The Termination Phase in Protein Synthesis is not Obligatorily Followed by the RRF/EF-G- Dependent Recycling Phase

Transcription

1 Article DCC YJMBI-65105; No. of pages: 11; 4C: 3, 4, 5, 6, 8 The Termination Phase in Protein Synthesis is not Obligatorily Followed by the RRF/EF-G- Dependent Recycling Phase Bo Qin 1,2,, Hiroshi Yamamoto 1,2,, Takuya Ueda 3, Umesh Varshney 4 and Knud H. Nierhaus 1,2 1 - Max-Planck-Institut für molekulare Genetik, Ihnestrasse 73, D Berlin, Germany 2 - Institut für Medizinische Physik und Biophysik, Charité, Charitéplatz 1, Berlin, Germany 3 - Department of Medical Genome Sciences, Graduate School of Frontier Sciences, The University of Tokyo, Kashiwa, Chiba Prefecture , Japan 4 - Department of Microbiology and Cell Biology, Indian Institute of Science, Bangalore, , India Correspondence to Hiroshi Yamamoto and Knud H. Nierhaus: yamamoto.hiroshi@charite.de Edited by Ruben L. Gonzalez Abstract It is general wisdom that termination of bacterial protein synthesis is obligatorily followed by recycling governed by the factors ribosomal recycling factor (RRF), EF-G, and IF3, where the ribosome dissociates into its subunits. In contrast, a recently described 70S-scanning mode of initiation holds that after termination, scanning of 70S can be triggered by fmet-trna to the initiation site of a downstream cistron. Here, we analyze the apparent conflict. We constructed a bicistronic mrna coding for luciferases and showed with a highly resolved in vitro system that the expression of the second cistron did not at all depend on the presence of active RRF. An in vivo analysis cannot be performed in a straightforward way, since RRF is essential for viability and therefore, the RRF gene cannot be knocked out. However, we found an experimental window, where the RRF amount could be reduced to below 2.5%, and in this situation, the expression of the second cistron of a bicistronic luciferase mrna was only moderately reduced. Both in vitro and in vivo results suggested that RRF-dependent recycling is not an obligatory step after termination, in agreement with the previous findings concerning 70S-scanning initiation. In this view, recycling after termination is a special case of the general RRF function, which happens whenever fmet-trna is not available for triggering 70S scanning Published by Elsevier Ltd. Introduction It is generally accepted that in eubacteria, termination of translation of a cistron is followed by a ribosomal recycling factor (RRF)- and EF-G-mediated dissociation of ribosomes from mrna termed as the recycling phase of protein synthesis, a process supported by IF3 [1 5]. RRF is an essential factor [6], universal in Eubacteria but absent in Achaea and Eukaryota [7]. X-ray and cryo-electron microscopic studies have shed light on the mechanism of dissociation of post-termination complexes [8 13]. It is thought that the recycling phase is an obligatory step following the termination phase in eubacteria and is required for providing 30S subunits for initiating the next round of translation of a cistron. An obligatory recycling phase after termination cannot easily be reconciled with numerous hints over the last decades for a re-initiation, where 70S ribosomes rather than 30S subunits were conjectured to be involved [14 17], although no mechanistic features of a 70S re-initiation were reported. Recently, we have demonstrated a frequent initiation mode in bacteria, the so-called 70S-scanning initiation, according to which 70S ribosomes do not dissociate after termination but rather scan along the mrna until reaching the initiation site of the downstream cistron of the same mrna. 30S scanning was not observed; binding of fmet-trna triggers 70S scanning, which occurs in the absence of energy-rich compounds (e.g., GTP) and seems to be driven by unidimensional diffusion [18]. A recent report about 70S ribosomes / 2016 Published by Elsevier Ltd. J Mol Biol (2016) xx, xxx xxx

2 2 Termination Phase in Protein Synthesis with tethered subunits preventing full dissociation into ribosomal subunits demonstrated that these ribosomes were able to synthesize active proteins [19]. The apparent conflict between these two mechanisms, RRF-dependent recycling versus 70S-scanning initiation, is analyzed here. If the recycling phase is an essential prerequisite for the subsequent initiation event, depleting the RRF pool should result in a collapse of protein synthesis. Here, we show in vitro and in vivo with suitable bicistronic mrna constructs that RRF is not essential for the expression of a downstream cistron. Results The RRF/EF-G-dependent recycling phase is thought to be an obligatory and essential step after translation termination [20]. First, we tested whether the RRF used in the following in vitro experiments is active. To this end, we applied a test reported by Zavialov et al. [3], who used a 7-, 22-, and 75-fold excess of RRF in the presence of IF3 and EF-G. We incubated a 70S complex containing a deacylated trna in the P site with EF-G GTP, IF3, and different amounts of RRF (5- and 15-fold excess of 70S ribosomes; Fig. 1). The addition of RRF triggered a clear dissociation effect, viz. the RRF preparation we used is active. In the next experiment, we asked whether or not the translation of the second cistron in a bicistronic mrna depends on the presence of RRF. A modified PURE system was applied [21], which allows protein synthesis in the presence or complete absence of the factors IF1, IF3, and RRF. Our PURE system is further characterized by a lowered total Mg 2+ concentration (8.5 instead of 13 mm), which yields a free Mg 2+ concentration of 2.3 mm Mg 2+. The low Mg 2+ concentration is responsible for an extremely slow equilibrium rate between 70S ribosomes and the subunits during the incubation of the translation assay (Fig. S4 in Ref. [18]); during the first 15 min of incubation, no equilibrium is observed between 70S and subunits. We designed a bicistronic mrna coding for luciferases from Renilla (Rluc) and firefly (Fluc), respectively (Fig. 2, top). The mrna is similar to that used in Ref. [18], where we have shown that blocking either the expression of the first cistron or the intercistronic runway for scanning 70S ribosomes by hybridizing antisense oligo-dnas strongly reduces the expression of the second cistron (Fig. 1 in Ref. [18]). Furthermore, we introduced a secondary structure hiding the Shine Dalgarno (SD) sequence in front of the second cistron (Fig. 2). Such a secondary structure prevents 30S binding but can be resolved by scanning 70S ribosomes (see Fig. 2C of Ref. [18]). Another feature of the mrna shown in Fig. 2 is a fusion of the second cistron Fluc with a part of the sequence of the regulatory protein SecM (SecM ), which blocks translating ribosomes on the mrna [22]. Therefore, each ribosome will Fig. 1. Post-termination complexes containing a deacylated trna in programmed 70S ribosomes were incubated with the indicated components. When RRF was added (lanes 4 and 5) in 5 or 15 fold excess over the 70S ribosomes, dissociation occurs.

3 Termination Phase in Protein Synthesis 3 translate the Fluc cistron only once; afterwards, a recycling phase cannot take place (see Fig. 2B in Ref. [18], where the same Fluc construct was used; the control in Fig. S3B demonstrated that Fluc was present exclusively as peptidyl trna). The SecM sequence is long enough to ensure that the entire Fluc protein can pass through the ribosome exit tunnel before stalling occurs, allowing an undisturbed Fluc folding. Before we consider the expected outcome of the experiments using the mrna shown in Fig. 2, we should consider an important difference between 30S and 70S initiation. In 70S ribosomes, mrna is not located in an open cleft as in free 30S subunits but rather in a tunnel; therefore, free 70S ribosomes cannot directly initiate at internal initiation sites of an mrna (like the Fluc initiation site) but must thread the mrna from the 5 -end (see Fig. 2B of Ref. [18]). [23,24,18,25,26]. When we add 70S ribosomes to our modified PURE system in the presence of the mrna shown in Fig. 2, ribosomes will scan down to the initiation site of the first cistron and translate Rluc. If the RRF/ EF-G/IF3-dependent recycling phase compulsorily follows the translation termination of the first Rluc cistron, all ribosomes will leave the mrna and dissociate into subunits, resulting in a block of Fluc synthesis, because 30S-binding initiation is prevented by the secondary structure hiding the SD sequence. In contrast, if 70S scanning occurs after termination of the Rluc synthesis, 70S ribosomes will scan down the intercistronic region of the mrna, resolve the secondary structure with the SD sequence, and translate the Fluc cistron, regardless of whether or not RRF is present. Translation assays were performed with 70S ribosomes. Rluc and Fluc synthesis was monitored in the presence and absence of RRF, IF1, and IF3. In the absence of both initiation factors, negligible amounts of luciferases were synthesized (Fig. 2); in their presence, the yield of both luciferases was independent of the presence of RRF. The clear-cut results indicate that the RRF/EF-G recycling phase is not an obligatory step after translation of a cistron in our in vitro system. To seek more evidence, we turned to in vivo systems and tried to manipulate the RRF amounts in the cells; due to its essential nature, the gene cannot be knocked out. By chance, we found an experimental window, through which this problem could be tackled. We worked with an Escherichia coli strain carrying a temperature-sensitive RRF. Originally isolated by the Kaji group (LJ14 strain; [27]), the strain used here was modified; it carries Tn10 conferring tetracycline resistance with ~25% linkage to the frr locus as selection marker [28]. Fig. 2. Expression of the bicistronic mrna coding for Rluc and Fluc in the PURE system containing 70S ribosomes and all necessary components, and IF1/IF3 and the ribosomal recycling factor RRF when indicated (molar ratio of RRF:70S = 1:1). IR, intercistronic region; the blue stretch of the IR contains a heteropolymeric sequence of 20 nt that is 21 nt away from the upstream Rluc stop-codon and 40 nt from the downstream Fluc initiation AUG codon [18].

4 4 Termination Phase in Protein Synthesis We confirmed the observation of Hirokawa et al. [29] that after a shift from permissive (30 C) to non-permissive temperatures (43 C), temperaturesensitive RRF (RRF ts ) was strongly diminished within 2 h as demonstrated by Western blotting using antibodies against RRF (Fig. 3a, lane 2). The wild-type (WT) strain showed normal levels of RRF. Most surprisingly, after an overnight incubation even at the permissive temperature of 30 C, the mutant frr ts strain has lost its RRF in contrast to the WT strain (Fig. 2A, lanes 7 and 8 versus 5 and 6). A possible explanation is that the active conformer of RRF ts requires transient interactions with the ribosome to maintain its active state. If so, it might lose both its active conformation and compact structure in the absence of significant protein synthesis during the stationary phase and is thus degraded. Fig. 3. Western blots analyzing the amounts of RRF normalized via the L2 band. (a) RRF ts disappears when incubated at the non-permissive temperature of 43 C for 120 min (lane 2) or overnight (16 h) at the permissive temperature of 30 C (lanes 7 and 8). Lane 3, control L2; lane 4, control RRF, the isolated RRF contains a His-tag and therefore bands a bit higher than the cellular RRF. (b) Western blot from kinetics at 25 C (permissive temperature) from 15 to 240 min after an overnight incubation (16 h) at permissive temperature (30 C). (c) Relative amounts of RRF (RRF ts /RRF WT ) derived from the band intensities of (B) and normalized via the intensities of the corresponding L2 bands (green columns). In addition, the corresponding OD 600 values from the cell suspensions of WT and RRF ts strains are shown. In the next experiment, we tested how much time the RRF ts strain needs to replenish the RRF factor pool again after a 30 C overnight incubation. After diluting the overnight culture 100-fold in fresh LB medium, the incubation was continued at the permissive temperature of 25 C. The density of the culture was measured at distinct time points and the amount of RRF assessed by Western blotting. The first faint RRF band begins to appear after 60 min, and only after 240 min, it reaches a level similar to that in the strain WT for frr (Fig. 3b and c). The onset of growth after the lag phase is observed only after 60 min, when the RRF levels exceed 20% of the WT levels, and after 240 min, the RRF ts strain has reached the growth rate of the WT cells (Fig. 3c). The results indicate that RRF is virtually absent during the first 30 min, thus providing a time window of at least 25 min at 25 C for testing protein synthesis under negligible concentrations of RRF in vivo. We transformed both WT and RRF ts strains with a bicistronic Rluc/Fluc reporter plasmid (sequence in Fig. 4, top), but this time lacking both the secondary structure hiding the SD sequence and the Fluc SecM fusion used in the experiment shown in Fig. 2. After an overnight incubation at 30 C, cells were diluted in fresh LB medium. IPTG was added immediately to induce expression of the luciferases, and incubation was continued at 25 C for 25 min. At this time point, the RRF levels were at least 40-fold lower in the RRF ts strain than in the WT strain (Fig. 4a, green bars). Despite a negligible amount of RRF, an arrest of protein synthesis was not observed in the RRF ts strain. The expression ratio Fluc/Rluc was only moderately reduced by less than 30% (yellow-red bars in Fig. 4a, see the measured values in Fig. 4b). It is clear that an RRF-dependent recycling phase after translation of the first cistron is not a prerequisite for translating the downstream cistron. One important aspect of the RRF-dependent recycling phase was thought to provide 30S subunits for the 30S mode of initiation [4,30]. If so, such a dramatic RRF reduction as that seen in Fig. 4a should deprive the cells for free ribosomal subunits. To test this point, we analyzed sister aliquots of mutant and WT cell lysates from the experiment shown in Fig. 4 in a sucrose gradient under conditions that hardly change the equilibrium between ribosomal subunits and 70S ribosomes at 0 C [20 mm Hepes (ph 7.6) at 0 C, 6 mm Mg 2+, and 30 mm K + ] [31]. The relative molar ratios of (70S plus polysomes):50s:30s are practically constant in WT cells (normal RRF amounts, Fig. 5a) and in cells deprived of RRF (Fig. 5b; see the comparison in Fig. 5c). Although protein expression is damped after stationary phase, these results do not support the view that the RRF-mediated recycling phase is an essential source for providing ribosomal subunits for the 30S mode of initiation.

5 Termination Phase in Protein Synthesis 5 Fig. 4. In vivo expression of the bicistronic mrna shown above (a) in the WT and RRF ts strains. After an overnight incubation of 30 C for 19 h, the cell suspensions were diluted in fresh LB medium containing 1 mm IPTG for triggering the expression of the luciferases during an incubation of 25 C for 25 min. (a) Comparison of the RRF amounts (green bars) with the luciferase ratio Fluc/Rluc (yellow-red bars). (b) Absolute values of the luciferases in WT and RRF ts cells normalized via the input of A 260 units of the S30 lysate (for details, see Materials and Methods). Discussion The RRF-dependent recycling phase does not obligatorily follow termination RRF can increase oligopeptide synthesis two- to threefold as shown in highly resolved in vitro tests using mrnas coding for oligo-peptides with six or less residues [32 34]. Translation experiments with short peptides are designed for an extreme recycling activity and may not reflect physiological synthesis of proteins with an average length of 300 to 400 aa in bacteria. Concerning the synthesis of larger proteins, two different types of RRF effects were reported: (i) in vitro synthesis of the coat protein from the RNA phage R17 performed with fractions of S100 enzymes was stimulated up to 1.5-fold by addition of RRF fractions [32], a result not arguing for an essential role of the RRF recycling phase. However, these results might not be conclusive because the system was rather crude. Similar results were reported, when unaffected expression of β-galactosidase was observed in a strain harboring RRF ts under semipermissive conditions (39 C), depriving the pool of active RRF and also arguing against an essential role of the RRF-dependent recycling phase. However, the residual RRF amounts in the cells were not checked [35]. (ii) It was reported that after termination, a re-initiation of protein synthesis might occur directly after the stop codon [32] or in any reading frame [27] under conditions of RRF deprivation ( unscheduled translation ). The observed effects were very weak; for example, expression of β-galactosidase lacking some N-terminal amino acids revealed a signalto-noise ratio of 4 (Fig. 6 in Ref. [27]), whereas normal

6 6 Termination Phase in Protein Synthesis Fig. 5. Sucrose gradient patterns from the samples analyzed for luciferase expression shown in Fig. 4. (a) WT strain; (b) RRF ts strain; (c) Relative amounts of ribosomes and ribosomal subunits. The areas above the red lines were determined; all areas of one A 260 profile were taken as 100%. β-galactosidase expression can achieve a 1000-fold ratio. Therefore, only a minority of ribosomes undergo unscheduled translation. These interesting but inconclusive indications prompted us to analyze the effects of RRF on protein synthesis under highly controlled in vitro conditions. We designed a bicistronic mrna coding for Rluc and Fluc with the features described in the Results section. The features assure that the translation of the second Fluc cistron depends on the translation of the first Rluc cistron, because direct 30S-binding initiation of the Fluc cistron was prevented by hiding the SD sequence in the secondary structure that, however, could be resolved by scanning 70S ribosomes [18]. Consequently, Fluc expression would depend exclusively on 70S-scanning initiation. At least three scenarios could be considered: (i) If the RRF/EF-G-dependent recycling phase were an obligatory and essential step after translation of a cistron (here Rluc), forcing all ribosomes to leave the mrna and to dissociate into subunits, there would be no expression of the downstream cistron Fluc. (ii) If recycling would not be obligatory but at least would be able to efficiently compete with 70S scanning, Fluc expression would be sharply reduced in the presence of RRF. (iii) Finally, the mrna construct is also a highly sensitive test of the unscheduled translation :if in the absence of RRF the ribosomes would be quantitatively subjected to unscheduled translation in any frame as proposed, the fraction of scanning 70S ribosomes initiating the second Fluc cistron in a defined frame would be strongly reduced, thus also strongly impairing the amount of synthesized Fluc. None of the predictions could be confirmed: in vitro translation in the highly defined PURE system demonstrated that the yield of both luciferases was practically the same, independent of presence or absence of RRF (Fig. 2). The following conclusions can be drawn from these results: (i) an RRF-dependent recycling is not an obligatory step after translation termination of a cistron. (ii) Unscheduled translation does not play an important role in the absence of RRF. (iii) 70S scanning is majority after termination of Rluc cistron translation. (iv) Assuming a frequent RRF/EF-G recycling of 70S ribosomes after the synthesis of Rluc, 70S ribosomes will dissociate into subunits, and thus, less 70S ribosome will be available for the scanning process and the translation of the downstream Fluc cistron. Observing identical amounts of both Rluc and Fluc in the presence or absence of RRF means that 70S scanning is the prevailing event after translation of the Rluc cistron under our experimental conditions. These surprising results prompted us to check again the effects of RRF on protein synthesis in vivo in virtual absence of RRF. It has been reported that RRF ts is degraded at non-permissive temperature and is present at permissive temperature at lower amounts than RRF in the WT strain [29]. In extending these observations, we found that RRF ts was practically absent after an overnight incubation at permissive temperature (stationary-phase cells, Fig. 3a). After diluting the overnight cell suspension in fresh LB medium and continuing incubation at permissive temperature, RRF ts did not recover for more than 30 min (Fig. 3b), which opened an experimental window for an analysis in vivo. The essential outcome of our analysis was that at least 40-fold reduction of the cellular RRF amount severely blocked growth (Fig. 3c) but impaired protein synthesis only moderately by about 30% (Fig. 4a and b). Forcing the RRF level down to 2.5% of the WT amounts did not affect the relative amounts of ribosomal subunits in vivo (Fig. 5).

7 Termination Phase in Protein Synthesis 7 We observed in vitro and in vivo that the absence of RRF only slightly affects protein synthesis. These results suggest that the RRF-dependent recycling phase does not necessarily follow translation termination in bacteria. A recent report [18] indicates that instead of recycling 70S ribosomes stay on the mrna and scan to the initiation site of the downstream cistron. Which RRF activity might be responsible for the factors' essential feature? An extensive literature about RRF has been accumulated over the last five decades. Revisiting the published RRF data in depth will be presented elsewhere. Here, we will give only few arguments suggesting a broader, functional spectrum of RRF, where a possible RRF-depending recycling of post-termination ribosomes is not obligatorily. The first hint comes from an assay testing the activity of RRF. In this assay, the conversion of polysomes to monosomes/70s ribosomes is monitored [1]. Efficient conversion required the removal of the peptidyl group from the ribosome-bound trna by puromycin [36]. Therefore, the RRF assay includes a treatment with puromycin before the conversion is triggered by RRF and EF-G. The resulting 70S complexes carrying deacylated trnas are considered to be models for post-termination ribosomes ready for the recycling phase. This assumption is not valid for most ribosomes, because either they are vacant ribosomes or will have a sense codon at the A site in contrast to post-termination ribosomes, which have a stop codon at the A site. Therefore, 70S ribosomes carrying a deacylated trna at the P site (P-tRNA) are a target for RRF and EF-G. A genetic screen gathered further evidence; it revealed a suppressor mutation of the temperaturesensitive phenotype of the peptidylhydrolase (Pth ts ). The mutation was located in the promoter region of the RRF gene frr [37] and caused low amounts of RRF. Pth is another essential gene in bacteria [38] and hydrolyses peptidyl-trnas rather than fmet-trna, which had accidentally fallen from the ribosome [39,40] or are still present on stalled ribosomes (see next paragraph). It follows that low amounts of RRF can prevent the accumulation of peptidyl-trna in the cytosol in a Pth ts background. This observation suggests that RRF directly or indirectly is involved in peptidyl-trna release from the ribosome in cooperation with EF-G and Pth. Targeting stalled ribosomes would be an important if not an essential RRF function. In fact, it has been reported that RRF participates in the release of peptidyl-trnas from stalled ribosomes in the elongation phase [41]. Another report directly demonstrated a peptidyl-trna release in vivo [42] carrying long peptides (about 350 residues) on stalled ribosomes. The authors also demonstrated that Pth can hydrolyze peptidyl-trna on stalled ribosomes, when a sense codon is at the A site. Possibly, Pth can also hydrolyze peptidyl-trna on the ribosome, when a peptidyl-trna resides at the A site. IF3 and RF3 seem to cooperate with RRF as indicated in the following observations. Growth defects of a Pth ts strain at non-permissive temperature could be healed not only by a reduced RRF synthesis as expected but also to a lesser extent with a deleted RF3 gene; restoration of viability of the Pth ts strain was optimal with both reduced RRF synthesis and deletion of the RF3 gene [37]. A similar full restoration of the Pth ts phenotype was reported from a strain containing a functionally compromised IF3 together with the mentioned insertion in the promoter of the RRF gene, causing low cellular levels of RRF [43]. Furthermore, the same authors demonstrated that overexpression of IF3 could heal the temperature-sensitive phenotype of RRF ts [43]. Curing the temperature-sensitive feature of RRF ts indicates that IF3 is linked to an essential RRF function in addition to its involvement in the RRF post-termination recycling activity. It follows that the RRF-dependent rescue of stalled ribosomes consists of the triad RRF, EF-G GTP, and IF3. There are two more systems, which can hydrolyze peptidyl-trna on the ribosome preferentially on stalled ribosomes. The factor ArfB (YaeJ) is able to hydrolyze peptidyl-trnas at the P site codon independently [44,45]. The same is true for the cooperative action of ArfA (YdhL) and RF2, which however preferentially cleave peptidyl-trna, when the A site does not contain a codon [45,46]. Therefore, ArfA and RF2 are an alternative for the tmrna/smb rescue system, which requires a peptidyl-trna at the P site and a codon-free A site [47]. All three systems, viz. PtH, ArfB, and ArfA/RF2, are targeting stalled ribosomes and generating 70S complexes with a deacylated P-tRNA and a sense codon at the A site, which are ready targets for the suggested RRF/EF-G/IF3 rescue system possibly essential for the viability of the bacterial cell. The post-termination complex is similar, because it also carries a deacylated P-tRNA but rather a stop codon at the A site. We suggest that the apparent conflict between recycling and 70S scanning can be reconciled in the following way. If an fmet-trna for any reason is not available to chase the deacylated P-tRNA of a post-termination complex and trigger 70S scanning, the stalled post-termination complex will be recycled (Fig. 6). Even if fmet-trna is available for 70S scanning, there is a case for RRF/EF-G recycling. We have seen 70S scanning on intercistronic regions with a length of up to ~60 nt, and more than 70% of the intercistronic regions are shorter than 30 nt [18]. It seems that 70S scanning

8 8 Termination Phase in Protein Synthesis Fig. 6. Simplified picture of the general role of RRF reconciling recycling and 70S scanning after termination. We hypothesize that the general role of RRF is resolving stalled ribosomes regardless whether a sense or a stop codon is present at the A site. A ribosome stalled in translation can be recognized by numerous factors mentioned in the text. These factors trigger the release or hydrolysis of the peptidyl-trna at the P site and somehow also the release of the E-tRNA. The product is a complex of programmed 70S ribosome with a deacylated P-tRNA, which is recognized by the triad RRF/ EF-G GTP/ IF3 and released from the mrna and thus ready to enter a new initiation phase again. It is not yet clear, whether the released 70S has to dissociate first before entering the initiation phase or straight initiates as a 70S ribosome. If a stop codon is at the A site, the post-termination complex binds an fmet-trna, which triggers 70S scanning to the initiation site of the downstream cistron. In case fmet-trna is not available, RRF, EF-G GTP, and IF3 provoke the known recycling reactions. may work for the most of genes. However, the successful 70S scanning depends on optimal SD sequence in front of AUG codon (see Fig. S7 of Ref. [18]). This result indicates that 70S scanning is difficult to manage without SD sequence even if the intercistronic region is below ~ 60 nt. Moreover, we have not yet analyzed the maximum distance of 70S scanning. How 70S scanning and the RRF/ EF-G GTP-dependent recycling compete after the termination event still remains to be elucidated. Materials and Methods Buffers, ribosomes, mrnas H 20 M 6 N 30 SH 4, 20 mm Hepes-KOH (ph 7.6) at 0 C, 6mMMg(Ac) 2, 30 mm NH 4 (Ac), and 4 mm β-mercaptoethanol; H 20 M 4.5 K 150 SH 4 Spd 2 Spm 0.05,20mM Hepes-KOH (ph 7.6) at 0 C, 4.5 mm Mg(Ac) 2, 150 mm KAc, 4 mm β-mercaptoethanol, 2 mm spermidine, and 0.05 mm spermine; SDS sample buffer, 60 mm Tris HCl (ph 6.8), 2% SDS, 10% glycerol, 200 mm β- mercaptoethanol, and 0.05% bromophenol blue. The isolation of ribosomes, the use of the highly resolved PURE system, and the construction of the bicistronic mrnas coding for Rluc and Fluc used in vitro (Fig. 2) have been constructed in the following way: Rluc and Fluc-SecM were amplified from plasmids pet23c R-IR-F and pet23c Fluc-SecM using the primers T7 promoter and Rluc-rev or Fluc-fw and T7 terminator (see Ref. [18]). Primers Rluc-rev and Fluc-fw were 5 -phosphorylated prior to amplification. The two genes were ligated and cloned using BglII and EcoRI into pet23c (Novagen); the resulting plasmid was named pet23c R-Fluc-SecM. The intercistronic region with the hidden SD sequence (IR-HSD) was introduced as double-stranded oligonucleotides providing 5 overhangs compatible to

9 Termination Phase in Protein Synthesis 9 BamHI and NcoI restriction sites via ligation into pet23c R-Fluc-SecM. Oligonucleotides used: IR-HSD sense (5 -GATCCACCCCACCCCACCCCGCAG GATCGAGCGCAGACTGAC CCCACCCCACCCCACCCCTCTCCATTAGGA GAACTAC-3 ) and IR-HSD antisence (5 -CATGGTAGTT CTCCTAATGGAGAGGGGTGGGGTGGGGTG GGGTCAGTCTGCGCTCGATCCTGCGGGGTG GGGTGGGGTG-3 ). The plasmid construction for the expression of the luciferases from the bicistronic mrna used in vivo (Fig. 4) has been described [18]. Testing the activity of RRF We incubated 10 pmol 70S with 20 pmol trna Phe in the presence of poly(u) (final concentration 8 mg/ml) in buffer H 20 M 4.5 K 150 Me 4 Spd 2 Sp 0.05 at 37 C for 15 min in a reaction volume of 10 μl (post-termination state). Then, 40 pmol EF-G, 150 pmol IF3, 0.5 mm GTP, and RRF at different amounts (0 or 50 or 150 pmol) were added, increasing the reaction volume to 100 μl of the same buffer; an incubation at 37 C for 20 min followed. The reaction mixtures were loaded on 10 30% sucrose gradients in binding buffer and centrifuged at 24,000 rpm for 20 h in a SW40 rotor. The ribosome profiles were monitored at A 260. Experiments with a strain harboring a RRF ts Generation of E. coli LJ14.1 (Tn10) and LJ14.2 (Tn10 frr ts ) was described earlier [28]. E. coli LJ14.1 and LJ14.2 (WT and RRF ts strains, respectively) were grown overnight in LB medium (ampicillin 25 mg/l and tetracycline 12.5 mg/l) at 30 C. After determining the OD 600, the suspension was diluted (starting OD 600 = 0.05) with 400 ml LB medium with LB medium (ampicillin 25 mg/l and tetracycline 12.5 mg/l), and incubation continued at 25 C. At various time points (Fig. 3B and C), aliquots were taken (100 ml at the beginning, later 50 ml) and the cells harvested. SDS-sample buffer was added and proteins were denatured at 95 C for 5 min. The samples (corresponding to 0.1 OD 600 ) were loaded onto 15% polyacrylamide gels for assessment of the RRF amounts by Western blot. The Western blot was performed as described [18] using polyclonal antibodies against RRF and the ribosomal protein L2 as input control and for normalizing the RRF bands. Expressing Renilla and firefly luciferases in vitro Construction of the bicistronic mrna coding for Rluc and Fluc (shown in Fig. 2, top) and in vitro expression in a modified PURE system have been described in Ref. [18]. The modified PURE system The system, as well as the purified His-tagged IF1 and IF3, was provided by Dr. T. Ueda, Tokyo. The expression in the PURE system was performed according to Ref. [48] with the following modifications. The final concentration of ribosomes in the reaction was 0.5 μm. The amount of translation factors was accordingly reduced; our modified PURE system lacked IF1, IF3, and RRF, if not indicated. T7 polymerase, as well as CTP and UTP were omitted from the reaction mixture; GTP and ATP were present in 2 mm each. The final Mg2+ concentration was decreased to 8.5 mm. Note that the free Mg2+ concentration in our PURE system was about 2.5 mm in the presence of ATP and GTP (each 2 mm), which binds about 1 to 1.5 mm Mg2+ per mm NTP [49]. Thisisneartothein vivo conditions [50]. The reaction was incubated for 2 h at 30 C. Expressing Renilla and firefly luciferases in WT and RRF ts strains We thawed 60 μl competent E. coli cells (WT and RRF ts ) on ice and then transformed them with the plasmids carrying the bicistronic mrna coding for Rluc and Fluc (1 μl with approximately 10 ng of DNA); then, these cells were mixed and well kept on ice for 5 min. Electroporation was performed with 1.5 kv, 200 Ω resistance, and C = 25 μf capacity by applying pulsed electrical field for about 4 5 ms, which causes pore formation and enables foreign plasmid to enter into the cells. Then, 0.5 ml fresh LB medium were added into the cuvettes and incubated at 30 C for 1 h in 1.5 ml Eppendorf tubes. We plated 0.1 ml culture on LB plates with ampicillin 25 mg/l and tetracycline 12.5 mg/l and incubated it overnight for single colonies to grow. E. coli cells (WT and the RRF ts strain) were grown overnight in 30 ml LB medium containing ampicillin (25 mg/l) and tetracycline (12.5 mg/l) at 30 C. After determining the OD 600, the suspension was diluted with 200 ml LB medium (0.2 OD 600 ) containing ampicillin (25 mg/l), tetracycline (12.5 mg /L), and IPTG (1 mm) and incubated at 25 C for 25 min. The cells were harvested, resuspended in H 20 M 4.5 K 150 SH 4 Spd 2 Spm 0.05, supplemented with lysozyme (0.4 mg/ml) and 10 μl DNase (RNase free), and subjected to three freeze thaw cycles to lyse the cells. The membranes were removed by two low-speed centrifugation steps (9000 g for 5 min and g for 30 min). Then, 5 μl of the resulting S30 (diluted 100-times) was taken for measuring dual luciferase activity by Dual-Glo Luciferase Assay System (Promega); the chemiluminescence was measured in a Centro Microplate-Luminometer LB 960 (Berthold Technologies). Expression values of Rluc and Fluc in WT and the RRF ts strain were normalized via the A 260 input into the reaction mixture. The RRF amount was determined by a Western blot as described [18] using polyclonal antibodies against RRF and the ribosomal protein L2 as input control and for normalizing the RRF bands. Sucrose gradients of the cell lysates (S 30) We loaded 1 A 260 unit of S30 of E. coli cells (WT and RRF ts ) onto 10 30% sucrose gradient made in H 20 M 6 K 30 SH 4 buffer. A centrifugation with the swing-out rotor SW55 was performed (at 39,000 rpm for 135 min), and the gradient was recorded by gradient runner (LKB). Gradient runner is an instrument that measures the absorbance of samples at A 260, while the sucrose gradient is pumped out. Based on the absorbance value, the curve was plotted and the peak areas were analyzed with the software Origin8 (OriginLab)

10 10 Termination Phase in Protein Synthesis Acknowledgments We thank Drs. Akira and Hideko Kaji for both supplying us with RRF antibodies and intensive discussions, and Drs. Markus Pech, Gene Center, University of Munich, Germany, and John Achenbach, NOXXON Pharma AG, Berlin, Germany, for help and discussions. We thank Profs. M. Vingron, Max-Planck-Institut für Molekulare Genetik in Berlin-Dahlem, and C.M.T. Spahn, Institute of Medical Physics and Biophysics at the Charité, Berlin, for generous support. This work was supported by the China Scholarship Council (File No ) to B.Q., the HFSPO Research Grant Funds (HFSP-Ref. RGP0008/2014) to K.H.N. and T.U., and the Alexander-von-Humboldt Foundation (grant GAN STP-2) to H.Y., the Max-Planck-Gesellschaft and the Verein zur Förderung junger Wissenschaftler (Berlin). Financial Disclosure: This work was supported by the Max-Planck-Gesellschaft and the Verein zur Förderung junger Wissenschaftler (Berlin). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Received 27 November 2015; Received in revised form 18 May 2016; Accepted 22 May 2016 Available online xxxx Keywords: protein synthesis; termination of protein synthesis; ribosome recycling; RRF function; recycling phase not obligatory B.Q. and H.Y. contributed equally to this work. Abbreviations used: RRF, ribosomal recycling factor; Rluc, Renilla luciferase; Fluc, firefly luciferase; SD, Shine Dalgarno; RRF ts, temperature-sensitive RRF; WT, wild-type; Pth, peptidylhydrolase; Pth ts, temperature-sensitive Pth. References [1] A. Hirashima, A. Kaji, Role of elongation factor G and a protein factor on the release of ribosomes from messenger ribonucleic acid, J. Biol. Chem. 248 (1973) [2] A. Rao, U. Varshney, Specific interaction between the ribosome recycling factor and the elongation factor G from Mycobacterium tuberculosis mediates peptidyl-trna release and ribosome recycling in Escherichia coli, EMBO J. 20 (2001) [3] A.V. Zavialov, V.V. Hauryliuk, M. Ehrenberg, Splitting of the posttermination ribosome into subunits by the concerted action of RRF and EF-G, Mol. Cell. 18 (2005) [4] G. Hirokawa, R.M. Nijman, V.S. Raj, H. Kaji, K. Igarashi, A. Kaji, The role of ribosome recycling factor in dissociation of 70S ribosomes into subunits, RNA. 11 (2005) [5] G. Hirokawa, N. Iwakura, A. Kaji, H. Kaji, The role of GTP in transient splitting of 70S ribosomes by RRF (ribosome recycling factor) and EF-G (elongation factor G), Nucleic Acids Res. 36 (2008) [6] L. Janosi, I. Shimizu, A. Kaji, Ribosome recycling factor (ribosome releasing factor) is essential for bacterial growth, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) [7] L. Janosi, R. Ricker, A. Kaji, Dual functions of ribosome recycling factor in protein biosynthesis: disassembling the termination complex and preventing translational errors, Biochimie 78 (1996) [8] M. Selmer, S. Al-Karadaghi, G. Hirokawa, A. Kaji, A. Liljas, Crystal structure of Thermotoga maritima ribosome recycling factor: a trna mimic, Science. 286 (1999) [9] R.K. Agrawal, M.R. Sharma, M.C. Kiel, G. Hirokawa, T.M. Booth, C.M. Spahn, et al., Visualization of ribosome-recycling factor on the Escherichia coli 70S ribosome: functional implications, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) [10] N. Gao, A.V. Zavialov, W. Li, J. Sengupta, M. Valle, R.P. Gursky, et al., Mechanism for the disassembly of the posttermination complex inferred from cryo-em studies, Mol. Cell 18 (2005) [11] A. Weixlbaumer, S. Petry, C.M. Dunham, M. Selmer, A.C. Kelley, V. Ramakrishnan, Crystal structure of the ribosome recycling factor bound to the ribosome, Nat. Struct. Mol. Biol. 14 (2007) [12] R.D. Pai, W. Zhang, B.S. Schuwirth, G. Hirokawa, H. Kaji, A. Kaji, J.H. Cate, Structural insights into ribosome recycling factor interactions with the 70S ribosome, J. Mol. Biol. 376 (2008) [13] T.Yokoyama,T.R.Shaikh,N.Iwakura,H.Kaji,A.Kaji,R.K. Agrawal, Structural insights into initial and intermediate steps of the ribosome-recycling process, EMBO J. 31 (2012) [14] A. Sarabhai, S. Brenner, A mutant which reinitiates the polypeptide chain after chain termination, J. Mol. Biol. 27 (1967) [15] M.R. Adhin, J. van Duin, Scanning model for translational reinitiation in eubacteria, J. Mol. Biol. 213 (1990) [16] H.U. Petersen, A. Danchin, M. Grunberg-Manago, Toward an understanding of the formylation of initiator trna methionine in prokaryotic protein synthesis. II. A two-state model for the 70S ribosome, Biochemistry 15 (1976) [17] K. Saito, L.C. Mattheakis, M. Nomura, Post-transcriptional regulation of the str operon in Escherichia coli. Ribosomal protein S7 inhibits coupled translation of S7 but not its independent translation, J. Mol. Biol. 235 (1994) [18] H. Yamamoto, D. Wittek, R. Gupta, B. Qin, T. Ueda, R. Krause, et al., 70S-scanning initiation is a novel and frequent initiation mode of ribosomal translation in bacteria, Proc. Natl. Acad. Sci. U. S. A. 113 (2016) E1180 E1189. [19] C. Orelle, E.D. Carlson, T. Szal, T. Florin, M.C. Jewett, A.S. Mankin, Protein synthesis by ribosomes with tethered subunits, Nature (2015) [20] G. Hirokawa, N. Demeshkina, N. Iwakura, H. Kaji, A. Kaji, The ribosome-recycling step: consensus or controversy? Trends Biochem. Sci. 31 (2006) [21] Y. Shimizu, T. Ueda, PURE technology, Methods Mol. Biol. 607 (2010)

11 Termination Phase in Protein Synthesis 11 [22] S. Bhushan, T. Hoffmann, B. Seidelt, J. Frauenfeld, T. Mielke, O. Berninghausen, et al., SecM-stalled ribosomes adopt an altered geometry at the peptidyl transferase center, PLoS Biol. 9 (2011), e [23] A.G. Balakin, E.A. Skripkin, I.N. Shatsky, A.A. Bogdanov, Unusual ribosome binding properties of messenger RNA encoding bacteriophage lambda repressor, Nucleic Acids Res. 20 (1992) [24] I. Moll, G. Hirokawa, M.C. Kiel, A. Kaji, U. Blasi, Translation initiation with 70S ribosomes: an alternative pathway for leaderless mrnas, Nucleic Acids Res. 32 (2004) [25] A. Simonetti, S. Marzi, L. Jenner, A. Myasnikov, P. Romby, G. Yusupova, et al., A structural view of translation initiation in bacteria, Cell. Mol. Life Sci. 66 (2009) [26] S. Mohan, J.P. Donohue, H.F. Noller, Molecular mechanics of 30S subunit head rotation, Proc. Natl. Acad. Sci. U. S. A. 111 (2014) 13,325 13,330. [27] L. Janosi, S. Mottagui-Tabar, L. Isaksson, Y. Sekine, E. Ohtsubo, S. Zhang, et al., Evidence for in vivo ribosome recycling, the fourth step in protein biosynthesis, EMBO J. 17 (1998) [28] A. Seshadri, B. Dubey, M.H. Weber, U. Varshney, Impact of rrna methylations on ribosome recycling and fidelity of initiation in Escherichia coli, Mol. Microbiol. 72 (2009) [29] G. Hirokawa, H. Inokuchi, H. Kaji, K. Igarashi, A. Kaji, In vivo effect of inactivation of ribosome recycling factor fate of ribosomes after unscheduled translation downstream of open reading frame, Mol. Microbiol. 54 (2004) [30] R. Karimi, M. Pavlov, R. Buckingham, M. Ehrenberg, Novel roles for classical factors at the interface between translation termination and initiation, Mol. Cell. 3 (1999) [31] G. Blaha, N. Burkhardt, K.H. Nierhaus, Formation of 70S ribosomes: large activation energy is required for the adaptation of exclusively the small ribosomal subunit, Biophys. Chem. 96 (2002) [32] M. Ryoji, J.W. Karpen, A. Kaji, Further characterization of ribosome releasing factor and evidence that it prevents ribosomes from reading through a termination codon, J. Biol. Chem. 256 (1981) [33] M.Y. Pavlov, D.V. Freistroffer, J. MacDougall, R.H. Buckingham, M. Ehrenberg, Fast recycling of Escherichia coli ribosomes requires both ribosome recycling factor (RRF) and release factor RF3, EMBO J. 16 (1997) [34] Y. Shimizu, A. Inoue, Y. Tomari, T. Suzuki, T. Yokogawa, K. Nishikawa, T. Ueda, Cell-free translation reconstituted with purified components, Nat. Biotechnol. 19 (2001) [35] Y. Inokuchi, A. Hirashima, Y. Sekine, L. Janosi, A. Kaji, Role of ribosome recycling factor (RRF) in translational coupling, EMBO J. 19 (2000) [36] A. Hirashima, A. Kaji, Factor-dependent release of ribosomes from mrna. Requirement for 2 heat-stable factors, J. Mol. Biol. 65 (1972) [37] V. Heurgue-Hamard, R. Karimi, L. Mora, J. MacDougall, C. Leboeuf, G. Grentzmann, et al., Ribosome release factor RF4 and termination factor RF3 are involved in dissociation of peptidyl-trna from the ribosome, EMBO J. 17 (1998) [38] J.R. Menninger, Accumulation of peptidyl trna is lethal in Escherichia coli, J. Bacteriol. 137 (1979) [39] H. Kössel, U.L. RajBhandary, Studies on polynucleotides. LXXXVI. Enzymic hydrolysis of N-acylaminoacyl-transfer RNA, J. Mol. Biol. 35 (1968) [40] H. Kössel, Purification and properties of peptidyl-trna hydrolase from Escherichia coli, Biochim. Biophys. Acta 204 (1970) [41] N.S. Singh, R. Ahmad, R. Sangeetha, U. Varshney, Recycling of ribosomal complexes stalled at the step of elongation in Escherichia coli, J. Mol. Biol. 380 (2008) [42] S. Vivanco-Dominguez, J. Bueno-Martinez, G. Leon-Avila, N. Iwakura, A. Kaji, H. Kaji, G. Guarneros, Protein synthesis factors (RF1, RF2, RF3, RRF, and tmrna) and peptidyltrna hydrolase rescue stalled ribosomes at sense codons, J. Mol. Biol. 417 (2012) [43] N. Singh, G. Das, A. Seshadri, R. Sangeetha, U. Varshney, Evidence for a role of initiation factor 3 in recycling of ribosomal complexes stalled on mrnas in Escherichia coli, Nucleic Acids Res. 33 (2005) [44] Y. Handa, N. Inaho, N. Nameki, YaeJ is a novel ribosomeassociated protein in Escherichia coli that can hydrolyze peptidyl-trna on stalled ribosomes, Nucleic Acids Res. 39 (2011) [45] Y. Shimizu, ArfA recruits RF2 into stalled ribosomes, J. Mol. Biol. 423 (2012) [46] Y. Chadani, K. Ito, K. Kutsukake, T. Abo, ArfA recruits release factor 2 to rescue stalled ribosomes by peptidyl-trna hydrolysis in Escherichia coli, Mol. Microbiol. 86 (2012) [47] D.J. Ramrath, H. Yamamoto, K. Rother, D. Wittek, M. Pech, T. Mielke, et al., The complex of tmrna SmpB and EF-G on translocating ribosomes, Nature 485 (2012) [48] Y. Shimizu, T. Kanamori, T. Ueda, Protein synthesis by pure translation systems, Methods 36 (2005) [49] K.L. Manchester, Resolution with a protein synthesising system of conflicting estimates for the stability constant for KATP3- formation, Biochim. Biophys. Acta 630 (1980) [50] E.M. Froschauer, M. Kolisek, F. Dieterich, M. Schweigel, R.J. Schweyen, Fluorescence measurements of free [Mg2+] by use of mag-fura 2 in Salmonella enterica, FEMS Microbiol. Lett. 237 (2004)