(iii) Role of ribosomal components, ribosome biogenesis factors, translation factors and their coevolution in the fidelity of initiation and/or ribosome biogenesis.
(a) 16S rRNA (anti Shine-Dalgarno sequence) and the accuracy of i-tRNA selection in P-site: The interaction between the Shine-Dalgarno (SD) sequence often found upstream of the initiation codon in mRNAs, and anti-SD (aSD) sequence in the 3’ end of 16S rRNA, facilitates binding and placement of mRNA on 30S, and this interaction has a significant impact on the efficiency of initiation. An optimal strength of SD-aSD interaction is crucial for efficient and accurate initiation. Though previously believed that a high affinity between SD-aSD promotes translation initiation (Lim, Furuta, and Kobayashi 2012; Ma, Campbell, and Karlin 2002; Schurr, Nadir, and Margalit 1993; Starmer et al. 2006), it was later suggested that a sub-optimal level of SD-aSD affinity, which often occurs in E. coli , increases the efficiency of translation (Osterman et al. 2013; Wei, Silke, and Xia 2017). The correlation between the length of the SD sequence and the efficiency of translation follows a bell-shaped curve, with a maximum initiation efficiency in the range of 4-6 base pairs (Chen et al. 1994; Ma et al. 2002; Vimberg et al. 2007). Our suppressor analysis revealed that a C to T mutation at position 1536 towards the 3’ end of the 16S rRNA (inrrsC operon) resulting in an extended SD-aSD interaction of 8 bp (as opposed to the original 6 bp) with the reporter CATam1 mRNA, impaired scrutiny of i-tRNA, enabling initiation with the 3GC mutant i-tRNA (Shetty et al. 2014). This extended SD-aSD interaction of 8 bp mimics the conditions used in determining the X-ray crystal structure of the bacterial ribosome bound to the fMet-i-tRNA and mRNA (Selmer et al. 2006) wherein a weaker interaction of G1338 and A1339 was observed with the 3GC pairs. The extended SD-aSD interaction triggers a >2Å movement between the h26 and h28 of 16S rRNA, resulting in a change in the position of G1338 and A1339 (Korostelev et al. 2007; Selmer et al. 2006; Shetty et al. 2014). Thus, the SD-aSD context and its extent of the base pairing decide not only the rate of initiation but also the fidelity of initiation.
(b) Importance of ribosome maturation in accuracy of initiation: In the section on ‘Initiator tRNA abundance and ribosome maturation’, we described a novel role of i-tRNA particularly the 3GC pairs during the pioneering round of initiation, in the terminal stages of ribosome maturation. The work showed that the ribosomes when not completely mature (for example because of deletion of metZWV ) use the 3GC mutant i-tRNA for initiation. Ribosome biogenesis/maturation is an elaborate process requiring RNA/RNA, RNA/protein, and protein/protein interactions to process precursor rRNA molecules, nucleoside modifications, structural rearrangements, and interaction of r-proteins. A number of biogenesis factors are also required to assemble mature ribosomes in vivo (Kaczanowska and Rydén-Aulin 2007; Shajani, Sykes, and Williamson 2011). Interestingly, one of the suppressor strains (named A18) allowed us to uncover a new role for RluD in ribosome biogenesis and fidelity of initiation. RluD binds 50S and modifies U1911, U1915, and U1917 in H69 of 23S rRNA to pseudouridines.  In A18, RluD suffered an E265K mutation in its C-terminal tail. This mutation causes no detectable change in the biochemical activity of RluD as pseudouridine synthase but allows initiation with the 3GC mutant i-tRNA (Lahry et al. 2022). An earlier study where a catalytically dead RluD was used, inferred an alternate function of RluD in rescue ofE. coli growth (Gutgsell et al. 2001). The RluD (E265K) results in increased association of RbfA to 30S. The data suggested that the C-terminal tail of RluD facilitates release of RbfA from the 30S(Fig. 4) . The findings not only suggest the role of RluD in the fidelity of translation initiation via promoting 30S biogenesis but also resolve a long-held hypothesis of the alternate function of RluD. RbfA bound to 30S may affect the scrutiny of the 3GC pairs in the P-site possibly via G1338 and A1339 (Korostelev et al. 2007; Shetty et al. 2014) to allow initiation even with an i-tRNA mutated in its 3GC sequence. Interestingly, IF3 has also been shown to release RbfA from 30S (Sharma and Woodson 2020).
(c) Initiation factors in accuracy of i-tRNA selection: Initiation is assisted by three essential initiation factors, IF1, IF2, and IF3 which scrutinize the fidelity of translation initiation. The ribosome-bound IF2 recruits i-tRNA to 30S (Milon et al. 2010). IF1 prevents untimely access of tRNAs to the A-site and aids in positioning of IF2 bound i-tRNA in the P-site in P/I (the peptidyl/initiation) state, forming the 30S PIC (Antoun et al. 2006b, 2006a). Although IF1 and IF2 are independent proteins in bacteria, we showed that an insertion of 37 amino acids in IF2 (in mammalian mitochondria) substitutes for the independent requirement of IF1 in E. coli(Gaur et al. 2008). In case of IF3, it has been shown that it not only serves as an anti-association factor (preventing association of 30S and 50S in the absence of an mRNA) but also as a proof-reader of the 3GC pairs in i-tRNA on the 30S P-site (Ayyub, Dobriyal, and Varshney 2017; O’Connor et al. 2001). Biochemical analysis suggested that all of its (IF3) known activities (anti-association of the ribosomal subunits, prevention of pseudo-initiation complex formation, and prevention of initiation from leaderless mRNAs and non-canonical initiation codons) can be discharged by the C-terminal globular domain (CTD) alone (Dallas and Noller 2001; Godefroy-Colburn et al. 1975; Gualerzi and Pon 1990; Hartz et al. 1990; O’Connor et al. 2001; Singh et al. 2005; Tedin et al. 1999), and the globular N-terminal domain (NTD) was deemed to provide stability of binding to IF3 (Petrelli et al. 2001). Our in vivostudies on the individual domains led to characterization of a crucial role of the NTD in the fidelity i-tRNA selection (Ayyub et al. 2017). Subsequently, we showed that IF3 NTD interactions with the elbow region of i-tRNA are crucial for the movements of the two domains of IF3 and in the fidelity of P-site binding of i-tRNA (Singh et al. 2022). Other studies revealed that the genetic interactions between uS12 and IF3 also play a role in i-tRNA selection (Datta, Singh, et al. 2021).
(d) Coevolution of the translation apparatus: Initiator tRNAs in mycoplasma and rhizobia species possess variations in the 3GC sequence in their i-tRNA anticodon stems. For example, variations like A29-U41, G30-C40, G31-C39 (AU/GC/GC); G29-C41, G30-C40, G31-U39 (GC/GC/GU) or A29-U41, G30-C40, G31-U39 (AU/GC/GU), are of common occurrence in these organisms (Fig. 6A) . Computational analysis showed that unconventional nature of the 3GC pairs in these i-tRNAs is accompanied with changes in other components of the translation apparatus. The conserved C-terminal tail (SKR) of uS9 is represented by TKR sequence in mycoplasma, and uS13 possesses a longer C-terminal tail. A conserved R131 position in IF3, is represented by P, F or Y in mycoplasma harbouring i-tRNA with AU/GC/GU pairs. Initiation with non-AUG codons is also common in mycoplasma. When these features (P131 in IF3, TKR at C-terminal tail of uS9, and longer C-terminal tail of uS13) were tested in E. coli (Ayyub et al. 2018), they facilitated initiation with i-tRNAAU/GC/GU and 3GC mutant i-tRNA.
Additional understanding of coevolution of the translation apparatus, particularly the interplay between i-tRNA, IF3 and uS12 came from the homologous system of E. coli . We earlier showed that E. coli lacking native i-tRNA genes could be sustained on i-tRNAcg/GC/cg (Fig. 6A) . However, these strains grew poorly (Shetty et al. 2016). We later discovered that for sustenance on i-tRNAcg/GC/cg as the only source of i-tRNA, E. coli required V93A mutation in IF3. When this slow growing E. coli strain was repeatedly sub-cultured to acquire suppressor mutations to enable faster growth, it came up with additional mutations of either V32L or H76L in uS12. The V93A mutation in IF3 was the initial requirement to decrease the stringency of i-tRNA selection at the P-site, essential for survival with the only available i-tRNAcg/GC/cg, at the cost of compromised overall growth rate. The succeeding mutations in the genome that enhance growth, occurred in a r-protein, uS12 (V32L or H76L). This led to the discovery of cooperation between uS12, IF3 and i-tRNA for faithful initiation and a crosstalk between uS12, RRF, EFG and Pth (recycles tRNAs from peptidyl-tRNAs by hydrolysing the ester bond between the peptide and tRNA) modulating ribosome recycling to fine-tune the overall fidelity of translation initiation (Datta, Pillai, et al. 2021). The aforementioned systematic evolution demonstrated a paradigm for coevolution of the translation apparatus (Fig. 6B ).
(e) Ribosome recycling factor and the fidelity of initiation: Ribosome recycling, a vital step in protein synthesis, dissociates the post-termination complexes to make the ribosomal subunits available for another round of initiation (Kiel, Kaji, and Kaji 2007). We and others showed that the specific interactions between RRF and EFG, the two key factors involved at this step, are crucial in dissociation of the complexes consisting of mRNA-bound ribosomes harbouring deacylated tRNA (Fujiwara et al. 2004; Ito et al. 2002; Rao and Varshney 2001). Subsequently, it was shown that IF3 also plays an active role in ribosome recycling (Singh et al. 2005). The genetic interactions between RRF and Pth (Das and Varshney 2006) were seen in several independent investigations (Seshadri et al. 2009; Singh et al. 2005, 2008; Singh and Varshney 2004). These interactions strongly suggest that RRF, EFG and IF3 can act together on the pre-termination ribosomal complexes to release peptidyl-tRNAs. On the other hand, the structural investigations inferred that pre-termination ribosomal complexes are not optimal substrates for RRF binding (Schmeing and Ramakrishnan 2009; Weixlbaumer et al. 2008). However, given that RRF levels are high in cell (~20 µM) (Prabhakar et al. 2017), and multiple GTP molecules are often hydrolysed in carrying out a single round of recycling (Borg et al. 2016), it is feasible that the pre-termination ribosomal complexes, at least the ones carrying short peptidyl-tRNAs (e. g., at the interface of initiation and elongation steps or in early elongation), may be utilized as substrates by RRF (Fig. 7 ).
In this context, would recycling of the incorrectly assembled 70S complexes (carrying formyl-aminoacyl-i-tRNA) or the ones wherein the initial codons in the ORF were incorrectly decoded, by RRF increase the accuracy of initiation? Evidence for this came from our earlier studies where we genetically combined the folD122 allele (having G122D mutation in FolD) with the frr ts allele (temperature sensitive allele of RRF from E. coli LJ14) (Seshadri et al. 2009). As already described, folD122 allows initiation with the 3GC mutant i-tRNA. Interestingly, when we transduced E. coli LJ14 strain (reduced level of RRF) with folD122 allele, we noted increased initiation with the 3GC mutant i-tRNA, suggesting that deficiency of RRF allowed increased accommodation of the 3GC mutant i-tRNA in the ribosomal P-site (Singh et al. 2008) showing that RRF plays a role in accuracy of i-tRNA selection. Furthermore, we noted that uS12 (through its PNSA loop), shows genetic interaction with RRF and Pth, which are important in the fidelity of translation (Datta, Pillai, et al. 2021).
More importantly, as mentioned, in our coevolution study involving initiation with the i-tRNAcg/GC/cg, we noted that in the background of IF3 (V93A), while H76L mutation in uS12 improved the accuracy of i-tRNA selection, the V32L mutation compensated for the reduced fidelity of i-tRNA selection by ensuring a fidelity check by enhanced RRF function (Datta, Singh, et al. 2021). This investigation highlighted the importance of the genetic interactions between i-tRNA, IF3, uS12, EF-G, RRF, and Pth in maintaining the overall fidelity of initiation. It is satisfying that a more recent investigation further supports the role of RRF in peptidyl-tRNA drop-off in early stages of elongation as an additional quality control measure to maintain faithful translation (Nagao et al. 2023).