The conserved family of NXF proteins has been implicated in the export of messenger RNAs from the nucleus. In metazoans, NXFs heterodimerize with p15. The yeast genome encodes a single NXF protein (Mex67p), but there are multiple nxf genes in metazoans. Whether metazoan NXFs are functionally redundant, or their multiplication reflects an adaptation to a greater substrate complexity or to tissue-specific requirements has not been established. The Drosophila genome encodes one p15 homolog and four putative NXF proteins (NXF1 to NXF4). Here we show that depletion of the endogenous pools of NXF1 or p15 from Drosophila cells inhibits growth and results in a rapid and robust accumulation of polyadenylated RNAs within the nucleus. Fluorescence in situ hybridizations show that export of both heat-shock and non-heat-shock mRNAs, as well as intron-containing and intronless mRNAs is inhibited. Depleting endogenous NXF2 or NXF3 has no apparent phenotype. Moreover, NXF4 is not expressed at detectable levels in cultured Drosophila cells. We conclude that Dm NXF1/p15 heterodimers only (but not NXF2-NXF4) mediate the export of the majority of mRNAs in Drosophila cells and that the other members of the NXF family play more specialized or different roles.

There are 10 known putative pseudouridine synthase genes in Escherichia coli. The products of six have been previously assigned, one to formation of the single pseudouridine in 16S RNA, three to the formation of seven pseudouridines in 23S RNA, and three to the formation of three pseudouridines in tRNA (one synthase makes pseudouridine in 23S RNA and tRNA). Here we show that the remaining four putative synthase genes make bona fide pseudouridine synthases and identify which pseudouridines they make. RluB (formerly YciL) and RluE (formerly YmfC) make pseudouridine2605 and pseudouridine2457, respectively, in 23S RNA. RluF (formerly YjbC) makes the newly discovered pseudouridine2604 in 23S RNA, and TruC (formerly YqcB) makes pseudouridine65 in tRNAIle1 and tRNAAsp. Deletion of each of these synthase genes individually had no effect on exponential growth in rich media at 25 °C, 37 °C, or 42 °C. A strain lacking RluB and RluF also showed no growth defect under these conditions. Mutation of a conserved aspartate in a common sequence motif, previously shown to be essential for the other six E. coli pseudouridine synthases and several yeast pseudouridine synthases, also caused a loss of in vivo activity in all four of the synthases studied in this work.

Rpp21, a protein subunit of human nuclear ribonuclease P (RNase P) was cloned by virtue of its homology with Rpr2p, an essential subunit of Saccharomyces cerevisiae nuclear RNase P. Rpp21 is encoded by a gene that resides in the class I gene cluster of the major histocompatibility complex, is associated with highly purified RNase P, and binds precursor tRNA. Rpp21 is predominantly localized in the nucleoplasm but is also observed in nucleoli and Cajal bodies when expressed at high levels. Intron retention and splice-site selection in Rpp21 precursor mRNA regulate the intranuclear distribution of the protein products and their association with the RNase P holoenzyme. Our study reveals that dynamic nuclear structures that include nucleoli, the perinucleolar compartment and Cajal bodies are all involved in the production and assembly of human RNase P.

The mammalian signal recognition particle (SRP) catalytically promotes cotranslational translocation of signal sequence containing proteins across the endoplasmic reticulum membrane. While the S-domain of SRP binds the N-terminal signal sequence on the nascent polypeptide, the Alu domain of SRP temporarily interferes with the ribosomal elongation cycle until the translocation pore in the membrane is correctly engaged. Here we present biochemical and biophysical evidence for a hierarchical assembly pathway of the SRP Alu domain. The proteins SRP9 and SRP14 first heterodimerize and then initially bind to the Alu RNA 5′ domain. This creates the binding site for the Alu RNA 3′ domain. Alu RNA then undergoes a large conformational change with the flexibly linked 3′ domain folding back by 180° onto the 5′ domain complex to form the final compact Alu ribonucleoprotein particle (Alu RNP). We discuss the possible mechanistic consequences of the likely reversibility of this final step with reference to translational regulation by the SRP Alu domain and with reference to the structurally similar Alu RNP retroposition intermediates derived from Alu elements in genomic DNA.

Some studies suggest that the hepatitis C virus (HCV) internal ribosome entry site (IRES) requires downstream 5′ viral polyprotein-coding sequence for efficient initiation of translation, but the role of this RNA sequence in internal ribosome entry remains unresolved. We confirmed that the inclusion of viral sequence downstream of the AUG initiator codon increased IRES-dependent translation of a reporter RNA encoding secretory alkaline phosphatase, but found that efficient translation of chloramphenicol acetyl transferase (CAT) required no viral sequence downstream of the initiator codon. However, deletion of an adenosine-rich domain near the 5′ end of the CAT sequence, or the insertion of a small stable hairpin structure (ΔG = −18 kcal/mol) between the HCV IRES and CAT sequences (hpCAT) substantially reduced IRES-mediated translation. Although translation could be restored to both mutants by the inclusion of 14 nt of the polyprotein-coding sequence downstream of the AUG codon, a mutational analysis of the inserted protein-coding sequence demonstrated no requirement for either a specific nucleotide or amino acid-coding sequence to restore efficient IRES-mediated translation to hpCAT. Similar results were obtained with the structurally and phylogenetically related IRES elements of classical swine fever virus and GB virus B. We conclude that there is no absolute requirement for viral protein-coding sequence with this class of IRES elements, but that there is a requirement for an absence of stable RNA structure immediately downstream of the AUG initiator codon. Stable RNA structure immediately downstream of the initiator codon inhibits internal initiation of translation but, in the case of hpCAT, did not reduce the capacity of the RNA to bind to purified 40S ribosome subunits. Thus, stable RNA structure within the 5′ proximal protein-coding sequence does not alter the capacity of the IRES to form initial contacts with the 40S subunit, but appears instead to prevent the formation of subsequent interactions between the 40S subunit and viral RNA in the vicinity of the initiator codon that are essential for efficient internal ribosome entry.

Pre-messenger RNA splicing is a two-step process by which introns are removed and exons joined together. In yeast, the U5 snRNA loop 1 interacts with the 5′ exon before the first step of splicing and with the 5′ and 3′ exons before the second step. In vitro studies revealed that yeast U5 loop 1 is not required for the first step of splicing but is essential for holding the 5′ and 3′ exons for ligation during the second step. It is critical, therefore, that loop 1 contacts the 5′ exon before the first step of splicing to hold this exon following cleavage from the pre-mRNA. At present it is not known how U5 loop 1 is positioned on the 5′ exon prior to the first step of splicing. To address this question, we have used site-specific photoactivated crosslinking in yeast spliceosomes to investigate the interaction of U5 loop 1 with the pre-mRNA prior to the first step of splicing. We have found that the highly conserved uridines in loop 1 make ATP-dependent contacts with an approximately 8-nt region at the 5′ splice site that includes the invariant GU. These interactions are dependent on functional U2 and U6 snRNAs. Our results support a model where U5 snRNA loop 1 interacts with the 5′ exon in two steps during its targeting to the 5′ splice site.

In eubacteria, the biosynthesis of queuine, a modified base found in the wobble position (#34) of tRNAs coding for Tyr, His, Asp, and Asn, occurs via a multistep pathway. One of the key enzymes in this pathway, tRNA-guanine transglycosylase (TGT), exchanges the genetically encoded guanine at position 34 with a queuine precursor, preQ1. Previous studies have identified a minimal positive RNA recognition motif for Escherichia coli TGT consisting of a stable minihelix that contains a U-G-U sequence starting at the second position of its seven base anticodon loop. Recently, we reported that TGT was capable of recognizing the U-G-U sequence outside of this limited structural context. To further characterize the ability of TGT to recognize the U-G-U sequence in alternate contexts, we constructed mutants of the previously characterized E. coli tRNATyr minihelix. The U-G-U sequence was shifted to various positions within the anticodon loop of these mutants. Characterization of these analogs demonstrates that in addition to the normal U33G34U35 position, TGT can also recognize the U34G35U36 analog (UGU+1). The other analogs were not active. This indicates that the recognition of the U-G-U sequence is not strictly dependent upon its position relative to the stem. In E. coli, the full-length tRNA with a U34G35U36 anticodon sequence is one of the isoacceptors that codes for threonine. We found that TGT is able to recognize tRNAThr(UGU) but only in the absence of a uridine at position 33. U33, an invariant base present in all tRNAs, has been shown to strongly influence the conformation of the anticodon loop of certain tRNAs. We find that mutation of this base confers on TGT the ability to recognize U34G35U36, and suggests that loop conformation affects recognition. The fact that the other analogs were not active indicates that although TGT is capable of recognizing the U-G-U sequence in additional contexts, this recognition is not indiscriminate.

Association of U2 snRNP with the pre-mRNA branch region is the first ATP-dependent step in spliceosome assembly. The basis of this energy dependence is not known. Previously, we identified minimal intron-derived substrates that form complexes with U2 independent of ATP. Here, we identify the intron region linked to the ATP dependence of this step by comparing these substrates to longer RNAs that recapitulate the ATP requirement. This region needed to impose ATP dependence lies immediately 5′ to the branch site. Sequences ranging from 6 to 14 nt yield a near linear inhibitory effect on efficiency of complex formation with U2 snRNP, with 18 nt yielding near maximal ATP dependence. This region is not protected prior to U2 addition, and RNase H targeting of the region within nuclear extract converts an ATP-dependent substrate into an ATP-independent one. Within this region, there is no sequence specificity linked with the ATP requirement, as neither a specific sequence is needed, nor even nucleobases. These data and the results of other modifications suggest models in which the 18-nt region is a target for interactions with U2 snRNP in an ATP-bound or -activated conformation.

Ro RNPs are small cytoplasmic RNA–protein complexes of unknown function that have been found in all metazoan cells studied so far. In human cells, Ro RNPs consist of one of four small RNA molecules, termed hY RNAs and at least two well-characterized proteins, Ro60 and La. In previous Xenopus laevis oocyte microinjection studies, we showed that an intact Ro60 binding site (Stem-loop 1) is a prerequisite for efficient nuclear export of hY1 RNA, whereas an intact La-binding site promotes nuclear retention (Simons et al. RNA, 1996, 2:264–273). Here we present evidence that the distal half (Stem 2) of the conserved base-paired stem structure found in all hY RNAs also plays a critical role in the export process. A minimal RNA molecule containing this region, L1S2 RNA, competes effectively for the export of full-length hY1 RNAs and is itself exported very rapidly in a Ro60-independent and RanGTP-dependent manner. Mutational analyses of this RNA shows that a 5′/3′ terminal double-stranded stem structure (>10 bp) of no specific nucleotide sequence constitutes a novel nuclear export element (NEE). Cross-competition studies indicate that this type of NEE may also be involved in export of other classes of RNAs. Like full-length hY1 RNA, L1S2 RNA also competes for export of ET-202 RNA, an RNA that was selected for its efficient nuclear export in the presence of the nuclear transport inhibitor, VSV Matrix protein (Grimm et al. Proc Natl Acad Sci USA, 1997, 94:10122–10127). However, export of L1S2 RNA is strongly inhibited by VSV-M protein, showing that these RNAs use partially overlapping, but not identical export pathways. We propose that export of Y RNAs is mediated by two contiguous cis-acting elements in the 5′/3′ double-stranded stem region that is conserved between different Y RNAs.

We have determined the X-ray structures of six MS2 RNA hairpin–coat-protein complexes having five different substitutions at the hairpin loop base −5. This is a uracil in the wild-type hairpin and contacts the coat protein both by stacking on to a tyrosine side chain and by hydrogen bonding to an asparagine side chain. The RNA consensus sequence derived from coat protein binding studies with natural sequence variants suggested that the −5 base needs to be a pyrimidine for strong binding. The five −5 substituents used in this study were 5-bromouracil, pyrimidin-2-one, 2-thiouracil, adenine, and guanine. The structure of the 5-bromouracil complex was determined to 2.2 Å resolution, which is the highest to date for any MS2 RNA–protein complex. All the complexes presented here show very similar conformations, despite variation in affinity in solution. The results suggest that the stacking of the −5 base on to the tyrosine side chain is the most important driving force for complex formation. A number of hydrogen bonds that are present in the wild-type complex are not crucial for binding, as- they are missing in one or more of the complexes. The results also reveal the flexibility of this RNA-protein interface, with respect to functional group variation, and may be generally applicable to other RNA–protein complexes.

Previously we described an in vitro selection variant abbreviated SERF (in vitro selection from random rRNA fragments) that identifies protein binding sites within large RNAs. With this method, a small rRNA fragment derived from the 23S rRNA was isolated that binds simultaneously and independently the ribosomal proteins L4 and L24 from Escherichia coli. Until now the rRNA structure within the ternary complex L24–rRNA–L4 could not be studied due to the lack of an appropriate experimental strategy. Here we tackle the issue by separating the various complexes via native gel-electrophoresis and analyzing the rRNA structure by in-gel iodine cleavage of phosphorothioated RNA. The results demonstrate that during the transition from either the L4 or L24 binary complex to the ternary complex the structure of the rRNA fragment changes significantly. The identified protein binding sites are in excellent agreement with the recently reported crystal structure of the 50S subunit. Because both proteins play a prominent role in early assembly of the large subunit, the results suggest that the identified rRNA fragment is a key element for the folding of the 23S RNA during early assembly. The introduced in-gel cleavage method should be useful when an RNA structure within mixed populations of different but related complexes should be studied.

We identified a temperature-sensitive allele of small bristles (sbr), the Drosophila ortholog of human TAP/NXF-1 and yeast Mex67, in a screen for mutants defective in mRNA export. We show that sbr is essential for the nuclear export of all mRNAs tested in a wide range of tissues and times in development. High resolution and sensitive in situ hybridization detect the rapid accumulation of individual mRNA species in sbr mutant nuclei in particles that are distinct from nascent transcript foci and resemble wild-type export intermediates. The particles become more numerous and intense with increasing time at the restrictive temperature and are exported very rapidly after shifting back to the permissive temperature. The mRNA export block is not due indirectly to a defect in splicing, nuclear protein import, or aberrant nuclear ultrastructure, suggesting that in sbr mutants, mRNA is competent for export but fails to dock or translocate through NPCs. We conclude that NXF-1 is an essential ubiquitous export factor for all mRNAs throughout development in higher eukaryotes.

Using a series of point mutations in chimeric reporter gene constructs consisting of the 5′ regions of the Chlamydomonas chloroplast rbcL or atpB genes fused 5′ to the coding sequence of the bacterial uidA (GUS) gene, RNA-stabilizing sequence elements were identified in vivo in the 5′ untranslated regions (5′ UTRs) of transcripts of the chloroplast genes rbcL and atpB in Chlamydomonas reinhardtii. In chimeric rbcL 5′ UTR:GUS transcripts, replacement of single nucleotides in the 10-nt sequence 5′-AUUUCCGGAC-3′, extending from positions +38 to +47 relative to the transcripts' 5′ terminus, shortened transcript longevity and led to a reduction in transcript abundance of more than 95%. A similar mutational analysis of atpB 5′ UTR:GUS transcripts showed that the 12-nt atpB 5′ UTR sequence 5′-AUAAGCGUUAGU-3′, extending from position +31 to position +42, is important for transcript stability and transcript accumulation in the chloroplast of Chlamydomonas. We discuss how the 5′ UTR sequence elements, which are predicted to be part of RNA secondary structures, might function in RNA stabilization.

Two different transcription termination control mechanisms, the T box and S box systems, are used to regulate transcription of many bacterial aminoacyl-tRNA synthetase, amino acid biosynthesis, and amino acid transport genes. Both of these regulatory mechanisms involve an untranslated mRNA leader region capable of adopting alternate structural conformations that result in transcription termination or transcription elongation into the downstream region. Comparative analyses revealed a small RNA secondary structural element, designated the GA motif, that is highly conserved in both T box and S box leader sequences. The motif consists of two short helices separated by an asymmetric internal loop, with highly conserved GA dinucleotide sequences on either side of the internal loop. Site-directed mutagenesis of this motif in model T and S box leader sequences indicated that it is essential for transcriptional regulation in both systems. This motif is similar to the binding site of yeast ribosomal protein L30, the Snu13p binding sites found in U4 snRNA and box C/D snoRNAs, and two elements in 23S rRNA.

HIV-1 env expression from certain subgenomic vectors requires the viral regulatory protein Rev, its target sequence RRE, and a 5′ splice site upstream of the env open reading frame. To determine the role of this splice site in the 5′-splice-site-dependent Rev-mediated env gene expression, we have subjected the HIV-1 5′ splice site, SD4, to a mutational analysis and have analyzed the effect of those mutations on env expression. The results demonstrate that the overall strength of hydrogen bonding between the 5′ splice site, SD4, and the free 5′ end of the U1 snRNA correlates with env expression efficiency, as long as env expression is suboptimal, and that a continuous stretch of 14 hydrogen bonds can lead to full env expression, as a result of stabilizing the pre-mRNA. The U1 snRNA-mediated stabilization is independent of functional splicing, as a mismatch in position +1 of the 5′ splice site that led to loss of detectable amounts of spliced transcripts did not preclude stabilization and expression of the unspliced env mRNA, provided that Rev enables its nuclear export. The nucleotides capable of participating in U1 snRNA:pre-mRNA interaction include positions −3 to +8 of the 5′ splice site and all 11 nt constituting the single-stranded 5′ end of U1 snRNA. Moreover, env gene expression is significantly decreased upon the introduction of point mutations in several upstream GAR nucleotide motifs, which are mediating SF2/ASF responsiveness in an in vitro splicing assay. This suggests that the GAR sequences may play a role in stabilizing the pre-mRNA by sequestering U1 snRNP to SD4.

The U6 spliceosomal snRNA forms an intramolecular stem-loop structure during spliceosome assembly that is required for splicing and is proposed to be at or near the catalytic center of the spliceosome. U6atac snRNA, the analog of U6 snRNA used in the U12-dependent splicing of the minor class of spliceosomal introns, contains a similar stem-loop whose structure but not sequence is conserved between humans and plants. To determine if the U6 and U6atac stem-loops are functionally analogous, the stem-loops from human and budding yeast U6 snRNAs were substituted for the U6atac snRNA structure and tested in an in vivo genetic suppression assay. Both chimeric U6/U6atac snRNA constructs were active for splicing in vivo. In contrast, several mutations of the native U6atac stem-loop that either delete putatively unpaired residues or disrupt the putative stem regions were inactive for splicing. Compensatory mutations that are expected to restore base pairing within the stem regions restored splicing activity. However, other mutants that retained base pairing potential were inactive, suggesting that functional groups within the stem regions may contribute to function. These results show that the U6atac snRNA stem-loop structure is required for in vivo splicing within the U12-dependent spliceosome and that its role is likely to be similar to that of the U6 snRNA intramolecular stem-loop.

The decoding of UGA as a selenocysteine (Sec) codon in mammalian selenoprotein mRNAs requires a selenocysteine insertion sequence (SECIS) element in the 3′ untranslated region. The SECIS is a hairpin structure that contains a non-Watson–Crick base-pair quartet with a conserved G.A/A.G tandem in the core of the upper helix. Another essential component of the Sec insertion machinery is SECIS-binding protein 2 (SBP2). In this study, we define the binding site of SBP2 on six different SECIS RNAs using enzymatic and hydroxyl radical footprinting, gel mobility shift analysis, and phosphate-ethylation binding interference. We show that SBP2 binds to a variety of mammalian SECIS elements with similar affinity and that the SBP2 binding site is conserved across species. Based on footprinting studies, SBP2 protects the proximal part of the hairpin and both strands of the lower half of the upper helix that contains the non-Watson–Crick base pair quartet. Gel mobility shift assays showed that the G.A/A.G tandem and internal loop are critical for the binding of SBP2. Modification of phosphates by ethylnitrosourea along both strands of the non-Watson–Crick base pair quartet, on the 5′ strand of the lower helix and part of the 5′ strand of the internal loop, prevented binding of SBP2. We propose a model in which SBP2 covers the central part of the SECIS RNA, binding to the non-Watson–Crick base pair quartet and to the 5′ strands of the lower helix and internal loop. Our results suggest that the affinity of SBP2 for different SECIS elements is not responsible for the hierarchy of selenoprotein expression that is observed in vivo.

The U3 small nucleolar ribonucleoprotein (snoRNP) is composed of a small nucleolar RNA (snoRNA) and at least 10 proteins. The U3 snoRNA base pairs with the pre-rRNA to carry out the A0, A1, and A2 processing reactions that lead to the release of the 18S rRNA from the nascent pre-rRNA transcript. The yeast U3 snoRNA can be divided into a short 5′ domain (nt 1–39) and a larger 3′ domain (73 to the 3′ end) separated by a stretch of nucleotides called the hinge region (nt 40–72). The sequences required for pre-rRNA base pairing are found in the 5′ domain and hinge region whereas the 3′ domain is largely covered with proteins. Mpp10p, one of the protein components unique to the U3 snoRNP, plays a role in processing at the A1 and A2 sites. Because of its critical role in U3 snoRNP function, we determined which sequences in the U3 snoRNA are required for Mpp10p association. Unlike fibrillarin and all the previous U3 snoRNP components studied in this manner, sequences in the 3′ domain are not sufficient for Mpp10p association. Instead, a conserved sequence element in the U3 snoRNA hinge region is required, placing Mpp10p near the 5′ domain that carries out the pre-rRNA base-pairing interactions in the functional center of the U3 snoRNP.

We show that the problem of designing RNA sequences that can fold into multiple stable secondary structures can be transformed into a combinatorial optimization problem that can be solved by means of simple heuristics. Hence it is feasible to design RNA switches with prescribed structural alternatives. We discuss the theoretical background and present an efficient tool that allows the design of various types of switches. We argue that both the general properties of the sequence structure map of RNA secondary structures and the ease with which our design tool finds bistable RNAs strongly indicates that RNA switches are easily accessible in evolution. Thus conformational switches are yet another function for which RNA can be employed.

DNA harvested directly from complex natural microbial communities by PCR has been successfully used to predict RNase P RNA structure, and can potentially provide an abundant source of information for structural predictions of other RNAs. In this study, we utilized genetic variation in natural communities to test and refine the secondary and tertiary structural model for the bacterial tmRNA. The variability of proposed tmRNA secondary structures in different organisms and the lack of any predicted tertiary structure suggested that further refinement of the tmRNA could be useful. To increase the phylogenetic representation of tmRNA sequences, and thereby provide additional data for statistical comparative analysis, we amplified, sequenced, and compared tmRNA sequences from natural microbial communities. Using primers designed from gamma proteobacterial sequences, we determined 44 new tmRNA sequences from a variety of environmental DNA samples. Covariation analyses of these sequences, along with sequences from cultured organisms, confirmed most of the proposed tmRNA model but also provided evidence for a new tertiary interaction. This approach of gathering sequence information from natural microbial communities seems generally applicable in RNA structural analysis.