Adenosine receptors (ARs) trigger signal transduction pathways inside the cell when activated by extracellular adenosine. Selective modulation of the A3AR subtype may be beneficial in controlling diseases such as colorectal cancer and rheumatoid arthritis. Here, we report the synthesis and evaluation of β-D-apio-D-furano- and α-D-apio-L-furanoadenosines and derivatives thereof. Introduction of a 2-methoxy-5-chlorobenzyl group at N6 of β-D-apio-D-furanoadenosine afforded an A3AR antagonist (10c, Ki = 0.98 μM), while a similar modification of an α-D-apio-L-furanoadenosine gave rise to a partial agonist (11c, Ki = 3.07 μM). The structural basis for this difference was examined by docking to an A3AR model; the antagonist lacked a crucial interaction with Thr94. Keywords: G protein-coupled receptor, apionucleosides, Adenosine A3 receptor INTRODUCTION Transmembrane receptors coupled to G-proteins (heterotrimeric guanine nucleotide binding proteins) constitute a large receptor family, referred to as G protein-coupled receptors (GPCRs) or seven-transmembrane domain (7TM) receptors. Triggered by messenger molecules or signals outside the cell, GPCRs activate different signal transduction pathways inside the cell and, ultimately, cellular responses. The extracellular messengers range from photons to biogenic amines and other neurotransmitters to proteins. GPCRs are ubiquitous and involved in processes varying from directed chemotaxis[1] of small organisms (e.g. searching food for survival) to triggering apoptosis (programmed cell death) in large animals. They are the targets of almost 50% of marketed active pharmaceutical ingredients.[2] GPCRs activated by extracellular adenosine are classified as adenosine receptors (ARs), which are divided in four different subtypes, i.e. A1, A2A, A2B and A3 ARs.[3,4] The amino acid sequence similarity between the human (h) A3AR and hA1, hA2A, hA2B AR is 54 %, 48 % and 44 %, respectively.[5] The ARs use different signaling pathways; the A1 AR and A3 AR are preferentially coupled to Gi-proteins, and upon activation lead to adenylate cyclase inhibition, while the A2A AR and A2B AR are preferentially coupled to Gs-proteins and lead to adenylate cyclase activation. In some cells (e.g., mast cells), the A2BAR is dually coupled to Gs and Gq and consequently also mobilizes calcium and activates phospholipase C and MAPK. [3,6] Besides adenosine (1) itself, which is used clinically for the treatment of supraventricular tachycardia and in myocardial perfusion imaging,[7] only one AR-specific agent, the A2AAR agonist regedenoson, has so far been approved by the FDA. However, a relatively large group of AR ligands is currently under clinical evaluation.[8] With the exception of compound 2, which was reported in the mid-1980s as being inactive at A1 and A2ARs,[9] and the recently reported carbocyclic analogue 3,[10] a weak A3AR agonist, 4′-hydroxymethyl transposed nucleosides have not been investigated as AR ligands. This led us to employ a new and convenient method for the synthesis of apionucleosides from 1,2-O-isopropylidene-α-L-threose,[11] for the construction of suitably modified 9-(3-Chydroxymethyl-β-D-erythrofuranosyl)adenines (aka β-D-apio-D-furanoadenosines) (4-6) and 9-(3-C-hydroxymethyl-α-L-threofuranosyl)adenines (aka α-D-apio-L-furanoadenosines) (7-9) as potential A3AR modulators (Figure 1). Figure 1 Apio-type nucleosides previously tested for modulation of ARs (1-3) and target analogues of the present study (4-15). On the one hand, we envisaged to substitute the N6 position of β-D-apio-D-furanoadenosine 4 with substituted benzyl groups known to enhance A3AR affinity;[12] on the other hand we planned to substitute the well-known ethylcarboxamide moiety in place of the 4′-CH2OH group.[13] We decided to introduce a benzyl moiety at the N6 position and an N-alkylcarboxamide moiety at the 5′ position because this combination was shown to be conducive to selectivity at the A3AR. For example, N6-(3-iodobenzyl)-5′-N-methylcarboxamidoadenosine and its 2-chloro analogue (Cl-IB-MECA) display Ki values at A3AR of ~1 nM and ≥50-fold selectivity for this subtype. [14] Subsequently, an N6-(3-chloro-5-methoxybenzyl) substitution was shown to be beneficial for A3AR selectivity,[12] and we incorporated this moiety in the current target compounds. However, synthetic problems in introducing an ethylcarboxamide moiety motivated us to introduce a N-methyl or N-ethylcarbamoyloxymethyl group instead. Analogue modifications were carried out on α-D-apio-L-furanoadenosine 7. more...