Structural insights into the translational infidelity mechanism

Abstract

The decoding of mRNA on the ribosome is the least accurate process during genetic information transfer. Here we propose a unified decoding mechanism based on 11 high-resolution X-ray structures of the 70S ribosome that explains the occurrence of missense errors during translation. We determined ribosome structures in rare states where incorrect tRNAs were incorporated into the peptidyl-tRNA-binding site. These structures show that in the codon-anticodon duplex, a G·U mismatch adopts the Watson-Crick geometry, indicating a shift in the tautomeric equilibrium or ionization of the nucleobase. Additional structures with mismatches in the 70S decoding centre show that the binding of any tRNA induces identical rearrangements in the centre, which favours either isosteric or close to the Watson-Crick geometry codon-anticodon pairs. Overall, the results suggest that a mismatch escapes discrimination by preserving the shape of a Watson-Crick pair and indicate that geometric selection via tautomerism or ionization dominates the translational infidelity mechanism.

Figure 1. Miscoding of mRNA during protein synthesis.

(a) The scheme depicts an event when the ribosome miscodes the phenylalanine codon by leucyl-tRNA^Leu, which becomes translocated to the P-site and introduces an erroneous amino acid to a polypeptide chain. A, P and E define aminoacyl, peptidyl and exit tRNA-binding sites, respectively (b,c) Schematic representations of the 70S ribosome complexes where the G·U mismatch was modelled at the first (b) or second (c) position of the codon–anticodon duplex (framed). For each complex tRNA, which was used for the complex formation is specified together with the sequence of mRNA; in mRNA-1 and mRNA-2 the Shine–Dalgarno sequence is underlined. Cognate complexes used for comparisons are indicated below each schematic representation. The first nucleotides of the mRNA codons bound in the P- and A-sites are numbered (+1) and (+4), respectively.

Figure 2. The G·U mismatch forms Watson–Crick-like pairs at the first and second codon–anticodon positions in the P-site of the 70S ribosome.

(a) Region of codon–anticodon interactions in the P-site (framed) in the context of the full 70S ribosome; tRNA bound in the A-, P- and E-sites are represented by red, blue and green, respectively. (b,c) The G·U mismatch mimics Watson–Crick pair at the first (b; left) and second (c; left) positions of the codon–anticodon duplex (b,c; right). Putative hydrogen bonds formed by an anticipated tautomeric form of G or U have 2.8–3.2 Å length (see text). (d) Geometry of canonical Watson–Crick pair and non-canonical wobble pair formed by keto isomers (left panel). Watson–Crick-like pairs formed by rare enol tautomers of uracil or guanosine (right panel; marked by asterisk). In b and d, the following figures schemes of the codon–anticodon duplexes are indicated and arrows mark the described mismatches. All graphical representations were rendered by PyMol. For all figures, the density maps are contoured at 1.6−1.8 σ level.

Figure 3. The C·A mismatch does not form a stable pair in the 70S ribosome-decoding centre.

(a,b) The C·A mismatch at the first position of the codon–anticodon duplex in the absence (a) or presence (b) of the queuosine modification in the tRNA anticodon. In a, the left panel shows that A1493 in 16S rRNA prevents strong pairing interactions in the C·A mismatch by constraining its sugar-phosphate backbone by hydrogen bonding. In b, the left and middle panels demonstrate a lack of pairing in the C·A mismatch reflected by the weak electron density signal corresponding to the cytosine base and misshaping of the mini-helical structure due to displacement of the cytosine by queuosine (right). (c) The C·A mismatch at the second position of the codon–anticodon duplex; as in a and b, the left panel depicts conserved A1492 and G530 in 16S rRNA tightening around the mismatch and contorting it. The density maps are contoured at 1.8 σ level. In a–c, the right panels depict overall geometry of the mismatches in the codon–anticodon mini-helices.

Figure 4. The A·A mismatch in the decoding centre of the 70S ribosome.

The left panels demonstrate the A·A mismatch at the first (a) and second (b) positions of the codon–anticodon duplexes. As for the C·A mismatches, A1493 (a) and A1492 with G530 (b) in 16S rRNA constrain the sugar-phosphate backbones of the first and second mispairs by hydrogen bonding. In a, the exact position of the codon adenosine was not detectable and the figure shows one of the possible positions of this nucleobase; in b, the distances between adenosines exceed 3.6 Å demonstrating absence of strong interactions. In a and b, the right panels depict overall geometry of the mismatches in the codon–anticodon mini-helices.

Figure 5. Proposed mechanism of translational infidelity.

(a) Misincorporation of an amino acid by the ribosome. The main steps of the tRNA selection process are shown including hydrolysis of GTP (black asterisk) by elongation factor Tu on establishing codon–anticodon interactions in the decoding centre; (i–iv) indicate sequential steps of the process (see the text). (b) Conformation of the main nucleotides of the decoding centre without tRNA (left), bound by cognate or near-cognate aa-tRNA at the initial recognition step (middle) and at the final step of accommodation (right). The crucial nucleotides of 16S and 23S rRNA are shown in cyan and red, respectively. Ribosomal protein S12, which belongs to the shoulder domain of the small ribosomal subunit and additionally restricts the second codon–anticodon pair, is depicted in green. The three nucleotides of the mRNA codon in the A-site are numbered according to the standard system (see legend to Fig. 2). (c) Overall states of the small ribosomal subunit during selection of tRNA. The left panel pictures spontaneous movement of the shoulder domain (black arrows) when the decoding centre is unoccupied; the middle and right panels show that the shoulder is shifted and stabilized on binding of near-cognate tRNA during the initial selection step and further accommodation (see text); sh, h, pl denote the shoulder, head and platform domains of the small ribosomal subunit, respectively. (d) Strengthening of cognate tRNA binding by protein tails from the small (S) and large (L) ribosomal subunits. Fastening of cognate tRNA in the A-site is also represented by formation of the additional intersubunit bridge between protein L31 and proteins S13 and S19 (PDB codes 3I8H and 3I8I). The protein extensions are disordered when near-cognate tRNA is bound in the A-site (present work; the model of the 70S ribosome with tRNA₂^Leu bound to the UUU codons in the P- and A-sites). The absence of additional stabilization by the proteins can promote dissociation of near-cognate tRNA from the ribosome.