Synthesis and study of 5′-ester prodrugs of N6-cyclopentyladenosine, a selective A1 receptor agonist

Article abstract of DOI:10.1023/A:1011018730459

Purpose. A series of 5′-esters of N⁶-cyclopentyladenosine (CPA) were prepared with the aim to improve stability and bioavailability of selective A₁ agonists. Log P values, stability, affinity, and activity toward human adenosine A

Full text of DOI:10.1023/A:1011018730459

Pharmaceutical Research, Vol. 18, No. 4, 2001
Research Paper
scribed (3). These effects allow adenosine A1 agonists to pro-
duce ischemic tolerance and protection in neuronal and
cardiac tissues (4,5); moreover, selective A1 agonists have
been reported to increase survival of gerbils after ischemic
injury (6).
Synthesis and Study of 5؅-Ester
Prodrugs of N6-Cyclopentyladenosine,
a Selective A1 Receptor Agonist
Despite encouraging laboratory results on the central
nervous system, the side effects at other organ sites do not
allow the clinical use of receptor-subtype-selective adenosine
agonists (2,7). Selective A1 agonists appear poorly adsorbed
into the brain (8) and can be quickly degraded in vivo or in
whole blood: CPA is degraded in rat and human whole blood
with an half-life of approximately 20 minutes (9) and in con-
scious rats with an half-life of 7 minutes (10).
Alessandro Dalpiaz,1 Angelo Scatturin,1
Enea Menegatti,1 Fabrizio Bortolotti,1
Barbara Pavan,2 Carla Biondi,2 Elisa Durini,1 and
Stefano Manfredini1,3
The synthesis of esters is a general approach for devel-
oping alcohol prodrugs. This derivatization can provide an
increase of stability (11) and/or lipophilicity of the parent
compound, thus allowing latency and possibly improving the
diffusion through plasma membrane or lipid barriers. An ap-
proach aimed to study esters at the 5Ј-hydroxyl position has
been reported (12) and in this context, 5Ј-acetyl-CPA showed
very weak stability at physiological pH values.
We have recently demonstrated that susceptibility to es-
terases of nucleoside esters prodrugs is strongly dependent on
the bulk and lipophilicity of the ester substituent (13). We
have therefore investigated a prodrug approach for CPA with
the aim to enhance its stability in physiological fluids and thus
bioavailability.
Received December 10, 2000; accepted January 8, 2001
Purpose. A series of 5Ј-esters of N6-cyclopentyladenosine (CPA)
were prepared with the aim to improve stability and bioavailability of
selective A1 agonists. Log P values, stability, affinity, and activity
toward human adenosine A1 receptors were evaluated.
Methods. An appropriate synthetic procedure was adopted to avoid
concomitant deamination at position 6. Log P values were obtained
by the Mixxor system. The stability of CPA and its 5Ј-ester was
evaluated in human plasma and whole blood and analyzed with high-
performance liquid chromatography. The affinities to human A1 re-
ceptor expressed by N6-cyclohexyladenosine cells were obtained by
binding experiments. The activities were evaluated by measurements
of the inhibition of forskolin stimulated 3Ј-5Ј-cyclic adenosine mono-
phosphate, performing competitive binding assays.
Results. All prodrugs were more lipophilic than CPA, and their hy-
drolysis, in whole blood and in plasma, was found related, respec-
tively, to the length and hindrance of 5Ј-substituents. Affinity and
activity values indicated a very weak interaction toward adenosine A1
receptor of the intact prodrugs.
Conclusions. We propose 5Ј-esters of CPA, characterized by suitable
lipophilicity and elevated degree of stability in physiological fluids, as
possible canditates for CPA prodrugs.
MATERIALS AND METHODS
Materials
[3H]CHA (32.3 Ci/mmol) was obtained from NEN Re-
search Products (Boston, MA, USA). [3H]c-AMP (24 Ci/
mmol) was obtained from Amersham, Amity Srl (Milan,
Italy). CPA, N6-cyclohexyladenosine (CHA), and 1,3-
dipropyl-8-cyclopentylzanthine (DPCPX) were purchased
from RBI (Natick, MA, USA). ADA, c-AMP, Ro 20-1724,
and forskolin were obtained from Sigma (St. Louis, MO,
USA). HPLC grade solvents were purchased from Carlo
Erba Reagenti (Milan, Italy). CHO cells transfected with
adenosine A1human receptors (CHO A1) were a kind gift of
Prof. Peter Schofield (Garvan Institute of Medical Research,
Darlinghurst, Australia) (14,15).
KEY WORDS: N6-cyclopentyladenosine; adenosine A1 receptor;
prodrugs; stability; affinity; activity.
INTRODUCTION
Adenosine exerts its physiological effects by interaction
with membrane receptors, which have been classified into
four subtypes (1). A1activation has been found to produce
cardiac and neuronal excitability depression (2), and the abil-
ity of the prototype A1-selective agonist 5Ј-esters of N6-
cyclopentyladenosine (CPA) to inhibit ischemia has been de-
Chemistry
Melting points were determined with a Kofler apparatus
and are uncorrected. Utraviolet spectra were recorded on a
Kontron UVIKON 922 spectrometer. Reaction courses were
routinely monitored by thin-layer chromatography (TLC) on
silica gel precoated Durasil-25 UV254 Macherey–Nagel plates
with detection under 254 nm UV lamp and/or by spraying the
plates with 10% H2SO4/MeOH and heating. Nuclear mag-
netic resonance (1H-NMR) spectra were determined in
DMSO-d6 or CDCl3 solution with a Bruker AC-200 spec-
trometer and chemical shifts are given in ppm from internal
tetramethylsilane as a standard. Matrix-assisted laser desorp-
tion ionization time-of-flight (MALDI-TOF) spectra were
obtained on a Hewlett–Packard HPG2025A mass spectrom-
eter operating in a positive linear mode. Column chromatog-
raphy was performed with Macherey-Nagel 70–230 mesh
silica gel.
1 Department of Pharmaceutical Sciences, via Fossato di Mortara
17-19, Ferrara University, I-44100 Ferrara, Italy.
2 Department of Biology, General Physiology Section, via Borsari 46,
Ferrara University, I-44100 Ferrara, Italy.
3
To whom correspondence should be addressed. (e-mail:
stefano.manfredini@dns.unife.it)
ABBREVIATIONS: ADA, adenosine deaminase; c-AMP, 3Ј,5Ј-
cyclic adenosine monophosphate; CHA, N6-cyclohexyladenosine;
CPA, N6-cyclopentyladenosine; CNS, central nervous system;
DPCPX, 1,3-dipropyl-8-cyclopentylxanthine; 5Ј-CH-CPA, 5Ј-
cyclohexanoyl-CPA; 5Ј-PE-CPA, 5Ј-pentanoyl-CPA; 5Ј-Oc-CPA, 5Ј-
octanoyl-CPA; Ro 20-1724, 4-(3-butoxy-4-methoxybenzyl)-2-
imidazolidinone; TMOB-CPA, trimethoxybenzoyl-CPA.
531
0724-8741/01/0400-0531$19.50/0 © 2001 Plenum Publishing Corporation

532
Dalpiaz et al.
Preparation of N6-Cyclopentyl-5؅-O-Acyl-Adenosine
Kinetic Experiments in Human Plasma and Whole Blood
Derivatives (4a-d)
The compounds were incubated at 37°C, 3 ml of plasma
or whole blood were spiked with drug solutions resulting in
final concentrations of 2 ␮M. At regular time intervals 100 ␮l
of samples were withdrawn. Plasma or whole blood samples
were quenched in 200 ␮l of ice-cold ethanol or 500 ␮l of
ice-cold water, respectively, and stored at −20°C until analy-
sis.
Procedure A (3c-d)
Compound 2 (12) (400 mg, 1 mmol) and 4-DMAP (360
mg, 3 mmol) were dissolved in 15 ml of dry CH2Cl2 and the
appropriate chloride (2.3 mmol) was added drop-wise under
vigorous stirring and the reaction mixture was stirred at room
temperature for 3 hours.
CPA and Prodrugs Kinetic Analysis
Procedure B (3a)
Degradation of all the compounds was assessed using
reversed-phase high performance liquid chromatography
(RP-HPLC) separation. As for plasma samples, 50 ␮l of
4 и 10−5 M internal standard (CHA for CPA and 5Ј-PE-CPA;
5Ј-PE-CPA for 5Ј-Oc-CPA, 5Ј-CH-CPA and 5ЈTMOB-CPA)
were added. After 5 minutes of centrifugation at 9,000 g, 250
␮l of solution were reduced to dryness under a nitrogen
stream. Two hundred microliters of mobile phase were added
and, after centrifugation, 40 ␮l were injected into the HPLC
system. As for whole blood samples, 50 ␮l of 4 и 10−5 M in-
ternal standard and 50 ␮l of 3M sodium hydroxide for CPA
analysis or 50 ␮l of 10% sulfosalicylic acid for analysis of
prodrugs, were added. The samples were extracted twice with
1 ml of water-saturated ethyl acetate. After centrifugation,
the organic layer was reduced to dryness and after 100 ␮l
mobile phase addition, 40 ␮l were injected into the HPLC
system. This was from Jasco and consisted of a piston pump
and a variable wavelength UV detector set at 269 nm. Chro-
matography was performed at room temperature on a re-
versed-phase column (Hypersil BDS C-18 5U cartridge col-
umn, 150 mm × 4.6 mm I.D.; Alltech Italia Srl BV, Milan,
Italy). The mobile phase consisted of a ternary mixture of
acetonitrile, methanol and 10 mM acetate buffer (pH 4) with
a ratio of 4/40/56 (v/v/v) for CPA analysis and of 4/50/46 for
analysis of prodrugs . The flow rate was 1.2 ml/min. The re-
tention times of CPA, CHA, 5Ј-Pe-CPA, 5Ј-Oc-CPA, 5Ј-CH-
CPA and 5Ј-TMOB-CPA were 4, 7, 5, 20, 9, and 4 minutes,
respectively.
In the case of the valeroyl derivative the starting com-
pound was dissolved in 15 ml of anhydrous DMF and valeric
anhydride (269 ␮l, 1.34 mmol) was added drop-wise under
vigorous stirring, the reaction mixture was stirred at room
temperature for 20 hours.
When the reaction was complete (TLC: diethyl ether/
hexane 8:2), the mixture was diluted with CH2Cl2 and then
washed with water (3 × 100 ml); the organic phase, dried over
MgSO4, was evaporated to give 650–800 mg of a crude green
oil (3a-d). Compounds 3a-d were used without any further
purification for the next deprotection step, the crude com-
pounds were dissolved in a CHCl3/MeOH solution (24 ml, 1:1
mixture), and 4% TFA/H2O solution was added (320 ml, 32
ml of TFA solution every 0.1 theoretic mmol of expected
product) at 0°C. The reaction was stirred at 4°C for 72 hours
(TLC: AcOEt/hexane 8:2). The solvent was then evaporated,
below 30°C, and the residue was purified by flash chromatog-
raphy on silica gel (EtOAc/hexane linear gradient from 6:4 to
8:2). After concentration of the desired fractions the desired
compounds were obtained in 40–60% overall yield.
Compound 4a
Pale yellow foam; 253 mg, 60% yield; UV (MeOH) ␭
max
269 nm (␧: 9000), ␭
nm 232 (␧: 1900), ␭
212 nm (␧:
min
10100); MALDI MS (M + H)+421 Da.
max
Compound 4b
Pale yellow foam; 180 mg, 40% yield; UV (MeOH) ␭
max
269 nm (␧: 6500), ␭min240 nm (␧: 2500), ␭max213 nm (␧:
Log P Determination for CPA and Prodrugs
11200); MALDI MS (M + H)+463.7 Da.
Partition coefficients were determined by using
MIXXOR separator cylinders (Genex, Industrial Drive, Gai-
thersburg, MD, USA). After 40 strokes, the residual amounts
of the compounds in the aqueous phase were monitored by
injection (40 ␮l) in the same HPLC system previously de-
scribed. For CPA and 4c an octanol/water 1/1 ratio (5 ml each
one) was used. For compounds 4a and 4d an octanol/water 1/3
ratio (2.5 and 7.5 ml) was used, whereas for compound 4b an
octanol/water 1/9 ratio (1 and 9 ml) was used.
Compound 4c
White foam, 250 mg, 56% yield; UV (MeOH) ␭max269
nm (␧: 14800), ␭min 232 nm (␧: 2100), ␭max 214 nm (␧: 13900);
MALDI MS (M + H)+446.9, (M + Na)+ 468.9 Da.
Compound 4d
White solid, mp: 88–93°C, 270 mg, 53% yield; UV
(MeOH) ␭max 267 nm (␧: 14500), ␭min 235 nm (␧: 4400), ␭
max
217 nm (␧: 19600); MALDI MS (M + H)+531.1 Da. The NMR
data for compound 4a are given as an example below
Receptor Binding Assays
Cells were grown and membrane prepared as previously
described (16). Membrane aliquots containing 40 ␮g of pro-
Compound 4a
1H NMR (DMSO-d6) ␦ 8.29 (s, 1H, H8); 8.19 (s, 1H, H2); teins were incubated in 400 ␮l of 50 mM TRIS-HCl at 25°C
7.74 (d, J ס 7.83 Hz 1H, NH); 5.9 (d, J ס 4.63 Hz, 1H, H1Ј); for 105 minutes. Saturation experiments were carried out us-
5.58 (d, J ס 5.68 Hz, 1H, OH2); 5.37 (d, J ס 5.6 Hz, 1H, ing twelve different concentrations of [3H] CHA ranging from
OH3); 4.72–4.65 (m, 1H, CHNH); 4.45–4.05 (m, 5H, H2Ј, H5Ј, 1 to 100 nM. Displacement experiments on rat brain mem-
H3Ј, H5ЈЈ, H4Ј,); 2.26 (t, J ס 7.15 Hz 2H, CH2CO); 1.95–1.91, branes were performed in the presence of 20 nM [3H]CHA.
1.70–1.40 (m, 8H, cyclopentyl); 1.27–1.16 (m, 4H, 2xCH2); Non-specific binding was measured using 10 ␮M DPCPX.
0.81 (t, 3H, CH3, J ס 7.09 Hz).
Separation of bound from free radioligand was per-

Adenosine A1 Receptor, Agonist Prodrugs
533
formed as previously described (16). The stability of the pro- Log P
drugs in binding conditions have been evaluated incubating
for 105 minutes each compound (10 ␮M final concentration)
in 400 ␮l of TRIS-HCl, pH 7.4, at 25°C in the presence of one
membrane aliquot containing 40 ␮g of proteins. Forty micro-
liters of filtered solution where injected into the HPLC sys-
tem.
Peak areas were quantified using the integrator which
was calibrated with standard solutions of pure compounds.
Log P values have been calculated as
n + C ͒ − C
͑
o
w
Log P =
Cw
where C0 and Cw indicate the drug concentration before and
after partitioning, respectively n ס 1 for CPA (1) and 4c, n ס
2.5 for compounds 4a,d and n ס 10 for compound 4b (18).
Determination of Cyclic AMP Levels
Cells grown on 12-well plates were preincubated for 10
minutes at 37°C in fresh serum-free F12/DMEM containing 1
IU/ml of ADA and 10 ␮M Ro 20-1724 as phosphodiesterase
inhibitor. Then, eight different concentrations of CPA or pro-
drugs were added. After 5 minutes, forskolin was included (1
␮M final concentration, 500 ␮l final volume) The reaction
was stopped after 5 minutes by rapid removal of the medium
and the addition of 0.3 ml of ice-cold 0.1 N HCl to each well;
cells were then scraped, collected, and frozen at −80°C until
cAMP assay. The samples were centrifuged for 10 min at
2,000 g, and the supernatants were neutralized with 0.06 ml of
0.5 M Trizma base and assayed for cAMP content, as re-
ported by Brown et al. (17).
The stability of the prodrugs in c-AMP measurement
conditions have been evaluated incubating each compound
(10 ␮M final concentration) in 500 ␮l of serum-free F12/
DMEM on 12-well plates for 10 minutes in the presence of
confluent cells. 40 ␮l of filtered solution where injected into
the HPLC system.
Binding and c-AMP Assays
Data of binding experiments were obtained as previously
described (16). IC50 values in the c-AMP assay were obtained
by computer analysis of the concentration-inhibition curves
(Graph Pad Prism, San Diego, CA, USA).
RESULTS
Chemistry
According to a published synthetic strategy (12), CPA
(1) was reacted with triethylorthoformate in dioxane to give
the protected N6-cyclopentyl-adenosine (2) in 94% yield (Fig.
1). Next treatment of (2) with the appropriate acyl-chloride in
CH2Cl2 and DMAP (19) gave compounds 3b-d, whereas the
5Ј pentanoyl ester 3a was better obtained using the corre-
sponding anhydride in presence of a catalytic amount of
4-DMAP (12). The removal of the protecting group, was per-
formed on the crude ester derivatives, under mild conditions,
by using a 1:1 mixture of CHCl3/MeOH and 4% TFA to give
the final products 4a-d in good yields (two steps 40–60%).
Calculations
Biology
Table I reports the half life (t1/2) values of CPA (1) and
its prodrugs, obtained by regression analysis of semilogarit-
mic plots reported in Figure 2. No significant CPA (1) deg-
radation were observed in plasma, whereas t1/2 is about 15
minutes in whole blood, where adenosine kinase catalyzes the
phosphorilation of the 5Ј-hydroxyl group of CPA (9).
Kinetics
The peak areas in the chromatograms were quantified
using a chromatographic data processor (Chromatopac
C-R3A, Shimadzu, Kyoto, Japan). The half-life of each aden-
osine analogue was calculated from a semi-logarithmic plot of
the peak area ratio between the compound and internal stan-
dard, expressed as percentage, versus incubation time.
As for prodrugs, the hydrolysis rate of 5Ј-esters appeared
to be notably modulated by the substituents, the t1/2 values
Fig. 1. Synthesis of compounds 4a-d.

534
Dalpiaz et al.
Table I Kinetic (t1/2), Hydrophobic (Log P), Affinity (Ki), and Activity (IC50) data of CPA and Prodrugsa
t1/2plasma
(min)
t1/2 whole
blood (min)
Percent
Percent
Ligand
Log P
Ki (nM)
hydrolysisb
IC50 (nM)
1.5 0.1
hydrolysisc
1
n.s.
14
0
3
1.21
2.79
3.91
2.95
2.27
2.9 0.2
586 41
176 14
941 72
4260 230
–
n.s.
4a
4b
4c
4d
18
28
45
3
4
5
1.14 0.06
2.30 0.09
0.33 0.01
n.s.
70
76
82
4
5
6
2.59 0.09
3.97 0.16
1.56 0.08
n.s.
30
31
n.s.
4
4
n.s.
2590 120
a Stability data of compounds under experimental condition for affinity and c-AMP levels measurements are also reported. n.s. ס not
significant within 3 hours.
b Experimental binding condition.
c Experimental condition for c-AMP levels determination.
obtained for 5Ј-Pentanoyl-CPA (4a) in plasma (18 minutes) drugs showed intermediate affinity and % hydrolysis values
dramatically decreased in whole blood: after 1 minute, the as well (Table 1).
HPLC analysis of the blood sample showed complete disap-
All prodrugs appear able to fully inhibit forskolin-
pearance of the prodrug with the concomitant appearance of stimulated intracellular c-AMP levels, acting as full agonists;
CPA, that followed, upon 45 minutes, the characteristic deg- however, their potency is lower with respect to that of CPA
radation pattern. The t1/2 values obtained for 5Ј-Octanoyl- (Fig. 4). The IC50 value of CPA is 1.5 nM (Table 1), whereas
CPA (4b) in plasma and whole blood are around 29 min. A for prodrugs the values ranges from 70 to 2590 nM, the
similar t1/2 value (31 minutes) has been found for 5Ј- weaker compound is not hydrolysed (5Ј-TBMOB-CPA,
cyclohexanoyl-CPA (4c) in whole blood, whereas in plasma IC50ס 2590 nM), and the other three compounds (4a-c) show
this compound showed higher stability (t1/2 ס 45 minutes). comparable hydrolysis (1.56–3.97%) and thus comparable
Finally, the presence of the trimetoxyphenyl ring allowed us IC50 values (70–82 nM).
to obtain a compound with high stability. In fact, degradation
of TMOB-CPA (4d) was not significant within 3 hours. All
prodrugs are more hydrophobic than CPA (LogP ס 1.21)
with Log P values ranging from 2.27 to 3.91 (Table 1). [3]CHA
DISCUSSION
A series of 5Ј-esters of CPA have been prepared with the
was used as A1 selective radioligand for inhibition experi- aim to evaluate their chemico-physical properties and stabil-
ments, KD and BMAX values were 24 2 nM and 995 11 ity in physiological fluids. An appropriate synthetic procedure
fmol/mg protein. Figure 3 illustrates the inhibition experi- has been adopted in order to avoid concomitant deamination
ments of CPA and its prodrug. As reported in Table 1, the Ki at position N6. Log P values of CPA and its 5Ј-derivatives
value of CPA is 2.9 nM, according to the results obtained have been determined to evaluate the influence of the lipo-
from [3H]DPCPX inhibition experiments (16). Ki values of philicity on the capability to give rise to increased plasma and
prodrugs, were higher than that of CPA—ranging from 176 to whole blood concentrations with improved bioavailability as
4260 nM. In order to verify if the Ki values were related on a compared to the parent CPA (13). Moreover, affinity and
potential hydrolysis of prodrugs, their stability has been activity were measured and found related to the nature of the
evaluated as the percentage of hydrolysis of the compounds. substituent at position 5Ј: intact prodrugs, as 5Ј-TMOB-CPA
The weaker compound is not hydrolyzed (5Ј-TMOB-CPA − (4d), are not able to interact efficiently with the receptor and
4d − Ki ס 4260 nM) whereas the more potent prodrug thus to display significant activity in cells. Regarding hydro-
showed the higher hydrolysis percentage (5Ј-Octanoyl-CPA − lysed prodrugs, susceptibility to hydrolysis appeared to be
4b − Ki ס 176 nM, hydrolysis ס 2.3%). The other two pro- influenced by the chain lengths, indeed the 5Ј-octanoyl-CPA
Fig. 2. Time courses of CPA and its prodrugs in human plasma and whole blood.

Adenosine A1 Receptor, Agonist Prodrugs
535
Fig. 4. Inhibition of forskolin-stimulated c-AMP levels in CHO A1
Fig. 3. Competition experiments of CPA and its prodrugs for specific
[3H]CHA binding performed at 25°C on CHO A1 cell membranes.
cells by CPA and prodrugs.
(4b) was the better hydrolysed, in both experimental condi-
tions for affinity and activity determination, and thus showed
the best activity and affinity values among prodrugs. We hy-
pothesise that the activity displayed derive from the different
amount of CPA released by the parent prodrug.
University of Ferrara. The authors thank Prof. Peter
Schofield of the Garvan Institute of Medical research, Syd-
ney, Australia and Dr. Andrea Townsend-Nicholson of the
Department of Anatomy & Developmental Biology, Univer-
sity College, London, for providing the CHO A1 cells. The
authors thank the Laboratorio Analisi Estense of Ferrara,
Italy, for supplying fresh blood.
Another important result of this study is represented by
the influence of the substituent at 5Ј position on the hydro-
lysis in plasma and whole blood. 5Ј-Pentanoyl-CPA (4a) is
very rapidly hydrolysed in whole blood (Table 1) whereas is
somewhat preserved in plasma (t1/2 ס 18 minutes). Increasing
the chain length, plasma half-life (i.e., 5Ј-octanoyl-CPA, 4b) is
significantly increased (t1/2 ס 30 minutes) and, surprisingly, a
comparable stability in whole blood is also induced. On the
other hand, substituents with similar length and LogP, but
having a hindered structure (compare 5Ј-pentanoyl-, 4a, and
5Ј-cyclohexanoyl-CPA, 4c), induced increased plasma and
whole blood half-lives. However, similar stability in whole
blood and increased half life in plasma was observed with
respect to longer chains such as for the 5Ј-octanoyl-CPA (4b).
These results suggest that chain length strongly influences
erythrocyte activity (compare 5Ј-pentanoyl-, 4a, and 5Ј-
octanoyl-CPA, 4b) whereas steric hindrance contributes to an
increasing of plasma half-life. Of particular interest the ob-
servation that the octanoyl chain blocked erythrocyte hydro-
lysis as indicated by the coincident t1/2 values in plasma and
whole blood. This occurrence was not observed in the case of
the 5Ј-cyclohexanoyl-CPA (4c). It is thus not surprising that a
bulky and planar substituent as the TMOB-group induced a
great stability.
In conclusion, our results demonstrate that the substitu-
tion pattern at 5Ј-position can highly influence the stability in
physiological fluids of CPA derivatives. In our opinion, the
octanoyl ester chain is of particular significance because it
shows a selective inhibition of CPA release into erythrocytes.
This occurrence may be ascribed to a selective process con-
nected either to uptake or hydrolysis inside erythrocytes.
Upon substitution at 5Ј-position, CPA shows increased
lipophilicity, stability and a strong decrease of affinity and
activity. CPA properties are restored only after hydrolysis.
Taking all these results into account, we therefore propose
5Ј-esters of CPA, characterized by suitable lipophilicity and
high degree of stability in physiological fluids, as canditates
for CPA-prodrugs.
REFERENCES
1. B. B. Fredholm, M. P. Abbracchio, G. Burnstock, J. W. Daly, T.
K. Harden, K. A. Jacobson, P. Leff, and M. Williams. Nomen-
clature and classification of purinoceptors. Pharmacol. Rev.
46:143–156 (1994).
2. K. A. Jacobson, B. K. Trivedi, P. C. Churchill, and M. Williams.
Novel therapeutics acting via purinergic receptors. Biochem.
Pharmacol. 41:1399–1410 (1991).
3. R. E. Simpson, M. H. O’Regan, L. M. Perkins, and J. W. Phillis.
Excitatory transmitter amino acid release form the ischemic rat
cerebral cortex: Effects of adenosine receptor agonists and an-
tagonists. J. Neurochem. 58:1683–1690 (1992).
4. C. Heurteaux, I. Lauritzen, C. Widmann, and M. Lazdunski. Es-
sential role of adenosine, adenosine A1 receptors, and ATP-
sensitive K+ channels in cerebral ischemic preconditioning, Proc.
Natl. Acad. Sci. USA 92:4666–4670 (1995).
5. A. L. Tucker and J. Linde. Cloned receptors and cardiovascular
responses to adenosine, Cardiovasc. Res. 27:54–61 (1993).
6. D. K. von Lubitz, R. C. Lin, N. Bischofberger, M. Beenhakker,
M. Boyd, R. Lipartowska, and K. A. Jacobson. Protection against
ischemic damage by adenosine amine congener, a potent and
selective A1 receptor agonist. Eur. J. Pharmacol. 369:313–317
(1999).
7. D. K. von Lubitz. Adenosine and cerebral ischemia: Therapeutic
future or death of a brave concept? Eur. J. Pharmacol. 371:85–
102 (1999).
8. M. S. Brodie, K. Lee, B. B. Fredholm, L. Stahle, and T. V. Dun-
widdie. Central versus peripheral mediation of responses to aden-
osine receotir agonists: Evidence against a central mode of ac-
tion. Brain Res. 415:323–330(1987).
9. B. Pavan and A. P. IJzerman. Processing of adenosine receptor
agonists in rat and human whole blood. Biochem. Pharmacol.
56:1625–1632 (1998).
10. R. A. A. Mathoˆt, E. A. van Schaick, M. W. E. Langemeijer, W.
Soudijn, D. D. Breimer, A. P. IJzerman, and M. Danhof. Phar-
macokinetic-pharmacodynamic relationship of the cardiovascular
effects of the adenosine A1 receptor agonist N6-cyclopentyl-
adenosine (CPA) in the rat. J. Pharmacol. Exp. Ther. 268:616–
624 (1994).
11. G. Remaud, X. Zhou, J. Chattopadhyaya, M. Oivanen, H. Lo¨n-
nberg. The effect of protecting groups of the nucleobase and the
sugar mioeties on the acidic hydrolisis of the glycosidic bond of
2Ј-deoxyadenosine: A kinetic and NMR spectroscopic study. Tet-
raedron 34:4453–4461 (1987).
ACKNOWLEDGMENTS
This research was supported by Italian Ministry of Uni-
versity, Scientific and Technological Research and by the
12. M. C. Maillard, O. Nokodijevich, K. F. Lanoue, D. Berkich, X. Ji,
R. Bartus, and K. A. Jacobson. Adenosine receptor prodrugs:

536
Dalpiaz et al.
Synthesis and biological activity of derivatives of potent A1-
selective agonists, J. Pharm. Sci. 83:46–53 (1994).
Schofield, and A. P. IJzerman. Thermodynamics of full agonist,
partial agonist and antagonist binding to wild-type and mutant
adenosine A1 receptors. Biochem. Pharmacol. 56:1437–1445
(1998).
13. S. Manfredini, P. G. Baraldi, R. Bazzanini, F. Bortolotti, S. Ver-
tuani, N. Ashida, and H. Machida. Enzymatic synthesis of 2Ј-O-
acyl prodrugs of 1-(␤-D-arabinofuranosyl)-5(E)-(2-
bromovinyl)uracil and of 2Ј-O-acyl-araU, -araC and −araA. An-
tiviral Chem. & Chemother. 9:25–31 (1998).
14. A. Townsend-Nicholson and J. Shine. Molecular cloning and
characterization of a human brain A1 adenosine receptor cDNA.
Mol. Brain Res. 16:365–370 (1992).
17. B. L. Brown, R. P. Ekins, and J. M. D. Albano. Saturation assay
for cyclic AMP using endogenous binding protein. In P. Green-
gard, G. Robinson, and R. Paoletti (eds.), Advances in Cyclic
Nucleotide Research, Raven Press, New York, 1972 pp. 25–40.
18. J. I. Wells. Pharmaceutical Preformulation: The Physicochemical
Properties of Drug Substances, Halsted Press- John-Wiley &
Sons, New York, 1988.
15. A. Townsend–Nicholson and P. R. Shofield. A threonine residue
in the seventh transmembrane domain of the human A1 adeno-
sine receptor mediates specific agonistic binding. J. Biol. Chem. 19. V. J. Davisson, D. R. Davis, V. M. Dixit, and C. D. Poulter.
269:27900–27906 (1994).
16. A. Dalpiaz, A. Townsend–Nicholson, M. W. Beukers, P. R.
Synthesis of Nucleotide 5Ј-Diphosphates from 5Ј-O-Tosyl
Nucleosides, J. Org. Chem. 52:1794–1801 (1987).