Small protein B interacts with the large and the small subunits of a s
http://www.100md.com
《核酸研究医学期刊》
Université de Rennes I, Upres JE 2311, Inserm ESPRI, Biochimie Pharmaceutique 2 Avenue du Prof. Léon Bernard, 35043 Rennes, France
*To whom correspondence should be addressed. Tel: +332 2323 4851; Fax: +332 2323 4456; Email: bfelden@univ-rennes1.fr
ABSTRACT
During trans-translation, stalled bacterial ribosomes are rescued by small protein B (SmpB) and by transfer-messenger RNA (tmRNA). Stalled ribosomes switch translation from the defective messages to a short internal reading frame on tmRNA that tags the nascent peptide chain for degradation and recycles the ribosomes. We present evidences that SmpB binds the large and small ribosomal subunits in vivo and in vitro. The binding between SmpB and the ribosomal subunits is very tight, with a dissociation constant of 1.7 x 10–10 M, similar to its KD for the 70S ribosome or for tmRNA. tmRNA displaces SmpB from its 50S binding but not from the 30S. In vivo, SmpB is detected on the 50S when trans-translation is impaired by lacking tmRNA or a functional SmpB. SmpB contacts the large subunit transiently and early during the trans-translational process. The affinity of SmpB for the two ribosomal subunits is modulated by tmRNA in the course of trans-translation. It is the first example of two copies of the same protein interacting with two different functional sites of the ribosomes.
INTRODUCTION
Ribosomes that translate truncated mRNAs can stall and generate incomplete polypeptides. In bacteria, these problems are solved by a remarkable translational quality control surveillance system headed by an RNA–protein complex containing transfer-messenger RNA (known as tmRNA, SsrA RNA or 10Sa RNA) and small protein B (SmpB). Owing to a tRNA-like domain and an internal short coding sequence, tmRNA functions as both a tRNA and an mRNA (1). This process, termed trans-translation, allows the stuck ribosome to switch mRNA templates, resuming translation on the tmRNA internal reading frame and freeing the ribosome when it decodes tmRNA termination codon. After termination, the polypeptide encoded by the problematic mRNA gains a hydrophobic peptide tag encoded by tmRNA, leading to targeted proteolysis .
SmpB is an universal cofactor of tmRNA (4) and adopts a ?-barrel fold in solution (5). The protein binds the tRNA part of tmRNA in the elbow region on the D-like loop face (6) and has additional binding sites, including one around the first codon of tmRNA reading frame where translation resumes (7). The C-terminal tail of SmpB emerges from the ?-barrel opposite to the face that contacts tmRNA. Although having no defined structure in both free and bound states to tmRNA, it might be ordered when bound to the ribosome. C-terminal tail truncated Escherichia coli SmpB, despite being still capable of binding ribosomes and tmRNA (8,9), are inactive for trans-translation suggesting that this tail has an essential function.
E.coli SmpB enhances tmRNA aminoacylation (10) and is required for the stable association of tmRNA with ribosomes (4) and remains bound to a ‘ribosome–tmRNA’ complex isolated from cells with stalled translation at various positions within tmRNA reading frame (11). Therefore, SmpB has essential roles at early and later stages of trans-translation. Cryo-EM study of the initial entry of tmRNA into a stalled ribosome indicates that SmpB bridges tmRNA and the 50S subunit (12). We showed previously that SmpB can bind the 70S ribosome in the absence of tmRNA in vivo, triggering the subsequent recruitment of tmRNA to process trans-translation. Two molecules of SmpB are required for a single ribosome to trigger trans-transfer (13). Recent in vitro data indicate that SmpB binds 30S and 50S ribosomal subunits (14). Thus, trans-translation can be initiated in a different way than originally thought where SmpB in complex with tmRNA and EF-Tu binds to a stalled ribosome (15).
To study the functional role of SmpB proteins during trans-translation we looked at the localization of SmpB during the trans-translational process in vivo. Ribosomes from cells blocked at various stages of the trans-translational process were dissociated into ribosomal subunits at low magnesium concentration and the localization of SmpB on the 50S and 30S subunits separated in a sucrose gradient was assessed by western blotting. A similar procedure was used to show the localization of the Hsp15 protein on the 50S ribosomal subunit when translation was blocked by chloramphenicol in E.coli (16), the association of the DEVH-box protein Ski2W to the 40S ribosomal subunit (17) and the localization of the GCN2 protein kinase on the 60S ribosomal subunit (18). In this study, we demonstrate that SmpB binds both the small 30S and the large 50S ribosomal subunits in vivo and in vitro with similar affinities, with a dissociation constant of 1.7 x 10–10 M, similar to its KD for the 70S ribosome or for tmRNA. Dissociation constants between SmpB and the 70S, SmpB and the two ribosomal subunits, compared with those for tmRNA, were measured by surface plasmon resonance (SPR). Both the SPR and the competition assays in vitro indicate that tmRNA can displace SmpB from the 50S subunit, but not from the 30S one. In vivo, SmpB is detected on the 50S subunit only when trans-translation is impaired, either by omitting tmRNA from the reaction or by using a SmpB mutant that is defective in the first step of trans-translation. These results suggest that the binding of SmpB to the large ribosomal subunit is an early and transient step during trans-translation. Altogether, our data can be reconciled into a model that illustrates how two molecules of SmpB modulate their affinities towards each of the two ribosomal subunits versus tmRNA to trigger trans-translation, allowing the accommodation of tmRNA acceptor branch, trans-transfer and translation resumption on tmRNA reading frame.
MATERIALS AND METHODS
Strains and plasmids
E.coli smpB, ssrA and smpBssrA strains have been described previously (13). The promoter of smpB gene was amplified by PCR from genomic DNA of E.coli using the 5' primer (TAATATGAATTCTCCTGCTGT) and 3' primer (CGTCATAAGCTTCGTGAATCA) and then cloned into the ampicillin-resistant pBR322 plasmid at its EcoR1 and HindIII restriction sites to obtain the low-copy pBR322-pS vector. Wild-type smpB and an smpB gene construct deleted of the sequence coding for the 16 C-terminally amino acids were amplified by PCR and inserted into the HindIII and BamH1 site of the pBR322-pS vector. The truncated smpB gene was also cloned into either the pET-21a or pEt-42a to overexpress the 16-smpB protein tagged by six histidine to its C or N-terminal portion. His-tagged SmpB proteins were overexpressed and purified as previously described (13).
Complementation and in vitro trans-translation assays
Complementation with SmpB and 16-smpB was performed by the transformation of the smpB strain with either the pBR322-pS-SmpB or the pBR322-pS-16-smpB plasmids. Transformed cells were grown to mid-logarithmic phase at 37°C, diluted at an A600 of 0.01 in an LB broth containing ampicillin and incubated at 45°C. In vitro translation of poly(U) template and in vitro trans-translation was performed at 37°C in 35 μl with the PURE system described by Shimizu (19) using ribosomes from the smpBssrA strain (0.5 μM), His-tagged SmpB proteins (1 μM) and tmRNA aminoacylated with alanine (0.5 μM). Each of the final 50 μl reaction mixtures contained 15 pmol ribosomes, 30 pmol of either SmpB wt, delta 16-SmpB or no SmpB and 60 pmol elongation factor EF-Tu. Pure tRNAPhe was adjusted to 4 μM. To stall the ribosomes on a polyphenylalanine polypeptide, a polyUridine RNA (600 pmol) was translated in 47 μl reaction mixture for 30 min at 37°C in the presence of purified PheRS. Separately, 15 pmol tmRNA was aminoacylated by purified aminoacyl–tRNA synthetase in the presence of alanine (9.25 kBq) and this mixture was added to the first one to monitor the codon-independent step of trans-peptidation.
Determination of tmRNA and SmpB levels in subcellular fractions of E.coli
E.coli strains, grown in LB broth, were harvested to mid-logarithmic phase and lysed in buffer A (10 mM HEPES pH 7.5, 100 mM ammonium acetate, 10.5 mM magnesium acetate, 3mM B-mercaptoethanol) by freeze/thaw cycles. To block the in vivo translation, cells were incubated for 5 min at 37°C with 2 mM chloramphenicol and mixed with an identical volume of ice before centrifugation. The S30 (whole fraction), S100 (soluble material) and P100 (crude ribosome extract) fractions were obtained by differential centrifugations as previously described (13). For each fraction (S30, S100 et P100), an amount of RNAs and proteins corresponding to the same number of E.coli cells was analyzed by northern or western blotting. For northern hybridization, RNAs were separated by electrophoresis on a 1.5% (w/v) agarose gel containing 6.5% (v/v) formaldehyde and transferred in 10x SSC to nylon membrane by the capillary method. Pre-hybridization and hybridization with 32P-labeled DNA oligonucleotides complementary to tmRNA (5'-CGG GTA CGG GTA GGA TCG CAC ACC-3') or to 16S ribosomal RNA (5'-CCG TCC GCC ACT CGT CAG CAA-3') were carried out in ExpressHyb according to the protocol (Clontech). SmpB was imunodetected by western blotting using a rabbit polyclonal antibodies directed against His-tagged SmpB protein followed by chemiluminescence detection (Amersham Biosciences).
Sucrose density gradient centrifugation and analysis of 50S and 30S subunits
Sucrose purified ribosomes were dissociated into 50S and 30S subunits by diluting the 70S ribosome with 10-fold volume of Mg2+-free buffer. Ribosomal subunits were separated onto a 10–30% sucrose gradient (32 ml) in 10 mM HEPES pH 7.5, 100 mM sodium chloride, 1 mM magnesium acetate for 15 h at 25 000 r.p.m (Rotor SW32) and 4°C. RNAs were isolated from one-half of each fraction by extraction with phenol and precipitation with ethanol. Proteins were extracted from the other half of samples by TCA precipitation. SmpB and tmRNA were detected by western and northern blotting as described (13).
In vitro binding analysis
70S ribosomes, 50S and 30S ribosomal subunits from smpBssrA cells were purified by gradient sucrose centrifugation and incubated 10 min at 4°C with a 10-fold molar excess of His-tagged SmpB protein in buffer E (10 mM Tris–HCl pH 7.5, 100 mM ammonium chloride, 10 mM magnesium acetate and 3 mM 2-mercaptoethanol). The resulting 70S-SmpB, 50S-SmpB and 30S-SmpB complexes were purified further on a 10–30% sucrose gradient in buffer E to eliminate free SmpB proteins. 50S-SmpB subunit (10 pmol) and 30S-SmpB subunit (50 pmol) were incubated for 15 min at 37°C with increasing amount of tmRNA in 400 μl of buffer E and loaded on 10–30% sucrose gradient. Fractions either containing the 50S subunits, the 30S subunit or tmRNA (fractions including tmRNA and soluble proteins) were pooled. 70S-SmpB ribosome was incubated for 10 min at 4°C with a 4-fold molar excess of tmRNA in buffer E. The resulting 70S-SmpB-tmRNA complexes were purified on a 10–30% sucrose gradient in buffer E to eliminate free tmRNA. 70S-SmpB and 70S-SmpB-tmRNA were dissociated into 50S and 30S subunits as described above and separated onto 10–30% sucrose gradient. Proteins were extracted from one-half of samples by TCA precipitation, electrophoresed on 15% acrylamide Tris-glycine gel, transferred to PVDF membranes and SmpB was detected by western blotting. RNA from the other half of pooled fractions was prepared by phenol extraction and ethanol precipitation. TmRNA and 16S ribosomal RNA were detected by northern hybridization as described above.
Surface Plasmon Resonance (SPR) assays
Equilibrium constants of dissociation were measured by SPR using BIAcore X biosensor system (BIAcore). For real-time analyses of molecular interactions between the 70S ribosome or the individual subunits and the SmpB-His protein, anti-histidine antibody was first immobilized on a C1 sensor chip (BIAcore) by covalent linkage between activated carboxyl groups on the chip matrix and lysine residues in the anti-histidine antibody as described by the supplier (BIAapplications Handbook, BIAcore). The chips were then washed 3 times with 20 μl of 50 mM NaOH containing 0.1% SDS at a flow rate of 10 μl/min to stabilize the surface prior to the binding of His-tagged SmpB proteins. Chips were equilibrated in the continuous-flow buffer (10mM HEPES, pH 7.4, 150 mM KCl, 10 mM NH4Cl, 10 mM MgOAc, 3mM ?-mercaptoethanol, 0.05% surfactant P20). Purified His-tagged SmpB proteins were immobilized on one flow cell by injecting an aliquot of 3 nM solution protein in continuous-flow buffer. Various concentrations of 70S ribosome, 50S subunit, 30S subunit or tmRNA diluted in the running buffer were injected at a flow rate of 10 μl/min on 45–55 resonance units (RU) of SmpB proteins until equilibrium was reached. The Sensor Chip was regenerated between each injection by a 60 s pulse of 50 mM NaOH containing 0.1% SDS. Final curves were obtained by subtraction of the signal corresponding to the empty flow cell. For each concentration of 70S ribosome, 50S subunit, 30S subunit or tmRNA, the steady state binding (Req) were measured and used to calculate the equilibrium constants of dissociation following the equation .
where C is the 70S, 50S, 30S or tmRNA concentration, Rmax is the total surface binding capacity in RU and Req is the steady state binding level in RU obtained when equilibrium was reached for each concentrations of 70S ribosome, 50S subunit, 30S subunit or tmRNA. A plot of Req/C against Req at different concentrations gives a Scatchard plot from which KD (1/KA) can be calculated.
RESULTS
E.coli SmpB binds the 50S and the 30S ribosomal subunits in vivo
In the absence or presence of tmRNA, SmpB binds 70S ribosomes in vivo and in vitro (13). We have previously shown that SmpB was neither found associated to free 50S or free 30S subunits in wild type or tmRNA-deleted E.coli strains (delta ssrA strain) (13). To localize the binding site(s) of SmpB to the subunits of the 70S ribosome, we performed sucrose gradient ultra-centrifugation of purified 70S ribosomes, derived from both wild type and delta ssrA strains. The sucrose gradient experiments were performed at low (1 mM) magnesium salts in order to dissociate the 70S ribosome into its two ribosomal subunits (Figure 1A). Reducing the magnesium concentration dissociates the two ribosomal subunits (21). This procedure was used to purify the individual subunits to solve the structures of the two ribosomal subunits by X-ray crystallography (22,23). The localization of the 30S subunit onto the gradient was confirmed by northern blot hybridization against 16S rRNA (Figure 1B). The presence of tmRNA in the different fractions was monitored by northern blotting (Figure 1C). When the 70S ribosomes are dissociated into two subunits, the tmRNA is released from the ribosome to the soluble fraction (Figure 1C). On the other hand, using rabbit polyclonal antibodies directed against His-tagged SmpB, immunoblots show that endogenous SmpB mostly co-sediments with tmRNA, away from the ribosome. The presence of tmRNA and SmpB in the same fractions argues for a release of a tmRNA–SmpB complex during ribosome dissociation. However, a small fraction of the protein remains bound to the 30S subunit (Figure 1D). Interestingly, in the absence of tmRNA (delta ssrA strain), endogenous SmpB is not released from the ribosomal subunits after ribosome splitting, but remains bound predominantly to the 30S subunits (Figure 1E). Furthermore, a minor fraction of the protein is also bound to the 50S subunit. These results demonstrate that SmpB, in the context of 70S ribosome, has affinity to both ribosomal subunits in vivo.
Figure 1 Location of the SmpB protein in vivo when the ribosome dissociates into its ribosomal subunits. (A) Crude ribosomes from either wild type or delta ssrA cells were fractionated by sucrose gradient centrifugation at a low concentration of Mg2+ ions. The absorbances of each fraction from wild type or delta ssrA cells are shown. The 16S ribosomal RNA (B) and tmRNA (C) were detected by northern hybridization using complementary 32P-labeled DNA oligonucleotides. The presence of endogenous SmpB in a fraction on two of the sucrose gradient from wild-type (D) or delta ssrA (E) crude ribosomes was detected by western blotting using rabbit polyclonal antibody directed against a histidine-tagged SmpB.
E.coli SmpB binds in vitro the 30S, the 50S, the 70S and tmRNA with similar affinities
SmpB binds tmRNA (4), the ribosome (13) and each of the two ribosomal subunits separately (this study). In order to determine the equilibrium dissociation constants (KD) between SmpB and either tmRNA, the 30S, the 50S or the 70S ribosome, SPR analysis was performed as described in the Materials and Methods section. SmpB proteins were transiently coupled by their C-terminal His-tails to a flow chip preloaded covalently with a commercial anti-histidine antibody. Increasing concentrations of purified small and large ribosomal subunits, purified 70S ribosomes or purified native tmRNA were injected over the sensor chip functionalized with purified SmpB, until steady state binding level is reached. Three independent experiments were performed (Figure 2A–D are representative) and the dissociation constants derived from these experiments are presented in Figure 2E. Scatchard analysis was used to linearize the data from the saturation binding experiments (Figure 2, insets). The dissociation constants between SmpB and each of the purified ribosomal subunits are similar, 0.17 nM (Figure 2E). For the 70S binding, a curve is observed when the data are visualized on a Scatchard (Figure 2A, inset). Since the 70S ribosome has a 2.5 x 106 Da molecular weight, only one ribosome can bind to one molecule of SmpB immobilized onto the Chip which excludes a cooperative binding in this experiment but argues for a bivalent analyte model with the presence of two binding sites on the ribosome. The dissociation constants between SmpB and these two binding sites on the 70S ribosome are in the same affinity range (0.1–0.33 nM, Figure 2E). Since each ribosomal subunit can bind SmpB with a similar high affinity, the two binding sites identified on the 70S ribosome are consistent with the existence of an independent binding site on each ribosomal subunit. In the context of 70S ribosome, the two binding sites could be more or less accessible to the same extend as in isolated subunits leading to two different KDs. Alternatively, the binding of SmpB to one site could be stabilized by the overall structure of the 70S ribosome. Moreover, we can not exclude by this experiment a cooperative interaction of SmpB with both ribosome sites which can occur in solution. The KD between tmRNA and SmpB measured by SPR is 0.40 nM, consistent to those derived from filter binding assays (0.34 ± 0.18 nM, (8) or from gel mobility-shift assays . At the equilibrium, the affinities between SmpB and tmRNA, between SmpB and the 70S ribosome or between SmpB and each of the two ribosomal subunits are comparable (Figure 2E). These bindings were specific since the addition of a large excess of tRNA in the continuous flow (440 nM) did not impair the association of SmpB with 70S, 30S, 50S or tmRNA (data not shown). Indeed, SmpB is a RNA binding protein which can bind un-specifically to RNAs and especially to tRNAs (4). At this 440 nM competitor RNA concentration, we found Kds for the 70S, 50S, 30S and tmRNA to be 7.3, 4.8, 4.4 and 7.5 nM, respectively. Therefore, the affinity of SmpB for the ribosome, ribosomal subunits and tmRNA are specific and similar.
Figure 2 Determination of the dissociation constants between SmpB and either the 70S, the 30S, the 50S or tmRNA by SPR. A plot of the equilibrium binding response of (A) 70S, (B) 30S, (C) 50S and (D) tmRNA to SmpB versus concentration were obtained by the injection of increasing amounts of the purified ligands to the flow cell immobilized with SmpB. Req, in RU, was determined when the injection curve showed a plateau and thus equilibrium was reached. Insets: Scatchard analysis linearizing the data from the saturation binding experiments, to determine the KDs. For the 70S binding, a broken line is observed, favoring two binding sites. The R2 values for the linear fit were 0.99 and 0.95 for KD1 and KD2, respectively. (E) The dissociation constants at the equilibrium obtained from three independent experiments are comparable with one another. The standard deviation of each experimental data was indicated.
tmRNA can displace SmpB from the 50S subunit but not from the 30S subunit
Purified 30S or 50S subunits were bound to a SmpB-coated chip until saturation of the ligands and the dissociation of each subunit from SmpB was monitored by SPR, after adding an excess of tmRNA (Figure 3A). tmRNA could only displace the 50S and not the 30S subunit from its binding to SmpB (Figure 3A). To confirm these results, the affinities of SmpB to the 30S and 50S ribosomal subunits were compared with the one between SmpB and tmRNA using competition assays in vitro. For this, 30S or 50S ribosomes preloaded with SmpB and purified by a sucrose gradient were titrated with native tmRNA from a 1:1 to 1:10 ratio between the ribosomal subunit-SmpB complexes and the tmRNA concentrations. After incubation, the samples were loaded onto a sucrose gradient. Western blots show that SmpB binds each of the ribosomal subunit in vitro in the absence of tmRNA or additional cofactors (Figure 3B and C). Adding increasing amounts of tmRNA to each of the two ribosomal subunits preloaded with SmpB does not lead to ternary complex formation whereas a ternary complex can form between 70S, tmRNA and SmpB . At a 1:1 ratio between the 50S subunit and tmRNA, half SmpB stays bound to the 50S while the other half binds tmRNA, as expected from similar KDs between the SmpB-50S and the SmpB–tmRNA complexes (Figure 2E). At a 3- to 10-fold molar excess of tmRNA over SmpB, the protein is completely released from the 50S subunit and forms a unique complex with tmRNA in the soluble fraction (Figure 3B). Remarkably, a 10-fold excess of tmRNA over 30S ribosomes failed to pull SmpB from its ribosome-bound state to its tmRNA-bound state (Figure 3C). Thus, although SmpB has similar affinity for tmRNA and for the 30S subunit, tmRNA does not compete for SmpB binding to the 30S subunit suggesting that the conformation of the protein might be modified upon binding the 30S.
Figure 3 SmpB binds to the isolated 30S and 50S ribosomal subunits in vitro and tmRNA can compete the 50S out of the complex with SmpB, but not the 30S. (A) Competition between the individual ribosomal subunits and tmRNA for binding SmpB by SPR. The sensorgrams of 30S and 50S binding to SmpB were obtained by the injection of saturating amounts of the purified ligands to the flow cell immobilized with SmpB (association, a 2-fold higher signal for the 50S compared with the 30S reflecting mass differences between both subunits). After 100 s of dissociation, an excess of native tmRNA (400 nM) is added. Boxed inset: superimposition, before adding tmRNA, of the curves corresponding to the dissociation of both 30S and 50S subunits from SmpB; the addition of an excess of tmRNA dissociates the 50S–SmpB complex, but not the 30S–SmpB complex. His-tagged SmpB proteins were added to either the purified (B) 50S or the (C) 30S ribosomal subunits from delta smpB-ssrA cells, and the resulting 30S–SmpB or 50S–SmpB complexes were purified by sucrose gradient centrifugation. The complexes were incubated without (lanes 1 and 5) or with an increasing amount of purified native tmRNA (lanes 2–4 and 6–8) and loaded on a 10–30% (w/v) linear sucrose gradient. For each assay, fractions containing either the ribosomal subunits (lanes 1–4) or soluble molecules including tmRNA (lanes 5–8) were pooled and the amount of SmpB and tmRNA present in the ribosomal and tmRNA soluble fractions were monitored by western blotting (top panels) and by northern hybridization (bottom panels), respectively. The molar ratio between tmRNA and SmpB pre-bound ribosomal subunits in the samples loaded on sucrose gradient is indicated.
To access the affinity of SmpB for tmRNA or ribosomal subunits after in vitro 70S dissociation, 70S ribosomes were incubated with a 10-fold molar excess of His-tagged SmpB protein and purified by a sucrose density gradient centrifugation to eliminate the unbound SmpB proteins. The dissociation of the resulting 70S-SmpB complexes at low magnesium concentration reveals the presence of SmpB on the 50S and 30S subunits (Figure 4E) confirming the binding of two SmpB molecules per 70S ribosome (13) one on the 30S subunit and the other on the 50S subunit. Moreover, SmpB still bound to ribosomal subunits at low magnesium concentration indicates that magnesium did not influence the affinity of SmpB for either the 50S or the 30S subunits. The incubation of 70S-SmpB complexes with tmRNA leads to the recruitment of tmRNA without the uncoupling of SmpB from 70S ribosome . Upon dissociation at low-magnesium concentration of purified 70S-SmpB-tmRNA complexes, SmpB was released from 50S subunit in complex with tmRNA (Figure 4C and D) whereas the other SmpB molecule remains bound to the 30S subunit (Figure 4B and D). Thus, tmRNA binds only to 70S-SmpB complexes and then displaces SmpB from the 50S subunit independently of the magnesium concentration (Figures 3A, B and 4D).
Figure 4 SmpB binds the 30S and the 50S subunits in 70S ribosomes and SmpB is released from the 50S by tmRNA upon 70S dissociation. (A) Analysis of SmpB (lanes 1 and 2), tmRNA (lanes 3 and 4) and ribosomal 16S RNA (lanes 5 and 6) levels in purified 70S-SmpB (lanes 2, 4 and 6) or 70S-SmpB-tmRNA (lanes 1, 3 and 5) complexes. Purified 70S-SmpB or 70S-SmpB-tmRNA complexes were fractionated by sucrose gradient centrifugation at a low concentration of Mg2+ ions. The 16S ribosomal (B) RNA and (C) tmRNA were detected by northern hybridization using complementary 32P-labeled DNA oligonucleotides. The presence of SmpB in sucrose fractions from (D) 70S-SmpB-tmRNA or (E) 70S-SmpB complexes was detected by western blotting using rabbit polyclonal antibody directed against a histidine-tagged SmpB.
A SmpB protein mutant as a tool to investigate trans-translation
C-tail truncated SmpB variants are inactive in trans-translation but bind tmRNA or the 70S ribosomes and support the association of tmRNA with the stalled ribosomes. Also, the contribution of the tail on trans-translation is not due to an effect on SmpB stability (8,9). A recombinant E.coli SmpB missing 16 amino acids at its C-terminal tail (delta 16-SmpB) was cloned and sequenced. The phenotype of the SmpB defective strain at 45°C cannot be complemented when expressing delta 16-SmpB from a plasmid whereas a plasmid expressing wild-type SmpB does complement the growth defect (Figure 5A).
Figure 5 An inactive C-tail truncated SmpB remains bound to the ribosomal subunits but is not associated with tmRNA in the soluble fraction. (A) The growth defect of the delta SmpB strain at 45°C is not rescued by a C-tail truncated SmpB. The growth of the delta SmpB strain with either an empty plasmid (closed circle) or a plasmid allowing the expression of the C-tail truncated SmpB (open triangle) and a plasmid containing a wild-type SmpB strain (open square), is compared. (B) The truncated protein is inactive in an in vitro trans-translation assay showing the incorporation of alanine from tmRNA to stalled ribosomal complexes containing poly(U) mRNA. (C). Expression levels of tmRNA, truncated and wild type SmpB in the S30, S100 and P100 fractions detected by northern hybridization and western blotting, respectively. (D) Fractions from either wild type or cells expressing delta 16-SmpB were fractionated by sucrose gradient centrifugation at a low concentration of Mg2+ ions. The absorbance of each fraction from wild type or delta 16-SmpB cells is shown. tmRNA was detected by northern hybridization using complementary 32P-labeled DNA oligonucleotides in either the wild type (E) or delta 16-SmpB (G) cells. The presence of SmpB (F) or of delta 16-SmpB (H) in sucrose fractions from wild type or delta SmpB crude ribosomes was detected by western blotting using rabbit polyclonal antibody directed against a histidine-tagged SmpB.
In order to access the activity of this truncated SmpB protein into an in vitro trans-translation assay, we expressed and purified a delta 16-SmpB protein tagged by six histidines to its C or N-terminal portion. These truncated proteins histidine tagged binds both ribosomal subunits in vitro, as wild-type SmpB (data not shown) but are inactive in a ‘codon-independent’ transfer of a polypeptidyl-tRNAPhe to the alanine from alanyl-tmRNAAla in the ribosome P-site in vitro (Figure 5B). Thus, delta 16-SmpB protein is unable to complement the phenotype of the SmpB defective strain at 45°C because of its inability to perform the first trans-peptidation step in the trans-translation process.
Inactive C-tail truncated SmpB are not released with tmRNA when the ribosomes are dissociated
What is the location in vivo of the SmpB protein defective in trans-translation? The inactive C-tail truncated protein, expressed from a plasmid containing its endogenous promoter, is exclusively bound to the 70S ribosomes (P100) and has an expression level slightly lower than endogenous wild-type SmpB (Figure 5C). In the presence of this inactive truncated SmpB protein, tmRNA was found associated with the 70S ribosomes (P100, delta 16-SmpB, Figure 5C). However, in contrast to endogenous SmpB, higher amounts of tmRNA are in the soluble fraction in cells expressing the truncated SmpB (S100, Figure 5C). When the 70S ribosomes from the P100 fraction of the strain that expresses the inactive SmpB variant are dissociated into the 30S and the 50S ribosomal subunits, all the tmRNA molecules are released from the ribosomes without being associated with the mutant SmpB protein (Figure 5G). The inactive protein is bound to the 30S and the 50S ribosomal subunits, in a comparable ratio, between both subunits, with that observed with the strain that does not express tmRNA (see Figure 1E). Despite not working during trans-translation, the defective SmpB protein which was able to recruit tmRNA on stalled 70S ribosome still binds the two ribosomal subunits in vivo (Figure 5H). The presence of SmpB on the 50S subunit on 70S ribosomes which are all blocked in the early stages of trans-translation suggest that SmpB bind early and transiently to the 50S subunit.
The binding of endogenous SmpB to the 50S subunit is an early step in vivo
The results presented above suggest that the binding of SmpB to the large ribosomal subunit is an early and a transient step during trans-translation. If true, blocking the early steps of trans-translation in vivo might allow the detection of endogenous SmpB on the 50S subunit. To test that, translation and trans-translation are impaired in vivo by chloramphenicol that blocks aminoacyl-tRNAs and alanyl-tmRNAAla interactions with the peptidyltransferase (PT) center (24). Chloramphenicol leads to the enrichment of ribosomes blocked on the first step of trans-translation by blocking the late translational step of tmRNA. When these ribosomes blocked at various stages of trans-translation are purified and dissociated (Figure 6A), endogenous SmpB, detected by immunoblots, is in complex with the released tmRNA (Figure 6B) and with the 30S subunit (Figure 6C). This SmpB localization is in agreement with the presence of the protein bound to tmRNA and to the 30S subunit on 70S ribosomes engaged on the late step of trans-translation in chloramphenicol-untreated cells (Figure 1A). Moreover, a significant amount of SmpB is bound to the 50S subunits (Figure 6C) which represents the SmpB binding during the early step of trans-translation. Therefore, impairing translation and trans-translation in vivo with chloramphenicol allows to visualize the early step of trans-translation and transient interaction of SmpB with the large ribosomal subunit.
Figure 6 SmpB is detected on the 50S subunit when translation and trans-translation are blocked by chloramphenicol in vivo. (A) Crude ribosomes from wild-type cells which were fractionated by sucrose gradient centrifugation at a low concentration of Mg2+ ions. (B) tmRNA was detected by northern hybridization using complementary 32P-labeled DNA oligonucleotides. (C) The presence of endogenous SmpB was detected by western blotting using rabbit polyclonal antibody directed against a histidine-tagged SmpB.
DISCUSSION
In a previous work, we showed that SmpB is located exclusively on the 70S ribosomes in vivo and such complexes can recruit alanyl-tmRNA to trigger trans-translation (13). In the present study, we demonstrate that endogenous SmpB associates with both the small and the large ribosomal subunits in vivo and in vitro. In vivo, the association of SmpB with ribosomal subunits occurs only in the context of the 70S ribosome since SmpB was not detected in free 50S and 30S subunits either in the presence or the absence of tmRNA (13). Our in vivo data reveal that the dissociation of the 70S ribosomes from wild-type cells leads to the release of all tmRNA from the ribosome, in complex with SmpB. In these ribosomes engaged in trans-translation, a small fraction of the SmpB protein remains bound to the 30S subunit after ribosome dissociation. Wild-type SmpB protein is also found associated in vivo on the 50S subunit when the 70S ribosomes are purified and dissociated into subunits either from cells lacking tmRNA or from cells treated by chloramphenicol.
The ability of SmpB to interact in vivo with the two ribosomal subunits is in agreement with the in vitro data demonstrating that SmpB binds to the 30S and the 50S subunits, both prepared to high purity. The binding between SmpB and each of the ribosomal subunits is very tight and specific, with a dissociation constant of 1.7 x 10–10 M for each one, similar to its KD for the 70S and tmRNA. Although the SmpB protein binds each ribosomal subunit with similar affinity, the domains of the protein in interaction with the 30S and the 50S might be different, in agreement with the existence of two distinct clusters of conserved amino acids at the protein surface (5). SmpB has similar affinity in vitro for the two ribosomal subunits, for the 70S ribosome and for tmRNA but are found associated exclusively to the 70S in vivo which suggest that SmpB has high affinity for 70S ribosomes stalled on a problematic mRNA and argues for an early binding of the protein onto stalled ribosomes prior to the recruitment of the tmRNA.
SmpB possesses high affinity for the 30S subunit. This is consistent with the structural homology of its tertiary structure (5,25) to that of ribosomal protein S17 from the 30S subunit (5) and especially with that of initiation factor IF1 that transiently binds the A site within the 30S (26). Our biochemical results are in agreement with structural data collected at 11.8 ? resolution in the course of a recent cryo-electron study of stalled ribosomes in complex with SmpB in the absence of tmRNA (R. Gillet, S. Kaur, W. Li, M. Hallier, B. Felden, J. Frank, manuscript submitted). In that study, SmpB was located on the 30S subunit, towards the empty decoding center. Moreover, recent in vitro data indicate that SmpB footprints nucleotides that are in the vicinity of the P-site of the small subunit (14).
The binding of SmpB to the 50S subunit is consistent with cryo-EM data obtained with a ribosome from Thermus thermophilus, stalled at the end of a short mRNA and reacted with alanyl-tmRNAala, his-tagged SmpB and EF-Tu-GTP. In this pre-accommodated ribosomal complex, SmpB was initially found to be exclusively located on the 50S subunit (12). A pre-accommodation step reconstructed at a higher resolution, however, reveals the presence of two molecules of SmpB, one interacts with the large subunit while the second is directed towards the decoding site (DS) within the 30S subunit (R. Gillet, S. Kaur, W. Li, M. Hallier, B. Felden, J. Frank, manuscript submitted). The presence of more than one SmpB protein on ribosomal complexes agrees with our previous work showing that two molecules of SmpB are required for trans-transfer events in vitro (13).
In contrast to SmpB-70S preformed complexes (Figure 4A), neither the SmpB-50S nor the SmpB-30S complexes are capable of recruiting tmRNA in vitro . When the 70S ribosomes, the 50S or the 30S subunits compete with tmRNA for binding SmpB, tmRNA can only release SmpB from the large ribosomal subunit (Figure 3). Moreover, SmpB is released from the 50S subunits with tmRNA when the 70S ribosomes containing SmpB and tmRNA are dissociated in two subunits. Thus, while SmpB presents a similar affinity for ribosomal subunits and tmRNA, the binding of tmRNA on 70S-SmpB complexes leads to a modulation of SmpB affinity for the two ribosomal subunits versus tmRNA in these ribosomes not engaged in trans-translation. It suggests that SmpB is transiently associated to the 50S subunit during trans-translation but tightly and stably bound to the 30S subunit. In the presence of tmRNA, SmpB contacts the large subunit by a close interaction with helix 69 in the 23S RNA (12). To avoid steric clashes with the P-site tRNA when the accommodation proceeds, SmpB has to move away from the 50S subunit (6). The transient and early association of SmpB to the 50S subunit agrees with its location on wild-type ribosomes performing trans-translation, in which SmpB is only detected on the small subunit (Figure 1D). Moreover, SmpB on the 50S in vivo is detected when trans-translation is voluntarily impaired by chloramphenicol, which leads to the accumulation of 70S ribosomes blocked at the early stages of trans-translation. Since the association between SmpB and ribosomal subunits or tmRNA is independent of the magnesium concentration (Figures 3 and 4), the presence of SmpB on the 50S subunit in chloramphenicol-treated cells is not because of the low level of magnesium but reflects a stage during trans-translation.
This notion that SmpB binding to the 50S is a very early and transient step during trans-translation is reinforced further by the subunits localization of a mutant SmpB protein. Indeed, when wild type SmpB protein is replaced by C-tail truncated SmpB variant that is inactive in the first step of trans-translation, the truncated protein remains bound to both ribosomal subunits in vivo upon ribosome dissociation, without being released in complex with tmRNA. Mutant protein binds both ribosomal subunits and recruits tmRNA but is unable to trigger essential conformational rearrangement(s) within the stalled ribosome that are required for its binding and release with tmRNA from the stalled ribosome. This suggests that SmpB has to move away from the 50S subunit during the accommodation of alanyl-tmRNAAla into the large subunit.
What is the fate of SmpB on the 50S? Complex formation between tmRNA and SmpB was recently monitored by chemical and enzymatic probes combined to SPR measurements, and three independent binding sites of the protein onto tmRNA were identified (7). These data suggest that during trans-translation SmpB is able to bind tmRNA at different locations, to induce key structural changes within the tmRNA scaffold required for the sliding of tmRNA through the ribosome to be rescued. The protein could re-associate onto tmRNA structure at a binding site different from the one on the tRNA domain, to participate in resuming translation on tmRNA internal tag reading frame.
We also detect some of the protein bound to the 30S upon subunit dissociation, in agreement with a stably in vitro binding of the protein to the 30S, contrasting with its labile binding on the 50S in the presence of tmRNA. tmRNA has a tRNA accepting branch that has to fit within the PT center on the 50S, but lacks an anticodon stem–loop analog that fits into the DS of the 30S during the decoding of aminoacylated tRNAs in canonical translation. Assuming that SmpB binds at or near functional sites within the ribosome, it could rationalize the competition of tmRNA for the binding of SmpB to the large subunit but not on the small subunit. During trans-translation, however, when the location of SmpB on the ribosomal subunits from wild-type cells is monitored in vivo, the majority of the protein is released with tmRNA with only a minor fraction that remains attached to the 30S. It suggests that during trans-translation, the molecules of SmpB bound to the 30S either move onto tmRNA or are released from the ribosome.
Altogether, these data can be reconciled into a working model (Figure 7). During trans-translation, SmpB modulates its affinity for the ribosomal subunits versus tmRNA. When a ribosome stalls on a problematic message, SmpB would bind the 70S ribosome in association with tmRNA or prior to recruiting tmRNA. During pre-accommodation, SmpB is still bound to the large and the small ribosomal subunits. The accommodation results in the movement of both the alanylated-tmRNAAla and the SmpB protein bound to the 50S subunit. SmpB either moves to tmRNA or is released from the 50S subunit, leading to the entrance of the tmRNA acceptor branch within the PT center. Subsequently, there is a codon-independent transfer of the peptidyl residue stalled in the P-site to the alanine from tmRNA. The molecule of SmpB bound to the 30S either binds tmRNA at different site from the tRNA-like domain or is released from the ribosome. Indeed, the molecule of SmpB bound to the small subunit has to move to allow translation resumption on tmRNA reading frame. The molecule that has been recently identified by Shpanchenko et al (2005) based on monitoring in vivo ribosome–tmRNA complexes during trans-translation could come from that bound to either the 30S or the 50S (11). This report adds several important insights concerning the location, the functional and structural roles of the molecules of SmpB during bacterial ribosome rescue. Many important questions remain to be elucidated, and among them, when and how the SmpB proteins, alone or in complex with tmRNA, identify the ribosomes that are being stalled on problematic messages.
Figure 7 Schematic illustration of the location of the SmpB molecules during trans-translation. SmpB proteins bound to both ribosomal subunits and tmRNA when a ribosome is stalled on a problematic mRNA. During the accommodation, SmpB moves from the 50S subunit on the tmRNA to allow the entrance of tmRNA acceptor branch within the PT center. The transfer of the nascent polypeptide chain to the alanine charged on tmRNA requires that SmpB moves away from the 30S subunit. One SmpB is released from the ribosome and the other remains associated with tmRNA during translation of its internal reading frame.
ACKNOWLEDGEMENTS
The authors thank Drs L. Metzinger and R. Gillet for the critical reading of the manuscript. This work was supported by the ‘Région Bretagne’ (Programme d'accueil d'équipes en émergence) to Brice Felden. Funding to pay the Open Access publication charges for this article was provided by Region Bretagne.
REFERENCES
Keiler, K.C., Waller, P.R.H., Sauer, R.T. (1996) Role of a peptide tagging system in degradation of proteins synthesized from damaged messenger RNA Science, 271, 990–993 .
Saguy, M., Gillet, R., Metzinger, L., Felden, B. (2005) tmRNA and associated ligands: a puzzling relationship Biochimie, . 87, 897–903 .
Haebel, P.W., Gutmann, S., Ban, N. (2004) Dial tm for rescue: tmRNA engages ribosomes stalled on defective mRNAs Curr. Opin. Struct. Biol, . 14, 58–65 .
Karzai, A.W., Susskind, M.M., Sauer, R.T. (1999) SmpB, a unique RNA-binding protein essential for the peptide-tagging activity of SsrA (tmRNA) EMBO J, . 18, 3793–3799 .
Dong, G., Nowakowski, J., Hoffman, D.W. (2002) Structure of small protein B: the protein component of the tmRNA-SmpB system for ribosome rescue EMBO J, . 21, 1845–1854 .
Gutmann, S., Haebel, P.W., Metzinger, L., Sutter, M., Felden, B., Ban, N. (2003) Crystal structure of the transfer-RNA domain of transfer-messenger RNA in complex with SmpB Nature, 424, 699–703 .
Metzinger, L., Hallier, M., Felden, B. (2005) Independent binding sites of small protein B onto transfer-messenger RNA during trans-translation Nucleic Acids Res, . 33, 2384–2394 .
Jacob, Y., Sharkady, S.M., Bhardwaj, K., Sanda, A., Williams, K.P. (2005) Function of the SmpB tail in transfer-messenger RNA translation revealed by a nucleus-encoded form J. Biol. Chem, . 280, 5503–5509 .
Sundermeier, T.R., Dulebohn, D.P., Cho, H.J., Karzai, A.W. (2005) A previously uncharacterized role for small protein B (SmpB) in transfer messenger RNA-mediated trans-translation Proc. Natl Acad. Sci. USA, 102, 2316–2321 .
Barends, S., Karzai, A.W., Sauer, R.T., Wower, J., Kraal, B. (2001) Simultaneous and functional binding of SmpB and EF-Tu-TP to the alanyl acceptor arm of tmRNA J. Mol. Biol, . 314, 9–21 .
Shpanchenko, O.V., Zvereva, M.I., Ivanov, P.V., Bugaeva, E.Y., Rozov, A.S., Bogdanov, A.A., Kalkum, M., Isaksson, L.A., Nierhaus, K.H., Dontsova, O.A. (2005) Stepping transfer messenger RNA through the ribosome J. Biol. Chem, . 280, 18368–18374 .
Valle, M., Gillet, R., Kaur, S., Henne, A., Ramakrishnan, V., Frank, J. (2003) Visualizing tmRNA entry into a stalled ribosome Science, 300, 127–130 .
Hallier, M., Ivanova, N., Rametti, A., Pavlov, M., Ehrenberg, M., Felden, B. (2004) Pre-binding of small protein B to a stalled ribosome triggers trans-translation J. Biol. Chem, . 279, 25978–25985 .
Ivanova, N., Pavlov, M.Y., Bouakaz, E., Ehrenberg, M., Schiavone, L.H. (2005) Mapping the interaction of SmpB with ribosomes by footprinting of ribosomal RNA Nucleic Acids Res, . 33, 3529–3539 .
Rudinger-Thirion, J., Giege, R., Felden, B. (1999) Aminoacylated tmRNA from Escherichia coli interacts with prokaryotic elongation factor Tu RNA, 5, 1–4 .
Korber, P., Stahl, J.M., Nierhaus, K.H., Bardwell, J.C. (2000) Hsp15: a ribosome-associated heat shock protein Embo J, . 19, 741–748 .
Qu, X., Yang, Z., Zhang, S., Shen, L., Dangel, A.W., Hughes, J.H., Redman, K.L., Wu, L.C., Yu, C.Y. (1998) The human DEVH-box protein Ski2w from the HLA is localized in nucleoli and ribosomes Nucleic Acids Res, . 26, 4068–4077 .
Ramirez, M., Wek, R.C., Hinnebusch, A.G. (1991) Ribosome association of GCN2 protein kinase, a translational activator of the GCN4 gene of Saccharomyces cerevisiae Mol. Cell Biol, . 11, 3027–3036 .
Shimizu, Y. and Ueda, T. (2002) The role of SmpB protein in trans-translation FEBS Lett, . 514, 74–77 .
Hart, D.J., Speight, R.E., Cooper, M.A., Sutherland, J.D., Blackburn, J.M. (1999) The salt dependence of DNA recognition by NF-kappaB p50: a detailed kinetic analysis of the effects on affinity and specificity Nucleic Acids Res, . 27, 1063–1069 .
Noll, M. and Noll, H. (1976) Structural dynamics of bacterial ribosomes. V. Magnesium-dependent dissociation of tight couples into subunits: measurements of dissociation constants and exchange rates J. Mol. Biol, . 105, 111–130 .
Ban, N., Freeborn, B., Nissen, P., Penczek, P., Grassucci, R.A., Sweet, R., Frank, J., Moore, P.B., Steitz, T.A. (1998) A 9 A resolution X-ray crystallographic map of the large ribosomal subunit Cell, 93, 1105–1115 .
Clemons, W.M., Jr, Brodersen, D.E., McCutcheon, J.P., May, J.L., Carter, A.P., Morgan-Warren, R.J., Wimberly, B.T., Ramakrishnan, V. (2001) Crystal structure of the 30 S ribosomal subunit from Thermus thermophilus: purification, crystallization and structure determination J. Mol. Biol, . 310, 827–843 .
Schlunzen, F., Zarivach, R., Harms, J., Bashan, A., Tocilj, A., Albrecht, R., Yonath, A., Franceschi, F. (2001) Structural basis for the interaction of antibiotics with the peptidyl transferase centre in eubacteria Nature, 413, 814–821 .
Someya, T., Nameki, N., Hosoi, H., Suzuki, S., Hatanaka, H., Fujii, M., Terada, T., Shirouzu, M., Inoue, Y., Shibata, T., et al. (2003) Solution structure of a tmRNA-binding protein, SmpB, from Thermus thermophilus FEBS Lett, . 535, 94–100 .
Carter, A.P., Clemons, W.M., Jr, Brodersen, D.E., Morgan-Warren, R.J., Hartsch, T., Wimberly, B.T., Ramakrishnan, V. (2001) Crystal structure of an initiation factor bound to the 30S ribosomal subunit Science, 291, 498–501 .(Marc Hallier, Jessie Desreac and Brice F)
*To whom correspondence should be addressed. Tel: +332 2323 4851; Fax: +332 2323 4456; Email: bfelden@univ-rennes1.fr
ABSTRACT
During trans-translation, stalled bacterial ribosomes are rescued by small protein B (SmpB) and by transfer-messenger RNA (tmRNA). Stalled ribosomes switch translation from the defective messages to a short internal reading frame on tmRNA that tags the nascent peptide chain for degradation and recycles the ribosomes. We present evidences that SmpB binds the large and small ribosomal subunits in vivo and in vitro. The binding between SmpB and the ribosomal subunits is very tight, with a dissociation constant of 1.7 x 10–10 M, similar to its KD for the 70S ribosome or for tmRNA. tmRNA displaces SmpB from its 50S binding but not from the 30S. In vivo, SmpB is detected on the 50S when trans-translation is impaired by lacking tmRNA or a functional SmpB. SmpB contacts the large subunit transiently and early during the trans-translational process. The affinity of SmpB for the two ribosomal subunits is modulated by tmRNA in the course of trans-translation. It is the first example of two copies of the same protein interacting with two different functional sites of the ribosomes.
INTRODUCTION
Ribosomes that translate truncated mRNAs can stall and generate incomplete polypeptides. In bacteria, these problems are solved by a remarkable translational quality control surveillance system headed by an RNA–protein complex containing transfer-messenger RNA (known as tmRNA, SsrA RNA or 10Sa RNA) and small protein B (SmpB). Owing to a tRNA-like domain and an internal short coding sequence, tmRNA functions as both a tRNA and an mRNA (1). This process, termed trans-translation, allows the stuck ribosome to switch mRNA templates, resuming translation on the tmRNA internal reading frame and freeing the ribosome when it decodes tmRNA termination codon. After termination, the polypeptide encoded by the problematic mRNA gains a hydrophobic peptide tag encoded by tmRNA, leading to targeted proteolysis .
SmpB is an universal cofactor of tmRNA (4) and adopts a ?-barrel fold in solution (5). The protein binds the tRNA part of tmRNA in the elbow region on the D-like loop face (6) and has additional binding sites, including one around the first codon of tmRNA reading frame where translation resumes (7). The C-terminal tail of SmpB emerges from the ?-barrel opposite to the face that contacts tmRNA. Although having no defined structure in both free and bound states to tmRNA, it might be ordered when bound to the ribosome. C-terminal tail truncated Escherichia coli SmpB, despite being still capable of binding ribosomes and tmRNA (8,9), are inactive for trans-translation suggesting that this tail has an essential function.
E.coli SmpB enhances tmRNA aminoacylation (10) and is required for the stable association of tmRNA with ribosomes (4) and remains bound to a ‘ribosome–tmRNA’ complex isolated from cells with stalled translation at various positions within tmRNA reading frame (11). Therefore, SmpB has essential roles at early and later stages of trans-translation. Cryo-EM study of the initial entry of tmRNA into a stalled ribosome indicates that SmpB bridges tmRNA and the 50S subunit (12). We showed previously that SmpB can bind the 70S ribosome in the absence of tmRNA in vivo, triggering the subsequent recruitment of tmRNA to process trans-translation. Two molecules of SmpB are required for a single ribosome to trigger trans-transfer (13). Recent in vitro data indicate that SmpB binds 30S and 50S ribosomal subunits (14). Thus, trans-translation can be initiated in a different way than originally thought where SmpB in complex with tmRNA and EF-Tu binds to a stalled ribosome (15).
To study the functional role of SmpB proteins during trans-translation we looked at the localization of SmpB during the trans-translational process in vivo. Ribosomes from cells blocked at various stages of the trans-translational process were dissociated into ribosomal subunits at low magnesium concentration and the localization of SmpB on the 50S and 30S subunits separated in a sucrose gradient was assessed by western blotting. A similar procedure was used to show the localization of the Hsp15 protein on the 50S ribosomal subunit when translation was blocked by chloramphenicol in E.coli (16), the association of the DEVH-box protein Ski2W to the 40S ribosomal subunit (17) and the localization of the GCN2 protein kinase on the 60S ribosomal subunit (18). In this study, we demonstrate that SmpB binds both the small 30S and the large 50S ribosomal subunits in vivo and in vitro with similar affinities, with a dissociation constant of 1.7 x 10–10 M, similar to its KD for the 70S ribosome or for tmRNA. Dissociation constants between SmpB and the 70S, SmpB and the two ribosomal subunits, compared with those for tmRNA, were measured by surface plasmon resonance (SPR). Both the SPR and the competition assays in vitro indicate that tmRNA can displace SmpB from the 50S subunit, but not from the 30S one. In vivo, SmpB is detected on the 50S subunit only when trans-translation is impaired, either by omitting tmRNA from the reaction or by using a SmpB mutant that is defective in the first step of trans-translation. These results suggest that the binding of SmpB to the large ribosomal subunit is an early and transient step during trans-translation. Altogether, our data can be reconciled into a model that illustrates how two molecules of SmpB modulate their affinities towards each of the two ribosomal subunits versus tmRNA to trigger trans-translation, allowing the accommodation of tmRNA acceptor branch, trans-transfer and translation resumption on tmRNA reading frame.
MATERIALS AND METHODS
Strains and plasmids
E.coli smpB, ssrA and smpBssrA strains have been described previously (13). The promoter of smpB gene was amplified by PCR from genomic DNA of E.coli using the 5' primer (TAATATGAATTCTCCTGCTGT) and 3' primer (CGTCATAAGCTTCGTGAATCA) and then cloned into the ampicillin-resistant pBR322 plasmid at its EcoR1 and HindIII restriction sites to obtain the low-copy pBR322-pS vector. Wild-type smpB and an smpB gene construct deleted of the sequence coding for the 16 C-terminally amino acids were amplified by PCR and inserted into the HindIII and BamH1 site of the pBR322-pS vector. The truncated smpB gene was also cloned into either the pET-21a or pEt-42a to overexpress the 16-smpB protein tagged by six histidine to its C or N-terminal portion. His-tagged SmpB proteins were overexpressed and purified as previously described (13).
Complementation and in vitro trans-translation assays
Complementation with SmpB and 16-smpB was performed by the transformation of the smpB strain with either the pBR322-pS-SmpB or the pBR322-pS-16-smpB plasmids. Transformed cells were grown to mid-logarithmic phase at 37°C, diluted at an A600 of 0.01 in an LB broth containing ampicillin and incubated at 45°C. In vitro translation of poly(U) template and in vitro trans-translation was performed at 37°C in 35 μl with the PURE system described by Shimizu (19) using ribosomes from the smpBssrA strain (0.5 μM), His-tagged SmpB proteins (1 μM) and tmRNA aminoacylated with alanine (0.5 μM). Each of the final 50 μl reaction mixtures contained 15 pmol ribosomes, 30 pmol of either SmpB wt, delta 16-SmpB or no SmpB and 60 pmol elongation factor EF-Tu. Pure tRNAPhe was adjusted to 4 μM. To stall the ribosomes on a polyphenylalanine polypeptide, a polyUridine RNA (600 pmol) was translated in 47 μl reaction mixture for 30 min at 37°C in the presence of purified PheRS. Separately, 15 pmol tmRNA was aminoacylated by purified aminoacyl–tRNA synthetase in the presence of alanine (9.25 kBq) and this mixture was added to the first one to monitor the codon-independent step of trans-peptidation.
Determination of tmRNA and SmpB levels in subcellular fractions of E.coli
E.coli strains, grown in LB broth, were harvested to mid-logarithmic phase and lysed in buffer A (10 mM HEPES pH 7.5, 100 mM ammonium acetate, 10.5 mM magnesium acetate, 3mM B-mercaptoethanol) by freeze/thaw cycles. To block the in vivo translation, cells were incubated for 5 min at 37°C with 2 mM chloramphenicol and mixed with an identical volume of ice before centrifugation. The S30 (whole fraction), S100 (soluble material) and P100 (crude ribosome extract) fractions were obtained by differential centrifugations as previously described (13). For each fraction (S30, S100 et P100), an amount of RNAs and proteins corresponding to the same number of E.coli cells was analyzed by northern or western blotting. For northern hybridization, RNAs were separated by electrophoresis on a 1.5% (w/v) agarose gel containing 6.5% (v/v) formaldehyde and transferred in 10x SSC to nylon membrane by the capillary method. Pre-hybridization and hybridization with 32P-labeled DNA oligonucleotides complementary to tmRNA (5'-CGG GTA CGG GTA GGA TCG CAC ACC-3') or to 16S ribosomal RNA (5'-CCG TCC GCC ACT CGT CAG CAA-3') were carried out in ExpressHyb according to the protocol (Clontech). SmpB was imunodetected by western blotting using a rabbit polyclonal antibodies directed against His-tagged SmpB protein followed by chemiluminescence detection (Amersham Biosciences).
Sucrose density gradient centrifugation and analysis of 50S and 30S subunits
Sucrose purified ribosomes were dissociated into 50S and 30S subunits by diluting the 70S ribosome with 10-fold volume of Mg2+-free buffer. Ribosomal subunits were separated onto a 10–30% sucrose gradient (32 ml) in 10 mM HEPES pH 7.5, 100 mM sodium chloride, 1 mM magnesium acetate for 15 h at 25 000 r.p.m (Rotor SW32) and 4°C. RNAs were isolated from one-half of each fraction by extraction with phenol and precipitation with ethanol. Proteins were extracted from the other half of samples by TCA precipitation. SmpB and tmRNA were detected by western and northern blotting as described (13).
In vitro binding analysis
70S ribosomes, 50S and 30S ribosomal subunits from smpBssrA cells were purified by gradient sucrose centrifugation and incubated 10 min at 4°C with a 10-fold molar excess of His-tagged SmpB protein in buffer E (10 mM Tris–HCl pH 7.5, 100 mM ammonium chloride, 10 mM magnesium acetate and 3 mM 2-mercaptoethanol). The resulting 70S-SmpB, 50S-SmpB and 30S-SmpB complexes were purified further on a 10–30% sucrose gradient in buffer E to eliminate free SmpB proteins. 50S-SmpB subunit (10 pmol) and 30S-SmpB subunit (50 pmol) were incubated for 15 min at 37°C with increasing amount of tmRNA in 400 μl of buffer E and loaded on 10–30% sucrose gradient. Fractions either containing the 50S subunits, the 30S subunit or tmRNA (fractions including tmRNA and soluble proteins) were pooled. 70S-SmpB ribosome was incubated for 10 min at 4°C with a 4-fold molar excess of tmRNA in buffer E. The resulting 70S-SmpB-tmRNA complexes were purified on a 10–30% sucrose gradient in buffer E to eliminate free tmRNA. 70S-SmpB and 70S-SmpB-tmRNA were dissociated into 50S and 30S subunits as described above and separated onto 10–30% sucrose gradient. Proteins were extracted from one-half of samples by TCA precipitation, electrophoresed on 15% acrylamide Tris-glycine gel, transferred to PVDF membranes and SmpB was detected by western blotting. RNA from the other half of pooled fractions was prepared by phenol extraction and ethanol precipitation. TmRNA and 16S ribosomal RNA were detected by northern hybridization as described above.
Surface Plasmon Resonance (SPR) assays
Equilibrium constants of dissociation were measured by SPR using BIAcore X biosensor system (BIAcore). For real-time analyses of molecular interactions between the 70S ribosome or the individual subunits and the SmpB-His protein, anti-histidine antibody was first immobilized on a C1 sensor chip (BIAcore) by covalent linkage between activated carboxyl groups on the chip matrix and lysine residues in the anti-histidine antibody as described by the supplier (BIAapplications Handbook, BIAcore). The chips were then washed 3 times with 20 μl of 50 mM NaOH containing 0.1% SDS at a flow rate of 10 μl/min to stabilize the surface prior to the binding of His-tagged SmpB proteins. Chips were equilibrated in the continuous-flow buffer (10mM HEPES, pH 7.4, 150 mM KCl, 10 mM NH4Cl, 10 mM MgOAc, 3mM ?-mercaptoethanol, 0.05% surfactant P20). Purified His-tagged SmpB proteins were immobilized on one flow cell by injecting an aliquot of 3 nM solution protein in continuous-flow buffer. Various concentrations of 70S ribosome, 50S subunit, 30S subunit or tmRNA diluted in the running buffer were injected at a flow rate of 10 μl/min on 45–55 resonance units (RU) of SmpB proteins until equilibrium was reached. The Sensor Chip was regenerated between each injection by a 60 s pulse of 50 mM NaOH containing 0.1% SDS. Final curves were obtained by subtraction of the signal corresponding to the empty flow cell. For each concentration of 70S ribosome, 50S subunit, 30S subunit or tmRNA, the steady state binding (Req) were measured and used to calculate the equilibrium constants of dissociation following the equation .
where C is the 70S, 50S, 30S or tmRNA concentration, Rmax is the total surface binding capacity in RU and Req is the steady state binding level in RU obtained when equilibrium was reached for each concentrations of 70S ribosome, 50S subunit, 30S subunit or tmRNA. A plot of Req/C against Req at different concentrations gives a Scatchard plot from which KD (1/KA) can be calculated.
RESULTS
E.coli SmpB binds the 50S and the 30S ribosomal subunits in vivo
In the absence or presence of tmRNA, SmpB binds 70S ribosomes in vivo and in vitro (13). We have previously shown that SmpB was neither found associated to free 50S or free 30S subunits in wild type or tmRNA-deleted E.coli strains (delta ssrA strain) (13). To localize the binding site(s) of SmpB to the subunits of the 70S ribosome, we performed sucrose gradient ultra-centrifugation of purified 70S ribosomes, derived from both wild type and delta ssrA strains. The sucrose gradient experiments were performed at low (1 mM) magnesium salts in order to dissociate the 70S ribosome into its two ribosomal subunits (Figure 1A). Reducing the magnesium concentration dissociates the two ribosomal subunits (21). This procedure was used to purify the individual subunits to solve the structures of the two ribosomal subunits by X-ray crystallography (22,23). The localization of the 30S subunit onto the gradient was confirmed by northern blot hybridization against 16S rRNA (Figure 1B). The presence of tmRNA in the different fractions was monitored by northern blotting (Figure 1C). When the 70S ribosomes are dissociated into two subunits, the tmRNA is released from the ribosome to the soluble fraction (Figure 1C). On the other hand, using rabbit polyclonal antibodies directed against His-tagged SmpB, immunoblots show that endogenous SmpB mostly co-sediments with tmRNA, away from the ribosome. The presence of tmRNA and SmpB in the same fractions argues for a release of a tmRNA–SmpB complex during ribosome dissociation. However, a small fraction of the protein remains bound to the 30S subunit (Figure 1D). Interestingly, in the absence of tmRNA (delta ssrA strain), endogenous SmpB is not released from the ribosomal subunits after ribosome splitting, but remains bound predominantly to the 30S subunits (Figure 1E). Furthermore, a minor fraction of the protein is also bound to the 50S subunit. These results demonstrate that SmpB, in the context of 70S ribosome, has affinity to both ribosomal subunits in vivo.
Figure 1 Location of the SmpB protein in vivo when the ribosome dissociates into its ribosomal subunits. (A) Crude ribosomes from either wild type or delta ssrA cells were fractionated by sucrose gradient centrifugation at a low concentration of Mg2+ ions. The absorbances of each fraction from wild type or delta ssrA cells are shown. The 16S ribosomal RNA (B) and tmRNA (C) were detected by northern hybridization using complementary 32P-labeled DNA oligonucleotides. The presence of endogenous SmpB in a fraction on two of the sucrose gradient from wild-type (D) or delta ssrA (E) crude ribosomes was detected by western blotting using rabbit polyclonal antibody directed against a histidine-tagged SmpB.
E.coli SmpB binds in vitro the 30S, the 50S, the 70S and tmRNA with similar affinities
SmpB binds tmRNA (4), the ribosome (13) and each of the two ribosomal subunits separately (this study). In order to determine the equilibrium dissociation constants (KD) between SmpB and either tmRNA, the 30S, the 50S or the 70S ribosome, SPR analysis was performed as described in the Materials and Methods section. SmpB proteins were transiently coupled by their C-terminal His-tails to a flow chip preloaded covalently with a commercial anti-histidine antibody. Increasing concentrations of purified small and large ribosomal subunits, purified 70S ribosomes or purified native tmRNA were injected over the sensor chip functionalized with purified SmpB, until steady state binding level is reached. Three independent experiments were performed (Figure 2A–D are representative) and the dissociation constants derived from these experiments are presented in Figure 2E. Scatchard analysis was used to linearize the data from the saturation binding experiments (Figure 2, insets). The dissociation constants between SmpB and each of the purified ribosomal subunits are similar, 0.17 nM (Figure 2E). For the 70S binding, a curve is observed when the data are visualized on a Scatchard (Figure 2A, inset). Since the 70S ribosome has a 2.5 x 106 Da molecular weight, only one ribosome can bind to one molecule of SmpB immobilized onto the Chip which excludes a cooperative binding in this experiment but argues for a bivalent analyte model with the presence of two binding sites on the ribosome. The dissociation constants between SmpB and these two binding sites on the 70S ribosome are in the same affinity range (0.1–0.33 nM, Figure 2E). Since each ribosomal subunit can bind SmpB with a similar high affinity, the two binding sites identified on the 70S ribosome are consistent with the existence of an independent binding site on each ribosomal subunit. In the context of 70S ribosome, the two binding sites could be more or less accessible to the same extend as in isolated subunits leading to two different KDs. Alternatively, the binding of SmpB to one site could be stabilized by the overall structure of the 70S ribosome. Moreover, we can not exclude by this experiment a cooperative interaction of SmpB with both ribosome sites which can occur in solution. The KD between tmRNA and SmpB measured by SPR is 0.40 nM, consistent to those derived from filter binding assays (0.34 ± 0.18 nM, (8) or from gel mobility-shift assays . At the equilibrium, the affinities between SmpB and tmRNA, between SmpB and the 70S ribosome or between SmpB and each of the two ribosomal subunits are comparable (Figure 2E). These bindings were specific since the addition of a large excess of tRNA in the continuous flow (440 nM) did not impair the association of SmpB with 70S, 30S, 50S or tmRNA (data not shown). Indeed, SmpB is a RNA binding protein which can bind un-specifically to RNAs and especially to tRNAs (4). At this 440 nM competitor RNA concentration, we found Kds for the 70S, 50S, 30S and tmRNA to be 7.3, 4.8, 4.4 and 7.5 nM, respectively. Therefore, the affinity of SmpB for the ribosome, ribosomal subunits and tmRNA are specific and similar.
Figure 2 Determination of the dissociation constants between SmpB and either the 70S, the 30S, the 50S or tmRNA by SPR. A plot of the equilibrium binding response of (A) 70S, (B) 30S, (C) 50S and (D) tmRNA to SmpB versus concentration were obtained by the injection of increasing amounts of the purified ligands to the flow cell immobilized with SmpB. Req, in RU, was determined when the injection curve showed a plateau and thus equilibrium was reached. Insets: Scatchard analysis linearizing the data from the saturation binding experiments, to determine the KDs. For the 70S binding, a broken line is observed, favoring two binding sites. The R2 values for the linear fit were 0.99 and 0.95 for KD1 and KD2, respectively. (E) The dissociation constants at the equilibrium obtained from three independent experiments are comparable with one another. The standard deviation of each experimental data was indicated.
tmRNA can displace SmpB from the 50S subunit but not from the 30S subunit
Purified 30S or 50S subunits were bound to a SmpB-coated chip until saturation of the ligands and the dissociation of each subunit from SmpB was monitored by SPR, after adding an excess of tmRNA (Figure 3A). tmRNA could only displace the 50S and not the 30S subunit from its binding to SmpB (Figure 3A). To confirm these results, the affinities of SmpB to the 30S and 50S ribosomal subunits were compared with the one between SmpB and tmRNA using competition assays in vitro. For this, 30S or 50S ribosomes preloaded with SmpB and purified by a sucrose gradient were titrated with native tmRNA from a 1:1 to 1:10 ratio between the ribosomal subunit-SmpB complexes and the tmRNA concentrations. After incubation, the samples were loaded onto a sucrose gradient. Western blots show that SmpB binds each of the ribosomal subunit in vitro in the absence of tmRNA or additional cofactors (Figure 3B and C). Adding increasing amounts of tmRNA to each of the two ribosomal subunits preloaded with SmpB does not lead to ternary complex formation whereas a ternary complex can form between 70S, tmRNA and SmpB . At a 1:1 ratio between the 50S subunit and tmRNA, half SmpB stays bound to the 50S while the other half binds tmRNA, as expected from similar KDs between the SmpB-50S and the SmpB–tmRNA complexes (Figure 2E). At a 3- to 10-fold molar excess of tmRNA over SmpB, the protein is completely released from the 50S subunit and forms a unique complex with tmRNA in the soluble fraction (Figure 3B). Remarkably, a 10-fold excess of tmRNA over 30S ribosomes failed to pull SmpB from its ribosome-bound state to its tmRNA-bound state (Figure 3C). Thus, although SmpB has similar affinity for tmRNA and for the 30S subunit, tmRNA does not compete for SmpB binding to the 30S subunit suggesting that the conformation of the protein might be modified upon binding the 30S.
Figure 3 SmpB binds to the isolated 30S and 50S ribosomal subunits in vitro and tmRNA can compete the 50S out of the complex with SmpB, but not the 30S. (A) Competition between the individual ribosomal subunits and tmRNA for binding SmpB by SPR. The sensorgrams of 30S and 50S binding to SmpB were obtained by the injection of saturating amounts of the purified ligands to the flow cell immobilized with SmpB (association, a 2-fold higher signal for the 50S compared with the 30S reflecting mass differences between both subunits). After 100 s of dissociation, an excess of native tmRNA (400 nM) is added. Boxed inset: superimposition, before adding tmRNA, of the curves corresponding to the dissociation of both 30S and 50S subunits from SmpB; the addition of an excess of tmRNA dissociates the 50S–SmpB complex, but not the 30S–SmpB complex. His-tagged SmpB proteins were added to either the purified (B) 50S or the (C) 30S ribosomal subunits from delta smpB-ssrA cells, and the resulting 30S–SmpB or 50S–SmpB complexes were purified by sucrose gradient centrifugation. The complexes were incubated without (lanes 1 and 5) or with an increasing amount of purified native tmRNA (lanes 2–4 and 6–8) and loaded on a 10–30% (w/v) linear sucrose gradient. For each assay, fractions containing either the ribosomal subunits (lanes 1–4) or soluble molecules including tmRNA (lanes 5–8) were pooled and the amount of SmpB and tmRNA present in the ribosomal and tmRNA soluble fractions were monitored by western blotting (top panels) and by northern hybridization (bottom panels), respectively. The molar ratio between tmRNA and SmpB pre-bound ribosomal subunits in the samples loaded on sucrose gradient is indicated.
To access the affinity of SmpB for tmRNA or ribosomal subunits after in vitro 70S dissociation, 70S ribosomes were incubated with a 10-fold molar excess of His-tagged SmpB protein and purified by a sucrose density gradient centrifugation to eliminate the unbound SmpB proteins. The dissociation of the resulting 70S-SmpB complexes at low magnesium concentration reveals the presence of SmpB on the 50S and 30S subunits (Figure 4E) confirming the binding of two SmpB molecules per 70S ribosome (13) one on the 30S subunit and the other on the 50S subunit. Moreover, SmpB still bound to ribosomal subunits at low magnesium concentration indicates that magnesium did not influence the affinity of SmpB for either the 50S or the 30S subunits. The incubation of 70S-SmpB complexes with tmRNA leads to the recruitment of tmRNA without the uncoupling of SmpB from 70S ribosome . Upon dissociation at low-magnesium concentration of purified 70S-SmpB-tmRNA complexes, SmpB was released from 50S subunit in complex with tmRNA (Figure 4C and D) whereas the other SmpB molecule remains bound to the 30S subunit (Figure 4B and D). Thus, tmRNA binds only to 70S-SmpB complexes and then displaces SmpB from the 50S subunit independently of the magnesium concentration (Figures 3A, B and 4D).
Figure 4 SmpB binds the 30S and the 50S subunits in 70S ribosomes and SmpB is released from the 50S by tmRNA upon 70S dissociation. (A) Analysis of SmpB (lanes 1 and 2), tmRNA (lanes 3 and 4) and ribosomal 16S RNA (lanes 5 and 6) levels in purified 70S-SmpB (lanes 2, 4 and 6) or 70S-SmpB-tmRNA (lanes 1, 3 and 5) complexes. Purified 70S-SmpB or 70S-SmpB-tmRNA complexes were fractionated by sucrose gradient centrifugation at a low concentration of Mg2+ ions. The 16S ribosomal (B) RNA and (C) tmRNA were detected by northern hybridization using complementary 32P-labeled DNA oligonucleotides. The presence of SmpB in sucrose fractions from (D) 70S-SmpB-tmRNA or (E) 70S-SmpB complexes was detected by western blotting using rabbit polyclonal antibody directed against a histidine-tagged SmpB.
A SmpB protein mutant as a tool to investigate trans-translation
C-tail truncated SmpB variants are inactive in trans-translation but bind tmRNA or the 70S ribosomes and support the association of tmRNA with the stalled ribosomes. Also, the contribution of the tail on trans-translation is not due to an effect on SmpB stability (8,9). A recombinant E.coli SmpB missing 16 amino acids at its C-terminal tail (delta 16-SmpB) was cloned and sequenced. The phenotype of the SmpB defective strain at 45°C cannot be complemented when expressing delta 16-SmpB from a plasmid whereas a plasmid expressing wild-type SmpB does complement the growth defect (Figure 5A).
Figure 5 An inactive C-tail truncated SmpB remains bound to the ribosomal subunits but is not associated with tmRNA in the soluble fraction. (A) The growth defect of the delta SmpB strain at 45°C is not rescued by a C-tail truncated SmpB. The growth of the delta SmpB strain with either an empty plasmid (closed circle) or a plasmid allowing the expression of the C-tail truncated SmpB (open triangle) and a plasmid containing a wild-type SmpB strain (open square), is compared. (B) The truncated protein is inactive in an in vitro trans-translation assay showing the incorporation of alanine from tmRNA to stalled ribosomal complexes containing poly(U) mRNA. (C). Expression levels of tmRNA, truncated and wild type SmpB in the S30, S100 and P100 fractions detected by northern hybridization and western blotting, respectively. (D) Fractions from either wild type or cells expressing delta 16-SmpB were fractionated by sucrose gradient centrifugation at a low concentration of Mg2+ ions. The absorbance of each fraction from wild type or delta 16-SmpB cells is shown. tmRNA was detected by northern hybridization using complementary 32P-labeled DNA oligonucleotides in either the wild type (E) or delta 16-SmpB (G) cells. The presence of SmpB (F) or of delta 16-SmpB (H) in sucrose fractions from wild type or delta SmpB crude ribosomes was detected by western blotting using rabbit polyclonal antibody directed against a histidine-tagged SmpB.
In order to access the activity of this truncated SmpB protein into an in vitro trans-translation assay, we expressed and purified a delta 16-SmpB protein tagged by six histidines to its C or N-terminal portion. These truncated proteins histidine tagged binds both ribosomal subunits in vitro, as wild-type SmpB (data not shown) but are inactive in a ‘codon-independent’ transfer of a polypeptidyl-tRNAPhe to the alanine from alanyl-tmRNAAla in the ribosome P-site in vitro (Figure 5B). Thus, delta 16-SmpB protein is unable to complement the phenotype of the SmpB defective strain at 45°C because of its inability to perform the first trans-peptidation step in the trans-translation process.
Inactive C-tail truncated SmpB are not released with tmRNA when the ribosomes are dissociated
What is the location in vivo of the SmpB protein defective in trans-translation? The inactive C-tail truncated protein, expressed from a plasmid containing its endogenous promoter, is exclusively bound to the 70S ribosomes (P100) and has an expression level slightly lower than endogenous wild-type SmpB (Figure 5C). In the presence of this inactive truncated SmpB protein, tmRNA was found associated with the 70S ribosomes (P100, delta 16-SmpB, Figure 5C). However, in contrast to endogenous SmpB, higher amounts of tmRNA are in the soluble fraction in cells expressing the truncated SmpB (S100, Figure 5C). When the 70S ribosomes from the P100 fraction of the strain that expresses the inactive SmpB variant are dissociated into the 30S and the 50S ribosomal subunits, all the tmRNA molecules are released from the ribosomes without being associated with the mutant SmpB protein (Figure 5G). The inactive protein is bound to the 30S and the 50S ribosomal subunits, in a comparable ratio, between both subunits, with that observed with the strain that does not express tmRNA (see Figure 1E). Despite not working during trans-translation, the defective SmpB protein which was able to recruit tmRNA on stalled 70S ribosome still binds the two ribosomal subunits in vivo (Figure 5H). The presence of SmpB on the 50S subunit on 70S ribosomes which are all blocked in the early stages of trans-translation suggest that SmpB bind early and transiently to the 50S subunit.
The binding of endogenous SmpB to the 50S subunit is an early step in vivo
The results presented above suggest that the binding of SmpB to the large ribosomal subunit is an early and a transient step during trans-translation. If true, blocking the early steps of trans-translation in vivo might allow the detection of endogenous SmpB on the 50S subunit. To test that, translation and trans-translation are impaired in vivo by chloramphenicol that blocks aminoacyl-tRNAs and alanyl-tmRNAAla interactions with the peptidyltransferase (PT) center (24). Chloramphenicol leads to the enrichment of ribosomes blocked on the first step of trans-translation by blocking the late translational step of tmRNA. When these ribosomes blocked at various stages of trans-translation are purified and dissociated (Figure 6A), endogenous SmpB, detected by immunoblots, is in complex with the released tmRNA (Figure 6B) and with the 30S subunit (Figure 6C). This SmpB localization is in agreement with the presence of the protein bound to tmRNA and to the 30S subunit on 70S ribosomes engaged on the late step of trans-translation in chloramphenicol-untreated cells (Figure 1A). Moreover, a significant amount of SmpB is bound to the 50S subunits (Figure 6C) which represents the SmpB binding during the early step of trans-translation. Therefore, impairing translation and trans-translation in vivo with chloramphenicol allows to visualize the early step of trans-translation and transient interaction of SmpB with the large ribosomal subunit.
Figure 6 SmpB is detected on the 50S subunit when translation and trans-translation are blocked by chloramphenicol in vivo. (A) Crude ribosomes from wild-type cells which were fractionated by sucrose gradient centrifugation at a low concentration of Mg2+ ions. (B) tmRNA was detected by northern hybridization using complementary 32P-labeled DNA oligonucleotides. (C) The presence of endogenous SmpB was detected by western blotting using rabbit polyclonal antibody directed against a histidine-tagged SmpB.
DISCUSSION
In a previous work, we showed that SmpB is located exclusively on the 70S ribosomes in vivo and such complexes can recruit alanyl-tmRNA to trigger trans-translation (13). In the present study, we demonstrate that endogenous SmpB associates with both the small and the large ribosomal subunits in vivo and in vitro. In vivo, the association of SmpB with ribosomal subunits occurs only in the context of the 70S ribosome since SmpB was not detected in free 50S and 30S subunits either in the presence or the absence of tmRNA (13). Our in vivo data reveal that the dissociation of the 70S ribosomes from wild-type cells leads to the release of all tmRNA from the ribosome, in complex with SmpB. In these ribosomes engaged in trans-translation, a small fraction of the SmpB protein remains bound to the 30S subunit after ribosome dissociation. Wild-type SmpB protein is also found associated in vivo on the 50S subunit when the 70S ribosomes are purified and dissociated into subunits either from cells lacking tmRNA or from cells treated by chloramphenicol.
The ability of SmpB to interact in vivo with the two ribosomal subunits is in agreement with the in vitro data demonstrating that SmpB binds to the 30S and the 50S subunits, both prepared to high purity. The binding between SmpB and each of the ribosomal subunits is very tight and specific, with a dissociation constant of 1.7 x 10–10 M for each one, similar to its KD for the 70S and tmRNA. Although the SmpB protein binds each ribosomal subunit with similar affinity, the domains of the protein in interaction with the 30S and the 50S might be different, in agreement with the existence of two distinct clusters of conserved amino acids at the protein surface (5). SmpB has similar affinity in vitro for the two ribosomal subunits, for the 70S ribosome and for tmRNA but are found associated exclusively to the 70S in vivo which suggest that SmpB has high affinity for 70S ribosomes stalled on a problematic mRNA and argues for an early binding of the protein onto stalled ribosomes prior to the recruitment of the tmRNA.
SmpB possesses high affinity for the 30S subunit. This is consistent with the structural homology of its tertiary structure (5,25) to that of ribosomal protein S17 from the 30S subunit (5) and especially with that of initiation factor IF1 that transiently binds the A site within the 30S (26). Our biochemical results are in agreement with structural data collected at 11.8 ? resolution in the course of a recent cryo-electron study of stalled ribosomes in complex with SmpB in the absence of tmRNA (R. Gillet, S. Kaur, W. Li, M. Hallier, B. Felden, J. Frank, manuscript submitted). In that study, SmpB was located on the 30S subunit, towards the empty decoding center. Moreover, recent in vitro data indicate that SmpB footprints nucleotides that are in the vicinity of the P-site of the small subunit (14).
The binding of SmpB to the 50S subunit is consistent with cryo-EM data obtained with a ribosome from Thermus thermophilus, stalled at the end of a short mRNA and reacted with alanyl-tmRNAala, his-tagged SmpB and EF-Tu-GTP. In this pre-accommodated ribosomal complex, SmpB was initially found to be exclusively located on the 50S subunit (12). A pre-accommodation step reconstructed at a higher resolution, however, reveals the presence of two molecules of SmpB, one interacts with the large subunit while the second is directed towards the decoding site (DS) within the 30S subunit (R. Gillet, S. Kaur, W. Li, M. Hallier, B. Felden, J. Frank, manuscript submitted). The presence of more than one SmpB protein on ribosomal complexes agrees with our previous work showing that two molecules of SmpB are required for trans-transfer events in vitro (13).
In contrast to SmpB-70S preformed complexes (Figure 4A), neither the SmpB-50S nor the SmpB-30S complexes are capable of recruiting tmRNA in vitro . When the 70S ribosomes, the 50S or the 30S subunits compete with tmRNA for binding SmpB, tmRNA can only release SmpB from the large ribosomal subunit (Figure 3). Moreover, SmpB is released from the 50S subunits with tmRNA when the 70S ribosomes containing SmpB and tmRNA are dissociated in two subunits. Thus, while SmpB presents a similar affinity for ribosomal subunits and tmRNA, the binding of tmRNA on 70S-SmpB complexes leads to a modulation of SmpB affinity for the two ribosomal subunits versus tmRNA in these ribosomes not engaged in trans-translation. It suggests that SmpB is transiently associated to the 50S subunit during trans-translation but tightly and stably bound to the 30S subunit. In the presence of tmRNA, SmpB contacts the large subunit by a close interaction with helix 69 in the 23S RNA (12). To avoid steric clashes with the P-site tRNA when the accommodation proceeds, SmpB has to move away from the 50S subunit (6). The transient and early association of SmpB to the 50S subunit agrees with its location on wild-type ribosomes performing trans-translation, in which SmpB is only detected on the small subunit (Figure 1D). Moreover, SmpB on the 50S in vivo is detected when trans-translation is voluntarily impaired by chloramphenicol, which leads to the accumulation of 70S ribosomes blocked at the early stages of trans-translation. Since the association between SmpB and ribosomal subunits or tmRNA is independent of the magnesium concentration (Figures 3 and 4), the presence of SmpB on the 50S subunit in chloramphenicol-treated cells is not because of the low level of magnesium but reflects a stage during trans-translation.
This notion that SmpB binding to the 50S is a very early and transient step during trans-translation is reinforced further by the subunits localization of a mutant SmpB protein. Indeed, when wild type SmpB protein is replaced by C-tail truncated SmpB variant that is inactive in the first step of trans-translation, the truncated protein remains bound to both ribosomal subunits in vivo upon ribosome dissociation, without being released in complex with tmRNA. Mutant protein binds both ribosomal subunits and recruits tmRNA but is unable to trigger essential conformational rearrangement(s) within the stalled ribosome that are required for its binding and release with tmRNA from the stalled ribosome. This suggests that SmpB has to move away from the 50S subunit during the accommodation of alanyl-tmRNAAla into the large subunit.
What is the fate of SmpB on the 50S? Complex formation between tmRNA and SmpB was recently monitored by chemical and enzymatic probes combined to SPR measurements, and three independent binding sites of the protein onto tmRNA were identified (7). These data suggest that during trans-translation SmpB is able to bind tmRNA at different locations, to induce key structural changes within the tmRNA scaffold required for the sliding of tmRNA through the ribosome to be rescued. The protein could re-associate onto tmRNA structure at a binding site different from the one on the tRNA domain, to participate in resuming translation on tmRNA internal tag reading frame.
We also detect some of the protein bound to the 30S upon subunit dissociation, in agreement with a stably in vitro binding of the protein to the 30S, contrasting with its labile binding on the 50S in the presence of tmRNA. tmRNA has a tRNA accepting branch that has to fit within the PT center on the 50S, but lacks an anticodon stem–loop analog that fits into the DS of the 30S during the decoding of aminoacylated tRNAs in canonical translation. Assuming that SmpB binds at or near functional sites within the ribosome, it could rationalize the competition of tmRNA for the binding of SmpB to the large subunit but not on the small subunit. During trans-translation, however, when the location of SmpB on the ribosomal subunits from wild-type cells is monitored in vivo, the majority of the protein is released with tmRNA with only a minor fraction that remains attached to the 30S. It suggests that during trans-translation, the molecules of SmpB bound to the 30S either move onto tmRNA or are released from the ribosome.
Altogether, these data can be reconciled into a working model (Figure 7). During trans-translation, SmpB modulates its affinity for the ribosomal subunits versus tmRNA. When a ribosome stalls on a problematic message, SmpB would bind the 70S ribosome in association with tmRNA or prior to recruiting tmRNA. During pre-accommodation, SmpB is still bound to the large and the small ribosomal subunits. The accommodation results in the movement of both the alanylated-tmRNAAla and the SmpB protein bound to the 50S subunit. SmpB either moves to tmRNA or is released from the 50S subunit, leading to the entrance of the tmRNA acceptor branch within the PT center. Subsequently, there is a codon-independent transfer of the peptidyl residue stalled in the P-site to the alanine from tmRNA. The molecule of SmpB bound to the 30S either binds tmRNA at different site from the tRNA-like domain or is released from the ribosome. Indeed, the molecule of SmpB bound to the small subunit has to move to allow translation resumption on tmRNA reading frame. The molecule that has been recently identified by Shpanchenko et al (2005) based on monitoring in vivo ribosome–tmRNA complexes during trans-translation could come from that bound to either the 30S or the 50S (11). This report adds several important insights concerning the location, the functional and structural roles of the molecules of SmpB during bacterial ribosome rescue. Many important questions remain to be elucidated, and among them, when and how the SmpB proteins, alone or in complex with tmRNA, identify the ribosomes that are being stalled on problematic messages.
Figure 7 Schematic illustration of the location of the SmpB molecules during trans-translation. SmpB proteins bound to both ribosomal subunits and tmRNA when a ribosome is stalled on a problematic mRNA. During the accommodation, SmpB moves from the 50S subunit on the tmRNA to allow the entrance of tmRNA acceptor branch within the PT center. The transfer of the nascent polypeptide chain to the alanine charged on tmRNA requires that SmpB moves away from the 30S subunit. One SmpB is released from the ribosome and the other remains associated with tmRNA during translation of its internal reading frame.
ACKNOWLEDGEMENTS
The authors thank Drs L. Metzinger and R. Gillet for the critical reading of the manuscript. This work was supported by the ‘Région Bretagne’ (Programme d'accueil d'équipes en émergence) to Brice Felden. Funding to pay the Open Access publication charges for this article was provided by Region Bretagne.
REFERENCES
Keiler, K.C., Waller, P.R.H., Sauer, R.T. (1996) Role of a peptide tagging system in degradation of proteins synthesized from damaged messenger RNA Science, 271, 990–993 .
Saguy, M., Gillet, R., Metzinger, L., Felden, B. (2005) tmRNA and associated ligands: a puzzling relationship Biochimie, . 87, 897–903 .
Haebel, P.W., Gutmann, S., Ban, N. (2004) Dial tm for rescue: tmRNA engages ribosomes stalled on defective mRNAs Curr. Opin. Struct. Biol, . 14, 58–65 .
Karzai, A.W., Susskind, M.M., Sauer, R.T. (1999) SmpB, a unique RNA-binding protein essential for the peptide-tagging activity of SsrA (tmRNA) EMBO J, . 18, 3793–3799 .
Dong, G., Nowakowski, J., Hoffman, D.W. (2002) Structure of small protein B: the protein component of the tmRNA-SmpB system for ribosome rescue EMBO J, . 21, 1845–1854 .
Gutmann, S., Haebel, P.W., Metzinger, L., Sutter, M., Felden, B., Ban, N. (2003) Crystal structure of the transfer-RNA domain of transfer-messenger RNA in complex with SmpB Nature, 424, 699–703 .
Metzinger, L., Hallier, M., Felden, B. (2005) Independent binding sites of small protein B onto transfer-messenger RNA during trans-translation Nucleic Acids Res, . 33, 2384–2394 .
Jacob, Y., Sharkady, S.M., Bhardwaj, K., Sanda, A., Williams, K.P. (2005) Function of the SmpB tail in transfer-messenger RNA translation revealed by a nucleus-encoded form J. Biol. Chem, . 280, 5503–5509 .
Sundermeier, T.R., Dulebohn, D.P., Cho, H.J., Karzai, A.W. (2005) A previously uncharacterized role for small protein B (SmpB) in transfer messenger RNA-mediated trans-translation Proc. Natl Acad. Sci. USA, 102, 2316–2321 .
Barends, S., Karzai, A.W., Sauer, R.T., Wower, J., Kraal, B. (2001) Simultaneous and functional binding of SmpB and EF-Tu-TP to the alanyl acceptor arm of tmRNA J. Mol. Biol, . 314, 9–21 .
Shpanchenko, O.V., Zvereva, M.I., Ivanov, P.V., Bugaeva, E.Y., Rozov, A.S., Bogdanov, A.A., Kalkum, M., Isaksson, L.A., Nierhaus, K.H., Dontsova, O.A. (2005) Stepping transfer messenger RNA through the ribosome J. Biol. Chem, . 280, 18368–18374 .
Valle, M., Gillet, R., Kaur, S., Henne, A., Ramakrishnan, V., Frank, J. (2003) Visualizing tmRNA entry into a stalled ribosome Science, 300, 127–130 .
Hallier, M., Ivanova, N., Rametti, A., Pavlov, M., Ehrenberg, M., Felden, B. (2004) Pre-binding of small protein B to a stalled ribosome triggers trans-translation J. Biol. Chem, . 279, 25978–25985 .
Ivanova, N., Pavlov, M.Y., Bouakaz, E., Ehrenberg, M., Schiavone, L.H. (2005) Mapping the interaction of SmpB with ribosomes by footprinting of ribosomal RNA Nucleic Acids Res, . 33, 3529–3539 .
Rudinger-Thirion, J., Giege, R., Felden, B. (1999) Aminoacylated tmRNA from Escherichia coli interacts with prokaryotic elongation factor Tu RNA, 5, 1–4 .
Korber, P., Stahl, J.M., Nierhaus, K.H., Bardwell, J.C. (2000) Hsp15: a ribosome-associated heat shock protein Embo J, . 19, 741–748 .
Qu, X., Yang, Z., Zhang, S., Shen, L., Dangel, A.W., Hughes, J.H., Redman, K.L., Wu, L.C., Yu, C.Y. (1998) The human DEVH-box protein Ski2w from the HLA is localized in nucleoli and ribosomes Nucleic Acids Res, . 26, 4068–4077 .
Ramirez, M., Wek, R.C., Hinnebusch, A.G. (1991) Ribosome association of GCN2 protein kinase, a translational activator of the GCN4 gene of Saccharomyces cerevisiae Mol. Cell Biol, . 11, 3027–3036 .
Shimizu, Y. and Ueda, T. (2002) The role of SmpB protein in trans-translation FEBS Lett, . 514, 74–77 .
Hart, D.J., Speight, R.E., Cooper, M.A., Sutherland, J.D., Blackburn, J.M. (1999) The salt dependence of DNA recognition by NF-kappaB p50: a detailed kinetic analysis of the effects on affinity and specificity Nucleic Acids Res, . 27, 1063–1069 .
Noll, M. and Noll, H. (1976) Structural dynamics of bacterial ribosomes. V. Magnesium-dependent dissociation of tight couples into subunits: measurements of dissociation constants and exchange rates J. Mol. Biol, . 105, 111–130 .
Ban, N., Freeborn, B., Nissen, P., Penczek, P., Grassucci, R.A., Sweet, R., Frank, J., Moore, P.B., Steitz, T.A. (1998) A 9 A resolution X-ray crystallographic map of the large ribosomal subunit Cell, 93, 1105–1115 .
Clemons, W.M., Jr, Brodersen, D.E., McCutcheon, J.P., May, J.L., Carter, A.P., Morgan-Warren, R.J., Wimberly, B.T., Ramakrishnan, V. (2001) Crystal structure of the 30 S ribosomal subunit from Thermus thermophilus: purification, crystallization and structure determination J. Mol. Biol, . 310, 827–843 .
Schlunzen, F., Zarivach, R., Harms, J., Bashan, A., Tocilj, A., Albrecht, R., Yonath, A., Franceschi, F. (2001) Structural basis for the interaction of antibiotics with the peptidyl transferase centre in eubacteria Nature, 413, 814–821 .
Someya, T., Nameki, N., Hosoi, H., Suzuki, S., Hatanaka, H., Fujii, M., Terada, T., Shirouzu, M., Inoue, Y., Shibata, T., et al. (2003) Solution structure of a tmRNA-binding protein, SmpB, from Thermus thermophilus FEBS Lett, . 535, 94–100 .
Carter, A.P., Clemons, W.M., Jr, Brodersen, D.E., Morgan-Warren, R.J., Hartsch, T., Wimberly, B.T., Ramakrishnan, V. (2001) Crystal structure of an initiation factor bound to the 30S ribosomal subunit Science, 291, 498–501 .(Marc Hallier, Jessie Desreac and Brice F)