Prp18 functions in the second step of pre-mRNA splicing, joining the spliceosome just prior to the transesterification reaction that creates the mature mRNA. Prp18 interacts with Slu7, and the functions of the two proteins are intertwined. Using the X-ray structure of Prp18, we have designed mutants in Prp18 that imply that Prp18 has two distinct roles in splicing. Deletion mutations were used to delineate the surface of Prp18 that interacts with Slu7, and point mutations in Prp18 were used to define amino acids that contact Slu7. Experiments in which Slu7 and mutant Prp18 proteins were expressed at different levels support a model in which interaction between the proteins is needed for stable binding of both proteins to the spliceosome. Mutations in an evolutionarily conserved region show that it is critical for Prp18 function but is not involved in binding Slu7. Alleles with mutations in the conserved region are dominant negative, suggesting that the resulting mutant prp18 proteins make proper contacts with the spliceosome, but fail to carry out a Prp18-specific function. Prp18 thus appears to have two separable roles in splicing, one in stabilizing interaction of Slu7 with the spliceosome, and a second that requires the conserved loop.

Archaea-like bacteria are prokaryotes but, in contrast, use eukaryotic-like systems for key aspects of DNA, RNA, and protein metabolism. mRNA is typically unstable in bacteria and stable in eukaryotes, but little information is available about mRNA half-lives in archaea. Because archaea are generally insensitive to antibiotics, examination of mRNA stability in the hyperthermophile, Sulfolobus solfataricus, required the identification of transcription inhibitors for half-life determinations. An improved lacS promoter-dependent in vitro transcription system was used to assess inhibitor action. Efficient inhibitors were distinguished as blocking both lacSp transcription in vitro and the incorporation of 3H-uracil into bulk RNA in vivo. Actinomycin D was the most stable and potent compound identified. A survey of transcript chemical half-lives normalized to levels of the signal recognition particle 7S RNA ranged from at least 2 h for tfb1, a transcription factor TFIIB paralog, to a minimum of 6.3 min for gln1, one of three glutamine synthetase paralogs. Transcript half-lives for other mRNAs were: 2 h, superoxide dismutase (sod); 37.5 min, glucose dehydrogenase (dhg1); 25 min, alpha-glucosidase (malA); and 13.5 min, transcription factor TFIIB-2 (tfb2) resulting in a minimum average half-life of 54 min. These are the first mRNA half-lives reported for a hyperthermophile or member of the crenarchaea. The unexpected stability of several transcripts has important implications for gene expression and mRNA degradation in this organism.

Radioactively labeled 4.5S RNA containing statistically distributed 4-thiouridine residues in place of normal uridine was prepared by T7 transcription. The ability of this modified 4.5S RNA to form a complex with the protein Ffh was demonstrated by a gel shift assay. The modified 4.5S RNA, with or without Ffh, was added to Escherichia coli ribosomes under various conditions, and crosslinking from the thiouridine residues was induced by irradiation at 350 nm. The crosslinked ribosomal components were analyzed by our standard procedures. Two clearly defined types of crosslinking were observed. The first was a crosslink to 23S rRNA, which was entirely dependent both on the presence of Ffh and a nascent protein chain in the 50S subunit. This crosslink was localized to nt ∼ 2828–2837 of the 23S rRNA, from position 84 of the 4.5S molecule. The second type of crosslinking, to the 30S ribosomal subunit, was independent of the presence of Ffh, and was found both with vacant 70S ribosomes or isolated 30S subunits. Here the crosslink was localized to the 3′-terminal region of the 16S rRNA, from positions 29–50 of the 4.5S RNA. Crosslinking to ribosomal protein S1 was also observed. The known crystal structure of the protein Ffh/4.5S RNA fragment complex was extrapolated by computer modeling so as to include the whole 4.5S molecule, and this was docked onto the ribosome using the crosslinking data. The results are discussed in terms of multiple functions and binding sites of the 4.5S RNA.

A 5-methyluridine (m5U) residue at position 54 is a conserved feature of bacterial and eukaryotic tRNAs. The methylation of U54 is catalyzed by the tRNA(m5U54)methyltransferase, which in Saccharomyces cerevisiae is encoded by the nonessential TRM2 gene. In this study, we identified four different strains with mutant forms of tRNACGASer. The absence of the TRM2 gene in these strains decreased the stability of tRNACGASer and induced lethality. Two alleles of TRM2 encoding catalytically inactive tRNA(m5U54)methyltransferases were able to stabilize tRNACGASer in one of the mutants, revealing a role for the Trm2 protein per se in tRNA maturation. Other tRNA modification enzymes interacting with tRNACGASer in the maturation process, such as Pus4p, Trm1p, and Trm3p were essential or important for growth of the tRNACGASer mutants. Moreover, Lhp1p, a protein binding RNA polymerase III transcripts, was required to stabilize the mutant tRNAs. Based on our results, we suggest that tRNA modification enzymes might have a role in tRNA maturation not necessarily linked to their known catalytic activity.

We constructed two aptamers, each of which contains a 7-nt-long loop complementary to the anticodon loop of a suppressor tRNA. One of these aptamers can form a stable bimolecular complex with the suppressor tRNA in vitro and protects the 7 nt in the suppressor's anticodon loop from RNase S1. An Escherichia coli strain, carrying an amber mutation in the lac Z gene, produces β-galactosidase only if the suppressor is present; the aptamer's coexpression in the cell inhibits the activity of the suppressor tRNA. Moreover, in E. coli extract, the aptamer partially inhibits the read-through of the stop codon on the part of the suppressor tRNA. These results point to a novel strategy that need not be limited to the suppressor tRNA. By constructing appropriate inducible aptamers, it may well be possible to effectively control translation in vivo.

The internal ribosome entry site (IRES) of the hepatitis C virus (HCV) RNA is known to interact with the 40S ribosomal subunit alone, in the absence of any additional initiation factors or Met-tRNAi. Previous work from this laboratory on the 80S and 48S ribosomal initiation complexes involving the HCV IRES showed that stem-loop III, the pseudoknot domain, and some coding sequence were protected from pancreatic RNase digestion. Stem-loop II is never protected by these complexes. Furthermore, there is no prior evidence reported showing extensive direct binding of stem-loop II to ribosomes or subunits. Using direct analysis of RNase-protected HCV IRES domains bound to 40S ribosomal subunits, we have determined that stem-loops II and III and the pseudoknot of the HCV IRES are involved in this initial binding step. The start AUG codon is only minimally protected. The HCV-40S subunit binary complex thus involves recognition and binding of stem-loop II, revealing its role in the first step of a multistep initiation process that may also involve rearrangement of the bound IRES RNA as it progresses.

All mitochondrial tRNAs of the kinetoplastid protozoan Leishmania tarentolae are encoded in the nucleus and are imported from the cytosol into the mitochondrion. We previously reported the partitioning of five tRNAs and found that all were shared between the two compartments to different extents. To increase our knowledge of the tRNAs of this organism, and to attempt to understand the signals involved in their subcellular localization, a method to RT-PCR amplify new tRNAs was developed. Various tRNAs were 3′ polyadenylated and reverse transcribed with a sequence-tagged primer. The cDNA was tagged by ligation to an anchor oligonucleotide, and the resulting double-tagged cDNA was amplified by PCR. Four new tRNAs were obtained, bringing to 20 the total number of L. tarentolae tRNAs identified to date. The subcellular localization of 17 tRNAs was quantitatively analyzed by two-dimensional gel electrophoresis and northern hybridization. In general, the previously suggested operational classification of tRNAs into three groups (mainly cytosolic, mainly mitochondrial, and shared between the two compartments) is still valid, but the relative abundance of each tRNA in the cytosol or mitochondrion varied greatly as did the level of expression.

Recent studies on endogenous SsrA-tagged proteins have revealed that the tagging could occur at a position corresponding to the normal termination codon. During the study of SsrA-mediated LacI tagging (Abo et al., EMBO J, 2000 19:3762–3769), we found that a variant LacI (LacIΔC1) lacking the last C-terminal amino acid residue is efficiently tagged in a stop codon-dependent manner. SsrA tagging of LacIΔC1 occurred efficiently without LacI binding to the lac operators at any one of three stop codons. The C-terminal (R)LESG peptide of LacIΔC1 was shown to trigger the SsrA tagging of an unrelated protein (CRP) when fused to its C terminus. Mass spectrometry analysis of the purified fusion proteins revealed that SsrA tagging occurs at a position corresponding to the termination codon. The alteration of the amino acid sequence but not the nucleotide sequence of the C-terminal portion eliminated the tagging. We also showed that the tagging-provoking sequences cause an efficient translational readthrough at UGA but not UAA codons. In addition, we found that C-terminal dipeptides known to induce an efficient translation readthrough could cause an efficient tagging at stop codons. We conclude that the amino acid sequence of nascent polypeptide prior to stop codons is a major determinant for the SsrA tagging at all three stop codons.

The α-globin mRNA contains a C-rich stability element (CRE) in its 3′ untranslated region (3′ UTR) which is critical for the stability of this long-lived mRNA. A protein complex, termed the α-complex, forms on the CRE and has been shown to contribute to stabilization of the mRNA by at least two mechanisms, first by interacting with the poly(A)-binding protein (PABP) to prevent deadenylation, and second by protecting the mRNA from attack by an erythroid endoribonuclease. In this report, we demonstrate that the α-globin 3′ UTR can confer stability on a heterologous mRNA in cells, and this stability is dependent on the α-complex. Moreover, the stability was exclusively detected with cytoplasmic mRNA, suggesting that the regulation of α-globin mRNA stability is a cytoplasmic event. An additional mechanism by which the α-complex can confer stability on an RNA in vitro was also identified and shown to involve inhibition of 3′ to 5′ exonucleolytic degradation. Furthermore, using an in vitro mRNA decay system, we were able to follow the demise of the α-globin RNA and demonstrate that the decay was initiated by deadenylation followed by 3′-to-5′ decay carried out by the exosome and ultimately hydrolysis of the residual cap structure.

The Ran protein regulates nucleocytoplasmic transport mediated by the karyopherin family of nuclear transport factors. Ran is converted to the active, GTP bound form in the nucleus and then binds to a conserved domain found in all karyopherins. This interaction induces cargo binding for exportins and cargo release for importins. In either case, the Ran·GTP is then transported to the cytoplasm by the karyopherin, where it is hydrolyzed to Ran·GDP. To ask whether Ran could function as a nuclear mRNA export factor, we fused Ran to the MS2 coat protein and inserted MS2 RNA-binding sites into an unspliced cat mRNA that is normally sequestered in the nucleus. Coexpression of MS2-Ran induced cat mRNA export and CAT enzyme expression as effectively as, for example, an MS2-Rev fusion protein. MS2-Ran dependent nuclear mRNA export was reduced by inhibitors specific for Crm1, but not blocked as was seen with MS2-Rev. Consistent with the hypothesis that Crm1 is not the only karyopherin cofactor for MS2-Ran mediated mRNA export, we show that not only Crm1 but also CAS, transportin, importin β and exportin t can all export mRNA from the nucleus when tethered via the MS2 RNA-binding domain. In contrast, two shuttling hnRNPs, hnRNP A1 and hnRNP K, proved unable to function as nuclear RNA export factors when expressed as MS2 fusions. Together, these data argue that karyopherins that normally function to transport proteins into or out of the nucleus are also capable of exporting tethered mRNA molecules.

Patients with systemic autoimmune disorders produce autoantibodies against sequence-specific conformational RNA epitopes on U1 snRNA, 28S rRNA, and transfer RNAs. The molecular basis for immunological reactivity with these highly abundant and stable RNAs is not understood. Here, we report the existence of discrete RNA epitopes in messenger RNAs that are generally less abundant and less stable than snRNAs and tRNAs. An iterative selection and amplification procedure using pooled autoimmune patient sera identified immunoreactive mRNA species. Following deconvolution of the pools to identify the reactive sera, several mRNAs recognized by these autoantibodies were cloned and sequenced. Detailed analysis using one particular serum indicated reactivity against the messages encoding alternative splicing factor (ASF/SF2) and calmodulin. Deletion and site-directed mutagenesis determined that an epitope recognized by this serum is located in a 17-base stem-loop structure common to both messages. This serum was then used to immunoprecipitate native mRNAs encoding ASF/SF2 and calmodulin from total HeLa cell RNA. Our results demonstrate that despite its low abundance and instability, messenger RNA is capable of reacting with autoantibodies generated during an autoimmune response. These data are consistent with direct presentation as a model to explain the generation of RNA conformation-specific autoantibodies.

The stability of a trans-acting delta ribozyme was studied under various conditions. Although in vitro (i.e., in the presence of protein extracts) this delta ribozyme appears to be only slightly more stable than a hammerhead ribozyme, in vivo (i.e., after cell transfection) it exhibits an outstanding stability that manifests itself in the calculated half-life of over 100 h regardless of the means of transfection. The P2 stem, which includes both the 5′ and 3′ ends, is shown to play a critical role in this stability. Direct mutagenesis of the most nuclease susceptible nucleotides failed to generate a more stable ribozyme that retained the same catalytic potential. Clearly, delta ribozyme appears to be well adapted to the human cell environment, and is therefore ideal for the development of a gene-inactivation system.

Group II introns are self-splicing RNAs that also act as retroelements in bacteria, mitochondria, and chloroplasts. Group II introns were identified in Escherichia coli in 1994, but have not been characterized since, and, instead, other bacterial group II introns have been studied for splicing and mobility properties. Despite their apparent intractability, at least five distinct group II introns exist naturally in E. coli strains. To illuminate their function and learn how the introns have dispersed in their natural host, we have investigated their distribution in the ECOR reference collection. Two introns were cloned and sequenced to complete their partial sequences. Unexpectedly, southern blots showed all ECOR strains to contain fragments and/or full-length copies of group II introns, with some strains containing up to 15 intron copies. One intron, E.c.I4, has two natural homing sites in IS629 and IS911 elements, and the intron can be present in one, both, or neither homing site in a given strain. Nearly all strains that contain full-length introns also contain unfilled homing sites, suggesting either that mobility is highly inefficient or that most full-length copies are nonfunctional. The data indicate independent mobility of the introns, as well as mobility via the host DNA elements, and overall, the pattern of intron distribution resembles that of IS elements.

Translation of the hepatitis C virus genomic RNA is mediated by an internal ribosome entry site (IRES). The 330-nt IRES RNA forms a binary complex with the small 40S ribosomal subunit as a first step in translation initiation. Here chemical probing and 4-thiouridine-mediated crosslinking are used to characterize the interaction of the HCV IRES with the HeLa 40S subunit. No IRES-18S rRNA contacts were detected, but several specific crosslinks to 40S ribosomal proteins were observed. The identity of the crosslinked proteins agrees well with available structural information and provides new insights into HCV IRES function. The protein-rich surface of the 40S subunit thus mediates the IRES–ribosome interaction.

The dsRNA-activated protein kinase PKR is involved in signal transduction pathways that mediate cellular processes as diverse as cell growth and differentiation, the stress response, and apoptosis. PKR was originally described as an interferon-inducible eIF2α kinase involved in the antiviral defense mechanism of the cell. The interaction of the kinase with specific viral RNAs has been studied in much detail, but information about cellular mRNAs, which are able to bind and activate PKR, is scarce. In search for such cellular mRNAs, we developed a cloning strategy to identify individual mRNA species from the dsRNA-rich fraction of Daudi cell poly(A)+ RNA. Two out of five cDNA clones we obtained contained sequences derived from the mRNA of the translationally controlled tumor protein P23/TCTP, indicating that this mRNA is present in the dsRNA-rich fraction. Secondary structure predictions and gel electrophoretic mobility investigations on P23/TCTP transcripts confirmed the potential of this mRNA to form extensive secondary structure. A full-length P23 transcript, but not a truncated version thereof, was able to bind to PKR in vitro and in vivo. Transient transfection experiments in human 293 cells showed that coexpression of full-length P23 mRNA leads to partial inhibition of the expression of a β-galactosidase reporter gene in trans. Additional coexpression of a dominant negative mutant of PKR or of adenovirus VA1 RNA suppressed this inhibition, indicating that it is mediated by PKR. Studies on P23/TCTP expression in cells from PKR-knockout mice suggest that P23/TCTP mRNA translation is regulated by PKR. Hence, our results demonstrate that the mRNA of P23/TCTP may both activate PKR and be subject to translational regulation by this kinase.

The mRNA poly(A) tail serves different purposes, including the facilitation of nuclear export, mRNA stabilization, efficient translation, and, finally, specific degradation. The posttranscriptional addition of a poly(A) tail depends on sequence motifs in the 3′ untranslated region (3′ UTR) of the mRNA and a complex trans-acting protein machinery. In this study, we have replaced the 3′ UTR of the yeast TRP4 gene with sequences encoding a hammerhead ribozyme that efficiently cleaves itself in vivo. Expression of the TRP4-ribozyme allele resulted in the accumulation of a nonpolyadenylated mRNA. Cells expressing the TRP4-ribozyme mRNA showed a reduced growth rate due to a reduction in Trp4p enzyme activity. The reduction in enzyme activity was not caused by inefficient mRNA export from the nucleus or mRNA destabilization. Rather, analyses of mRNA association with polyribosomes indicate that translation of the ribozyme-containing mRNA is impaired. This translational defect allows sufficient synthesis of Trp4p to support growth of trp4 cells, but is, nevertheless, of such magnitude as to activate the general control network of amino acid biosynthesis.

We identified a novel gene, Ks-1, which is expressed preferentially in the small-type Kenyon cells of the honeybee brain. This gene is also expressed in some of the large soma neurons in the brain and in the suboesophageal ganglion. Reverse transcription–polymerase chain reaction experiments indicated that Ks-1 transcripts are enriched in the honeybee brain. cDNA cloning revealed that the consensus Ks-1 cDNA is over 17 kbp and contains no significant open reading frames. Furthermore, fluorescent in situ hybridization revealed that Ks-1 transcripts are located in the nuclei of the neural cells, accumulating in some scattered spots. These findings demonstrate that Ks-1 encodes a novel class of noncoding nuclear RNA and is possibly involved in the regulation of neural functions.

There are two major components of Escherichia coli ribosomes directly involved in selection and binding of mRNA during initiation of protein synthesis—the highly conserved 3′ end of 16S rRNA (aSD) complementary to the Shine–Dalgarno (SD) domain of mRNA, and the ribosomal protein S1. A contribution of the SD-aSD and S1-mRNA interactions to translation yield in vivo has been evaluated in a genetic system developed to compare efficiencies of various ribosome-binding sites (RBS) in driving β-galactosidase synthesis from the single-copy (chromosomal) lacZ gene. The in vivo experiments have been supplemented by in vitro toeprinting and gel-mobility shift assays. A shortening of a potential SD-aSD duplex from 10 to 8 and to 6 bp increased the β-galactosidase yield (four- and sixfold, respectively) suggesting that an extended SD-aSD duplex adversely affects translation, most likely due to its redundant stability causing ribosome stalling at the initiation step. Translation yields were significantly increased upon insertion of the A/U-rich S1 binding targets upstream of the SD region, but the longest SD remained relatively less efficient. In contrast to complete 30S ribosomes, the S1-depleted 30S particles have been able to form an extended SD-aSD duplex, but not the true ternary initiation complex. Taken together, the in vivo and in vitro data allow us to conclude that S1 plays two roles in translation initiation: It forms an essential part of the mRNA-binding track even when mRNA bears a long SD sequence, and through the binding to the 5′ untranslated region, it can ensure a substantial enhancing effect on translation.

The C-to-U editing of apolipoprotein-B (apo-B) mRNA is catalyzed by an enzyme complex that recognizes an 11-nt mooring sequence downstream of the editing site. A minimal holoenzyme that edits apo-B mRNA in vitro has been defined. This complex contains apobec-1, the catalytic subunit, and apobec-1 complementation factor (ACF), the RNA-binding subunit that binds to the mooring sequence. Here, we show that ACF binds with high affinity to single-stranded but not double-stranded apo-B mRNA. ACF contains three nonidentical RNA recognition motifs (RRM) and a unique C-terminal auxiliary domain. In many multi-RRM proteins, the RRMs mediate RNA binding and an auxiliary domain functions in protein–protein interactions. Here we show that ACF does not fit this simple model. Based on deletion mutagenesis, the RRMs in ACF are necessary but not sufficient for binding to apo-B mRNA. Amino acids in the pre-RRM region are required for complementing activity and RNA binding, but not for interaction with apobec-1. The C-terminal 196 amino acids are not absolutely essential for function. However, further deletion of an RG-rich region from the auxiliary domain abolished complementing activity, RNA binding, and apobec-1 interaction. The auxiliary domain alone did not bind apobec-1. Although all three RRMs are required for complementing activity and apobec-1 interaction, the individual motifs contribute differently to RNA binding. Point mutations in RRM1 or RRM2 decreased the Kd for apo-B mRNA by two orders of magnitude whereas mutations in RRM3 reduced binding affinity 13-fold. The pairwise expression of RRM1 with RRM2 or RRM3 resulted in moderate affinity binding.

The initiation of enteroviral positive-strand RNA synthesis requires the presence of a functional ribonucleoprotein complex containing a cloverleaf-like RNA secondary structure at the 5′ end of the viral genome. Other components of the ribonucleoprotein complex are the viral 3CD proteinase (the precursor protein of the 3C proteinase and the 3D polymerase), the viral 3AB protein and the cellular poly(rC)-binding protein 2. For a molecular characterization of the RNA-binding properties of the enteroviral proteinase, the 3C proteinase of coxsackievirus B3 (CVB3) was bacterially expressed and purified. The recombinant protein is proteolytically active and forms a stable complex with in vitro-transcribed cloverleaf RNA of CVB3. The formation of stable complexes is also demonstrated with cloverleaf RNA of poliovirus (PV) 1, the first cloverleaf of bovine enterovirus (BEV) 1, and human rhinovirus (HRV) 2 but not with cloverleaf RNA of HRV14 and the second cloverleaf of BEV1. The apparent dissociation constants of the protein:RNA complexes range from approx. 1.7 to 4.6 μM. An electrophoretic mobility shift assay with subdomain D of the CVB3 cloverleaf demonstrates that this RNA is sufficient to bind the CVB3 3C proteinase. Binding assays using mutated versions of CVB3 and HRV14 cloverleaf RNAs suggest that the presence of structural features rather than a defined sequence motif of loop D are important for 3C proteinase–RNA interaction.