Identifying the ligated amino acid of archaeal tRNAs based on positions outside the anticodon

Abstract

Proper recognition of tRNAs by their aminoacyl-tRNA synthetase is essential for translation accuracy. Following evidence that the enzymes can recognize the correct tRNA even when anticodon information is masked, we search for additional nucleotide positions within the tRNA molecule that potentially contain information for amino acid identification. Analyzing 3936 sequences of tRNA genes from 86 archaeal species, we show that the tRNAs' cognate amino acids can be identified by the information embedded in the tRNAs' nucleotide positions without relying on the anticodon information. We present a small set of six to 10 informative positions along the tRNA, which allow for amino acid identification accuracy of 90.6% to 97.4%, respectively. We inspected tRNAs for each of the 20 amino acid types for such informative positions and found that tRNA genes for some amino acids are distinguishable from others by as few as one or two positions. The informative nucleotide positions are in agreement with nucleotide positions that were experimentally shown to affect the loaded amino acid identity. Interestingly, the knowledge gained from the tRNA genes of one archaeal phylum does not extrapolate well to another phylum. Furthermore, each species has a unique ensemble of nucleotides in the informative tRNA positions, and the similarity between the sets of positions of two distinct species reflects their evolutionary distance. Hence, we term this set of informative positions a "tRNA cipher." It is tempting to suggest that the diverging code identified here might also serve the aminoacyl tRNA synthetase in the task of tRNA recognition.

Intrinsic Properties of tRNA Molecules as Deciphered via Bayesian Network and Distribution Divergence Analysis

“…Our approach, as implemented in this study, is conceptually similar to the recent data-driven analyses of tRNA databases [ 26 , 27 ]. Galili and colleagues used a decision-tree based classifier to perform a variable selection (embedded in classification) to derive the minimal set of tRNA identity/informative positions for each amino acid in Arc domain.…”

“…Depending on the extent of decision tree pruning, single decision tree-based classifiers can also be relatively blind to the rare variants (which is not a drawback that BNs share). In our opinion, it would be a very fruitful research direction to adapt the analyses in [ 26 ] to more robust ensemble classifiers, such as Random Forests with double-loop cross-validation [ 28 ]. At this time, however, we believe that the robustness and exhaustiveness of the variable sets generated by our approach is preferable to the compactness of the sets generated by a single decision tree—especially in the context of the exploratory, hypothesis-generating analysis activity.…”

“…In statistical terms, the latter is simply too false negative-prone. Galili et al [ 26 ] clearly recognize the shortcomings of such “greedy” algorithms as CART; however, they conclude that combinatorial efficiency and compactness of the results outweigh such drawbacks. We will re-emphasize here that CART and BNs, in this context (of predominantly exploratory analytic activity), complement (rather than compete with) each other.…”

Protection of Pepper Plants from Drought by Microbacterium sp. 3J1 by Modulation of the Plant's Glutamine and α-ketoglutarate Content: A Comparative Metabolomics Approach

“…Other amino acids were found to increase in the plant when the microorganism was present such as histidine, alanine, isoleucine, leucine, proline, arginine, or tyrosine with increases between 6- and 1.6-fold. Amino acids of the aromatic family like histidine, or tyrosine have been described to take part in maize and wheat plants response to drought (Harrigan et al, 2007 ; Less and Galili, 2008 ; Bowne et al, 2012 ; Witt et al, 2012 ; Galili et al, 2016 ). While histidine has been described as part of the response to abiotic stress (Harrigan et al, 2007 ), tyrosine is synthesized through the shikimate pathway.…”

The structural basis of the genetic code: amino acid recognition by aminoacyl-tRNA synthetases

“…Mischarged tRNAs due to failed cognate tRNA detection can hardly be corrected by the respective aaRS, but mistranslation may still be avoided by cross-editing of the cognate aaRS 71 . The evolutionary older 72 acceptor stem is highly necessary for specificity, whereas tRNAs are still correctly detected when the evolutionary younger anticodon information is masked 73 . Additionally, certain informative nucleotide motifs in the tRNA are relevant for the aaRS to couple the cognate tRNA and amino acid, differing only in as few as one position between the 20 types 15 , 73 , 74 .…”