Maximum-likelihood Analysis of Single-Molecule Data on Elongation in an In Vitro Eukaryotic Translation System

The ribosome plays a central role in translation of the genetic code into amino acid sequences during synthesis of polypeptides. During each cycle of peptide elongation, the ribosome must discriminate between correct and incorrect aminoacyl-tRNAs according to the codon present in its A-site. Ribosomes rely on a complex sequence of proofreading mechanisms to minimize erroneous selection of incorrect aminoacyl-tRNAs that would lead to mistakes in translation. These mechanisms have been studied extensively in prokaryotic organisms, but eukaryotic elongation is less well understood. We applied single-molecule fluorescence resonance energy transfer to this problem. smFRET yields rich datasets that can be used to interrogate the mechanisms of the eukaryotic ribosome. Traditional analyses, however, often reduce such data to first moments, for example, mean waiting time as a function of tRNA concentration. We outline a principled, maximum-likelihood approach that objectively fits the entire distributions of waiting times for a proposed kinetic model. Our approach is computationally inexpensive and amenable to bootstrap estimation of errors. We compared accommodation of a Tryptophan-aminoacyl-tRNA into the ribosomal A-site containing either a cognate or near-cognate codon and unexpectedly found faster tRNA binding of the near-cognate tRNA after initial unsuccessful tRNA sampling and a strong negative correlation between the ambient concentration of near-cognate aminoacyl-tRNA and the efficiency of tRNA accommodation. These novel characteristics of near-cognate interaction with the eukaryotic ribosome suggest that rejection of a near-cognate tRNAs leads to an altered ribosomal configuration that assists in rejecting subsequent incorrect tRNA interactions.