New irreversible adenosine A1 antagonists based on FSCPX

Article abstract of DOI:10.1016/S0960-894X(02)00639-X

FSCPX (1) and its amide analogue (2) have been reported to exhibit potent and selective irreversible antagonism of the A₁ adenosine receptor (A₁AR) when used in in vitro biological preparations. In order to obtain an irreversible A

Full text of DOI:10.1016/S0960-894X(02)00639-X

Bioorganic & Medicinal Chemistry Letters 12 (2002) 3179–3182
New Irreversible Adenosine A1 Antagonists Based on FSCPX
Anthony R. Beauglehole,a Stephen P. Bakerb and Peter J. Scammellsa,*
aDepartment of Medicinal Chemistry, Victorian College of Pharmacy, Monash University, 381 Royal Parade, Parkville,
Victoria 3052, Australia
bDepartment of Pharmacology and Therapeutics, University of Florida, PO Box 100267, Gainesville, FL 32610, USA
Received 23 April 2002; accepted 23 July 2002
Abstract—FSCPX (1) and its amide analogue (2) have been reported to exhibit potent and selective irreversible antagonism of the
A1 adenosine receptor (A1AR) when used in in vitro biological preparations. In order to obtain an irreversible A1AR antagonist
with improved stability, analogues of FSCPX incorporating the chemoreactive 4-(fluorosulfonyl)phenyl moiety separated from
the xanthine pharmacophore by a ketone linkage were explored. Compounds 4a–c exhibited improved affinity for the A1AR and
concentration-dependent irreversible binding to the A1AR.
# 2002 Elsevier Science Ltd. All rights reserved.
Adenosine receptors (AR’s) are distributed widely
throughout the mammalian body where they mediate
the many functions of the purine nucleoside
adenosine.1ꢀ3 These responses are mediated via the four
distinct AR subtypes, namely A1, A2A, A2B, and A3.4,5
These effects include release of neurotransmitters,6
inhibition of neuronal firing, inhibition of lipolysis, car-
diac depression and vasodilation,7 vasoconstriction and
immunosuppresant effects.3
compounds exhibited high affinity for the A1AR, high
selectivity for the A1AR over the A2AAR, and irrever-
sibly blocked the binding of [3H]-8-cyclopentyl-1,3-
dipropylxanthine ([3H]CPX) to the A1AR of DDT1
MF-2 cells (Table 1).13ꢀ16
Irreversible alkylating ligands are useful tools for the
study of receptor structure and function, including the
measurement of receptor reserve8ꢀ10 and the identifica-
tion and mapping of ligand-binding sites.11,12 Desirable
characteristics of irreversible ligands include high effi-
ciency of incorporation into the receptor allowing for
the control of receptor concentrations over a wide
range, a pharmacophore that provides affinity and
selectivity for the receptor and a reactive moiety that is
reactive enough to provide covalent incorporation into
the receptor but is stable enough that it can be used
under a wide variety of biological conditions.
Although compounds 1 and 2 bind to the A1AR irre-
versibly, the ester linkage of FSCPX (1) and the amide
linkage of 2 have the potential to be cleaved by esterase or
amidase enzymes, when used in vivo or in preparations
containing significant enzyme activity. Cleavage of the
fragment containing the reactive moiety would
obviously render the ligand inactive in terms of irrever-
sible binding. Consistent with this possibility is the
recent report indicating that FSCPX is rapidly degraded
in rat blood.17
Fluorosulfonyl substituted xanthines have been synthe-
sised where the chemoreactive 4-(fluorosulfonyl)phenyl
moiety is attached to N3 of the xanthine ring via ester
(FSCPX, 1)13 and amide (compound 2)14 linkages. All
In order to obtain an irreversible A1AR antagonist with
increased stability toward enzyme cleavage, we prepared
an analogue with an alkyl chain [–(CH2)5–] between the
4-(fluorosulfonyl)phenyl and the xanthine ring.14 While
*Corresponding author. Tel.:+61-3-9903-9542; fax:+61-3-9903-9582;
e-mail: peter.scammels@vcp.monash.edu.au
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00639-X

3180
A. R. Beauglehole et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3179–3182
Table 1. Ligand binding to the A1AR from DDT1 MF2 cells
Compd
IC50 versus [3H]DPCPX
(nM)
% Loss of specific [3H]DPCPX binding
(concn)b
KD [3H]DPCPX binding post-xanthine
(nMc)
1a
2a
4a
4b
4c
11.8ꢂ3.2
21ꢂ7
67ꢂ3 (5) 77ꢂ4 (100)
74ꢂ4 (5) 83ꢂ3 (100)
51ꢂ3 (5) 73ꢂ0.7 (100)
56ꢂ8 (5) 78ꢂ3 (100)
50ꢂ5 (5) 81ꢂ2 (100)
0.33ꢂ0.05 0.71ꢂ0.04
0.36ꢂ0.06 0.31ꢂ0.09
0.53ꢂ0.02 0.46ꢂ0.06
0.40ꢂ0.04 0.31ꢂ0.06
0.53ꢂ0.02 0.41ꢂ0.04
8.7ꢂ3.9
16.8ꢂ3.5
3.9ꢂ0.9
aData taken from Beauglehole et al.14
bDDT1 MF-2 cells were incubated without (control) and with the indicated xanthine derivative at the concentration (nM) in parentheses, and after
10cell wash cycles, membranes were prepared and assayed for [ 3H]DPCPX Bmax and KD. Control Bmax from individual experiments ranged from
267 to 593 fmol/mg protein.
cControl KD was 0.49ꢂ0.06 nM. Data are the meanꢂSE, n=3–6.
this linker was expected to be relatively inert, the com-
pound exhibited poor irreversible binding character-
istics. This was attributed to the electronic effect of the
linker on the reactivity of the fluorosulfonyl group to
substitution reactions. It was predicted that a ketone
linkage would be more resistant to enzymatic cleavage
and would possess similar electronic properties to
FSCPX (1) and its amide analogue 2. In order to maximise
pharmacophore affinity and selectivity for the A1AR, target
xanthines possessing 8-cyclopentyl, 8-cyclohexyl, and
8-(norborn-2-yl) substitution were targeted.18
resulted in successful coupling with concurrent hydro-
lysis of the fluorosulfonyl moiety to the corresponding
sulfonic acid. As the fluorosulfonyl moiety is relatively
stable in water (commonly recrystallised from aqueous
ethanol), formation of the sulfonic acid presumably
involved initial formation of the iodosulfonyl prior to
hydrolysis. This result was somewhat surprising as the
fluorosulfonyl proved to be stable to refluxing sodium
iodide in acetone. Therefore, it was predicted that for-
mation of the iodosulfonyl was occurring during con-
centration on workup. Successful alkylation was achieved
by pouring the reaction mixture into a solution of silver
nitrate to remove free iodide ions immediately after the
reaction was complete. Subsequent SEM deprotection with
aqueous hydrochloric acid afforded the target compounds
4a–c23ꢀ25 in good yield (52–89%).
3-Benzyl xanthines with the appropriate C8 substituent
(3a–c) were synthesised via literature methods.19,20
Optically pure (2S)-endo-norborn-5-ene-2-carboxylic
acid, prepared using methodology developed by Helm-
chen et al.21 was used in the synthesis of 3c. Scheme 1
depicts the synthetic route used to prepare targets with a
ketone linkage separating the 4-(fluorosulfonyl)phenyl
moiety from the xanthine ring. Alkylation with [2-(tri-
methylsilyl)ethoxy]methyl chloride (SEM-Cl) and
K2CO3 then protected N-7 of the 3-benzyl-8-cyclo-
alkyl-1-propylxanthines (3a–c, 45–97%). Subsequent
debenzylation at N3 with 10% palladium on carbon and
ammonium formate in refluxing methanol22 (40–89%)
provided a convenient attachment point for N3 sub-
stituents. In the case of compound 3c, reduction of the
alkene moiety of the 8-norborn-5-en-2-yl group pro-
ceeded concomitantly with the debenzylation. Attach-
ment of the chemoreactive N3 substituent with the alkyl
iodide 5 and potassium carbonate in DMF provided the
penultimate compounds in moderate yields (34–60%).
Initial attempts at N3 alkylation using an identical pro-
cedure to that used in the preparation of compound 214
(xanthine, alkyl halide, potassium carbonate and DMF)
Scheme 2 depicts construction of the N3 substituent
and formation of the ketone functionality via a highly
functionalised organocopper reagent. Following acti-
vation of the zinc with 4 mol% of 1,2-dibromoethane
and 3 mol% of trimethylsilyl chloride, a regioselective
zinc insertion occurs into the carbon iodide bond of
1-chloro-4-iodobutane.26 GC showed no evidence of
zinc insertion into the carbon-chlorine bond. The
organozinc iodide is then transformed into the orga-
nocopper species via a transmetallation reaction with
.
the THF soluble copper salt CuCN 2LiCl, to form
the zinc–copper reagent, tentatively represented as
RCu(CN)ZnX.26 The zinc–copper reagent displays an
enhanced reactivity toward organic electrophiles, and
reacts rapidly with acyl chlorides in THF at 0 ꢁC over
4–12 h to form ketones.26 Compound 6 was produced in
an overall yield of 54% with respect to 4-(fluor-
osulfonyl)benzoyl chloride. Alkyl chlorides of this type
failed to alkylate N3 under a variety of conditions. In
order to overcome this problem a halogen exchange was
performed on 6 to increase the reactivity of the alkyl
halide for subsequent substitution.
Scheme 2. Reagents and cꢁonditions: (i) Zn, THF, 45 ꢁC, 2.5 h; (ii)
FO2SC6H4COCl, ꢀ78 to 0 ꢁC, 4 h; (v) NaI, acetone.
Scheme 1. Reagents and conditions: (i) K2CO3, SEM-Cl, DMF; (ii)
Pd/C, NH4CO2H, MeOH; (iii) K2CO3, I(CH2)4C(O)C6H4SO2F (5),
DMF; (iv) aq HCl.
ꢁ
ꢁ
ꢁ
.
CuCN 2LiCl, ꢀ60 C to 0 C, 5 min; (iii) ꢀ78 C to 0 C, 4 h; (iv)

A. R. Beauglehole et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3179–3182
3181
Initially, efforts to synthesise the bromo analogue of 5
were focused on benzylic oxidation of 3-(4-(fluoro-
sulfonyl)phenyl)pentyl bromide due to its availability in
our laboratory. Attempts at benzylic oxidation with
selenium dioxide in dioxane27,28 and sodium bismuthate
in acetone29 returned starting material, while ceric
ammonium nitrate in acetic acid30 resulted in oxidation
of the benzylic position to the corresponding alcohol.
separately and joined prior to the final deprotection
step. All synthetic steps proceeded in moderate to
excellent yield, making this approach readily amenable
to gram scale preparation of the desired targets. The
xanthines 4a–c were found to be highly potent A1AR
antagonists that appear to bind to the receptor in an
irreversible manner. These characteristics are advanta-
geous in their use as irreversible ligands for the A1AR.
Furthermore, because 4a–c contain a biologically stable
ketone linker between the reactive moiety and the xan-
thine ring, it is likely that they will be useful for in vivo
studies to further elucidate the function and regulation
of the A1AR.
All target xanthines were tested for biological activity via
their ability to inhibit the binding of [3H]-8-cyclopentyl-
1,3-dipropylxanthine ([3H]DPCPX) to the A1AR of
DDT1 MF-2 cells as described in detail previously.14,31
Affinity was determined as the concentration of ligand
required to inhibit radioligand binding to the receptor
by 50% (IC50). Irreversible binding was determined by
incubating the cells with antagonist for 30min at 30 ꢁC
followed by 10cell washing cycles to remove unbound
ligand and then determining the binding maximum
(Bmax) and equilibrium dissociation constant (KD) for
the radioligand using cell membranes.32ꢀ35
Acknowledgements
The authors thank the Australian Research Council for
financial support and Assoc. Prof. Robert Singer (St
Mary’s University, Halifax) for helpful discussions con-
cerning the organozinc chemistry.
As previously reported, FSCPX (1) and its amide ana-
logue 2 have low nanomolar potency for binding to the
A1AR, and reduce the receptor content in a concentration
dependent manner.13,14 The reduction in receptor content
is consistent with 1 and 2 binding irreversibly to the
A1AR. Compounds 4a–c containing a ketone linkage
separating the 4-(fluorosulfonyl)phenyl moiety and
xanthine ring, exhibited high affinity for the A1AR as its
IC50 for the inhibition of [3H]DPCPX binding were 8.7,
16.8, and 3.9 nM for the 8-cyclopentyl, 8-cyclohexyl,
and 8-(2S-endo)-norborn-2-yl, respectively (Table 1).
These potencies are in the same range as that reported
for 1 and 2.14 Previously, it has been shown that the
linker between the 4-(fluorosulfonyl)phenyl and the
xanthine ring contribute to the ligand affinity for the
A1AR. Thus, an ester (1) and amide (2) in the linker
were about 10- and 5-fold more potent for the receptor
than an alkyl linker.14 The present data show that
replacement of the amide or ester with a more stable
ketone in the linker does not decrease the affinity of the
ligand for the receptor.
References and Notes
1. Daly, J. W. J. Med. Chem. 1982, 25, 197.
2. Jacobson, K. A.; van Galen, P. J. M.; Williams, M. J. Med.
Chem. 1992, 35, 407.
3. Poulsen, S.-A.; Quinn, R. J. Bioorg. Med. Chem. 1998, 6,
619.
4. Fredholm, B. B.; Abbracchio, M. P.; Burnstock, G.; Daly,
J. W.; Harden, T. K.; Jacobson, K. A.; Leff, P.; Williams, M.
Pharmacol. Rev. 1994, 46, 143.
5. Alexander, S. P. H.; Peters, J. A. Trends Pharmacol. Sci.
1998, 7.
6. Williams, M. Neurochem. Int. 1989, 14, 249.
7. Belardinelli, L.; Linden, J.; Berne, R. M. Prog. Cardiovasc.
Dis. 1989, 32, 73.
8. Venter, J. C. Mol. Pharmacol. 1979, 16, 429.
9. Posner, P.; Peterson, C. V.; Pitha, J.; Baker, S. P. Eur. J.
Pharmacol. 1984, 100, 373.
10. Nelson, C. A.; Muther, T. F.; Pitha, J.; Baker, S. P. J.
Pharm. Exp. Ther. 1986, 237, 830.
11. Birdsall, N. J. M.; Burgen, A. S. V.; Hulme, E. C. Br. J.
Pharmacol. 1979, 66, 337.
12. Dickinson, K. E. J.; Heald, S. L.; Jeffs, P. W.; Lefkowitz,
R. J.; Caron, M. G. Mol. Pharmacol. 1985, 27, 499.
13. Scammells, P. J.; Baker, S. P.; Belardinelli, L.; Olsson,
R. A. J. Med. Chem. 1994, 37, 2704.
14. Beauglehole, A. R.; Baker, S. P.; Scammells, P. J. J. Med.
Chem. 2000, 43, 4973.
15. Srinivas, M.; Shryock, J. C.; Scammells, P. J.; Ruble, J.;
Baker, S. P.; Belardinelli, L. Mol. Pharmacol. 1996, 50, 196.
16. Baker, S. P.; Scammells, P. J.; Belardinelli, L. Br. J.
Pharmacol. 2000, 130, 1156.
17. van Muijlwijk-Koezen, J. E.; Timmerman, H.; van der
Sluis, R. P.; van de Stolpe, A. C.; Menge, W. M. P. B.; Beu-
kers, M. W.; van der Graaf, P. H.; de Groote, M.; IJzerman,
A. P. Bioorg. Med. Chem. Lett. 2001, 11, 815.
18. Schimada, J.; Suzuki, F.; Hiromi, N.; Ishii, A. J. Med.
Chem. 1992, 35, 924.
Pretreatment of DDT1 cells with 5 or 100 nM of 4a, 4b,
or 4c reduced the receptor content of the cell mem-
branes by 50–56 and 73–81%, respectively, with no
change in the KD for [3H]DPCPX binding to the receptors
remaining. These data suggest that compounds 4a–c
produce a concentration-dependent decrease in the
A1AR of DDT1 cells by binding to the receptor in an
irreversible manner. The lack of change in the KD value
for [3H]DPCPX binding after pretreatment with the
higher concentration of 4a–c indicates that these com-
pounds wash out of the cells relatively easily. This is in
contrast to FSCPX (1) that has previously been reported
to be more resistant to washout.14 The ease of cell
washout for 4a–c provides a distinct advantage for these
irreversible antagonists as compared to FSCPX.
19. Papesch, V.; Schroeder, E. F. J. Org. Chem. 1951, 1879.
20. Speer, J. H.; Raymond, A. L. J. Am. Chem. Soc. 1953, 75,
114.
21. Poll, T.; Sobazak, A.; Hartmann, H.; Helmchen, G. Tet-
rahedron Lett. 1985, 26, 3095.
An efficient synthetic approach has been developed for
3-substituted xanthines of this type in which the xan-
thine and reactive moiety plus linker are constructed

3182
A. R. Beauglehole et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3179–3182
22. Holschbach, M. H.; Fein, T.; Krummeich, C.; Lewis,
R. G.; Wutz, W.; Schwabe, U.; Unterlugauer, D.; Olsson,
R. A. J. Med. Chem. 1998, 41, 555.
H70), 2.36 (br s, 1H, H40), 2.76 (br s, 1H, H10), 3.17 (t, 2H,
CH2CH2CH2CH2CO), 3.35 (m, 1H, H20), 4.01 (t, 2H,
CH2CH2CH3), 4.25 (t, 2H, CH2CH2CH2CH2CO), 8.11, 8.19
(2ꢃd, 4H, phenyl), 12.47 (br s, 1H, N7H), 13C NMR (CDCl3)
d 11.4, 20.5, 21.3, 24.1, 27.3, 29.2, 32.6, 37.1, 38.5, 40.2, 40.7,
42.5, 43.6 (ꢃ2), 106.6, 128.8, 129.0, 136.4 (d), 142.1, 147.5,
150.7, 155.2, 157.4, 198.1, Anal. (C26H31FN4O5S) C, H, N:
calcd, 58.85, 5.89, 10.56, found 58.79, 5.97, 10.20.
26. Knochel, P.; Jones, P.; Langer, F. In Organozinc Reagents.
A Practical Approach; Knochel, P., Jones, P., Eds.; Oxford:
New York, 1999; p 179.
1
23. Characterisation of 4a: H NMR (CDCl3) d 0.97 (t, 2H,
CHCH3), 1.64–1.99, 2.11–2.17 (m, 14H, CH2CH2CH3,
CH2CH2CH2CH2CO, CH2CH2CH2CH2CO, 4ꢃcyclopentyl
CH2), 3.18 (t, 2H, CH2CH2CH2CH2CO), 3.28 (tt, 1H, CH),
4.02 (t, 2H, CH2CH2CH3), 4.24 (t, 2H, CH2CH2CH2CH2CO),
8.12, 8.20(2 ꢃd, 4H, phenyl), 12.52 (br s, 1H, N7H), 13C NMR
(CDCl3) d 11.4, 20.6, 21.4, 25.5, 27.4, 32.5, 38.6, 40.0, 43.6,
43.8, 106.5, 128.9, 129.1, 136.5 (d), 142.2, 147.3, 150.7, 155.4,
158.7, 198.2, Anal. (C24H24FN4O5S) C, H, N: calcd, 57.13,
5.79, 11.1, found, 57.29, 5.94, 10.72.
27. Postowsky, J. J.; Lugowkin, B. P. Chem. Ber. 1935, 68,
852.
28. Liu, M.-C.; Lin, T.-S.; Cory, J. G.; Cory, A. H.; Sartorelli,
A. C. J. Med. Chem. 1996, 39, 2586.
1
24. Characterisation of 4b: H NMR (CDCl3) d 0.97 (t, 2H,
CH2CH2CH3),
1.23–2.05
(m,
16H,
CH2CH2CH3,
CH2CH2CH2CH2CO, CH2CH2CH2CH2CO, 50cyclohexyl
CH2), 2.86 (tt, 1H, CH), 3.17 (t, 2H, CH2CH2CH2CH2CO),
4.04 (t, 2H, CH2CH2CH3), 4.22 (t, 2H, CH2CH2CH2CH2CO),
8.12, 8.19 (2ꢃd, 4H, phenyl), 12.52 (br s, 1H, N7H), 13C NMR
(CDCl3) d 11.4, 20.5, 21.3, 25.6, 25.9, 27.2, 31.4, 38.4, 38.7,
42.9, 43.3, 106.5, 128.8, 129.0, 136.4 (d), 142.2, 148.7, 151.0,
29. Banik, B. K.; Venkatraman, M. S.; Mukhopadhyay, C.;
Becker, F. F. Tetrahedron Lett. 1998, 39, 7247.
30. Laing, S. B.; Sykes, P. J. J. Chem. Soc. C 2915, 1968.
31. Ramkumar, V.; Barrington, W. W.; Jacobson, K. A.;
Stiles, G. L. Mol. Pharmacol. 1989, 37, 149.
32. Bradford, M. M. Anal. Biochem. 1976, 72, 248.
33. Scatchard, G. Ann. N. Y. Acad. Sci. 1949, 51, 660.
34. Lopes, L. V.; Cunha, R. A.; Ribeiro, J. A. J. Neurochem.
1999, 73, 1733.
1
1
.
.
155.5, 159.7, 198.1, Anal. (C25H31FN4O5S 2H2O 2EtOAc) C,
H, N: calcd, 56.73, 6.35, 9.80, found, 56.73, 6.22, 9.99.
1
25. Characterisation of 4c: H NMR (CDCl3) d 0.98 (t, 2H,
CH2CH2CH3),
1.10–1.97
(m,
14H,
CH2CH2CH3,
35. Zhang, J.; Belardinelli, L.; Jacobson, K. A.; Otero, D. H.;
Baker, S. P. Mol. Pharmacol. 1997, 52, 491.
CH2CH2CH2CH2CO, CH2CH2CH2CH2CO, H30, H50, H60,