A G.U base pair in the eukaryotic selenocysteine tRNA is important for interaction with SePF, the putative selenocysteine-specific elongation factor.

In Escherichia coli, selenocysteine biosynthesis and incorporation into selenoproteins requires the action of four gene products, including the specialized selenocysteine tRNA(Sec) and elongation factor SELB, different from the universal EF-Tu. In this regard, the situation is less clear in eukaryotes, but we previously reported the existence of SePF, a putative SELB homologue. The secondary structure of the tRNA(Sec) differs slightly in eukaryotes, due to a change in the lengths of several stems. Two non-Watson-Crick base pairs, G5a x U67b and U6 x U67, reside in the acceptor stem and are conserved in the course of evolution. Since it has already been reported that changing them to Watson-Crick base pairs did not affect the serylation or selenylation levels of tRNA(Sec), we asked whether these non-Watson-Crick base pairs are required for the interaction with SePF. To this end, tRNA(Sec) variants carrying Watson-Crick changes at these positions were tested for their ability to maintain the interaction with SePF. In these assays, the tRNA(Sec)-SePF interaction was determined by the protective action it confers against hydrolysis of the amino acid ester bond, under basic conditions. All the changes introduced at U6 x U67 did not significantly affect the interaction. Interestingly, however, the G5a x U67b to G5a-C67b substitution was sufficient, by itself, to lead to unprotection of the ester bond. Therefore, our finding strongly suggests that SePF is unable to interact with a tRNA(Sec) mutant version carrying a Watson-Crick G5a-C67b instead of the wild-type G5a x U67b base pair, establishing that G5a x U67b constitutes a structural determinant for SePF interaction.