In recent years, a growing number of mammalian genes have been shown to undergo alternative splicing in response to extracellular stimuli. However, the factors and pathways involved in such signal-induced alternative splicing are almost entirely unknown. Here we describe a novel method for identifying candidate trans-acting factors that are involved in regulating mammalian alternative splicing, using the activation-induced alternative splicing of the human CD45 gene in T cells as a model system. We generated a cell line that stably expresses a CD45 minigene-based GFP reporter construct, such that the levels of green-fluorescent protein (GFP) expressed in the cell reflect the splicing state of the endogenous CD45 gene. Following mutagenesis of this cell line, and multiple rounds of selection for cells that displayed aberrant levels of GFP expression, we isolated several cell lines that are at least partially defective in their ability to support regulated alternative splicing of endogenous CD45 pre-mRNA in response to cell stimulation. Thus we have successfully isolated mutants in a mammalian alternative splicing pathway through use of a somatic cell-based genetic screen. This study clearly demonstrates the feasibility of using genetic screens to further our understanding of the regulation of mammalian splicing, particularly as it occurs in response to environmental cues.

The Fragile X mental retardation gene (FMR1) contains a polymorphic trinucleotide CGG repeat in the 5′ untranslated region (UTR) of the FMR1 messenger. We have characterized three lymphoblastoid cell lines derived from unrelated male carriers of a premutation that overexpress FMR1 mRNA and show reduced FMRP level compared to normal cells. The analysis of polysomes/mRNPs distribution of mRNA in the cell lines with a premutation shows that the polysomal association of FMR1 mRNA, which is high in normal cells, becomes progressively lower with increasing CGG repeat expansion. In addition, we could detect a very low level of FMR1 mRNA in a lymphoblastoid cell line from a patient with a full mutation. In this case, FMR1 mRNA is not at all associated with polysomes, in agreement with the complete absence of FMRP. The impairment of FMR1 mRNA translation in patients with the Fragile X syndrome with FMR1 premutation is the cause of the lower FMRP levels that leads to the clinical involvement.

Sm and Sm-like (LSm) proteins form heptameric complexes that are involved in various steps of RNA metabolism. In yeast, the Lsm1–7 complex functions in mRNA degradation and is associated with several enzymes of this pathway, while the complex LSm2–8, the composition of which largely overlaps with that of LSm1–7, has a role in pre-mRNA splicing. A human gene encoding an LSm1 homolog has been identified, but its role in mRNA degradation has yet to be elucidated. We performed subcellular localization studies and found hLSm1 predominantly in the cytoplasm. However, it is not distributed evenly; rather, it is highly enriched in small, discrete foci. The endogenous hLSm4 is similarly localized, as are the overexpressed proteins hLSm1–7, but not hLSm8. The foci also contain two key factors in mRNA degradation, namely the decapping enzyme hDcp1/2 and the exonuclease hXrn1. Moreover, coexpression of wild-type and mutant LSm proteins, as well as fluorescence resonance energy transfer (FRET) studies, indicate that the mammalian proteins hLSm1–7 form a complex similar to the one found in yeast, and that complex formation is required for enrichment of the proteins in the cytoplasmic foci. Therefore, the foci contain a partially or fully assembled machinery for the degradation of mRNA.

Box H/ACA small nucleolar ribonucleoprotein particles (snoRNPs) contain four essential proteins, Cbf5p, Gar1p, Nhp2p, and Nop10p, each of which, with the exception of Gar1p, is required for box H/ACA snoRNA accumulation. Database searches identified a novel essential protein, which we termed Naf1p, with a region of homology to the RNA-binding domain of Gar1p and other features in common with hnRNP-like proteins. Naf1p is localized to the nucleus and is not a stable component of the H/ACA snoRNPs, but it is required for the accumulation of all box H/ACA snoRNAs. This requirement is not at the level of snoRNA transcription initiation or termination. Naf1p shows in vitro RNA-binding activity and also binds directly to Cbf5p and Nhp2p. Naf1p was shown to bind to the CTD in vivo in a two-hybrid assay, and the phosphorylated CTD, but not the nonphosphorylated CTD, was shown to precipitate tagged Naf1p from a cell lysate. We propose that Naf1p is recruited to the CTD of RNA polymerase II and binds to nascent box H/ACA snoRNAs promoting snoRNP assembly.

U2 is the most extensively modified of all spliceosomal snRNAs. We previously showed that at least some of the internally modified nucleotides in U2 snRNA are required for snRNP biogenesis and pre-mRNA splicing. Recent work from several laboratories suggests that nuclear guide RNAs facilitate U2 snRNA internal modification, including pseudouridylation and 2′-O-methylation. Here, we present a novel approach to identifying guide RNAs for U2 pseudouridylation. Several Xenopus oocyte nuclear RNAs were affinity selected with U2 snRNA substituted with 5-fluorouridine, a pseudouridylation inhibitor that sequesters pseudouridylases. One of these RNAs was sequenced and found to be a novel RNA of 134 nt. This small RNA contains an H/ACA motif and folds into a typical H/ACA RNA structure, and its authenticity as an H/ACA RNA was confirmed by immunoprecipitation analysis. The RNA contains two guide sequences for pseudouridylation (Ψ) of U2 snRNA at positions 34 and 44 in the branch-site recognition region, and we demonstrate that this RNA indeed guides the formation of Ψ34 and Ψ44 in U2 using a Xenopus oocyte reconstitution system. Therefore, this novel RNA was designated pugU2-34/44, for pseudouridylation guide for U2 snRNA U34 and U44. Intranuclear localization analyses indicate that pugU2-34/44 resides within the nucleoplasm rather than nucleoli or Cajal bodies where other guide RNAs have been localized. Our results clarify the mechanism of U2 snRNA pseudouridylation in Xenopus oocytes, and have interesting implications with regard to the intranuclear localization of U2 snRNA pseudouridylation.

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.

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 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.

Classical swine fever virus (CSFV) is a member of the pestivirus family, which shares many features in common with hepatitis C virus (HCV). It is shown here that CSFV has an exceptionally efficient cis-acting internal ribosome entry segment (IRES), which, like that of HCV, is strongly influenced by the sequences immediately downstream of the initiation codon, and is optimal with viral coding sequences in this position. Constructs that retained 17 or more codons of viral coding sequence exhibited full IRES activity, but with only 12 codons, activity was ∼66% of maximum in vitro (though close to maximum in transfected BHK cells), whereas with just 3 codons or fewer, the activity was only ∼15% of maximum. The minimal coding region elements required for high activity were exchanged between HCV and CSFV. Although maximum activity was observed in each case with the homologous combination of coding region and 5′ UTR, the heterologous combinations were sufficiently active to rule out a highly specific functional interplay between the 5′ UTR and coding sequences. On the other hand, inversion of the coding sequences resulted in low IRES activity, particularly with the HCV coding sequences. RNA structure probing showed that the efficiency of internal initiation of these chimeric constructs correlated most closely with the degree of single-strandedness of the region around and immediately downstream of the initiation codon. The low activity IRESs could not be rescued by addition of supplementary eIF4A (the initiation factor with ATP-dependent RNA helicase activity). The extreme sensitivity to secondary structure around the initiation codon is likely to be due to the fact that the eIF4F complex (which has eIF4A as one of its subunits) is not required for and does not participate in initiation on these IRESs.

The aim of the present study was to identify functional antisense oligodeoxynucleotides (ODNs) against the rat glutathione S-transferase Mu (GSTM) isoforms, GSTM1 and GSTM2. These antisense ODNs would enable the study of the physiological consequences of GSTM deficiency. Because it has been suggested that the effectiveness of antisense ODNs is dependent on the secondary mRNA structures of their target sites, we made mRNA secondary structure predictions with two software packages, Mfold and STAR. The two programs produced only marginally similar structures, which can probably be attributed to differences in the algorithms used. The effectiveness of a set of 18 antisense ODNs was evaluated with a cell-free transcription/translation assay, and their activity was correlated with the predicted secondary RNA structures. Four phosphodiester ODNs specific for GSTM1, two ODNs specific for GSTM2, and four ODNs targeted at both GSTM isoforms were found to be potent, sequence-specific, and RNase H-dependent inhibitors of protein expression. The IC50 value of the most potent ODN was approximately 100 nM. Antisense ODNs targeted against regions that were predicted by STAR to be predominantly single stranded were more potent than antisense ODNs against double-stranded regions. Such a correlation was not found for the Mfold prediction. Our data suggest that simulation of the local folding of RNA facilitates the discovery of potent antisense sequences. In conclusion, we selected several promising antisense sequences, which, when synthesized as biologically stable oligonucleotides, can be applied for study of the physiological impact of reduced GSTM expression.