Synthesis and Pharmacological Evaluation of Identified and Putative Metabolites of the A1 Adenosine Receptor Antagonist 8-Cyclopentyl-3-(3-fluoropropyl)-1-propylxanthine (CPFPX)

ChemMedChem

10.1002/cmdc.201600592

Synthesis and Pharmacological Evaluation of Identified and Putative Metabolites of the A1

Adenosine Receptor Antagonist Cyclopentyl-3-(3-fluoropropyl)-1-propylxanthine (CPFPX)

1

2

1

Marcus H. Holschbach *, Dirk Bier , Wiebke Sihver , Annette Schulze , Bernd Neumaier

1

Institut für Neurowissenschaften und Medizin-Nuklearchemie (INM-5), Wilhelm-Johnen-Straße,

Forschungszentrum Jülich GmbH, D-52428 Jülich, Germany

2

Institut für Radiopharmazeutische Krebsforschung, Helmholtz-Zentrum Dresden-Rossendorf,

Bautzener Landstraße 400, D-01328 Dresden

*

Address for correspondence: Tel. +49 2461 61-6961; Fax +49 2461 61-2535; e-mail

m.holschbach@fz-juelich.de

Dedicated to Professor Dr. Ray A. Olsson, a good friend and excellent teacher, in

acknowledgement of his contributions to purine chemistry

1

ChemMedChem

10.1002/cmdc.201600592

Abstract

18

The A adenosine receptor (A AR) antagonist [ F]cyclopentyl-3-(3-fluoropropyl)-1-

1

8

propylxanthine ([ F]CPFPX), used in imaging human brain A ARs by PET, is stable in the brain

1

but rapidly undergoes transformation into one major (M1) and several minor metabolites in blood.

This report describes the synthesis of putative metabolites of CPFPX as standards for the

identification of those metabolites.

Analysis by (radio)HPLC revealed that extracts of human liver microsomes incubated with

1

8

n.c.a.[ F]CPFPX contained the major metabolite, M1, as well as radioactive metabolites

1

corresponding to derivatives functionalized at the cyclopentyl moiety, but no N -despropyl species

3

or metabolites resulting from functionalizations of the N -fluoropropyl chain.

3

The putative metabolites displaced the binding of [ H]CPFPX to the A AR in pig brain cortex at

1

3

K s between 1.9 and 380 nM and the binding of [ H]ZM 241385 to the A AR in pig striatum at

i

2A

1

K s greater than 180 nM. One metabolite, a derivative functionalized at the -position of the N -

i

propyl chain, showed high affinity (K 2 nM) to and very good selectivity (> 9000) for the A AR.

i

1

8

Keywords: A adenosine receptor ligand, [ F]CPFPX, in-vitro metabolism, binding properties,

1

ligand design

2

ChemMedChem

10.1002/cmdc.201600592

1

. Introduction

18

The radiofluorinated A adenosine receptor (A AR) antagonist [ F]8-cyclopentyl-3-(3-

1

18

fluoropropyl)-1-propylxanthine ([ F]CPFPX) (chart) is the lead diagnostic agent to image that

1

,2

receptor in human brain by means of positron emission tomography (PET). Although

18

[

F]CPFPX is not metabolized in the brain, it rapidly undergoes metabolism in the periphery to

3

,4

one main metabolite, M1 (chart), and a number of more polar metabolites.

Chart Structures of radiofluorinated CPFPX and the main metabolite M1.

The rate of metabolism (metabolic stability) of a drug indicates the need for structural

modifications to reduce metabolism. Metabolite identification followed by detection of soft spots

provides information on the sites to be modified or blocked and thus indicates possible chemical

modifications or even new scaffolds. Overall, the information obtained serves as a basis for taking

key decisions in the developmental phase of a ligand.

The aim of the present work was to synthesize and evaluate identified and putative CPFPX-

metabolites with regard to the development of metabolically stabilized second generation (follow

up) ligands. Thus, in the present study putative oxidative (phase I) metabolites of CPFPX were

synthesized to provide reference standards for the studies directed toward elucidating the

structures of the authentic drug metabolites found in human liver microsome preparations.

Pharmacological evaluation of this putative metabolite family was also performed to determine

receptor binding affinity and subtype selectivity in comparison to the parent compound.

3

ChemMedChem

10.1002/cmdc.201600592

18

Since [ F]CPFPX for human applications is prepared under no-carrier-added (n.c.a.) conditions,

19

the amounts of [ F]CPFPX administered are in the low-nanogram range, thus making direct

identification of the radioactive metabolites by LC-MS from human blood samples impossible. In

recent work it has been shown that human liver microsomes are a convenient and reliable model

4

for the Phase I metabolism of CPFPX.

A comparison by HPLC of extracts from human plasma obtained from blood samples drawn 30

min post tracer injection with extracts from human liver microsomes incubated with unlabelled

CPFPX for 30 min showed that the microsome extracts contained compounds having

chromatographic behavior similar to all of the radioactive metabolites found in human plasma.

LC-MS showed that they had m / z ratios of hydroxy-, oxo-, and unsaturated derivatives but did

4

not allow unambiguous structure determination of the metabolites. The compounds described in

the present work should allow to bridge that gap. In liver microsomes, CPFPX was shown to

undergo monohydroxylations on three positions of the cyclopentane ring. The synthesized

reference compounds should give detailed information about the alcohol regioisomers formed in

vitro.

5

Cytochrome P450-mediated oxidation in the liver is the main route of alkylxanthine metabolism,

3

and, presumably, that of CPFPX as well. Cytochrome P450-mediated oxidations of sp carbons

6

-8

such as those in the propyl and cyclopentyl groups of CPFPX have been described in detail.

Those consist of hydroxylations and oxidations of methyl and methylene groups except those

adjacent (alpha) to heteroatoms such as N, O or S. In the latter instance the initially formed

hydroxylation product is unstable and the alkyl group is eliminated by hydrolytic cleavage to form

the desalkyl xanthine. According to these rules, there are a number of plausible metabolites of

CPFPX. Table 1 shows some possible functionalization positions and -products. Besides the

already identified metabolites many of these are the targets of the syntheses described below.

4

ChemMedChem

10.1002/cmdc.201600592

Table 1. Possible CYP 450 functionalization products of CPFPX

+

m / z [M+H]

Species Represents

_

______________________________________________________________________________________________________________________________________________

species found in human blood

3

23

39

37

CPFPX

CPFPX-OH

CPFPX=O

intact ligand

sp -carbon hydroxylation

oxygenation

dehydrogenation

3

23 – 2 x n

unsaturated species

species not found human blood

3

2

3

54

81

62

CPFPX-COOH

oxidation of -CH₃₁

CPFPX-despropyl

CPFPX-desfluoropropyl

poly-OH species

poly=O species

dealkylation of N -propyl

3

dealkylation of N -fluoropropyl

multiple hydroxylation

23 + 16 x n

23 + 14 x n

multiple oxygenation

Since all the metabolites found in human blood contain radiofluorine and because there is no

3

significant radiofluorine deposition in bone, it appears that the N -fluoropropyl moiety of

18

[

F]CPFPX resists cleavage and further metabolism. The only carbon in the fluoropropyl chain

2

prone to functionalization is carbon C (beta to fluorine) to give the respective β-hydroxy / oxo

derivative, which LC-MS did not detect. Accordingly, any stable metabolites must contain

1

8

alterations of the N -propyl and / or C -cyclopentyl substituents. Analysis by LC-MS indicated

that these metabolites are exclusively Phase I but not Phase II metabolites, that is, they reflect

hydroxylation and oxidation but not conjugation.

5

ChemMedChem

10.1002/cmdc.201600592

With regard to the development of new potent A AR selective radioligands and with respect to

1

possible fine tuning of their lipophilicities the synthesized target compounds should be useful for

further development.

6

ChemMedChem

10.1002/cmdc.201600592

2

. Results and discussion

.1. Chemistry

Earlier work using extracts from human liver microsome preparations incubated with unlabelled

4

CPFPX had already identified CPFPX metabolites by HPLC and LC-MS analysis. Scheme 1

outlines the synthesis of 8-cyclopentyl-3-(3-fluoropropyl)-7-pivaloyloxymethylxanthine, 13, the

key intermediate for the synthesis of 1-substituted 8-cyclopentyl-3-(3-fluoropropyl)xanthines.

1

3

7

This synthesis exploited the different reactivities of xanthine N , N , and N toward alkylation,

which is 3 ≈ 7 >> 1. The first stages of the synthesis prepared 3-benzyl-8-cyclopentylxanthine, 6,

9

-11

in several steps from benzylurea, 1, and a cyanoacetic ester.

The synthesis of 6 differed in

1

2

some details from the approach used earlier. The condensation of benzylurea with methyl

cyanoacetate in sodium methoxide rather than with cyanoacetic acid in acetic anhydride cyclized

the cyanoacetylurea in one step rather than requiring a separate cyclization in alkali. Nitrosation

of 6-amino-1-benzyluracil 2 used isoamyl nitrite rather than NaNO and the acylation of 5,6-

2

diaminouracil 4 employed cyclopentane carboxylic acid, a water soluble carbodiimide (EDCI) and

4

-dimethylaminopyridine (DMAP) as a catalyst rather than cyclopentane carbonyl chloride and

Et N. Collectively, these modifications improved the overall yield of 6 to 40 %. Alkylation of 6

3

7

with one equivalent of chloromethyl pivalate (POM-Cl) for 3 hours protected N by

regioselectively forming the 7-pivaloyloxymethyl (POM) derivative, 7. Since this reaction is

apparently without literature precedent, it was carried out primarily to confirm the preferential

7

1

alkylation of N over N . As expected, using a 3-fold excess of POM-Cl and prolonging the

alkylation of 7 for additional 5 hours produced the 1,7-bis-POM derivative, 8. Alternatively,

alkylation of 6 overnight with 3.5 equivalents of POM-Cl gave 8 in one step. Debenzylation by

catalytic transfer hydrogenation (CTH) over Pd-C using ammonium formate as the hydrogen

13

donor generated the 1,7-bis-POM, 3-unsubstituted xanthine 9. Cleavage of both POM groups

gave 8-cyclopentylxanthine, 10. The regioselective introduction of a 3-fluoropropyl group through

alkylation of 9 with 1-bromo-3-fluoropropane formed the protected 3-(3-fluoropropyl)xanthine,

1

1. Alkaline hydrolysis removed the POM groups, giving one of the putative metabolites, 8-

7

ChemMedChem

10.1002/cmdc.201600592

7

cyclopentyl-3-(3-fluoropropyl)xanthine, 12. A short-term alkylation with POM-Cl protected N of

1

2 to give key intermediate 13.

Scheme 1 Synthesis of compound 10 and key intermediate 13

Scheme 2 summarizes the syntheses of the target xanthines. Alkylation of 13 with 3-bromo-1-

propanol gave the 1-(3-hydroxypropyl)xanthine 14a and, after deprotection, the 7-H xanthine,

1

5a. Alkylation of 13 with 2-(2-bromoethyl)-1,3-dioxolane and double deprotection gave the 1-

8

ChemMedChem

10.1002/cmdc.201600592

(3-oxopropyl)xanthine 15b. Alkylation of 13 with chloroacetone and deprotection of the product

afforded 1-(2-oxopropyl)xanthine 15c. Alkylation with chloroacetone followed by reduction with

NaBH and deprotection without isolation of intermediates yielded the 1-(2-hydroxypropyl)-7H-

4

xanthine, 15d, as a mixture of the R and S enantiomers. Alkylation of 1-propyl-8-cyclopentyl-7-

pivaloylmethylxanthine 16 with epifluorohydrin in the presence of a catalytic amount of base

1

4,15

gave, after deprotection, racemic 3-(3-fluoro-2-hydroxypropyl)-1-propylxanthine, 17.

Scheme 2 Synthesis of putative metabolites functionalized at the N-1 and N-3 propyl chains

Scheme 3 depicts the synthesis of the xanthines modified at the cyclopentyl ring. Since the

presence of double bonds in the cyclopentenyl moieties and an O-benzyl protecting group in the

cyclopentyl building block used in the synthesis of compounds 24a-c was incompatible with the

9

ChemMedChem

10.1002/cmdc.201600592

hydrogenation conditions necessary to remove an amide benzyl protecting group, the acid labile

1

6

2

,4-dimethoxybenzyl (DMB) protecting group was alternatively installed. Thus, carbodiimide /

DMAP induced condensation of the key intermediate 5,6-diamino-1-(2,4-dimethoxybenzyl)-3-

propyluracil 18 with with either cyclopent-2-enecarboxylic acid or the isomeric cyclopent-3-

enecarboxylic acid followed by alkali induced ring closure of amides 19a and 19b as described

1

6

for the synthesis of 6 furnished xanthines 20a and 20b. Xanthine 20c was obtained by the

reaction of diamine 18 with cis-2-benzyloxycyclopentane-1-carboxylic acid using the mixed

anhydride method followed by alkali induced cyclization of the intermediate amide 19c. Cis-2-

benzyloxycyclopentane-1-carboxylic acid was synthesized by benzylation of commercially

available cis-ethyl 2-hydroxyoxycyclopentane-1-carboxylate with benzyl-2,2,2-

1

7

trichloroacetimidate followed by ester hydrolysis.

7

POM protection of N in xanthines 20a-c yielded compounds 21a-c which were subsequently

3

deprotected at N by hydrolysis of the DMB groups with TFA to provide 22a-c. Alkylation with

1

-bromo-3-fluoropropane furnished derivatives 23a-c which upon alkaline hydrolysis of the POM

7

protecting groups at N gave the unsaturated target compounds 24a and 24b and the benzyl ether

1

8

2

4c. Alcohol 25 was synthesized by hydroboration of cyclopentene derivative 24a with 9-BBN,

followed by oxidation with hydrogen peroxide, as a mixture of the cis and trans isomers. The

regioisomeric cis-alcohol 26 was obtained by methanesulfonic acid induced debenzylation of 24c.

Scheme 3 Synthesis of putative metabolites functionalized at the cyclopentyl ring

1

0

ChemMedChem

10.1002/cmdc.201600592

1

ChemMedChem

10.1002/cmdc.201600592

Analysis of the xanthines by electrospray ionization-mass spectroscopy detected, in addition to

+

the [M + H] species, a second species of m / z 60 units greater. This appears to be an adduct of

+

the [M + H] species with acetic acid, added to the solvent to increase sensitivity. All the

1

7

xanthines but not their precursors formed these adducts, whether or not N and N were protected.

2

.2. Human liver microsomes

Human liver microsome stability is meanwhile a standard indicator for predicting in vivo

1

9,20

pharmacokinetic parameters of a compound.

To generate a chromatographic profile of

18

[

F]CPFPX metabolites formed in vitro, human liver microsomes were incubated with

1

8

n.c.a.[ F]CPFPX and analyzed by (radio)HPLC. The figure shows superimposed chromatograms

1

8

from the incubation of n.c.a.[ F]CPFPX with human liver microsomes (red radioactive trace) and

the same sample spiked with the potential metabolites (blue UV trace). Due to the minute amounts

1

8

of non radioactive CPFPX present in [ F]CPFPX preparations of low activity (± 500 KBq) the

radiolabeled isotopomer and its metabolites are not visible with UV detection at 275 nm. Thus,

simultaneous chromatography of the non radioactive and radioactive components allows the direct

and unambigous comparison between the synthesized compounds and the in vitro formed

metabolites in a single HPLC run.

1

2

ChemMedChem

10.1002/cmdc.201600592

Figure (Radio)HPLC chromatogram of a human liver microsome preparation extract after

1

8

metabolization of [ F]CPFPX. The extract was subsequently spiked with a mixture of the non-

radioactive putative metabolites 12, 15a-d, 17, 24a-b, 25 and 26 and analyzed in a single run.

Blue: UV trace, red: radioactive trace.

1

3

ChemMedChem

10.1002/cmdc.201600592

2

.2.1. Derivatives formed through multiple functionalizations

2

1

The most prominent metabolite belonging to this class has already been described earlier. It is

18

the main metabolite of [ F]CPFPX in human blood, M1, most likely arising from

dehydrogenation, double allylic oxygenation and loss of water from the intermediately formed

geminal diol (ketone hydrate) resulting in a highly conjugated double bond system.

2

.2.2. Unsaturated derivatives

Representatives of unsaturated metabolites containing double bonds in the alicyclic cyclopentane

4

ring are compounds 24a and 24b. Earlier MS investigations had identified these metabolites

+

with an exact mass of 322 [M+H] , which is 2 Da lower than CPFPX. A decrease of two mass

2

units corresponds to a dehydrogenation reaction. The inserts in the figure indicate the presence

of 24a and 24b, albeit to a small extent.

2

.2.3. Dealkylated derivatives

Compounds 10 and 12 belong to this group of metabolites. In contrast to the naturally occurring

5

methylxanthines where N-demethylation is the main metabolic pathway, no metabolites

1

8

originating from N-dealkylation of [ F]CPFPX could be found. Moreover, since compound 10

does not contain fluorine-18, it would not be detectable in the radioactive channel. The absence of

3

compound 12 makes the existence of compound 10 as well as that of the N -dealkylated congener

most unlikely.

2

.2.4. Oxidative derivatives

.2.4.1. Alkyl chains

Metabolites functionalized at the N-1-propyl and/or at the N-3-fluoropropyl side chains of the

xanthine core are compounds 15a-d and 17, respectively. Whereas the N-1-propyl derived

metabolites 15a and 15d are formed, the putative metabolites 17, originating from hydroxylation

1

4

ChemMedChem

10.1002/cmdc.201600592

2

3

of C in the N -fluoropropylchain, and ketone 15c are completely absent. These results are in

4

agreement with earlier findings. The observation that compound 17 is not detected might be

exploited as a potential strategy in drug design to mask alkyl chain metabolic hotspots through

2

3,24

incorporation of a fluorine atom at the terminal carbon (-fluorination).

2

.2.4.2. Cyclopentyl ring

The in vitro formation of oxidative metabolites corresponding to the functionalization of

2

5,26

cyclopentyl groups has scarcely been described in the literature.

In principle there are three chemically different carbon atoms in the cyclopentyl ring of CPFPX

1

2(5)

3(4)

prone to monohydroxylation, namely C , C , and C . MS studies using extracts of microsome

preparations incubated with unlabeled CPFPX have revealed the existence of three hydroxylated

4

regioisomers. Consequently ring-hydroxylated metabolites 25 and 26 have been synthesized. In

the first instance one would expect a chromatographic match of the UV-HPLC peaks of 25 and 26

with those of the radioactive trace. However, only compound 25 matched with a radioactivity

signal in the chromatogram in the figure. It is well known that stereoisomers are observed in

certain disubstituted (and higher substituted) cyclic compounds. Thus, substituted cycloalkanes

must be viewed as three-dimensional configurations in order to consider the spatial orientations of

the substituents. In non planar ring systems like cyclopentane substituents can adapt equatorial

and/or axial orientations resulting in configurational isomers (cis-trans isomers) which sometimes

can be separated by chromatographic methods. The reaction sequence for the synthesis of

compound 25 from the cyclopent-3-en precursor 24a (hydroboration-oxidation) yields a mixture

2

7-29

of cis and trans stereoisomers with a preference for the trans isomer (cis:trans ~ 20:80),

whereas the cis configuration of compound 26 is predetermined by the starting material cis-2-

hydroxycyclopentanecarboxylic acid. Cis-trans isomerism might explain the marginal difference

in the HPLC retention times of 26 (cis) and the putative radioactive metabolite (trans). This leads

to the assumption that the hydroxyl functions on the cyclopentyl moiety in metabolites 25 and 26

assume trans orientation.

These findings are in good agreement with the results of a study on the in vitro metabolism of the

25

cyclic nucleotide phosphodiesterase type IV inhibitor RP 73401. The authors clearly showed

1

5

ChemMedChem

10.1002/cmdc.201600592

that the predominant ring hydroxylated metabolite formed by human liver microsomes was the

trans alcohol, with the cis isomer constituting less than 3% of the amount of total

hydroxycyclopentyl metabolite formed.

2

.2.5. Binding studies

In binding studies (table 2) using pig brain membranes all examined compounds had lower

affinities (higher K values) for the A and A adenosine receptors than the parent compound

i

1

2A

CPFPX (K 0.47 nM). Differences in K values for the A AR ranged from 1.9 to 380 nM. The

i

1

most efficient deactivation was found for metabolite M1, the main metabolite in human blood,

showing a 800 fold reduction in affinity compared to CPFPX, but four compounds, 12, 15a, 15d,

and 17 showed higher subtype selectivity (A /A ) for the A AR. The four mentioned compounds

2

A

1

are exclusively functionalized at the N-alkyl chains of the xanthine heterocycle. Functionalization

at the cyclopentyl ring invariably led to a distinct reduction in subtype selectivity, with the main

metabolite M1 being the most unselective compound. Interestingly compound 15a has both a high

affinity for the A AR (K 2 nM) and a tenfold higher subtype selectivity (> 9000) compared to

1

i

CPFPX (800), making 15a a useful hydrophilic A AR ligand for the examination of this receptor

1

outside the brain. Moreover, because of its potentially low unspecific binding, 15a might be a

good candidate for in vitro experiments.

Table 2. Affinity of CPFPX metabolites 10, 12, 15a-d, 17, 24a-b, 25 and 26 at pig brain

membranes.

A AR (cortex)

1

A R (striatum)

2A

3

a

3

a

Cpd

[ H]CPFPX, K [nM]

[ H]ZM241385, K [nM]

selectivity, A /A

1

i

2A

CPFPX

0.47

375

(279 to 504)

798

21

(0.36 to 0.61)

M1

378

273 to 525)

7842

(2538 to 24229)

(

1

0

147

>100 000

>700

1

6

ChemMedChem

10.1002/cmdc.201600592

(81 to 265)

1

2

34

>100 000

>10 000

>3000

>9000

>600

>650

>1500

1020

303

(24 to 47)

5a

5b

5c

1.12

(0.97 to 1.28)

16.6

13.6 to 20.1)

(

15

(11 to 21)

(

±)15d

±)17

6.5

(4.8 to 8.8)

(

1.9

1939

(1021 to 3684)

(1.5 to 2.4)

2

4a

4b

2.8

847

(550 to 1305)

(1.1 to 2.9)

6.9

185

27

(2.8 to 17)

(133 to 258)

(

a

±)25

6.5

189

(117 to 306)

29

(3.7 to 12)

±)26

17.6

453

(381 to 507)

26

(13.4 to 22)

Mean and 95% confidence limits of at least three separate assays.

1

7

ChemMedChem

10.1002/cmdc.201600592

3

. Conclusion

1

8

This report describes the in vitro metabolic functionalization of the A AR antagonist [ F]CPFPX

1

in human liver microsome preparations. Ten putative metabolites were synthesized by the

classical Traube synthesis followed by a series of derivatization steps and evaluated as A - and

1

A adenosine receptor ligands. Metabolic profiling by co-chromatography of the synthesized

2

A

standards and extracts from human liver microsome incubations revealed, additionally to the

known main metabolite M1, the formation of one metabolite resulting from hydroxylation of

1

the N -propyl chain and two metabolites arising from oxygenation of the cyclopentyl substituent.

The present data suggest that monohydroxylations of the cyclopentyl moiety are stereoselective to

give the respective trans alcohols. Furthermore two minor metabolites, unsaturated in the

cyclopentyl ring, were formed.

The affinity data of the metabolites for the A - and A adenosine receptor indicate that enzymatic

1

2A

1

functionalization of either the pendent N -side chain or the cyclopentane ring results in the

formation of phase I metabolites with lower affinities to both receptor subtypes than the parent

compound CPFPX.

Two points can be made about the potential future significance of this study:



the metabolites are biologically active via a broad affinity range although less affine than

the parent compound CPFPX



the biologically highly active and subtype selective hydroxylated metabolite 15a provides

a convenient starting point for further drug development and the altered lipophilicity of

1

5a may offer advantages in terms of unspecific binding and its use as a radioligand.

1

8

ChemMedChem

10.1002/cmdc.201600592

4

. Experimental Section

Melting points were measured on a Büchi Melting Point B 545 apparatus (Büchi Labortechnik,

Flawil, Switzerland) and are uncorrected. Elemental analyses were performed by the Central

Division of Analytical Chemistry at the Forschungszentrum Jülich. Analyses indicated by the

symbols of the elements are within ± 0.4 % of the theoretical values (table 3). Low resolution

mass spectra (MS) were obtained in the ElectroSpray Ionisation (ESI+) mode on a Thermo

Finnigan Surveyor MSQ mass spectrometer (Thermo Fisher Scientific GmbH, Dreieich,

Germany). Thin layer chromatography (TLC) employed 0.2 mm precoated silica sheets (4 x 8 cm,

Alugram SIL G/UV254, Macherey-Nagel, Düren, Germany) developed with ethyl acetate / hexane

5

0 / 50 (v / v). Visualization was by quenching of UV fluorescence at 254 nm and iodine staining.

®

For flash chromatography a Grace Reveleris iES flash chromatography system equipped with

®

RevealX™ detection, allowing for multisignal (UV/ELSD) collection, and Reveleris flash silica

1

13

19

cartridges (size 40 µm) as stationary phase was employed. H, C, and F NMR spectra were

recorded at 200.13, 50.32, and 188.31 MHz, respectively, by means of a Bruker DPX-200 Avance

instrument (Bruker Bio Spin GmbH, Rheinstetten, Germany) in 5% solution at 25°C. Chemical

1

13

shifts are reported in ppm relative to the residual solvent signal ( H NMR: δ 7.26; C NMR: 77.0

1

13

for CDCl , H NMR: δ 2.52; C NMR: 50.32 for DMSO-d ). Multiplicities are reported by using

3

6

the following abbreviations: s; singlet, d; doublet, t; triplet, q; quartet, m; multiplet, br; broad.

n

Coupling constants J (in Hertz) for protons are given in the form J , those for carbons as J .

HX

CF

The protons were assigned with the aid of COSY, and the carbons with the aid of DEPT and

HMQC experiments. All solvents and reagents in the highest state of purity available were

obtained commercially (Sigma-Aldrich, Taufkirchen, Germany; Alfa Aesar, Karlsruhe, Germany)

and used as supplied by the vendor. Methanol for CTH was dried by refluxing over Mg turnings

and distillation. Ammonium formate was dried in a dessiccator over P O at room temperature

4

10

and ambient pressure. For CTH equal weights of substrate and 10% Pd-C and a 10-fold molar

excess of ammonium formate was used. Times and temperatures are indicated in individual

experiments. Purification of dimethylformamide (DMF) consisted of distillation from calcium

hydride and storage in a lightproof container over 4Å molecular sieves. Dichloromethane and

toluene were dried by storage over 4Å molecular sieves. Other solvents, chemicals and

biochemical reagents were obtained from and used as supplied by Sigma-Aldrich (Taufkirchen,

1

9

ChemMedChem

10.1002/cmdc.201600592

Germany). Pig brains for membrane preparations were from a local abattoir. The syntheses of

3

18

21,30

[

H]CPFPX and [ F]CPFPX were by the methods cited.

Reactions were carried out at

ambient temperature (± 23°C) under an atmosphere of argon unless otherwise stated.

Table 3. Elemental Analyses of New Compounds

Cpd

7

Formula

Carbon

Hydrogen

Nitrogen

Oxygen

C H N O Calcd

65.08

65.11

6.65

6.61

13.20

13.14

15.08

15.16

2

3

28

4

Found

8

9

C H N O Calcd

64.67

64.71

7.11

7.04

10.40

10.48

17.62

17.77

29

38

4

6

Found

C H N O Calcd

58.91

58.88

7.19

7.24

12.49

12.53

21.40

21.32

22

32

4

6

Found

1

2

3

C H FN O Calcd

59.04

58.97

7.33

7.29

11.02

10.94

18.88

19.01

25

37

4

6

Found

C H FN O Calcd

55.70

55.75

6.11

6.20

19.99

19.88

11.42

11.52

13

17

4

2

Found

C H FN O Calcd

57.85

57.87

6.90

6.92

14.20

14.16

16.22

16.27

19

27

4

Found

5a C H FN O Calcd

56.79

56.74

6.85

6.91

16.56

16.38

14.18

14.21

16

23

4

3

Found

5b C H FN O Calcd

57.13

56.98

6.29

6.14

16.66

16.62

14.32

14.24

16

21

4

3

Found

5c

C H FN O Calcd

57.13

57.21

6.29

6.23

16.66

16.72

14.32

14.24

16

21

4

3

Found

5d C H FN O Calcd

56.79

56.81

6.85

6.79

16.56

16.51

14.18

14.24

16

23

4

3

Found

6

7

C H FN O Calcd

56.79

56.81

6.85

6.87

16.56

16.48

14.18

14.22

16

23

4

3

Found

C H FN O Calcd

56.79

56.86

6.85

6.81

16.56

16.52

14.18

14.24

16

23

4

3

Found

2

0

ChemMedChem

10.1002/cmdc.201600592

1

2

9a C H N O Calcd

61.67

61.73

6.59

6.51

13.08

12.99

18.67

18.74

22

28

4

5

Found

9b C H N O Calcd

61.67

61.69

6.59

6.56

13.08

13.14

18.67

18.65

22

28

4

5

Found

0a C H N O Calcd

64.37

64.43

6.38

6.44

13.65

13.52

15.59

15.44

22

26

4

Found

0b C H N O Calcd

64.37

64.40

6.38

13.65

13.66

15.59

15.53

22

26

4

Found

0c

C H N O Calcd

67.16

67.03

6.61

6.57

10.80

10.74

15.43

15.48

2

9

34

4

5

Found

1a C H N O Calcd

64.10

63.99

6.92

6.81

10.68

10.52

18.30

18.19

28

36

4

6

Found

1b C H N O Calcd

64.10

64.04

6.92

6.88

10.68

10.63

18.30

18.24

28

36

4

6

Found

1c

C H N O Calcd

66.44

66.34

7.01

6.98

8.85

8.83

17.70

17.54

3

5

44

4

7

Found

2a C H N O Calcd

60.95

61.03

7.00

6.90

14.96

14.72

17.09

17.01

19

26

4

Found

2b C H N O Calcd

60.95

60.99

7.00

7.05

14.96

14.88

17.09

17.07

19

26

4

Found

2c

C H N O Calcd

64.71

64.64

7.10

7.14

11.61

11.52

16.58

16.66

2

6

34

4

5

Found

3a C H FN O Calcd

60.81

61.00

7.19

7.11

12.89

12.83

14.73

14.65

22

31

4

Found

3b C H FN O Calcd

60.81

60.77

7.19

7.16

12.89

12.96

14.73

14.81

22

31

4

Found

3c

C H FN O Calcd

64.19

64.07

7.24

7.19

10.32

10.33

14.74

14.69

2

9

39

4

5

Found

4a C H FN O Calcd

59.99

60.02

6.61

6.81

17.49

17.42

9.99

10.14

16

21

4

2

Found

2

1

ChemMedChem

10.1002/cmdc.201600592

2

4b C H FN O Calcd

59.99

6.61

6.57

17.49

17.53

9.99

10.06

16

21

4

2

Found

4c

5

C H FN O Calcd

64.47

64.37

6.82

6.69

13.08

12.93

11.20

11.19

23

29

4

3

Found

C H FN O Calcd

56.79

56.72

6.85

6.95

16.56

16.55

14.18

14.11

16

23

4

3

Found

6

C H FN O Calcd

56.79

56.87

6.85

6.89

16.56

16.53

14.18

14.26

1

6

23

4

3

Found

3

1

6

-Amino-1-benzyluracil (2).

A solution of sodium methanolate in methanol (30%, 357 mL, 2 mol) was diluted with methanol

to a final volume of 600 mL and heated to reflux. Under efficient stirring N-benzylurea (1, 100 g,

0

.66 mol) was added, followed by cyanoacetic acid methyl ester (66 g, 0.66 mol). The reaction

mixture was stirred at reflux for 6 h, cooled to room temperature and evaporated. Stirring in

boiling water (1400 mL) dissolved the syrupy yellow residue and the hot solution was carefully

acidified with acetic acid (CAUTION! VIGOROUS FOAMING!). After cooling the voluminous

precipitate was filtered off, washed with water and dried at 120°C. Recrystallization from 50%

aqueous ethanol gave an analytical sample. Yield: 89 g (62%), mp 286°C (50% EtOH).

1

H-NMR (DMSO-d ):  4.68 (s, 1H, C-5-H), 5.08 (s, 2H, Ph-CH ), 6.89 (s, 2H, NH ), 7.27 (m,

6

2

1

3

5

1

H, C H ), 10.69 (s , 1H, NH). C-NMR (DMSO-d ):  44.4 (Ph-CH ), 76.4 (C-5), 127.2, 128,

6

5

br

6

2

29.3 (Ph-C-4, C-2, C-3), 137.5 (Ph-C-1), 152.3 (C-2), 156.7 (C-6), 163.3 (C-4). MS (ESI+) m/z:

+

[

M+H] Calcd for C H N O 218.22; Found 218.23.

1

11

3

2

3

2

6

-Amino-5-nitroso-1-benzyluracil (3).

Crude 2 (4.34 g, 20 mmol) was suspended by stirring in ethanol (100 mL) and isoamyl nitrite (5

mL, 36 mmole) was added dropwise. Acidification with ethanolic HCl (2.5 N, 1 mL) changed the

color from yellow to violet. The reaction mixture was stirred for 0.5 h and cooled in the

refrigerator for 2 h. The microcrystalline solid collected by filtration was air-dried. Yield: 4.5 g

(92%), mp > 300°C (violet crystals).

1

H-NMR (DMSO-d ):  5.17 (s, 2H, Ph-CH ), 6.89 (s, 2H, NH ), 7.31 (m, 5H, C H ), 11.28 (s ,

6

2

6

5

br

3

13

4

2

3

1

H, N -H). C-NMR (DMSO-d ):  44.2 (Ph-CH ), 127.4, 128.3, 129.4 (Ph-C , C , C ), 135.5

6

2

ChemMedChem

10.1002/cmdc.201600592

1

5

2

6

4

+

(

Ph-C ), 139.5 (C ), 147.7 (C ), 149.8 (C ), 161.3 (C ). MS (ESI+) m/z: [M+H] Calcd for

C H N O 247.22; Found 247.23.

11

10

4

3

1

5

1

-Benzyl-5,6-diaminouracil (4).

Compound 3 (7.4 g 30 mmol) was dissolved in NH OH (10%, ca. 70 ml) at 50-60°C to give a

4

brownish solution. The addition of solid Na S O (10 g, 90 mmol) in small portions over 0.3 h

2

4

discharged the color. Stirring for 2 h longer at 50 - 60°C, followed by cooling, precipitated

greenish-grey product, which was filtered off by suction, washed with a small amount of ice-cold

water and dichloromethane, and used immediately in the next step. Yield 6.3 g (90%). MS (ESI+)

+

m/z: [M+H] Calcd for C H N O 232.24; Found 233.24.

1

12

4

2

N-(6-Amino-1-benzyl-2,4-dioxo-3-propyl-1,2,3,4-tetrahydropyrimidin-5-yl)-

1

5

cyclopentanecarboxamide (5).

Crude 4 (2.4 g, 10 mmol) was suspended in DMF (50 mL) and 1-(3-dimethylaminopropyl)-3-

ethylcarbodiimide hydrochloride (EDAC, 1.92 g, 10 mmol) and a catalytic amount (~ 150 mg) of

4

-dimethylaminopyridine (DMAP) was added. Under efficient stirring cyclopentanecarboxylic

acid (1.08 mL, 10 mmol) was added, and the mixture was stirred for 20 h at ambient temperature.

During this time the greenish suspension became a clear brown solution, and sometimes a yellow

precipitate formed. The mixture was poured onto ice-water (150 mL), the yellow precipitate was

collected by filtration, washed with water (50 mL), and oven-dried at 120°C. Yield 2.82 g (86%).

1

H-NMR (DMSO-d ): 1.46 – 1.84 (m, 8H, Cp-CH ), 2.77 (m, 1H, Cp-CH), 5.13 (s, 2H, Ph-

6

2

3

13

CH ), 6.44 (s, 2H, NH ), 7.34 (m, 5H, C H ), 8.22 (s, 1H, NHCO), 10.74 (s, 1H, N -H). C-NMR

2

6

5

(

DMSO-d ):  26.6 (Cp-CH ), 30.8 (Cp-CH ), 44.9 (Cp-CH), 53.6 (Ph-CH ), 89.2 (C ), 127.2,

6

2

4

2

3

1

2

6

4

1

28, 129.3 (Ph-C , C , C ), 137.5 (Ph-C ), 151.2 (C ), 153.5 (C ), 160.7 (C ), 176.9 (C=O). MS

+

(

ESI+) m/z: [M+H] Calcd for C H N O 329.37; Found 329.39.

1

7

20

4

3

1

5

3

-Benzyl-8-cyclopentyl-1H-purine-2,6(3H,7H)-dione (6).

Compound 5 was refluxed for 4 h in aqueous 2N NaOH (oil bath 115°C) and cooled to room

temperature. Bringing the clear brownish solution to pH 4 - 5 with concentrated HCl precipitated

a solid that was filtered off, washed with water and triturated with boiling ethanol to remove

2

3

ChemMedChem

10.1002/cmdc.201600592

coloured impurities. Filtration and oven drying gave almost colourless product. The xanthine was

recrystallized from acetic acid (± 25 mL / g), filtered, washed with ethanol and oven-dried. Yield

2

.43 g (91%), mp > 300°C. TLC: Rf 0.21.

1

H-NMR (DMSO-d ):  1.61 – 2.01 (m, 8H, Cp-CH ), 3.12 (m, 1H, Cp-CH), 5.09 (s, 2H, Ph-

6

2

1

7

13

CH ), 7.33 (m, 5H, C H ), 11.07 (s, 1H, N -H) 13.18 (s, 1H, N -H). C-NMR (DMSO-d ):  25.9

2

6

5

6

5

4

(

Cp-CH ), 32.8 (Cp-CH ), 39.6 (Cp-CH), 45.8 (Ph-CH ), 107.4 (C ), 128.2, 128.5, 129.3 (Ph-C ,

2 2 2

2

3

1

8

4

2

6

+

C , C ), 138.1 (Ph-C ), 150.1 (C ), 151.7 (C ), 155.1 (C ), 158.7 (C ). MS (ESI+) m/z: [M+H]

Calcd for C H N O 311.35; Found 311.32.

1

7

18

4

2

3

-Benzyl-8-cyclopentyl-7-pivaloyloxymethyl-1H-purine-2,6(3H,7H)-dione (7).

A slurry of 6 (3.1 g, 10 mmol), K CO (1.38 g, 10 mmol) and chloromethyl pivalate (POM-Cl,

2

3

1

.45 mL, 10 mmol) in dry DMF (40 mL) was stirred at 60°C for 3 h. The reaction mixture was

poured into ice water and neutralized with 2N HCl. Precipitated product was filtered off and

washed with water. The solid was treated with boiling EtOH (50 mL), the mixture was filtered,

and the solvent was evaporated in vacuo to give an oily yellow residue. Adding MeOH (± 5 mL),

scratching with a glass rod and cooling crystallized the product. The off-white solid was filtered

off, washed with a small volume of ice cold MeOH and air-dried. Yield: 3.7 g (92 %), mp: 174-

1

75°C (MeOH).

In some preparations traces of the 1,7-diPOM compound formed, hence the crude product mixture

was purified by flash cromatography eluting with 50% ethyl acetate in hexane. TLC: Rf 0.65.

1

7

H-NMR (DMSO-d ):  1.13 (s, 9H, N -POM-CH ), 1.64 – 2.11 (m, 8H, Cp-CH ), 3.40 (m, 1H,

6

3

2

7

1

Cp-CH), 5.09 (s, 2H, Ph-CH ), 6.20 (s, 2H, N -POM-CH ), 7.30 (m, 5H, C H ), 11.26 (s, 1H, N -

2

6

5

1

3

7

H). C-NMR (DMSO-d ):  25.9 (Cp-CH ), 27.4 (N -POM-CH )), 33 (Cp-CH ), 36.4 (Cp-CH),

6

2

3

2

7

5

3

9.1 (N -POM-C(CH ) ), 45.7 (Ph-CH ), 68.1 (N -POM-CH ), 106.9 (C ), 128.5, 129, 129.2 (Ph-

3 3 2 2

4

2

3

1

8

4

2

6

C , C , C ), 137.8 (Ph-C ), 150.3 (C ), 155.5 (C ), 155 (C ), 160.1 (C ), 177.4 (C=O). MS (ESI+)

+

m/z: [M+H] Calcd for C H N O 425.49; Found 425.51.

2

3

28

4

3

-Benzyl-8-cyclopentyl-1,7-(dipivaloyloxymethyl)-1H-purine-2,6(3H,7H)-dione (8).

A slurry of 7 (424 mg, 1 mmol), K CO (152 mg, 1.1 mmol) and POM-Cl (216 µL, 1.5 mmol) in

2

3

dry DMF (4 mL) was stirred at 60°C for 5 h. The reaction mixture was poured into ice water and

2

4

ChemMedChem

10.1002/cmdc.201600592

neutralized with 2N aqueous HCl. The turbid supernatant was decanted from the gummy

precipitate and the gum was dissolved in hot MeOH. Addition of water precipitated the product as

pale yellowish crystals. The product was recrystallized from 70% aqueous MeOH. Yield: 450 mg

(83%), mp: 128-129°C (70% MeOH). TLC: Rf 0.87.

1

7

H-NMR (DMSO-d ):  1.09 (s, 9H, N -POM-CH ), 1.13 (s, 9H, N -POM-CH ), 1.66 – 2.14 (m,

6

3

1

8

2

H, Cp-CH ), 3.40 (m, 1H, Cp-CH), 5.17 (s, 2H, Ph-CH ), 5.88 (s, 2H, N -POM-CH ), 6.22 (s,

2 2 2

7

13

H, N -POM-CH ), 7.33 (m, 5H, C H ), C-NMR (DMSO-d ):  26.2 (Cp-CH ), 27.4 (POM-

2

6

5

6

2

CH ), 27.5 (POM-CH ), 33 (Cp-CH ), 36.4 (Cp-CH), 39.1 (POM-C(CH ) ), 39.2 (POM-

3

2

3 3

5

C(CH ) ), 45.8 (Ph-CH ), 65.7 (POM-CH ), 68.2 (POM-CH ), 106.2 (C ), 128.5, 129, 129.30 (Ph-

3

2

4

2

3

1

8

4

2

6

C , C , C ), 137.31 (Ph-C ), 149.1 (C ), 151 (C ), 153.5 (C ), 161.2 (C ), 177.1 (C=O), 177.4

+

(

C=O). MS (ESI+) m/z: [M+H] Calcd for C H N O 539.64; Found 539.66.

2

9

38

4

6

NOTE: Compound 8 can also be prepared in one step in a similar yield by the reaction of 6 (1 eq)

with POM-Cl (3.5 eq) and K CO (3.5 eq) in DMF (4 mL / mmol xanthine) for 24 h at 60°C,

2

3

dilution with water, extraction with DCM, drying of the organic layer over Na SO , evaporation

2

4

and crystallizing the clear oil from 70% aqueous MeOH.

8

-Cyclopentyl-1,7-(dipivaloyloxymethyl)-1H-purine-2,6(3H,7H)-dione (9).

A 50 mL round-bottomed flask was charged with 8 (538 mg, 1 mmol), dry 10% Pd/C (538 mg)

and dry ammonium formate (631 mg 10 mmol). Dry MeOH (25 mL) was added and under

efficient stirring the flask was immersed in an oil bath preheated to 140°C. After 2h TLC

indicated the absence of starting material. The catalyst was removed by filtration over Celite™,

the filter cake was washed with DCM (3 x 20 mL) and acetone (3 x 20 mL) and the combined

filtrate and washings were evaporated in vacuo to give an analytically pure white solid in

quantitative yield, mp 138-139°C (MeOH). TLC: Rf 0.73.

1

7

H-NMR (DMSO-d ):  1.01 (s, 9H, N -POM-CH ), 1.06 (s, 9H, N -POM-CH ), 1.59 – 2.01 (m,

6

3

1

7

3

8

H, Cp-CH ), 3.43 (m, 1H, Cp-CH), 5.84 (s, 2H, N -POM-CH ), 6.18 (s, 2H, N -POM-CH ) N -H

2 2 2

1

3

appears as a broad signal between 3 and 5 ppm. C-NMR (DMSO-d ):  26.3 (Cp-CH ), 27.4

6

2

(

POM-CH ), 27.6 (POM-CH ), 32.9 (Cp-CH ), 36.30 (Cp-CH), 39.1 (POM-C(CH ) ), 39.2

3

2

3 3

5

8

4

2

(

POM-C(CH ) ), 65.6 (POM-CH ), 68 (POM-CH ), 105.9 (C ), 153.9 (C ), 155 (C ), 160.9 (C ),

3

2

5

ChemMedChem

10.1002/cmdc.201600592

6

+

1

77.2 (C ), 177.7 (C=O), 177.8 (C=O). MS (ESI+) m/z: [M+H] Calcd for C H N O 449.51;

22 32 4 6

Found 449.52

8

-Cyclopentyl-1H-purine-2,6(3H,7H)-dione (10).

A solution of 9 (448 mg, 1 mmol) in 7N methanolic NH (15 mL) was stirred for 48 h at ambient

3

temperature. Evaporation in vacuo gave a solid residue that was recrystallized from MeOH. Yield

2

15 mg (95%) colourless crystals, m.p. > 300°C (MeOH). TLC: Rf 0.00.

1

H-NMR (DMSO-d ):  1.61 – 2.03 (m, 8H, Cp-CH ), 3.09 (m, 1H, Cp-CH),3.60 (s , 1H, NH),

6

2

br

1

3

1

1.38 (s , 2H, 2 x NH). C-NMR (DMSO-d ):  25.9 (Cp-CH ), 32.7 (Cp-CH ), 49.5 (Cp-CH),

br

6

2

5

8

4

2

6

+

07.2 (C ), 149.9 (C ), 152.3 (C ), 156.1 (C ), 158.6 (C ). MS (ESI+) m/z: [M+H] Calcd for

C H N O 221.23; Found 221.22

10

12

4

2

8

-Cyclopentyl-1,7-(dipivaloyloxymethyl)-3-(3-fluoropropyl)-1H-purine-2,6(3H,7H)-dione

11).

A slurry of 9 (448 mg, 1 mmol), K CO (152 mg, 1.1 mmol) and 1-bromo-3-fluoropropane (137

(

2

3

µL, 1.5 mmol) in dry DMF (4 mL) was stirred for 24 h at 60°C. Pouring the reaction mixture into

ice water and neutralizing with 2N aqueous HCl precipitated the product as a gum that solidified

on standing. The solid was recrystallized from MeOH. Yield: 482 mg (95%), mp: 116-117°C

(MeOH). TLC: Rf 0.69.

1

7

H-NMR (DMSO-d ):  1.11 (s, 9H, N -POM-CH ), 1.12 (s, 9H, N -POM-CH ), 1.67 – 2.21 (m,

6

3

1

0H, Cp-CH + N -CH -CH ), 3.43 (m, 1H, Cp-CH), 4.13 (t, 2H, N -CH ), 4.51 (dt, J = 47.2, 5,7

2 2 2 2

1

7

13

Hz, 2H, CH -F), 5.88 (s, 2H, N -POM-CH ), 6.23 (s, 2H, N -POM-CH ). C-NMR (DMSO-d ):

2

6

2



26.2 (Cp-CH ), 27.4 (POM-CH ), 27.5 (POM-CH ), 29.5 ( J = 19.2 Hz, CH -CH F), 32.9 (Cp-

2 3 3 2 2

3

CH ), 36.4 (Cp-CH), 39.1 (POM-C(CH ) ), 39.1 ( J = 5.5 Hz, N -CH ), 39.2 (POM-C(CH ) ),

2

3 3

2

3 3

1

5

8

6

5.7 (POM-CH ), 68.1 (POM-CH ), 82.6 ( J = 162.0 Hz, CH -F), 106.2 (C ), 149.2 (C ), 151.1

2 2 2

4

2

6

19

(

C ), 159.6 (C ), 161.1 (C ), 177.1 (C=O), 177.5 (C=O). F-NMR (DMSO-d ):  – 219.16. MS

6

+

(

ESI+) m/z: [M+H] Calcd for C H FN O 509.58; Found 509.60.

2

5

37

4

6

8

-Cyclopentyl-3-(3-fluoropropyl)-1H-purine-2,6(3H,7H)-dione (12).

2

6

ChemMedChem

10.1002/cmdc.201600592

A solution of 11 (508 mg, 1 mmol) in a mixture of EtOH (10 mL) and aqueous NaOH (2N, 2 mL,

mmol) was stirred at room temperature for 8 h. Diluting the reaction mixture with water and

4

neutralization with acetic acid precipitated the deprotected purine. Collecting the product by

filtration, washing with water and drying furnished colourless crystals of 12. Yield 275 mg (98%),

mp 263-264°C (50% MeOH). TLC: Rf 0.23.

1

3

H-NMR (DMSO-d ):  1.63 – 2.02 (m, 10H, Cp-CH + N -CH -CH ), 3.14 (m, 1H, Cp-CH), 4.04

6

2

3

1

7

(

t, 2H, N -CH ), 4.48 (dt, J = 47.2, 5,7 Hz, 2H, CH -F), 10.98 (s, 1H, N H), 13.09 (s, 1H, N H).

2

1

3

2

C-NMR (DMSO-d ):  25.9 (Cp-CH ), 29.7 ( J = 19.2 Hz, CH -CH F), 32.8 (Cp-CH ), 39.6

6

2

3

1

5

8

(

Cp-CH), 39.7 ( J = 5.5 Hz, N -CH ), 82.7 ( J = 162.0 Hz, CH -F), 107.4 (C ), 150.1 (C ), 151.8

2 2

4

2

6

19

+

(

C ), 155.2 (C ), 158.6 (C ). F-NMR (DMSO-d ):  – 218.83. MS (ESI+) m/z: [M+H] Calcd

6

for C H FN O 281.30; Found 281.32.

1

3

17

4

2

8

-Cyclopentyl-7-pivaloyloxymethyl-3-(3-fluoropropyl)-1H-purine-2,6(3H,7H)-dione (13).

A slurry of 12 (2.8 g, 10 mmol) and K CO (1.38 g, 10 mmol) in dry DMF (40 mL) was stirred

2

3

for 0.5 h at ambient temperature. POM-Cl (1.53 mL, 10.5 mmol) was added dropwise and the

mixture was stirred at 60°C for 1 h. The reaction mixture was poured into ice water, neutralized

with 2N HCl and extracted with diethyl ether (2 x 100 mL). Evaporation of the solvent gave

semisolid material that was purified by flash chromatography eluting with 50% ethyl acetate in

hexane. The 1,7-diPOM compound (100 mg) eluted first, followed by product 13. Evaporation of

the solvent gave colourless crystals, 2.9 g (74%). An analytical sample was recrystallized from

aqueous 80% MeOH, mp: 146-147°C (80% MeOH). TLC: Rf 0.35.

1

7

3

H-NMR (DMSO-d ):  1.12 (s, 9H, N -POM-CH ), 1.64 – 2.12 (m, 10H, Cp-CH + N -CH -

6

3

2

3

CH ), 3.38 (m, 1H, Cp-CH), 4.04 (t, 2H, N -CH ), 4.51 (dt, J = 47.2, 5,7 Hz, 2H, CH -F), 6.19 (s,

2

7

1

13

2

H, N -POM-CH ), 11.16 (s, 1H, N -H). C-NMR (DMSO-d ):  26.1 (Cp-CH ), 27.4 (POM-

2

6

2

7

CH )), 29.6 ( J = 19.2 Hz, CH -CH F), 33 (Cp-CH ), 36.4 (Cp-CH), 39 (N -POM-C(CH ) ), 39.5

3

2

3 3

3

1

5

8

(

J = 5.5 Hz, N -CH ), 68 (POM-CH ), 82.7 ( J = 162.0 Hz, CH -F), 106.9 (C ), 149.85 (C ),

2

4

2

6

19

1

51.4 (C ), 155.1 (C ), 160.1 (C ), 177.2 (C=O). F-NMR (DMSO-d ):  – 218.96. MS (ESI+)

6

+

m/z: [M+H] Calcd for C H FN O 395.44; Found 395.45.

1

9

27

4

2

7

ChemMedChem

10.1002/cmdc.201600592

8

2

-Cyclopentyl-7-pivaloyloxymethyl-3-(3-fluoropropyl)-1-(3-hydroxypropyl)-1H-purine-

,6(3H,7H)-dione (14a).

A mixture of K CO (166 mg, 1.2 mmol) and 13 (394 mg, 1 mmol) in dry DMF (5 mL) was

2

3

stirred for 0.5 h, 1-bromo-3-hydroxypropane (134 µL, 1.5 mmol) was added and the suspension

was stirred at 70°C for 2 h. After cooling to room temperature the mixture was poured into ice

water (100 mL) and the aqueous solution was extracted with diethyl ether (2 x 75 mL). The

pooled organic extracts were washed with water (100 mL), dried over anhydrous Na SO , filtered

2

4

and evaporated to give a clear, colourless oil. The addition of 40% aqueous methanol (1 mL)

followed by scratching with a glass rod gave a semi-solid material which crystallized upon

standing overnight in an open vial. Filtration and drying in air gave 14a that was pure by TLC as

colourless, waxy crystals 450 mg, 98%, mp. 80-81°C. TLC: Rf 0.25.

1

7

1

H-NMR (CDCl ):  1.20 (s, 9H, N -POM-CH ), 1.58 – 2.31 (m, 12H, Cp-CH + N -CH -CH +

3

2

3

1

N -CH -CH ), 3.25 (m, 1H, Cp-CH), 3.53 (br s, 3H, CH OH + OH), 4.19 + 4.29 (2t, 4H, N -CH

2

3

7

13

+

N -CH ), 4.53 (dt, J = 47.2, 5,7 Hz, 2H, CH -F), 6.24 (s, 2H, N -POM-CH ). C-NMR

2 2 2

2

1

(

CDCl ):  26.2 (Cp-CH ), 27.3 (POM-CH ), 29.6 ( J = 19.2 Hz, CH -CH F), 31.2 (N -CH -

3

2

3

2

3

1

CH ), 33.1 (Cp-CH ), 36.7 (Cp-CH), 38.1 (POM-C(CH ) ), 40.7 ( J = 5.5 Hz, N -CH ), 40.8 (N -

2

3 3

2

1

5

8

CH ), 58.9 (CH -OH), 66.9 (POM-CH ), 82.3 ( J = 162.0 Hz, CH -F), 106.7 (C ), 148.75 (C ),

2

4

2

6

19

1

52 (C ), 155.5 (C ), 160.62 (C ), 177.7 (C=O). F-NMR (CDCl ):  – 220.48. MS (ESI+) m/z:

3

+

[

M+H] Calcd for C H FN O 453.52; Found 453.54.

2

33

4

5

8

-Cyclopentyl-3-(3-fluoropropyl)-1-(3-hydroxypropyl)-1H-purine-2,6(3H,7H)-dione (15a).

A solution of 14a (452 mg, 1 mmol) in methanolic ammonia (7N, 20 mL) was stirred for 72 h at

room temperature. Evaporation and recrystallization of the residue from 80% aqueous MeOH

yielded 15a as colorless crystals. Yield 285 mg (84%), m.p. 213-214°C (80% MeOH). TLC: Rf

0

.06.

1

3

H-NMR (DMSO-d ):  1.55 – 2.21 (m, 12H, Cp-CH + N -CH -CH + N -CH -CH ), 3.14 (m,

6

2

1

3

1

5

=

H, Cp-CH), 3.46 (m, 2H, CH OH), 3.93 + 4.11 (2t, 4H, N -CH + N -CH ), 4.58 (m + t, J = 47.2,

2 2 2

7

13

2

,7 Hz, 3H, CH -F + OH), 13.13 (s, 1H, N -H). C-NMR (DMSO-d ):  25.9 (Cp-CH ), 29.6 ( J

2

6

2

1

19.2 Hz, CH -CH F), 31.9 (N -CH -CH ), 32.8 (Cp-CH ), 39.2 (N -CH ), 39.6 (Cp-CH), 59.7

2

1

5

8

4

2

(

CH -OH), 82.8 ( J = 162.0 Hz, CH -F), 107.1 (C ), 148.4 (C ), 151.6 (C ), 154.8 (C ), 158.7

2

8

ChemMedChem

10.1002/cmdc.201600592

6

19

+

(

C ). F-NMR (DMSO-d ):  – 218.84. MS (ESI+) m/z: [M+H] Calcd for C H FN O 339.38;

6

16 23

4

3

Found 339.39.

8

-Cyclopentyl-3-(3-fluoropropyl)-1-(3-oxopropyl)-1H-purine-2,6(3H,7H)-dione (15b).

A mixture of 8-cyclopentyl-3-(3-fluoropropyl)-7-pivaloyloxymethylxanthine 13 (197 mg, 0.5

mmol) and K CO (83 mg, 0.6 mmol) in dry DMF (5 mL) was stirred for 0.5 h, then 2-

2

3

bromoethyl-1,3-dioxolane (88 µL, 0.75 mmol) and a catalytic amount of KI (25 mg) were added,

and the suspension was stirred at room temperature for 24 h. The mixture was poured onto ice-

water (100 mL) and the aqueous solution was extracted with diethyl ether (2 x 75 mL). The

pooled organic extracts were washed with water (100 mL), dried over anhydrous Na SO , filtered

2

4

and evaporated in vacuo to give a clear yellow oil. The oil was dissolved in EtOH (10 mL) and 2N

aqueous NaOH (1 mL, 2 mmol) was added. After stirring for 1 h at room temperature the mixture

was poured into ice-water (100 mL), acidified to pH 4 by the addition of 2N aqueous HCl, and

extracted with diethyl ether (2 x 100 mL). The pooled organic phases were dried over anhydrous

Na SO , filtered and the solvent evaporated in vacuo to give a colourless powder. This was taken

2

4

up in 0.5N aqueous HCl (10 mL) and the turbid mixture was stirred 2 h at reflux. After cooling

the solution was adjusted to pH 5-6 with 1N aqueous NaOH and precipitated 15b was extracted

into dichloromethane (2 x 25 mL). The pooled organic extracts were washed with water (50 mL),

dried over anhydrous Na SO , filtered and evaporated to leave a solid residue. Recrystallization

2

4

from 50% aqueous MeOH gave off-white crystals. Yield 148 mg (88%), m.p. 223-224°C (50%

1

3

1

MeOH. H-NMR (CDCl ):  1.51 - 2.19 (m, 10H, Cp-CH + N -CH -CH ), 2.83 (m, 3H, N -CH -

3

2

3

1

CH ), 3.24 (m, 1H, Cp-CH), 4.32 (t, 2H, N -CH ), 4.57 (dm, J = 47.2, 5,7 Hz, 4H, CH -F + N -

2

7

13

2

CH ), 9.87 (s, 1H, CHO), 12.28 (s , 1H, N -H). C-NMR (CDCl ):  25.8 (Cp-CH ), 29.5 ( J =

2

br

3

2

1

3

1

4

1

9.2 Hz, CH -CH F), 32.8 (Cp-CH ), 36.1 (N -CH ), 40.2 (Cp-CH), 40.9 ( J = 5.5 Hz, N -CH ),

2 2 2 2 2

1

5

8

4

2

3 (N -CH -CH ), 82.3 ( J = 162.0 Hz, CH -F),106.9 (C ), 149.6 (C ), 151.3 (C ), 155.6 (C ),

2

6

19

+

60.3 (C ), 200.4 (C=O). F-NMR (CDCl ):  - 220.50. MS (ESI+) m/z: [M+H] Calcd for

3

C H FN O 337.36; Found 337.38.

16

21

4

3

8

-Cyclopentyl-3-(3-fluoropropyl)-1-(2-oxopropyl)-1H-purine-2,6(3H,7H)-dione (15c).

2

9

Article Doi

Synthesis and Pharmacological Evaluation of Identified and Putative Metabolites of the A1 Adenosine Receptor Antagonist 8-Cyclopentyl-3-(3-fluoropropyl)-1-propylxanthine (CPFPX)

DOI: 10.1002/cmdc.201600592

Source and publish data:

Authors:

Article abstract of DOI:10.1002/cmdc.201600592

Full text of DOI:10.1002/cmdc.201600592

Products guided by the article

R&D Labs maybe for 200556-91-8

Relevant to this article

Hot Product