SYNTHESIS OF N⁶-SUBSTITUTED 3′-UREIDOADENOSINE DERIVATIVES AS HIGHLY POTENT AGONISTS AT THE MUTANT A₃ ADENOSINE RECEPTOR

Abstract

Several N⁶-substituted 3′-ureidoadenosine derivatives were efficiently synthesized starting from d-glucose for the development of H272E mutant A₃ adenosine receptor (AR) agonists. Among compounds tested, 3′-ureido-N ⁶-(3-iodobenzyl)adenosine (2c) exhibited the highest binding affinity (K_i = 0.22 μM) at the H272E mutant A₃ AR without binding to the natural A₃ AR.

INTRODUCTION

Selective A₃ adenosine receptor (AR) full agonists show high therapeutic potentials in the treatment of cardiac and cerebral ischemia^[1] and cancer^[2] but, the ubiquitous presence of adenosine receptors throughout the body hindered them from being developed as clinically useful agents. Thus, Jacobson et al.^[3] have demonstrated a re-engineered G protein-coupled receptor, mutant A₃ AR in which the His residue (H272) of natural A₃ AR, strongly hydrogen-bonded to the 3′-hydroxyl group of adenosine,^[4] was mutated to a negatively charged residue, Glu or Asp. This mutant A₃AR called neoceptor can recognize only a specifically designed A₃ agonist, but not the native A₃ agonist.

Thus, for the purpose of developing optimal agonists at the mutant A₃AR, we modified the 3′-hydroxyl group of adenosine (1) to the 3′-ureido group, because the 3′-ureido group is able to form highly favorable electrostatic interaction with the Glu residue of H272E mutant A₃ adenosine receptor. Herein, we report the synthesis of 3′-ureidoadenosine derivatives and their binding affinity at the mutant A₃ adenosine receptor (Figure 1).

FIGURE 1.

Rationale for the design of the desired nucleosides 2a–d.

RESULTS AND DISCUSSION

Synthesis of the target nucleosides 2a–d started from the known intermediate 3^[5] which was derived from d-glucose, as depicted in Scheme 1.

SCHEME 1.

Reagents and conditions: (a) NH₄OH, 1,4-dioxane, rt or MeNH₂, 1,4-dioxane, rt or 3-iodobenzylamine-HCl or phenetylamine-HCl, Et₃N, EtOH, 50°C, then NaOMe, MeOH, rt; (b) TBSCl, imidazole, DMF, rt; (c) Ph₃P, NH₄OH/H₂O, THF, rt; (d) chloroacetyl isocyanate, DMF, 0°C, then NaOMe, MeOH, rt; (e) TBAF, THF, rt.

Treatment of 3 with various appropriate amines followed by deacetylation with sodium methoxide gave the 3′-azido-N⁶-substituted adenosine derivatives 4a–d in high yields. Protection of two hydroxyl groups in 4a–d with t-butyldimethylsilyl (TBS) group followed by catalytic hydrogenation of azido group with triphenylphosphine in aqueous ammonium hydroxide afforded the 3′-amino derivatives 5a–d, respectively. Conversion of 3′-amino group into 3′-ureido group was accomplished by treating 5a–d with chloroacetyl isocyanate followed by reacting with sodium methoxide to give the 3′-ureido compounds 6a–d, respectively. Removal of TBS groups in 6a–d with TBAF yielded the final nucleosides 2a–d, respectively.

Binding affinities of all synthesized 3′-ureido derivatives at the wild-type A₃ AR as well as H272E mutant A₃ AR were measured using a radioligand binding assay. All compounds did not show any significant binding affinities to all subtypes of wild-type adenosine receptors. However, compound 2c had no effect on the WT A₃AR but bound to the H272E mutant receptor with a K_i value of 0.22 μM.

In summary, we have synthesized the 3′-ureidoadenosine derivatives, among which compound 2c formed a favorable electrostatic interaction only at the H272E mutant A₃ AR (neoceptor), not at the WT A₃AR. This selective ligand (neoligand)-neoceptor approach will expedite the development of clinically useful organ-specific compounds.