A New Class of Fluorinated A2A Adenosine Receptor Agonist with Application to Last-Step Enzymatic [18F]Fluorination for PET Imaging

Article abstract of DOI:10.1002/cbic.201700382

The A_2A adenosine receptor belongs to a family of G-coupled protein receptors that have been subjected to extensive investigation over the last few decades. Due to their prominent role in the biological functions of the heart, lungs, CNS and brain, they have become a target for the treatment of illnesses ranging from cancer immunotherapy to Parkinson's disease. The imaging of such receptors by using positron emission tomography (PET) has also been of interest, potentially providing a valuable tool for analysing and diagnosing various myocardial and neurodegenerative disorders, as well as offering support to drug discovery trials. Reported herein are the design, synthesis and evaluation of two new 5′-fluorodeoxy-adenosine (FDA)-based receptor agonists (FDA-PP1 and FDA-PP2), each substituted at the C-2 position with a terminally functionalised ethynyl unit. The structures enable a synthesis of ¹⁸F-labelled analogues by direct, last-step radiosynthesis from chlorinated precursors using the fluorinase enzyme (5′-fluoro-5′-deoxyadenosine synthase), which catalyses a transhalogenation reaction. This delivers a new class of A_2A adenosine receptor agonist that can be directly radiolabelled for exploration in PET studies.

Full text of DOI:10.1002/cbic.201700382

Accepted Article
Title: A novel class of fluorinated A2A adenosine receptor agonist
with application to last step enzymatic [18F]fluorination for PET
imaging
Authors: David O'Hagan, Phillip T Lowe, Sergio Dall'Angello, Matteo
Zanda, Adriaan P Ijzerman, and Thea Mulder-Krieger
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemBioChem 10.1002/cbic.201700382
Link to VoR: http://dx.doi.org/10.1002/cbic.201700382
A Journal of
www.chembiochem.org

10.1002/cbic.201700382
ChemBioChem
A novel class of fluorinated A2A adenosine receptor agonist with
application to last step enzymatic [18F]fluorination for PET imaging
Phillip T. Lowe,[a] Sergio Dall'Angelo,[b] Thea Mulder-Krieger,[c] Adriaan P. IJzerman,[c] Matteo Zanda,[b]
and David O'Hagan*[a]
[a] Dr. P. T. Lowe, Prof. D. OʹHagan
School of Chemistry and Centre for Biomolecular Sciences
University of St. Andrews, North Haugh, St. Andrews, Fife, KY16 9ST (UK)
E-mail: do1@st-andrews.ac.uk
[b] Dr. S. DallʹAngelo, Prof. M. Zanda
John Mallard Scottish PET Centre
School of Medicine, Medical Sciences and Nutrition
University of Aberdeen, Foresterhill, Aberdeen, AB25 2ZD (UK)
E-mail: m.zanda@abdn.ac.uk
[c] Dr. T. Mulder-Krieger, Prof. A. IJzerman
Leiden University, Leiden Academic Centre for Drug Research,
Medicinal Chemistry
Gorlaeus Laboratories, Einsteinweg 55, 2333 CC Leiden
E-mail: ijzerman@lacdr.leidenuniv.nl
Abstract: The A2A adenosine receptor belongs to a family of G-coupled protein receptors that have
been subjected to extensive investigation over the last few decades. Due to their prominent role in
the biological functions of the heart, lungs, CNS and brain, they have become a target for the
treatment of illnesses ranging from cancer immunotherapy to Parkinson’s disease. The imaging of
such receptors using positron emission tomography (PET) has also been of interest, potentially
providing a valuable tool to analyse and diagnose various myocardial and neurodegenerative
disorders, as well as offering support to drug-discovery trials. Reported herein is the design, synthesis
and evaluation of two novel 5'-fluorodeoxy-adenosine (FDA) based receptor agonists, each
substituted at the C-2 position with a terminally functionalised ethynyl unit. The structures enable a
18
synthesis of F-labelled analogues via direct, last-step, radiosynthesis from chlorinated precursors
using the fluorinase enzyme (5'-fluoro-5'-deoxyadenosine synthase) which catalyses a
transhalogenation reaction. This delivers a new class of A2A adenosine receptor agonist which can be
directly radiolabelled for exploration in PET studies.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
Introduction
Adenosine is a purine nucleoside that presents ubiquitously throughout the body and is involved in
numerous vital biological functions.[1] Along with its involvement in important physiological processes
as a substructure of endogenous molecules such as ATP[2] or RNA, adenosine acts at an extracellular
level, where it is the native ligand for adenosine receptors. These adenosine receptors are divided into
four subtypes: A1, A2A, A2B, and A3, each belonging to the broader family of G-protein-coupled
receptors (GPCRs).[3] Adenosine signalling is widespread throughout mammalian organisms and each
class of adenosine receptor possess distinct distributions and control over a broad spectrum of
physiological and pathophysiological functions.[4]
Several studies have been conducted into the various subclasses of adenosine receptors, mostly
relating to the inherent wider biological implications of their stimulation.[4-5] The A2A adenosine
receptors are responsible for the modulation of secondary messenger pathways, in particular the
regulation of intracellular cyclic adenosine monophosphate (cAMP) biosynthesis.[6] Upon binding of
adenosine to the adenosine A2A receptor, and following its subsequent activation, levels of adenylyl
cyclase are enhanced, thus stimulating cAMP production. A2A adenosine receptors are found in
numerous regions throughout the body, but they are most notably expressed in the immune cells of
the spleen, leukocytes, blood platelets, and in the heart where adenosine is responsible for regulating
the vasodilation of the coronary arteries, thus mediating coronary blood flow. Adenosine receptors
are also significant in particular regions of the brain including the thymus, striatopallidal GABAergic
neurons and the olfactory bulb, where they play an important role in the regulation of glutamate and
dopamine release.[1, 7] A2A adenosine receptors have been the subject of increased focus in recent
years due to their association with numerous pathogenic, myocardial and neurological diseases. As
such, they have become a therapeutic target for treatment of disorders ranging from inflammation,
ischemia reperfusion injury, insomnia, infectious diseases, depression to CNS disorders and
Parkinson's disease.[1, 7-8]
Considerable effort has been made to establish structure-activity relationships of adenosine-based
ligands with respect to their stimulation of adenosine receptors, and from these studies a general
pharmacophore has been established.[9] Predominant approaches to develop A2A adenosine receptor
agonists originate through modification of adenosine itself, and from these attempts it has been
acknowledged that the adenosine scaffold must be largely maintained as the structural basis for
agonist activity (see Figure 1). Furthermore, modifications to distinct regions of this scaffold can have
discrete and significant effects on the binding and selectivity of adenosine analogues. Concerning the
A2A adenosine receptor specifically, it has been shown that the 2ʹ- and 3ʹ-hydroxyl groups on the ribose
sugar are essential for full agonist activity, although modifications at the C-2 and N-6 positions of the
adenine base (which often bring rise to increased metabolic stability), as well as modifications to the
5’ position of the ribose ring are tolerated. However, N-6 substitution generally decreases A2A
adenosine receptor potency, and in many instances substitution at this position enhances the affinity
and selectivity of A1 and A3 receptor binding.[7, 9a]
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
A
B
Figure 1. (A) General adenosine receptor agonist pharmacophore. (B) Examples of A2A adenosine receptor
agonists.
Herein we describe the synthesis of novel C-2 ethynyl adenosines 1 and 2 modified at C-5’ of the ribose
ring with a fluorodeoxy moiety (see Figure 2), along with assessment of their affinities for the human
A2A adenosine receptor and their agonist behaviour.
Figure 2. Structures of novel fluorinated A2A adenosine receptor agonists 1 and 2.
The fluorine in these agonists presents an immediate opportunity to develop radiotracers for PET
imaging using the fluorine-18 isotope, as the 5'-fluorodeoxyadenosine motif is compatible with the
enzymatic introduction of [18F]fluorine from [18F]fluoride. The wild type fluorinase enzyme[10] (5'-
fluoro-5'-deoxyadenosine synthase) catalyses the reaction between S-adenosyl-L-methionine (SAM)
and fluoride ion, generating 5ʹ-fluoro-5ʹ-deoxyadenosine (FDA) and L-(S)-methionine (L-Met).
However, the enzyme has been shown to catalyse a transhalogenation reaction using C-2 acetylene
18
substituted 5'-chlorodeoxyadenosine substrates to prepare F-labelled C-2 acetylene substituted 5'-
fluorodeoxyadenosines under experimentally benign conditions (buffers at pH 7.8) as illustrated in
Scheme 1.[11] This has led to the strategy of tethering PEGylated peptide cargo (R group in Scheme 1)
to permit last step [18F]radiolabelling of cancer targeting peptides.[11-12]
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
Scheme 1. Fluorinase-catalysed transhalogenation reactions with C-2 modified 5'-chloro-5'-deoxyadenosine
substrates.
This study encompasses a strategy for both the chemical and enzymatic syntheses of 1 and 2, which
are derivatives of the adenosine receptor agonists in Figure 1.[13] As the fluoromethyl group at C-5' of
this structural class of agonist had the potential to render them amenable to enzymatic fluorination
from their chlorinated precursors using [18F]fluoride, we demonstrated this in the radiochemical
synthesis of [18F]1, exemplifying the potential application of this class of A2a adenosine receptor
agonist as tracers for clinical imaging.[9e, 14]
Results and Discussion
Synthesis
The synthesis of 1 and 2, along with their chlorinated precursors 19 and 24, required an approach
involving a Sonogashira coupling reaction between the C-2 iodinated adenosine scaffold and the
appropriate functionalised alkyne coupling partner. For this study, 5ʹ-fluoro-5ʹ-deoxy-2-iodoadenosine
13 and 5ʹ-chloro-5ʹ-deoxy-2-iodoadenosine 11 were synthesised using previously reported
strategies[12b] as illustrated in Scheme 2. Briefly, 8 was synthesised from guanosine using a two-step
procedure first involving the per acetylation of the ribose moiety, followed by chlorination at C-6 of
the guanine base using POCl3. 8 was then iodinated at C-2 to afford 9, which was subjected to
simultaneous amination and deprotection to give 2-Iodoadenosine 10. With 10 in hand, treatment
with thionyl chloride followed by treatment with ammonia in aqueous methanol afforded the 5ʹ-
chloro-2-iodinated coupling partner 11. Alternatively, acetonide protection of 11, before fluorination
with TsF and TBAF, followed by deprotection with TFA, afforded 13 in good yield.
The formation of alkyne 18 was achieved from commercially available N-Boc-4-piperidinemethanol
14. Introduction of a tosyl group to N-Boc-4-piperidinemethanol 14 afforded 15 which, after flash
chromatography, allowed for the insertion of
a
terminal alkyne using lithium
acetylide/ethylenediamine complex in DMSO. Removal of the Boc group gave the free amine 17,
which followed by treatment with TEA and methyl chloroformate afforded 18 in good yield.
Sonogashira cross-coupling was then performed using an excess of the alkyne 18 over 11 in the
presence of a Pd2(dba)3 catalyst, CuI, and trimethylamine, in DMF. The product was subjected to C18
cartridge purification, followed by semi-prep HPLC to afford the coupled product 19 in good yield and
18
high purity. In line with the proposed strategy of using the fluorinase to generate F analogues of
these A2A adenosine receptor agonists, 1 was prepared enzymatically by transhalogenation of 19 in
good yield as illustrated in Scheme 3 (See supporting Information for full experimental details for
preparative transhalogenation).
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
Scheme 2. Synthesis of 5ʹ-chloro-5ʹ-deoxy-2-iodoadenosine 11 and 5ʹ-fluoro-5ʹ-deoxy-2-iodoadenosine 13.
Reagents and conditions: a) Ac2O, pyridine, DMF. b) POCl3, Et4NCl, dimethylaniline, CH3CN. c) CuI, I2, CH2I2,
isoamyl nitrite, THF. d) NH3, MeOH. e) SOCl2, pyridine, CH3CN, then NH3, MeOH, H2O. f) Acetone, 2,2-DMP, HClO4.
g) TBAF, TsF, THF, then TFA, MeOH.
Scheme 3. Synthesis of 19 and 1. Reagents and conditions: a) Tosyl chloride, Et3N, THF. b) Lithium acetylide,
ethylenediamine complex, DMSO. c) TFA; d) Et3N, Methyl chloroformate. e) Pd2(dba)3, Et3N, CuI, DMF; f)
Fluorinase, l-Se-Met, KF, phosphate buffer (pH 7.8).
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
Product 2 was acquired from a similar a strategy commencing with 1-Boc-piperazine 20. Introduction
of the acetylene was accomplished using TEA and propargyl bromide, which after purification afforded
21 in very good yield. As before, the methyl carbamate moiety was introduced using successive TFA
deprotection of the Boc protected amine, followed by treatment with methyl chloroformate in the
presence of an excess of base to afford 23. With 23 in hand Sonogashira cross-coupling was performed
using the previously established protocol, with both the 5ʹ-fluoro-5ʹ-deoxy-2-iodoadenosine 13 and
5ʹ-chloro-5ʹ-deoxy-2-iodoadenosine 11 coupling partners as illustrated in Scheme 4. Purification by C18
cartridge followed by semi prep HPLC afforded 2 and 24 in good yields and in high purity
Scheme 4. Synthesis of 2 and 24. Reagents and conditions: a) Propargyl bromide, Et3N; b) TFA. c) Et3N, Methyl
chloroformate. d) Pd2(dba)3, Et3N, CuI, DMF.
Radioligand displacement assays on Hek293 membranes expressing the human A2A receptor
With 1 and 2 in hand their relative affinities to the A2A adenosine receptor and any potential agonist
activity was evaluated. Radioligand displacement experiments at the human A2A adenosine receptor
were performed with a range of concentrations of each potential ligand, in the presence of [3H]-
ZM241385. As a reference, CGS21680, a known agonist which possesses a high affinity to the A2A
adenosine receptor,[15] was also included in these experiments (Figure 3). Subsequent concentration-
response curves were well behaved, and revealed the affinities of 1 and 2 for the human adenosine
A2A receptor to be both in the nanomolar range, with Ki values of 39 nM and 176 nM, respectively.
Encouragingly, the affinity of the reference agonist CGS21680 under these conditions was also 39 nM,
notably comparable to that of 1 (see also Table 1). It follows that the fluoromethyl group at the 5'-
position is well tolerated by the A2A adenosine receptor. A prerequisite for any viable PET tracer is a
high affinity for the target receptor; with this in mind the [18F] analogue of 1 was deemed the most
promising of the two new agonists as a potential PET tracer, and as such was explored in hot
radiolabelling studies.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
1 2 5
1 0 0
7 5
5 0
2 5
0
1
2
C G S 2 16 80
1 + N a C l
2 + N a C l
C G S 2 1 6 8 0 + N a C l
1
0
9
8
7
6
5
4
3
-
-
-
-
-
-
-
1
1
-
-
L o g (co m p e tito r) (M )
Figure 3. Concentration-response curves of 1, 2 and the control CGS21680 under standard conditions and in
the presence of 1M NaCl on Hek293 cell membranes stably expressing the hA2A receptor.
To determine the agonist behavior of 1 and 2 each experiment was repeated in the presence of 1M
NaCl. The binding of a sodium ion to the allosteric binding pocket of the human A2A adenosine receptor
stabilises the inactive conformation, and this results in an observed reduction in orthosteric binding
of the agonists.[16] During these experiments the presence of sodium ion dramatically decreased the
affinities of 1 and 2 to (sub)micromolar values (Table 1), consistent with both compounds acting as
agonists of the human A2A adenosine receptor.[17]
Table 1. Affinities of 1, 2 and the control CGS21680 to the A2A adenosine receptor in the absence
and presence of sodium ion.
Compound
pKi ± SEM (control)
7.42 ± 0.07
pKi ± SEM (+ 1 M NaCl)
6.19 ± 0.02
1
2
6.76 ± 0.01
5.52 ± 0.06
7.43 ± 0.09
6.56 ± 0.07
CGS21680
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
Evaluation of 1 and 2 as substrates for fluorinase mediated transhalogenation
In order to realise the application of last step enzyme catalysed 18F-fluorination of 19 and 24 towards
PET imaging of A2A adenosine receptors, transhalogenation reactions with each of these chlorinated
substrates were first explored with the fluorinase under ‘cold’ labelling conditions using [19F]fluoride.
In a typical experiment each 5'-chloro-5ʹ-deoxy-adenosine substrate 19 and 24 was incubated with the
fluorinase enzyme (0.7 mg mL-1) in the presence of L-Se-Met (75 μM), potassium fluoride (50 mM) in
phosphate buffer, at pH 7.8 (see the Experimental Section and Supporting Information for full
experimental details and Figures). In order to monitor the progress of each assay, analytical time
course experiments were conducted by removing aliquots from each reaction at specified time points.
The enzyme was first removed by heat precipitation, followed by centrifugation, and conversions were
determined by HPLC. Both 19 and 24 proved to be very good substrates for fluorinase mediated
transhalogenation to generate 1 and 2 respectively (see Figures 4 and 5). The identities of both
fluorinated products were confirmed by both mass spectrometry, and against fully characterised
synthetic reference samples. The success of these reactions demonstrates that the methyl
piperizine/piperidine-1-carboxylate groups of 19 and 24 do not negatively interfere with the action of
the fluorinase enzyme, despite being on a short tether to the 5'-chloro-5ʹ-deoxy-adenosine
recognition/substrate motif. This opened up the prospect for direct ‘last-step’ [18F]fluorine-labelling
using fluorinase catalysis for PET.
19
1
Reference
0 h
0.5 h
1 h
2 h
4 h
6 h
8 h
22 h
Figure 4. HPLC time course (UV, 254 nm) of the incubation of 19, green (tR = 10.9 min), with the fluorinase, l-Se-
Met, KF, phosphate buffer (pH 7.8) at 37 °C. Traces show the formation of 1, yellow (tR = 9.9 min), and the
consumption of 19. For full conditions see the Experimental Section and Supporting Information.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
24
2
0 h
0.5 h
1 h
2 h
3 h
4 h
5 h
6 h
7 h
20 h
Figure 5. HPLC time course (UV, 254 nm) of the incubation of 24, green (tR = 4.5 min), with the fluorinase, l-Se-
Met, KF, phosphate buffer (pH 7.8) at 37 °C. Traces show the formation of 2, yellow (tR = 2.8 min), and the
consumption of 24. For full conditions see the Experimental Section and Supporting Information.
There have been numerous studies probing the substrate specificity of the fluorinase enzyme. To date,
these have revealed only a limited tolerance at N-6 and C-2 of the adenine base,[18] or more elaborate
additions to the acetylene terminus of elongated PEG chains projecting from C-2.[11] 19 and 24
represent a new class of fluorinase substrate, whereby a more bulker cyclic substituent is incorporated
just one methylene unit removed from the C-2 alkyne of the ClDEA motif, thus demonstrating a
development in fluorinase substrate specificity.
Last step enzymatic 18F labelling of 19 to [18F]1
The ability to generate 18F-labelled analogues of human A2A adenosine receptor agonists provides an
opportunity to study a variety of physiological disorders in vivo, as well as offering support to drug
discovery trials targeting this receptor. As a proof of principle study, a transhalogenation reaction was
performed with 19 in the presence of [18F]fluoride, to observe its conversion to [18F]1. Radiolabelling
trials of this type are typically conducted using similar conditions to that described with [19F]fluoride,
but in this instance an aqueous [18F]fluoride solution obtained directly from the cyclotron is used.
Another significant distinction is that under hot labelling conditions the aqueous solution of
[18F]fluoride is generated in GBq as a dilute solution in [18O]water. This solution is then utilised directly
at MBq levels. Thus the final [18F]fluoride ion concentrations are necessarily very low, in the pico-molar
range, and as such these fluorinase-catalyzed radiochemical reactions are no longer catalytic as the
enzyme (micromolar) is in considerable excess over the [18F]fluoride (picomolar). In the event 19,
[18F]fluoride ion, L-Se-Met and fluorinase enzyme were incubated at 37 ◦C in phosphate buffer (pH 7.8)
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
(see Supporting Information for full experimental details and Figures). Product [18F]1 was generated
in approximately 10% radiochemical conversion after a 30 min incubation. A semi-prep HPLC trace
(radiochemistry detector) of the reaction mixture at 30 min is shown in Figure 6.
A
B
[18F]1 (collected)
Figure 6. (A) Reaction scheme of the fluorinase catalysed transhalogenation of 19 to [18F]1. (B) HPLC radioactivity
trace of a fluorinase catalysed conversion of 19 to [18F]1 in the presence of 18F-, l-Se-Met in phosphate buffer
(pH 7.8), at 37 °C after 30 min incubation. [18F]1 was observed at tR = 10.6 min.
After a 30 min reaction the enzyme was heat denatured and cleanly removed by centrifugation. The
clarified supernatant was then purified using semi-preparative HPLC to obtained [18F]1 (see Figure 6).
Once collected, [18F]1 was then diluted with water and loaded onto a C18 reverse phase cartridge,
washed with water and eluted with ethanol to reveal pure [18F]1 as illustrated in Figure 7. A typical
procedure from [18F]fluoride (205 MBq) to EtOH elution of [18F]1 (7 MBq) took 1 h, 50 min, and with a
radiochemical yield of 3.5% (decay uncorrected) and afforded a radiochemical purity of >99%.
[18F]1
Figure 7. Analytical radiochemical HPLC trace for [18F]1 (after semi-prep HPLC purification).
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
This fluorinase mediated 18F-labelling methodology provides a novel route for generating a new class
of human A2A adenosine receptor targeting PET tracer candidates, while offering the principle
advantages of an enzyme mediated synthesis (aqueous and ambient conditions, neutral pH). A
common drawback with the development of small molecular weight PET tracers from known agonists
is that it often necessitates the addition of a structurally bulky radiolabelled prosthetic or chelating
group. The addition of such a group can often perturb the binding and pharmacokinetics of the ligand,
and in the case of the human A2A adenosine receptor SAR studies have shown a relatively strict
structural scaffold must be maintained when developing adenosine based agonists. This fluorinase
mediated strategy allows for smoother transition from the discovery of any new 5'-fluorodeoxy-
adenosine (FDA) based receptor agonists of this class to the development of its corresponding 18F-PET
tracer analogue.
Conclusions
This study demonstrates a new class of A2A adenosine receptor agonist which is designed around a 5'-
fluorodeoxy-adenosine (FDA) scaffold. From this class, two novel agonists were synthesized and their
affinity and agonist behavior against the human A2A adenosine receptor was determined. The
substrate scope of the fluorinase enzyme was further explored and expanded in the fluorinase
mediated synthesis of [18F]1, which demonstrated an efficient radiolabeling strategy that avoids the
need to sacrifice structural integrity in the development of PET tracers from known agonists. Such a
strategy can be applied to future A2A adenosine receptor agonists of this structural class.
Experimental Section
Cold Transhalogenation assay of 19 to 1, and 24 to 2
In a total reaction volume of 1000 μL (in 50 mM phosphate buffer, at pH 7.8), recombinant fluorinase
(0.7 mg/mL−1) was incubated with 19 (0.1 mM) or 24 (0.2 mM), L-SeMet (0.075 mM) and KF (50 mM)
at 37 °C. Samples (50 μL) were periodically removed, the protein precipitated by heating at 95 °C for
5 min, before being clarified by centrifugation (13 000 rpm, 10 min). Samples of the supernatant (40
μL) were removed for analysis by HPLC. HPLC analysis was performed on a Shimadzu Prominence
system using a Kinetix 5μm XB-C18 100A (150 mm × 4.6 mm) column and a guard cartridge. Mobile
phase: 0.05% TFA in water (solvent A) and 0.05% TFA in MeCN (solvent B); Linear Gradient: 15%
solvent B to 95% solvent B over 25 min, 95% for 5 min, and back to 15% B for 10 min to re-equilibrate
the column. Flow rate: 1 mL/min−1; Detection: 254 nm; Injection volume: 40 μL.
[18F] Labelling of 19 to [18F]1
18
A typical F- labelling experiment of 24 was performed as follows: L-selenomethionine (40 μL of a
2mM solution in water) and compound 24 (0.6 mg in 100 μL of water) were added successively to an
Eppendorf tube containing a solution of fluorinase (5 mg in 50 mM phosphate buffer, 60 μL). The
contents were mixed well with a pipette and to this mixture was added [18F-]fluoride in [18O]water
(205 MBq, 50 μL), making a total volume of 250 μL. The contents were again well mixed and incubated
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
at 37 °C for 30 min. After this time the reaction was stopped and the mixture denatured by heating at
95 °C for 5 min, before being clarified by centrifugation (13 000 rpm, corresponding to 16060 g, 5 min).
The supernatant was injected into a Shimadzu Prominence HPLC system equipped with a quaternary
pump, a degasser, a diode array detector and a radioactivity detector using a Phenomenex Kingsorb
C18 (250 × 10.00 mm, 5μ) column and a guard cartridge. Mobile phase: 0.05% TFA in water (solvent
A) and 0.05% TFA in MeCN (solvent B); Linear Gradient: 15% solvent B to 38% solvent B over 16 min,
95% for 5 min, and back to 15% B for 10 min to re-equilibrate the column. Flow rate: 5 mL/min−1. The
radioactive fraction corresponding to the reference of [18F]1 was collected, diluted with water (50 mL)
and loaded onto a preactivated Waters Oasis HLB® Cartridge (conditioned with 2 mL EtOH, 5 mL
water). The cartridge was washed with 20 mL of water and the desired product was collected by
eluting with 0.5 mL of ethanol, to give about 7 MBq (3.5 %, decay uncorrected) of >99% pure product
of [18F]1.
Radioligand displacement assays on Hek293 membranes expressing the human A2A receptor
Radioligand displacement experiments were performed using 6 concentrations of competing ligand in
the presence of 1.7 nM [3H]ZM241385. At this concentration total radioligand binding did not exceed
10% of that added to prevent ligand depletion. The incubations were performed under standard
conditions and in the presence of 1M NaCl to test the effect of NaCl on agonist affinity. Nonspecific
binding was determined in the presence of 100 µM NECA (5ʹ-(N-Ethylcarboxamido)adenosine).
Membrane aliquots containing 30 µg of protein were incubated in a total volume of 100 µL of assay
buffer (25 mM Tris-HCl, pH 7.4) at 25 oC for 2 h to ensure equilibrium was reached.
Incubations were terminated by rapid vacuum filtration to separate the bound and free radioligand
through prewetted 96-well GF/B filter plates using a PerkinElmer Filtermate-harvester (Perkin Elmer,
Groningen, the Netherlands) after the indicated incubation time. Filters were subsequently washed
o
12 times with ice-cold wash buffer (25 mM Tris-HCl, pH 7.4). The plates were dried at 55 C and
MicroscintTM-20 cocktail (Perkin Elmer, Groningen, The Netherlands) was added. After 3 h the filter-
bound radioactivity was determined by scintillation spectrometry using a 2450 MicroBeta Microplate
Counter (Perkin Elmer, Groningen, The Netherlands).
See Supporting Information for experimental detail on: Compound synthesis, characterisation, large
scale enzymatic synthesis of 1, fluorinase overexpression and purification, and HEK293hA2AR cell
culture, membrane preparation and data analysis
Acknowledgements
We thank the Engineering and Physical Sciences Research Council, UK, for a research grant.
Keywords: adenosine receptors · biocatalysis · fluorinase · 18F labelling · positron emission
tomography
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
References
[1]
a) A. N. Druri, A. Szent-György, J. Physiol. 1929, 68, 213; b) K. A. Jacobson, Z.-G. Gao,
Nat. Rev. Drug Discov. 2006, 5, 247-264.
[2]
[3]
J. R. Knowles, Annu. Rev. Biochem. 1980, 49, 877-919.
a) B. B. Fredholm, A. P. Ijzerman, K. A. Jacobson, K. N. Klotz, J. Linden, Pharmacol. Rev. 2001,
53, 527-552; b) B. B. Fredholm, , A. P. Ijzerman, K. A. Jacobson, J. Linden, C. E. Müller,
Pharmacol. Rev. 2011, 63, 1-34.
[4]
[5]
B. B. Fredholm, J. F. Chen, S. A. Masino, J. M. Vaugeois, Annu. Rev. Pharmacol. Toxicol. 2005,
45, 385-412.
a) C. E. Muller, T. Scior, Pharm. Acta Helv. 1993, 68, 77; b) H. K. Eltzschig, M. V. Sitkovsky, S. C.
Robson, N. Engl. J. Med. 2012, 367, 2322-2333; c) K. Fuxe, S. Ferré, M. Canals, M. Torvinen, A.
Terasmaa, D. Marcellino, S. R. Goldberg, W. Staines, K. X. Jacobsen, C. Lluis, A. S. Woods, L. F.
Agnati, R. Franco, J. Mol. Neurosci. 2005, 26, 209.
[6]
[7]
[8]
D. Van Calker, M. Muller, B. Hamprecht, J. Neurochem. 1979, 33, 999.
M. de Lera Ruiz, Y.-H. Lim, J. Zheng, J. Med. Chem. 2014, 57, 3623-3650.
a) G. Hasko, J. Linden, B. Cronstein, P. Pacher, Nat. Rev. Drug Discov. 2008, 7, 759-770; b) J.-F.
Chen, H. K. Eltzschig, B. B. Fredholm, Nat. Rev. Drug Discov. 2013, 12, 265-286; c) K. A.
Jacobson, Z. G. Gao, Nat. Rev. Drug Discov. 2006, 5, 247; d) M. T. Armentero, A. Pinna, S. Ferre,
J. L. Lanciego, C. E. Muller, R. Franco, Pharmacol. Ther. 2011, 132, 280.
a) J. M. Rieger, M. L. Brown, G. W. Sullivan, J. Linden, T. L. Macdonald, J. Med. Chem. 2001, 44,
531-539; b) S. Basu, D. A. Barawkar, S. Thorat, Y. D. Shejul, M. Patel, M. Naykodi, V. Jain, Y.
Salve, V. Prasad, S. Chaudhary, I. Ghosh, G. Bhat, A. Quraishi, H. Patil, S. Ansari, S. Menon, V.
Unadkat, R. Thakare, M. S. Seervi, A. V. Meru, S. De, R. K. Bhamidipati, S. R. Rouduri, V. P. Palle,
A. Chug, K. A. Mookhtiar, J. Med. Chem. 2017, 60, 681-694; c) J. D. Hothersall, D. Guo, S. Sarda,
R. J. Sheppard, H. M. Chen, W. Keur, M. J. Waring, A. P. Ijzerman, S. J. Hill, I. L. Dale, P. B.
Rawlins, Mol. Pharmacol. 2017, 91, 25-38; d) O. Saku, M. Saki, M. Kurosawa, K. Ikeda, T.
Takizawa, N. Uesaka, Bioorg. Med. Chem. Lett. 2010, 20, 1090; e) C. E. Müller, K. A. Jacobson,
Biochim. Biophys. Acta, Biomembr. 2011, 1808, 1290-1308; f) J. Carlsson, L. Yoo, Z.-G. Gao, J.
J. Irwin, B. K. Shoichet, K. A. Jacobson, J. Med. Chem. 2010, 53, 3748-3755.
D. O'Hagan, C. Schaffrath, S. L. Cobb, J. T. G. Hamilton, C. D. Murphy, Nature 2002, 416, 279-
279.
[9]
[10]
[11]
[12]
S. Thompson, Q. Zhang, M. Onega, S. McMahon, I. Fleming, S. Ashworth, J. H. Naismith, J.
Passchier, D. O'Hagan, Angew. Chem., Int. Ed. Engl. 2014, 53, 8913-8918.
a) Q. Zhang, S. Dall'Angelo, I. N. Fleming, L. F. Schweiger, M. Zanda, D. O'Hagan, Chem. - Eur.
J. 2016, 22, 10998-11004; b) S. Thompson, I. N. Fleming, D. O'Hagan, Org. Biomol. Chem. 2016,
14, 3120-3129.
[13]
[14]
W. Dai, S. L. Hale, R. Nayak, R. A. Kloner, Open Cardiovasc. Med. J. 2009, 3, 166-172.
a) O. Barret, J. Hannestad, C. Vala, D. Alagille, A. Tavares, M. Laruelle, D. Jennings, K. Marek,
D. Russell, J. Seibyl, G. Tamagnan, J. Nucl. Med. 2015, 56, 586-591; b) W. D. Heiss, K. Herholz,
J. Nucl. Med. 2006, 47, 302-312.
[15]
[16]
A. J. Hutchison, R. L. Webb, H. H. Oei, G. R. Ghai, M. B. Zimmerman, M. Williams, J. Pharmacol.
Exp. Ther. 1989, 251, 47-55.
a) Z. G. Gao, A. P. Ijzerman, Biochem. Pharmacol. 2000, 60, 669; b) H. Gutiérrez-de-Terán, A.
Massink, D. Rodríguez, W. Liu, Gye W. Han, J. S. Joseph, I. Katritch, L. H. Heitman, L. Xia, A. P.
Ijzerman, V. Cherezov, V. Katritch, R. C. Stevens, Structure 2013, 21, 2175-2185.
a) V. P. Jaakola, M. Griffith, M. A. Hanson, V. Cherezov, E. Y. T. Chien, J. R. Lane, A. P. Ijzerman,
R. C. Stevens, Science 2008, 322, 1211; b) F. Noël, F. M. do Monte, J. Pharmacol. Toxicol.
Methods 2017, 84, 51-56.
[17]
[18]
W. L. Yeo, X. Chew, D. J. Smith, K. P. Chan, H. Sun, H. Zhao, Y. H. Lim, E. L. Ang, Chem. Commun.
2017, 53, 2559-2562.
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700382
ChemBioChem
Entry for the Table of Contents
FULL PAPER
Phillip T. Lowe,[a] Sergio Dall'Angelo,[b]
Thea Mulder-Krieger,[c] Adriaan P.
IJzerman,[c] Matteo Zanda,[b] and David
O'Hagan*[a]
The synthesis and evaluation of a new
class of A2a adenosine receptor agonist,
designed around a 5'-fluorodeoxy-
adenosine scaffold, is reported. The
fluoromethyl group at the C-5' position
of this class allows for the
Page No. – Page No.
radiosynthesis of 18F-PET tracer
analogues via enzymatic fluorination of
chlorinated precursors with
A novel class of fluorinated A2A
adenosine receptor agonist with
application to last step enzymatic
[18F]fluorination for PET imaging
[18F]fluoride.
This article is protected by copyright. All rights reserved.