The key steps of mammalian translation are initiation, elongation, termination and recycling

Translation can be divided into four parts: 1) Initiation, which is the binding of the very first tRNA, positioning of the mRNA on the small ribosomal subunit, and joining of the large subunit; 2) iterative elongation cycles, wherein the amino acids coded for in the mRNA are successively connected to form the peptide chain; 3) termination, resulting in the release of the peptide chain from the ribosome, and 4) recycling, which is splitting of the associated subunits into small and large one such that they are ready to be fed in a new round of translation (Figure 9).

Initiation places the start codon in the P site by ribosome scanning

Initiation is the first step of translation. Here, the small subunit is prepared for joining with the large subunit (Figure 9, Figure 10). This preparation includes binding of an initiator tRNAi Met to the small subunit P site as well as binding of the mRNA and positioning of the mRNA start codon into the P site under assistance of initiation factors. In the next step, the large subunit joins this 40S•mRNA•tRNAi Met complex, the remaining initiation factors dissociate, and elongation can start with the incorporation of the tRNA carrying the amino acid that is coded for in the second codon.

Eukaryotic and bacterial systems profoundly differ at the stage of initiation. Bacterial 16S rRNA possesses a sequence, to which the characteristic Shine Dalgarno sequence of the mRNA aligns. This alignment facilitates positioning of the mRNA start codon in the P site of the 30S subunit (Shine and Dalgarno, 1974). Initiation factor 1 (IF1) binds at the 30S A site and induces a structural change in the 30S subunit (Allen et al., 2005; Carter et al., 2001; Milon et al., 2008). A ternary complex consisting of fMet-tRNAi fMet (the first amino acid in bacterial proteins is always formyl-methionine) and IF2•GTP binds to the P site, while initiation factor 3 (IF3) ensures that it is an initiator tRNA and not an elongator tRNA that binds (Milon et al., 2008). Positioning of the initiator tRNA in the P site is followed by joining of the large subunit, GTP-hydrolysis of IF2, and finally dissociation of the initiation factors (Simonetti et al., 2008).

In contrast, in eukaryotes, the positioning of the start codon into the 40S P site occurs after ‘ribosome scanning’ of the mRNA (Jackson et al., 2010; Kozak, 1999; Shine and Dalgarno, 1974). The protein machinery required for eukaryotic initiation is by far more complex than in bacteria, where only three initiation factors are needed (Hashem and Frank, 2018; Jackson et al., 2010; Kozak, 1999; Shirokikh and Preiss, 2018).

Eukaryotic initiation starts with the formation of the 43S preinitiation complex, which consists of the 40S subunit, the ternary complex eIF2•GTP•Met•tRNAi Met, eIF3, eIF1 and eIF1A. This 43S complex is ready to bind to the mRNA, which can be circularized via the 5’-bound eIF4F that connects to the 3’ end by binding to PABPs. eIF1 and eIF1A induce an open latch conformation of the 40S that facilitates mRNA binding, while eIF4G plays a key role in loading the mRNA on the 43S preinitiation complex via interactions with eIF3. The main role of the now assembled 48S complex is the scanning of the mRNA until finding a start codon in the correct environment (Jackson et al., 2010; Kozak, 1999). It is eIF1 that enables the 48S complex to discriminate the eligible start codon against other codons or start codons with poor nucleotide context. The establishment of codon- anticodon base pairing between the mRNA start codon and the tRNA leads to eIF1 dissociation, allowing GTP hydrolysis and dissociation of eIF2 under assistance of its GTPase activating protein eIF5. Finally, eIF5B mediates the joining of the 60S subunit and the assembly of the elongation competent 80S ribosome is completed when GTP hydrolysis of eIF5B leads to its dissociation from the 80S (Figure 10).

Additionally to this classical initiation pathway, in the eukaryotic system a group of RNA structures called IRESs (Internal ribosomal entry sites) are able to employ alternative initiation pathways partially or fully independent from the 5’ cap and/or initiation factors (Pelletier and Sonenberg, 1988; Yamamoto et al., 2017).

Elongation is at the heart of translation

Elongation is an iterative process which is aimed at polymerization of the peptide chain until all amino acids encoded in the mRNA are incorporated into the peptide chain. It consists of the steps decoding/tRNA selection, peptidyl transfer and translocation. At each codon between start and stop, the ribosome must undergo one full elongation cycle with the result of one amino acid being added to the peptide chain.

The very first elongation cycle takes place right after initiation and has as its starting point the assembled 80S ribosome with a Met- tRNAi Met in the P site. The A site is empty and must be occupied by a tRNA carrying the amino acid that is encoded next in the mRNA. The selection of the correct tRNA requires eEF1A (EF-Tu in bacteria), a translational GTPase which reaches the ribosome as ternary complex in association with a tRNA and GTP. The interaction of the ribosome with the ternary complex is referred to as ‘decoding’. Here, the tRNA anticodon is brought to the small subunit A site (‘decoding center’) and interacts with the mRNA codon. Correct base pairing in case of complementary codons triggers conformational changes of the ribosome that in turn result in GTP-hydrolysis and eEF1A dissociation (Budkevich et al., 2014; Schmeing and Ramakrishnan, 2009; Schuette et al., 2009; Voorhees et al., 2010).

The dissociation of eEF1A allows the tRNA to be fully accommodated in the 60S A site. There, the CCA-end of the tRNA with the aminoacylated amino acid is positioned in proximity to the CCA-end of the P-site tRNA that carries the first amino acid, or in later rounds of elongation the entire peptide chain built so far. During peptide bond formation, this amino acid/peptide chain from the P-site tRNA is transferred to the A-site tRNA. As consequence, the ribosome contains a peptidyl tRNA in the A site, a deacylated tRNA in the P site, and in case of later elongation rounds, also a deacylated E-site tRNA (Behrmann et al., 2015). The ribosomal conformation at this point is characterized by an unrotated (canonical) 40S subunit that is rolled relative to the 60S subunit (Budkevich et al., 2011, 2014). It is referred to as the classical PRE state, where PRE means pre- translocation.

Translocation prepares the ribosome for a new elongation cycle

While the main task of the first half of the elongation cycle is the polymerization of the peptide chain, the second half serves as the preparation for the next elongation cycle. Translocation is the movement of both tRNAs as well as the mRNA, forming the tRNA2•mRNA module, from the A- and P- to the P- and E sites, respectively. It is catalyzed by eEF2 (EF-G in bacteria) and leads from the pre-translocational (PRE) to the post-translocational (POST)-state ribosome. The POST-state ribosome can then accept a new tRNA in its A site. Alternatively, in case of the very last elongation cycle, the ribosome does not bind a new tRNA, but enters termination (see below).

eEF2/EF-G belongs to the group of translational GTPases and is a five-domain protein (Figure 11). Mammalian eEF2 is very similar to its homologs from other domains of life (Supplemental Figure 6) The G-domain (or domain I) is responsible for GTP hydrolysis and possesses the earlier described characteristic switch loops and G-motifs. Domains 2 and 3 are part of a bridge between the G-domain and the small ribosomal subunit when the factor is bound to the ribosome. Domain 4 is a mimicry of a tRNA anticodon arm and protrudes into the A-site of the ribosome. Domain 4 carries three functionally important loops: loop 1, facing the tRNA (Ramrath et al., 2013), loop 2, facing the small subunit body/platform (Ramrath et al., 2013), and loop 3 in the middle (Figure 11B). A unique feature of domain 4 of the eukaryotic and archaeal elongation factors eEF2 is a diphthamide modification on histidine 715 in mammalia (H699 in yeast) in loop 3 of domain 4 (Oppenheimer and Bodley, 1981).

EF-G/eEF2 function can be specifically targeted by antibiotic agents. Fusidic acid binds between the G-domain and domain 3 of EF-G. It occupies the place of the Pi after its release following GTP-hydrolysis (Figure 11) and prevents EF-G’s dissociation from the ribosome. The concrete impact of fusidic acid on eukaryotic eEF2 is not fully understood yet. Sordarin binds between domain 2 and 3 of eEF2 and prevents dissociation of eEF2 in yeast (Figure 11) (Spahn et al., 2004a).

The diphthamide modification on loop 3 of domain 4 of eEF2 is known for being the target of toxins like diphtheria toxin and exotoxin A. When it is ribosylated, eEF2 cannot function in translation (Davydova and Ovchinnikov, 1990; Oppenheimer and Bodley, 1981).

There exists a current model on translocation based on bacterial and eukaryotic structures, however, a detailed understanding of the mechanism has not been (fully) achieved yet. Studies of bacterial translocation intermediates reveal the following sequence of events:

First, tRNA movement on the large 50S subunit occurs spontaneously after peptide bond formation which alters tRNA affinities and drives subunit rotation (Cornish et al., 2008; Valle et al., 2003). This subunit rotation is reversible and coupled to fluctuations between classical A/A, P/P states and hybrid A/P, P/E states of the tRNAs and has as well been visualized in eukaryotic ribosomes (Agirrezabala et al., 2008; Behrmann et al., 2015; Blanchard et al., 2004; Budkevich et al., 2011; Moazed and Noller, 1989; Munro et al., 2007).

Only after the tRNAs have reached their hybrid positions, EF-G/ eEF2 comes into play. It contributes to the irreversibility and directionality of the translocation reaction. eEF2/EF-G probably binds to the rotated ribosome (Brilot, 2013). According to Rodnina et al., 1997, already at this early stage and before tRNA movement has started, GTP-hydrolysis takes place. The next state visualized is the TI (translocation intermediate)-POST state, where the tRNAs are already translocated on the 30S body/platform and adopt chimeric hybrid (ap/P, pe/E) positions. The ribosome at this state adopts a partly rotated conformation and exhibits a high-degree head swivel (Ramrath et al., 2013; Ratje et al., 2010; Zhou et al., 2014) (Figure 12). After completion of translocation, EF-G dissociates. The ribosome can now accept the next tRNA.

It is not clear how exactly EF-G/eEF2 and GTP-hydrolysis contribute to translocation. Two opposing models were suggested, 1) the ‘Brownian ratchet’ model, according to which binding of EF-G/ eEF2 to the ribosome suffices to deflect the ribosome’s natural, thermodynamically driven propensity to backrotate, resulting in translocation (reviewed in (Spirin, 2009)). 2) The power stroke model, in which the energy of GTP-hydrolysis is used by eEF2/EF-G to actively push the tRNAs in the direction of translocation (Rodnina et al., 1997; Chen et al., 2016).

Termination releases the peptide chain and is followed by recycling

Usually, the last codon of the mRNA is followed by a stop codon (UAA, UAG or UGA). When this stop codon is positioned into the A site, no canonical tRNA will match it. The stop codon is recognized by class-1 release factor eRF1 (RF1 and RF2 in bacteria) which facilitates release of the peptide chain from the P-site tRNA. The activity of eRF1 is supported by class-2 release factor eRF3 (RF3 in bacteria), which belongs to the group of translational GTPases.

After termination, the ribosomal subunits dissociate and are then reused for the next round of translation. This process is called ribosome recycling. In bacteria, ribosomal recycling factor (RRF), EF-G and IF3 act together to disassemble the 80S ribosome (Hirashima and Kaji, 1970). In eukaryotes, ABCE1 splits the 80S ribosome (Jackson et al., 2012; Pisarev et al., 2010). Ligatin (also known as eIF2D) and DENR (density regulated protein) have been found to promote dissociation of tRNA and mRNA from the small subunit in eukaryotes (Skabkin et al., 2013).

Some evidence points towards an alternative event after termination. It is referred to as ‘reinitiation’: Here, recycling and dissociation of the 60S subunit takes place as well, but the 40S subunit does not leave the mRNA. Instead, it continues with scanning and thus can translate a next open reading frame (ORF; see Figure 4B) (Jackson et al., 2012; Skabkin et al., 2013).

mRNA quality control prevents production of degenerated proteins

Corrupt mRNA will lead to defective protein products. To avoid mistakes in protein translation, the cell has developed multi-level control mechanisms to check for integrity and correctness of the mRNA. Already mRNA transcription and maturation are susceptible to errors, and control points at this level are there to detect and eliminate defective mRNA products. In the nucleus, aberrant mRNAs are degraded in the 5’ to 3’ direction by exoribonuclease Xrn2 and in the 3’ to 5’ direction by the nuclear exosome before reaching the cytoplasm (Fasken and Corbett, 2009). In the cytoplasm, the ribosome and protein factors are involved in controlling the translated mRNA in three main ways:

Nonsense-mediated mRNA decay takes place when there is a premature stop codon. Downstream of the premature stop codon, the exon junction complex contains the upstream frameshift proteins (UPF) 2 and 3. Upon recognition of the stop codon, UPF1 is recruited. The interaction of UPF1 with UPF2 activates the mRNA degradation process (Isken and Maquat, 2007).

Non-stop mRNA decay is induced by the absence of a stop codon. The ribosome is stalled at the 3’ end of the mRNA because termination factors are not recruited (Isken and Maquat, 2007). Such stalled ribosomes are recognized by a mechanism that is not yet clearly understood, but apparently involving Dom34•Hbs1 (Hilal et al., 2016; Tsuboi et al., 2012). The exosome is recruited to the ribosome and degrades the mRNA.

In some cases, the ribosome cannot continue translation due to stable secondary structure. The mechanism that rescues a ribosome in this situation is referred to as no-go mRNA decay. In yeast, a Dom34•Hbs1•GTP complex is able to mediate dissociation of such stalled ribosomes (Becker et al., 2011; Shoemaker et al., 2010).