
D1
D2
D3
D4
D5
H1
H2
H3
H4
M1
0.571
—
0.019
0.013
0.308
—
—
—
—
—
M2
M3
M4
M5
GABAA
σ1
σ2
DOR
KOR
MOR
PBR
DAT
NET
SERT
BZP
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
Table 2 Human β-AR subtype binding inhibition and adenylyl cyclase
antagonism in CHO cells for compound 2
—
—
—
—
β-AR
Binding inhibition
Cellular adenylyl cyclase
subtype
(IC50, μM)
antagonism (IC50, μM)
α1B
α1D
α2A
—
—
1.147
β1
β2
β3
0.0017 0.0003
0.0014 0.0005
0.24 0.064
0.017 0.014
0.0069 0.0027
NDa
—
a
Means test compound shows <50% inhibition at 10 μM in the
a
ND = not determined.
primary radioligand competition binding assay.
This journal is © The Royal Society of Chemistry 2018
Med. Chem. Commun.

View Article Online
Research Article
MedChemComm
the University of North Carolina at Chapel Hill and Project
Officer Jamie Driscoll at NIMH, Bethesda MD, USA.
9 (a) D. W. Boulton and J. P. Fawcett, Br. J. Clin. Pharmacol.,
1996, 41, 35; (b) P. W. Ind, Thorax, 1997, 52, 839; (c) D. W.
Cockcroft and V. A. Swystun, Thorax, 1997, 52, 845; (d) B. J.
Lipworth, et al., Thorax, 1997, 52, 849.
Conflicts of interest
10 P. Pratesi, et al., Farmaco, Ed. Sci., 1960, 15, 3.
11 For recent examples of conformationally restricted ligands,
see: (a) C. Jellimann, et al., J. Med. Chem., 2000, 43, 4051; (b)
J. Våbenø, et al., Bioorg. Med. Chem., 2005, 13, 1977; (c) T. H.
McLean, et al., J. Med. Chem., 2006, 49, 5794; (d) C. Sorbi,
et al., ChemMedChem, 2009, 4, 393; (e) R. V. S. Nirogi, et al.,
J. Enzyme Inhib. Med. Chem., 2012, 27, 443; ( f ) C. R. Moyes,
et al., J. Med. Chem., 2014, 57, 1437; (g) H. Torres-Gomez,
et al., Eur. J. Med. Chem., 2013, 70, 78; (h) M. McMullan,
et al., Eur. J. Med. Chem., 2016, 123, 48.
12 S. Rasamizafy, et al., J. Nat. Prod., 1987, 50, 674.
13 A. F. Ku and G. D. Cuny, Org. Lett., 2015, 17, 1134.
14 A. F. Ku and G. D. Cuny, J. Org. Chem., 2016, 81, 10062.
15 K. E. McCrea and S. J. Hill, Br. J. Pharmacol., 1993, 110,
619.
The authors declare no competing interest.
Notes and references
1 (a) E. S. Lander, et al., Nature, 2001, 409, 860; (b) J. C.
Venter, et al., Science, 2001, 291, 1304.
2 (a) P. B. Molinoff, Drugs, 1984, 28, 1; (b) B. K. Kobilka,
Biochim. Biophys. Acta, 2007, 1768, 794; (c) D. E. Gloriam,
et al., BMC Genomics, 2007, 8, 338; (d) M. A. Hanson, et al.,
Structure, 2008, 16, 897; (e) Y.-M. Sung, et al., Proc. Natl.
Acad. Sci. U. S. A., 2016, 113, 3539.
3 (a) V. T. Martin, et al., Headache, 2006, 46, 125; (b) L.
Diatchenko, et al., Am. J. Med. Genet., Part B, 2006, 141, 449;
(c) J. S. Skouen, et al., Eur. J. Pain, 2012, 16, 1232.
4 (a) F. J. Prime, Drugs, 1971, 1, 269; (b) H. S. Nelson, N. Engl.
J. Med., 1995, 333, 499; (c) M. Johnson, J. Allergy Clin.
Immunol., 2006, 117, 18; (d) N. Hizawa, J. Clin. Pharm. Ther.,
2009, 34, 631.
16 B. Feve, et al., Proc. Natl. Acad. Sci. U. S. A., 1994, 91,
5677.
17 (a) T. Frielle, et al., Proc. Natl. Acad. Sci. U. S. A., 1987, 84,
7920; (b) X. M. Zhou, et al., Cell. Signalling, 1995, 7, 207.
18 (a) O. E. Brodde, et al., J. Cardiovasc. Pharmacol., 1984, 6,
1184; (b) H. E. Hopkinson, et al., Br. J. Pharmacol.,
2000, 131, 124.
19 (a) J. Besnard, et al., Nature, 2012, 492, 215; (b) W. K.
Kroeze, et al., Nat. Struct. Mol. Biol., 2015, 22, 362.
20 (a) Y. Cheng and W. H. Prusoff, Biochem. Pharmacol.,
1973, 22, 3099; (b) P. J. Munson and D. Rodbard, J. Recept.
Res., 1988, 8, 533; (c) R. Z. Cer, et al., Nucleic Acids Res.,
2009, 37, W441.
5 (a) J. J. Baldwin, et al., J. Med. Chem., 1979, 22, 687; (b) N.
Ono and M. Kaneko, Life Sci., 1995, 57, 345; (c) J. Jakubetz,
et al., J. Cardiovasc. Pharmacol., 1999, 33, 461; (d) R. C.
McLean, et al., Am. J. Cardiol., 2012, 110, 1001.
6 S. Mittal, et al., Science, 2017, 357, 891.
7 (a) K. Leftheris and M. Goodman, J. Med. Chem., 1990, 33,
216; (b) V. Quirke, Med. Hist., 2006, 50, 69; (c) M. Audet and
M. Bouvier, Nat. Chem. Biol., 2008, 4, 397; (d) A. M. Ring,
et al., Nature, 2013, 502, 575; (e) L. J. Martin, et al., J. Pain,
2015, 16, 1321; ( f ) H. C. S. Chan, et al., Sci. Rep., 2016, 6,
34736.
21 (a) A. S. Pupo, et al., Pharmacol. Res., 2016, 112, 49; (b) Z.
Rankovic, et al., Bioorg. Med. Chem. Lett., 2016, 26, 241; (c)
B. A. Evans, et al., Br. J. Pharmacol., 2010, 159, 1022; (d)
M. T. Drake, et al., J. Biol. Chem., 2008, 283, 5669.
8 (a) R. Howe and B. S. Rao, J. Med. Chem., 1968, 11, 1118; (b)
Y. Q. Wen, et al., Chemistry, 2014, 20, 3806.
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2018