Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M600167200 on February 15, 2006

J. Biol. Chem., Vol. 281, Issue 15, 10561-10566, April 14, 2006
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
281/15/10561    most recent
M600167200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tanner, D. R.
Right arrow Articles by Buskirk, A. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tanner, D. R.
Right arrow Articles by Buskirk, A. R.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Genetic Analysis of the Structure and Function of Transfer Messenger RNA Pseudoknot 1*

Douglas R. Tanner, Jonathan D. Dewey, Mickey R. Miller, and Allen R. Buskirk1

From the Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84602

Received for publication, January 6, 2006 , and in revised form, February 14, 2006.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
tmRNA rescues stalled ribosomes in eubacteria by forcing the ribosome to abandon its mRNA template and resume translation with tmRNA itself as a template. Pseudoknot 1 (pk1), immediately upstream of this coding region in tmRNA, is a structural element that is considered essential for tmRNA function based on the analysis of pk1 mutants in vitro. pk1 binds near the ribosomal decoding site and may make base-specific contacts with tmRNA ligands. To study pk1 structure and function in vivo, we have developed a genetic selection that ties the life of Escherichia coli cells to tmRNA activity. Mutation of pk1 at 20% per base and selection for tmRNA activity yielded sequences that retain the same pseudoknot fold. In contrast, selection of active mutants from 106 completely random sequences identified hairpin structures that functionally replace pk1. Rational design of a hairpin with increased stability using an unrelated sequence yielded a tmRNA mutant with nearly wild-type activity. We conclude that the role of pk1 in tmRNA function is purely structural and that it can be replaced with a variety of hairpin structures. Our results demonstrate that in the study of functional RNAs, the inactivity of a mutant designed to destroy a given structure should not be interpreted as proof that the structure is necessary for RNA function. Such mutations may only destabilize a global fold that could be formed equally well by an entirely different, stable structure.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
tmRNA2 plays an important role in quality control of protein synthesis in eubacteria (1, 2). This dual-function RNA acts as both tRNA and mRNA to rescue ribosomes stalled on broken templates or at certain nascent peptide sequences (3). Aminoacylated tmRNA enters the A-site of stalled ribosomes and transfers alanine to the growing peptide chain as would a normal tRNA. The ribosome then resumes translation with tmRNA as the template, adding a 10-amino-acid tag to the nascent polypeptide. As a result of tmRNA action, stalled ribosomes are released and recycled, and the aborted protein product is marked for destruction by proteases (1, 2).

In the trans-translation model of tmRNA function described above, the ribosome switches templates while synthesizing a single polypeptide. How does tmRNA position itself inside the ribosome for translation to resume in the correct frame? The global fold of tmRNA places the first codon near the decoding center of the ribosome; a short single-stranded sequence immediately upstream of this initial codon determines the precise frame (4-6). The global structure of tmRNA is dominated by several pseudoknot structures surrounding the tag-encoding sequence (7, 8). The Escherichia coli tmRNA, for example, has four pseudoknots (see Fig. 1).

Only 11 nucleotides upstream of the resume codon, pseudoknot 1 (pk1), may be involved with positioning the template in the ribosomal A-site. The cryoelectron microscopy structure of the 70 S ribosome bound to tmRNA, and its partner protein SmpB reveals that although pk2-4 are looped around the beak of the 30 S subunit without forming extensive contacts with the ribosome, pk1 is bound intimately between the beak and the decoding center on the 30 S subunit (9). Valle et al. (9) speculate that pk1 is pulled toward the decoding center as tmRNA transitions from the initial binding complex visualized by cryoelectron microscopy to full accommodation in the A-site resulting in Ala transfer and tmRNA translocation to the ribosomal P-site.

In vitro studies support the conclusion that pk1 plays a crucial role in tmRNA function. Although pk2-4 are dispensable for trans-translation in vitro, replacing pk1 with single-stranded sequence destroys tmRNA function (10). Mutations designed to disrupt the base pairing of pseudoknot 1 helices reduce tmRNA-mediated tagging dramatically (11). The alteration of single-stranded loop sequences lowers activity, leading to proposals that these nucleotides form Mg2+ binding sites (11) or make base-specific contacts with the ribosome (12).

One limitation in the study of the trans-translation system has been the lack of a robust in vivo selection for tagging activity. The ssrA gene that encodes tmRNA in E. coli is not essential for growth under laboratory conditions (2). One report of a genetic selection takes advantage of tmRNA-mediated tagging and proteolysis of the Arc repressor, allowing derepression of the kanR gene and survival on kanamycin (4). The authors characterized 2451 colonies by replica plating and reported high background (~0.1%); this screen is not robust enough to search through large collections of mutants (libraries) of tmRNA or other components of the tagging machinery.

To extend the study of tmRNA structure and function to a relevant in vivo context, we have created a genetic selection that ties tmRNA function to the life of an E. coli cell. This selection allows the characterization of millions of mutants of tmRNA in a single experiment. Here we describe the application of this selection to identifying the structural and functional requirements for pseudoknot 1. How does the pk1 sequence determine its structure and what role does pk1 play in positioning tmRNA correctly inside the ribosome for resumption of translation on tmRNA?


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
DNA oligonucleotides were synthesized on an ABI Expedite 8909 or purchased from Sigma-Genosys. Enzymes were purchased from New England Biolabs.

Plasmid Construction—The tmRNA-K1 plasmid expresses the altered tmRNA gene from its own (ssrA) promoter on a low copy (p15A) origin with a tetracycline resistance marker. It was created by altering the tag sequence (ANDENYALAA) to code for the last 14 amino acids of KanR (ANKLQFHLMLDEFF). Nucleotides 90-137 of tmRNA, from the resume codon to the end of helix 5, were replaced with the sequence GCAAATAAACTGCAGTTTCATTTGATGCTCGATGAGTTCTTCTAATAACAGAATCTCATC. This sequence creates a stem-loop structure similar to helix 5 following the tag template. It contains a PstI cleavage site for cloning mutant pk1 regions immediately upstream. The tmRNA-K1 plasmid was created by PCR amplification of pKW11 (13) using the primers GAGCATCAAATGAAACTGCAGTTTATTTGCGACTATTTTTTGCGGCTTTTTAC and GATGAGTTCTTCTAATAACAGAATCTCATCCCTCTCTCCCTAGCCTCC, followed by phosphorylation and blunt end ligation.


Figure 1
View larger version (37K):
[in this window]
[in a new window]
  		FIGURE 1.
Secondary structure of E. coli tmRNA. Four pseudoknots (pk1-4) dominate the global tmRNA structure. The tag template lies between pk1 and pk2, with the resume and stop codons marked with white boxes. TLD, tRNA-like domain. Adapted from Ref. 24.

 
Truncated kanR constructs were expressed from the araBAD promoter on a multicopy (pBR322 origin) vector with a beta-lactamase marker. The first version of the selection, used in the 20% mutagenesis of wild-type pk1, induced stalling with the sequence Pro255-Pro256-Opal. The kanR-DPP plasmid was cloned by the amplification of the kanR gene with primers CATATGGCTAGCATGAGCCATATTCAACGGGAAAC and CGGCTTTTTCAAAAATATGGTATTGATCCGCCTTGAGAATTCGAGCTC, digestion with NheI and EcoRI, and ligation into the pBAD-GFP vector.

To combine the two plasmids into a single selection vector, the tmRNA expression cassette was PCR-amplified from tmRNA-K1 with primers CCGCTACGGTCCGAGAACTGTGAATGCGCAAACC and TAGCGAAGATCTTAAATCCTGGTGTCCCTGTTG. The kanR-DPP plasmid was PCR-amplified with primers CGACCGAGATCTTCGCTACGTGACTGGGTCATG and AGTTCTCGGACCGTAGCGGAGTGTATACTGGCTTAAC. These fragments were digested with BglII and RsrII and ligated together. QuikChange mutagenesis and removal of SphI and PstI sites from the resulting plasmid, pBad-KT2, allow these unique sites surrounding the pseudoknot 1 sequence to be used for cloning. The sequence GAATAGAGGCCTTCAACTCCGCGGATACTA was then inserted between SphI and PstI, replacing wild-type pk1, and making tmRNA inactive. This "dummy insert" also provides StuI and SacII sites to assist in library creation.

A second, more active version of the selection vector was generated by replacing the C-terminal amino acids Asp254-Pro255-Pro256 (kanR-DPP) with Ser254-Glu255-Pro256 in the truncated kanR gene. This kanR-SEP sequence was discovered by cloning six random codons after Ile253 of the truncated kanR{Delta}15 gene in pBad-KT2 and selecting for kanamycin resistance under more stringent conditions: 37 °C and 15 µg/ml kanamycin. Addition of a chloramphenicol resistance cassette into the BglII site provided a second antibiotic marker to reduce contamination, yielding the plasmid p16Dum-Cat.

tmRNA expression vectors for the bacteriophage assays were created by PCR amplification of the pKW11 plasmid and blunt end cloning. The procedure was similar to that described above to make tmRNA-K1, except that no change was made to the tag template; the only difference from wild-type tmRNA is the pseudoknot 1 sequence. The {Delta}ssrA control was made by removing the entire tmRNA sequence by digestion of pKW11 with NcoI and EcoRI, followed by treatment with the Klenow fragment of DNA polymerase I to create blunt ends for religation. The pk1 sequences are as follows: 50CUC, CCTCGGGCGGTTGGCCTCGTAAAAAGCCGC; pk1L, CGAGGGCCGC; M20, AAAAGCGTCCCGTTAGGGACGGTGGGAATA; M20-2, AAAAGCATCCCGTTAGGGATGTTGGGATA; RD2, AAACAGCCCGGGGAACCGGGCTGAATAAA.

Selection of Mutant pk1 Libraries—The conservative mutagenesis of pk1 was accomplished by mutating the 30 nucleotides of pk1 at 20% per base. The oligonucleotide CAAGGTGCATGCCGAGGGGCGGTTGGCCTCGTAAAAAGCCGCAAAAAATAGTCGCAAATAAACTGCAGTTTCAT was synthesized with mixed phosphoramidites incorporated into the positions underlined. These mixes included 80% of the wild-type base and ~7% each of the other three bases. A short primer bound to a constant 3'-region was extended by the Klenow fragment of DNA polymerase I to generate double-stranded DNA. This DNA was cleaved by SphI and PstI and ligated into the pBad-KT2 vector, and the plasmid library was introduced into DH10B by electroporation. Following plasmid purification, the resulting library was introduced into the selection strain, X90 ssrA::cat (14). The cells were induced with 2% arabinose for 3 h at 30 °C, plated onto media containing ampicillin, 2% arabinose, and 15 µg/ml kanamycin, and grown for 48 h at 25 °C.

The second library, with fully randomized pk1 sequences, was generated in a similar manner except it incorporated N30 (25% each base) in place of the underlined sequence above. This library was cloned into the second generation selection vector p16Dum-Cat and selected as above. The third library, with 20% mutagenesis of M20, replaced the underlined sequence with mixed phosphoramidites as above but based on the M20 sequence. The mutant inserts were cloned into p16Dum-Cat and selected on media containing ampicillin, 2% arabinose, and 50 µg/ml kanamycin at 25 °C. Surviving clones from these libraries were recloned, transformed into fresh cells, grown up from single colonies, and reassayed under selection conditions to ensure that the tmRNA mutant in question is responsible for the tagging activity.

Phage Efficiency of Plating (EOP) Assays—X90 ssrA::cat cells carrying a pKW11-derivative expressing pk1 mutants of tmRNA were grown overnight in media with tetracycline at 37 °C. The cells were washed in 2x YT media, diluted, and grown in 2x YT with 10 mM MgSO4 for 3 h at 37 °C to an A600 of ~1. The cells were then washed with 10 mM MgSO4, resuspended to an A600 of 0.5, and measured out into 200-µl aliquots. 1 µl of phage {lambda}immP22 c2-dis at various dilutions was added to an aliquot, incubated at 37 °C for 15 min, mixed with 3 ml of top agar (2x YT, 10 mM MgSO4, and 0.7% agarose), and grown overnight at 37 °C. Plaques on tmRNA mutants were smaller than those formed on wild-type tmRNA-containing cells as reported previously (15). The number of plaques were counted for four different trials as reported in Fig. 5.


Figure 2
View larger version (21K):
[in this window]
[in a new window]
  		FIGURE 2.
Genetic selection for tmRNA activity. Ribosomes stall on a truncated kanR template (kanR{Delta}15) at the Glu-Pro-(Opal) sequence. Active tmRNA molecules with a mutant template sequence add the final 15 amino acids of KanR (shown in "red") to the nascent polypeptide (yellow). These C-terminal 15 residues form a structurally critical helix in the crystal structure of the homologous Aph(3')-IIIa protein (16). tmRNA function is linked to KanR activity and cellular survival, yielding 106-fold enrichment of active sequences.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Development of a Selection for tmRNA Activity—The function of tmRNA in vivo is to tag mistranslated proteins with an 11-amino-acid sequence. If this tag could be altered to confer protein function rather than proteolysis, the activity of the tagged protein would reflect the level of tmRNA function. To provide a genetic selection, the protein must be required for cell survival and have essential residues in its C-terminal 10-20 amino acids. In the absence of tmRNA activity, the truncated protein would be inactive, and the cells would die; in the presence of an active tmRNA variant, the protein could be tagged with the essential amino acids to form a functional enzyme and confer cell survival (Fig. 2).

The selection protein we chose is the E. coli kanamycin resistance protein (KanR). Analysis of the homologous Aph(3')-IIIa·kanamycin co-crystal structure (Fig. 2) reveals that the C-terminal helix of 15 residues (shown in red) plays an important role both structurally and catalytically in binding the substrate (16). Deletion of these residues leads to loss of function; cells expressing the KanR C-terminal deletion (kanR{Delta}15) show no more kanamycin resistance than cells lacking the kanR gene.

Natural tmRNA sequences encoding the peptide tag vary in length and composition and can be altered without loss of function (13, 17). We replaced the natural tag template with a sequence encoding ANKLQFHLMLDEFF and expressed tmRNA from its own promoter from the tmRNA-K1 plasmid. Following transfer of Ala from the tmRNA itself, the addition of these 14 residues encoded by tmRNA restores the missing 15 amino acids that make up the C-terminal helix of KanR. We found that the identity of the resume codon was an important factor; GCA was significantly more active than AUG, the wild-type kanR codon (data not shown). All known tmRNA template sequences contain GNN resume codons (17). Presumably G at this position is important for the recognition of the proper frame for reinitiation of translation by the ribosome.

For the altered tag to complete the KanR protein correctly, ribosomes must stall at exactly the right position on the kanR{Delta}15 mRNA. Ribosomes stall during termination at inefficient stop codons such as the opal codon (UGA) when the protein sequence terminates in proline (3). The E. coli YbeL protein, for example, ends in Glu-Pro-(Opal) and is tagged with good efficiency (up to 40%) by tmRNA precisely at the position of the stop codon (3). Two mutations in kanR{Delta}15, N255E, and Met257Opal, create a Glu-Pro-(Opal) sequence in the kanR mRNA that acts as a signal to induce ribosome stalling, yielding the kanR-SEP selection gene. Based on the structure of Aph(3')-IIIa, the KanR protein is expected to have a surface-exposed loop of five residues, ending in Pro256, preceding the C-terminal helix (16). This loop permits the introduction of the Glu-Pro-Ala scar introduced by these changes and transfer of Ala from aminoacylated tmRNA without affecting KanR function.

Tagging of the truncated KanR polypeptides on stalled ribosomes by the altered tmRNA generates full-length, functional KanR protein, conferring kanamycin resistance to cells. We co-transformed an E. coli strain lacking tmRNA with the kanR-SEP and tmRNA-K1 plasmids. When plated onto media containing 15 µg/ml kanamycin, all of the co-transformants survive at 37 °C. Under the same conditions, bacteria with kanR-SEP but lacking the modified tmRNA-K1 survive at the level of 5 colony forming units in 107 plated. The enrichment ratio is roughly a factor of 103 better than the previous genetic selection for tmRNA function (4).

Conservative Mutagenesis of pk1—This genetic selection for transtranslation provides a method to rapidly identify active tmRNA sequences from libraries of millions of mutants. Conserved bases in the sequences of active mutants reveal which positions contain information required for function. This strategy of mutagenesis and selection was used to characterize the relationship of the pk1 sequence to its structure and function.

In creating a library of pk1 mutants, we biased the sequences toward the wild-type pk1 sequence to retain the overall pk1 fold thought to be required for tmRNA function. We randomized the sequence of pk1 at the rate of 20% per base (see "Materials and Methods"). Analysis of mutants prior to selection revealed that an average of seven mutations were introduced at random positions in the 30 nucleotides of pk1. The library of ~4 x 107 tmRNA mutants was introduced with the selection gene into an E. coli strain lacking tmRNA. Following selection at 25 °C on 15 µg/ml kanamycin, 104-105 colonies survived.


Figure 3
View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 3.
Consensus sequence of active pseudoknot 1 mutants. The height at each position indicates the informational content required for function. The wild-type sequence was mutated at 20% per base, so differences in height and frequency are more significant than they appear, as the sequence is heavily biased toward wild-type. Stem 1 is colored orange, and stem 2 is blue. Created by Weblogo (25).

 
The comparison of 67 pk1 sequences from the selected tmRNAs gives a clear picture of what positions are tolerant to mutation. The average number of mutations per pk1 sequence was 1.4, significantly lower than the average of 7 in preselection sequences. This was expected given the sensitivity of the pk1 structure to mutation. The results in Fig. 3 indicate the information required at each position by the total height (e.g. position 61 has low information content and is very tolerant of mutations). Because the greatest possible mutagenesis at each position is 20%, the sequence is heavily biased toward the wild-type sequence. The information content is therefore artificially inflated and the differences are more significant than they appear.

The two helical regions, stem 1 (nucleotides 49-53 and 63-67) and stem 2 (nucleotides 55-59 and 74-78), are not equally important in the pk1 structure (Fig. 3). Stem 1 is absolutely invariant, with all the selected clones maintaining all five possible base pairs. The wobble pair introduced by the mutation A51G is active; likewise, mutations at the other positions are always accompanied by covarying mutations that re-establish the base pair (e.g. G50C and C66G). In contrast, only three of the five base pairs of stem 2 are conserved; uncompensated mutations of the outer G55-C78 and U59-G74 base pairs are tolerated.

Contrary to conclusions based on in vitro work (11, 12), loop 2 (nucleotides 60-62) can be mutated in vivo without affecting trans-translation. The earlier finding that loop 2 mutants form pseudoknots but display inhibited activity led to the speculation that these bases form an Mg2+ or ribosome binding site. The strict requirement for G or U at position 61 and G at 62 in vitro was not observed in our in vivo assay; G61C and G62U mutants were viable.

There is disagreement in the literature whether or not the first nucleotide in loop 2, U60, pairs with A73 to extend stem 2 by one base pair (7, 8, 11). Chemical probing showed that U60 is accessible, but A73 is protected from modifications (11). Similarly, our findings suggest that U60 can be mutated to the other bases without loss of function, but A73 is highly conserved. We propose that U60 is unpaired and is part of loop 2, whereas A73 forms interactions elsewhere.

The conservation of the adenosines in loop 3 (nucleotides 69-73) is striking, particularly at the 3'-end of that sequence. Analysis of the pseudoknot data base shows that adenosine is highly enriched in loop 3 of pseudoknots (63.9% A) particularly at the 3'-end (18). These adenosines may make A-minor motif interactions (19) with the stem 1 helix, as seen in the high-resolution structures of several pseudoknots (20, 21). Such interactions are consistent with the in vitro finding that mutations in loop 3 destabilize stem 1 in tmRNA pk1 (11). Formation of A-minor motif interactions may place constraints on the sequence of stem 1, particularly the G52-C64 and G53-C63 pairs. These positions are invariant in all of our analyzed sequences. Construction of a covariant mutant (G49U51C53-G63A65C67) that alters these base-pairing patterns in stem 1 did not restore activity in vitro (11).

Random Mutagenesis of pk1—Having learned what the requirements are for pk1 to fold correctly, we wondered whether more radical mutations could form alternate structures that function in place of pk1. To isolate active sequences far away in sequence space from wild-type pk1, we created a library of 1 x 106 mutants in which the entire pk1 sequence was completely randomized (N30). This library covered but a tiny fraction of the theoretical diversity of 430 or ~1018. Following selection at 25 °C on 15 µg/ml kanamycin, roughly 50 colonies survived. Sequencing of tmRNA sequences revealed that these mutants bear no resemblance to the wild-type pk1 sequence. None form pseudoknot structures by inspection or analysis by the prediction software pknotsRG-mfe (22). Instead, the mutant structures are predicted by the mfold algorithm (23) to form simple hairpin-loop structures (e.g. mutant M20, Fig. 4). All mfold free energy calculations and structural predictions were performed at 37 °C.

The activity of mutants like M20 that replace pk1 with a hairpin shows that pk1 is not required for tmRNA activity. M20 is significantly less active than wild-type tmRNA but fails to confer resistance to high kanamycin concentrations (50 µg/ml) or at higher temperatures (37 °C), conditions under which cells containing wild-type tmRNA display 100% survival. To create a second generation library, we mutated the M20 sequence at 20% per base over the 30 nucleotides replacing the pk1 sequence. A library of 2 x 106 M20 derivatives was subjected to selection on plates containing 50 µg/ml kanamycin (kan50) at 25 °C. Of the eight surviving clones, M20-2 is the most active, with 100% survival on kan50 at 25 °C and 50% survival on kan15 at 37 °C. The predicted secondary structure of M20-2 contains two more base pairs than its M20 parent and is predicted to be slightly more stable thermodynamically (-12.4 kJ/mol versus -11.5 kJ/mol, respectively).

Quantification of pk1 Mutant Activity—Our initial analyses of M20 and M20-2 function were performed with the KanR assay and in the tmRNA context (a mutated tag sequence) in which the mutants were evolved. To correlate our assay with others in the literature and to obtain more quantitative measures of the activity of the mutants, we introduced the evolved pk1 sequences into otherwise wild-type tmRNA (containing the natural template encoding ANDENYALAA). tmRNA activity was measured by the efficiency of plaque formation by the hybrid bacteriophage {lambda}immP22 c2-dis. This phage only forms plaques on bacteria with intact trans-translation systems (2, 15); cells expressing wild-type tmRNA from a low copy plasmid (pKW11) support the growth of ~1 x 105 more plaques than cells expressing no tmRNA.

In addition to the wild-type and {Delta}ssrA controls, we assayed two pk1 mutants that were characterized previously in vitro. The first, pk1L, replaces pk1 with a single-stranded sequence CGAGGGCCGC. This mutant is from the study reporting that although pk2, pk3, and pk4 were dispensable for tmRNA function in vitro, pk1 was essential (10). The second, 50CUC, maintains the pk1 sequence but is designed to destroy the folding of stem 1 by preventing the core three bases from pairing. The 50CUC mutant displays a 10-fold reduction in activity in vitro (11). EOP assays reveal that these two mutants are indistinguishable from the tmRNA knock-out, {Delta}ssrA (Fig. 5). As measured by this assay, a missing or misfolded pk1 region renders tmRNA completely inactive in vivo.


Figure 4
View larger version (13K):
[in this window]
[in a new window]
  		FIGURE 4.
Secondary structures of active pk1 replacements. The M20 mutant was selected from a randomized sequence (N30) replacing the wild-type pk1 sequence 49-79. 20% mutagenesis of the M20 sequence and selection at higher stringency yielded M20-2. RD2 was rationally designed using an unrelated sequence to form a hairpin structure with increased stability. Structures and thermodynamic stabilities were predicted by the mfold algorithm at 37 °C (23).

 


Figure 5
View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 5.
Analysis of mutant tmRNA activity. The hybrid bacteriophage {lambda}immP22 c2-dis only forms plaques on cells expressing active tmRNA (15). Data are expressed as EOP with wild-type tmRNA taken as EOP = 1. The del(ssrA) mutant lacks tmRNA; mutants pk1L (10) and 50CUC (11) destroy pseudoknot 1 folding and tmRNA function in vitro. Error bars represent the S.D. of four independent experiments.

 
The M20 clone evolved from random sequence is ~1000-fold more active than the missing or misfolded pk1 controls, suggesting that its hairpin structure is sufficient to restore tmRNA stability and function (Fig. 5). M20 is, however, ~100-fold weaker than wild-type tmRNA. The M20-2 second-round clone shows an activity 4-fold higher than M20. The cutoff for survival in the high stringency KanR assay must therefore lie between M20 and M20-2 activity.

The fact that the improved M20-2 mutant has more predicted base pairs and thermodynamic stability led to the hypothesis that further improvements in stability of the hairpin could improve activity. If a hairpin structure is all that is required to replace the wild-type pk1, perhaps there is nothing special about the hairpins that we evolved, and any stable hairpin structure could functionally replace pk1. To test these hypotheses, we rationally designed an eight-base-pair stem with a tetraloop, utilizing a sequence substantially different from the evolved M20 and M20-2 clones. This rationally designed hairpin, RD2, is predicted to be significantly more stable (-19.8 kJ/mol versus -11.5 kJ/mol) than the M20 hairpin (23). EOP assays show that RD2 is nearly as active as wild-type tmRNA, 29-fold more active than M20 (Fig. 5). These findings suggest that the stability of the structure replacing pseudoknot 1 plays an key role in the overall function of tmRNA mutants but that the specific sequence used is not as important.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Previous in vitro studies suggest that tmRNA pseudoknot 1 (pk1) is essential to the tagging function in trans-translation. It must form a pseudoknot structure (10, 11), may make specific contacts to the ribosome (12), and may play a role in positioning the template region of tmRNA near the decoding center for proper continuation of translation (9), although the precise choice of frame is determined by the single-stranded sequence between the resume codon and pk1 (4-6).

To study the requirements for pk1 structure and its relation to tmRNA function in vivo, we developed a genetic selection that ties the life of an E. coli cell to tmRNA activity. KanR polypeptides lacking an essential C-terminal sequence are stalled on ribosomes; if rescued by an altered tmRNA that codes for the missing amino acids, the ribosomes produce full-length, functional KanR, and the cell survives on kanamycin plates. Cells lacking tmRNA activity are killed. This selection provides the means to identify rare active mutants among large libraries of tmRNA, giving 106-fold enrichment of active sequences. Work is also underway to use this genetic selection to characterize other components of the trans-translation process.

The generation of two libraries with different levels of mutagenesis yields apparently contradictory conclusions. The first library, generated by a conservative mutagenesis scheme (20% per base), supports the previous in vitro findings that mutations that destabilize the pseudoknot structure of pk1 destroy tmRNA activity. Very few mutations were tolerated; the stem 1 region allowed covarying mutations only, maintaining all five of its base pairs, whereas stem 2 allowed the loss of its outer two base pairs. The U60G61G62 bases of loop 2 are tolerant to mutation, suggesting that they do not form critical Mg2+ binding sites or bind external ligands such as ribosomal components as previously proposed (11, 12). U60 does not pair with A73, instead, A73 and other conserved adenine bases at the 3'-end of loop 3 probably form A-minor motif interactions with the end of stem 1. These findings about the sequence requirements of the pk1 structure add a new layer of detail to what was previously known but are largely consistent with previous in vitro findings. They reinforce the idea that a well folded pseudoknot 1 structure is essential to tmRNA function.

In contrast, results from the second library, in which pk1 was replaced by completely random sequence (N30), support the conclusion that any stable secondary structure is sufficient. Rare mutants were isolated that substitute hairpin-loop structures in place of pk1. A second round of mutagenesis and selection improved the activity of the best of these clones, M20, by increasing hairpin stability (mutant M20-2). Rational design of an even more stable hairpin, using an unrelated sequence, created a mutant (RD2), which is nearly as active as wild-type tmRNA. The sequence of the hairpin appears to be much less important than its thermodynamic stability. These surprising results were confirmed with a standard EOP assay using phage {lambda}immP22 c2-dis. This assay demonstrates that our results are not because of an artifact of the KanR assay or the altered tmRNA tag sequence used in the selection.

What is the resolution of these apparently paradoxical conclusions? Low levels of mutagenesis of the wild-type pk1 sequence either destroys or retains the pk1 fold, but total randomization allows alternate stable structures far away in sequence space to be identified. We propose that the only requirement for the pk1 region of tmRNA is that it forms a stable structure that prohibits other deleterious global misfolding events. Alternate stable conformations were detected by NMR and denaturation profiles previously in vitro (11, 12); the role of stable pseudoknots is probably to prevent these structures from folding. A pseudoknot structure is not necessary however.

This finding has important consequences for mutational analyses of structures in functional RNAs. The inactivity of a mutant designed to destroy a given structure does not logically require that the structure is necessary for function, as is too often assumed. The mutation may cause global folding problems that in turn render the molecule inactive. This does not prove that the given structure could not be replaced by a stable structure quite different from itself. In our case, the inactivity of a mutant pseudoknot does not prove that the structure has to be a pseudoknot; it only demonstrates that the region in question must form a stable fold. Stability and global folding should therefore be an important consideration in designing mutations to test hypotheses about the role of specific structures in RNA and in interpreting such studies. Our findings also highlight the importance of analyzing a large number of mutants and the power of genetic approaches to do so in an efficient manner.


   FOOTNOTES
 
* This work was supported by funds from Brigham Young University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 To whom correspondence should be addressed: C203 BNSN, Dept. of Chemistry and Biochemistry, Brigham Young University, Provo, UT 84602. Tel.: 801-422-1967; Fax: 801-422-0153; E-mail: buskirk{at}chem.byu.edu.

2 The abbreviations used are: tmRNA, transfer messenger RNA; pk1, pseudoknot 1; A-site, ribosomal aminoacyl site; P-site, ribosomal peptidyl site; EOP, efficiency of plating. Back


   ACKNOWLEDGMENTS
 
We thank Robert Sauer and co-workers for the pKW11 plasmid and the X90 ssrA::cat strain and David Friedman for the {lambda}immP22 c2-dis phage.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Karzai, A. W., Roche, E. D., and Sauer, R. T. (2000) Nat. Struct. Biol. 7, 449-455[CrossRef][Medline] [Order article via Infotrieve]
  2. Withey, J. H., and Friedman, D. I. (2003) Annu. Rev. Microbiol. 57, 101-123[CrossRef][Medline] [Order article via Infotrieve]
  3. Hayes, C. S., Bose, B., and Sauer, R. T. (2002) J. Biol. Chem. 277, 33825-33832[Abstract/Free Full Text]
  4. Williams, K. P., Martindale, K. A., and Bartel, D. P. (1999) EMBO J. 18, 5423-5433[CrossRef][Medline] [Order article via Infotrieve]
  5. Lee, S., Ishii, M., Tadaki, T., Muto, A., and Himeno, H. (2001) RNA (N. Y.) 7, 999-1012
  6. Lim, V. I., and Garber, M. B. (2005) J. Mol. Biol. 346, 395-398[Medline] [Order article via Infotrieve]
  7. Williams, K. P., and Bartel, D. P. (1996) RNA (N. Y.) 2, 1306-1310
  8. Felden, B., Himeno, H., Muto, A., McCutcheon, J. P., Atkins, J. F., and Gesteland, R. F. (1997) RNA (N. Y.) 3, 89-103
  9. Valle, M., Gillet, R., Kaur, S., Henne, A., Ramakrishnan, V., and Frank, J. (2003) Science 300, 127-130[Abstract/Free Full Text]
  10. Nameki, N., Tadaki, T., Himeno, H., and Muto, A. (2000) FEBS Lett. 470, 345-349[CrossRef][Medline] [Order article via Infotrieve]
  11. Nameki, N., Felden, B., Atkins, J. F., Gesteland, R. F., Himeno, H., and Muto, A. (1999) J. Mol. Biol. 286, 733-744[CrossRef][Medline] [Order article via Infotrieve]
  12. Nameki, N., Chattopadhyay, P., Himeno, H., Muto, A., and Kawai, G. (1999) Nucleic Acids Res. 27, 3667-3675[Abstract/Free Full Text]
  13. Roche, E. D., and Sauer, R. T. (2001) J. Biol. Chem. 276, 28509-28515[Abstract/Free Full Text]
  14. Keiler, K. C., Waller, P. R., and Sauer, R. T. (1996) Science 271, 990-993[Abstract]
  15. Withey, J., and Friedman, D. (1999) J. Bacteriol. 181, 2148-2157[Abstract/Free Full Text]
  16. Fong, D. H., and Berghuis, A. M. (2002) EMBO J. 21, 2323-2331[CrossRef][Medline] [Order article via Infotrieve]
  17. Gueneau de Novoa, P., and Williams, K. P. (2004) Nucleic Acids Res. 32, D104-108[Abstract/Free Full Text]
  18. Aalberts, D. P., and Hodas, N. O. (2005) Nucleic Acids Res. 33, 2210-2214[Abstract/Free Full Text]
  19. Nissen, P., Ippolito, J. A., Ban, N., Moore, P. B., and Steitz, T. A. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 4899-4903[Abstract/Free Full Text]
  20. Su, L., Chen, L., Egli, M., Berger, J. M., and Rich, A. (1999) Nat. Struct. Biol. 6, 285-292[CrossRef][Medline] [Order article via Infotrieve]
  21. Michiels, P. J., Versleijen, A. A., Verlaan, P. W., Pleij, C. W., Hilbers, C. W., and Heus, H. A. (2001) J. Mol. Biol. 310, 1109-1123[CrossRef][Medline] [Order article via Infotrieve]
  22. Reeder, J., and Giegerich, R. (2004) BMC Bioinformatics 5, 104[CrossRef][Medline] [Order article via Infotrieve]
  23. Zuker, M. (2003) Nucleic Acids Res. 31, 3406-3415[Abstract/Free Full Text]
  24. Burks, J., Zwieb, C., Muller, F., Wower, I., and Wower, J. (2005) BMC Mol. Biol. 6, 14[Medline] [Order article via Infotrieve]
  25. Crooks, G. E., Hon, G., Chandonia, J. M., and Brenner, S. E. (2004) Genome Res. 14, 1188-1190[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 RNAHome page
I. K. Wower, C. Zwieb, and J. Wower
Escherichia coli tmRNA lacking pseudoknot 1 tags truncated proteins in vivo and in vitro
RNA, January 1, 2009; 15(1): 128 - 137.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
281/15/10561    most recent
M600167200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tanner, D. R.
Right arrow Articles by Buskirk, A. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tanner, D. R.
Right arrow Articles by Buskirk, A. R.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2006 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement