initiation of bacterial protein synthesis with wild …...initiation of bacterial protein synthesis...

Initiation of bacterial protein synthesis with wildtype and mutated variants of initiation factor 2

Michael Y. Pavlov, Suparna Sanyal and Måns Ehrenberg

111. Introduction

In-frame translation of code triplets (codons) of mRNAs into the peptide chains that define the cell’s proteome requires accurate positioning of the mRNA start codon in the ribosomal P site during initiation of protein synthesis. Initiation in bacteria begins on the small (30S) ribosomal subunit. It is generated through dissociation of the bacterial ribosome into its small and large (50S) subunits. Ribosome dissociation is promot-ed by ribosomal recycling factor, RRF, and elongation factor G (EF-G) (Pavlov et al., 1997; Karimi et al., 1999; Peske et al., 2005; Pavlov et al., 2008; Savelsbergh et al., 2009) following termination of protein synthesis by a class-1 release factor (RF1/2) (Freistroffer et al., 2000) and RF1/2 recycling by the G-protein RF3 (Freistroffer et al., 1997; Zavialov et al., 2001).

Initiation starts with the binding of mRNA to the 30S subunit, after which the start codon, AUG, is iden-tified among all possible base triplets of mRNA and placed in the P site of the 30S subunit (Ringquist et al., 1992; Studer and Joseph, 2006). The daunting problem of finding the correct AUG codon for initiation among a manifold of in-frame or out-of-frame AUG triplets in an mRNA sequence is solved with help of the Shine and Dahlgarno (SD) sequence in the mRNA (Shine and Dalgarno, 1974; Londei, 2005). This sequence interacts with the anti-SD sequence in the 16S rRNA of the 30S subunit, which positions AUG codon clos-est to the SD sequence in the neighborhood of the P site of the 30S subunit. Precise P-site positioning is ac-complisheD by the formylated version of Met-charged initiator tRNA, fMet-tRNAi (fMet-tRNAfMET CAU ), through its high affinity to the P site of the 30S subunit and to the AUG codon (Steitz and Jakes, 1975; Londei, 2005). As a result, the P site becomes programmed with the AUG codon and the A site with the second codon in

the open reading frame (ORF) of the protein to be syn-thesized (Hartz et al., 1989). Subsequent docking of the 50S subunit to the mRNA-programmed and fMet-tRNAi-containing 30S subunit leads to formation of the 70S initiation complex (70S IC), which completes initiation of translation.

Addition of fMet-tRNAi and mRNA with proper spacing between SD sequence and initiation codon (Ringquist et al., 1992) to the empty 30S subunit cor-rectly positions initiator tRNA and initiation codon in the P site of the 30S.mRNA.fMet-tRNAi

pre-initiation complex (30S PIC) (Hartz et al., 1989). Subsequent ad-dition of the 50S subunit to the 30S PIC leads to for-mation of an elongation competent 70S IC (Hartz et al., 1989). However, addition of a mixture of elongator tRNAs and fMet-tRNAi to mRNA-programmed 30S subunits results in a heterogeneous set of 30S PICs complexes containing initiator and elongator tRNAs paired to their cognate codons in the P site, generat-ing a multiplicity of mRNA reading frames (Hartz et al., 1989). Therefore, this minimal system for initiation does not provide accurate selection of fMet-tRNAi into the 30S PIC and, hence, fails to correctly position the start codon of mRNA on the ribosome. The required selectivity is provided by the three initiation factors (IF1, IF2 and IF3) (Hartz et al., 1989). These factors use features of initiator fMet-tRNAi, which separate it from all elongator tRNAs (RajBhandary, 1994; Schmitt et al., 1996), to ensure binding of fMet-tRNAi, but not other tRNAs, including the AUG-reading elonga-tor tRNAMet, to 30S subunits containing mRNA, IF1, IF2•GTP, and IF3 (Wintermeyer and Gualerzi, 1983; Pon and Gualerzi, 1984; Canonaco et al., 1986; Antoun et al., 2006 a).

The roles of initiation factors in selective bind-ing of fMet-tRNAi into the 30S PIC have been under extensive study during several decades (reviewed in

M. V. Rodnina et al. (eds.), Ribosomes© Springer-Verlag/Wien 2011

130 Chapter 11 Initiation of bacterial protein synthesis with wild type and mutated variants of initiation factor 2

(Gualerzi and Pon, 1990; Gualerzi et al., 2000; Boe lens and Gualerzi, 2002; Marintchev and Wagner, 2004; Laursen et al., 2005). In contrast, the effects of initia-tion factors on the kinetics of 50S subunit docking to the 30S PIC for 70S IC formation have received little attention in the past (Grunberg-Manago et al., 1975; Chaires et al., 1981). Stopped-flow experiments with detection of scattered light were, however, used in pi-oneering work to demonstrate that complex formation between vacant 30S subunits and 50S subunits follows the law of mass action (Wishnia et al., 1975). More re-cently, we used this technique to study the effects of initiation factors, GTP and fMet-tRNAi on the kinetics of subunit joining (Antoun et al., 2003; Antoun et al., 2004; Antoun et al., 2006 b; Antoun et al., 2006 a).

2. Roles of initiation factors in rapid and accurate formation of the 30S PIC

The roles of initiation factors for rapid and selective binding of fMet-tRNAi to the 30S subunit and sub-sequent rapid docking of the 50S subunit have been systematically studied in (Antoun et al., 2006 b; An-toun et al., 2006 a). One conclusion from this work is that all three initiation factors, and, in particular, IF2 greatly favor the preferential binding of fMet-tRNAi into the 30S PIC, in line with previous results from a model system for initiation in which N-acetyl-Phe-tRNAPhe was used instead of fMet-tRNAi (Gualerzi et al., 1979; Wintermeyer and Gualerzi, 1983; Canonaco et al., 1986; Gualerzi and Wintermeyer, 1986). Inter-estingly, we have found that IF3 effectively destabilizes the binding of all tRNAs, including fMet-tRNAi, to the mRNA-programmed 30S subunit by increasing the rate constant for tRNA dissociation from the 30S sub-unit much more than the association rate constant (An-toun et al., 2006 a). In particular, IF3 alone increases the rate of fMet-tRNAi association to 30S subunits pro-grammed with an unstructured mXR7 mRNA (encod-ing an MFTI tetrapeptide) with a strong SD sequence about 5-fold but increases its dissociation rate about 100-fold, implying a 20-fold reduction in fMet-tRNAi affinity (association equilibrium constant) to the mR-NA-programmed 30S subunit (Antoun et al., 2006 a). Surprisingly, IF3-induced reduction of the elongator Phe-tRNAPhe affinity to an mXR7 mRNA-programmed 30S subunit was somewhat smaller, about 10-fold (An-toun et al., 2006 a). These results apparently contradict earlier findings that IF3 reduces the 30S subunit-af-finity of initiator tRNA much less than the affinity of

elongator tRNAs (Risuleo et al., 1976). We note, how-ever, that selective recognition of initiator tRNA by IF3 seems to depend on three consecutive G-C base pairs in its anticodon stem (Hartz et al., 1989; RajBhandary, 1994; O’Connor et al., 2001; Hoang et al., 2004; Lan-caster and Noller, 2005) and that tRNAPhe from E. coli contains two out of these three G-C base pairs. It is therefore possible that IF3 reduces the affinity of other elongator tRNAs, lacking all three G-C base-pairs in the anticodon stem, much more than the affinity of initiator tRNA and tRNAPhe (Hartz et al., 1989; Hoang et al., 2004; Lancaster and Noller, 2005). It has been suggested that selection of G-C containing tRNAs de-pends on IF3-induced removal of the S9 protein tail from the P site of the 30S PIC, which allows for inter-action of these G-C base-pairs with the G1338/A1339 bases of 16S rRNA (Dallas and Noller, 2001; Lancaster and Noller, 2005). The IF3-dependent de-positioning of the S9 tail may both interrupt the non-specific in-teraction of S9 with tRNA in the P site (Hoang et al., 2004) and broaden the passage to the P site, which could explain the increased dissociation and associa-tion rate constants for tRNA interaction with the 30S subunit in the presence of IF3 (Antoun et al., 2006 a). It has also been suggested that in the absence of IF3 all tRNAs interact with S9 instead of G1338/A1339, which could give them similar and high affinity to the 30S P site, ensuring high processivity and uniform rate of protein elongation (Antoun et al., 2006 a).

IF3-induced acceleration of tRNA association to and dissociation from the 30S subunit is amplified by the addition of IF1 (Antoun et al., 2006 a), in line with previous observations (Pon and Gualerzi, 1984; Gualerzi and Pon, 1990). IF1 binding to the A site al-ters global conformation of the 30S subunit (Carter et al., 2001). It seems therefore that both IF1 and IF3 af-fect the kinetics of tRNA binding to the 30S PIC by changing the conformation of the 30S subunit, instead of making direct contacts with the P site-bound tRNA (Gualerzi and Pon, 1990; Carter et al., 2001; Dallas and Noller, 2001).

Addition of IF2 to the 30S PIC containing IF1, IF3, and mXR7 mRNA increases about 60-fold the rate constant for fMet-tRNAi association and decreases about 3-fold the dissociation rate constant. At the same time, IF2 addition has little effect on these rate constants for non-formylated Met-tRNAi and elon-gator Phe-tRNAPhe (Antoun et al., 2006 a). These re-sults underline the importance of the formyl group of fMet-tRNAi as its major structural element recognized by IF2 (Gualerzi and Pon, 1990; RajBhandary, 1994;

Michael Y. Pavlov, Suparna Sanyal and Måns Ehrenberg 131

Schmitt et al., 1996). Accordingly, addition of all three initiation factors accelerates fMet-tRNAi association to and dissociation from the 30S subunit about 400- and 300-fold, respectively, leaving the affinity of initi-ator tRNA binding virtually unaltered (Antoun et al., 2006 a). In contrast, the three initiation factors increase the association and dissociation rates of elongator Phe-tRNAPhe by about 5- and 300- fold, respectively, reduc-ing by 60-fold the affinity of this tRNA to the 30S PIC. This strong reduction of the 30S binding affinity for tRNAs other than fMet-tRNAi by initiation factors ac-counts for a major part of the preferential selection of fMet-tRNAi into the 30S PIC (Antoun et al., 2006 a).

3. Roles of initiation factors and initiator tRNA in subunit joining

The presence of fMet-tRNAi on the 30S subunit strong-ly affects the rate of subunit joining. In the absence of IF3 and presence of only IF2 or IF2 and IF1, ad-dition of fMet-tRNAi accelerates the effective rate, kc , of subunit docking 60- or 30-fold, respectively. These results suggest that fMet-tRNAi may provide an addi-tional complementary surface to increase the rate of binding of the 50S subunit. Alternatively, fMet-tRNAi may drive IF2 on the 30S subunit into a conformation that is optimal for rapid subunit joining (Antoun et al., 2006 a; Simonetti et al., 2008). The latter scenario is in line with our recent results as discussed below in Sec-tion 8. In the absence of IF3, kc is directly proportional to the 50S subunit concentration, implying a rate con-stant for subunit dissociation close to zero (Antoun et al., 2006 a). Under these conditions the overall rate of dipeptide formation after mixing 30S PICs with 50S subunits and ternary complexes composed of elon-gation factor EF-Tu, GTP, and aminoacyl-tRNA (aa-tRNA) cognate to the second codon of XR7 mRNA is 27 s−1 and significantly smaller than the rate of subunit joining (42 s−1) measured by light scattering (Antoun et al., 2006 b). The delay between subunit joining and peptide bond formation is due to the kinetic steps that separate these two events, which include hydrolysis of GTP on IF2, release of IF2•GDP from the 70S IC and the ternary complex binding to the 70S IC (Antoun et al., 2003; Antoun et al., 2006 b).

In the presence of IF1, IF2, and IF3, the rate and ex-tent of subunit joining are virtually zero in the absence of fMet-tRNAi, but increase to 3 s−1 and near 100 % 70S formation, respectively, by initiator tRNA addition (Antoun et al., 2006 b). The rate of 3 s−1 corresponds

well to the rate of dipeptide formation (3.6 s−1) meas-ured in a parallel quench-flow experiment, indicating that in the presence of IF3, subunit joining is rate limit-ing for di-peptide formation (Antoun et al., 2006 b). In-creasing IF3 concentration reduces the rate of subunit joining, suggesting that dissociation of IF3 precedes 70S complex formation (Antoun et al., 2006 b). Add-ition of IF3 decreases the affinity of fMet-tRNAi to the 30S PIC about 20-fold implying (from the detailed bal-ance constraint) that fMet-tRNAi similarly reduces the affinity of IF3 to the 30S subunit, thereby promoting rapid dissociation of the factor (Antoun et al., 2006 b). Hence, fMet-tRNAi binding to the 30S PIC accelerates subunit joining by two distinct mechan isms. Firstly, fMet-tRNAi drives IF2 into a conformation, which promotes subunit joining, and, secondly, fMet-tRNAi alleviates inhibition of subunit joining by IF3 (Antoun et al., 2006 b).

The kinetic model depicted in Figure 1 and quan-tified by Eq. 1 explains how the average time, 1/kc, for subunit joining depends on the association rate con-stant for 50S subunit docking, k50S, the 50S concentra-tion, [50S], the IF3 concentration, [IF3], and the rate constants qIF3 and kIF3 for IF3 dissociation from and as-sociation to the 30S PIC, respectively:

(1)

This model predicts a linear relation between 1/kc and [IF3]. It also predicts a minimal subunit join-ing time in the limit of large [50S]-values, equal to the average time (1/qIF3) of IF3 release from the 30S PIC, as observed experimentally (Antoun et al., 2006 b).

Removal of IF1 from the 30S PIC increases the sub-unit docking rate about 3-fold, possibly due to reduced affinity of IF3 to the 30S PIC by an increased value of qIF3 (Antoun et al., 2006 b). Such an increase in qIF3 due to the IF1 removal was later observed by another group (Milon et al., 2008). IF1 may also favor a 50S-docking-inactive conformation of the 30S PIC (Milon et al., 2008). These two explanations of the adverse ef-fect of IF1 on subunit docking rate are not mutually exclusive, since the putative, IF1-induced inactive 30S PIC conformation (Milon et al., 2008) may have higher affinity to IF3 than the active form of the subunit.

It has been suggested (Milon et al., 2008) that the kinetics of subunit docking and, in particular, the im-pact of IF1 on this kinetics depends on the strength of the SD-antiSD interaction (Milon et al., 2008). Here, we have compared the kinetics of 50S subunit-dock-


ing to 30S PICs programmed with mRNAs contain-ing strong (XR7) or weak (SD022) SD-sequence and found the impacts of initiation factors and GTP on the rate of subunit docking for the two types of mRNA to be very similar. We found, in particular, that addition of IF1 and/or IF3 to the 30S subunit alters the subunit docking kinetics similarly for the strong and weak SD-mRNAs (Table 1; Figure 2A, B). When, furthermore, the IF3 concentration increases in the presence of IF1 and IF2, the subunit docking rate considerably de-creases for both mRNA types (Table 1). When GTP is

replaced by GDP, the rate of 50S subunit docking to 30S PICs programmed with our weak-SD mRNA, and containing fMet-tRNAi along with all three initiation factors, decreases 100-fold (Figure 2D, Table 1). All these results are perfectly in line with previous obser-vations with the strong-SD XR7 mRNA (Antoun et al., 2004; Antoun et al., 2006 b). The biphasic kinetics of subunit joining with our weak-SD mRNA (Figure 2B) is similar to that previously observed in the presence of IF1, also with a weak SD-mRNA (Milon et al., 2008).

In summary, the effects of initiation factors on the rate of subunit joining observed here (Figure 2; Table 1) and in previous (Antoun et al., 2006 a; An-toun et al., 2006 b) experiments are similar to those observed by Milon et al (2008) for their weak-SD mRNAs. The major difference between the results of the two groups is the 200-fold reduction in subunit joining rate upon IF1 addition they observe for 30S PICs programmed with strong SD-mRNA, contrasting the three-fold reduction observed by us for both types of mRNA (Figure 2; Table 1). A temperature shift from 37 °C, routinely used by us, to 20 °C, used by (Milon et al., 2008), does not qualitatively alter the impact of IF1 and/or IF3 on the kinetics of subunit joining with our strong-SD mRNA (Figure 2C, Table 1), implying that the different results are not due to different tempera-tures. The discrepancies may, we suggest, be caused by different structural features of the mRNAs, unrelated to their SD sequences or, possibly, by different buffer conditions.

Table 1 Eff ective rates, kc, of the 50S subunits docking to 30S PICs assembled with fMet-tRNAi, IF2, and diff erent combina-tions of IF1 and IF3

Conditions IF2 IF2 and IF1

IF2 and IF3

IF2, IF1 and IF3 or IF3(a)

mXR7, 37oC, GTP

37 s−1 40 s−1 22 s−1 8.5 s−1/4.7 s−1 (a)

mXR7, 20oC, GTP

21 s−1 24 s−1 12 s−1 3.6 s−1/ 2.1 s−1 (a)

mSD022, 37oC, GTP

30 s−1 39 s−1 17 s−1 5.1 s−1/3.3 s−1 (a)

mSD022, 37oC, GDP

1.5 s−1 1.5 s−1 0.12 s−1 0.05 s−1

30S subunits were programmed with mXR7 (strong SD) or mSD022 (weak SD) mRNAs. (a)IF3 was added in 0.4 μM (IF3) or in 0.8 μM (IF3(a)) concentrations. The effective rate kc is defined here as the in-verse of the time at which 50 % of the 70S initiation complexes have been formed after the subunit mixing (Zorzet et al., 2010).

Fig. 1 Kinetic scheme of initiation of protein synthesis starting from fMet-tRNAi, 50S subunits, and tRNA-free 30S PICs contain-ing three initiation factors and mRNA. fMet-tRNAi binds to 30S PIC with the second order rate constant ktRNA and dissociates with rate constant qtRNA. Binding of fMet-tRNAi to 30S PIC destabilized

IF3 binding increasing qIF3. The presence of IF1 stabilizes IF3 bind-ing decreasing qIF3. 50S subunit associates with 30S PIC containing tRNA with the second order rate constant k50S. In the absence of IF3, kc=k50S∙[50S].


Fig. 2 Kinetics of 70S initiation complex formation after rapid mixing of 50S subunits with 30S PICs containing fMet-tRNAi, IF2, and different combinations of IF1 and IF3. (A) 0.37 μM 50S subu-nits were mixed in a stopped-flow apparatus with a 30S PIC mixture containing 0.31 μM 30S subunits, 0.6 μM fMet-tRNA, 0.7 μM XR7 (strong SD) mRNA supplemented with 1 μM IF2 (◆); with 0.5 μM IF2 and 0.5 μM IF1 ( ); with 1 μM IF2 and 0.4 μM IF3 (▲); with 0.5 μM IF2, 0.5 μM IF1 and 0.4 μM IF3 (▲▲) or with 0.5 μM IF2, 0.5 μM IF1 and 0.8 μM IF3 (■) as indicated in the figure. After the mixing, light scattering at 430 nm was monitored to follow subunit asso-ciation. The experiment was conducted at 37 °C in a polymix-like

buffer, LS4, contained 95 mM KCl, 5 mM NH4Cl, 0.5 mM CaCl2, 8 mM putrescine, 1 mM spermidine, 30 mM HEPES pH 7.5, 1 mM DTE, 1 mM GTP, 1 mM ATP, 6 mM Mg(OAc)2, and 2 mM PEP, supplemented with 1 μg/ml pyruvate kinase (PK) and 0.1 μg/ml myokinase (MK). Since each ATP or GTP molecule chelate one Mg2+ cation, the free Mg2+ concentration in the L4 buffer is about 4 mM (Pavlov et al., 2008). (B) The same as (A) but 30S PICs con-tained 0.8 μM mSD022 (weak SD) mRNA. (C) The same as (A) but the experiment was conducted at 20 °C. (D) The same as (B) but the LS4 buffer contained 1 mM GDP instead of GTP, and PK /MK enzymes were excluded. All concentrations are after mixing.

A B

C D


4. Roles of GTP and GTP hydrolysis in initiation of mRNA translation

An important question in initiation of protein synthe-sis concerns the role played by GTP in the function of IF2 that belongs to the same family of translational GTPases as EF-Tu, EF-G, RF3, and SelB (Bourne et al., 1991; Hauryliuk et al., 2008; Paleskava et al., 2010). It was early concluded that IF2•GTP is essential for ini-tiation of mRNA translation and that GTP hydrolysis is required for release of IF2 from the 70S IC to al-low for binding of ternary complex and formation of the first peptide bond (Benne et al., 1973; Fakunding and Hershey, 1973; Grunberg-Manago et al., 1975). The importance of GTP hydrolysis for fast release of IF2 from the ribosome has later been questioned by Pon et al (1985), who from binding experiments pro-posed that the GDP- and GTP-bound forms of IF2 have similar and high affinities to the 70S ribosome (Pon et al., 1985). However, their experimental setup did not exclude hydrolysis of GTP on ribosome bound IF2 (Luchin et al., 1999; Grigoriadou et al., 2007 a), implying that the reported affinity of IF2∙GTP reflects the association rate of IF2∙GTP to the 70S ribosome, while the dissociation rate is more complex, involving GTP hydrolysis and dissociation of either IF2∙GTP or IF2∙GDP. The authors (Pon et al., 1985) do not discuss the possibility that, due to rapid GTP hydrolysis (Lu-chin et al., 1999; Grigoriadou et al., 2007 a), their experi-ment might have been insensitive to any difference in dissociation rate constants for IF2∙GTP and IF2∙GDP. It is, in other words, possible, that their data report the IF2∙GDP affinity to the 70S ribosome in both the GTP and the GDP case. Moreover, the high IF2∙GDP affin-ity to the 70S ribosome (Pon et al., 1985) could be due to their particular experimental conditions. Indeed, the overall rate of 27 s−1, measured for dipeptide for-mation upon mixing 30S PICs with 50S subunits and ternary complexes (Antoun et al., 2006 b), provides a lower limit for the rate of IF2∙GDP dissociation from the 70S IC. This suggests that under physiological con-ditions (37 °C and low free Mg2+ concentration) the af-finity of IF2∙GDP to the 70S ribosome is low.

The pivotal importance of GTP and its hydrolysis for rapid formation of the elongation competent 70S IC from a 30S PIC and 50S subunit has been demon-strated by stopped-flow experiments with monitor-ing light scattering (Antoun et al., 2003). The subunit docking rate decreases 20-fold when GDP replaces GTP but only two-fold when the non-hydrolyzable GTP analogue GDPNP replaces GTP on the naturally

truncated -form of IF2 in the 30S PIC. In subsequent work ( Antoun et al., 2004), we used the full-length -form of IF2 to demonstrate a 60-fold decrease in the rate of docking of the 50S subunit to the 30S PIC when GTP was swapped for GDP and, as before, a two-fold rate decrease when GTP was swapped for GDPNP. This shows that GTP, but not GTP hydrolysis, is required for rapid subunit docking. Moreover, the 70S IC formed either with IF2 and GDPNP or with a small amount of IF2∙GTP following incubation with IF2 and GDPNP is virtually inactive in dipeptide formation. From these results we concluded that the GTP (GDPNP) form of IF2 dissociates very slowly from the 70S IC, thereby blocking the entry of ternary complex to the A site of the ribosome (Antoun et al., 2003), in line with earlier conclusions (Dubnoff et al., 1972; Benne et al., 1973).

In summary, our results demonstrate that rapid 70S IC formation requires GTP, but not GTP hydroly-sis, and that GTP hydrolysis leads to rapid release of IF2•GDP from the 70S IC swiftly followed by dipep-tide formation.

5. Alternative models of subunit joining and 70S IC formation

A recent study with E. coli components, complement-ed with IF2 from B. stearothermophilus, has confirmed that both fMet-tRNAi and IF2 are required on the 30S subunit for its rapid docking to the 50S subunit and that subunit joining rates with GTP and its non-hydrolysable analogue (GDPCP) are similar (Grigo-riadou et al., 2007 a; Grigoriadou et al., 2007 b). The dipeptide formation rate of 0.2 s−1, measured uponmixing 30S PICs with 50S subunits and ternary com-plexes, was much smaller than the rate of subunit asso-ciation due to the small rate of dissociation of thermo-philic IF2•GDP from E. coli 70S IC (Grigoriadou et al., 2007 a). Such slow dissociation of thermophilic IF2·GDP may also explain the earlier results by Tom-sic et al. (2000), who found very similar and small (0.2 s−1) rates of peptide bond formation after mixing 50S subunits with 30S PICs containing either IF2•GDP or IF2•GTP. They concluded that neither the presence of GTP nor GTP hydrolysis is important for formation of the first peptide bond in a nascent protein (Tomsic et al., 2000). However, slow dissociation of B. stearo-thermophilus IF2 from the E. coli 70S IC could have been rate-limiting for peptide bond formation, thereby masking any difference in subunit joining rate between the GTP and GDP cases.


Recent experiments have led to the proposal of a novel kinetic model of subunit joining (Grigoriadou et al., 2007 a). According to this model, docking of the 50S subunit to the 30S PIC leads to an unstable 70S IC, which initially can undergo rapid dissociation into subunits but eventually becomes stable along a com-plex kinetic pathway. No effect of IF3 on the rate of for-mation of 70S IC was observed in these experiments (Grigoriadou et al., 2007 b). In contrast, more recent (Milon et al., 2008) and our previous (Antoun et al., 2006 b) observations show that, in the absence of IF3, ribosomal subunits do not dissociate after the initial joining, and that addition of IF3 to the 30S PIC greatly reduces the rate of subunit docking. These contrasting results put in question details of the kinetic model sug-gested by (Grigoriadou et al., 2007 a; Grigoriadou et al., 2007 b), as further discussed below in section 5.1.

It was concluded from stopped-flow data based on fluorescence detection that IF3 dissociates from the 70S IC after subunit joining (Milon et al., 2008) and not before, as previously suggested (Antoun et al., 2006 b). IF3 consists of globular N- and C-domains connected by a long, helical linker (Dallas and Noller, 2001; Petrelli et al., 2001). When the isolated C-do-main of IF3 is bound to the 30S subunit, its docking with the 50S subunit is inhibited (Petrelli et al., 2001). From these observations, we speculate that spontane-ous dissociation of the C-domain of IF3 from the 30S PIC may allow for 70S IC formation, even though its N-domain remains transiently bound on the 30S sub-unit. Such transient IF3 binding after C-domain dis-sociation and 70S complex formation may explain why IF3 dissociation appears slower than subunit joining (Milon et al., 2008). The strong dependence of the rate of subunit joining on IF3-concentration (Antoun et al., 2006 b) (Figure 2; Table 1) could then be explained on the further assumption that the C-domain dissociation permits binding of a second molecule of IF3 to the vac-ated C-domain binding site on the 30S subunit.

5.1. Roles of GTP hydrolysis and Pi release in subunit joining

It is an established fact that the GTP molecule on IF2 is hydrolyzed during initiation of translation. However, the timing of GTP hydrolysis on IF2 and subsequent Pi release from IF2 in relation to subunit joining has become controversial. That is, it was recently claimed that stable subunit association occurs after GTP hy-drolysis and that phosphate release follows GTP hy-

drolysis after a long (> 0.1 s) lag period (Grigoriadou et al., 2007 a; Grigoriadou et al., 2007 b). Our experi-ments at high 50S subunit concentration in Figure 3 show, in contrast, that subunit docking is almost com-plete at the onset of GTP hydrolysis and that Pi release very rapidly follows GTP hydrolysis. Indeed, the time corresponding to the rapid phase of subunit joining is about 6 ms, while the times for GTP hydrolysis and Pi release after subunit mixing are 25 ms and 32 ms, respectively (Figure 3).

That stable subunit joining does not require GTP hydrolysis has been shown more directly in very recent experiments (Huang et al., in preparation), where the rate of subunit joining was monitored with the GTPase impaired IF2-mutants V400G and H448E (Luchin et al., 1999). These mutants have much slower rates of GTP hydrolysis and Pi release than wild type IF2, while the rates of subunit docking promoted by the mutated and wild type variants of IF2 are virtually the same. The rates of subunit joining are, furthermore, strictly proportional to the 50S subunit concentration for wild

Fig. 3 Comparison of kinetics of subunit association (red trace), GTP hydrolysis ( , black trace) and Pi release (blue trace) at large excess of 50S subunits over 30S PICs. Mix A containing 0.5 μM 30S subunit, 1 μM MLI mRNA, 1 μM fMet-tRNAfMet, 1 μM IF1, 1 μM IF2 and 1μM GTP, pre-incubated in 37 ºC for 5 minutes was mixed rapidly with pre-incubated mix B containing 2.5 μM 50S subunit in a stopped-flow apparatus and light scattering at 430 nm was moni-tored to follow subunit association at 37 ºC. MLI mRNA coding for Met-Leu-Ile tripeptide had the sequence identical with XR7 mRNA outside the coding region (Huang et al., 2010). In a parallel setup, 10 μM of MDCC labeled phosphate binding protein (PBP-MDCC) was added in the mix B (as described in Huang et al. 2010) and Pi release was measured by the increase of fluorescence at 464 nm (lEx = 425 nm). For GTP hydrolysis, 20 μM of [3H]GTP was used in mix A in otherwise identical reaction condition in quench flow. The amount of GDP produced in every time point was estimated by separating GTP and GDP in TLC or in a Mono-Q column attached to FPLC.


Mohr et al., 2002; Diaconu et al., 2005). The ribosomal stalk in E. coli is composed of four copies of flexible L12 proteins (Chandra Sanyal and Liljas, 2000; Mulder et al., 2004), which interact with translation factors us-ing their C-terminal domains (Diaconu et al., 2005; Helgstrand et al., 2007). Here, we have studied the role of the stalk proteins in IF2 recognition during initia-tion of mRNA translation (Huang et al., 2010).

We have compared the docking rates of 30S PICs or 30S subunits to native and L12-deficient 50S sub-units. The rates of 50S subunit docking to vacant 30S subunits or to 30S PICs lacking IF2 were unaffected by L12 depletion (Figure 4A). However, in the presence of IF2 in the 30S PIC the docking rate was 40 times smaller for L12 depleted than for native 50S subunits (Figure 4B), demonstrating that L12 speeds up sub-unit joining in an IF2-dependent manner. We have also monitored the extents of GTP hydrolysis and Pi release of IF2-containing 30S PICs as they dock with native or L12 deficient 50S subunits. We found that the time spans between a subunit-docking event and GTP hydrolysis or Pi release are very similar for native and L12 depleted 50S subunits (Huang et al., 2010). These data suggest that L12 is essential for rapid docking of the 50S subunit to an IF2•GTP-containing 30S PIC, but not for subsequent GTP hydrolysis and Pi release.

type and mutants of IF2, implying negligible dissoci-ation of the 70S IC into subunits before GTP hydro-lysis (Huang et al., in preparation).

The reasons why our previous (Antoun et al., 2003; Antoun et al., 2006 b) and present (Huang et al., in preparation; Figure 3) observations differ qualitatively from those by (Grigoriadou et al., 2007 a; Grigoriadou et al., 2007 b) are unclear. The experiments were car-ried out at different temperatures (37° vs. 20 °C) and had different buffer conditions. Furthermore, homolo-gous IF2 variants from E. coli were used in one case (Antoun et al., 2003; Antoun et al., 2006 b) (Huang et al., in preparation), while heterelogous IF2 variants from B. stearothermophilus were used in the other case (Grigoriadou et al., 2007 a; Grigoriadou et al., 2007 b), but also other explanations for the discrepancies are conceivable.

5.2. IF2 interaction with the L12 protein of the 50S subunit promotes fast subunit joining

It has been suggested that ribosomal stalk proteins (L7/L12) are involved in the recruitment of translational GTPases, EF-Tu and EF-G to the ribosome and in ac-celeration of GTP hydrolysis (Savelsbergh et al., 2000;

Fig. 4 Role of the ribosomal stalk proteins (L12) in subunit asso-ciation. (A) The kinetics of 70S ICs formation after rapid mixing of 0.5 μM 50S subunits (native 50S – red trace, L12-depleted 50S – green trace and L12 reconstituted 50S subunits – blue trace) with a 30S PIC mixture containing 0.5 μM 30S, 2 μM IF1, 2 μM MLI mRNA, 2 μM fMet-tRNAfMet, 100 μM GTP. MLI mRNA is described

in the caption of Fig. 3 After the mixing in a stopped flow appar-atus, light scattering at 430 nm was monitored to follow subunit association at 37 ºC. (B) the same as (A), but the 30S PIC mixture contained 2 μM IF2. Individual mixtures were pre-incubated at 37 ºC for 10 minutes before mixing

A B


Thus, despite its involvement in stimulation of GTP hydrolysis on the elongation factors EF-G and EF-Tu (Savelsbergh et al., 2000; Mohr et al., 2002; Diaconu et al., 2005), L12 is not directly involved in stimulation of GTP hydrolysis on IF2 (Huang et al., 2010).

6. Accuracy of initiator tRNA selection in initiation of mRNA translation

In addition to a large rate of initiation of mRNA trans-lation, the accuracy of fMet-tRNAi selection in the initiation process is crucial for rapid growth and high fitness of bacterial populations (Zorzet et al., 2010). Light scattering data on 50S subunit docking to 30S PICs, containing fMet-tRNAi, Met-tRNAi, tRNAi or Phe-tRNAPhe, in combination with kinetics and equi-librium data on the binding of these tRNAs into the 30S PIC have been used to estimate the accuracy of initiator tRNA selection into the 70S IC (Antoun et al., 2006 a). This means, in particular, estimation of kcat /Km parameters for the incorporation of any one of these tRNAs in the 70S IC starting from 50S subunits, free tRNAs and tRNA-free 30S PICs (Figure 1). The kcat /Km parameter for the kinetic scheme in Figure 1 is in each case the association rate constant, ktRNA, for binding of a tRNA to the 30S PIC, multiplied by the probability, PtRNA, that the tRNA ends up in the 70S IC, rather than dissociates from the 30S PIC:

(2)

Here, qtRNA is the rate constant for dissociation of tRNA from the 30S PIC and kc is a compounded rate constant for the docking of 50S subunits to tRNA-con-taining 30S PICs (see Figure 1). Equation (1) defines kc in terms of elemental rate constants in the kinetic scheme of Figure 1. The accuracy, A, of initiation is the ratio of the kcat /KM parameters for cognate and non-cognate tRNAs:

(3)

The value of A gives the ratio between the number of 70S complexes formed with cognate (fMet-tRNAi) and non cognate (e. g., Phe-tRNAPhe) tRNAs at equal

free concentrations. A “non-cognate tRNA” here refers to a tRNA other than fMet-tRNAi, either bound to the AUG initiation codon (e. g. Met-tRNAi or tRNAi) or to another codon (e. g., Phe-tRNAPhe bound to its cognate UUC codon adjacent to AUG). It follows from Equa-tion (3) that in the limit of rapid subunit docking (i. e., kc >> qtRNA), A=kc tRNA/knc tRNA is only determined by the association rate constants for fMet-tRNAi and its non-cognate competitor, implying low accuracy. In the lim-it of slow subunit docking (i. e. kc<<qtRNA), A = (kc tRNA/knc tRNA)×(qnc tRNA/qc tRNA)×(kc c/knc c,) implying high accuracy (Antoun et al., 2006 a). The later relation for A explains why IF2 plays such a pivotal role in the accuracy of initiation: firstly, IF2 selectively increases the affinity of formylated initiator tRNA to the 30S PIC in relation to non-formylated initiator and elongator tRNAs and, secondly, IF2 selectively increases the rate of 50S dock-ing to 30S PICs containing formylated initiator tRNA (Antoun et al., 2006 a).

Equation (3) predicts the accuracy of initiator tRNA selection to increase when the dissociation rate constants for initiator and non-initiator tRNAs in-crease by the very same factor (Antoun et al., 2006 a). IF1 and IF3 greatly increase the accuracy of initiation by such a purely kinetic mechanism, i. e. by uniformly increasing the rate constant (qtRNA) for tRNA dissoci-ation from the 30S subunit. Moreover, IF1 and IF3 also greatly reduce the rate constant (kc) of subunit docking for all tRNAs, which according to Equation (3), would further increase accuracy of initiation (Antoun et al., 2006 a).

In summary, our results show that the presence of all three initiation factors in the 30S PIC is required to achieve the highest accuracy of initiation (Antoun et al., 2006 a). The accuracy enhancement by IF1/IF3, achieved by the purely kinetic mechanism described above, may be further increased by IF3-induced pref-erential dissociation of elongator tRNAs lacking G-C base-pairs in their anticodon stem from the 30S PIC (Risuleo et al., 1976; Gualerzi et al., 1979; Hoang et al., 2004). In addition to their roles in enhancing the ac-curacy of initiation, IF1/IF3 also increase the rate of tRNA binding to the 30S PIC, thereby considerably increasing the rate and extent of formation of fMet-tRNAi-containing 70S ICs from 50S subunits, fMet-tRNAi and tRNA-free 30S PICs (Antoun et al., 2006 b; Antoun et al., 2006 a).


7. Initiation with non-formylated Met-tRNAi

Formylation deficiency in bacteria is normally caused by loss-of-function mutations in the formyl-methio-nyl-transferase (FMT) gene or other genes involved in single-carbon metabolism and leads to slow growth phenotype (Guillon et al., 1992; Steiner-Mosonyi et al., 2004). In line with this, Met-tRNAi binds much more slowly to the 30S PIC than fMet-tRNAi and subse-quent docking to the 50S subunit is slower for the Met-tRNAi-containing than for the fMet-tRNAi-containing 30S PICs (Antoun et al., 2006 a). Formylation deficient bacteria often acquire secondary mutations that in-crease their growth rate (Apfel et al., 2001; Nilsson et al., 2006). A previously unknown class of such muta-tions was recently identified (Zorzet et al., 2010). Here, point mutations distal to the fMet-tRNAi binding do-main IV, and thus unrelated to tRNA binding, display strong (A-type, mutations located in domain III of IF2) or weak (B-type, mutations located in domains III, G or in the N-terminus of IF2) growth complementation. The A-type mutations in IF2 increase the effective rate, (kcat /Km), of Met-tRNAi binding to the 30S PIC by 3-fold. They also increase the maximal rate (kcat) of docking of Met-tRNAi-containing 30S PIC to the 50S subunit by four-fold. Such an increased in kcat /Km and kcat parameters by mutations (Table 2) accounts for the growth complementary phenotype of the A- and B-type IF2 mutants in a formylation-deficient back-ground (Zorzet et al., 2010). In contrast, kcat /Km for fMet-tRNAi incorporation into 70S complex was not affected by mutations in IF2 (Table 2). Accordingly, it follows from kcat /Km values in Table 2 and Eq. 3 that the accuracy of initiation with wild type IF2 (A=36) is three-fold higher than with A-type IF2 (A=12). Consequently, these mutations increase the speed of initi ation with Met-tRNAi at the cost of reduced ac-curacy in fMet-tRNAi over Met-tRNAi selection. Fur-thermore, the overall rate of aberrant initiation with an elongator tRNA increased even more, five-fold, by

the A-type mutations (Zorzet et al., 2010). Such an ex-cessive form ation of aberrant 70S ICs with elongator tRNAs may explain why A-type IF2 mutations have a considerable fitness cost in the formyl ation proficient background (Zorzet et al., 2010).

We have found that the rate, kc, of subunit docking with wild type IF2 was 9 s−1 with fMet-tRNAi, 1.3 s−1 with Met-tRNAi and 0.18 s−1 with tRNAi, whereas with A-type IF2 the corresponding docking rates were 10 s−1, 6.6 s−1 and to 2.2 s−1, respectively. These results im-ply that A-type IF2 is much less sensitive to removal of the formyl group and methionine from initiator tRNA than wild-type IF2 (Zorzet et al., 2010). The docking rate reduction upon formyl group removal was here about seven-fold in contrast to a 50-fold reduction in our previous work (Antoun et al., 2006 a). This dis-crepancy is, we suggest, caused by deacylation of Met-tRNAi to tRNAi in our previous study (Antoun et al., 2006 a). The much larger kc value for subunit docking with Met-tRNAi estimated here implies that the accu-racy of fMet-tRNAi selection over Met-tRNAi should be close to the ratio kc tRNA/knc tRNA of the association rates of fMet-tRNAi and Met-tRNAi to the 30S PIC. This es-timates A as ≈ 46 using the ktRNA parameters measured in (Antoun et al., 2006 a), a value reasonably close to A ≈ 36 calculated here from directly measured kcat /Km parameters (Table 2).

8. IF2 mutants active in subunit joining in the absence of tRNA or GTP

The subunit joining properties of A-type IF2 were also studied in the absence of IF3 (Pavlov et al., 2011). In a particular experiment the subunit dock-ing rate was about 40 s−1 for all IF2 variants in the pres-ence of fMet-tRNAi and IF1. Removal of fMet-tRNAi reduced the docking rate 20-fold (from 37 to 1.9 s−1) for wild-type IF2, but only by 40 % (from 41 to 25 s−1) for A-type IF2 (Table 3). This result shows that rapid docking is restored after fMet-tRNAi removal by the A-

Table 2 Kinetic parameters of 70S IC formation from 50S subunits, tRNA-free 30S PICs and Met-tRNAi (or fMet-tRNAi) with diff erent IF2 s in the 30S PIC

kcat, Met-tRNAi (kcat/KM)Met-tRNAi (kcat/KM)fMet-tRNAi Accuracy

WT IF2 1 s−1 0.55 μM−1s−1 20 μM−1s−1 36

B-type IF2 2 s−1 0.9 μM−1s−1 20 μM−1s−1 22

A-type IF2 4 s−1 1.65 μM−1s−1 20 μM−1s−1 12


type mutations, implying that fMet-tRNAi per se is not required for fast subunit docking. Its role is, we sug-gest, to induce an IF2 conformation which promotes rapid subunit docking (Simonetti et al., 2008). From this follows that the phenotype of the A-type and B-type mutations is explained by their ability to shift the equilibrium between an inactive and an active form of IF2 towards the active conformation of the factor. In line with this interpretation, we have found that when GTP is swapped for GDP in the presence of fMet-tRNAi the subunit joining rate is reduced 60-fold (from 37 to 0.6 s−1) with wild type IF2 but less than two-fold (from 41 to 26 s−1) with A-type IF2 in the 30S PIC (Table 3). This implies that in the presence of fMet-tRNAi a large fraction of A-type, but not wild- type, IF2 is driven into the active docking conformation in the presence of GDP. In the absence of guanine nucleotide, the rate of subunit docking was slow even with the A-type IF2 mutant (Table 3), implying that addition of GDP to the apo-form of IF2 shifts the conformational equilibrium towards the active form, albeit to a lesser extent than GTP. In summary, these results mean that activation of wild type IF2 requires the presence of both GTP and fMet-tRNAi, while activation of A-type IF2 requires the presence of only one of these activating ligands. In fact, activation of IF2 and its mutants provide a strik-ing example of conditional activation of a GTP-bind-ing protein by GTP and other ligands (Hauryliuk et al., 2008).

9. Conclusion

It is now generally accepted that IF2•GTP promotes selective binding of fMet-tRNAi into the 30S PIC and that the rapid docking of the 50S subunit requires the presence of both fMet-tRNAi and IF2•GTP in the 30S PIC. Accordingly, IF2•GTP favors fMet-tRNAi se-lection into the 70S IC at two steps of the initiation process, thereby ensuring high accuracy of translation initiation. Importantly, our results show that IF2 mu-

tants that increase the speed of initiation with non-formylated tRNA are over-activated in the sense that they can promote fast subunit joining in the presence of only one activating ligand: either GTP or fMet-tRNAi, whereas wild-type IF2 requires the presence of both these ligands for its full activation. Moreover, these IF2-activating mutations decrease the accuracy of authentic initiation with formylated fMet-tRNAi and reduce the fitness of otherwise wild-type bacte-ria, explaining why evolution has optimized wild-type IF2 for full activation by the simultaneous presence of GTP and fMet-tRNAi. A deeper understanding of IF2-function in initiation may be eventually provided by structurally based explanations of how GTP and fMet-tRNAi activate IF2 on the 30S subunit and how IF2 mutants complementing formylation deficiency can by-pass the two-ligand requirement for activation.

Acknowledgements

The experiments presented in Figures 3 and 4 were conducted by Chenhui Huang and Chandra S. Man-dava. We also thank Chenhui Huang for preparation of Figures 3 and 4. The research grants from the Swed-ish Research Council (project support to M. E. and S. S. and Linne support (URRC) to M. E.) and NIH/USA to M. E. and S. S., and from Göran Gustafsson Stiftelse, Carl Tryggers Stiftelse and Wenner-Gren Stif-telse to S. S. are thankfully acknowledged.

Table 3 Eff ects of A-type mutations in IF2 on the eff ective rate, kc, of the 50S subunit docking to 30S PICs assembled with fMet-tRNAi and either GTP, or GDP, or no G-nucleotide, or with only GTP without tRNA. Th e eff ective rate kc is defi ned here as the inverse of the time at which 50 % of the 70S ICs have been formed aft er the subunit mixing (Zorzet et al., 2010)

fMet-tRNAi

GTPfMet-tRNAi

GDPfMet-tRNAi

no G-nucleotideno tRNAGTP

WT IF2 37 s−1 0.6 s−1 0.5 s−1 1.9 s−1

A-type IF2 41 s−1 26 s−1 1.8 s−1 25 s−1


References

Antoun A, Pavlov MY, Andersson K, Tenson T, Ehrenberg M (2003) The roles of initiation factor 2 and guanosine triphosphate in initiation of protein synthesis. Embo J 22 : 5593 – 5601

Antoun A, Pavlov MY, Lovmar M, Ehrenberg M (2006 a) How ini-tiation factors maximize the accuracy of tRNA selection in initiation of bacterial protein synthesis. Mol Cell 23 : 183–193

Antoun A, Pavlov MY, Lovmar M, Ehrenberg M (2006 b) How ini-tiation factors tune the rate of initiation of protein syn thesis in bacteria. Embo J 25 : 2539 – 2550

Antoun A, Pavlov MY, Tenson T, Ehrenberg MM (2004) Ribo-some formation from subunits studied by stopped-flow and Rayleigh light scattering. Biol Proced Online 6 : 35 – 54

Apfel CM, Locher H, Evers S, Takacs B, Hubschwerlen C, Pirson W, Page MG, Keck W (2001) Peptide deformylase as an antibac-terial drug target: target validation and resistance develop-ment. Antimicrob Agents Chemother 45 : 1058–1064

Benne R, Naaktgeboren N, Gubbens J, Voorma HO (1973) Re-cycling of initiation factors IF-1, IF-2 and IF-3. Eur J Bio-chem 32 : 372 – 380

Boelens R, Gualerzi CO (2002) Structure and function of bacterial initiation factors. Curr Protein Pept Sci 3 : 107–119

Bourne HR, Sanders DA, McCormick F (1991) The GTPase superfamily: conserved structure and molecular mechanism. Nature 349 : 117–127

Canonaco MA, Calogero RA, Gualerzi CO (1986) Mechanism of translational initiation in prokaryotes. Evidence for a dir ect effect of IF2 on the activity of the 30 S ribosomal sub unit. FEBSL ett 207 : 198 – 204

Carter AP, Clemons WM, Jr., Brodersen DE, Morgan-Warren RJ, Hartsch T, Wimberly BT, Ramakrishnan V (2001) Crystal structure of an initiation factor bound to the 30S ribo somal subunit. Science 291 : 498 – 501

Chaires JB, Pande C, Wishnia A (1981) The effect of initiation fac-tor IF-3 on Escherichia coli ribosomal subunit association kinetics. J Biol Chem 256 : 6600 – 6607

Chandra Sanyal S and Liljas A (2000) The end of the beginning: structural studies of ribosomal proteins. Curr Opin Struct Biol 10 : 633 – 636

Dallas A, Noller HF (2001) Interaction of translation initiation factor 3 with the 30S ribosomal subunit. Mol Cell 8 : 855 – 864

Diaconu M, Kothe U, Schlunzen F, Fischer N, Harms JM, Tonevit-sky AG, Stark H, Rodnina MV, Wahl MC (2005) Structural basis for the function of the ribosomal L7/12 stalk in factor binding and GTPase activation. Cell 121 : 991–1004

Dubnoff JS, Lockwood AH, Maitra U (1972) Studies on the role of guanosine triphosphate in polypeptide chain initiation in Escherichia coli. JB iol Chem 247 : 2884 – 2894

Fakunding JL, Hershey JW (1973) The interaction of radioactive initiation factor IF-2 with ribosomes during initiation of protein synthesis. JB iol Chem 248 : 4206 – 4212

Freistroffer DV, Kwiatkowski M, Buckingham RH, Ehrenberg M (2000) The accuracy of codon recognition by polypeptide release factors. Proc Natl Acad Sci USA 97 : 2046 – 2051

Freistroffer DV, Pavlov MY, MacDougall J, Buckingham RH, Ehren-berg M (1997) Release factor RF3 in E. coli accelerates the dissociation of release factors RF1 and RF2 from the ribo-some in a GTP-dependent manner. Embo J 16 : 4126 – 4133

Grigoriadou C, Marzi S, Kirillov S, Gualerzi CO, Cooperman BS (2007 a) A quantitative kinetic scheme for 70 s translation initiation complex formation. J Mol Biol 373 : 562 – 572

Grigoriadou C, Marzi S, Pan D, Gualerzi CO, Cooperman BS (2007 b) The Translational Fidelity Function of IF3 During

Transition from the 30 SI nitiation Complex to the 70 S Ini-tiation Complex. J Mol Biol 373 : 551– 561

Grunberg-Manago M, Dessen P, Pantaloni D, Godefroy-Colburn T, Wolfe AD, Dondon J (1975) Light-scattering stud-ies showing the effect of initiation factors on the revers-ible dissociation of Escherichia coli ribosomes. J Mol Biol 94 : 461– 478

Gualerzi C, Risuleo G, Pon C (1979) Mechanism of the spontaneous and initiation factor 3-induced dissociation of 30 S. aminoa-cyl-tRNA. polynucleotide ternary complexes. J Biol Chem 254 : 44 – 49

Gualerzi C, Wintermeyer W (1986) Prokaryotic initiation factor 2 acts at the level of the 30 S ribosomal subunit. A fluorescence stopped-flow study. FEBS Lett 202 : 1– 6

Gualerzi CO, Brandi L, Caserta E, LaTeana A, Spurio R, Tomsic J, Pon C (2000) Translation initiation in bacteria. In Garett RA, Liljas A, Matheson AT, Moore PB, Noller HF (eds.), The Ribosome: Structure, Function, Anti biotics and Cellu-lar Interactions. ASM Press, Washington DC, pp 477– 494

Gualerzi CO, Pon CL (1990) Initiation of mRNA translation in prokaryotes. Biochemistry 29 : 5881– 5889

Guillon JM, Mechulam Y, Schmitter JM, Blanquet S, Fayat G (1992) Disruption of the gene for Met-tRNA(fMet) formyltrans-ferase severely impairs growth of Escherichia coli. J Bacteriol 174 : 4294 – 4301

Hartz D, McPheeters DS, Gold L (1989) Selection of the ini tiator tRNA by Escherichia coli initiation factors. Genes Dev 3 : 1899–1912

Hauryliuk V, Hansson S, Ehrenberg M (2008) Co-factor dependent conformational switching of GTPases. Biophys J 95 : 1704–1715

Helgstrand M, Mandava CS, Mulder FA, Liljas A, Sanyal S, Akke M (2007) The ribosomal stalk binds to translation factors IF2, EF-Tu, EF-G, RF3 via a conserved region of the L12 C-terminal domain. J Mol Biol 365 : 468 – 479

Hoang L, Fredrick K, Noller HF (2004) Creating ribo somes with an all-RNA 30S subunit P site. Proc Natl Acad Sci USA 101 : 12 439–12 443

Huang C, Mandava CS, Sanyal S (2010) The ribosomal stalk plays a key role in IF2-mediated association of the ribosomal sub-units. J Mol Biol 399 : 145–153

Karimi R, Pavlov MY, Buckingham RH, Ehrenberg M (1999) Novel roles for classical factors at the interface between translation termination and initiation. Mol Cell 3 : 601– 609

Lancaster L, Noller HF (2005) Involvement of 16S rRNA nucleo-tides G1338 and A1339 in discrimination of initiator tRNA. Mol Cell 20 : 623 – 632

Laursen BS, Sorensen HP, Mortensen KK, Sperling-Petersen HU (2005) Initiation of protein synthesis in bacteria. Microbiol Mol Biol Rev 69 : 101–123

Londei P (2005) Evolution of translational initiation: new insights from the archaea. FEMS Microbiol Rev 29 : 185 – 200

Luchin S, Putzer H, Hershey JW, Cenatiempo Y, Grunberg-Manago M, Laalami S (1999) In vitro study of two dominant inhibi-tory GTPase mutants of Escherichia coli translation initiation factor IF2. Direct evidence that GTP hydrolysis is necessary for factor recycling. J Biol Chem 274 : 6074 – 6079

Marintchev A, Wagner G (2004) Translation initiation: structures, mechanisms and evolution. Q Rev Biophys 37 : 197– 284

Milon P, Konevega AL, Gualerzi CO, Rodnina MV (2008) Kin etic checkpoint at a late step in translation initiation. Mol Cell 30 : 712–720

Mohr D, Wintermeyer W, Rodnina MV (2002) GTPase activ ation of elongation factors Tu and G on the ribosome. Biochemistry 41 : 12520–12528


Mulder FA, Bouakaz L, Lundell A, Venkataramana M, Liljas A, Akke M, Sanyal S (2004) Conformation and dynamics of ri-bosomal stalk protein L12 in solution and on the ribosome. Biochemistry 43 : 5930 – 5936

Nilsson AI, Zorzet A, Kanth A, Dahlstrom S, Berg OG, Andersson DI (2006) Reducing the fitness cost of anti biotic resistance by amplification of initiator tRNA genes. Proc Natl Acad Sci USA 103 : 6976 – 6981

O’Connor M, Gregory ST, Rajbhandary UL, Dahlberg AE (2001) Altered discrimination of start codons and initiator tRNAs by mutant initiation factor 3. RNA 7 : 969 – 978

Paleskava A, Konevega AL, Rodnina MV (2010) Thermody-namic and kinetic framework of selenocysteyl-tRNASec recognition by elongation factor SelB. J Biol Chem 285 : 3014 – 3020

Pavlov MY, Zorzet A, Andersson DI, Ehrenberg M (2011) Activ-ation of initiation factor 2 by ligands and mutations for rapid docking of ribosomal subunits. Embo J 30 : 289–301

Pavlov MY, Antoun A, Lovmar M, Ehrenberg M (2008) Comple-mentary roles of initiation factor 1 and ribosome recycling factor in 70S ribosome splitting. Embo J 27 : 1706–1717

Pavlov MY, Freistroffer DV, MacDougall J, Buckingham RH, Ehren-berg M (1997) Fast recycling of Escherichia coli ribosomes requires both ribosome recycling factor (RRF) and release factor RF3. Embo J 16 : 4134 – 4141

Peske F, Rodnina MV, Wintermeyer W (2005) Sequence of steps in ribosome recycling as defined by kinetic analysis. Mol Cell 18 : 403 – 412

Petrelli D, LaTeana A, Garofalo C, Spurio R, Pon CL, Gualerzi CO (2001) Translation initiation factor IF3: two domains, five functions, one mechanism? Embo J 20 : 4560 – 4569

Pon CL, Gualerzi CO (1984) Mechanism of protein biosynth esis in prokaryotic cells. Effect of initiation factor IF1 on the initial rate of 30 S initiation complex formation. FEBS Lett 175 : 203 – 207

Pon CL, Paci M, Pawlik RT, Gualerzi CO (1985) Structure-func-tion relationship in Escherichia coli initiation factors. Bio-chemical and biophysical characterization of the interac-tion between IF-2 and guanosine nucleotides. J Biol Chem 260 : 8918 – 8924

Raj Bhandary UL (1994) Initiator transfer RNAs. J Bacteriol 176 : 547– 552

Ringquist S, Shinedling S, Barrick D, Green L, Binkley J, Stormo GD, Gold L (1992) Translation initiation in Escherichia coli: sequences within the ribosome-binding site. Mol Microbiol 6 : 1219–1229

Risuleo G, Gualerzi C, Pon C (1976) Specificity and properties of the destabilization, induced by initiation factor IF-3, of ternary complexes of the 30-S ribosomal subunit, amino acyl-tRNA and polynucleotides. Eur J Biochem 67 : 603 – 613

Savelsbergh A, Mohr D, Wilden B, Wintermeyer W, Rodnina MV (2000) Stimulation of the GTPase activity of translation elongation factor G by ribosomal protein L7/12. J Biol Chem 275 : 890 – 894

Savelsbergh A, Rodnina MV, Wintermeyer W (2009) Distinct func-tions of elongation factor G in ribosome recycling and trans-location. RNA 15 : 772–780

Schmitt E, Guillon JM, Meinnel T, Mechulam Y, Dardel F, Blan-quet S (1996) Molecular recognition governing the initia-tion of translation in Escherichia coli. A review. Biochimie 78 : 543 – 554

Shine J, Dalgarno L (1974) The 3’-terminal sequence of Escherichia coli 16S ribosomal RNA: complementarity to nonsense tri-plets and ribosome binding sites. Proc Natl Acad Sci USA 71 : 1342–1346

Simonetti A, Marzi S, Myasnikov AG, Fabbretti A, Yusupov M, Gualerzi CO, Klaholz BP (2008) Structure of the 30S transla-tion initiation complex. Nature 455 : 416 – 420

Steiner-Mosonyi M, Creuzenet C, Keates RA, Strub BR, Mangroo D (2004) The Pseudomonas aeruginosa initiation factor IF-2 is responsible for formylation-independent protein initiation in P. aeruginosa. J Biol Chem 279 : 52 262 – 52 269

Steitz JA, Jakes K (1975) How ribosomes select initiator regions in mRNA: base pair formation between the 3’ terminus of 16S rRNA and the mRNA during initiation of protein synthesis in Escherichia coli. Proc Natl Acad Sci USA 72 : 4734 – 4738

Studer SM, Joseph S (2006) Unfolding of mRNA secondary struc-ture by the bacterial translation initiation complex. Mol Cell 22 : 105–115

Tomsic J, Vitali LA, Daviter T, Savelsbergh A, Spurio R, Striebeck P, Wintermeyer W, Rodnina MV, Gualerzi CO (2000) Late events of translation initiation in bacteria: a kinetic analysis. Embo J 19 : 2127– 2136

Wintermeyer W, Gualerzi C (1983) Effect of Escherichia coli initia-tion factors on the kinetics of N-Acphe-tRNAPhe binding to 30S ribosomal subunits. A fluorescence stopped-flow study. Biochemistry 22 : 690 – 694

Wishnia A, Boussert A, Graffe M, Dessen PH, Grunberg-Manago M (1975) Kinetics of the reversible association of ribo somal subunits: stopped-flow studies of the rate law and of the ef-fect of Mg2+. J Mol Biol 93 : 499 – 515

Zavialov AV, Buckingham RH, Ehrenberg M (2001) A post-termination ribosomal complex is the guanine nucleo-tide exchange factor for peptide release factor RF3. Cell 107 : 115–124

Zorzet A, Pavlov MY, Nilsson AI, Ehrenberg M, Andersson DI (2010) Error-prone initiation factor 2 mutations reduce the fitness cost of antibiotic resistance. Mol Microbiol 75 : 1299–1313

initiation of bacterial protein synthesis with wild …...initiation of bacterial protein synthesis...

Documents