Fluorescent-Labeled Selective Adenosine A2B Receptor Antagonist Enables Competition Binding Assay by Flow Cytometry

Article abstract of DOI:10.1021/acs.jmedchem.7b01627

Fluorescent ligands represent powerful tools for biological studies and are considered attractive alternatives to radioligands. In this study, we developed fluorescent antagonists for A_2B adenosine receptors (A_2BARs), which are targeted by antiasthmatic xanthines and were proposed as novel targets in immuno-oncology. Our approach was to merge a small borondipyrromethene (BODIPY) derivative with the pharmacophore of 8-substituted xanthine derivatives. On the basis of the design, synthesis, and evaluation of model compounds, several fluorescent ligands were synthesized. Compound 29 (PSB-12105), which displayed high affinity for human, rat, and mouse A_2BARs (K_i = 0.2-2 nM) and high selectivity for this AR subtype, was selected for further studies. A homology model of the human A_2BAR was generated, and docking studies were performed. Moreover, 29 allowed us to establish a homogeneous receptor-ligand binding assay using flow cytometry. These compounds constitute the first potent, selective fluorescent A_2BAR ligands and are anticipated to be useful for a variety of applications.

Full text of DOI:10.1021/acs.jmedchem.7b01627

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Fluorescent-labeled selective adenosine A2B receptor
antagonist enables competition binding assay by flow cytometry
Meryem Köse, Sabrina Gollos, Tadeusz Karcz, Amelie Fiene, Fabian Heisig, Andrea
Behrenswerth, Katarzyna J. Kiec-Kononowicz, Vigneshwaran Namasivayam, and Christa E Müller
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01627 • Publication Date (Web): 22 Apr 2018
Downloaded from http://pubs.acs.org on April 23, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Fluorescentꢀlabeled selective adenosine A receptor
2
B
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
antagonist enables competition binding assay by
flow cytometry
1≠
1≠
2
1
1
Meryem Köse, * Sabrina Gollos, Tadeusz Karcz, Amelie Fiene, Fabian Heisig, Andrea
1
2
1
Behrenswerth, Katarzyna KiećꢀKononowicz, Vigneshwaran Namasivayam, and
1
Christa E. Müller *
1
PharmaCenter Bonn, Pharmaceutical Institute, Pharmaceutical Chemistry I, University of Bonn,
2
An der Immenburg 4, Dꢀ53121 Bonn, Germany, Department of Technology and Biotechnology
of Drugs, Faculty of Pharmacy, Jagiellonian University Medical College, Medyczna 9, 30ꢀ688,
Kraków, Poland
≠
authors contributed equally
KEYWORDS: GPCR, adenosine A2B receptor, antagonist, BODIPY, fluorescent ligand,
fluorophore, synthesis
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Fluorescent ligands represent powerful tools for biological studies and are considered as
attractive alternatives to radioligands. In this study, we developed fluorescent antagonists for A2B
adenosine receptors (A ARs), which are targeted by antiꢀasthmatic xanthines and were
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
B
proposed as novel targets in immunoꢀoncology. Our approach was to merge a small
borondipyrromethene (BODIPY) derivative with the pharmacophore of 8ꢀsubstituted xanthine
derivatives. Based on the design, synthesis and evaluation of model compounds several
fluorescent ligands were synthesized. Compound 29 (PSBꢀ12105), which displays high affinity
for human, rat and mouse A ARs (K 0.2ꢀ2 nM) and high selectivity versus all other AR
2
B
i
subtypes, was selected for further studies. A homology model of the human A AR was
2
B
generated and docking studies were performed. Moreover, 29 served to establish a homogeneous
receptorꢀligand binding assay using flow cytometry. These compounds constitute the first potent,
selective fluorescent A AR ligands and are anticipated to be useful for a variety of applications.
2
B
ACS Paragon Plus Environment
2

Page 3 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Adenosine acts as a signaling molecule by activating G proteinꢀcoupled membrane receptors
1
(GPCRs) termed A , A , A and A adenosine receptors (ARs). These receptors are
1
2A
2B
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ubiquitously expressed, but their expression levels and subtype distribution is organꢀ and tissueꢀ
specific. Moreover, AR expression is regulated under pathological conditions, e.g. upregulation
of A ARs is observed under inflammatory, ischemic and hypoxic conditions. The specific
2B
blockade of the A AR subtype has been proposed as a novel strategy, e.g. for the treatment of
2
B
2ꢀ6
asthma, inflammatory bowel disease, pain, diabetes, sickle cell disease, and cancer. The
effectiveness of A AR antagonists in preclinical cancer models has recently attracted much
2
B
attention since they do not only prevent suppression of immune cells by adenosine in the
microenvironment of cancer tissues, similarly as A AR antagonists, but they additionally
2
A
7
ꢀ9
display direct antiꢀproliferative effects on cancer cells and inhibit angiogenesis. Moreover, we
recently showed that A ARs can form homomeric assemblies and heteromeric complexes with
2
B
the closely related A AR subtype. The A AR is thereby able to completely block A AR
2
A
2B
2A
10
signaling.
In drug development the direct measurement of receptorꢀligand affinity provides important
3
131
information. Radioligands are typically used to obtain these data, in particular Hꢀ or Iꢀlabeled
18
11
compounds that are used for in vitro studies, while Fꢀ or Cꢀlabeled tracers are suitable for in
vivo studies employing positron emission tomography (PET). Due to the high costs of
radioligands, the risk in handling radioactive materials, and the problem of producing radioactive
3
waste with a long halfꢀlife of more than 12 years in case of the frequently employed Hꢀlabeled
ligands, the development of fluorescent ligands has gained importance and may become the
future method of choice, especially for in vitro studies. Fluorescent ligands enable the
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
11
application of powerful techniques including fluorescence polarization (FP), fluorescence
1
2ꢀ14
15,16
correlation spectroscopy (FCS),
fluorescence resonance energy transfer (FRET),
flow
1
7
18
cytometry and fluorescence recovery after photobleaching (FRAP) experiments. Several
1
1ꢀ17,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
fluorescentꢀlabeled ligands for ARs, agonists and antagonists, have already been described
19ꢀ37
and some of them were shown to be useful tools for studying the receptors on a molecular
3
8,39
level
. Fluorescentꢀlabeled adenosine and Nꢀethylcarboxamidoadenosine (NECA) derivatives
6
were prepared by attaching fluorophores via long linkers to the adenine N ꢀ or C2ꢀposition (see
compounds 1-9, Figure 1). However, most of the obtained fluorescent agonists also activate A1ꢀ,
1
2ꢀ14, 21ꢀ24, 31ꢀ32, 34, 40ꢀ42
A ꢀ and A ARs and are thus nonꢀselective.
Selective fluorescent A AR
3
2
A
3
agonists were developed by coupling of a potent and selective A agonist (MRS3558) with the
3
26
27
fluorophores Cy5 (7, MRS5218) , squaraineꢀrotaxane (8, MRS5243) , or 4ꢀpyrene (9,
29
MRS5704) (for structures see Figure 1). The determined affinities at the human A AR were in
3
the nanomolar range (17.2 nM, 239 nM and 68.3 nM, respectively).
Fluorescent AR antagonists have also been prepared (see, for examples, compounds 10-12 in
Figure
2.
These
were
based
on
the
xanthine
derivative
8ꢀ[4ꢀ[[[[(2ꢀ
aminoethyl)amino]carbonyl]methyl]oxy]phenyl]ꢀl,3ꢀdipropylxanthine (XAC), e.g. compound 10,
the pyrazolotriazolopyrimidine derivative 2ꢀ(2ꢀfuranyl)ꢀ7ꢀ[3ꢀ(4ꢀmethoxyphenyl)propyl]ꢀ7Hꢀ
pyrazolo[4,3ꢀe][1,2,4]triazolo[1,5ꢀc]pyrimidinꢀ5ꢀamine
triazoloquinazoline derivative 9ꢀchloroꢀ2ꢀ(2ꢀfuryl)[1,2,4]triazolo[1,5ꢀc]quinazolinꢀ5ꢀamine
(CGS15943), e.g. compounds 11 and 12, to which fluorescent dyes were attached.
Compound 11 (MRS5449) was found to be a potent A AR antagonist (K 6.4 nM, human A AR)
(SCH442416),
or
the
related
1
1,13,17,33,37
3
i
3
but displayed low selectivity versus the other AR subtypes (K 73.0 nM at the human A AR,
i
2A
17
87.0 nM at the human A AR) (see Figure 2). Moreover, fluorescentꢀlabeled dendrimers were
1
ACS Paragon Plus Environment
4

Page 5 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
25,27
developed as ligands for A ꢀ, A ꢀ or A ARs.
Recently, the first fluorescenceꢀbased binding
1
2A
3
1
7,39
assays for ARs – specifically for the A AR subtype – were developed.
So far, no A2Bꢀ
3
selective fluorescent AR ligands have been reported.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Typically used fluorophores include fluorescein, rhodamine, coumarine and
43
borondipyrromethene (BODIPY) derivatives . Traditional fluorophores such as dansyl, pyrene
and FITC are fraught with significant drawbacks including pH sensitivity and short fluorophore
excitation wavelength interfering with autofluorescence of cells and tissues, which hinders their
use as molecular probes. BODIPY derivatives display a number of advantages, e.g. they are
photochemically stable, show high absorption coefficients and high fluorescence quantum yields
resulting high peak intensity. Moreover, they display λmax values of ≥ 500 up to 630 nm
depending on the substitution pattern. All of these properties make them ideally suitable
4
4ꢀ47
fluorophores for biological studies.
We previously reported on the synthesis of small
BODIPY derivatives bearing a variety of functional groups, which can readily be coupled to
target molecules. Moreover, we had demonstrated their utility by synthesizing a fluorescentꢀ
4
8
labeled A AR agonist.
3
In the present study, we designed and developed the first fluorescent A AR ligands by
2
B
integrating a BODIPY into the pharmacophore of an A ꢀselective AR antagonist. This led to
2
B
compounds with high affinity and A AR subtypeꢀselectivity. Moreover, we applied one of the
2
B
new fluorescent A AR ligands for establishing a flow cytometryꢀbased A AR fluorescence
2
B
2B
binding assays, which is performed on living cells and may be used to replace radioligand
binding studies.
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
SO3-
N
Cl
Cl
Cl
HN
N
HN
HN
N
N
N
N
N
N
N
N
N
H
H
N
N
H
N
O
N
O
N
O
H3C
O
H3C
H3C
H
N
N
5
SO3-
N
OH OH
H
OH OH
OH OH
O
N
N
N
Rotaxane
(MRS5218)26
8 (MRS5243)27
9 (MRS5704)29
7
Cy5-labeled adenosine analog
K human A ): no binding
Squaraine rotaxane-labeled adenosine analog
Ki (human A1): no binding
4-Pyrenyl-labeled adenosine analog
i (
1
Ki (human A ): no binding
1
K human A2A): 4730 nM
Ki (human A2A): no binding
Ki (human A2A): 3110 nM
i (
K (human A ): 17.2 nM
Ki (human A3): 239 nM
Ki (human A ): 68.3 nM
3
i
3
Figure 1. Examples of fluorescent labeled AR agonists
ACS Paragon Plus Environment
6

Page 7 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Structures of selected fluorescent adenosine receptor antagonists
A fluorescent receptor ligand with ideal properties for the desired biological studies would
combine the following properties: (i) high wavelength of excitation (> 350 nm) and emission, (ii)
waterꢀsolubility, (iii) high absorption coefficient and high fluorescence quantum yield resulting
in high peak intensity, (iv) small size in order not to interfere with binding to the target, and (v)
photochemical stability (under basic, reductive and oxidative conditions).
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
RESULTS AND DISCUSSION
Design and synthesis of model compounds
Our approach was to design a fluorescent A ꢀselective AR antagonist by integrating a small
2
B
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
fluorophore into the target structure as part of its pharmacophore. This is different from most
previous approaches which utilized the attachment of a mostly large fluorophore via a long linker
to the AR ligand (see Figure 1 and 2). We had previously discovered that 3ꢀunsubstituted 8ꢀ
phenylꢀ1ꢀpropylxanthine derivatives generally displayed higher A AR affinity and subtypeꢀ
2
B
selectivity than the corresponding 1,3ꢀsubstituted derivatives. Based on the known structureꢀ
4
9
activity relationships for xanthine derivatives, and supported by homology modeling and
5
0
docking studies (also see molecular modeling studies described below) we designed
fluorescentꢀlabeled derivatives in which the fluorophore was to be introduced into the paraꢀ
position of the 8ꢀphenyl ring, at which larger substituents are known to be well tolerated and can
even contribute to high affinity and selectivity. A small (carboxymethyl)oxy linkage was
selected to connect the fluorophore to the 8ꢀphenylꢀ1ꢀpropylxanthine core structure, which would
then allow to introduce an aminoꢀsubstituted fluorophore by amide formation. We had
previously developed a synthetic access to a toolbox of
small,
functionalized
greenꢀemitting
BODIPY
derivatives, including aminoꢀsubstituted compounds
such as 13ꢀ17, which can readily be coupled to target
4
8
molecules (see Figure 3).
In order to probe whether lipophilic, aromatic residues
like BODIPY derivatives were actually tolerated at that position by the A AR, and in particular
2
B
to investigate the optimal linker length between the xanthine core and the attached ring system,
ACS Paragon Plus Environment
8

Page 9 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
we initially prepared several model compounds. Instead of the precious BODIPY derivatives we
attached a simple phenyl ring. Even a small BODIPY derivative is considerably more lipophilic
and larger than a phenyl ring, and therefore, it has to be kept in mind that this constitutes a
simplification. Compounds 25-28 were synthesized starting from 6ꢀaminouracil (18) via 1ꢀ
propylpurineꢀ2,6ꢀdioneꢀ8ꢀphenoxyacetic acid (24) in analogy to published procedures (Scheme
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
1ꢀ54
1
).
Scheme 1. Synthesis of model compounds 25-28 substituted with a phenyl ring instead of a
fluorescent BODIPY residue
a
Reaction conditions: (i) Hexamethyldisilazane (HMDS), (NH ) SO , reflux for 2h; (ii) propyl
4
2
4
bromide, reflux for 1.5 h; (iii) addition of Na S O in H O, followed by saturated aq. NaHCO
3
2
2
3
2
b
solution at 0°C. Dimethylformamide (DMF)/H O, NaNO , concd. aq. HCl, 70°C, 1 h.
2
2
c
d
Methanol (MeOH), PtO /H (42 psi), rt, 2 h. Methylꢀ(4ꢀformylphenoxy)acetate, ethanol, reflux,
2
2
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
e
f
3
0 min. Thionyl chloride, heating from 0°C to 70 °C, 2 h. DMF, 0.1ꢀN aq. Na CO solution,
2
3
g
1
00°C, 30 min. NꢀMethylmorpholine, DMF; addition of isobutyl chloroformate and the
appropriate phenylalkylamine at ꢀ22°C, warming to rt for 3h.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
ꢀAminoꢀ3ꢀpropyluracil (19) was obtained in 81 % yield by regioselective alkylation of 6ꢀ
5
3
aminouracil (18) with propyl bromide according to a method developed by Müller. Compound
9 was nitrosated with NaNO /acetic acid (AcOH) and subsequently reduced with H /PtO to
1
2
2
2
obtain 5,6ꢀdiaminoꢀ3ꢀpropyluracil (21). Since 21 is sensitive to humidity and to oxidation by air
it was immediately reacted with methylꢀ(4ꢀformylphenoxy) acetate yielding 22 in excellent yield
(98 %). Ring closure was performed by treatment with thionyl chloride at 70°C yielding xanthine
derivative 23. The methyl ester function was cleaved by heating with base (Na CO ) in
2
3
DMF/H O to obtain the free carboxylic acid 24. Amide formation of 24 was achieved by the
2
addition of the appropriate phenylalkylamine in the presence of Nꢀmethylmorpholine and
isobutyl chloroformate leading to the desired products 25-28.
Synthesis of BODIPY-labeled A AR antagonists 29-33
2B
Since different linker lengths were tolerated, but structureꢀactivity relationships of the phenylꢀ
substituted model compounds and the BODIPYꢀsubstituted analogs appeared to be somewhat
different, we decided to prepare a series of BODIPYꢀlabeled xanthine derivatives. The aminoꢀ
4
8
substituted BODIPY derivatives 13-17 were synthesized as previously described and
subsequently coupled with xanthine 24 in the presence of 2ꢀ(6ꢀchloroꢀ1Hꢀbenzotriazolꢀ1ꢀyl)ꢀ
1
,1,3,3ꢀtetramethyluronium hexafluorophosphate (HCTU), 1ꢀhydroxybenzotriazole (HOBt) and
Nꢀmethylmorpholine as a base (for 29 and 32) or by reaction with isobutyl chloroformate in the
ACS Paragon Plus Environment
10

Page 11 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
presence of Nꢀmethylmorpholine (for 30, 31 and 33) (see Scheme 2). The former method
provided somewhat higher yields. The fluorescent ligands 29-33 were purified by flash
chromatogaphy on silica gel yielding 11ꢀ40 % of pure products (for details see Experimental
Section).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 2. Synthesis of the fluorescent ligands 29-33
a
Reaction conditions: HCTU, HOBt, Nꢀmethylmorpholine, dry DMF, rt, 16h for the preparation
of 29 (40% yield) and 32 (30% yield); isobutyl chloroformate, Nꢀmethylmorpholine, dry DMF, rt
3h, for the synthesis of 30 (22% yield), 31 (17% yield), and 33 (11% yield).
Analysis of the products
All products were analyzed by HPLCꢀ(DAD)UVꢀ(ESI)MS analysis and showed at least 95 %
1
13
purity. The structures of the new compounds were confirmed by H and C NMR spectra. The
fluorescent properties of 29 were determined by fluorescence spectroscopy. The absorption
maximum was at 497 nm, and the emission at 498 nm (for absorption and emission spectra see
Figure S1 in Supporting Information). The fluorescence quantum yield was measured in ethanol
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with rhodamine 6G as a reference compound and was found to be 0.94, close to 1, which is very
high as expected for this kind of BODIPY dye.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Biological evaluation
Radioligand binding studies
The synthesized model compounds 25-28 as well as the fluorescent ligands 29-33 were
investigated in radioligand binding studies at the human A AR stably expressed in Chinese
2
B
hamster ovary (CHO) cells versus the antagonist radioligand [³H]PSBꢀ603 (Table 1). In order to
assess the compounds’ selectivity, radioligand binding studies at A AR, A AR and A ARs were
1
2A
3
additionally performed using the following radioligands: [³H]CCPA (A ), [³H]MSXꢀ2 (A ), and
1
2A
[
³H]PSBꢀ11 (A ). For comparison, data for the unlabeled standard A AR antagonist 8ꢀ(4ꢀ(4ꢀ(4ꢀ
3 2B
chlorophenyl)piperazineꢀ1ꢀsulfonyl)phenyl)ꢀ1ꢀpropylxanthine
(34, PSBꢀ603)
and
the
nonselective antagonist caffeine (35), obtained in the same assay, are provided (see Table 1).
All tested model compounds displayed high affinity for the A AR with K values in the low
2
B
i
nanomolar range. The model compound with a linker length of four methylene units was the
most potent A2B antagonist (28, K = 3.30 nM), however derivatives with a methylene or an
i
ethylene linker were similarly potent (25, K 7.00 nM; 26, K 7.45 nM). Only the compound with
i
i
a propylene chain appeared to be somewhat less potent 27, K 27.9 nM, 8ꢀfold difference
i
compared to 28. All model compounds were shown to be highly (> 150ꢀfold) selective versus the
A AR subtype. Selectivity versus the A AR was also high (> 100ꢀfold) for most compounds
2
A
3A
with the exception of 25, the derivative with the methylene linker, which still displayed 38ꢀfold
selectivity for A2B over A . However, selectivity appeared to be much lower versus the A AR,
3
1
ACS Paragon Plus Environment
12

Page 13 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
although in this case – like in the case of the A AR – data for the human A AR were compared
2
A
2B
with data obtained at the rat A AR. This appeared to be accurate enough for a preliminary
1
assessment of model compounds, but may actually underestimate the compounds’ A selectivity
2
B
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
for the human A AR since xanthine derivatives are often more potent at rat as compared to
2
B
49
human A ARs. In summary, compound 26 showed the best profile among the tested model
1
compounds, i.e. high potency, short linker chain, and high selectivity, at least versus A2A and
A ARs, and thus an ethylene linker appeared to be most promising. Therefore, we synthesized
3
the corresponding BODIPYꢀlabeled xanthine derivative 29. In addition, we prepared the
fluorescent ligands 30ꢀ33 with longer linker lengths up to ten methylene units and evaluated
them in radioligand binding assays (see Table 1).
Table 1. Adenosine receptor binding affinities
K ± SEM (nM)
i
Compound
n
A AR
A AR
[ H]MSX-2
A AR
[³H]PSB-603
A AR
[³H]PSB-11
1
2A
2B
3
3
3
[
H]CCPA
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Human
Human
Human (h),
Human
recombinant
(h)
recombinant (h) recombinant (h) rat (r) or
or rat brain
cortex (r)
or rat brain
striatum (r)
mouse (m)
recombinant
5
5
55
55
56
49
3
4
-
-
> 10,000 (h, r)
> 10,000 (h, r)
> 10,000 (h, r)
0.553 (h) 5
0.351 (m)
> 10,000 (h)
> 13,300 (h)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
(
PSBꢀ603)
57a
55
3
5
23,400 (h)
28,700 (r)
33,800 (h)5
6
(
Caffeine)
30,000 (r)
5
6
23,000 (m)
25
1
2
3
4
27.2 ± 4.1 (r)
7.85 ± 2.46 (r)
35.3 ± 5.7 (r)
29.6 ± 4.0 (r)
> 10,000 (r)
> 10,000 (r)
> 10,000 (r)
498 ± 132 (r)
7.00 ± 1.39 (h)
7.45 ± 2.43 (h)
27.9 ± 9.1 (h)
3.30 ± 1.24 (h)
266 ± 61 (h)
> 10,000 (h)
26
27
28
2,990 ± 438 (h)
1,410 ± 234 (h)
2
9
2
≥ 10,000 (h)
422 ± 35 (r)
> 10,000 (h)
> 10,000 (r)
1.83 ± 0.76 (h)
0.180 ± 0.067 (r)
> 10,000 (h)
(
PSB-12105)
1
.32 ± 0.26 (m)
3
0
3
4
5
168 ± 13 (r)
178 ± 23 (r)
2150 ± 392 (r)
1570 ± 277 (r)
11.6 ± 3.4 (h)
24.4 ± 6.6 (h)
5.90 ± 2.16 (h)
904 ± 155 (h)
719 ± 116 (h)
4550 ± 858 (h)
31
32
714 ± 272 (h)
77.9 ± 12.8 (r)
58.7 ± 14.4 (h)
477 ± 68 (r)
3
3
10 > 10,000 (h)
27 ± 21 (r)
1600 ± 140 (h)
> 10,000 (r)
7.82 ± 1.94 (h)
≥ 10,000 (h)
2
a
3
3
Human A2AAR vs. [ H]8ꢀcyclopentylꢀ1,3ꢀdipropylxanthine ([ H]DPCPX)
As already observed for the phenylꢀsubstituted model compounds 25ꢀ28, the linker length had
also only a minor effect on A2B affinity of the BODIPYꢀlabeled xanthines. All derivatives,
including the ones with a longer C5ꢀ (32) or C10ꢀlinker (33), displayed high A AR affinity with
2
B
low nanomolar K values. However, the SARs showed subtle differences, not only regarding A
i
2B
ACS Paragon Plus Environment
14

Page 15 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
affinity but also with respect to selectivity versus the other AR subtypes. The A ꢀselectivity of
2
B
the BODIPYꢀlabeled xanthines versus the A AR was generally higher (compare 26/29, 27/30
1
and 28/32) than that of the model compounds. The fluorescent ligand 29 showed the highest A2B
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
affinity of all investiged compounds with a K value of 1.83 nM for the human A AR (see Table
i
2B
1
and Figure 4A). Moreover, similarly to one of the most potent and selective A2B antagonists,
the standard antagonist 34 (for structure and data see Table 1), 29 was found to be highly
>1,000ꢀfold) selective versus the human A , A and A AR subtypes. In order to evaluate
(
1
2A
3
potential species differences, 29 was further examined in radioligand binding assays using
membrane preparations of CHO cells stably overexpressing rat or mouse A ARs, respectively.
2
B
The fluorescent ligand 29 also displayed high affinity for rat and mouse A ARs, with K values
2
B
i
of 0.180 nM (rat A AR) and 1.32 nM (mouse A AR) (Table 1, Figure S5, Figure S6 in
2B
2B
Supporting Information). This means that 29 is similarly potent at A ARs of all three species.
2
B
Figure S6 summarizes the obtained radioligand binding data for the fluorescent A AR
2
B
antagonist 29. Given the wellꢀknown fact that xanthine derivatives lacking a substituent at N3
49
are inactive at rodent A ARs it can be concluded that 29 represents a very potent and highly
3
selective fluorescent A AR ligand which should not only be useful for studies in human cells
2B
and tissues, but also in rodent cells, tissues and animal models.
cAMP accumulation studies
As a next step, functional properties of the model compounds 25-27 and of the fluorescent
ligand 29 were studied employing cAMP accumulation assays. The model compounds as well as
the BODIPYꢀlabeled xanthine derivative 29 inhibited agonistꢀ (NECAꢀ, 10 µM) -induced cAMP
accumulation in A AR expressing CHO cells in a concentrationꢀdependent manner and to the
2
B
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
same maximal extent as the standard A AR antagonist 34. The fluorescent compound 29
2
B
showed an IC value of 262 nM, while the standard antagonist 34 displayed an IC value of
5
0
50
18.3 nM (see Figure 4B). The IC values determined for the model compounds were as follows:
5
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
35 nM for 25, 141 nM for 26 and 174 nM for 27. The apparently lower potency of the tested
antagonists 25-27, 29 and 34 in functional assays as compared to radioligand binding studies can
be explained by the high agonist concentration that was used in the assay to stimulate the
A ARs, which required relatively high concentrations of the competitive antagonists to block its
2
B
effect. These data clearly show that the model compounds as well as the fluorescent ligand 29
functionally behave as A AR antagonists. Moreover, the most potent antagonist in radioligand
2
B
binding studies, 34, was also the most potent antagonist in the functional assay (K 0.553 nM;
i
IC 18.3 nM).
5
0
A
B
Figure 4. A. Competition binding experiments of 29 at membrane preparations of recombinant
CHO cells expressing human A ARs. A K value of 1.83 ± 0.76 nM was determined. Data
2
B
i
points represent means ± SEM of three independent experiments. B. Concentrationꢀdependent
inhibition of NECA (10 µM)ꢀinduced cAMP accumulation by 25, 26, 27, and 29 and by the
ACS Paragon Plus Environment
16

Page 17 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
standard A AR antagonist 34 in CHO cells expressing the human A AR. IC values were 135
2
B
2B
50
±
45 nM (25), 141 ± 30 nM (26), 174 ± 38 nM (27), 262 ± 48 nM (29) and 18.3 ± 2.1 nM (34).
Data points represent means ± SEM from three independent experiments performed in
duplicates.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Confocal microscopy
Confocal microscopy was performed to investigate the imaging properties of the fluorescent
ligand 29. Incubation of CHO cells recombinantly expressing the human A AR (CHOꢀA )
2
B
2B
with 29 (100 nM) resulted in labeling of the cell membranes (Figure 5A), which was absent in
the control experiments using wildꢀtype (wt) CHO cells treated with 29 under the same
conditions (Figure 5D). To further confirm that the observed labeling was A ARꢀspecific, the
2
B
CHOꢀA2B cells were preꢀincubated with the selective A AR antagonist 34. In fact, preꢀ
2B
incubation with 34 (100 nM) for 30 min completely prevented membrane labeling (Figure 5B).
Similarly, preꢀtreatment with 1,3ꢀdipropylꢀ8ꢀcyclopentylxanthine (DPCPX, 1 µM) led to reduced
fluorescence although to a lower extent (Figure 5D).
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Confocal imaging. CHOꢀA cells (AꢀC) were grown on sterile chambered coverslips
2
B
and subsequently incubated with 29 (100 nM) for 15 min at 37°C, (A) in the absence, or (B) in
the presence of 34 (100 nM, preꢀincubation for 30 min), or (c) in the presence of DPCPX (1 µM,
preincubation for 30 min). As a control, CHO wt cells were treated with 29 under the same
conditions (D). All confocal images and their corresponding transmitted light images (lower
row) were taken using identical microscope settings regarding laser power, detector offset and
gain. Images shown are from a single representative experiment out of three similar experiments.
Scale bars = 100 µm.
Flow cytometry-based A AR binding assays
2
B
Finally, we utilized the new fluorescent A AR antagonist 29 to develop a homogenous flow
2
B
cytometryꢀbased fluorescence binding assay for the A AR using living CHO cells
2
B
recombinantly expressing the human A AR. The optimal cell number in a sample vial providing
2
B
the best signal window was determined to be in a range of 250,000 to 500,000 cells per vial.
ACS Paragon Plus Environment
18

Page 19 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Using Hank’s balanced salt solution (HBSS) as a buffer was found to result in a higher
fluorescence signal to noise ratio than Dulbecco’s Modified Eagle Medium (DMEMꢀF12) and
was therefore utilized as an incubation buffer in all subsequent experiments. For the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
measurement of nonspecific binding various AR ligands were tested, namely the A ꢀselective
1
AR antagonist DPCPX, the nonꢀselective AR agonist NECA and the A ꢀselective AR
2
B
antagonist PSBꢀ603 (34). In these preliminary experiments, the unlabeled ligands were coꢀ
applied with the fluorescent ligand 29. DPCPX was not regarded suitable for the determination
of nonspecific binding, since in initial experiments we could not detect specific binding of 29
using DPCPX as a competitor. Contrary to our expectations, the fluorescence signal even
increased in the presence of DPCPX. The reason for that might be the formation of aggregates
between the fluorescent ligand 29 and DPCPX at higher DPCPX concentration (10 µM)
5
8
disturbing the assay. However, as expected, the nonselective A AR agonist NECA as well as
2
B
the highly selective A AR antagonist 34 were clearly able to displace the fluorescent ligand 29
2
B
from its binding site and were thus found to be suitable for the determination of the nonspecific
binding. In notransfected CHOꢀK1 cells no specific binding could be determined showing that
the detected specific binding was merely due to the A AR, and 29 specifically labels A ARs.
2B
2B
Subsequently, we performed competition binding experiments and determined concentrationꢀ
dependent inhibition of binding of the fluorescent A AR ligand 29 (1 nM) by the standard
2
B
antagonist 34 (Figure 6D). An IC50 value of 0.534 nM was calculated for 34, which is in
5
5
excellent agreement with its K value determined in radioligand binding assays (0.553 nM).
i
After initial FACS experiments performed on a BD LSRII flow cytometer, we subsequently
adapted it to a 96ꢀwell format using a guava easyCyte HT sampling flow cytometer (for details
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
see Experimental Section). In preliminary measurements we observed fast association kinetics
for 29. Therefore, we subsequently changed the protocol and preꢀincubated the cells with the
compounds that were investigated for competition. This change in experimental conditions
allowed us to observe displacement of 29 by DPCPX, which had not inhibited 29 binding in
preliminary FACS experiments in which both compounds had been coꢀadministered. The
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
standard A2B antagonist 34 displayed an IC value of 0.739 nM which is not significantly
50
different from the IC50 value of 0.534 nM previously determined in flow cytometry assays
without preꢀincubation. Thus, preꢀincubation does not appear to be required for all antagonists, it
likely depends on the binding kinetics of each compound in comparison to the fluorescent ligand.
In all subsequent experiments, the antagonists were preꢀincubated for 30 min at 37°C with the
cells before adding the fluorescent ligand, followed by an additional incubation for 20 min.
Various standard A AR antagonists were examined for their ability to inhibit the binding of 29.
2
B
As shown in Figure 6B, all tested A AR antagonists displaced 29 with IC values ranging
2
B
50
between 1 and 100 nM.
These results indicate that the new flow cytometryꢀbased competition binding assay with the
newly developed A ꢀselective fluorescent antagonist 29 provides a convenient alternative
2B
method to radioligand binding assays, potentially adaptable to highꢀthroughput screening. Thus,
our fluorescent A AR antagonist 29 represents a new powerful tool for A AR studies.
2
B
2B
A
B
ACS Paragon Plus Environment
20

Page 21 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Fluorescence competition binding experiments using CHO cells recombinantly
expressing human A ARs. A. Experiments were performed using a BD LSRII flow cytometer.
2
B
The fluorescent ligand 29 was displaced by the A ꢀselective standard antagonist 34 in a
2
B
concentrationꢀdependent manner as determined by flow cytometry. An IC50 value of 0.534 ±
.155 nM was determined for 34. Data points represent means ± SEM from 5 independent
0
experiments, each performed in duplicates. B. Experiments were performed in a 96ꢀwell format
using a guava easyCyte HT sampling flow cytometer. The determined IC50 values were as
follows: 0.739 ± 0.622 nM (34), 1.06 ± 0.71 nM ((8ꢀ(4ꢀ(4ꢀ(4ꢀchlorobenzyl)piperazineꢀ1ꢀ
sulfonyl)phenyl)ꢀ1ꢀpropylxanthine (PSBꢀ0788)), 3.80 ± 1.15 nM (8ꢀcyclopentylꢀ1,3ꢀ
dipropylxanthine (DPCPX)), 23.5 ± 12.7 nM (8ꢀ[4ꢀ(4ꢀbenzylpiperazideꢀ1ꢀsulfonyl)phenyl]ꢀ1ꢀ
propylxanthine (PSBꢀ601)), 60.7 ± 27.4 nM (9ꢀchloroꢀ2ꢀ(2ꢀfuranyl)[1,2,4]triazolo[1,5ꢀ
c]quinazolinꢀ5ꢀamine (CGSꢀ15943)), and 94.9 ± 76.6 nM (4ꢀ(2ꢀ[7ꢀaminoꢀ2ꢀ(2ꢀfuryl)ꢀ
[1,2,4]triazoloꢀ[2,3ꢀa][1,3,5]triazinꢀ5ꢀylamino]ethyl)ꢀphenol (ZMꢀ241385)). Data points
represent means ± SEM from 3ꢀ5 independent experiments, each performed in duplicates.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
All values determined by the fluorescenceꢀbased competition assay using flow cytometry were in
5
5,56
the same range as those previously determined by radioligand binding studies.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Molecular modeling studies
The design of the potent and selective fluorescent A AR antagonist 29 was supported by
2
B
docking studies using a homology model of the human A AR. We had previously constructed
2B
5
0
such a model based on the crystal structure of the human A AR (2.6 Å) in complex with the
2
A
59
subtypeꢀselective antagonist ZM241385 (PDB ID: 3EML). In the present study, the homology
model of the human A AR was updated using the crystal structure of the human A AR
2
B
2A
obtained at a higher resolution (1.8 Å) in complex with ZM241385 (PDB ID: 4EIY) as a
60
template. The binding pose of 29 predicted by the docking studies showed that the xanthine
core with the attached 8ꢀphenyl group closely overlaps with the crystallographic pose of xanthine
amine congener (XAC, 36) observed in the Xꢀray structure its complex with the A AR with
2
A
6
1
(
PDB ID: 3REY, resolution 3.31 Å) (see Figure 7). A recently reported crystallographic pose
for another xanthine derivative, which was originally developed by our group, 1ꢀbutylꢀ8ꢀ
hexahydroꢀ2,5ꢀmethanopentalenꢀ3a(1H)ꢀyl)ꢀ3,7ꢀdihydroꢀ3ꢀ(3ꢀhydroxypropyl)ꢀ1Hꢀpurineꢀ2,6ꢀ
dione (PSBꢀ36, 37, Figure 7), in complex with the human A AR showed a different, alternative
(
2
A
binding mode: the hydroxypropyl substituent attached to the N3 of 37 pointed towards the
position of the 8ꢀphenyl substituent of the BODIPYꢀlabeled 29. The 8ꢀ(3ꢀnoradamantyl) residue
6
2
of 37 was positioned between the C6ꢀcarbonyl and the N7ꢀH of the xanthine core of XAC (36).
The reason for the shifted pose of the xanthine core can be explained by the larger substituents at
N1 (butyl) and N3 (hydroxpropyl) in the 8ꢀnoradamantylxanthine derivative 37 as compared to
29 and 36, which would cause a steric clash with Met270.
ACS Paragon Plus Environment
22

Page 23 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. Proposed binding mode of 29. (A) Docking pose of the antagonist 29 (represented in
stick and carbon atoms colored in magenta) in the active site of the A AR (in ribbon
2
B
presentation and colored in continuum spectrum). (B) Docked pose of 29 in the binding pocket
of the A AR. (C) The predicted pose of 29 observed in the performed docking studies compared
2
B
to the binding pose of XAC (36) (carbon atoms colored in cyan, PDB ID: 3REY) and PSBꢀ36
37) (carbon atoms colored in orange, PDBID:5N2R) as observed in the Xꢀray structures in
complex with the A AR. (D) Comparison of the structures of 29 (PSBꢀ12105), 36 (XAC) and
(
2
A
3
7 (PSBꢀ36). Oxygen atoms are colored in red, nitrogen atoms in blue, hydrogen atoms in silver
white, boron atoms in light pink and fluorine atoms in light cyan.
ACS Paragon Plus Environment
24

Page 25 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 8. Proposed binding interactions of 29. (A) The docked pose of 29 (in magenta) and the
amino acid residues in the binding pocket of the A AR (pale green) are compared to 36 (XAC)
2
B
(
cyan) and the amino acid residues in the binding pocket of the A AR. Amino acids differing in
2A
A AR and A AR are colored in red. (B) A 2D interaction diagram of compound 29 with the
2
A
2B
human A AR and of 36 with the human A AR are depicted. The binding pose representation
2B
2A
and the color code is as described in Figure 7.
ACS Paragon Plus Environment
26

Page 27 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
As shown in Figure 8, the heterocyclic core structure of 29 likely forms one of the key πꢀπ
stacking interactions with Phe173 and utilizes the hydrophobic surface provided by Ile276 in the
A AR. The C6ꢀcarbonyl and the N7ꢀH of the xanthine core form key hydrogen bond
2
B
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
interactions with Asn254 and Glu174, respectively. The residues Phe173, Asn254 and Ile276 are
conserved among the different subtypes of human ARs, and Glu174 is common in three subtypes
(
A , A , A ), but replaced by valine in the A AR at the corresponding position (Val169, A AR
1 2A 2B 3 3
numbering, see Sequence Alignment in Supplementary Figure S9). In the crystal structure of the
human A AR complexed with XAC, the corresponding amino acid for Glu174 in the A AR
2
A
2B
(
Glu169 in A AR numbering) is rotated around by 150° and does not form hydrogen bond
2A
interactions with the N7ꢀH of the xanthine core of XAC. Instead, in the crystal structure with
ZM541385, Glu169 of the A AR forms an interaction with the free exocylic amino group
2
A
(
NH ) of the heterocyclic core. These observations further support the selection of the high
2
resolution crystal structure of the A AR (4EIY.pdb) as a template for the homology model of
2
A
the A AR. The propyl substituent at N1 of the xanthine derivative 29 binds within the pocket
2
B
formed by Met182, Val250 and His251. Among these residues, Val250 is unique for the A AR
2
B
which might contribute to the selectivity of the compounds for the A AR. The phenyl group at
2
B
C8 is coꢀplanar with the heterocyclic core and forms a weak hydrophobic interaction with Ile67.
The linker chain length contributes to the fold, and the carboxamide group in the linker forms
hydrogen bonding interactions with the main chain of Ser68 and the sideꢀchain of Lys279. The
interaction between the keto group of the carboxamide function and Lys279 possibly contributes
tothe selectivity of compound 29 for the A AR, in addition to Val250 in the heterocyclic core
2
B
binding pocket. In comparison to other subtypes of human ARs, both Val250 and Lys279 are
unique for the A AR (Figure 8). Additionally, another unique residue, namely Asn273 in the
2
B
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A AR, is in closer proximity to Lys279 and near the binding pocket of 36. According to the
2
B
model, the BODIPY residue of 29 binds within the pocket formed by the residues Leu3, Gln6,
Asp7, and Leu69 and extends to the surface between transmembrane regions TM1 and TM7 of
the receptor. Among these residues interacting with BODIPY, Gln6 and Asp7 are unique for the
A AR and might contribute to the compound’s selectivity. Thus, the docking studies provide a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
B
rational explanation for the interaction of 29 with the A AR and for its selectivity versus the
2B
other AR subtypes, which will have to be confirmed by future mutagenesis studies.
CONCLUSIONS
In summary, we have developed a fluorescent antagonist for the A AR, compound 29, in which
2
B
a BODIPY fluorophore is integrated into the A2B antagonist pharmacophore structure. The
fluorescent ligand shows equally high A AR affinity of around 1ꢀ2 nM for human, rat and
2
B
mouse A ARs combined with high A AR subtype selectivity. It possesses excellent
2
B
2B
fluorescence characteristics, in particular high fluorescence quantum yields. A refined homology
model of the human A AR was built based on a recently published highꢀresolution Xꢀray
2
B
structure of the A AR, and the binding of 29 to the orthosteric binding site was rationalized.
2A
The fluorescent antagonist ligand 29 was successfully used to establish a straightforward
homogeneous fluorescence binding assay using flow cytometry. The new A AR fluorescence
2
B
assay was found to be reliable and may be used to substitute radioligand binding studies. Future
studies will be directed towards evaluating 29 for further applications, e.g. in native cells and
tissues.
ACS Paragon Plus Environment
28

Page 29 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EXPERIMENTAL SECTION
General methods. All reactions were performed in dry solvents, unless otherwise indicated.
Dichloromethane (DCM) was freshly distilled over CaH prior to use. All other reagents were
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
used as obtained from various providers (Acros, Aldrich, Fluka, Merck, Sigma) unless otherwise
noted. The reactions were monitored by thin layer chromatography (TLC) using silica gelꢀcoated
aluminum sheets (0.2 mm layer, nanoꢀsilica gel 60 with fluorescence indicator UV254 (Merck,
Darmstadt, Germany)). Column chromatography was performed on Merck silica gel 60 (mesh
size 0.040ꢀ0.063 mm). Melting points were measured in open capillary tubes on a Wepa Apotec
capillary melting point apparatus (Wepa, HöhrꢀGrenzhausen, Germany) and are uncorrected.
1
13
NMR spectra were recorded on a Bruker Avance DRX 500 spectrometer. Hꢀ and C NMR data
1
were collected on a Bruker Avance 500 MHz NMR spectrometer at 500 MHz ( H), or 125 MHz
1
3
(
C), respectively. The employed solvents are given below. Data are presented as follows;
chemical shift, multiplicity (s, singlet; d, doublet; dd, double doublet; t, triplet; q, quartet; m,
multiplet; br, broad; br s, broad singlet), coupling constant, and integration.
The purities of isolated products were determined by ESIꢀmass spectra obtained on an LCMS
instrument (Applied Biosystems API 2000 LCMS/MS, HPLC Agilent 1100) using the following
procedure: the compounds were dissolved at a concentration of 1.0 mg/mL in acetonitrile
containing 2 mM ammonium acetate. Then, 10 ꢁL of the sample were injected into an HPLC
®
column (MachereyꢀNagel Nucleodur 3 µ C18, 50 x 2.00 mm). Elution was performed with a
gradient of water/acetonitrile (containing 2 mM ammonium acetate) from 90:10 to 0:100 for 20
min at a flow rate of 300 ꢁL/min, starting the gradient after 10 min. UV absorption was detected
from 200 to 950 nm using a diode array detector (DAD). Purity of all compounds was
determined at 254 nm. The purity of the compounds was ≥95%.
ACS Paragon Plus Environment
29