Investigation of the tRNA recognition oftRNA-guanine transglycosylase from Escherichia coli.

Abstract

Previous studies have identified a minimal recognition motif for E. coli tRNA-guanine transglycosylase (TGT), a key enzyme involved in the postranscriptional modification of tRNA with queuosine. This minimal recognition motif consists of a U-G-U sequence (at positions 33-35) within a seven-base loop and requires an RNA hairpin structure. To further understand the tRNA recognition by E. coli TGT, a series of tRNAs and tRNA analogues were studied. The results from studies of eight queuosine cognate tRNAs from E. coli and S. cerevisiae suggest that no primary or secondary structures other than the minimal recognition motif have any significant effects on TGT recognition. To further characterize this minimal recognition motif, several tRNA minihelix with different substitutions in the 2$\sp\prime$ position of the TGT reaction site, i.e., G34, were evaluated. Results from these studies imply that the 2$\sp\prime$-hydroxyl group of the guanosine at the reaction site is not crucial for TGT reaction but may be involved in orienting the molecule in the active site for optimal catalysis. To address whether the discrimination of TGT against queuosine-noncognate tRNAs is based purely on the nucleotide sequence at positions 33-35 or involves other negative recognition elements, a noncognate-cognate chimeric tRNA (yeast tRNA$\sp{\rm Phe{-}Asp})$ was investigated. During the course of this study, it was found serendipitously that a noncognate tRNA (yeast tRNA$\sp{\rm Phe})$ and its extended T$\Psi$C arm minihelix analogue (SCFMH(T$\Psi$C)) (both containing U-G-U in the T$\Psi$C stem/loop junction) were substrates for TGT. This discovery reveals that effective TGT recognition of its RNA substrate may not require the entire U-G-U sequence to be in the loop region as long as the reaction site can be introduced to the TGT active site properly. The exact guanine incorporation site in those queuosine-noncognate tRNA analogues, however, has not yet been identified. These results indicate that it is mechanistically feasible that the E. coli tRNA-guanine transglycosylase may modify other nucleic acids (RNA and DNA) in vivo. This finding has significant implications regarding the physiological roles of the queuosine modification.