Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

The SsrA–SmpB system for protein tagging, directed degradation and ribosome rescue

Abstract

Bacteria contain a remarkable RNA molecule — known alternatively as SsrA RNA, tmRNA, or 10Sa RNA — that acts both as a tRNA and as an mRNA to direct the modification of proteins whose biosynthesis has stalled or has been interrupted. These incomplete proteins are marked for degradation by cotranslational addition of peptide tags to their C-termini in a reaction that is mediated by ribosome-bound SsrA RNA and an associated protein factor, SmpB. This system plays a key role in intracellular protein quality control and also provides a mechanism to clear jammed or obstructed ribosomes. Here the structural, functional and phylogenetic properties of this unique RNA and its associated factors are reviewed, and the intracellular proteases that act to degrade the proteins tagged by this system are also discussed.

Main

SsrA RNA is highly conserved throughout eubacteria and was discovered in Escherichia coli when a 10S RNA fraction was resolved into two species, the 10Sa or SsrA (small stable RNA A) molecule and the 10Sb RNA, which is the catalytic subunit of ribonuclease P1,2. Sequence similarities between parts of SsrA and tRNAs were initially recognized based on the ssrA gene sequence from Mycobacterium tuberculosis3. Subsequent determination of the E. coli and Bacillus subtilis sequences demonstrated convincing homologies with alanyl tRNA, including the presence of an acceptor stem and T-arm4,5. The most compelling evidence for tRNA-like functionality, however, was provided by the demonstration that purified SsrA RNA could be charged with alanine by alanyl-tRNA synthetase4,5.

The ability of SsrA RNA to function as an mRNA was discovered through an elegant protein-chemical analysis of a set of aberrant translation products that accumulated in inclusion bodies during expression of the murine interleukin-6 (IL-6) protein in E. coli6. Partial IL-6 proteins, truncated to differing extents, were all found to contain the C-terminal sequence AANDENYALAA. This sequence was not IL-6 derived. All amino acids except the first alanine of this 11-residue tag, however, were encoded by an open reading frame in SsrA RNA. Moreover, no modification of IL-6 sequences was observed in cells bearing a defective ssrA gene. At this point, both tRNA-like and mRNA-like properties of SsrA had been established but the biological function of SsrA and the mechanism of C-terminal modification were still obscure.

A potential biological role for SsrA was suggested7 by the observation that the C-terminal pentapeptide of the tag sequence (YALAA) was similar to other hydrophobic C-terminal sequences (for example, WVAAA) known to cause degradation of bacterial proteins8. Indeed, model proteins with a DNA-encoded AANDENYALAA sequence placed at their C-terminus were rapidly degraded, but proteins in which the tag was changed to AANDENYALDD were not degraded7. These results suggested strongly that the SsrA system functioned to mark certain proteins for proteolytic destruction. Combining the known tRNA and mRNA activities of SsrA with the possibility that this system might function to ensure the degradation of proteins produced by incomplete translation led to the tmRNA model described below.

tmRNA model

In bacteria, the signals required for initiation of translation are located near the 5′ end of the mRNA, and an in-frame UAG, UAA, or UGA termination codon is required to recruit the protein factors needed for release of the nascent polypeptide chain from the ribosome. This need for active polypeptide release creates a potential problem. How do ribosomes deal with mRNA fragments — generated by nuclease cleavage or incomplete transcription — that have no termination codons in the protein reading frame? Translation of such mRNA fragments should initiate normally and continue until the ribosome reaches the 3′ end. At this point, neither continuation of translation nor normal protein chain release would be viable options and the ribosome should stall, leaving the partially synthesized protein chain attached to the P-site tRNA. This causes two problems. The first is clearing the ribosome to permit translation of new mRNAs. The second is dealing with the potentially deleterious consequences of releasing a partial protein, possibly with poor solubility properties or unregulated activities, into the cell.

The tmRNA model7 (Fig. 1) postulates that alanine-charged SsrA RNA recognizes stalled ribosomes, binds like a tRNA to the ribosomal A site, and then donates the charged alanine to the nascent polypeptide chain in a standard transpeptidation reaction. The stalled mRNA is then replaced by SsrA RNA, which acts as a surrogate message to direct translation of the degradation tag. Translation terminates normally at a stop codon that follows the SsrA peptide reading frame. The final translation product of this process carries the 11-residue degradation tag at its C-terminus and hence becomes a substrate for intracellular proteases. In theory, this system should provide an effective quality control mechanism to deal with incomplete and potentially harmful protein fragments and also a way to rescue ribosomes stalled on damaged mRNAs to permit further rounds of protein synthesis.

Figure 1: The tmRNA model for SsrA mediated tagging of proteins7.
figure 1

A ribosome stalls on an incomplete or untranslatable message, leading to the recruitment of aminoacylated-SsrA RNA to the ribosomal A site and transfer of the nascent chain to the alanine-charged tRNA domain of SsrA. A message switching event then replaces the faulty mRNA by an open reading frame (shown in magenta) within SsrA, which is translated until a stop codon is reached and the tagged protein is released to be degraded by proteases.

Considerable evidence supports the general tmRNA model for SsrA RNA function. For example, model proteins translated from mRNAs without stop codons are modified by C-terminal addition of the SsrA-encoded peptide tag and are rapidly degraded7. If the final residues of the SsrA peptide reading frame are mutated to encode DD instead of AA, then incorporation of the mutant ANDENYALDD tag into target proteins can be detected immunochemically and by mass spectrometry9,10. (Such DD-tagged proteins are relatively stable to degradation and far easier to detect than AA-tagged proteins.) Mutation of the key G:U wobble base pair in the proposed tRNA-like acceptor stem of SsrA decreases charging with alanine and tagging in vitro and also decreases activity in vivo4,11,12. Moreover, wild-type SsrA copurifies with 70S ribosomes but charging mutants of SsrA do not11,13,14. Finally, bacterial translation extracts supplemented with SsrA-DD RNA synthesized in vitro display DD-tagging of a protein synthesized from an mRNA without stop codons9, and translation systems programmed with poly(U) incorporate alanine and other SsrA tag amino acids even though this RNA codes only for phenylalanine11. When the charging determinants of SsrA are changed to those for histidyl-tRNA, histidine is also incorporated during poly(U) directed translation15.

SsrA synthesis, processing, and modification

SsrA is encoded by a single gene in E. coli and transcription appears to be controlled by a σ70-type promoter and a rho-independent transcriptional terminator4,16. In E. coli, the primary SsrA transcript is 457 nucleotides and processing of this precursor RNA is required to generate the mature molecule of 363 bases4. Processing is critical for generating both the mature 5′ and 3′ ends of the SsrA acceptor stem, which in turn are required for charging. Seven nucleotides are removed from the 5′ end of the E. coli SsrA transcript by RNase P cleavage to generate the mature 5′ terminus4. Processing of the 3′ end of this transcript seems to involve multiple RNases17,18,19,20. In E. coli, an initial endonucleolytic cleavage appears to be made by the RNase III and/or RNase E enzymes with subsequent removal of additional 3′ bases by the RNase T and RNase PH exonucleases. Mature SsrA, like all tRNAs, has a CCA sequence at its 3′ end. In B. subtilis and some other bacteria, this CCA trinucleotide is not encoded in the gene but is added to the processed SsrA transcript by tRNA nucleotidyltransferase5,21. E. coli SsrA RNA also contains pseudouridine and 5-methyluridine at positions analogous to those where the same base modifications are found in tRNAs1,22.

SsrA structure

Our present knowledge of SsrA RNA structure is based almost exclusively on phylogenetic comparisons and chemical modification experiments23,24. The intact RNA molecule is too large for NMR studies and crystals have not yet been obtained. Fig. 2 shows a predicted secondary structure for SsrA RNA21,23,24,25. The 5′ and 3′ ends of the RNA form an alanyl-tRNA-like domain, including an acceptor stem, a T-arm, and a D-loop with no stem. Instead of an anticodon stem-loop, a long disrupted stem connects the tRNA-like domain to the rest of the SsrA molecule. Upon leaving the 5′ portion of this stem, there is a pseudoknot (Ψ1), the peptide reading frame followed by a stop codon in a hairpin loop, and then a series of three additional pseudoknots (Ψ2–Ψ4). Mutational analyses and NMR studies of an SsrA fragment support the existence and functional importance of the Ψ1 pseudoknot26.

Figure 2: Predicted secondary structure of E. coli SsrA RNA21,23,24,25,26,27,28,29,30,31,32,33,34,35, highlighting its tRNA and mRNA properties including the partial tRNAala-like domain (the secondary structure of tRNAala is shown for comparison) and the degradation tag coding sequence.
figure 2

The four pseudoknots are labeled Ψ1–Ψ4. Figure adapted from the tmRNA website21.

Associated protein factors

All known biological activities of SsrA RNA require SmpB (small protein B), a protein of 160 residues that binds specifically and with high affinity to SsrA RNA27. In E. coli, SmpB is not required for SsrA expression, processing, or charging with alanine by alanyl-tRNA synthetase in vitro. Nevertheless, SsrA does not associate stably with 70S ribosomes in cells lacking SmpB. Therefore, formation of the SsrA–SmpB complex appears to be critical for at least one step after aminoacylation of SsrA RNA with alanine but prior to the transpeptidation reaction that couples this alanine to the nascent chain. Following purification of 70S ribosomes and dissociation of the 30S and 50S subunits, SmpB remains associated with SsrA (A.W.K., unpublished observation), suggesting that SmpB may also be required for later steps in the tmRNA model. SsrA does not appear to have a traditional anticodon stem-loop or to require codon–anticodon interactions for ribosome binding. It is conceivable that SmpB or other unidentified factors might act as an anticodon-arm mimic and provide additional contacts required for high affinity ribosome binding. There are precedents for a protein mimic of a tRNA anticodon stem-loop in elongation factor G28,29 and ribosome recycling factor30. Alternatively, SmpB might mediate recognition of other ribosomal elements required for SsrA-mediated function, facilitate some reactions directly, and/or serve a structural role in maintaining SsrA RNA in an active three-dimensional conformation. At present, there are no known SmpB functions that are independent of its role in supporting SsrA activity.

From a structural perspective, relatively little is known about SmpB. The isolated protein is monomeric and has a circular dichroism spectrum expected for a protein that is largely composed of β-sheet27. Sequence searches reveal SmpB orthologs in all complete and many partial bacterial genomes but do not result in statistically significant matches with other proteins. Threading searches do, however, suggest an intriguing potential structural homology between E. coli SmpB and c-H-Ras p21, a GTP-binding protein (A.W.K., unpublished observation).

SsrA RNA is recognized and charged by alanyl-tRNA synthetase4 and the resulting aminoacylated SsrA forms a ternary complex with elongation factor Tu (EF-Tu) and GTP in vitro31,32. Neither the kinetics of the charging reaction nor EF-Tu–GTP binding is affected by binding of SmpB to SsrA RNA, suggesting that the binding site for SmpB does not overlap with those of alanyl-tRNA synthetase or EF-Tu. However, none of these binding sites have been mapped. Complex formation with EF-Tu–GTP protects the labile ester linkage of aminoacylated SsrA31 and, by analogy with tRNAs, this interaction is also presumed to be important for delivery of SsrA to ribosomes. When tRNA–EF-Tu–GTP complexes bind to the ribosomal A site, GTP hydrolysis is stimulated by cognate codon–anticodon interactions, leading to a conformational change and dissociation of EF-Tu–GDP. This allows the ribosome-bound tRNA to participate in the transpeptidation reaction. Complexes with noncognate tRNAs hydrolyze GTP more than 1,000-fold more slowly and these complexes generally dissociate before GTP is hydrolyzed and a translational error is made33. Because SsrA does not appear to have a traditional anticodon stem-loop, it is unclear what determines the rate of GTP hydrolysis and release by EF-Tu and whether this plays any role in assuring the proper recognition of stalled ribosomes and/or the disruption of interactions with actively translating ribosomes.

SsrA function

There are important unanswered questions about mechanisms at almost every step of the tmRNA model. Are messages without in-frame stop codons produced in the cell by premature termination of normal transcription, by completed transcription of a gene containing a transcriptional terminator prior to any in-frame stop codons, and/or by mRNA degradation? It is known that mRNAs in E. coli have an average half-life of 2–3 min, and endoribonucleases or 3′→5′ exoribonucleases probably generate mRNA fragments continuously34. Run-off translation from mRNAs lacking stop codons results in tagging at protein positions corresponding to the final encoded residue, the penultimate residue, the antepenultimate residue, and so forth10, which could be explained by message trimming by exonucleases.

There may also be additional mRNA determinants that lead to tagging in the context of full-length mRNAs. For example, some ribosomes may reach the 3′ end of mRNAs because normal stop codons are translated by suppressor tRNAs. In addition, ribosome stalling at rare codons has been shown to cause tagging when the cognate tRNA is scarce9. It is unclear, in this instance, whether ribosome stalling by itself results in tagging at an internal mRNA site or if stalling results in ribosome directed cleavage of the mRNA at the rare codon which, in turn, results in tagging. If prolonged ribosome stalling is sufficient to recruit SsrA, then tagging may also occur when ribosomes stall at abasic or other damaged sites in mRNAs, pause because of stable mRNA secondary structures, or stop because charged tRNA is scarce. A variety of endogenous proteins are tagged by the SsrA–SmpB system in E. coli9, and identifying the sites of protein truncation and tag addition in these molecules may provide additional information about the features of mRNAs that result in tagging.

How does the SsrA–SmpB–EF-Tu complex recognize impaired ribosomes needing rescue without interfering with ribosomes engaged in normal translation? If all SsrA mediated tagging ultimately occurs on ribosomes stalled at the 3′ end of mRNAs, then the absence of mRNA in the A site or 3′ untranslated mRNA could be the major recognition determinant either directly or because steric clashes are eliminated (Fig. 3a). This simple model is appealing because in normal translation, the A site is always occupied and untranslated 3′ mRNA segments are always present, potentially explaining why SsrA does not interfere with routine translation. Alternatively, the ribosome may have an internal clock that detects prolonged stalling and signals a conformational change, enabling entry of the SsrA complex and perhaps loosening the hold of the ribosome on the mRNA (Fig. 3b). A clock of this type could also play a role in other translational events that require stalling for frame shifting or hopping on the same mRNA35,36.

Figure 3: Potential mechanisms for SsrA recruitment.
figure 3

a, SsrA interacts with ribosomes stalled at the extreme 3′ end of a mRNA either by directly recognizing the empty mRNA track in the A site or because untranslated 3′ mRNA, which might occlude SsrA, is absent. b, Prolonged ribosome stalling at an untranslatable sequence is detected by a ribosomal clock, leading to a conformational change in the ribosome. This structural change allows entry of SsrA RNA and could also facilitate the departure of the original mRNA.

The structure of the ribosome shows a very tight fit of the P-site tRNA between the 30S and 50S subunits37. How does SsrA RNA, which is roughly four times as large as a normal tRNA, fit into the same sites on the ribosome and participate in transpeptidation and message switching? The SsrA acceptor stem presumably interacts with the peptidyl transferase center just like a tRNA during transpeptidation, but the nature of subsequent events is unclear. It is not known, for example, whether SsrA itself undergoes translocation to the ribosomal P site, how the SsrA message is engaged, or how the damaged mRNA is released. The way in which the ribosome identifies the proper resume codon to start translation of the SsrA peptide-coding region must also be novel because initiator fMet-tRNA is not involved. In addition, because the first alanine added by SsrA is not coded in either the departing message or in SsrA, there must be some discontinuity in the translation process during the mRNA to SsrA message switch. Mutagenesis experiments and phylogenetic comparisons show that the identity of the resume codon can be changed but also reveal that certain bases in this region are highly conserved and hence are presumably required for setting the reading frame10. Insertion experiments also suggest that elements upstream of the tag sequence are more important for determining the reading frame than downstream elements. Once the reading frame is determined, is SsrA translated and released like a normal mRNA or do portions of SsrA have to be actively disengaged from the ribosome? Is the tagged polypeptide released normally or is there a special mechanism that ensures immediate degradation? Proteins tagged cotranslationally by the SsrA system seem to degraded more rapidly than proteins with DNA-encoded SsrA tags, suggesting that there may be additional factors recruited by SsrA–SmpB that modulate degradation during the tagging process38.

Degradation of SsrA-tagged proteins

The proteases that degrade SsrA-tagged proteins are an important part of the SsrA quality control system. Three different E. coli proteases, ClpXP, ClpAP, and FtsH (HflB), execute energy dependent degradation of SsrA tagged proteins in the bacterial cytoplasm as well as in purified systems38,39. ClpXP and ClpAP are multicomponent enzymes40,41, in which ClpP, a tetradecameric serine protease, forms stacked heptameric rings that enclose a chamber with a diameter of 50 Å (ref. 42;Fig. 4a). The crystal structure of ClpP42 reveals that the active sites are arrayed on the interior surface of this chamber with access controlled by axial pores too small to allow passage of native proteins. ClpX and ClpA are hexameric ATPases, which interact with ClpP to mediate substrate binding and regulate protease activity40,41. Although structures for neither ClpX nor ClpA are known, sequence homology suggests that the ATPase domains of both proteins would resemble the D2 domain of N-ethylmaleimide sensitive fusion protein (NSF)43. Indeed, the recent crystal structure of a closely related ATPase, HslU (ClpY), confirms this structural homology44 (Fig. 4a). ClpA and ClpX belong to the Clp/Hsp100 family of ATPases. FtsH is a membrane-bound zinc protease45 that belongs to the AAA superfamily of ATPases, a group distantly related to the Clp/Hsp100 proteins43. FtsH may degrade SsrA tagged membrane proteins as well as some cytoplasmic proteins38,39.

Figure 4: Proteases that degrade SsrA tagged proteins.
figure 4

a, Center: cartoon representation of ClpXP or ClpAP complexes containing a hexamer of ClpX or ClpA ATPase (blue) and the tetradecamer of the ClpP peptidase (green). Top: two views of those portions of the hexameric HslU (ClpY) ATPase44 that share sequence homology with ClpX and ClpA43. Bottom: two views of the structure of the ClpP peptidase42. The left view emphasizes the small axial pore that controls entrance to the proteolytic chamber. The right view is rotated by 90° and several subunits have been removed to allow the central chamber and several active site catalytic triads (magenta) to be seen. b, Stepwise model for binding, catalyzed denaturation, translocation, and degradation of a SsrA tagged substrate by the ClpXP complex. c, Model for binding, spontaneous denaturation, and degradation of an SsrA tagged substrate by the Tsp protease.

The SsrA peptide tag does not affect the structure or thermodynamic stability of attached proteins38, suggesting that it causes degradation simply by providing a recognition site for protease binding. Indeed, ClpX and ClpA have distinct domains of 100 residues, called sensor and substrate discrimination (SSD) domains, that bind specifically to SsrA tagged proteins and certain other substrates46,47. SsrA tagged proteins with dramatically different structures and stabilities also bind to ClpX with roughly the same interaction constants, suggesting that the peptide tag is the principal determinant of binding to these enzymes67. Following binding to ClpX or ClpA, SsrA tagged native proteins are actively denatured and translocated into ClpP for degradation48,49 (Fig. 4b; ref. 67). This ensures that SsrA tagged proteins in the bacterial cytoplasm can be efficiently degraded irrespective of their intrinsic stabilities. ClpX and ClpA accelerate the rate of denaturation of green fluorescent protein bearing a C-terminal SsrA tag more than 106-fold49,67 (unpublished observations), and presumably actively catalyze denaturation by exerting force on bound native substrates through conformational changes that accompany ATP binding or hydrolysis. ClpXP and ClpAP are also highly processive enzymes40, and thus SsrA-tagged cytoplasmic proteins can be degraded to small peptides.

SsrA tagged proteins with signal sequences are exported to the periplasmic compartment of E. coli where they are degraded by the Tsp (Prc) protease7. Tsp is monomeric, ATP-independent (ATP is not generated in the periplasm), and bears no sequence homology to any of the cytoplasmic proteases50. A region of Tsp with homology to mammalian PDZ domains51 mediates recognition of SsrA tagged substrates (N. Walsh and R.T.S., unpublished observation). This fact makes sense as SsrA tagged proteins are modified at their C-termini and PDZ domains are specialized for C-terminal peptide binding51. Tsp is believed to bind the C-terminal tails of its substrates and to wait until spontaneous protein unfolding exposes the polypeptide chain to the protease active site52 (Fig. 4c). Although this mechanism would suffice to degrade proteins with modest stabilities, hyperstable SsrA tagged substrates should be degraded very slowly and tagged proteins might not be degraded to small peptide fragments. Several other periplasmic proteases also have PDZ-like domains (DegP, DegQ, and DegS) and might participate in degradation of some SsrA tagged substrates. In fact, the PDZ-like domain of DegS binds to the SsrA peptide tag with submicromolar affinity (N. Walsh and R.T.S., unpublished observation), and overexpression of DegQ and DegS can complement tsp defective phenotypes53. Some degradation of SsrA tagged proteins has also been observed in tsp strains10.

The fact that proteases in different intracellular compartments can degrade SsrA tagged proteins obviously increases the reach of this surveillance system by ensuring that tagged proteins can be proteolyzed irrespective of their subcellular localization. In any given cellular compartment, it is also possible that having multiple proteases degrade SsrA tagged substrates increases the overall efficiency of this quality control system or, in some cases, allows degradation of specific classes of tagged substrates.

Phylogenetic insights

SsrA and SmpB are found in all complete bacterial genomes (Table 1) and have also been identified in many genomes that have only been partially sequenced. Alignments of many SsrA and SmpB sequences can be found at the tmRNA website21 (http://www.indiana.edu/~tmrna/) and in the COG database54 (http://www3.ncbi.nlm.nih.gov/COG/aln/COG0691.aln), respectively. The α-proteobacteria Caulobacter crescentus and Rickettsia prowazekii contain smpB genes but recognizable ssrA genes were not initially identified21,55. The presence of SmpB in these species strongly suggests, however, that they also contain SsrA. Indeed, recent studies reveal the presence of unusual two-piece SsrA molecules in both C. crescentus and R. prowazekii68. The SsrA–SmpB system seems, therefore, to be a universal feature of eubacteria, and its preservation even in a bacterium like Mycoplasma genitalium, which has fewer than 500 genes, argues strongly for an important biological role. The most highly conserved features of SsrA involve the alanyl-tRNA-like 5′ and 3′ ends, some properties of the degradation tag sequence, and, for the most part, the overall secondary structure21.

Table 1 Phylogenetic distribution in complete eubacterial genomes of RNAs and proteins associated with the SsrA–SmpB system1

There is no evidence for SsrA or SmpB in the archaeabacteria or in eukaryotes, with the exception of some organelles. The separation of nuclear transcription and cytoplasmic translation in eukaryotes may permit other mechanisms to identify and eliminate damaged mRNAs before they leave the nucleus. Furthermore, efficient translation of most eukaryotic mRNAs requires both 5′ and 3′ information56, and thus mRNAs that are cleaved after entering the cytoplasm may never be translated. The SsrA–SmpB system is similar to the eukaryotic ubiquitin system in using degradation tags to target proteins to specific proteases. The ClpXP and ClpAP complexes even share some architectural similarities with the eukaryotic 26S proteasome57. The SsrA and ubiquitin systems are very different, however, in several important ways. Notably, the SsrA system acts cotranslationally to ensure the degradation of incomplete protein chains, whereas the ubiquitin system acts post-translationally to target intact proteins for proteolytic destruction.

The SsrA tag sequences are reasonably well conserved in different bacteria (Table 1). The C-terminal residue is always alanine. Moreover, the C-terminal pentapeptide is generally quite hydrophobic and is usually preceded by a cluster of charged or polar residues10. The evolutionary conservation of the general properties of the tag sequences indicates that the tagging function of SsrA together with the associated degradation of tagged proteins must be biologically important. Of the proteases known to degrade SsrA tagged proteins in E. coli, ClpXP, ClpAP, and Tsp are distributed reasonably widely but only FtsH is present in all of the completely sequenced bacterial species (Table 1). It seems likely, therefore, that FtsH is responsible for degradation of SsrA tagged proteins in species such as Mycoplasma, which lack the other proteases. Interestingly, in these bacteria the tag sequences also have distinct sequence features (for example, a penultimate Phe in place of Ala), suggesting that they may be optimized for interactions with FtsH.

EF-Tu and alanyl-tRNA synthetase are present in all bacteria and obviously play essential roles in normal translation. If SsrA originated at the same ancient time as tRNAs and the translational machinery, then alanine tRNA synthetase would have evolved under the constraint of interacting with both tRNAala and SsrA RNA. This may explain, in part, why the charging determinants of this enzyme are both very simple and different from many other synthetase·tRNA pairs15,58. Some endoribonucleases, such as RNase E, which help process the ssrA transcript in E. coli, appear to be missing in many bacteria (Table 1), suggesting that the mechanism of processing is species dependent.

Physiological roles of SsrA

The SsrA–SmpB system is not required for growth of E. coli but is essential in Neisseria gonorrhoeae, M. genitalium, and M. pneumoniae59,60. Why is this system required in some bacteria and not others? One possibility is that species like E. coli have another, redundant system that can serve the same general function. In fact, E. coli can release proteins synthesized from mRNAs without in-frame stop codons by an ssrA independent pathway7. Alternatively, processes such as clearing stalled ribosomes and protein quality control may be more critical in some bacteria or under certain adverse conditions because errors leading to ribosome stalling are far more frequent. The latter possibility could, for example, explain why the SsrA–SmpB system is dispensable for normal growth of Salmonella typhimurium but is required for the survival of this bacterium within macrophages61,62. SsrA defective E. coli do display phenotypes including slow growth at high temperature4, reduced motility4, slow recovery from carbon starvation63, an increase in the activity of an unidentified protease64, failure to plate hybrid λimmP22 phage65 (λ phage with the immunity region of P22 phage), an inability to induce temperature-sensitive phage Mu lysogens27, and increases in the activities of several different DNA binding proteins66. Interestingly, some of these phenotypes can be largely complemented by SsrA mutants with mutant peptide tags that are poorly recognized by proteases, suggesting that ribosome clearance may be more important than degradation for these defects12.

How does elimination of the SsrA–SmpB system result in these cellular phenotypes? In one case, there is an appealing model. Recent studies suggest that Lac repressor binds to an operator site within the C-terminal coding region of its gene, causing premature termination of transcription, SsrA tagging of the truncated protein, and degradation of this tagged protein fragment (T. Abo & H. Aiba, pers. comm.). In the absence of SsrA, these repressor fragments would not be tagged and degraded, and could, therefore, contribute to increased DNA binding activity. This is the first instance in which a specific cellular phenotype — here, increased Lac repressor activity in SsrA defective E. coli — has been mechanistically linked via a specific mRNA and protein to the tmRNA model. SsrA RNA was initially proposed to inhibit DNA binding proteins such as Lac repressor by direct complex formation66 but this model fails to explain why SmpB is required for SsrA activity27. Unfortunately, specific protein or mRNA targets for tagging, degradation, or ribosome clearance have not been identified for most SsrA phenotypes.

Summary

The SsrA–SmpB system rescues bacterial ribosomes on which synthesis of incomplete proteins has stalled and tags these proteins for subsequent degradation. ATP dependent proteases recognize, denature, and degrade proteins tagged by this system in the cytoplasm and ATP independent proteases mediate degradation in the periplasm. Although general features have been suggested for the processes of cotranslational tagging and degradation, relatively little is known about the detailed structures or mechanisms of the molecular machines that mediate these reactions.

References

  1. Ray, B.K. & Apirion, D. Characterization of 10S RNA: a new stable RNA molecule from Escherichia coli. Mol. Gen. Genet. 174, 25–32 (1979).

    CAS  PubMed  Article  Google Scholar 

  2. Subbarao, M.N. & Apirion, D. A precursor for a small stable RNA (10Sa RNA) of Escherichia coli. Mol. Gen. Genet. 217, 499–504 (1989).

    CAS  PubMed  Article  Google Scholar 

  3. Tyagi, J.S. & Kinger, A.K. Identification of the 10Sa RNA structural gene of Mycobacterium tuberculosis. Nucleic Acids Res. 20, 138 (1992).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  4. Komine, Y., Kitabatake, M., Yokogawa, T., Nishikawa, K. & Inokuchi, H. A tRNA-like structure is present in 10Sa RNA, a small stable RNA from Escherichia coli. Proc. Natl Acad. Sci. USA 91, 9223–9227 (1994).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  5. Ushida, C., Himeno, H., Watanabe, T. & Muto, A. tRNA-like structures in 10Sa RNAs of Mycoplasma capricolum and Bacillus subtilis. Nucleic Acids Res. 22, 3392–3396 (1994).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  6. Tu, G.F., Reid, G.E., Zhang, J.G., Moritz, R.L. & Simpson, R.J. C-terminal extension of truncated recombinant proteins in Escherichia coli with a 10Sa RNA decapeptide. J. Biol. Chem. 270, 9322–9326 (1995).

    CAS  PubMed  Article  Google Scholar 

  7. Keiler, K.C., Waller, P.R. & Sauer, R.T. Role of a peptide tagging system in degradation of proteins synthesized from damaged messenger RNA. Science 271, 990–993 (1996).

    CAS  PubMed  Article  Google Scholar 

  8. Parsell, D.A., Silber, K.R. & Sauer, R.T. Carboxy-terminal determinants of intracellular protein degradation. Genes Dev. 4, 277–286 (1990).

    CAS  PubMed  Article  Google Scholar 

  9. Roche, E.D. & Sauer, R.T. SsrA-mediated peptide tagging caused by rare codons and tRNA scarcity. EMBO J. 18, 4579–4589 (1999).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  10. Williams, K.P., Martindale, K.A. & Bartel, D.P. Resuming translation on tmRNA: a unique mode of determining a reading frame. EMBO J. 18, 5423–5433 (1999).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  11. Himeno, H. et al. In vitro trans translation mediated by alanine-charged 10Sa RNA. J. Mol. Biol. 268, 803–808 (1997).

    CAS  PubMed  Article  Google Scholar 

  12. Withey, J. & Friedman, D. Analysis of the role of trans-translation in the requirement of tmRNA for lambda-immP22 growth in Escherichia coli. J. Bacteriol. 181, 2148–2157 (1999).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  13. Komine, Y., Kitabatake, M. & Inokuchi, H. 10Sa RNA is associated with 70S ribosome particles in Escherichia coli. J. Biol. Chem. (Tokyo) 119, 463–467 (1996).

    CAS  Article  Google Scholar 

  14. Tadaki, T., Fukushima, M., Ushida, C., Himeno, H. & Muto, A. Interaction of 10Sa RNA with ribosomes in Escherichia coli. FEBS Lett. 399, 223–226 (1996).

    CAS  PubMed  Article  Google Scholar 

  15. Nameki, N., Tadaki, T., Muto, A. & Himeno, H. Amino acid acceptor identity switch of Escherichia coli tmRNA from alanine to histidine in vitro. J. Mol. Biol. 289, 1–7 (1999).

    CAS  PubMed  Article  Google Scholar 

  16. Oh, B.K., Chauhan, A.K., Isono, K. & Apirion, D. Location of a gene (ssrA) for a small, stable RNA (10Sa RNA) in the Escherichia coli chromosome. J. Bacteriol. 172, 4708–4709 (1990).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  17. Srivastava, R.K., Miczak, A. & Apirion, D. Maturation of precursor 10Sa RNA in Escherichia coli is a two-step process: the first reaction is catalyzed by RNase III in presence of Mn2+. Biochimie 72, 791–802 (1990).

    CAS  PubMed  Article  Google Scholar 

  18. Srivastava, R.A., Srivastava, N. & Apirion, D. Characterization of the RNA processing enzyme RNase III from wild type and overexpressing Escherichia coli cells in processing natural RNA substrates. Int. J. Biochem. 24, 737–749 (1992).

    CAS  PubMed  Article  Google Scholar 

  19. Li, Z., Pandit, S. & Deutscher, M.P., 3′ exoribonucleolytic trimming is a common feature of the maturation of small, stable RNAs in Escherichia coli. Proc. Natl Acad. Sci. USA 95, 2856–2861 (1998).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  20. Lin-Chao, S., Wei, C.L. & Lin, Y.T., RNase E is required for the maturation of ssrA RNA and normal ssrA RNA peptide-tagging activity. Proc. Natl Acad. Sci. USA 96, 12406–12411 (1999).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  21. Williams, K.P. The tmRNA Website. Nucleic Acids Res. 28, 168–161 (2000).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  22. Felden, B., et al. Presence and location of modified nucleotides in Escherichia coli tmRNA: structural mimicry with tRNA acceptor branches. EMBO J. 17, 3188–3196 (1998).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  23. Williams, K.P. & Bartel, D.P. Phylogenetic analysis of tmRNA secondary structure. RNA 2, 1306–1310 (1996).

    CAS  PubMed  PubMed Central  Google Scholar 

  24. Felden, B., et al. Probing the structure of the Escherichia coli 10Sa RNA (tmRNA). RNA 3, 89–103 (1997).

    CAS  PubMed  PubMed Central  Google Scholar 

  25. Zwieb, C. & Wower, J. tmRDB (tmRNA database). Nucleic Acids Res. 28, 169–170 (2000).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  26. Nameki, N., Chattopadhyay, P., Himeno, H., Muto, A. & Kawai, G. An NMR and mutational analysis of an RNA pseudoknot of Escherichia coli tmRNA involved in trans-translation. Nucleic Acids Res. 27, 3667–3675 (1999).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  27. Karzai, A.W., Susskind, M.M. & Sauer, R.T. SmpB, a unique RNA-binding protein essential for the peptide-tagging activity of SsrA (tmRNA). EMBO J. 18, 3793–3799 (1999)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  28. Aevarsson, A., et al. Three-dimensional structure of the ribosomal translocase: elongation factor G from Thermus thermophilus. EMBO J. 13, 3669–3677 (1994).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  29. Nyborg, J., et al. Macromolecular mimicry in protein biosynthesis. Fold. Des. 2, S7–11 (1997).

    CAS  PubMed  Article  Google Scholar 

  30. Selmer, M., Al-Karadaghi, S., Hirokawa, G., Kaji, A. & Liljas, A. Crystal structure of Thermotoga maritima ribosome recycling factor: a tRNA mimic. Science 286, 2349–2352 (1999).

    CAS  PubMed  Article  Google Scholar 

  31. Rudinger-Thirion, J., Giege, R. & Felden, B. Aminoacylated tmRNA from Escherichia coli interacts with prokaryotic elongation factor Tu. RNA 5, 989–992 (1999).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  32. Barends, S., Wower, J. & Kraal, B. Kinetic parameters for tmRNA binding to alanyl-tRNA synthetase and elongation factor Tu from Escherichia coli. Biochemistry 39, 2652–2658 (2000).

    CAS  PubMed  Article  Google Scholar 

  33. Rodnina, M.V., Pape, T., Fricke, R. & Wintermeyer, W. Elongation factor Tu, a GTPase triggered by codon recognition on the ribosome: mechanism and GTP consumption. Biochem. Cell Biol. 73, 1221–1227 (1995).

    CAS  PubMed  Article  Google Scholar 

  34. Ehretsmann, C.P., Carpousis, A.J. & Krisch, H.M. mRNA degradation in procaryotes. FASEB J. 6, 3186–3192 (1992).

    CAS  PubMed  Article  Google Scholar 

  35. Farabaugh, P. J. Programmed translational frameshifting. Microbiol. Rev. 60, 103–134 (1996).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  36. Gesteland, R.F. & Atkins, J.F. Recoding: dynamic reprogramming of translation. Annu. Rev. Biochem. 65, 741–768 (1996).

    CAS  PubMed  Article  Google Scholar 

  37. Cate, J.H., Yusupov, M.M., Yusupova, G.Z., Earnest, T.N. & Noller, H.F. X-ray crystal structures of 70S ribosome functional complexes. Science 285, 2095–2104 (1999).

    CAS  PubMed  Article  Google Scholar 

  38. Gottesman, S., Roche, E., Zhou, Y. & Sauer, R.T. The ClpXP and ClpAP proteases degrade proteins with carboxy-terminal peptide tails added by the SsrA-tagging system. Genes Dev. 12, 1338–1347 (1998).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  39. Herman, C., Thevenet, D., Bouloc, P., Walker, G.C. & D'Ari, R. Degradation of carboxy-terminal-tagged cytoplasmic proteins by the Escherichia coli protease HflB (FtsH). Genes Dev. 12, 1348–1355 (1998).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  40. Wickner, S., Maurizi, M.R. & Gottesman, S. Posttranslational quality control: folding, refolding, and degrading proteins. Science 286, 1888–1893 (1999).

    CAS  PubMed  Article  Google Scholar 

  41. Porankiewicz, J., Wang, J. & Clarke, A.K. New insights into the ATP-dependent Clp protease: Escherichia coli and beyond. Mol. Microbiol. 32, 449–458 (1999).

    CAS  PubMed  Article  Google Scholar 

  42. Wang, J., Hartling, J.A. & Flanagan, J.M. The structure of ClpP at 2.3 Å resolution suggests a model for ATP-dependent proteolysis. Cell 91, 447–456 (1997).

    CAS  PubMed  Article  Google Scholar 

  43. Neuwald, A.F. The hexamerization domain of N-ethylmaleimide-sensitive factor: structural clues to chaperone function. Structure Fold. Des. 7, R19–23 (1999).

    CAS  PubMed  Article  Google Scholar 

  44. Bochtler, M., et al. The structures of HsIU and the ATP-dependent protease HsIU-HsIV. Nature 403, 800–805 (2000).

    CAS  PubMed  Article  Google Scholar 

  45. Akiyama, Y., Yoshihisa, T. & Ito, K. FtsH, a membrane-bound ATPase, forms a complex in the cytoplasmic membrane of Escherichia coli. J. Biol. Chem. 270, 23485–23490 (1995).

    CAS  PubMed  Article  Google Scholar 

  46. Levchenko, I., Smith, C.K., Walsh, N.P., Sauer, R.T. & Baker, T.A. PDZ-like domains mediate binding specificity in the Clp/Hsp100 family of chaperones and protease regulatory subunits. Cell 91, 939–947 (1997).

    CAS  PubMed  Article  Google Scholar 

  47. Smith, C.K., Baker, T.A. & Sauer, R.T. Lon and Clp family proteases and chaperones share homologous substrate-recognition domains. Proc. Natl Acad. Sci. USA 96, 6678–6682 (1999).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  48. Hoskins, J.R., Pak, M., Maurizi, M.R. & Wickner, S. The role of the ClpA chaperone in proteolysis by ClpAP. Proc. Natl Acad. Sci. USA 95, 12135–12140 (1998).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  49. Weber-Ban, E.U., Reid, B.G., Miranker, A.D. & Horwich, A.L. Global unfolding of a substrate protein by the Hsp100 chaperone ClpA. Nature 401, 90–93 (1999).

    CAS  PubMed  Article  Google Scholar 

  50. Silber, K.R., Keiler, K.C. & Sauer, R.T. Tsp: a tail-specific protease that selectively degrades proteins with nonpolar C termini. Proc. Natl Acad. Sci. USA 89, 295–299 (1992).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  51. Ponting, C.P. Evidence for PDZ domains in bacteria, yeast, and plants. Protein Sci. 6, 464–468 (1997).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  52. Keiler, K.C. & Sauer, R.T. Sequence determinants of C-terminal substrate recognition by the Tsp protease. J. Biol. Chem. 271, 2589–2593 (1996).

    CAS  PubMed  Article  Google Scholar 

  53. Bass, S., Gu, Q. & Christen, A. Multicopy suppressors of prc mutant Escherichia coli include two HtrA (DegP) protease homologs (HhoAB), DksA, and a truncated R1pA. J. Bacteriol. 178, 1154–1161 (1996).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  54. Tatusov, R.L., Galperin, M.Y., Natale, D.A. & Koonin, E.V. The COG database: a tool for genome-scale analysis of protein functions and evolution. Nucleic Acids Res. 28, 33–36 (2000).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  55. Felden, B., Gesteland, R.F. & Atkins, J.F. Eubacterial tmRNAs: everywhere except the alpha-proteobacteria? Biochim. Biophys. Acta 1446, 145–148 (1999).

    CAS  PubMed  Article  Google Scholar 

  56. Preiss, T. & Hentze, M.W. From factors to mechanisms: translation and translational control in eukaryotes. Curr. Opin. Genet. Dev. 9, 515–521 (1999).

    CAS  PubMed  Article  Google Scholar 

  57. Kessel, M., et al. Homology in structural organization between E. coli ClpAP protease and the eukaryotic 26 S proteasome. J. Mol. Biol. 250, 587–594 (1995).

    CAS  PubMed  Article  Google Scholar 

  58. Schimmel, P., Giege, R., Moras, D. & Yokoyama, S. An operational RNA code for amino acids and possible relationship to genetic code. Proc. Natl Acad. Sci. USA 90, 8763–8768 (1993).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  59. Huang, C., Wolfgang, M.C., Withey, J., Koomey, M. & Friedman, D.I. Charged tmRNA but not tmRNA-mediated proteolysis is essential for Neisseria gonorrhoeae viability. EMBO J. 19, 1098–1107 (2000).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  60. Hutchison, C.A. et al. Global transposon mutagenesis and a minimal Mycoplasma genome. Science 286, 2165–2169 (1999).

    CAS  Article  PubMed  Google Scholar 

  61. Baumler, A.J., Kusters, J.G., Stojiljkovic, I. & Heffron, F. Salmonella typhomurium loci involved in survival within macrophages. Infect. Immun. 62, 1623–1630 (1994).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  62. Julio, S.M., Heithoff, D.M. & Mahan, M.J. ssrA (tmRNA) plays a role in Salmonella enterica serovar typhimurium pathogenesis. J. Bacteriol. 182, 1558–1563 (2000).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  63. Oh, B.K. & Apirion, D. 10Sa RNA, a small stable RNA of Escherichia coli, is functional. Mol. Gen. Genet. 229, 52–56 (1991).

    CAS  PubMed  Article  Google Scholar 

  64. Kirby, J.E., Trempy, J.E. & Gottesman, S. Excision of a P4-like cryptic prophage leads to Alp protease expression in Escherichia coli. J. Bacteriol. 176, 2068–2081 (1994).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  65. Retallack, D.M., Johnson, L.L. & Friedman, D.I. Role for 10Sa RNA in the growth of lambda-P22 hybrid phage. J. Bacteriol. 176, 2082–2089 (1994).

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  66. Retallack, D.M. & Friedman, D.I. A role for a small stable RNA in modulating the activity of DNA-binding proteins. Cell 83, 227–235 (1995).

    CAS  PubMed  Article  Google Scholar 

  67. Kim, Y.I., Burton, R.E., Burton, B.M., Sauer, R.T. & Baker, T.A. Dynamics of substrate denaturation and translocation by the ClpXP degradation machine. Mol. Cell 5, 639–648 (2000).

    CAS  PubMed  Article  Google Scholar 

  68. Keiler, K.C., Shapiro, L. & Kelly, K.P. tmRNA's which encode proteolysis inducing tags are found in all bacterial genomes: a two-piece tmRNA functions in Caulobacter. Proc. Natl. Acad. Sci. USA, in the press.

Download references

Acknowledgements

We thank K. Williams, N. Walsh, K. Keiler, T. Baker, T. RajBhandary, and H. Aiba for communication of unpublished information and for helpful discussions.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Robert T. Sauer.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Karzai, A., Roche, E. & Sauer, R. The SsrA–SmpB system for protein tagging, directed degradation and ribosome rescue. Nat Struct Mol Biol 7, 449–455 (2000). https://doi.org/10.1038/75843

Download citation

  • Received:

  • Accepted:

  • Issue Date:

  • DOI: https://doi.org/10.1038/75843

Further reading

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing