Aminoacyl-tRNA synthetases attach amino acids to the 3-ends of their cognate tRNAs. This aminoacylated, or"charged," tRNA then enters the ribosome where its attached cognate amino acid is transferred to the growing peptide chain (Figure). This is accomplished with a very high degree of specificity, making this system an excellent model for probing the principles of RNA-protein interactions and substrate recognition.
My research has focused on the recognition of the terminal base pairs of the acceptor stem of tRNAAla by E. coli alanyl-tRNA synthetase (AlaRS). Short RNA duplexes that mimic the acceptor stem of tRNAAla contain the critical G3:U70 base pair and are efficient substrates for AlaRS (Figure).We have used atomic group "mutagenesis," in which modified bases are incorporated into RNA via solid-phase chemical synthesis, to make subtle changes in the terminal base pairs and determine the effect on aminoacylation efficiency. It was previously observed in our lab that a G1:C72 (WT) to C1:G72 base pair transversion completely eliminated aminoacylation by AlaRS.1 
We then constructed over thirty variants at this position to probe specific features of the wild-type G1:C72 and inactive C1:G72 base pair.2 These experiments failed to identify a single functional group that was important for recognition of the wild-type base pair. We observed that a carbonyl oxygen atom blocked aminoacylation whether presented in the context of a purine or pyrimidine in the major groove at position 72. We also probed potential conformational effects of the base pair transversion at this position, and determined that there may be subtle conformational changes at position A73, the first single-stranded nucleotide that stacks directly onto this base pair. We have also applied this atomic group "mutagenesis" approach to other positions in the top of the acceptor stem, and observe predominately major groove recognition at the discriminator base, A73 (work done by Abbey Fischer3), and positions 1:72 and 2:71. Minor groove recognition of the G:U wobble pair has been well-established at position 3:70;4,5 similarly, minor groove recognition is apparent at position 4:69 as well.6
Not only must the synthetases recognize the correct tRNAs, they must also select the correct amino acid out of all 20. 
The aminoacylation reaction provides the last check on the accurate pairing of amino acid to trinucleotide codon (anti-codon on the tRNA). Therefore, any mistakes made by a synthetase in selecting the amino acid must be corrected through one or more of a number of mechanisms. My second major project is to probe proofreading activity in prolyl-tRNA synthetase (ProRS). I have determined that E. coli, Methanococcus jannaschii, and human ProRS are capable of misactivating non-cognate amino acids, though their substrate profiles differ slightly. I have also found that E. coli ProRS possesses hydrolytic editing (removal of the misactivated amino acid) and deacylation (removal of the amino acid from a mischarged tRNA) activities.7 We are now attempting to locate the editing site on E. coli ProRS via site-directed mutagenesis, focusing on the large insertion domain between motifs 2 and 3 (Figure) that is missing in the eukaryotic ProRSs. We are also attempting to identify a deacylation signal on E. coli tRNAPro.
Funding for these projects has been provided by the National Institutes of Health, a Louise T. Dosdall Fellowship from the University of Minnesota, and a Department of Chemistry Fellowship.
References