The A2A adenosine receptor has an extended cytoplasmic carboxyl terminal tail (c-tail) that contains putative docking sites for accessory cellular proteins. The binding of two candidate interaction partners to the receptor c-tail, these are (i) the MAGUK (membrane-associated guanylate kinase-like domain) protein SAP102 (synapse associated protein of 102 kDa) and (ii) the ubiquitin-specific protease USP4 was evaluated. The yeast-two-hybrid test was employed to assess interaction with the receptor c-tail at its original length and with several variants that had been truncated from the distal end.
Various subdomains (PDZ, SH and GUK) were isolated from the multi-domain protein SAP102 and tested separately; the substrate recognition domain (DUSP) was used in case of USP4. For SAP102 a series of interaction tests indicated that the receptor c-tail binds - by virtue of a five-residue motif located at position 382-386 (that is 22 amino acids removed from the c-terminal end) - to the GUK domain. Signal transduction was examined by a receptor mutant where the original five-residue motif (DVELL) had been supplanted by an alternative sequence (RVRAA) such that the receptor could not bind to SAP102. In HEK293 cells neither expression of the receptor nor receptor-dependent formation of cAMP or activation of the MAP-kinase ERK1/2 differed between mutant and wild-type receptor. In contrast to SAP102, the binding site of the USP4-DUSP domain could not be unambiguously mapped on the receptor c-tail. Progressive truncation of the human c-tail identified a candidate interaction site that however was not detected in the highly homologous sequence of the mouse species orthologue. Thus, the conclusion is that (i) the yeast-two-hybrid interaction test may have limitations requiring the results to be re-evaluated by an independent experimental approach and that (ii) the consequence of mutating the SAP102 interaction site in the receptor should be revealed when the mutant receptor is expressed in a neuronal cell.