New fluorescent adenosine A1-receptor agonists that allow quantification of ligand-receptor interactions in microdomains of single living cells

Article abstract of DOI:10.1021/jm061279i

Fluorescence spectroscopy is becoming a valuable addition to the array of techniques available for scrutinizing ligand-receptor interactions in biological systems. In particular, scanning confocal microscopy and fluorescence correlation spectroscopy (FCS) allow the noninvasive imaging and quantification of these interactions in single living cells. To address the emerging need for fluorescently labeled ligands to support these technologies, we have developed a series of red-emitting agonists for the human adenosine A₁-receptor that, collectively, are N⁶-aminoalkyl derivatives of adenosine or adenosine 5′-N-ethyl carboxamide. The agonists, which incorporate the commercially available fluorophore BODIPY [630/650], retain potent and efficacious agonist activity, as demonstrated by their ability to inhibit cAMP accumulation in chinese hamster ovary cells expressing the human adenosine A₁-receptor. Visualization and confirmation of ligand-receptor interactions at the cell membrane were accomplished using confocal microscopy, and their suitability for use in FCS was demonstrated by quantification of agonist binding in small areas of cell membrane.

Full text of DOI:10.1021/jm061279i

782
J. Med. Chem. 2007, 50, 782-793
New Fluorescent Adenosine A1-Receptor Agonists That Allow Quantification of
Ligand-Receptor Interactions in Microdomains of Single Living Cells
Richard J. Middleton,†,‡ Stephen J. Briddon,‡,§ Yolande Cordeaux,§ Andrew S. Yates,† Clare L. Dale,† Michael W. George,|
Jillian. G. Baker,§ Stephen J. Hill,*,§ and Barrie Kellam*,†
School of Pharmacy, Centre for Biomolecular Sciences, and School of Chemistry, UniVersity of Nottingham,
UniVersity Park, Nottingham, United Kingdom, NG7 2RD, and Institute of Cell Signalling, Medical School, UniVersity of Nottingham,
Nottingham, United Kingdom, NG7 2UH
ReceiVed NoVember 2, 2006
Fluorescence spectroscopy is becoming a valuable addition to the array of techniques available for scrutinizing
ligand-receptor interactions in biological systems. In particular, scanning confocal microscopy and
fluorescence correlation spectroscopy (FCS) allow the noninvasive imaging and quantification of these
interactions in single living cells. To address the emerging need for fluorescently labeled ligands to support
these technologies, we have developed a series of red-emitting agonists for the human adenosine A1-receptor
that, collectively, are N6-aminoalkyl derivatives of adenosine or adenosine 5′-N-ethyl carboxamide. The
agonists, which incorporate the commercially available fluorophore BODIPY [630/650], retain potent and
efficacious agonist activity, as demonstrated by their ability to inhibit cAMP accumulation in chinese hamster
ovary cells expressing the human adenosine A1-receptor. Visualization and confirmation of ligand-receptor
interactions at the cell membrane were accomplished using confocal microscopy, and their suitability for
use in FCS was demonstrated by quantification of agonist binding in small areas of cell membrane.
Introduction
monophosphate (cAMP). Our current understanding of A1-AR
pharmacology has arisen predominantly from the use of
conventional techniques for investigating ligand-receptor in-
teractions, such as radioligand binding and standard assays of
receptor function, which involve metabolic labeling. These
experiments require large numbers of cells and give a population
average for the pharmacological parameters they determine.
However, it is now evident that A1-ARs are not uniformly
distributed at the cell surface but, instead, are organized within
membrane compartments and microdomains,3 providing a
mechanism by which intracellular signaling can be orchestrated
in different areas of an individual cell. Studying the pharmacol-
ogy of such receptor subpopulations is beyond the scope of such
traditional radioligand binding assays.
G-protein coupled receptors (GPCRsa) represent the largest
family of cell-surface receptors mediating cellular communica-
tion and are a major target for drugs in current clinical use.1
The endogenous ligands for GPCRs are a diverse range of
hormones, transmitters, autocrine factors, and even photons. In
each case, however, the receptor transduces the binding of ligand
to an intracellular signal via activation of a heterotrimeric
guanosine triphosphate-binding protein (G-protein). As a con-
sequence of this, a range of downstream intracellular signals
are activated, resulting in both short-term effects (e.g., changes
in cellular calcium levels) and long-term effects (e.g., gene
transcription). The nucleoside adenosine is one such ligand, for
which there are four characterized GPCRs (A1-, A2A-, A2B-, and
A3-receptors).2 The adenosine A1-receptor (A1-AR) is respon-
sible for mediating the physiological effects of adenosine in
diverse tissues such as the brain, heart, kidney, and adipocytes2
and signals via G-proteins of the Gi/o family to inhibit adenylyl
cyclase and thus reduce cellular levels of cyclic adenosine
The introduction of fluorescence-based techniques such as
confocal microscopy and fluorescence correlation spectroscopy
(FCS) has progressed the study of GPCR pharmacology to the
single cell level4 and to the point where domain-specific receptor
pharmacology can realistically be investigated. Specifically, FCS
records fluctuations in photon emission as a fluorescent molecule
diffuses through a diffraction limited excitation volume (∼0.5
fL) created by a highly focused laser beam. Statistical analysis
of such fluctuations allows the concentration and diffusional
characteristics of these fluorescent molecules to be determined.5
Because FCS is noninvasive, sensitive, and quantitative, it is
ideal for use in living cells and has been employed successfully
to measure the diffusion of both genetically and chemically
labeled proteins.6 The ability of FCS to distinguish between fast-
moving small molecule ligands and their slower moving
receptor-bound forms has been exploited to perform single-cell
pharmacology.5b,7,8 To apply this technique to the study of
GPCRs, there is a clear prerequisite to fluorescently label the
receptors and the molecules that interact with them. In particular,
for the quantification of ligand-receptor interactions, there is
a specific need for fluorescent ligands with appropriate photo-
chemical and pharmacological properties.9 To this end, we have
recently demonstrated this approach in the synthesis and
* To whom correspondence should be addressed. Tel.: +44-115-951-
3026 (B.K.); +44-115-8230082 (S.J.H.). Fax: +44-115-9513412 (B.K.);
+44-115-8230081 (S.J.H.). E-mail: barrie.kellam@nottingham.ac.uk (B.K.);
stephen.hill@nottingham.ac.uk (S.J.H.).
† School of Pharmacy.
‡ These authors contributed equally to this work.
§ Institute of Cell Signalling, Medical School.
| School of Chemistry.
a Abbreviations: cAMP, cyclic adenosine monophosphate; AR, adenosine
receptor; BAIB, bis(acetoxy)iodobenzene; BODIPY 630/650 or BY630,
6-(((4,4-difluoro-5-(2-thienyl)-4-bora-3a,4a-diaza-s-indacene-3-yl)styryl-
oxy)acetyl)amino-hexanoic acid; BODIPY 630/650-X, SE, 6-(((4,4-difluoro-
5-(2-thienyl)-4-bora-3a,4a-diaza-s-indacene-3-yl)styryloxy)acetyl)amino-
hexanoic acid succinimidyl ester; CHO, chinese hamster ovary cell; CPA,
N6-cyclopentyladenosine; DIEA, di-isopropylethylamine; DMF, N,N-di-
methylformamide; DMP, 2,2-dimethoxypropane; DPCPX, 1,3-di-n-propyl-
8-cyclopentylxanthine; FCS, fluorescence correlation spectroscopy; GPCRs,
G-protein coupled receptors; NECA, adenosine 5′-N-ethyl carboxamide; τD,
diffusion time; TEMPO, 2,2,6,6,tetramethyl-1-piperidinyloxy free radical;
THF, tetrahydrofuran; XAC, xanthine amine congener; Z, benzyloxycar-
bonyl.
10.1021/jm061279i CCC: $37.00 © 2007 American Chemical Society
Published on Web 01/24/2007

Fluorescent Adenosine A1-Receptor Agonists
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4 783
Scheme 1a
a Reagents and conditions: (a) DMP, tosic acid, acetone, rt, 18 h, 60%; (b) BAIB, TEMPO, MeCN/water (1:1), rt, 2 h, 83%; (c) (i) SOCl2, DMF, CHCl3,
reflux, 5 h (ii) EtNH2, THF, 5 °C to rt, 30 min, 60-90%; (d) Z-Cl, CHCl3, rt, 18 h, 35-65%; (e) 5a-d, DIEA, EtOH reflux, 18 h, 70-85%; (f) 1 M
HCl(aq)/1,4-dioxane (1:1), 70 °C, 4 h, 60-70%; (g) H2/Pd/C, MeOH/water/AcOH (9.0:0.9:0.1), rt, 3 h, quantitative; (h) BODIPY-X-630/650-OSu, DMF, rt,
4 h, 75-85%.
characterization of a novel fluorescent antagonist, XAC-BY630
((E)-N-(2-(2-(4-(2,6-dioxo-1,3-dipropyl-2,3,6,7-tetrahydro-1H-
purin-8-yl)phenoxy)acetamido)ethyl)-6-(2-(4-(2-(4,4-difluoro-
4,4a-dihydro-5-(thiophen-2-yl)-4-bora-3a,4a-diaza-s-indacene-
3-yl)vinyl)phenoxy)acetamido)hexanamide), for the A1-AR.10
This compound retained the antagonist properties of xanthine
amine congener (XAC), while also allowing visualization of
the A1-AR and detection of antagonist-receptor complexes in
small areas of cell membrane using FCS. It was our intention,
therefore, to complement this approach by developing fluores-
cent ligands for the A1-AR receptor that possess both appropriate
fluorescent properties for FCS but also have agonist activity at
the receptor.
A large proportion of fluorescent agonists for GPCRs thus
far described are peptide-based.11 For these ligands, the
conjugated fluorophore constitutes a relatively small component
of the overall molecule and can be positioned such that it is
well separated from the pharmacophore. However, in the case
of GPCRs activated by biogenic amines, the ligand binding sites
are buried within the transmembrane regions of the receptor
and interact with much smaller molecules. Consequently,
appropriate positioning of the fluorophore within the final
conjugate is critical to retain both receptor binding affinity and
efficacy. Given the high fidelity of ligand-receptor interactions,
it is imperative that new fluorescent derivatives are characterized
as novel pharmacological entities in their own right rather than
assuming that they retain the properties of the parent ligand.12
Our goal to generate fluorescent agonists for the human A1-
AR initiated a review of the structure-activity relationships of
the native ligand adenosine and its derivatives. For this receptor,
modifications to the N6-position of adenosine and the structurally
similar agonist adenosine 5′-N-ethyl carboxamide (NECA) are
well tolerated, yielding ligands that retain biological activity.13
This allows insertion of N6-aminoalkyl spacers of various chain
lengths to yield the corresponding functionalized congeners that
are amenable to fluorescent labeling. Examples of fluorescently
labeled derivatives of agonists targeted at the A1-AR have been
described using this approach, but their agonist activity, to the
best of our knowledge, is yet to be reported. Additionally, the
UV-excited fluorophores employed in these studies are unsuit-
able for use with FCS.14 In comparison, the use of a fluorophore
that possesses a longer excitation wavelength will help to
minimize autofluoresence and maximize flexibility for use in
multicolor applications with fluorescent protein labels.
In this paper, we describe the synthesis of new red-emitting
fluorescent agonists for the A1-AR. We show that they are potent
and efficacious as agonists, and we demonstrate their use in
monitoring and quantifying agonist-receptor interactions in
single living cells. Specifically, the congeners belong to a series
of N6-aminoalkyl NECAs and an N6-aminoalkyladenosine.
Results and Discussion
Synthesis. Following the general theme of Olsson,13a inosine
1 was protected as the corresponding 2′,3′-isopropylidene ketal
with 2,3-dimethoxypropane and tosic acid (Scheme 1). Crystal-
lization of ketal 2 from water provided a convenient purification
and proved to be a key determinant in its subsequent efficient
oxidation to 3. While most nucleoside 5′-hydroxymethyl oxida-
tions of this type have relied on metal-based reagents (Mn, Cr),
Epp and Widlanski have reported a facile preparation utilizing
the in situ generation of the N-oxoammonium salt derived from
TEMPO.15
Under these conditions and with high concentrations of
water,16 aliphatic alcohols are converted to their respective
carboxylic acids, and the oxidation of 2′,3′-isopropylidene-
nucleosides (adenosine, uridine, cytidine, and guanosine) was
reported in high yield.15 Furthermore, the reaction generates only
acetic acid and iodobenzene as byproducts, which are easily
removed from the precipitated product by trituration to yield
the acids in analytical purity.
Initial attempts of the synthesis of acid 3 from noncrystalline
2 resulted in a successful transformation but required long
reaction times and afforded lower purities than anticipated.15
In contrast to the previously investigated nucleosides, addition
of base or lowering of the reaction temperature did not elicit a
substantial improvement.15 However, substituting crystalline 2
into the reaction afforded acid 3 in 2 h in high yield and
analytical purity after trituration of the crude product with
acetone and diethyl ether. To the best of our knowledge, this
constitutes the first extension of this methodology to inosine
nucleosides and represents a major improvement in the chemical

784 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4
Middleton et al.
Scheme 2a
a Reagents and conditions: (a) Ac2O, pyridine, 40 °C, 1 h, 97%; (b) POCl3, N,N-dimethylaniline, reflux, 5 min, 85%; (c) (i) 5b, DIEA, EtOH reflux, 18
h; (ii) NH3/MeOH, 0 °C, 2 h, 66%; (d) H2/Pd/C, MeOH/water/AcOH (9.0:0.9:0.1), rt, 3 h, quantitative; (g) BODIPY-X-630/650-OSu, N,N-DMF, rt, 4 h,
75%.
repertoire available for the synthesis of adenosine derivatives.
Indeed, high-yielding, multigram preparations of 3 are now
routinely carried out in our laboratories without the need for
chromatographic or laborious purifications.
Removal of the ketal protecting group was effected with 1
N HCl in aqueous dioxane at 70 °C in moderate to good yields,
with diols 7a-d readily purified by preparative layer chroma-
tography. Hydrogenolysis of the N-benzyloxycarbonyl group
effected a near-quantitative recovery of amines 8a-d in
sufficient purity to be labeled directly. Surprisingly, initial
attempts at the Z-deprotection yielded N-methyl derivatives of
the desired primary amines 8a-d, which were observed by mass
spectrometry. Similar alkylations have been described in the
literature and arise from the conversion of methanol to
formaldehyde in the presence of the palladium catalyst and
subsequent reductive amination under the prevailing reducing
conditions.17 This alkylation was easily suppressed by inclusion
of aqueous acid in the solvent mixture.
Conjugation of an excess of the congeners 8a-d with the
commercially available amine-reactive fluorophore 6-(((4,4-
difluoro-5-(2-thienyl)-4-bora-3a,4a-diaza-s-indacene-3-yl)styry-
loxy)acetyl)amino-hexanoic acid, succinimidyl ester (BODIPY
630/650-X, SE; Molecular Probes) proceeded smoothly, and the
fluorescently labeled ligands 9a-d were isolated by preparative
layer chromatography and their purity was confirmed by RP-
HPLC with photodiode-array detection from 190 to 800 nm.
The synthesis of the N6-aminobutyladenosine derivative 14
followed a similar synthetic strategy (Scheme 2). The potential
for this ligand to interact with other nonreceptor adenosine
binding sites (e.g., adenosine transporters) is likely to be higher
than the corresponding uronamide derivatives; 14 was synthe-
sized, however, to test whether there was a benefit in terms of
increased potency or efficacy (cf. the selectivity of N6-
cyclopentyladenosine (CPA) for the A1-AR).18 Peracetylation
of inosine under standard conditions was followed by chlorina-
tion at C6 using POCl3. Displacement of the resultant 6-chloro
group with the mono-Z-protected diamine 5b also effected
partial deacetylation via simple O f N acyl transfer, with the
nucleoside mono- and diacetates being clearly distinguished by
thin layer chromatography (TLC). Global acetate deprotection
was therefore effected using an alcoholic solution of ammonia
to afford N6-(benzyloxycarbonylaminobutyl)adenosine 12 in
good yield. Hydrogenolysis and fluorophore conjugation fol-
lowed without incident, providing the fluorescent derivative 14
after preparative layer chromatography.
For the introduction of C5′- and N6-substituents, there exist
a number of reagents that may be used sequentially to effect
the required chlorination and carboxyl activation/amidation
reactions. We chose, however, to attempt simultaneous chlorina-
tion at these positions and exploit the susceptibility of the
resultant acyl and aryl halides toward regioselective amine
displacement under temperature control. This approach, as
described by Olsson et al., using DMF/SOCl2,13a has been both
reproduced13b,c,e,14a,b and described as not favorable under the
reported reaction conditions.13d Following minor experimental
modifications, we were able to isolate the uronamide derivative
4 in good yield following low temperature in situ ethylamine-
mediated aminolysis of the C4′-acid chloride. The reaction
remains highly sensitive to the quality of reagents employed,
however, and demands carefully determined reaction conditions
in return for a high-yielding transformation.
Displacement of the 6-chloro group with a variety of
monoprotected alkyl diamines 5a-d furnished the N6-ami-
noalkyl NECAs 6a-d thus introducing the linker unit and
chemical functionality for conjugation with the fluorophore. The
mono-Z-protected diamines 5a-d are available via the reaction
of a large excess of the corresponding diamines with benzyl
chloroformate under pseudo-dilution conditions. An aqueous
acid extraction effects good separation of the desired linkers
5a-d from the excess of diamine and bis-Z-protected byproduct.
It was our objective at this stage to remove all protecting
groups prior to conjugation of the ligands with the fluorophore;
this strategy has two distinct advantages. Due to the high cost
of the commercially available fluorophores and, hence, the
micromole conjugation reactions commonly employed, it is
desirable to conjugate as the last synthetic step of a sequence.
This reduces the synthetic loss of valuable intermediates and
eliminates exposure of the often-fragile fluorophores to the
rigorous conditions of chemical deprotections. Second, fully
deprotected congeners present the opportunity for the facile,
single-step synthesis of variously labeled ligands bearing a range
of fluorophores or other “tagging” moieties. We envisaged initial
ketal deprotection more appropriate, leaving the protected amine
of 7a-d as a suitable handle to assist purification of the resultant
diol.
We have previously labeled a small molecule GPCR antago-
nist with the commercially available fluorophore BODIPY [630/
650] and demonstrated its suitability for use in FCS.10 This

Fluorescent Adenosine A1-Receptor Agonists
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4 785
Figure 1. Radioligand binding and functional analysis of fluorescent adenosine derivatives in CHO-A1H cells. (a) Inhibition of [3H]-cAMP
accumulation by 9a-d. Stimulation of [3H]-cAMP accumulation (from [3H]-adenine) is expressed as a percentage of that obtained with 3 µM
forskolin. Each point represents mean ( s.e. mean obtained from three independent experiments, performed in triplicate with 9a (0), 9b ((), 9c
(b), and 9d (2). (b) Competition radioligand binding data in CHO-A1H cells using [3H]-DPCPX as radioligand. Total binding is shown in the
filled bar and nonspecific binding (in the presence of 10 µM XAC) is shown in the open bar. Data are shown as mean ( s.e. mean triplicates within
an experiment representative of eight (9a-d) or six (14) performed. The key to symbols is as described for (a) and 14 (9). (c) Competition
radioligand binding data in intact CHO-A1H cells for XAC (b) and 9b (O) using [3H]-DPCPX as radioligand. Total and nonspecific binding are
illustrated as in (b). Data are shown as mean ( s.e. mean triplicates within an experiment representative of four performed. (d) Experiments
performed as in (c) but with cells permeabilized with 0.01% Triton-X100.
Table 1. Concentration-Response and Ligand-Binding Parameters for Fluorescent Agonists Acting at the Human A1-Adenosine Receptora
[3H]-cAMP accumulation
[3H]-DPCPX binding
-log EC50
(A)
% maximal inhibition
of forskolin response
-log KD
(B)
% maximal inhibition
of specific binding
ligand
n
n
A-B
NECA
9a
9b
9c
9d
9.44b
98b
6.24 ( 0.04
6.61 ( 0.10
83.3 ( 1.2
50.9 ( 2.4
8
8
8
8
8
6
3.20
2.55
1.67
9.16 ( 0.08
8.47 ( 0.08
9.15 ( 0.12
7.92 ( 0.15
8.29 ( 0.05
92 ( 5%
99 ( 1%
96 ( 3%
77 ( 4%
96 ( 2%
4
3
3
4
3
6.80 ( 0.11
49.3 ( 1.7
c
c
c
c
14
6.65 ( 0.09
48.1 ( 3.9
1.64
a Concentration-response parameters were determined from inhibition of forskolin-stimulated (3 µM) [3H]-cAMP accumulation. Ligand-binding parameters
were determined from inhibition of the binding of 1 nM [3H]-DPCPX. Values are mean ( s.e. mean from n separate experiments. b The values for NECA
were taken from ref 20. c The inhibition curves for 9c and 9d were insufficiently defined at concentrations up to 10-5 M to determine ligand-binding
parameters.
fluorescent moiety is bright, photostable, and has low triplet-
state population.19 In addition, cells do not exhibit high levels
of autofluorescence or suffer from significant phototoxicity when
exposed to its peak excitation wavelength. To ensure that
conjugation of the fluorophore to the precursor molecules had
not affected its photochemical properties, we measured the
excitation/emission maxima of the five conjugates 9a-d and
14. None of these values differed significantly from that of the
free fluorophore (637/658 nm; data not shown).
determine the potency and efficacy of the fluorescent agonists
9a-d and 14 at the human A1-AR, we evaluated their ability
to inhibit forskolin-stimulated cAMP accumulation in CHO cells
expressing this receptor (CHO-A1H cells), as described previ-
ously.20 All five compounds produced a concentration-dependent
inhibition of cAMP accumulation with an EC50 in the nanomolar
range (Figure 1, Table 1).21 The striking feature of these data
was that all molecules displayed agonist activity. Apart from
9d, all compounds produced over 90% inhibition of the forskolin
response. This was similar to that previously obtained with
NECA.20 However, 9d appeared to be a lower efficacy agonist
in this system and interestingly was the least potent. To
demonstrate that these responses were mediated via the A1-
Pharmacology
Inhibition of [3H]-cAMP Accumulation and [3H]-DPCPX
Binding in Chinese Hamster Ovary (CHO)-A1H Cells. To

786 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4
Middleton et al.
Table 2. -Log EC50 Values for NECA or 9b Stimulation of
[3H]-Cyclic AMP Accumulation (A2A, A2B) and Inhibition of
Forskolin-Stimulated (3 µM) Cyclic AMP Accumulation (A1, A3) in
CHO K1 Cells Expressing the Human A1, A2A, A2B, or A3 Adenosine
Receptora
the linker (Table 1). This is likely to be related to agonist
efficacy rather than affinity because the ligand binding affinities
were quite similar for 9a, 9b, and 14. To gain some insight
into this, the difference between the -log EC50 and -log KD
values have been presented in Table 1 as an indicator of agonist
efficacy (on a logarithmic scale). It is notable that 9b and 14
share very similar efficacies and potencies on this basis.
-Log EC50 (n)
adenosine receptor
NECA
9.44b
7.76 ( 0.02 (3)
5.44 ( 0.22 (3)
8.34 ( 0.12 (4)
9b
A1
8.47 ( 0.08 (3)
6.76 ( 0.06 (3)
5.69 ( 0.18 (3)
8.57 ( 0.22 (5)
Confocal Imaging of Ligand Binding. Having demonstrated
the high potency of 9a-d in functional assays, we were able
to use these compounds to visualize ligand-receptor interactions
using confocal microscopy (Figure 2a). These experiments were
performed at 22 °C to minimize receptor internalization and
thus maximize the signal seen on the cell membrane. Incubation
of CHO-A1H cells with 9a-d resulted in a clear labeling of
the cell membrane, which was time- and concentration-
dependent. When added at 100 nM, membrane binding was seen
as little as 5 min after addition of the drug and did not increase
substantially after this time (data not shown). The degree of
membrane binding seen following addition of each compound
(100 nM, 10 min) was dependent on the length of the linker,
with 9b and 9c showing the highest and 9d showing the lowest
level of specific binding (Figure 2a). None of the compounds
showed any substantial intracellular binding at the times and
concentrations tested. In each case, a substantial proportion of
this binding was prevented by preincubation with DPCPX (100
nM, 30 min, 37 °C) and, hence, was specific binding to the
A1-AR. The binding of 9a-d (100 nM, 10 min) to the A1-AR
was investigated further using CHO cells expressing the A1-
AR fused on its C-terminus to the Topaz variant of green
fluorescent protein.10 Again, for each compound, significant
membrane binding was seen, and this binding co-localized with
the fluorescent membrane-localized receptor. Preincubation of
cells with DPCPX (100 nM) caused a significant decrease in
both overall membrane binding and, more specifically, a large
decrease in the co-localization seen between ligand and receptor
(illustrated for 9b in Figure 2b).
A2A
A2B
A3
a Values represent mean ( s.e. mean from n separate experiments, each
performed in triplicate. b The value for NECA was taken from ref 20.
AR, parallel experiments were performed with 9b and 14
following preincubation of the cells with the selective A1-AR
antagonist 1,3-di-n-propyl-8-cyclopentylxanthine (DPCPX; pKB
) 7.85 ( 0.01 and 7.91 ( 0.06, n ) 3, respectively).21 The
affinity values obtained from these functional assays for DPCPX
are consistent with those previously demonstrated in this cell
line.20 It was evident from these data that 14 displayed similar
agonist potency and efficacy when compared to the correspond-
ing uronamide derivative 9b. As a consequence, further studies
on variations in linker length were undertaken solely on the
uronamide series.
To gain some insight into the selectivity of this series of
compounds for the different adenosine receptors, we evaluated
the ability of 9b to stimulate cAMP accumulation via adenosine-
A2A and -A2B receptors expressed in CHO-K1 cells and to inhibit
forskolin-stimulated cAMP via adenosine-A3 receptors expressed
in the same cell background (Table 2). These data show that
9b has a very similar potency on Gi-coupled A3- and A1-
receptors, but is much less potent as an agonist of the Gs-coupled
A
2A- and A2B-receptors.
Whole cell binding studies with [3H]-DPCPX in CHO-A1H
cells yielded a KD value for [3H]-DPCPX of 1.92 ( 0.23 nM
(n ) 6). NECA inhibited the specific binding (defined as that
displaceable by 10 µM XAC) of [3H]-DPCPX by 83.3 ( 1.5%
to yield a -log KD value of 6.24 ( 0.04 (n ) 8; Table 1).
Compounds 9a, 9b, and 14 inhibited [3H]-DPCPX binding to
reach a clear plateau at about 50% of the specific binding
displaced by XAC (Figure 1b; Table 1). Estimation of the -log
KD values for this component of binding indicated that these
three ligands showed similar affinities for the A1-receptor.
Unfortunately, there was not sufficient displacement of [3H]-
DPCPX binding at concentrations of 9c and 9d below 10 µM
to allow confident estimation of ligand-binding parameters for
these two compounds (Figure 1b). A potential explanation for
the partial displacement of specific [3H]-DPCPX binding by 9a,
9b, and 14 is that both [3H]-DPCPX and XAC are able to cross
the cell membrane and bind to intracellular A1-receptors,
whereas 9a, 9b, and 14 are relatively less membrane permeable.
To test this hypothesis, we directly compared the degree of
maximum displacement of specific [3H]-DPCPX binding ob-
tained in intact CHO-A1H cells with that achieved in CHO-
A1H cells that had been permeabilized with 0.01% Triton-X100
(Figure 1c,d). These data confirmed that the maximum displace-
ment of specific [3H]-DPCPX binding (defined with XAC)
obtained with 9b increased from 49.3 ( 1.7 to 75.6 ( 2.5%
(Figure 1c,d) with no significant change in the apparent -log
Kd values (6.80 ( 0.11 and 7.13 ( 0.08; n ) 8 and 4,
respectively).
Quantification of Ligand Binding Using FCS. To quantify
the binding of these ligands to the A1-AR in nonspecified
membrane microdomains of single cells, we used FCS. This
technique analyzes the fluctuation in fluorescence signal as a
fluorescent species diffuses through a small, defined confocal
detection volume. Autocorrelation analysis of these fluctuations
provides information on both the diffusion characteristics of the
molecule (related to mass) and also the average number of
molecules in the detection volume during the measurement (i.e.,
concentration). For the purposes of detecting ligand-receptor
interactions, the detection volume can be positioned on the upper
membrane of the cell. Here fluctuations from both fast-moving
free-ligand molecules and slower-moving receptor-bound mol-
ecules will be detected simultaneously and their concentrations
calculated. Thus, both (actual) free and bound ligand concentra-
tions can be quantified in a small (∼0.1 µm2) area of
membrane.5b Compounds 9a-d all proved suitable for FCS
measurements in aqueous solution, showing monophasic auto-
correlation curves, with diffusion times of 69 ( 3 µs, 71 ( 3
µs, 72 ( 4 µs, and 74 ( 4 µs, respectively (n ) 4; data not
shown). All had triplet state populations of less than 7%,
equivalent to the free fluorophore. FCS measurements were also
taken on the upper membrane of CHO-A1H cells which had
been incubated with 9b as previously described.22 Following
positioning in x and y planes, the excitation volume was
positioned in the upper membrane using intensity scanning in
the z plane, where the peak corresponding to the upper
membrane was clearly visible (Figure 3a).
It was evident in the functional cAMP assays that the
structure-activity relationship for the uronamide series dem-
onstrated a dependence of potency on the odd/even nature of

Fluorescent Adenosine A1-Receptor Agonists
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4 787
Figure 2. Visualizing fluorescent agonist binding using live cell confocal microscopy. (a) Top panel: CHO-A1H cells were incubated with fluorescent
ligand (100 nM, 10 min, 22 °C) and single equatorial confocal images taken as previously described.10 Lower panel: Similar experiments were
performed on cells preincubated with the nonfluorescent A1-AR antagonist, DPCPX (100 nM, 30 min, 37 °C). Images shown are from a single
experiment within which identical settings for laser power, offset, and gain were applied. These data are representative of five (upper) or four
(lower) experiments performed. (b) Experiments as above were performed using CHO cells expressing an A1-AR fused on its C-terminus to Topaz
fluorescent protein, with images obtained by simultaneous excitation with 488 and 633 nm, as previously described.10 Ligand (red), receptor (green),
and overlay images are shown, with yellow indicating co-localized pixels. For ease of viewing, these pixels are indicated in black in the final panel.
The experiment shown is representative of three performed.
When positioned 2 µm above the cell membrane, the
autocorrelation curve was monophasic, with a diffusion time
of 67 ( 1 µs (n ) 16), equivalent to that seen for this compound
in aqueous solution (Figure 3b). However, positioning of the
volume on the upper cell membrane results in an autocorrelation
curve containing two further slow-diffusing components, which
over a range of concentrations of 9b (2-20 nM) were 9.5 (
0.3 ms (τD2) and 267 ( 27 ms (τD3; n ) 47; Figure 3c). Similar
diffusion components were identified when using a fluorescent
antagonist ligand for the A1-AR10 and also for 9a, c, and d (data
not shown). To determine whether these diffusion components
represented receptor-ligand complexes, measurements were
made on cells exposed to 9b (5 nM) in the presence or absence
of DPCPX (100 nM, 30 min, 37 °C) and the amount of τD2 and
ments in CHO-A2B cells (Figure 4a) demonstrated that, at this
concentration of 9b, there was no significant binding to the A2B-
AR in CHO cell membranes. Therefore, using the low concen-
trations of ligand required for FCS, it should be achievable to
selectively measure the binding to the A1-AR in cells endog-
enously expressing the A2B-AR. Binding of 9b to the CHO-A1
cells consisted of approximately 40% τD2 and 60% τD3 (Figure
4b). DPCPX significantly reduced the amount of both of these
binding components, suggesting that both τD2 and τD3 represent
the diffusion of ligand-receptor complexes. This is consistent
with previous reports of agonist binding using peptide ligands.8,23
Diffusion times for τD2 and τD3 were slightly faster in the cells,
which had been pretreated with DPCPX (τD2 ) 8.1 ( 0.2 and
7.1 ( 0.2 ms and τD3 ) 243 ( 10 and 151 ( 7 ms, in control
and DPCPX-treated cells, P < 0.05 and P < 0.01, respectively).
Though the exact nature of these species is yet to be determined,
they may represent, for instance, diffusion of the receptor within
different membrane microdomains or the diffusion of the
receptor superimposed on that of a microdomain.
τD3 quantified. As shown in Figure 4a, the level of total binding
on CHO-A1H cell membranes was significantly greater than
that seen in native CHO-K1 cells. This binding was to the A1-
AR because it was significantly inhibited by preincubation with
a selective concentration of DPCPX (100 nM). Similar experi-

788 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4
Middleton et al.
Figure 3. Monitoring binding of 9b (5 nM) to CHO-A1H cell membranes using FCS. (a) The confocal detection volume was positioned in x and
y using a live epi-fluorescent image of the cells (white cross). Intensity scanning in z allowed positioning of the volume on (UM) or 2 µm above
(EC) the intensity peaks corresponding to the upper cell membrane. (b) Fluorescence fluctuations detected in the EC position and their subsequent
autocorrelation analysis. Curve fitting indicated these data to show a monophasic decay curve with τD1 ) 60 µs (100%). Residuals of the fit are
shown in the lower graph. (c) Equivalent analysis with data collected on the UM position. The curve was best fit with a three component model
showing τD1 ) 60 µs (35%), τD2 ) 8.6 ms (35%), and τD3 ) 109 ms (30%). Data are from an individual cell representative of (b) 16 and (c) 47
experiments performed.
BODIPY 630/650-X, SE (Molecular Probes). No protecting
group manipulations were required following installation of the
fluorophore thereby obviating any deleterious side-reactions or
loss of yield that may result from the often harsh deprotection
conditions employed in chemical synthesis. Mindful of the
potential effects the fluorophore may exert on the interactions
of these ligands with the A1-AR, coupled with the fact that the
binding site for adenosine is buried within the transmembrane
domain of the GPCR, an evaluation of the pharmacology of
the new fluorescent derivatives was undertaken. All of the
molecules displayed potent agonist activity, despite a tripling
of molecular weight when compared to the native receptor
agonist. We were further able to utilize these molecules to
visualize ligand-receptor interactions using confocal micros-
copy. Incubation of CHO cells expressing the A1-AR with the
fluorescent agonists resulted in clear labeling of the cell
membrane, with a large proportion of this binding prevented
by preincubation with a nonfluorescent A1-AR antagonist,
DPCPX. The highest level of membrane binding was displayed
by the modified NECA ligands containing a four- or five-carbon
linker. Specific labeling of the A1-AR was further confirmed
using a CHO cell line expressing a C-terminal GFP-fused A1-
AR. In this example, clear membrane binding of 9b was
Figure 4. Quantifying binding of 9b to CHO cell membranes using
observed, which co-localized with the membrane fraction of
the fluorescent receptor. Quantification of agonist binding in a
small (∼0.1 µm2) area of cell membrane was achieved using
the noninvasive technique of FCS. This revealed the existence
of at least two types of agonist-receptor complexes with
differing diffusion characteristics. Further investigations will
concentrate on identifying the nature of these complexes, which
may give us further insight into the macrostructure of the A1-
AR and its associated signaling molecules. It will allow us to
localize these measurements to specific membrane domains and
determine the pharmacology of the A1-AR subpopulations within
them.
FCS. (a) CHO-A1H, CHO-A2B, or native CHO-K1 cells were
incubated with 9b (5 nM, 10 min) in the absence (Ctrl) or following a
30 min preincubation with antagonist (100 nM DPCPX (+DP) for
CHO-A1H cells; 1 µM XAC (+XAC) for CHO-A2B cells). FCS
measurements were then taken on the upper cell membrane, and the
total binding was quantified. Results are presented as mean ( s.e. mean,
with the number of cells indicated above the column in parenthesis.
*P < 0.001 vs CHO-K1 binding †P < 0.001 vs control for CHO-A1H
cells. (b) The results from CHO-A1H cells in (a), indicating the relative
levels of the τD2 and τD3 components and their sensitivity to preincu-
bation with DPCPX. **P < 0.01 vs control.
Conclusions
We have successfully synthesized and characterized five new
red-emitting fluorescent agonists, 9a-d and 14, for the human
adenosine A1-receptor (A1-AR). Installation of a range of
alkylamine linkers at the N6-position of either adenosine or
NECA generated a range of fully deprotected congeners,
functionally equipped to undergo a high yielding chemoselective
acylation with the commercially available fluorochrome
Experimental Section
General Chemistry Methods: Chemicals and solvents (HPLC
grade) were purchased from the standard suppliers and were used
without further purification. BODIPY fluorophores were obtained
from Molecular Probes, Cambridge Bioscience, Cambridge, U.K.
Merck Kieselgel 60, 230-400 mesh, for flash chromatography was

Fluorescent Adenosine A1-Receptor Agonists
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4 789
supplied by Merck KgaA (Darmstadt, Germany), and deuterated
solvents were purchased from Goss International Limited (England).
Unless otherwise stated, reactions were carried out at ambient
temperature. Reactions were monitored by TLC on commercially
available precoated aluminum-backed plates (Merck Kieselgel 60
CH3), 4.67 (1H, d, J ) 1.7, C4′H) 5.36-5.42 (2H, m, C2′H and
C3′H), 6.29 (1H, s, C1′H), 8.02 (1H, s, adenine CH), 8.29 (1H, s,
adenine CH), 12.41 (1H, br s, NH); δC (100.2 MHz, DMSO-d6)
25.3, 26.9 (2 × CH3), 84.0, 84.2 (C2′, C3′), 85.9 (C4′), 90.1 (C1′),
113.1 (ketal 4°), 124.7 (4°), 140.3 (adenine CH), 145.8 (adenine
CH), 148.5 (4°), 157.7 (4°), 171.2 (COOH); Anal. (C13H14N4O6)
C, H, N; m/z (ES+) 323 (MH)+, 137 (M-ribose)+.
F254). Visualization was by examination under UV light (254 and
366 nm). General staining employed KMnO4. A solution of
ninhydrin (EtOH) was used for the visualization of primary amines.
All organic extracts collected after aqueous workup procedures were
dried over anhydrous magnesium sulfate, filtered under gravity,
and evaporated to dryness. Organic solvents were evaporated in
vacuo at a temperature of <35 °C (water bath). Purification using
preparative layer chromatography was carried out using Fluka silica
gel GF254 on glass plates (200 mm × 200 mm × 1 mm).
6-Chloro-6-deoxy-5′-ethylamino-2′,3′-isopropylidene-5′-oxo-
5′-deoxyinosine (4): Under rigorously dry conditions and an inert
atmosphere, acid 3 (967 mg, 3 mmol) was suspended in CHCl3
(15 mL) to which was added N,N-DMF (581 µL, 7.5 mmol) and
SOCl2 (1.09 mL, 15 mmol). The suspension was placed in a hot
oil bath and maintained at a gentle reflux for 5 h. The resultant
solution was evaporated, and the yellow oil was dissolved in THF
(20 mL) at 5 °C. Ethylamine (2.0 M solution in THF, 3.75 mL,
7.5 mmol) was added dropwise, stirred at 5 °C for 15 min, and
allowed to warm to room temperature. The solvent was evaporated,
and the residue was dissolved in DCM (25 mL) and washed with
water (2 × 20 mL) and saturated brine solution (2 × 20 mL). The
organic fraction was dried and evaporated to leave a yellow solid
that was purified by column chromatography on silica (5% MeOH-
DCM) to give the title compound 4 (941 mg, 85%) as a yellow
Melting points were recorded on a Gallenkamp 3A 3790
apparatus and are uncorrected. FT-infrared spectra were recorded
as thin films or KBr discs in the range of 4000-600 cm-1 using
an Avatar 360 Nicolet FT-IR spectrophotometer. Optical rotation
was measured on a Bellingham-Stanley ADP220 polarimeter.
Lyophilization, from either water or 1,4-dioxane, was carried out
on an Edwards Modulyo FD freeze drier. Mass spectra ((ES/AP)
were recorded by Dr Cath Ortori, School of Pharmacy, University
of Nottingham, on a Micromass Platform instrument, using either
direct injection or liquid chromatography-mass spectrometry (LC-
MS). HRMS TOF-ES mass spectra were recorded on a Waters 2795
Separation Module/Micromass LCT platform. Microanalytical data
were recorded by Mr. T. Spencer, School of Chemistry, University
of Nottingham. Proton nuclear magnetic resonance spectra were
recorded on a Bruker-AMX 250 (250 MHz) or a Bruker-AV 400
(400 MHz) spectrometer. Carbon nuclear magnetic resonance
spectra were recorded at 62.9 and 100.6 MHz, respectively. Unless
otherwise stated, spectra were recorded in CDCl3. Chemical shifts
(δ) are recorded in ppm with reference to the chemical shift of the
deuterated solvent/an internal tetramethylsilane standard. Coupling
constants (J) are recorded in hertz, and the signal multiplicities are
described by the following: s, singlet; d, doublet; t, triplet; q,
quartet; br, broad; m, multiplet; dd, doublet of doublets. Spectra
were assigned using appropriate COSY, DEPT, and HMQC
sequences. Analytical reverse-phase high performance liquid chro-
matography (RP-HPLC) was performed on a Waters Millennium
995 LC system using columns, gradients, and flow rates as
described. The eluent was monitored using photodiode array
detection. Mobile phases were solvent A, water, and solvent B,
acetonitrile, degassed by helium bubble and sonication, respectively.
2′,3′-Isopropylideneinosine (2): Inosine (5.36 g, 0.02 mol) and
tosic acid monohydrate (3.80 g, 0.02 mol) were suspended in a
mixture of 2,2-dimethoxypropane (50 mL) and acetone (200 mL)
and stirred for 18 h. Sodium hydrogen carbonate (2.52 g, 0.02 mol)
and water (40 mL) were added, and the suspension was stirred for
15 min. The suspension was evaporated to constant volume, and
the crude product was recrystallized from the residual water,
yielding the acetonide 2 (3.71 g, 60%) as white needles: mp 266-
19
solid: mp 94-96 °C; [R]D -12.9 (c 0.50 in CHCl3); δH (400
MHz) 0.80 (3H, t, J ) 7.3, CH2CH3), 1.41 (3H, s, CH3), 1.64 (3H,
s, CH3), 2.90-3.10 (2H, m, CH2CH3), 4.74 (1H, d, J ) 2.0, C4′H),
5.45 (1H, dd, J ) 6.2, 2.3, C2′H), 5.54 (1H, dd, J ) 6.2, 2.0, C3′H),
6.23 (1H, d, J ) 2.3, C1′H), 8.25 (1H, s, adenine CH), 8.75 (1H, s,
adenine CH); δC (100.2 MHz) 14.2 (CH2CH3), 25.0, 26.9 (2 ×
CH3), 33.9 (CH2CH3), 82.9, 83.4 (C2′, C3′), 86.7 (C4′), 92.0 (C1′),
114.1 (ketal 4°), 132.3 (4°), 144.7 (adenine CH), 150.9 (4°), 151.9
(4°), 152.2 (adenine CH), 168.1 (CdO); m/z (ES-) 366 (M-H)-,
153 (M-ribose)-; m/z HRMS (TOF ES+) C15H19ClN5O4 (MH)+
calcd, 368.1126; found, 368.1104.
N1-Benzyloxycarbonyl-1,n-alkyldiamines (n ) 3, 4, 5, 8; 5a-
d):27 A solution of benzyl chloroformate (1 equiv) in CHCl3 was
added dropwise (3 h) to a stirred solution of 1,n-diaminoalkane
(5-10 equiv) in CHCl3 at 0 °C. The reaction was allowed to warm
to room temperature and stirred for 18 h. The reaction was
concentrated, and the residue was partitioned between EtOAc and
water. The organic layer was washed with water (×3), and the
combined aqueous phases were discarded. The organic layer was
washed with 2 M HCl(aq) (×3), and the aqueous layers were
combined, adjusted to pH 12 (NaOH), and saturated with NaCl.
Following extraction with EtOAc (×3), the combined organic
fractions were washed with brine (×2), dried, and evaporated to
yield the title compounds 5a-d (35-65%), which were used
without further purification.
N1-Benzyloxycarbonyl-1,4-butanediamine (5b):28 A pale yel-
low wax; δH (250 MHz) 1.48-1.33 (6H, m, C2H2, C3H2, NH2),
2.65 (2H, t, J ) 6.0, C4H2), 3.14 (2H, q, J ) 6.0, C1H2), 5.06 (2H,
s, benzyl CH2), 5.74 (1H, s, carbamate NH), 7.31 (5H, s, aromatics);
δC (69.2 MHz) 27.3 (CH2), 30.7 (CH2), 41.4 (CH2, C4), 41.7 (CH2,
C1), 66.3 (benzyl CH2), 128.0 (CH), 128.4 (CH), 136.8 (4°, ipso),
156.6 (4°, CdO); m/z (ES+) 223 (MH)+; FT-IR (thin film) 3327,
22
268 °C (lit. 266 °C;24 267 °C25); [R]D -67.1 (c 0.59 in MeOH)
20
(lit. [R]D -66.9 (c 0.8 in MeOH25)); δH (400 MHz, DMSO-d6)
1.32 (3H, s, CH3), 1.54 (3H, s, CH3), 3.56 (2H, m, C5′H2), 4.23
(1H, m, C4′H), 4.93 (1H, dd, J ) 6.1, 2.4, C3′H), 5.26 (1H, dd, J
) 6.1, 2.9, C2′H), 6.10 (1H, d, J ) 2.9, C1′H), 8.09 (1H, s, adenine
CH), 8.30 (1H, s, adenine CH), 12.53 (1H, br s, NH); δC (100.2
MHz, DMSO-d6) 25.1, 26.9 (2 × CH3), 61.4 (CH2), 81.2, 83.7,
86.6, 89.6 (C1′, C2′, C3′, C4′), 113.0 (ketal 4°), 125.4 (4°), 138.7
(adenine CH), 146.0 (adenine CH), 147.7 (4°), 156.5 (CdO); m/z
(ES+) 309 (MH)+, 137 (M-ribose)+.
3031, 2935, 2864, 1686, 1537, 1454, 1254, and 1027 cm-1
.
N6-(4-Benzyloxycarbonylaminobutyl)-5′-ethylamino-2′,3′-iso-
propylidene-5′-oxo-5′-deoxyadenosine (6b): Chloride 4 (300 mg,
0.82 mmol) was dissolved in ethanol (15 mL) to which 5b (503
mg, 2.46 mmol) and DIEA (0.17 mL, 0.98 mmol) was added. The
solution was maintained at reflux under an inert atmosphere for 20
h. The reaction mixture was concentrated, and the residue was
purified by column chromatography (10% MeOH-DCM) to yield
6b (125 mg, 69%) as a colorless oil/foam: δH (400 MHz) 0.80
(3H, t, J ) 7.2, CH2CH3), 1.31 (3H, s, CH3), 1.55 (3H, s, CH3),
1.58-1.66 (4H, m, 2 × CH2), 3.03 (2H, m, CH2CH3), 3.19 (2H,
m, CH2), 3.58 (2H, m, CH2) 4.62 (1H, d, J ) 1.3, C4′H), 5.02 (2H,
s, benzyl CH2), 5.25 (2H, m, C2′H and C3′H), 5.95 (1H, s, C1′H),
6.25 (1H, br s, NH), 7.07 (1H, br s, NH), 7.20 (5H, m, phenyl
CH), 7.72 (1H, s, adenine CH), 8.24 (1H, s, adenine CH); δc (100.2
MHz) 14.3 (CH2CH3), 25.1, 26.9 (2 × CH3), 27.1 (2 × CH2), 33.9
(CH2CH3), 40.0, 40.5 (2 × CH2), 66.7 (benzyl CH2), 82.4, 83.6
2′,3′-Isopropylidene-5′-oxoinosine (3): Acetonide 2 (3.08 g, 10
mmol), TEMPO (313 mg, 2 mmol), and iodosobenzene diacetate
(7.09 g, 22 mmol) were dissolved in MeCN/H2O (1:1, 50 mL) and
stirred, with the exclusion of light, for 4 h. The solvents were
carefully evaporated from the resultant suspension, and the reaction
residue was sequentially triturated with acetone and diethyl ether
to yield the acid 3 (2.67 g, 83%) as a white powder: mp 224-229
25
°C (lit. 272-274 °C;13a 252 °C26); [R]D -59.1 (c 0.66 in 1 M
NaOH); δH (400 MHz, DMSO-d6) 1.34 (3H, s, CH3), 1.52 (3H, s,

790 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4
Middleton et al.
(C2′, C3′), 85.7 (C4′), 92.0 (C1′), 114.5 (4°), 128.1 (4°), 128.5 (4°),
136.6 (4°), 139.2 (adenine CH), 153.3 (4°), 155.1 (4°), 156.6
(adenine CH), 168.7 (CdO); m/z HRMS (TOF ES+) C27H35N7O6
(MH)+ calcd, 554.2728; found, 554.2765.
MHz, DMSO-d6) 1.08, (3H, t, J 7.2, ethyl CH3), 1.19-1.26 (2H,
m, CH2), 1.40-1.62 (8H, m, 4 × CH2), 2.03 (2H, t, J ) 7.4, CH2),
3.02-3.27 (6H, m, 2 × CH2, ethyl CH2), 3.48 (2H, m, CH2), 4.12-
4.16 (1H, m, C3′H), 4.31 (1H, d, J ) 1.4, C4′H), 4.54 (2H, s,
BODIPY CH2), 4.59-4.64 (1H, m, C2′H), 5.54 (1H, d, J ) 6.5,
C2′-OH), 5.74 (1H, d, J ) 4.2, C3′-OH), 5.97 (1H, d, J ) 7.8, C1′H),
6.96 (1H, d, J ) 4.2), 7.08 (2H, d, J ) 8.8), 7.27-7.30 (3H, m),
7.38 (1H, d, J ) 4.4), 7.42 (1H, br s), 7.61 (1H, s), 7.62 (2H, d, J
) 8.7), 7.72-7.76 (2H, m), 7.83 (1H, dd, J ) 5.0, 0.9), 7.94-
8.01 (1H, m), 8.04 (1H, dd, J ) 3.8, 0.9), 8.12 (1H, t, J ) 5.7),
8.27 (1H, br s), 8.39 (1H, s, adenine CH), 8.91 (1H, t, J ) 5.6,
N6-(4-Benzyloxycarbonylaminobutyl)-5′-ethylamino-5′-oxo-5′-
deoxyadenosine (7b): Adenosine derivative 6b (261 mg, 0.47
mmol) was dissolved in 2 M HCl(aq)/1,4-dioxane (1:1, 4 mL), placed
in a 70 °C oil bath, and stirred for 1 h. The resultant solution was
adjusted to ∼pH 8 (satd NaHCO3(aq)), saturated with NaCl, and
extracted with EtOAc (3 × 5 mL). The combined organic fractions
were dried and evaporated, and the crude product was purified by
preparative layer chromatography (10% MeOH-DCM) to give the
title compound 7b (160 mg, 66%) as a white solid; mp 90-91 °C
(MeOH); δH (250 MHz, DMSO-d6) 1.08 (3H, t, J ) 7.2, CH2CH3),
1.45-1.62 (4H, m, 2 × CH2), 3.02 (2H, app q, CH2) 3.22 (2H,
app quintet, CH2CH3), 3.40-3.55 (2H, m, CH2), 4.14 (1H, m, C3′H),
4.31 (1H, d, J ) 1.1, C4′H), 4.61 (1H, m, C2′H), 4.99 (2H, s, benzyl
CH2), 5.56 (1H, d, J ) 6.5, C2′-OH), 5.76 (1H, d, J ) 4.2, C2′-
OH), 5.96 (1H, d J ) 7.6, C1′H), 7.26-7.34 (6H, m, phenyl CH
and carbamate NH), 8.01 (1H, br s, NH), 8.27 (1H, s, adenine CH),
8.39 (1H, s, adenine CH), 8.94 (1H, t, J ) 5.6, amide NH); δc
(69.2 MHz, DMSO-d6) 14.9 (CH2CH3), 26.6, 27.1 (2 × CH2), 33.4
(CH2CH3), 39.5, 40.3 (2 × CH2), 65.3 (benzyl CH2), 72.2 (C2′),
73.3 (C3′), 84.9 (C4′), 88.0 (C1′), 120.2 (4°), 127.9 (4°), 128.5 (phenyl
CH), 137.5 (adenine CH), 140.6 (4°), 148.4 (4°), 152.3 (adenine
CH), 154.9 (4°), 156.3 (4°), 169.3 (amide CdO); m/z HRMS (TOF
ES+) C24H31N7O6 (MH)+ calcd, 514.2415; found, 514.2371.
N6-(Aminoalkyl)-5′-ethylamino-5′-oxo-5′-deoxyadenosines (8a-
d): Adenosine derivatives 7a-d (50 mg) were dissolved in MeOH/
H2O/AcOH (9:0.9:0.1, 5 mL), to which was added 10% Pd/C (10
mg). The flask was evacuated, filled with hydrogen (balloon), and
stirred vigorously for 2-3 h. The reaction mixture was filtered
through celite, and the celite was washed with MeOH. The
combined organic filtrates were evaporated, and the resultant oil
was evaporated again from MeCN (2 × 15 mL) to give the title
compounds 8a-d (quantitative) as colorless oils/foams.
amide NH); Rt 13.61 min (Jones C8, 250 × 4.6 mm, 1 mL‚min-1
,
35-100% B, 30 min), Rt-2 9.31 min (Jones C4, 250 × 4.6 mm, 1
mL‚min-1, 35-100% B, 30 min; solvent A, H2O; solvent B,
MeCN); m/z HRMS (TOF ES+) C45H51BF2N10O7S (MH)+ calcd,
925.3803; found, 925.3842.
2′,3′,5′-Tri-O-acetylinosine (10): a solution of inosine (5.16 g,
19 mmol), pyridine (50 mL), and acetic anhydride (50 mL) was
heated at 50 °C for 12 h. Upon cooling, the solution was filtered,
and the filtrate was evaporated under high vacuum to leave a pale
yellow solid. The solid was triturated with diethyl ether and filtered
to yield the crude product as a white solid (7.37 g). Recrystallization
from methanol (150 mL) yielded the title compound 10 as a white
crystalline solid (6.11 g, 80%): mp 235 °C, MeOH (lit.30 241 °C);
[R]D24 -26.0 (c 1.0, MeOH) (lit.29 [R]D20 -36.5 (CHCl3)); δH (250
MHz) 2.10 (3H, s, OAc), 2.15 (3H, s, OAc), 2.16 (3H, s, OAc),
4.36-4.47 (3H, m, C4′H, C5′H), 5.54 (1H, t, J ) 5.4, C3′H), 5.90
(1H, t, J ) 5.6, C2′H), 6.18 (1H, d, J ) 5.5, C1′H), 8.03 (1H, s,
C8H), 8.32 (1H, s, C2H), 13.09 (1H, s, NH); δC (69.2 MHz) 20.4,
20.6, 20.8 (3 × OAc), 63.0 (C5′), 70.6, 73.4, 80.4 (C2′, C3′, C4′),
86.5 (C1′), 125.4 (4°), 138.6, 145.8 (C2′, C8′), 148.8 (4°), 159.0 (4°),
169.3, 169.6, 170.4 (3 × CO); FT-IR (KBr) 1748, 1702, 1588,
1376, 1227, 1053 cm-1; m/z (ES+) 395 (MH)+.
2′,3′,5′-Tri-O-acetyl-6-chloroinosine (11): A suspension of 10
(2.45 g, 6.22 mmol), N,N-dimethylaniline (0.83 mL, 6.53 mmol),
and phosphorus oxychloride (12.2 mL, 131 mmol) was stirred at
room temperature for 7 min under an atmosphere of nitrogen. The
flask was heated in a preheated oil bath and maintained at reflux
for 13 min. The solution was evaporated, and the resulting oil was
stirred in chloroform (20 mL) and ice (∼20 mL). The aqueous layer
was extracted with chloroform (2 × 25 mL). The combined organic
layers were washed with 2 M HCl (4 × 20 mL) and brine (2 × 20
mL), dried, and evaporated to yield 3 g of green oil. Purification
using silica chromatography (1:9 MeOH/DCM) yielded the title
N6-(4-Aminobutyl)-5′-ethylamino-5′-oxo-5′-deoxyadenosine
(ABEA; 8b): δH (400 MHz, MeOH-d4) 1.20 (3H, t, J ) 7.3,
CH2CH3), 1.75-1.81 (4H, m, 2 × CH2), 2.96-3.03 (2H, m, CH2),
3.36 (2H, q, J ) 7.3, CH2CH3), 3.66 (2H, m, CH2), 4.31 (1H, dd,
J ) 1.3, 4.7, C3′H), 4.47 (1H, d, J ) 1.3, C4′H), 4.74 (1H, dd, J )
4.8, 7.6, C2′H), 6.01 (1H, d, J ) 7.7, C1′H), 8.27 (2H, s, adenine
CH); δC (100.2 MHz, MeOH-d4) 15.1 (ethyl CH3), 22.6 (CH2),
25.9 (CH2), 27.6 (CH2), 35.1 (CH2), 40.4 (CH2), 73.5 (CH), 75.0
(CH), 86.4 (CH), 90.4 (CH), 121.7 (4°), 142.4 (CH), 149.3 (4°),
153.7 (CH), 156.4 (4°), 172.1 (4°); m/z HRMS (TOF ES+)
C16H25N7O4 (M + H)+ calcd, 380.2047; found, 380.2032.
Conjugation with BODIPY 630/650-X, SE: Ligands 8a-d
(15.1 µmol) were dissolved in N,N-DMF (1.0 mL) under an inert
atmosphere and with the exclusion of light. A solution of BODIPY
630/650-X, SE (Molecular Probes; 5.0 mg, 7.55 µmmol, 1 mL N,N-
DMF) was added, and the reaction was stirred for 4 h. The solution
was evaporated, and the crude product was purified by preparative
layer chromatography (10% MeOH-DCM) to give the title
compound 9a-d as purple powders (75-85%). When submitted
to analytical HPLC (Jones C8 column, [250 mm × 4.6 mm], 1
mL‚min-1, 35-100% B, 30 min, monitored using photodiode array
detection between 190 and 800 nm; mobile phases; solvent A, H2O;
solvent B, MeCN), all probes were observed to elute as single and
symmetrical peaks at the retention time Rt given below for each
compound. The purity of the ligands was confirmed by performing
a second analytical HPLC (Jones C4 column, [250 mm × 4.6 mm],
1 mL‚min-1, 35-100% B, 30 min, monitored using photodiode
array detection between 190 and 800 nm; mobile phases; solvent
A, H2O; solvent B, MeCN). Again, all target compounds eluted as
homogeneous single peaks. The identities of the compounds were
further analyzed by HRMS (TOF ES+).
24
compound 11 as a pale brown viscous oil (2.18 g, 85%): [R]D
-18.5 (c 1.3, MeOH); δH (250 MHz) 2.10 (3H, s, OAc), 2.13 (3H,
s, OAc), 2.17 (3H, s, OAc), 4.36-4.54 (3H, m, C4′H, C5′H), 5.67
(1H, t, J ) 5.2, C3′H), 5.97 (1H, t, J ) 5.3, C2′H), 6.27 (1H, d, J
) 5.5, C1′H), 8.37 (1H, s, adenine CH), 8.79 (1H, s, adenine CH);
δC (69.2 MHz) 20.1, 20.5, 20.8 (3 × OAc), 62.9 (C5′), 70.5, 73.1,
80.5, 86.9 (C1′, C2′, C3′, C4′), 132.3 (4°), 143.7 (adenine CH), 151.3
(4°), 151.6 (4°), 152.3 (adenine CH), 169.4, 169.6, 170.3 (3 × CO);
FT-IR (thin film) 1747, 1588, 1564, 1376, 1225, 1049 cm-1; m/z
(ES+) 413 (MH)+, 259 (M-purine)+.
N6-(Benzyloxycarbonylaminobutyl)adenosine (12): A solution
of chloride 11 (1.13 g, 2.7 mmol), diamine 5b (2.42 g, 10.9 mmol),
and DIEA (0.47 mL, 2.7 mmol) was refluxed in EtOH (40 mL)
for 18 h. The solution was evaporated, and the residue was dissolved
in an ice-cold, saturated methanolic solution of NH3. The solution
was stirred at 0 °C for 2 h, warmed to room temperature, and stirred
for an additional 2 h. The MeOH was evaporated to yield a solid.
Purification by silica chromatography (1:9 MeOH/DCM) yielded
1 g of yellow solid. Recrystallization from EtOH (30 mL) gave
the title compound 12 as a white solid (742 mg, 58%): mp 134
°C, EtOH; [R]D24 -48.0 (c 1.0, MeOH/water, 4:1); δH (250 MHz,
DMSO-d6) 1.40-1.60 (4H, m, 2 × CH2), 3.02 (2H, m, CH2), 3.37-
3.79 (4H, m, CH2, C5′H), 3.97 (1H, m, C4′H), 4.15 (1H, m, C3′H),
4.62 (1H, m, C2′H), 5.00 (2H, s, benzylic CH2), 5.22 (1H, d, J 4.5,
OH, exchanges D2O), 5.47 (2H, m, carbamate NH, OH, exchanges
D2O), 5.98 (1H, d, J ) 6.1, C1′H), 7.34 (5H, s, aromatics), 7.95
(1H, br s, NH, exchanges D2O), 8.21 (1H, s, purine C2H), 8.35
(2S,3S,4R,5R,E)-N-Ethyl-3,4-dihydroxy-5-(6-(4-(6-(2-(4-(2-
(4,4-difluoro-4,4a-dihydro-5-(thiophen-2-yl)-4-bora-3a,4a-diaza-
s-indacene-3-yl)vinyl)phenoxy)acetamido)hexanamido)butylami-
no)-9H-purin-9-yl)-tetrahydrofuran-2-carboxamide (9b): δH (400

Fluorescent Adenosine A1-Receptor Agonists
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4 791
(1H, s, purine C8H); δC (69.2 MHz, DMSO-d6) 26.6 (CH2), 27.1
(CH2), 39.5 (CH2), 41.0 (CH2), 61.9 (CH2, C5′), 65.3 (CH2,
benzylic), 70.9 (CH), 73.7 (CH), 86.1 (CH), 88.2 (CH), 120.0
(purine C5), 127.9 (CH, aromatic), 128.5 (CH, aromatic), 137.5 (4°,
ipso), 139.9 (purine C8), 148.0 (purine C6), 152.6 (purine C2), 154.9
(purine C4), 156.3 (carbamate CdO); m/z (ES+) 473 (MH)+, 342
(M-ribose)+; Anal. (C22H28O6N6‚0.5H2O) C, H, N.
incubated with 1 mL of Hank’s buffered salt solution (Sigma)
containing 20 mM HEPES (HHB) and [3H]adenine (Amersham
Biosciences, 37 kBq per well) for 2 h at 37 °C. Cells were washed
twice with HHB (1 mL) before incubation with the phosphodi-
esterase inhibitor rolipram (10 µM) for 15 min at 37 °C. Where
stated, cells were incubated with the antagonist DPCPX (100 nM).
In the case of CHO-A1H and CHO-A3 cells, agonists were added
for 5 min, prior to incubation with forskolin (3 µM) for an additional
10 min, in a final volume of 1 mL. For CHO-A2A and CHO-A2B
cells, agonists were added for 10 min in the absence of forskolin.
Incubations were terminated by the addition of concentrated HCl
(50 µL). [3H]-cAMP was isolated by single column alumina
chromatography, as previously described,31 and levels were deter-
mined by liquid scintillation counting.
N6-(Aminobutyl)adenosine (13): A suspension of 12 (250 mg,
0.53 mmol) and 50 mg of 10% Pd/C in 20 mL of solvent (9:1
MeOH/H2O with 1% AcOH) was stirred under hydrogen (balloon)
for 3 h. The reaction was filtered through a celite pad and was
concentrated in vacuo to yield 13 as a colorless oil. Under high
vacuum, the oil formed a white hygroscopic solid (219 mg,
quantitative): [R]D24 -21.0 (c 1.0, MeOH); δH (250 MHz, DMSO-
d6) 1.48-1.62 (4H, m, CH2), 1.79 (5H, br s, OH and NH2, reduces
D2O), 2.71 (2H, br t, CH2), 3.41 (2H, br s, CH2), 3.54 (1H, dd, J
) 12.0, 3.5, C5′aH), 3.67 (1H, dd, J ) 12.0, 3.5, C5′bH), 3.97 (1H,
m, C4′H), 4.16 (1H, m, C3′H), 4.62 (1H, t, J ) 5.2, C2′H), 5.90
(1H, d, J ) 6.1, C1′H), 7.95 (1H, br s, NH, exchanges D2O), 8.22
(1H, s, purine C2H), 8.38 (1H, s, purine C8H); δC (69.2 MHz,
DMSO-d6) 27.2 (CH2, br), 41.0 (CH2, br), 61.5 (CH2, C5′), 70.5
(CH), 73.5 (CH), 85.8 (CH), 87.8 (CH), 120.0 (purine C5), 139.5
(purine C8), 148.0 (purine C6), 152.2 (purine C2), 154.5 (purine
C4); m/z HRMS (FAB) C14H23N6O4 (MH)+ calcd, 339.1781; found,
339.1797.
Measurement of [3H]-DPCPX Binding to CHO-A1H Cells.
CHO-A1H cells were grown in white-sided 96-well view plates.
In saturation studies, once the cells were confluent, 100 µL of
serum-free media or serum-free media containing 20 µM XAC (to
define nonspecific binding) was added to each well, immediately
followed by 100 µL of serum-free media containing a range of
[3H]-DPCPX concentrations (0.004-38 nM). For competition
studies, the media was removed and replaced with 100 µL serum-
free media (i.e., DMEM/F12 containing 2 mM L-glutamine only)
containing the competing ligand at twice the final required
concentration. Immediately, 100 µL of serum-free media containing
[3H]-DPCPX was added to each well (hence, 1:2 dilutions in well)
to give final required concentrations. [3H]-DPCPX was used at a
concentrations of 1 nM. Total and nonspecific binding (as defined
by 10 µM XAC) were measured in every experiment. All plates
were incubated for 2 h at 37 °C in a humidified 5% CO2/95% air
atmosphere. After 2 h, media and drugs were removed, and the
cells were washed twice with 200 µL of cold PBS/well. A white
base was then added to the plate, followed by 100 µL of Microscint
20 per well, and a sealant film was placed over the wells. The plates
were then counted the following day on a Topcount (Packard) for
2 min per well at 21 °C. For studies in permeabilized cells,
incubation with ligands was performed in serum-free media
containing 0.01% Triton-X100 (under these conditions, trypan blue
enters all of the cells). At the end of the incubation period, the
cells and media from each well were transferred to a 96-well white
Millipore Multiscreen HTS FB filter plate, and the cells were
collected on the GF-B filter under vacuum. The filters were washed
twice with 200 µL of cold PBS/well, and scintillation counting was
performed as described above.
N-(4-(9-((2R,3R,4S,5R)-3,4-Dihydroxy-5-(hydroxymethyl)-tet-
rahydrofuran-2-yl)-9H-purin-6-ylamino)butyl)-6-(2-(2-(4,4-di-
fluoro-4,4a-dihydro-5-(thiophen-2-yl)-4-bora-3a,4a-diaza-s-in-
dacene-3-yl)vinyl)phenoxy)acetamido)hexanamide (14): A solution
of ABA 13 (4.9 mg, 14.5 µmol) and BODIPY 630/650-X, SE (5
mg, 7.6 µmol) was stirred in N,N-DMF (1 mL) in the dark, under
a nitrogen atmosphere, for 5 h. The DMF was removed in vacuo,
and the dark blue residue was purified by preparative layer
chromatography (15:85, MeOH/CHCl3) to yield the title compound
14 as a blue solid (5.0 mg, 75%). When submitted to analytical
HPLC (Jones C8 column, [250 mm × 4.6 mm], 1 mL‚min-1, 35-
100% B, 30 min, monitored using photodiode array detection
between 190 and 800 nm; mobile phases; solvent A, H2O; solvent
B, MeCN), 14 was observed to elute as a single and symmetrical
peak at retention time Rt given below. The purity of 14 was
confirmed by performing a second analytical HPLC (Jones C4
column, [250 mm × 4.6 mm], 1 mL‚min-1, 35%-100% B, 30
min, monitored using photodiode array detection between 190 and
800 nm; mobile phases; solvent A, H2O; solvent B, MeOH). Again,
14 eluted as a homogeneous single peak. The identity of 14 was
Data Analysis: EC50, IC50 (concentrations of drug producing
50% of the maximal stimulatory or inhibitory response), and pEC50
(negative logarithm of the EC50) values were obtained by computer-
assisted curve fitting (to a sigmoidal dose response) using the
computer program PRISM (GraphPAD, California). pKB values
were calculated from EC50 values obtained in the presence and
absence of antagonist using the Gaddum equation:
1
further analyzed by H NMR and HRMS (TOF ES+): δH (250
MHz, DMSO-d6) 1.69-1.15 (11H, m), 2.00 (2H, m), 3.21-3.0 (4H,
m), 3.70-3.45 (3H, m), 3.97 (1H, m, C4′H), 4.16 (1H, m, C3′H),
4.54 (2H, s, BODIPY CH2), 4.62 (1H, q, J ) 6.0, C2′H), 5.20 (1H,
d, J ) 4.6), 5.50-5.45 (2H, m), 5.90 (1H, d, J ) 6.0, C1′H), 6.96
(1H, d, J 4.3) 7.10-7.07 (2H, m), 7.43-7.27 (4H, m), 7.63-7.60
(3H, m), 7.79-7.72 (2H, m), 7.83 (1H, d, J ) 3.0), 7.90 (1H, br
s), 8.05 (1H, d, J 3.0), 8.22-8.15 (2H, m), 8.38 (1H, s, purine C8);
m/z HRMS (FAB) C43H49BF2N9O7S (MH)+ calcd, 884.3538; found,
pKB ) log(concentration ratio - 1) -
log(antagonist concentration)
884.3774; Rt 13.19 min (Jones C8, 250 × 4.6 mm, 1 mL‚min-1
,
35-100% B, 30 min), Rt-2 9.13 min (Jones C4, 250 × 4.6 mm, 1
mL‚min-1, 35-100% B, 30 min; solvent A, H2O; solvent B,
MeCN).
Saturation curves for specific [3H]-DPCPX binding were fitted to
the following equation:
General Pharmacology Methods: Measurement of [3H]-
cAMP Accumulation: The human adenosine A3 receptor cDNA,
cloned into pcDNA3.1+ (Invitrogen) was obtained from UMR
cDNA Resource Center (U.S.A.). This was transfected into CHO-
K1 cells (European Collection of Animal Cell Cultures, Porton
Down, Salisbury, U.K.) using lipofectamine (according to manu-
facturer’s instructions). Stably transfected CHO-K1 cells (CHO-
A3 cells) were selected using 500 µg/mL Geneticin (G418; Life
Technologies, Inc., Gaithersburg, MD) for two weeks. CHO-A3
cells resistant to G418 were subsequently cloned by dilution cloning.
Clonal CHO-A3 cells, CHO-A1H cells,20 CHO-A2B cells,30 or
CHO-A2A cells (a gift from Dr. K. N. Klotz, Wurzburg, Germany)
were grown in 24-well cluster dishes until confluent. Cells were
Bmax × KD
Specific binding )
([L] + KD)
where Bmax is the maximum specific binding, KD is the dissociation
constant of [3H]-DPCPX, and [L] is the concentration of [3H]-
DPCPX. Curves for inhibition of the binding of [3H]-DPCPX by
agonists were fitted to the following equation:
100 - Imax
% of uninhibited binding )
+ Imax
1 + ([A]/IC50)
where [A] is the concentration of competing ligand, IC50 is the

792 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4
Middleton et al.
(7) (a) Weisshart, K.; Jungel, V.; Briddon, S. J. The LSM 510 META
ConfoCor 2 System: An Integrated Imaging and Spectroscopic
Platform for Single-Molecule Detection. Curr. Pharm. Biotechnol.
2004, 5, 134-154. (b) Pramanik, A.; Olsson, M.; Langel, U.; Bartfai,
T.; Rigler, R. Fluorescence Correlation Spectroscopy Detects Galanin
Receptor Diversity on Insulinoma Cells Biochemistry 2001, 40,
10839-10845. (c) Hegener, O.; Jordan, R.; Haberlein, H. Dye-
Labeled Benzodiazepines: Development of Small Ligands for
Receptor Binding Studies Using Fluorescence Correlation Spectros-
copy J. Med. Chem. 2004, 47, 3600-3605.
(8) (a) Pramanik, A.; Rigler, R. Ligand-Receptor Interactions in the
Membrane of Cultured Cells Monitored by Fluorescence Correlation
Spectroscopy. Biol. Chem. 2001, 382, 371-378. (b) Zhong, Z.-H.;
Pramanik, A.; Ekberg, K.; Jansson, O. T.; Jo¨rnvall, H.; Wahren, J.;
Rigler, R. Insulin Binding Monitored by Fluorescence Correlation
Spectroscopy. Daibetologia 2001, 44, 1184-1188.
(9) McGrath, J. C.; Arribas, S.; Daly, C. J. Fluorescent Ligands for the
Study of Receptors. Trends Pharmacol. Sci. 1996, 17, 393-399 and
references therein.
(10) Briddon, S. J.; Middleton, R. J.; Cordeaux, Y.; Flavin, F. M.;
Weinstein, J. A.; George, M. W.; Kellam, B.; Hill, S. J. Quantitative
Analysis of the Formation and Diffusion of A(1)-Adenosine Recep-
tor-Antagonist Complexes in Single Living Cells. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 4673-4678.
concentration that inhibits the agonist-displaceable binding by 50%,
and Imax is the maximal inhibition of binding produced by the
agonist. Apparent agonist dissociation constants (KB) were then
determined from the following expression:
IC50 × KD
KB )
(KD + [L])
where [L] and KD are the concentration and dissociation constant
of [3H]-DPCPX, respectively, as determined from the saturation
binding studies.
Statistical significance was determined by Student’s unpaired t
test, unless otherwise stated. All data are presented as mean ( s.e.
mean. The n in the text refers to the number of separate experiments
performed.
Acknowledgment. We thank Julia A. Weinstein for help
with photophysical characterization measurements. This work
was supported by The Wellcome Trust (Grant Numbers 057199
and 066817) and The University of Nottingham. The molecules
described in this paper are the subject of an international patent
application, WO 2004/088312.
(11) (a) Daly, C. J.; McGrath, J. C. Fluorescent Ligands, Antibodies, and
Proteins for the Study of Receptors. Pharmacol. Ther. 2003, 100,
101-118. (b) Middleton, R. J.; Kellam, B. Fluorophore-Tagged
GPCR Ligands. Curr. Opin. Chem. Biol. 2005, 9, 517-525.
(12) Daly, C. J.; McGrath, J. C. Do Fluorescent Drugs Show You More
Than You Wanted To Know? Br. J. Pharmacol. 2003, 139, 187-
189.
Supporting Information Available: Full experimental proce-
dures and compound characterization data for compounds 5-9a,c,d.
Representative spectra for compounds 2-9b. This material is
available free of charge via the Internet at http://pubs.acs.org.
(13) (a) Olsson, R. A.; Kusachi, S.; Thompson, R. D.; Ukena, D.; Padgett,
W.; Daly, J. W. N-6-Substituted N-Alkyladenosine-5′-uronamides-
bifunctional Ligands Having Recognition Groups for A1 and A2
Adenosine Receptors. J. Med. Chem. 1986, 29, 1683-1689. (b)
Gallo-Rodriguez, C.; Ji, X.-d.; Melman, N.; Siegman, B. D.; Sanders,
L. H.; Orlina, J.; Fischer, B.; Pu, Q.; Olah, M. E.; van Galen, P. J.
M.; Stiles, G. L.; Jacobsen, K. A. Structure-Activity Relationships
of N-6-Benzyladenosine-5′-uronamides as A(3)-Selective Adenosine
Agonists. J. Med. Chem. 1994, 37, 636-646. (c) Gu¨ngo¨r, T.; Malabre,
P.; Teulon, J.-T.; Camborde, F.; Meignen, J.; Hertz, F.; Virone-Oddos,
A.; Caussade, F.; Cloarec, A. N-6-Substituted Adenosine Receptor
AgonistssSynthesis and Pharmacological Activity as Potent Anti-
nociceptive Agents. J. Med. Chem. 1994, 37, 4307-4316. (d) Ha,
S. B.; Nair, V. An Improved Approach to the Synthesis of Adenosine-
5′-N-ethyluronamides of Interest as Adenosine Receptor Agonists.
Tet. Lett. 1996, 37, 1567-1570. (e) Wright, D. M. J.; Baker, S. P.;
Stewart, S. G.; Scammells, P. J. N-6-(5,6-Epoxynorbornyl)adenosine
Analogs as A(1) Adenosine Agonists. Bioorg. Med. Chem. Lett. 1998,
8, 3647-3648.
References
(1) Ma, P.; Zemmel, R. Value of Novelty? Nat. ReV. Drug DiscoVery
2002, 1, 571-572.
(2) Ralevic, V. R.; Burnstock, G. Receptors for Purines and Pyrimidines.
Pharmacol. ReV. 1998, 50, 413-492.
(3) (a) Gines, S.; Ciruela, F.; Burgueno, J.; Casado, V.; Canela, E. I.;
Mallol, J.; Lluis, C.; Franco, R. Involvement of Caveolin in Ligand-
Induced Recruitment and Internalization of A(1) Adenosine Receptor
and Adenosine Deaminase in an Epithelial Cell Line. Mol. Phar-
macol. 2001, 59, 1314-1323. (b) Zajchowski, L. D.; Robbins, S.
M. Lipid Rafts and Little CavessCompartmentalized Signalling in
Membrane Microdomains. Eur. J. Biochem. 2002, 269, 737-752.
(c) Helmreich, E. J. M. Environmental Influences on Signal Trans-
duction Through Membranes: A Retrospective Mini-Review. Bio-
phys. Chem. 2003, 100, 519-534.
(4) (a) Baker, J. G.; Hall, I. P.; Hill, S. J. Pharmacology and Direct
Visualisation of BODIPY-TMR-CGP: A Long-Acting Fluorescent
Beta(2)-Adrenoceptor Agonist. Br. J. Pharmacol. 2003, 139, 232-
242. (b) Harikumar, K. G.; Pinon, D. I.; Wessels, W. S.; Prendergast,
F. G.; Miller, L. J. Environment and Mobility of a Series of
Fluorescent Reporters at the Amino Terminus of Structurally Related
Peptide Agonists and Antagonists Bound to the Cholecystokinin
Receptor. J. Biol. Chem. 2002, 277, 18552-18560. (c) Emmerson,
P. J.; Archer, S.; El-Hamouly, W.; Mansour, A.; Akil, H.; Medzi-
hradsky, F. Synthesis and Characterization of 4,4-Difluoro-4-bora-
3a,4a-diaza-s-indacene (BODIPY)-Labeled Fluorescent Ligands for
the mu Opioid Receptor. Biochem. Pharmacol. 1997, 54, 1315-
1322. (d) Bennett, V. J.; Simmons, M. A. Analysis of Fluorescently
Labeled Substance P Analogs: Binding, Imaging, and Receptor
Activation. BMC Chem. Biol. 2001, 1, 1-12.
(5) (a) Schwille, P. Fluorescence Correlation Spectroscopy and its
Potential for Intracellular Applications. Cell Biochem. Biophys. 2001,
34, 383-408. (b) Pramanik, A.; Rigler, R. FCS-Analysis of Ligand-
Receptor Interactions in Living Cells. In Fluorescence Correlation
Spectroscopy: Theory and Applications; Elson, E., Rigler, R., Eds.;
Springer: Heidelberg, 2001; pp 101-129.
(6) (a) Bacia, K.; Schwille, P. A Dynamic View of Cellular Processes
by In Vivo Fluorescence Auto- and Cross-Correlation Spectroscopy.
Methods 2003, 29, 74-85. (b) Bacia, K.; Scherfeld, D.; Kahya, N.;
Schwille, P. Fluorescence Correlation Spectroscopy Relates Rafts in
Model and Native Membranes. Biophys. J. 2004, 87, 1034-1043.
(c) Wang, Z.; Shah, J. V.; Chen, Z.; Sun, C. H.; Berns, M. W.
Fluorescence Correlation Spectroscopy Investigation of a GFP
Mutant-Enhanced Cyan Fluorescent Protein and Its Tubulin Fusion
in Living Cells with Two-Photon Excitation. J. Biomed. Opt. 2004,
9, 395-403. (d) Kim, S. A.; Heinze, K. G.; Waxham, M. N.;
Schwille, P. Intracellular Calmodulin Availability Accessed with
Two-Photon Cross-Correlation. Proc. Natl. Acad. Sci. U.S.A. 2004,
101, 105-110. (e) Bacia, K.; Majoul, I. V.; Schwille, P. Probing the
Endocytic Pathway in Live Cells Using Dual-Color Fuorescence
Cross-Correlation Analysis. Biophys. J. 2002, 83, 1184-1193.
(14) (a) Macchia, M.; Salvetti, F.; Barontini, S.; Calvani, F.; Gesi, M.;
Hamdan, M.; Lucacchini, A.; Pellegrini, A.; Soldani, P.; Martini, C.
Fluorescent Probes for Adenosine Receptors: Synthesis and Biology
of N-6-Dansylaminoalkyl-Substituted NECA Derivatives. Bioorg.
Med. Chem. Lett. 1998, 8, 3223-3228. (b) Macchia, M.; Salvetti,
F.; Bertini, S.; Di Bussolo, V.; Gattuso, L.; Gesi, M.; Hamdan, M.;
Klotz, K.-N.; Laragione, T.; Lucacchini, A.; Minutolo, F.; Nencetti,
S.; Papi, C.; Tuscano, D.; Martini, C. 7-Nitrobenzofurazan (NBD)
Derivatives of 5′-N-Ethylcarboxamidoadenosine (NECA) as New
Fluorescent Probes for Human A(3) Adenosine Receptors. Bioorg.
Med. Chem. Lett. 2001, 11, 3023-3026. (c) Jacobson, K. A.; Ukena,
D.; Padgett, W.; Kirk, K. L.; Daly, J. W. Molecular Probes for
Extracellular Adenosine Receptors. Biochem. Pharmacol. 1987, 36,
1697-1707.
(15) Epp, J. B.; Widlanski, T. S. Facile Preparation of Nucleoside-5′-
carboxylic Acids. J. Org. Chem. 1999, 64, 293-295.
(16) (a) De Mico, A.; Margarita, R.; Parlanti, L.; Vescovi, A.; Piancatelli,
G. A Versatile and Highly Selective Hypervalent Iodine (III)/2,2,6,6-
Tetramethyl-1-piperidinyloxyl-Mediated Oxidation of Alcohols to
Carbonyl Compounds. J. Med. Chem. 1997, 62, 6974-6977. (b) de
Nooy, A. E. J.; Besemer, A. C.; van Bekkum, H. On the Use of
Stable Organic Nitroxyl Radicals for the Oxidation of Primary and
Secondary Alcohols. Synthesis 1996, 1153-1174.
(17) (a) Chen, F. M. F.; Benoiton, N. L. Reductive N,N-Dimethylation of
Amino-Acid and Peptide Derivatives Using Methanol as Carbonyl
Source. Can. J. Biochem. 1978, 56, 150-152. (b) Benoiton, N. L.
On the Side-Reaction of N-Alkylation of Amino-Groups During
Hydrogenolytic Deprotection in Alcohol-Containing Solvents. Int.
J. Pept. Protein Res. 1993, 41, 611-611. (c) Mazaleyrat, J. P.; Xie,
J.; Wakselman, M. Selective Hydrogenolysis of the Benzyloxycar-
bonyl Protecting Group of N-Epsilon-Lysine in Cyclopeptides
Containing a Benzylic Phenyl Ether Function-Evidence for N-Epsilon-

Fluorescent Adenosine A1-Receptor Agonists
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 4 793
Methylated Lysine Side Products. Tetrahedron Lett. 1992, 33, 4301-
4302. (d) Filira, F.; Biondi, L.; Gobbo, M.; Rocchi, R. N-Alkylation
of Amino Acids During Hydrogenolytic Deprotection. Tetrahedron
Lett. 1991, 32, 7463-7464.
(25) Baddiley, J. Adenine 5′-Deoxy-5′-methylthiopentoside (Adenine
Thiomethyl Pentoside)sA Proof of Structure and Synthesis. J. Chem.
Soc. 1951, 1348-1351.
(26) Schmidt, R. R.; Fritz, H.-J. Synthesis and Structure of Inosine-5′-
carboxylic Acid and Derivatives. Chem. Ber. 1970, 103, 1867-1871.
(27) Atwell, G. J.; Denny, W. A. Monoprotection of R,ω-Alkanediamines
with the N-Benzyloxycarbonyl Group. Synthesis 1984, 1032-1033.
For an alternative synthesis, see Pittelkow, M.; Lewinsky, R.;
Christensen, J. B. Selective Synthesis of Carbamate Protected
Polyamines Using Alkyl Phenyl Carbonates. Synthesis 2002, 2195-
2202.
(28) For an alternative synthesis of 5b, see: (a) Blagborough, I. S.; Moya,
E.; Walford, S. P. Practical, Convergent Total Synthesis of Polyamine
Amide Spider Toxin NSTX-3. Tetrahedron Lett. 1996, 37, 551-
554. (b) Wa¨lchli-Schaer, E.; Eugster, C. H. Equisetic Alkaloids 17.
Synthesis of Threo-cis-threo-trans-dihydropalustrin and Erythro-cis-
erythro-trans-dihydropalustrin. HelV. Chim. Acta 1978, 61, 928-
935.
(29) Bredereck, H. U¨ ber Methylierte Nucleosiden und Purines und Ihre
Pharmakologische Wirkungen. I. Methylierung von Nucleosiden
durch Diazomethan. Chem. Ber. 1947, 80, 401-405.
(30) Alexander, S. P.; Cooper, J.; Shine, J.; Hill, S. J. Characterization of
the Human Brain Putative A2B Adenosine Receptor Expressed in
Chinese Hamster Ovary (CHO, A2B4) cells. Br. J. Pharmacol. 1996,
119, 1286-1290.
(18) Fredholm, B. B.; Ijzerman, A. P.; Jacobson, K. A.; Klotz, K.-N.;
Linden, J. International Union of Pharmacology. XXV. Nomenclature
and Classification of Adenosine Receptors. Pharmacol. ReV. 2001,
53, 527-552.
(19) Buschmann, V.; Weston, K. D.; Sauer, M. Spectroscopic Study and
Evaluation of Red-Absorbing Fluorescent Dyes. Bioconjugate Chem.
2003, 14, 195-204.
(20) Cordeaux, Y.; Briddon, S. J.; Megson, A. E.; McDonnell, J. M.;
Dickenson, J. M.; Hill, S. J. Influence of Receptor Number on
Functional Responses Elicited by Agonists Acting at the Human
Adenosine A(1) Receptor: Evidence for Signaling Pathway-Depend-
ent Changes in Agonist Potency and Relative Intrinsic Activity. Mol.
Pharmacol. 2000, 58, 1075-1084.
(21) For definitions and further details on data analysis, see Pharmacology
Methods in the Experimental Section.
(22) Briddon, S. J.; Middleton, R. J.; Yates, A. S.; George, M. W.; Kellam,
B.; Hill, S. J. Application of Fluorescence Correlation Spectroscopy
to the Measurement of Agonist Binding to a G-Protein Coupled
Receptor at the Single Cell Level. Faraday Discuss. 2004, 126, 197-
207.
(23) Rigler, R.; Pramanik, A.; Jonasson, P.; Kratz, G.; Jansson, O. T.;
Nygren, P.-Å.; Ståhl, S.; Ekberg, K.; Johansson, B.-L.; Uhle´n, S.;
Uhle´n, M.; Jo¨rnvall, H.; Wahren, J. Specific Binding of Proinsulin
C-Peptide to Human Cell Membranes. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 13318-13323.
(24) Hampton, A. Nucleotides 2. A New Procedure for Conversion of
Ribonucleosides to 2′,3′-O-Iisopropylidene Derivatives. J. Am. Chem.
Soc. 1961, 83, 3640-3645.
(31) Alvarez, R.; Daniels, D. V. A Single Column Method for the Assay
of Adenylate-Cyclase. Anal. Biochem. 1990, 187, 98-103.
JM061279I