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.

The Pescadillo protein was identified via a developmental defect and implicated in cell cycle progression. Here we report that human Pescadillo and its yeast homolog (Yph1p or Nop7p) are localized to the nucleolus. Depletion of Nop7p leads to nuclear accumulation of pre-60S particles, indicating a defect in subunit export, and it interacts genetically with a tagged form of the ribosomal protein Rpl25p, consistent with a role in subunit assembly. Two pre-rRNA processing pathways generate alternative forms of the 5.8S rRNA, designated 5.8SL and 5.8SS. In cells depleted for Nop7p, the 27SA3 pre-rRNA accumulated, whereas later processing intermediates and the mature 5.8SS rRNA were depleted. Less depletion was seen for the 5.8SL pathway. TAP-tagged Nop7p coprecipitated precursors to both 5.8SL and 5.8SS but not the mature rRNAs. We conclude that Nop7p is required for efficient exonucleolytic processing of the 27SA3 pre-rRNA and has additional functions in 60S subunit assembly and transport. Nop7p is a component of at least three different pre-60S particles, and we propose that it carries out distinct functions in each of these complexes.

We compared the gene expression patterns of fetal and adult porcine brains and identified a sequence tag that was more abundant in adult than in fetal brain. The RNA corresponding to the sequence tag has the highest expression level in adult cerebellum. Lower expression levels of the transcript were found in adult cerebrum, pituitary, and uterus, as well as in fetal brain, heart, intestine, kidney, and liver. The sequence tag was used to screen a cDNA library from adult porcine brain. Two independent clones of 2,273 nt and 1,701 nt were isolated. The shorter cDNA is a 5′-truncated form of the longer clone, and both clones have almost identical sequences with multiple start and stop codons in all three reading frames. Screening of two different human brain cDNA libraries with porcine cDNA probes resulted in four overlapping cDNA fragments, which were assembled to one contig of 2,336 nt in length. Like noncoding RNAs, the porcine and human sequences have no common conserved open reading frame and share stretches of high homology interrupted by stretches with almost no homology. The human and porcine RNAs were named UM 9(5)h and UM 9(5)p, respectively. They are part of larger transcripts, which are transcribed from single-copy genes, they have very similar tissue distributions, and their sequences are colinear with the respective genomic fragment.

The human La (SS-B) autoantigen is an abundantly expressed putative RNA chaperone, functioning in various intracellular processes involving RNA. To further explore the molecular mechanisms by which La functions in these processes, we performed large-scale immunoprecipitations of La from HeLa S100 extracts using the anti-La monoclonal antibody SW5. La-associated proteins were subsequently identified by sequence analysis. This approach allowed the identification of DDX15 as a protein interacting with La. DDX15, the human ortholog of yeast Prp43, is a member of the superfamily of DEAH-box RNA helicases that appeared to interact with La both in vivo and in vitro. The region needed for the interaction with La partly overlaps the DEAH-box domain of DDX15. Immunofluorescence data indicated that endogenous DDX15 accumulates in U snRNP containing nuclear speckles in HEp-2 cells. Surprisingly DDX15 also accumulates in the nucleoli of HEp-2 cells. Moreover, DDX15 and La seem to colocalize in the nucleoli. Regions of DDX15 involved in nuclear, nuclear speckle, and nucleolar localization are located within the N- and C-terminal regions flanking the DEAH-box. RNA coprecipitation experiments indicated that DDX15 is associated with spliceosomal U small nuclear RNAs in HeLa cell extracts. The possible functional implications of the interaction between La and DDX15 are discussed.

The crystal structure of ribosomal protein L5 from Thermus thermophilus complexed with a 34-nt fragment comprising helix III and loop C of Escherichia coli 5S rRNA has been determined at 2.5 Å resolution. The protein specifically interacts with the bulged nucleotides at the top of loop C of 5S rRNA. The rRNA and protein contact surfaces are strongly stabilized by intramolecular interactions. Charged and polar atoms forming the network of conserved intermolecular hydrogen bonds are located in two narrow planar parallel layers belonging to the protein and rRNA, respectively. The regions, including these atoms conserved in Bacteria and Archaea, can be considered an RNA–protein recognition module. Comparison of the T. thermophilus L5 structure in the RNA-bound form with the isolated Bacillus stearothermophilus L5 structure shows that the RNA-recognition module on the protein surface does not undergo significant changes upon RNA binding. In the crystal of the complex, the protein interacts with another RNA molecule in the asymmetric unit through the β-sheet concave surface. This protein/RNA interface simulates the interaction of L5 with 23S rRNA observed in the Haloarcula marismortui 50S ribosomal subunit.

The 3′ end of replication-dependent histone mRNAs terminate in a conserved sequence containing a stem-loop. This 26-nt sequence is the binding site for a protein, stem-loop binding protein (SLBP), that is involved in multiple aspects of histone mRNA metabolism and regulation. We have determined the structure of the 26-nt sequence by multidimensional NMR spectroscopy. There is a 16-nt stem-loop motif, with a conserved 6-bp stem and a 4-nt loop. The loop is closed by a conserved U[bull ]A base pair that terminates the canonical A-form stem. The pyrimidine-rich 4-nt loop, UUUC, is well organized with the three uridines stacking on the helix, and the fourth base extending across the major groove into the solvent. The flanking nucleotides at the base of the hairpin stem do not assume a unique conformation, despite the fact that the 5′ flanking nucleotides are a critical component of the SLBP binding site.

The crystal structure of the 19-mer RNA, 5′-GAAUGCCUGCGAGCAUCCC-3′ has been determined from X-ray diffraction data to 1.6 Å resolution by the multiwavelength anomalous diffraction method from crystals containing a brominated uridine. In the crystal, this RNA forms an 18-mer self-complementary double helix with the 19th nucleotide flipped out of the helix. This helix contains most of the target stem recognized by the bacteriophage Mu Com protein (control of mom), which activates translation of an unusual DNA modification enzyme, Mom. The 19-mer duplex, which contains one A[bull ]C mismatch and one A[bull ]C/G[bull ]U tandem wobble pair, was shown to bind to the Com protein by native gel electrophoresis shift assay. Comparison of the geometries and base stacking properties between Watson–Crick base pairs and the mismatches in the crystal structure suggest that both hydrogen bonding and base stacking are important for stabilizing these mismatched base pairs, and that the unusual geometry adopted by the A[bull ]C mismatch may reveal a unique structural motif required for the function of Com.

In maize plastids, transcripts are known to be modified at 27 C-to-U RNA editing sites, affecting the expression of 15 different genes. The relative contribution of editing efficiency versus transcript abundance in regulation of chloroplast gene expression has previously been analyzed for only a few genes. We undertook a comprehensive analysis of the editing efficiency of each of the 27 maize editing sites in 10 different maize tissues, which contain a range of plastid types including chloroplasts, etioplasts, and amyloplasts. Using a reproducible poisoned primer extension assay, we detected variation between RNA editing extent of different sites in the same transcript in the same tissue, and between the same site in different tissues. The most striking editing deficiency is in an editing site in ndhB that is edited at only 8% and 1% in roots and callus plastids respectively, whereas green leaf chloroplasts edit this site at 100%. Editing efficiencies of some sites are not affected by the developmental stages we examined and are always edited close to 80–100%. The relative amounts of transcripts of each of the 10 genes that exhibited variable editing extents were determined by real-time PCR. Seven genes exhibited over 100 times lower transcript abundance in either roots or tissue-cultured cells relative to green leaf tissue. The quantitative analysis indicates that a particular editing site can be efficiently edited over a large range of transcript abundance, resulting in no general correlation of transcript abundance and editing extent. The independent variation of editing efficiency of different sites within the same transcript fits with a model that postulates individual trans-acting factors specific to each editing site. Because tissues where editing efficiency at certain sites is low invariably also exhibited greatly decreased abundance of the transcripts carrying those sites, decrease in the amounts of particular RNAs rather than a lack of editing is predicted to have the most significant impact on gene expression under steady-state conditions. Our data is consistent with the hypothesis that the role of editing in angiosperm plastids is to correct otherwise detrimental mutations rather than to generate significant protein diversity.

Several cap-binding proteins from both the nucleus and cytosol have been identified that mediate processes such as pre-mRNA splicing, translation initiation, and mRNA turnover. Here we describe a novel cap-binding protein, pokeweed antiviral protein (PAP), a 29-kDa type I ribosome-inactivating protein (RIP) isolated from Phytolacca americana. In addition to depurinating the sarcin/ricin loop of the large rRNA, an activity common to all RIPs, we have reported recently that PAP depurinates capped, but not uncapped RNAs in vitro. Here we characterize this activity further and, using affinity chromatography, show that PAP binds to the m7Gppp cap structure. PAP UV-crosslinks to m7GpppG-capped luciferase mRNA more efficiently than GpppG-capped luciferase mRNA, indicating specificity for the methylated guanosine. We present evidence that PAP does not remove the cap structure or depurinate the m7Gppp as shown by primer extension of capped and uncapped luciferase transcripts incubated with PAP. Modeling studies of cap interaction with PAP predict that the cap structure would bind to the active site of PAP in a similar manner to guanine. We map the depurination sites on the capped luciferase RNA and illustrate that depurination occurs at specific adenine and guanine residues throughout the RNA sequence. Incubation of isolated ribosomes with PAP and increasing molar concentrations of m7GpppG relative to PAP resulted in a decrease in the level of rRNA depurination. Therefore, at elevated concentrations, the methylated cap structure competes with the adenine or guanine for binding to PAP, even though the affinity of PAP for capped message is almost fourfold lower than for rRNA. These results demonstrate that the activity of PAP is not limited to rRNA depurination, but that PAP binds to the cap structure and depurinates mRNAs downstream of the cap in vitro. These findings may have implications for understanding PAP activity in vivo.

Export of unspliced mRNA to the cytoplasm is required for the replication of all retroviruses. In simian type D retroviruses, the RNA export is mediated by the constitutive transport element (CTE) that binds the cellular nuclear export factor 1, NXF1(TAP). To search for potential cellular RNA substrates for NXF1, we have set up an in vitro selection procedure, using an RNA library expressed from total human genomic DNA. A sequence that was isolated most frequently as independent clones exhibits extensive homology to the 3′ untranslated region of expressed LINE1 (L1) retrotransposons. This region, termed L1-NXF1 binding element (L1-NBE) bears no structural resemblance to the viral CTE, but binds NXF1 as strongly as CTE, based on gel mobility shift competition assays. A deletion analysis of the NXF1 protein reveals that CTE and L1-NBE have different, but overlapping, binding domains on NXF1. Placed in an intron, L1-NBE is capable of mediating nuclear export of lariat RNA species in Xenopus laevis oocytes and of an unspliced HIV-1 derived RNA in human 293 cells, suggesting that it may function as a nuclear export element for the intronless L1 mRNA.

Selenoprotein synthesis in eukaryotes requires the selenocysteine insertion sequence (SECIS) RNA, a hairpin in the 3′ untranslated region of selenoprotein mRNAs. The SECIS RNA is recognized by the SECIS-binding protein 2 (SBP2), which is a key player in this specialized translation machinery. The objective of this work was to obtain structural insight into the SBP2-SECIS RNA complex. Multiple sequence alignment revealed that SBP2 and the U4 snRNA-binding protein 15.5 kD/Snu13p share the same RNA binding domain of the L7A/L30 family, also found in the box H/ACA snoRNP protein Nhp2p and several ribosomal proteins. In corollary, we have detected a similar secondary structure motif in the SECIS and U4 RNAs. Combining the data of the crystal structure of the 15.5 kD-U4 snRNA complex, and the SBP2/15.5 kD sequence similarities, we designed a structure-guided strategy predicting 12 SBP2 amino acids that should be critical for SECIS RNA binding. Alanine substitution of these amino acids followed by gel shift assays of the SBP2 mutant proteins identified four residues whose mutation severely diminished or abolished SECIS RNA binding, the other eight provoking intermediate down effects. In addition to identifying key amino acids for SECIS recognition by SBP2, our findings led to the proposal that some of the recognition principles governing the 15.5 kD-U4 snRNA interaction must be similar in the SBP2-SECIS RNA complex.

To explore possible mechanisms of intron-mediated enhancement of gene expression, the features of PAT1 intron 1 required to elevate mRNA accumulation were systematically tested in transgenic Arabidopsis. This intron is remarkably resilient, retaining some ability to increase mRNA accumulation when splicing was prevented by mutation of 5′ and 3′ splice sites, branchpoint sequences, or when intron U-richness was reduced. Enhancement was abolished by simultaneously eliminating branchpoints and the 5′ splice site, structures involved in the first two steps of spliceosome assembly. Although this suggests that the splicing machinery is required, intron splicing is clearly not enough to enhance mRNA accumulation. Five other introns were all efficiently spliced but varied widely in their ability to increase mRNA levels. Furthermore, PAT1 intron 1 was spliced but lost the ability to elevate mRNA accumulation when moved to the 3′ UTR. These findings demonstrate that splicing per se is neither necessary nor sufficient for an intron to enhance mRNA accumulation, and suggest a mechanism that requires intron recognition by the splicing machinery but also involves nonconserved intron sequences.

Maturation of the large subunit rRNAs includes a series of cleavages that result in removal of the internal transcribed spacer (ITS2) that separates mature 5.8S and 25/28S rRNAs. Previous work demonstrated that formation of higher order secondary structure within the assembling pre-ribosomal particle is a prerequisite for accurate and efficient pre-rRNA processing. To date, it is not clear which specific sequences or secondary structures are required for processing. Two alternative secondary structure models exist for Saccharomyces cerevisiae ITS2. Chemical and enzymatic structure probing and phylogenetic comparisons resulted in one structure (Yeh & Lee, J Mol Biol, 1990, 211:699–712) referred to here as the “hairpin model.” More recently, an alternate folded structure was proposed (Joseph et al., Nucleic Acids Res, 1999, 27:4533–4540), called here the “ring model.” We have used a functional genetic assay to examine the potential significance of these predicted structures in processing. Our data indicate that elements of both structural models are important in efficient processing. Mutations that prevent formation of ring-specific structures completely blocked production of mature 25S rRNA, whereas those that primarily disrupt hairpin elements resulted in reduced levels of mature product. Based on these results, we propose a dynamic conformational model for the role of ITS2 in processing: Initial formation of the ring structure may be required for essential, early events in processing complex assembly and may be followed by an induced transition to the hairpin structure that facilitates subsequent processing events. In this model, yeast ITS2 elements may provide in cis certain of the functions proposed for vertebrate U8 snoRNA acting in trans.

Slu7 and Prp18 act in concert during the second step of yeast pre-mRNA splicing. Here we show that the 382-amino-acid Slu7 protein contains two functionally important domains: a zinc knuckle (122CRNCGEAGHKEKDC135) and a Prp18-interaction domain (215EIELMKLELY224). Alanine cluster mutations of 215EIE217 and 221LELY224 abrogated Slu7 binding to Prp18 in a two-hybrid assay and in vitro, and elicited temperature-sensitive growth phenotypes in vivo. Yet, the mutations had no impact on Slu7 function in pre-mRNA splicing in vitro. Single alanine mutations of zinc knuckle residues Cys122, His130, and Cys135 had no effect on cell growth, but caused Slu7 function during pre-mRNA splicing in vitro to become dependent on Prp18. Specifically, zinc knuckle mutants required Prp18 in order to bind to the spliceosome. Compound mutations in both Slu7 domains (e.g., C122A-EIE, H130A-EIE, and C135A-EIE) were lethal in vivo and abolished splicing in vitro, suggesting that the physical interaction between Slu7 and Prp18 is important for cooperation in splicing. Depletion/reconstitution studies coupled with immunoprecipitations suggest that second step factors are recruited to the spliceosome in the following order: Slu7 → Prp18 → Prp22. All three proteins are released from the spliceosome after step 2 concomitant with release of mature mRNA.

Hepatitis delta virus (HDV) infection of individuals infected with hepatitis B virus (HBV) is associated with more severe liver damage and an increased risk of fulminant disease. HDV is a single-stranded RNA virus that encodes a single protein, the delta antigen, which is expressed in two forms, small (S-HDAg) and large (L-HDAg). Here we show that although HDV ribonucleoproteins are mainly detected in the nucleus, they are also present in the cytoplasm of cells infected with HDV or transfected with HDV cDNA. Making use of an heterokaryon assay, we demonstrate that HDV ribonucleoproteins shuttle continuously between the nucleus and the cytoplasm. In the absence of HDV RNA, both forms of the delta antigen are retained in the nucleus, whereas in the absence of the delta antigen, HDV RNA is predominantly detected in the cytoplasm. Coexpression of HDV RNA and S-HDAg (which binds to the viral RNA and contains a nuclear localization signal) results in nuclear accumulation of the viral RNA. This suggests that HDV RNA mediates export of viral particles to the cytoplasm whereas the delta antigen triggers their reimport into the nucleus.

Posttranscriptional modifications were mapped in helices 90–92 of 23S rRNA from the following phylogenetically diverse organisms: Haloarcula marismortui, Sulfolobus acidocaldarius, Bacillus subtilis, and Bacillus stearothermophilus. Helix 92 is a component of the ribosomal A-site, which contacts the aminoacyl-tRNA during protein synthesis, implying that posttranscriptional modifications in helices 90–92 may be important for ribosome function. RNA fragments were isolated from 23S rRNA by site-directed RNase H digestion. A novel method of mapping modifications by analysis of short, nucleotide-specific, RNase digestion fragments with Matrix Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI-MS) was utilized. The MALDI-MS data were complemented by two primer extension techniques using reverse transcriptase. One technique utilizes decreasing concentrations of deoxynucleotide triphosphates to map 2′-O-ribose methylations. In the other, the rRNA is chemically modified, followed by mild alkaline hydrolysis to map pseudouridines (Ψs). A total of 10 posttranscriptionally methylated nucleotides and 6 Ψs were detected in the five organisms. Eight of the methylated nucleotides and one Ψ have not been reported previously. The distribution of modified nucleotides and their locations on the surface of the ribosomal peptidyl transferase cleft suggests functional importance.

Chronic hepatitis C virus (HCV) infections are a significant cause of morbidity and mortality worldwide. Interferon-α2b treatment, alone or in combination with ribavirin, eliminates HCV from some patients, but patients infected with HCV genotype 1 viruses are cured less frequently than patients infected with HCV genotype 2 or 3 viruses. We report that HCV mRNA was detected and destroyed by the interferon-regulated antiviral 2′-5′ oligoadenylate synthetase/ribonuclease L pathway present in cytoplasmic extracts of HeLa cells. Ribonuclease L cleaved HCV mRNA into fragments 200 to 500 bases in length. Ribonuclease L cleaved HCV mRNA predominately at UA and UU dinucleotides within loops of predicted stem-loop structures. HCV mRNAs from relatively interferon-resistant genotypes (HCV genotypes 1a and 1b) have fewer UA and UU dinucleotides than HCV mRNAs from more interferon-sensitive genotypes (HCV genotypes 2a, 2b, 3a, and 3b). HCV 2a mRNA, with 73 more UA and UU dinucleotides than HCV 1a mRNA, was cleaved by RNase L more readily than HCV 1a mRNA. In patients, HCV 1b mRNAs accumulated silent mutations preferentially at UA and UU dinucleotides during interferon therapy. These results suggest that the sensitivity of HCV infections to interferon therapy may correlate with the efficiency by which RNase L cleaves HCV mRNA.

High-affinity binding sites for the hnRNP A1 protein stimulate the use of a distal 5′ splice site in mammalian pre-mRNAs. Notably, strong A1-mediated shifts in splice site selection are not accompanied by equivalent changes in the assembly of U1 snRNP-containing complexes on competing 5′ splice sites. To explain the above results, we have proposed that an interaction between hnRNP A1 molecules bound to high-affinity sites loops out the internal 5′ splice site. Here, we present additional evidence in support of the looping out model. First, replacing A1 binding sites with sequences that can generate a loop through RNA duplex formation activates distal 5′ splice site usage in an equivalent manner. Second, increasing the distance between the internal 5′ splice site and flanking A1 binding sites does not compromise activation of the distal 5′ splice site. Similar results were obtained with pre-mRNAs carrying inverted repeats. Using a pre-mRNA containing only one 5′ splice site, we show that splicing is repressed when flanked by two high-affinity A1 binding sites or by inverted repeats, and that inactivation of the internal 5′ splice site is sufficient to elicit a strong increase in the use of the distal donor site. Our results are consistent with the view that the binding of A1 to high-affinity sites promotes loop formation, an event that would repress the internal 5′ splice site and lead to distal 5′ splice site activation.

RNA-binding proteins are involved in the regulation of many aspects of eukaryotic gene expression. Targeted interference with RNA–protein interactions could offer novel approaches to modulation of expression profiles, alteration of developmental pathways, and reversal of certain disease processes. Here we investigate a decoy strategy for the study of the αCP subgroup of KH-domain RNA-binding proteins. These poly(C)-binding proteins have been implicated in a wide spectrum of posttranscriptional controls. Three categories of RNA decoys to αCPs were studied: poly(C) homopolymers, native mRNA-binding sites, and a high-affinity structure selected from a combinatorial library. Native chemistry was found to be essential for αCP decoy action. Because αCP proteins are found in both the nucleus and cytoplasm, decoy cassettes were incorporated within both nuclear (U1 snRNA) and cytoplasmic (VA1 RNA) RNA frameworks. Several sequences demonstrated optimal decoy properties when assayed for protein-binding and decoy bioactivity in vitro. A subset of these transcripts was shown to mediate targeted inhibition of αCP-dependent translation when expressed in either the nucleus or cytoplasm of transfected cells. Significantly, these studies establish the feasibility of developing RNA decoys that can selectively target biologic functions of abundant and widely expressed RNA binding proteins.

Ribonuclease P (RNase P) is a ribonucleoprotein that requires magnesium ions to catalyze the 5′ maturation of transfer RNA. To identify interactions essential for catalysis, the properties of RNase P containing single sulfur substitutions for nonbridging phosphodiester oxygens in helix P4 of Bacillus subtilis RNase P were analyzed using transient kinetic experiments. Sulfur substitution at the nonbridging oxygens of the phosphodiester bond of nucleotide U51 only modestly affects catalysis. However, phosphorothioate substitutions at A49 and G50 decrease the cleavage rate constant enormously (300–4,000-fold for P RNA and 500–15,000-fold for RNase P holoenzyme) in magnesium without affecting the affinity of pre-tRNAAsp, highlighting the importance of this region for catalysis. Furthermore, addition of manganese enhances pre-tRNA cleavage catalyzed by B. subtilis RNase P RNA containing an SP phosphorothioate modification at A49, as observed for Escherichia coli P RNA [Christian et al., RNA, 2000, 6:511–519], suggesting that an essential metal ion may be coordinated at this site. In contrast, no manganese rescue is observed for the A49 Sp phosphorothioate modification in RNase P holoenzyme. These differential manganese rescue effects, along with affinity cleavage, suggest that the protein component may interact with a metal ion bound near A49 in helix P4 of P RNA.