br Acknowledgements br Introduction Protein translation dema

Acknowledgements
Introduction Protein translation demands high fidelity. There are number of molecular checkpoints to fulfill this demand. Among these, insuring that a particular tRNA gets aminoacylated by its conjugate amino acid, which in turn is catalyzed by a particular aminoacyl-tRNA synthetase (aaRS), is important [1]. Typically, a tRNA needs to be specific to its cognate aaRS, and vice versa. Understanding the molecular basis of specificity in the aminoacylation reaction machinery is an active field of research [2], [3], [4], [5], [6], [7] that can yield important clues about molecular mechanisms of specific and non-specific aaRS–tRNA interaction. Experimentally, one way to probe aaRS–tRNA interaction is selective mutation of nucleotides (in tRNA) or amino ceramidase residues (in aaRS) and studying the effect of the mutations on aminoacylation efficiency. Experiments along these lines have been performed on tRNAGlu in the bacterium Escherichia coli, where tRNAGlu nucleotides were extensively mutated and the glutamylation efficiency of mutant tRNAGlu, catalyzed by glutamyl-tRNA synthetase (GluRS), were measured. This has yielded a nucleotide identity set of E. coli tRNAGlu – nucleotides that play a critical role in maintaining optimum glutamylation efficiency [2]. As shown in Fig. 1a, the identity elements in E. coli tRNAGlu are clustered in the anticodon loop, the acceptor arm and the augmented D-helix in tRNAGlu. In this paper we focus on the identity elements present at the acceptor arm and the augmented D-helix in tRNAGlu. Bacterial GluRS constitutes of a N-terminal catalytic domain and a C-terminal anticodon-binding domain [8]. The catalytic domain interacts with the acceptor arm and the augmented D-helix in tRNAGlu (Fig. 2a) [3]. Limited mutational studies have been performed on the catalytic domain in a few bacterial GluRSs. Mutation of a residue (C100Y), close to the acceptor stem of tRNAGlu in the Zn-binding SWIM domain in E. coli GluRS resulted in a variant with a slightly lower affinity for l-Glu suggesting that the SWIM domain participates in correctly positioning the tRNA acceptor end in the active site [9]. Mutational studies on Thermus thermophilus GluRS have also yielded a set of residues whose mutation affects the glutamylation efficiency [8]. Among the set of mutated catalytic domain residues, mutation of four residues, E282, S299, K309 and W312, were shown to play an important role in glutamylation of tRNAGlu. All four residues are close to the D-helix of tRNAGlu as shown in Fig. 2a. Limited mutational studies have also been performed on GluRS1 and GluRS2 from Helicobacter pylori where GluRS1 stands for the canonical GluRS and GluRS2 corresponds to a non-canonical version that only glutamylates tRNAGln[10]. In addition to mutational studies, glutamylation efficiencies of a domain-deleted or a domain-swapped version of E. coli GluRS have also been reported – from our lab [11], [12] and by Lapointe and co-workers [13]. These studies show that the isolated N-terminal catalytic domain of E. coli GluRS is still capable of glutamylating tRNAGlu, albeit with a reduced efficiency. Even when tRNAGlu identity elements are known, as in E. coli tRNAGlu, there is no guarantee that the identity set is universal among all bacterial tRNAGlu. As shown in Fig. 1b, a simple comparison of E. coli (gamma-proteobacteria) tRNAGlu identity elements with corresponding nucleotide sequences in tRNAGlu from four other bacteria – T. thermophilus (deinococcus-thermus), T. elongatus (cyanobacteria), Bacillus subtilis (firmicutes) and Mycobacterium tuberculosis (actinobacteria) — illustrates this point. The U2·A71 identity in the acceptor stem seems to be specific to the proteobacterium E. coli while a non-augmented D-helix (A13·A22) only occurs for the firmicute B. subtilis. It is interesting that B. subtilis GluRS cannot charge the augmented D-helix containing E. coli tRNAGlu[14] supporting the idea that the tRNAGlu identity elements in E. coli and B. subtilis are different. Indirect evidence that the distribution of tRNA identity elements in E. coli tRNAGlu may not be universal also comes from studies on GluRS2, a non-canonical GluRS, from glutamylation assays on chimeric tRNAs in H. pylori[15] (an epsilon proteobacteria) and from glutamylation assays on tRNAGlu isoacceptors of Acidithiobacillus ferrooxidans[16] (a gamma proteobacteria).