In mammals, as in all organisms examined, mRNAs that prematurely terminate translation are abnormally reduced in abundance by a mechanism called nonsense-mediated mRNA decay (NMD) or mRNA surveillance (for reviews, see Maquat, 1995, 1996; Ruiz-Echevarria & Peltz, 1996; Li & Wilkinson, 1998; Culbertson, 1999; Hentze & Kulozik, 1999; Hilleren & Parker, 1999). This mechanism is thought to have evolved to eliminate nonsense-containing RNAs that arise as a consequence of (1) mutations in germ-line or somatic DNA or (2) routine abnormalities in gene expression due to abnormalities in, for example, transcription initiation, splicing, and somatic rearrangements of the type that characterize the immunoglobulin and T-cell receptor genes. The elimination of nonsense-containing mRNAs protects cells from the potentially deleterious effects of the encoded truncated proteins, which can manifest new or dominant-negative functions (Kazazian et al., 1992; Pulak & Anderson, 1993; Hall & Thein, 1994; Cali & Anderson, 1998). In addition to eliminating abnormal transcripts, NMD also regulates the expression of certain mRNAs that are not abnormal. Examples in mammalian cells are provided by certain selenoprotein mRNAs that terminate translation at a UGA codon for the inefficiently incorporated amino acid selenocysteine (Sec; Moriarty et al., 1997, 1998). Other examples will undoubtedly resemble natural substrates found in other organisms such as the alternatively spliced mRNAs of Caenorhabditis elegans that retain an internal exon harboring an in-frame termination codon (Morrison et al., 1997), and transcripts of Saccharomyces cerevisiae that contain a small open reading frame upstream of the primary open reading frame (Leeds et al., 1992; Pierrat et al., 1993).

The wobble GoU pairs have been implicated in several biological processes where RNA molecules play a key role. We review the geometrical and conformational properties of wobble GoU pairs on the basis of available crystal structures of RNAs at high resolution. The similarities with the wobble A+oC pairs and UoU pairs are illustrated, while the differences with the recently discovered bifurcated G[bull ]U pairs are contrasted.

Here we report further characterization of an in vitro assay system for exon ligation by the human spliceosome in which the 3′ splice site AG is supplied by a different RNA molecule than that containing the 5′ splice and branch sites. By varying the time during splicing reactions when the 3′ splice site AG is made available to the splicing machinery, we show that AG recognition need not occur until after lariat formation. Thus an early AG recognition event required for spliceosome formation and lariat formation on some mammalian introns is not required for exon ligation. Depletion/add-back studies and cold competitor challenge experiments reveal that commitment of a 3′ splice site AG to exon ligation requires NTP hydrolysis. Because it both physically and kinetically uncouples exon ligation from spliceosome assembly and lariat formation, the bimolecular system will be a valuable tool for further mechanistic analysis of the second step of splicing.

The organization of the 5′ terminus region in the 16S rRNA was investigated using a series of RNA constructs in which the 5′ terminus was extended by 5 nt or was shortened to give RNA molecules that started at positions −5, +1, +5, +8, +14, or +21. The structural and functional effects of the 5′ extension/truncations were determined after the RNAs were reconstituted. 30S subunits containing 16S rRNA with 5′ termini at −5, +1, +5, +8 and +14 had similar structures (judged by UV-induced crosslinking) and exhibited a gradual reduction in tRNA binding activity compared to that seen with 30S subunits reconstituted with native 16S rRNA. To create the 5′ terminal site-specific photocrosslinking agent, the reagent azidophenacylbromide (APAB) was attached to the 5′ terminus of 16S rRNA through a guanosine monophosphorothioate and the APA-16S rRNAs were reconstituted. Crosslinking carried out with the APA revealed sites in six regions around positions 300–340, 560, 900, 1080, the 16S rRNA decoding region, and at 1330. Differences in the pattern and efficiency of crosslinking for the different constructs allow distance estimates for the crosslinked sites from nucleotide G9. These measurements provide constraints for the arrangement of the RNA elements in the 30S subunit. Similar experiments carried out in the 70S ribosome resulted in a five- to tenfold lower frequency of crosslinking. This is most likely due to a repositioning of the 5′ terminus upon subunit association.

RNA polymerase II is implicated in the RNA-templated RNA synthesis during replication of viroids and Hepatitis Delta Virus (HDV); however, neither the RNA template nor protein factor requirements for this process are well defined. We have developed an in vitro transcription system based on HeLa cell nuclear extract (NE), in which a segment of antigenomic RNA corresponding to the left-hand tip region of the HDV rod-like structure serves as a template for efficient and highly specific RNA synthesis. Accumulation of the unique RNA product is highly sensitive to α-amanitin in HeLa NE and only partially sensitive to this drug in NE from PMG cells that contain an allele of the α-amanitin-resistant subunit of pol II, strongly suggesting pol II involvement in this reaction. Detailed analysis of the RNA product revealed that it represents a chimeric molecule composed of a newly synthesized transcript covalently attached to the 5′ half of the RNA template. Selection of the start site for transcription is remarkably specific and depends on the secondary structure of the RNA template, rather than on its primary sequence. Some features of this reaction resemble the RNA cleavage-extension process observed for pol II-arrested complexes in vitro. A possible involvement of the described reaction in HDV replication is discussed.

In higher plants, RNA–DNA interactions can trigger de novo methylation of genomic sequences via a process that is termed RNA-directed DNA methylation (RdDM). In potato spindle tuber viroid (PSTVd)-infected tobacco plants, this process can potentially lead to methylation of all C residues at symmetrical and nonsymmetrical sites within chromosomal inserts that consist of multimers of the 359-bp-long PSTVd cDNA. Using PSTVd cDNA subfragments, we found that genomic targets with as few as 30 nt of sequence complementarity to the viroid RNA are detected and methylated. Genomic sequencing analyses of genome-integrated 30- and 60-bp-long PSTVd subfragments demonstrated that de novo cytosine methylation is not limited to the canonical CpG, CpNpG sites. Sixty-base-pair-long PSTVd cDNA constructs appeared to be densely methylated in nearly all tobacco leaf cells. With the 30-bp-long PSTVd-specific construct, the proportion of cells displaying dense transgene methylation was significantly reduced, suggesting that a minimal target size of about 30 bp is necessary for RdDM. The methylation patterns observed for two different 60-bp constructs further suggested that the sequence identity of the target may influence the methylation mechanism. Finally, a link between viroid pathogenicity and PSTVd RNA-directed methylation of host sequences is proposed.

Human Ro ribonucleoproteins (RNPs) are autoantigenic particles of unknown function(s) that consist of a 60-kDa protein (Ro60) associated with one hY RNA (hY1–5). Using a modified yeast three-hybrid system, named RNP interaction trap assay (RITA), we cloned a novel Ro RNP-binding protein (RoBPI), based on its property to interact in vivo in yeast with an RNP complex made of recombinant Ro60 (rRo60) protein and hY5 (rhY5) RNA. RoBPI cDNA contains three conserved RNA recognition motifs (RRM) and is present as a family of isoforms differing slightly at their 5′ end. The 2.0-kb RoBPI mRNA was detected in all human tissues tested. Highly homologous cDNA sequences were found in banks of expressed sequence tags (ESTs) from mice. Two-hybrid, three-hybrid, and RITA experiments respectively established that 60 kDa RoBPI did not interact in yeast with rRo60 alone, with rhY5 RNA alone, or with bait RNPs consisting of rRo60 and recombinant hY1, hY3, or hY4 RNAs. RoBPI coimmunoprecipitated with Ro RNPs from HeLa cell extracts and partially colocalized with Ro60 in nuclei of cultured cells. Because hY5 RNA and RohY5 RNPs are recent evolutionary additions seen only in primates, but RoBPI seems more conserved, their interaction may represent a gain of function for Ro RNPs. Alternatively, interaction of RohY5 RNPs with RoBPI may have no functional bearing, but may underlie some of the unique biochemical and immunological properties of these RNPs.

Myotonic dystrophy is caused by an expanded CTG repeat in the 3′ untranslated region of the DM protein kinase (DMPK) gene. The expanded repeat triggers the nuclear retention of mutant DMPK transcripts, but the resulting underexpression of DMPK probably does not fully account for the severe phenotype. One proposed disease mechanism is that nuclear accumulation of expanded CUG repeats may interfere with nuclear function. Here we show by thermal melting and nuclease digestion studies that CUG repeats form highly stable hairpins. Furthermore, CUG repeats bind to the dsRNA-binding domain of PKR, the dsRNA-activated protein kinase. The threshold for binding to PKR is ∼15 CUG repeats, and the affinity increases with longer repeat lengths. Finally, CUG repeats that are pathologically expanded can activate PKR in vitro. These results raise the possibility that the disease mechanism could be, in part, a gain of function by mutant DMPK transcripts that involves sequestration or activation of dsRNA binding proteins.

Eukaryotic nucleoli contain a large family of box C/D small nucleolar ribonucleoprotein complexes (snoRNPs) that are involved in processing and site-specific methylation of pre-rRNA. Several proteins have been reported to be common factors of box C/D snoRNPs in lower and higher eukaryotes; nevertheless none of them has been clearly shown to directly interact with RNA. We previously identified in Xenopus laevis, by means of UV crosslinking in vivo, two proteins associated with box C/D snoRNAs, fibrillarin and p68. Here we show that fibrillarin interacts directly and specifically with the U16 box C/D snoRNA in a X. laevis oocyte nuclear extract and that it does not require p68 for binding. Specific binding is also obtained with a recombinant fibrillarin demonstrating that the protein is able to bind directly and specifically to U16 snoRNA by itself.

Generation of RNA dimeric form of the human immunodeficiency virus type 1 (HIV-1) genome is crucial for viral replication. The dimerization initiation site (DIS) has been identified as a primary sequence that can form a stem-loop structure with a self-complementary sequence in the loop and a bulge in the stem. It has been reported that HIV-1 RNA fragments containing the DIS form two types of dimers, loose dimers and tight dimers. The loose dimers are spontaneously generated at the physiological temperature and converted into tight dimers by the addition of nucleocapsid protein NCp7. To know the biochemical process in this two-step dimerization reaction, we chemically synthesized a 39-mer RNA covering the entire DIS sequence and also a 23-mer RNA covering the self-complementary loop and its flanking stem within the DIS. Electrophoretic dimerization assays demonstrated that the 39-mer RNA reproduced the two-step dimerization process, whereas the 23-mer RNA immediately formed the tight dimer. Furthermore, deletion of the bulge from the 39-mer RNA prevented the NCp7-assisted tight-dimer formation. Therefore, the whole DIS sequence is necessary and sufficient for the two-step dimerization. Our data suggested that the bulge region regulates the stability of the stem and guides the DIS to the two-step dimerization process.

Rpb5-H147R is an AT-GC transition replacing CAC(His) by CGC(Arg) at a conserved and critical position of ABC27 (Rpb5p), one of the five common and essential subunits shared by all three eukaryotic RNA polymerases. This mutation is viable at 25 °C, but has a lethal phenotype at 34 °C. A search for dosage-dependent suppressors identified five distinct clones that all bear a copy of the tRNAHisGUG gene. Suppression was also observed with a small genomic insert bearing this tRNA gene and no other coding sequences, under conditions where there is a sevenfold increase in the cellular concentration of tRNAHisGUG. Overexpressing tRNAArgICG, which normally decodes the suppressed CGC codon, counteracted suppression. Suppression is codon specific because it was abolished when replacing CGC by its synonymous codons CGA, CGU, or AGA, but was not detectably affected by several nucleotide substitutions modifying the surrounding sequence and is thus largely insensitive to the nucleotide context. It is proposed that overexpressing tRNAHisGUG extends its decoding properties from CAC(His) to the noncognate CGC(Arg) codon through an illegitimate U[bull ]G pairing at the middle base of the anticodon. Accordingly, tRNAHisGUG would compete with tRNAArgICG for chain elongation and generate a significant level of misreading errors under normal growth conditions.

The SRm160/300 splicing coactivator, which consists of the serine/arginine (SR)-related nuclear matrix protein of 160 kDa and a 300-kDa nuclear matrix antigen, functions in splicing by promoting critical interactions between splicing factors bound to pre-mRNA, including snRNPs and SR family proteins. In this article we report the isolation of a cDNA encoding the 300-kDa antigen and investigate the activity of it and SRm160 in splicing. Like SRm160, the 300-kDa antigen contains domains rich in alternating S and R residues but lacks an RNA recognition motif; the protein is accordingly named “SRm300.” SRm300 also contains a novel and highly conserved N-terminal domain, several unique repeated motifs rich in S, R, and proline residues, and two very long polyserine tracts. Surprisingly, specific depletion of SRm300 does not prevent the splicing of pre-mRNAs shown previously to require SRm160/300. Addition of recombinant SRm160 alone to SRm160/300-depleted reactions specifically activates splicing. The results indicate that SRm160 may be the more critical component of the SRm160/300 coactivator in the splicing of SRm160/300-dependent pre-mRNAs.

In trypanosomatids, the majority of the guide (g) RNAs that provide the information for U-insertion/deletion RNA editing are encoded by minicircles that are catenated into large networks. In contrast, in the distantly related cryptobiid Trypanoplasma borreli, gRNA genes appear to reside in large 180-kb noncatenated DNA circles. To shed light on the evolutionary history and function of the minicircle network, we have analyzed minicircle organization in the free-living bodonid Bodo saltans, which is more closely related to trypanosomatids than T. borreli. We identified 1.4-kb circular DNAs as the B. saltans equivalent of minicircles via sequence analysis of 4 complete minicircles, 14 minicircle fragments, and 14 gRNAs. We show that each minicircle harbors two gRNA gene cassettes of opposite polarity residing in variable regions of about 200 nt in otherwise highly conserved molecules. In the conserved region, B. saltans minicircles contain a putative bent helix sequence and a degenerate dodecamer motif (CSB-3). Electron microscopy, sedimentation, and gel electrophoresis analyses showed no evidence for the existence of large minicircle networks in B. saltans, the large majority of the minicircles being present as circular and linear monomers (85–90%) with small amounts of catenated dimers and trimers. Our results provide the first example of a kinetoplastid species with noncatenated, gRNA gene-containing minicircles, which implies that the creation of minicircles and minicircle networks are separate evolutionary events.

Messenger RNAs are exported from the nucleus as large ribonucleoprotein complexes (mRNPs). To date, proteins implicated in this process include TAP/Mex67p and RAE1/Gle2p and are distinct from the nuclear transport receptors of the β-related, Ran-binding protein family. Mex67p is essential for mRNA export in yeast. Its vertebrate homolog TAP has been implicated in the export of cellular mRNAs and of simian type D viral RNAs bearing the constitutive transport element (CTE). Here we show that TAP is predominantly localized in the nucleoplasm and at both the nucleoplasmic and cytoplasmic faces of the nuclear pore complex (NPC). TAP interacts with multiple components of the NPC including the nucleoporins CAN, Nup98, Nup153, p62, and with three major NPC subcomplexes. The nucleoporin-binding domain of TAP comprises residues 508–619. In HeLa cells, this domain is necessary and sufficient to target GFP-TAP fusions to the nuclear rim. Moreover, the isolated domain strongly competes multiple export pathways in vivo, probably by blocking binding sites on the NPC that are shared with other transport receptors. Microinjection experiments implicate this domain in the export of specific CTE-containing RNAs. Finally, we show that TAP interacts with transportin and with two proteins implicated in the export of cellular mRNAs: RAE1/hGle2 and E1B-AP5. The interaction of TAP with nucleoporins, its direct binding to the CTE RNA, and its association with two mRNP binding proteins suggest that TAP is an RNA export mediator that may bridge the interaction between specific RNP export substrates and the NPC.