The tRNA identity landscape for aminoacylation and beyond

Abstract

tRNAs are key partners in ribosome-dependent protein synthesis. This process is highly dependent on the fidelity of tRNA aminoacylation by aminoacyl-tRNA synthetases and relies primarily on sets of identities within tRNA molecules composed of determinants and antideterminants preventing mischarging by non-cognate synthetases. Such identity sets were discovered in the tRNAs of a few model organisms, and their properties were generalized as universal identity rules. Since then, the panel of identity elements governing the accuracy of tRNA aminoacylation has expanded considerably, but the increasing number of reported functional idiosyncrasies has led to some confusion. In parallel, the description of other processes involving tRNAs, often well beyond aminoacylation, has progressed considerably, greatly expanding their interactome and uncovering multiple novel identities on the same tRNA molecule. This review highlights key findings on the mechanistics and evolution of tRNA and tRNA-like identities. In addition, new methods and their results for searching sets of multiple identities on a single tRNA are discussed. Taken together, this knowledge shows that a comprehensive understanding of the functional role of individual and collective nucleotide identity sets in tRNA molecules is needed for medical, biotechnological and other applications.

Figure 1.

Cloverleaf folding of tRNA and its three-dimensional L-shaped organization. The color code highlights the structural domains in tRNA. (A) The standard cloverleaf structure of cytosolic tRNAs and the conventional numbering system are used. Conserved nucleotides are explicitly indicated. The variable region (nucleotides 44–48) encompasses the long extra arm of tRNA^Leu, tRNA^Ser and tRNA^Tyr. The · symbol indicates Watson–Crick base pairings (including G·U pairs); dotted gray lines indicate other pairings important for tRNA L-shaped architecture. (B) Three-dimensional L-shaped structure of tRNA^Phe (1ehz) showing the folding of the different arms.

Figure 2.

Schematic representation of the distribution of identity elements across the 20 isoacceptor tRNA families. The tRNAs are presented by class and subclass of aaRSs. Invariant positions are represented in gray. The positions in red are the conserved or nearly conserved strong identity elements. Positions in yellow are weak determinants or determinants not conserved in the three domains of life. The red dashed lines indicate tertiary interactions involved in identity. For tRNA^Leu, the yellow double arrow indicates the low importance of the sequence elements in the extra arm; for tRNA^Ser, the red double arrow indicates that the size rather than the sequence of the extra arm is the major identity element.

Figure 3.

Positions occupied by identity elements in canonical tRNA cloverleaf folding, including identity-determining post-transcriptional modifications. Conserved and semi-conserved nucleotides, as well as tertiary interactions (broken green lines) are shown. The size of red bullets schematizes the extent of occupation in the canonical tRNA sequence (large for the four heavily occupied positions, medium for the eight significantly occupied positions and small for the 32 poorly occupied positions). The standard numbering of positions is used as in Figure 1. Variable region (nucleotides 44–48) with up to 16 nts for extra arms (47a to 47p). Modifications characterized as identity determinants for aminoacylation in anticodon loops are shown next to identity positions; they are displayed in standard abbreviations.

Figure 4.

A panel of atypical RNA folds present in tRNAs aminoacylated by aaRSs or only recognized by aaRSs. (A) E. coli tRNA^Sec. (B) M. barkeri tRNA^Pyl. (C) Turnip yellow mosaic virus (TYMV) tRNA-like structure (TLS) with valine-charging capacity. (D) E. coli thrS operator. Experimentally characterized identity elements are in red font. For easier comparison, the numbering of atypical tRNA folds is as in canonical tRNAs with, for example, positions 73 and 34–36 for discriminator bases and anticodons. For TYMV and mRNA TLSs, sequence numbering is from 3′ to 5′ ends of the molecules with starts at A₁ (in the -CCA_OH accepting end of the viral TLS and the A₁UG triplet next to the Shine and Dalgarno sequence in the mRNA TLS).

Figure 5.

Typical 2D folds in the four mt-RNA structural families indicating partial conservation of universal features necessary for canonical 3D folding, as found in representative mammalian and/or nematode mt-RNAs. (A) Canonical-like cloverleaf folding, (B) folding missing D arm, (C) folding missing T arm, and (D) folding missing both D and T arms. The numbering is based on that of canonical tRNAs. Conserved and semi-conserved nucleotides are in blue bold scripts. Experimentally characterized identity determinants are in red font (in bold for major determinants). Predicted determinants in R. culicivorax tRNA^Arg, based on E. coli tRNA^Arg, are in red italic font.

Figure 6.

Generalized L-shaped structure of RNAs recognized and/or aminoacylated by aaRSs with structural requirements necessary for function. The drawing schematizes how the amino acid acceptor (7 bp) and anticodon (5 bp) branches are connected in canonical tRNAs by two linkers L1 (from U₈ to N₂₆ with the D arm) and L2 (N₄₄ to N₄₈ with the T arm and either the small or large variable region). In atypical and mt-tRNAs, this architecture shows important peculiarities, with distal stem regions of 5–8 and 4–9 bp for acceptor and anticodon branches, respectively, pseudoknotted acceptor stems and great diversity in L1 and L2 sequences. The conserved pairings are shown. The orientation of the two arms is variable and the distance between the CCA_OH end and the anticodon triplet ranges between 45 and 75 Å.

Figure 7.

Examples of interactions between major identity determinants in tRNA with cognate aaRS as seen in crystal structures of tRNA:aaRSs complexes. (A) A₂₀ in the E. coli tRNA^Arg:ArgRS complex (5b63). (B) U₃₅ and C₃₆ in the S. cerevisiae tRNA^Asp:AspRS complex (1asy). (C) A₇₃ in the S. cerevisiae tRNA^Asp:AspRS complex (1asy). (D) G₃₄ in the T. thermophilus tRNA^Phe:PheRS complex (1eiy).

Figure 8.

Examples of position importance in bacterial tRNAs for (A) arginine and (B) alanine identities. Importance of positions in tRNA sequences is quantitatively evaluated by relative entropy (RE) values. Adapted from Branciamore, S. et al. (2018) Intrinsic properties of tRNA molecules as deciphered via Bayesian network and distribution divergence analysis. Life (Basel), 8, E5. (385). Profiles in (A) and (B) are adapted from panels c of Supplementary Figures S1 and S2 of reference (385). Arrows have been added to highlight the positions with the highest RE values associated with aminoacylation identity. According to the CC BY 4.0 license (https://creativecommons.org/licenses/by/4.0/).

Figure 9.

Pathogenic human mt-tRNAs, with most known mutations highlighted. (A) mt-tRNA^Leu(UUR) and (B) mt-tRNA^Lys cloverleaves with base modifications indicated. The location of mutations is indicated by arrows (black arrows for mutations with confirmed pathogenetic status, gray arrows for mutations of ‘likely pathogenic’ status). Red arrows highlight confirmed pathogenic mutations affecting identity elements. Data and status are from MITOMAP.