StreptoTag: a novel method for the isolation of RNA-binding proteins.

We describe a fast and simple one-step affinity-purification method for the isolation of specific RNA-binding proteins. An in vitro-transcribed hybrid RNA consisting of an aptamer sequence with high binding specificity to the antibiotic streptomycin and a putative protein-binding RNA sequence is incubated with crude extract. After complex formation, the sample is applied to an affinity column containing streptomycin immobilized to Sepharose. The binding of the in vitro-assembled RNA-protein complex to streptomycin-Sepharose is mediated by the aptamer RNA and the specifically bound proteins are recovered from the affinity matrix by elution with the antibiotic. Employing two well-characterized RNA-protein interactions, we tested the performance of this new method. The spliceosomal U1A protein and the bacteriophage MS2 coat protein could be isolated via their appropriate RNA motif containing hybrid RNA from crude yeast extracts in high yield and purity after only one round of affinity purification. As the purification principle is independent of the extract source, this new affinity chromatography strategy that makes use of an in vitro-selected antibiotic-binding RNA as a tag, "StreptoTag," should be applicable to extracts from other organisms as well. Therefore, we propose StreptoTag to be a versatile tool for the isolation of unknown RNA-binding proteins.

ASH1 mRNA localization in yeast involves multiple secondary structural elements and Ash1 protein translation.

Localization of ASH1 mRNA to the distal cortex of daughter but not mother cells at the end of anaphase is responsible for the two cells' differential mating-type switching during the subsequent cell cycle. This localization depends on actin filaments and a type V myosin (She1/Myo4). The 3' untranslated region (3' UTR) of ASH1 mRNA is reportedly capable of directing heterologous RNAs to a mother cell's bud [1] [2]. Surprisingly, however, its replacement has little or no effect on the localisation of ASH1 mRNA. We show here that, unlike all other known localization sequences that have been found in 3' UTRs, all the elements involved in ASH1 mRNA localization are located at least partly within its coding region. A 77 nucleotide region stretching from 7 nucleotides 5' to 67 nucleotides 3' of the stop codon of ASH1 mRNA is sufficient to localize mRNAs to buds; the secondary structure of this region, in particular two stems, is important for its localizing activity. Two regions entirely within coding sequences, both sufficient to localize green fluorescent protein (GFP) mRNA to growing buds, are necessary for ASH1 mRNA localization during anaphase. These three regions can anchor GFP mRNA to the distal cortex of daughter cells only inefficiently. The tight anchoring of ASH1 mRNA to the cortex of the daughter cell depends on translation of the carboxy-terminal sequences of Ash1 protein.

Translational fidelity: error-prone versus hyper-accurate ribosomes.

Decoding mRNA is a multistep process involving the RNA and protein components of the ribosome, and external factors; little is known about the mechanism, however. New evidence suggests that a central region in small ribosomal RNA switches between two helices in translation to maintain translational fidelity.

Identification of specific Rp-phosphate oxygens in the tRNA anticodon loop required for ribosomal P-site binding.

tRNA binding to the ribosomal P site is dependent not only on correct codon-anticodon interaction but also involves identification of structural elements of tRNA by the ribosome. By using a phosphorothioate substitution-interference approach, we identified specific nonbridging Rp-phosphate oxygens in the anticodon loop of tRNA(Phe) from Escherichia coli which are required for P-site binding. Stereospecific involvement of phosphate oxygens at these positions was confirmed by using synthetic anticodon arm analogues at which single Rp- or Sp-phosphorothioates were incorporated. Identical interference results with yeast tRNA(Phe) and E. coli tRNA(fMet) indicate a common backbone conformation or common recognition elements in the anticodon loop of tRNAs. N-ethyl-N-nitrosourea modification-interference experiments with natural tRNAs point to the importance of the same phosphates in the loop. Guided by the crystal structure of tRNA(Phe), we propose that specific Rp-phosphate oxygens are required for anticodon loop ("U-turn") stabilization or are involved in interactions with the ribosome on correct tRNA-mRNA complex formation.

In vitro selection of a viomycin-binding RNA pseudoknot.

BACKGROUND: The peptide antibiotic viomycin inhibits ribosomal protein synthesis, group I intron self-splicing and self-cleavage of the human hepatitis delta virus ribozyme. To understand the molecular basis of RNA binding and recognition by viomycin, we isolated a variety of novel viomycin-binding RNA molecules using in vitro selection. RESULTS: More than 90% of the selected RNA molecules shared one continuous highly conserved region of 14 nucleotides. Mutational analyses, structural probing, together with footprinting experiments by chemical modification, and Pb2+-induced cleavage showed that this conserved sequence harbours the antibiotic-binding site and forms a stem-loop structure. Moreover, the loop is engaged in a long-range interaction forming a pseudoknot. CONCLUSIONS: A comparison between the novel viomycin-binding motif and the natural RNA target sites for viomycin showed that all these segments form a pseudoknot at the antibiotic-binding site. We therefore conclude that this peptide antibiotic has a strong selectivity for particular RNA pseudoknots.

Identification of 2'-hydroxyl groups required for interaction of a tRNA anticodon stem-loop region with the ribosome.

Synthetic RNA stem loops corresponding to positions 28-42 in the anticodon region of tRNA(Phe) bind efficiently in an mRNA-dependent manner to ribosomes, whereas those made from DNA do not. In order to identify the positions where ribose is required, the anticodon stem-loop region of tRNA(Phe) (Escherichia coli) was synthesized chemically using a mixture of 2'-hydroxyl- and 2'-deoxynucleotide phosphoramidites. Oligonucleotides whose ribose composition allowed binding were retained selectively on nitrocellulose filters via binding to 30S ribosomal subunits. The binding-competent oligonucleotides were submitted to partial alkaline hydrolysis to identify the positions that were enriched for ribose. Quantification revealed a strong preference for a 2'-hydroxyl group at position U33. This was shown directly by the 50-fold lower binding affinity of a stem loop containing a single deoxyribose at position U33. Similarly, defective binding of the corresponding U33-2'-O-methyl-substituted stem-loop RNA suggests that absence of the 2'-hydroxyl group, rather than an altered sugar pucker, is responsible. Stem-loop oligoribonucleotides from different tRNAs with U33-deoxy substitutions showed similar, although quantitatively different effects, suggesting that intramolecular rather than tRNA-ribosome interactions are affected. Because the 2'-hydroxyl group of U33 was shown to be a major determinant of the U-turn of the anticodon loop in the crystal structure of tRNA(Phe) in yeast, our finding might indicate that the U-turn conformation in the anticodon loop is required and/or maintained when the tRNA is bound to the ribosomal P site.

Inhibition of the self-cleavage reaction of the human hepatitis delta virus ribozyme by antibiotics.

Human hepatitis delta virus (HDV) poses a health threat in populations where chronic hepatitis B is endemic. It is a single-stranded RNA virus of 1700 nucleotides and both genomic and antigenomic sequences contain ribozymes which are important for viral replication. Using ribozyme constructs we show that several classes of antibiotics inhibit the self-cleavage reaction of the HDV ribozyme. Antibiotics of the aminoglycoside, peptide and tetracycline classes all inhibit HDV cleavage in vitro at micromolar concentrations. Neomycin (an aminoglycoside) inhibits HDV self-cleavage with a Ki value of 28 (+/- 10) microM. Neomycin inhibition can be reversed by increasing magnesium ion concentration in a competitive manner. Lead acetate cleaves positions G76, A42 and G28, which surround the ribozyme cleavage site. Both Mg2+ and neomycin prevent lead cleavage. Footprinting experiments using base-specific chemical probes revealed enhanced modifications of a set of bases by neomycin, overlapping with the above mentioned lead cleavages. These observations may indicate that neomycin directly displaces divalent metal ions essential for catalysis.

A novel RNA motif for neomycin recognition.

BACKGROUND: Antibiotics can interfere with RNA activity. Translation of RNA by the prokaryotic ribosome, self-splicing of group I introns, HIV replication and hammerhead ribozyme cleavage are inhibited by the aminoglycoside neomycin B. To explore the molecular basis by which small molecules such as antibiotics inhibit RNA function, we undertook an in vitro selection to obtain a variety of RNA molecules with the capacity to recognize neomycin. RESULTS: The majority of the RNA molecules selected to specifically bind neomycin share a region of nucleotide sequence homology. From chemical probing and covariations among different clones we show that in all sequences this region folds into a hairpin structure, which from footprinting and partial alkaline hydrolysis experiments is shown to be the neomycin-binding site. Neomycin is recognized with high affinity (Kd approximately equal to 100 nM) and high specificity (> 100-fold higher affinity for neomycin than for paromomycin). CONCLUSIONS: The fact that RNAs containing the consensus sequence, as well as sequences that display variations within this region, specifically recognize neomycin suggests that a structural motif rather than a particular nucleotide sequence is required for neomycin recognition. We propose that a hairpin stem-loop structural motif, which might feature a widened major groove, may be a prerequisite for neomycin recognition. This structural pattern can be extrapolated to other natural neomycin-responsive RNAs.

Identification of bases in 16S rRNA essential for tRNA binding at the 30S ribosomal P site.

Previous studies suggest that the mechanism of action of the ribosome in translation involves crucial transfer RNA (tRNA)-ribosomal RNA (rRNA) interactions. Here, a selection scheme was developed to identify bases in 16S rRNA that are essential for tRNA binding to the P site of the small (30S) ribosomal subunit. Modification of the N-1 and N-2 positions of 2-methylguanine 966 and of the N-7 position of guanine 1401 interfered with messenger RNA (mRNA)-dependent binding of tRNA to the P site. Modification of the same positions as well as of the N-1 and N-2 positions of guanine 926 interfered with mRNA-independent binding of tRNA at high magnesium ion concentration. These results suggest that these three bases are involved in intermolecular contacts between ribosomes and tRNA.

Methylation interference experiments identify bases that are essential for distinct catalytic functions of a group I ribozyme.

Methylation interference experiments reveal bases involved in three different catalytic functions of the T4-phage derived sunY self-splicing intron. RNA molecules methylated at the N-7 position of the guanine at the cofactor binding site are inactive in cofactor-dependent splicing and 3' splice-site hydrolysis. In contrast, 5' splice-site hydrolysis occurs despite methylation at this position. Specific adenines that have been implicated in docking of the P1 stem to the catalytic core are shown to be important for cofactor-dependent splicing and essential for 5' splice-site hydrolysis. Similarly, methylation of bases in the P9.0 stem, as well as C56 in J5/4, interferes with 3' splice-site hydrolysis and with the splicing reaction. All of the bases identified as important for the overall splicing reaction are also identified as essential for either the 5' or 3' splice-site hydrolysis reactions, and vice versa. It is inferred that the bases implicated in 5' and 3' splice-site hydrolysis are involved in specific interactions of the 5' and 3' splice site, respectively, with the catalytic core.

Footprinting the sites of interaction of antibiotics with catalytic group I intron RNA.

Aminoglycoside inhibitors of translation have been shown previously to inhibit in vitro self-splicing by group I introns. Chemical probing of the phage T4-derived sunY intron shows that neomycin, streptomycin, and related antibiotics protected the N-7 position of G96, a universally conserved guanine in the binding site for the guanosine cofactor in the splicing reaction. The antibiotics also disrupted structural contacts that have been proposed to bring the 5' cleavage site of the intron into proximity to the catalytic core. In contrast, the strictly competitive inhibitors deoxyguanosine and arginine protected only the N-7 position of G96. Parallels between these results and previously observed protection of 16S ribosomal RNA by aminoglycosides raise the possibility that group I intron splicing and transfer RNA selection by ribosomes involve similar RNA structural motifs.

RNA as a catalyst: natural and designed ribozymes.

RNA can catalyse chemical reactions through its ability to fold into complex three-dimensional structures and to specifically bind small molecules and divalent metal ions. The 2'-hydroxyl groups of the ribose moieties contribute to this exceptional reactivity of RNA, compared to DNA. RNA is not only able to catalyse phosphate ester transfer reactions in ribonucleic acids, but can also show amino-acyl esterase activity, and is probably able to promote peptide bond formation. Bearing its potential for functioning both as a genome and as a gene product, RNA is suitable for in vitro evolution experiments enabling the selection of molecules with new properties. The growing repertoire of RNA catalysed reactions will establish RNA as a primordial molecule in the evolution of life.

Lead cleavage sites in the core structure of group I intron-RNA.

Self-splicing of group I introns requires divalent metal ions to promote catalysis as well as for the correct folding of the RNA. Lead cleavage has been used to probe the intron RNA for divalent metal ion binding sites. In the conserved core of the intron, only two sites of Pb2+ cleavage have been detected, which are located close to the substrate binding sites in the junction J8/7 and at the bulged nucleotide in the P7 stem. Both lead cleavages can be inhibited by high concentrations of Mg2+ and Mn2+ ions, suggesting that they displace Pb2+ ions from the binding sites. The RNA is protected from lead cleavage by 2'-deoxyGTP, a competitive inhibitor of splicing. The two major lead induced cleavages are both located in the conserved core of the intron and at phosphates, which had independently been demonstrated to interact with magnesium ions and to be essential for splicing. Thus, we suggest that the conditions required for lead cleavage occur mainly at those sites, where divalent ions bind that are functionally involved in catalysis. We propose lead cleavage analysis of functional RNA to be a useful tool for mapping functional magnesium ion binding sites.

Non-competitive inhibition of group I intron RNA self-splicing by aminoglycoside antibiotics.

Aminoglycoside antibiotics inhibit self-splicing of group I intron RNA in vitro at concentrations as low as 10(-6) M. The sites of interaction and the mechanism of inhibition have yet to be determined. A comparative study of inhibition by different 2-deoxystreptamine analogues reveals structural features of the aminoglycoside antibiotics required for their interaction and effect on group I introns. Complete antibiotic inhibition of the two steps of splicing was not reversed at high concentrations of guanosine, indicating a non-competitive inhibition. A mutant group I intron in which the conserved guanosine nucleotide of the G-binding site had been replaced by an adenosine, was sensitive to the antibiotics providing direct evidence that the antibiotics do not interact with the G-binding site in the same way as the guanine base. In addition kinetic analyses of the self-splicing process in the presence of antibiotic inhibitors supported a non-competitive mechanism of the mixed type for inhibition of the ribozyme.

Evolution of secondary metabolite production: potential roles for antibiotics as prebiotic effectors of catalytic RNA reactions.

It has been proposed that organic molecules related to known secondary metabolites have existed since the beginning of biochemical evolution and were present in primordial soups. Under primitive earth conditions certain of these molecules may have played roles as effectors in prebiotic reactions, especially those involving catalytic RNA (ribozymes). We demonstrate that a number of antibiotic-related secondary metabolites bind to group I introns and either inhibit splicing reactions or promote the formation of intron oligomers. This is consistent with the functional co-evolution of catalytic RNA and secondary metabolites as antibiotic inhibitors of translation, and supports the notion of an evolutionary relationship between group I introns and ribosomal RNA.

Antibiotic inhibition of group I ribozyme function.

The discovery of catalytically active RNA has provided the basis for the evolutionary concept of an RNA world. It has been proposed that during evolution the functions of ancient catalytic RNA were modulated by low molecular weight effectors, related to antibiotics, present in the primordial soup. Antibiotics and RNA may have coevolved in the formation of the modern ribosome. Here we report that a set of aminoglycoside antibiotics, which are known to interact with the decoding region of the 16S ribosomal RNA of Escherichia coli, inhibit the second step of splicing of the T4 phage-derived td intron. Thus catalytic RNA seems to interact not only with a mononucleotide and an amino acid, but also with another class of biomolecules, the sugars. Splicing of other group I introns but not group II introns was inhibited. The similarity in affinity and specificity of these antibiotics for group I introns and rRNAs may result from recognition of evolutionarily conserved structures.

Streptomycin inhibits splicing of group I introns by competition with the guanosine substrate.

Streptomycin is an aminocyclitol glycoside antibiotic, which interferes with prokaryotic protein synthesis by interacting with the ribosomal RNA. We report here that streptomycin is also able to inhibit self splicing of the group I intron of the thymidylate synthase gene of phage T4. The inhibition is kinetically competitive with the substrate guanosine. Streptomycin and guanosine have in common a guanidino group, which has been shown to undergo hydrogen bonds with the ribozyme (Bass & Cech, Biochemistry, 25, 1986, 4473). The inhibitory effect of streptomycin extends to other group I introns, but does not affect group II introns. Mutating the bulged nucleotide in the conserved P7 secondary structure element of the td intron alters the affinity of the ribozyme for both guanosine and streptomycin. Myomycin, an antibiotic with similar effects on protein synthesis as streptomycin, is also able to inhibit splicing. In contrast, bluensomycin, which is structurally related to streptomycin, but contains only one guanidino group does not inhibit splicing. We discuss these findings in support of an evolutionary model that stresses the antiquity of antibiotics (J. Davies, Molecular Microbiology 4, 1990, 1227).

Effects of mutations of the bulged nucleotide in the conserved P7 pairing element of the phage T4 td intron on ribozyme function.

The P7 element of group I introns contains a semiconserved "bulged" nucleotide, a C in group IA introns (nt 870 in the td intron) and an A in group IB introns [Cech, T.R. (1988) Gene 73, 259-271]. Variants U870, G870, and A870, isolated by a combination of in vitro and in vivo genetic strategies, indicate that C and A at position 870 are consistent with splicing whereas U and G are not. Although mutants G870 and U870 could be activated in vitro by increasing the Mg2+ concentration, their Km for GTP at pH 7 was 20-100-fold elevated, and they were unable to undergo site-specific hydrolysis. The dependence of the mutants on high guanosine concentrations could be substantially overcome by an increase in pH, suggesting that a tautomeric change, which makes U and G mimic C and A, is responsible for restoring function. In contrast to the striking Km effect, Vmax for the mutants differed by less than a factor of 2 from the wild type. Furthermore, streptomycin, an aminoglycoside antibiotic that competes with guanosine for its binding site, inhibited splicing of the U870 and G870 constructs at least as well as of the C870 and A870 variants, indicating that the guanosine-binding site of the mutants is proficient at interacting with a guanidino group. While our experiments argue against a hydrogen-bonding interaction between the C6-O of the cofactor and C4-NH2 of the bulged nucleotide, they are consistent with other models in which the C4-NH2 and/or N3 groups of the bulged C are involved in establishing an active ribozyme.

MRS3 and MRS4, two suppressors of mtRNA splicing defects in yeast, are new members of the mitochondrial carrier family.

When present in high copy number plasmids, the nuclear genes MRS3 and MRS4 from Saccharomyces cerevisiae can suppress the mitochondrial RNA splicing defects of several mit- intron mutations. Both genes code for closely related proteins of about Mr 32,000; they are 73% identical. Sequence comparisons indicate that MRS3 and MRS4 may be related to the family of mitochondrial carrier proteins. Support for this notion comes from a structural analysis of these proteins. Like the ADP/ATP carrier protein (AAC), the mitochondrial phosphate carrier protein (PiC) and the uncoupling protein (UCP), the two MRS proteins have a tripartite structure; each of the three repeats consists of two hydrophobic domains that are flanked by specific amino acid residues. The spacing of these specific residues is identical in all domains of all proteins of the family, whereas spacing between the hydrophobic domains is variable. Like the AAC protein, the MRS3 and MRS4 proteins are imported into mitochondria in vitro and without proteolytic cleavage of a presequence and they are located in the inner mitochondrial membrane. In vivo studies support this mitochondrial localization of the MRS proteins. Overexpression of the MRS3 and MRS4 proteins causes a temperature-dependent petite phenotype; this is consistent with a mitochondrial function of these proteins. Disruption of these genes affected neither mitochondrial functions nor cellular viability. Their products thus have no essential function for mitochondrial biogenesis or for whole yeast cells that could not be taken over by other gene products. The findings are discussed in relation to possible functions of the MRS proteins in mitochondrial solute translocation and RNA splicing.