Evolutionarily conserved structures in related RNA molecules contain coordinated variations (covariations) of paired nucleotides. Analysis of covariations is a very powerful approach to deduce phylogenetically conserved (i.e., functional) conformations, including tertiary interactions. Here we discuss conserved RNA folding pathways that are revealed by covariation patterns. In such pathways, structural requirements for alternative pairings cause some nucleotides to covary with two different partners. Such “coupled” covariations between three or more nucleotides were found in various types of RNAs. The analysis of coupled covariations can unravel important features of RNA folding dynamics and improve phylogeny reconstruction in some cases. Importantly, it is necessary to distinguish between multiple covariations determined by mutually exclusive structures and those determined by tertiary contacts.

Uridylate residues (Us) are inserted and deleted at precise positions in mitochondrial transcripts of Trypanosoma brucei. These sequence changes are determined by interactions with small guide RNAs (gRNAs) that are complementary to edited sequence. Adenylate (A) and guanylate (G) residues in gRNAs across from editing sites pair with inserted Us. We evaluate whether sequence bias exists in the bases surrounding insertion sites. Upon analyzing all reported insertion sites in T. brucei, we find that the predicted base pairs flanking insertion sites show a strong bias. Specifically, guiding As and Gs tend to be flanked by cytosine residues and Us. This bias is expected if precise base-pair interactions at the editing site determine the number of inserted Us.

Rrp5p is the only ribosomal RNA processing trans-acting factor that is required for the synthesis of both 18S and 5.8S rRNAs in Saccharomyces cerevisiae. Mutational analyses have characterized modified forms of Rrp5p that either affect formation of 18S rRNA by inhibiting cleavage at sites A0/A1/A2, or synthesis of 5.8S rRNA by inhibiting cleavage at site A3. Here, we examine the rRNA maturation process associated with a RRP5 bipartite allele that codes for two noncontiguous parts of the protein. This slow-growing bipartite mutant has a unique rRNA-processing phenotype that proceeds without endonucleolytic cleavage at site A2. In wild-type cells, the A2 cleavage takes place on the 32S pre-rRNA and is responsible for the formation of 20S and 27SA2 species, the precursors of mature 18S and 5.8S/25S rRNAs, respectively. In the bipartite strain, such precursors were not detectable as judged by Northern analysis or in vivo labeling. They were replaced by the aberrant 21S species and the bypassing 27SA3 precursor, both descended from direct cleavage of 32S pre-rRNA at site A3, which provides an alternative rRNA maturation pathway in this strain. The 21S pre-rRNA is the sole detectable and most likely available precursor of 18S rRNA in this particular strain, indicating that 18S rRNA can be directly produced from 21S. Furthermore, 21S species were found associated with 43S preribosomal particles as similarly observed for the 20S pre-rRNA in the wild-type cells.

Few point mutations have been described that specifically inhibit the second step of group II intron splicing. Furthermore, the effects of such mutations are typically not apparent unless the mutations are studied in the context of a substrate that harbors a very short 5′ exon. Truncation of the 5′ exon slows the second step of splicing. Once the second step has been slowed, the effects of point mutations can be seen. We report the unexpected observation that the deletion of a conserved GA dinucleotide dramatically inhibits the second step of splicing, even when the mutation is studied in the context of a full-length substrate. In contrast, we find that this mutation does not significantly affect the first step of splicing, unless the mutation is studied in combination with a second point mutation that is known to inhibit the first step. Even in that context, the effect of the GA deletion mutation on the first step is modest. These observations, together with the inferred location of the GA dinucleotide in the three-dimensional structure of the intron, suggest that this dinucleotide plays a particularly important role in the second step of splicing.

Chemically modified nucleotide analogs have gained widespread popularity for probing structure–function relationships. Among the modifications that were incorporated into RNAs for assessing the role of individual functional groups, the phenyl nucleotide has displayed surprising effects both in the contexts of the hammerhead ribozyme and pre-mRNA splicing. To examine the conformational properties of this hydrophobic base analog, we determined the crystal structure of an RNA double helix with incorporated phenyl ribonucleotides at 1.97 Å resolution. In the structure, phenyl residues are engaged in self-pairing and their arrangements suggest energetically favorable stacking interactions with 3′-adjacent guanines. The presence of the phenyl rings in the center of the duplex results in only moderate changes of the helical geometry. This finding is in line with those of earlier experiments that showed the phenyl analog to be a remarkably good mimetic of natural base function. Because the stacking interactions displayed by phenyl residues appear to be similar to those for natural bases, reduced conformational restriction due to the lack of hydrogen bonds with phenyl as well as alterations in its solvent structure may be the main causes of the activity changes with phenyl-modified RNAs.

The Suppressor of forked protein is the Drosophila homolog of the 77K subunit of human cleavage stimulation factor, a complex required for the first step of the mRNA 3′-end-processing reaction. We have shown previously that wild-type su(f) function is required for the accumulation of a truncated su(f) transcript polyadenylated in intron 4 of the gene. This led us to propose a model in which the Su(f) protein would negatively regulate its own accumulation by stimulating 3′-end formation of this truncated su(f) RNA. In this article, we demonstrate this model and show that su(f) autoregulation is tissue specific. The Su(f) protein accumulates at a high level in dividing tissues, but not in nondividing tissues. We show that this distribution of the Su(f) protein results from stimulation by Su(f) of the tissue-specific utilization of the su(f) intronic poly(A) site, leading to the accumulation of the truncated su(f) transcript in nondividing tissues. Utilization of this intronic poly(A) site is affected in a su(f) mutant and restored in the mutant with a transgene encoding wild-type Su(f) protein. These data provide an in vivo example of cell-type-specific regulation of a protein level by poly(A) site choice, and confirm the role of Su(f) in regulation of poly(A) site utilization.

The 3′ ends of animal replication-dependent histone mRNAs are formed by endonucleolytic cleavage of the primary transcripts downstream of a highly conserved RNA hairpin. The hairpin-binding protein (HBP) binds to this RNA element and is involved in histone RNA 3′ processing. A minimal RNA-binding domain (RBD) of ∼73 amino acids that has no similarity with other known RNA-binding motifs was identified in human HBP [Wang Z-F et al., Genes & Dev, 1996, 10:3028-3040]. The primary sequence identity between human and Caenorhabditis elegans RBDs is 55% compared to 38% for the full-length proteins. We analyzed whether differences between C. elegans and human HBP and hairpins are reflected in the specificity of RNA binding. The C. elegans HBP and its RBD recognize only their cognate RNA hairpins, whereas the human HBP or RBD can bind both the mammalian and the C. elegans hairpins. This selectivity of C. elegans HBP is mostly mediated by the first nucleotide in the loop, which is C in C. elegans and U in all other metazoans. By converting amino acids in the human RBD to the corresponding C. elegans residues at places where the latter deviates from the consensus, we could identify two amino acid segments that contribute to selectivity for the first nucleotide of the hairpin loop.

Human endogenous retrovirus K (HERV-K) is the name given to an ∼30-million-year-old family of endogenous retroviruses present at >50 copies per haploid human genome. Previously, the HERV-K were shown to encode a nuclear RNA export factor, termed K-Rev, that is the functional equivalent of the H-Rev protein encoded by human immunodeficiency virus type 1. HERV-K was also shown to contain a cis-acting target element, the HERV-K Rev response element (K-RRE), that allowed the nuclear export of linked RNA transcripts in the presence of either K-Rev or H-Rev. Here, we demonstrate that the functionally defined K-RRE coincides with a statistically highly significant unusual RNA folding region and present a potential RNA secondary structure for the ∼416-nt K-RRE. Both in vitro and in vivo assays of sequence specific RNA binding were used to map two primary binding sites for K-Rev, and one primary binding site for H-Rev, within the K-RRE. Of note, all three binding sites map to discrete predicted RNA stem-loop subdomains within the larger K-RRE structure. Although almost the entire 416-nt K-RRE was required for the activation of nuclear RNA export in cells expressing K-Rev, mutational inactivation of the binding sites for K-Rev resulted in the selective loss of the K-RRE response to K-Rev but not to H-Rev. Together, these data strongly suggest that the K-RRE, like the H-RRE, coincides with an extensive RNA secondary structure and identify specific sites within the K-RRE that can recruit either K-Rev or H-Rev to HERV-K RNA transcripts.

Through a genetic screen to search for factors that interact with Prp17/Cdc40p, a protein involved in both cell cycle progression and pre-mRNA splicing, we identify three novel factors, which we call Syf1p, Syf2p, and Syf3 (SYnthetic lethal with cdc Forty). Here we present evidence that all three proteins are spliceosome associated, that they associate weakly or transiently with U6 and U5 snRNAs, and that Syf1p and Syf3p (also known as Clf1p) are required for pre-mRNA splicing. In addition we show that depletion of Syf1p or Syf3p results in cell cycle arrest at the G2/M transition. Thus, like Prp17/Cdc40p, Syf1p and Syf3p are involved in two distinct cellular processes. We discuss the likelihood that Syf1p, Syf2p, and Syf3p are components of a protein complex that assembles into spliceosomes and also regulates cell cycle progression.

The translation of mammalian selenoprotein mRNAs requires the 3′ untranslated region that contains a selenocysteine insertion sequence (SECIS) element necessary for decoding an in-frame UGA codon as selenocysteine (Sec). Selenoprotein biosynthesis is inefficient, which may be due to competition between Sec insertion and termination at the UGA/Sec codon. We analyzed the polysome distribution of phospholipid hydroperoxide glutathione peroxidase (PHGPx) mRNA, a member of the glutathione peroxidase family of selenoproteins, in rat hepatoma cell and mouse liver extracts. In linear sucrose gradients, the sedimentation velocity of PHGPx mRNA was impeded compared to CuZn superoxide dismutase (SOD) mRNA, which has a coding region of similar size. Selenium supplementation increased the loading of ribosomes onto PHGPx mRNA, but not CuZn SOD mRNA. To determine whether the slow sedimentation velocity of PHGPx mRNA is due to a block in elongation, we analyzed the polysome distribution of wild-type and mutant mRNAs translated in vitro. Mutation of the UGA/Sec codon to UGU/cysteine increased ribosome loading and protein synthesis. When UGA/Sec was replaced with UAA or when the SECIS element core was deleted, the distribution of the mutant mRNAs was similar to the wild-type mRNA. Addition of SECIS-binding protein SBP2, which is essential for Sec insertion, increased ribosome loading and translation of wild-type PHGPx mRNA, but had no effect on the mutant mRNAs. These results suggest that elongation is impeded at UGA/Sec, and that selenium and SBP2 alleviate this block by promoting Sec incorporation instead of termination.

The structural requirements for assembly of functional class II transfer RNA core regions have been examined by sequence analysis and tested by reconstruction of alternative folds into the tertiary domain of Escherichia coli tRNA2Gln. At least four distinct designs have been identified that permit stable folding and efficient synthetase recognition, as assessed by thermal melting profiles and glutaminylation kinetics. Although most large variable-arm tRNAs found in nature possess an enlarged D-loop, lack of this feature can be compensated for by insertion of nucleotides either 3′ to the variable loop or within the short acceptor/D-stem connector region. Rare pyrimidines at nt 9 in the core region can be accommodated in the class II framework, but only if specific nucleotides are present either in the D-loop or 3′ to the variable arm. Glutaminyl-tRNA synthetase requires one or two unpaired uridines 3′ to the variable arm to efficiently aminoacylate several of the class II frameworks. Because there are no specific enzyme contacts in the tRNAGln core region, these data suggest that tRNA discrimination by GlnRS depends in part on indirect readout of RNA sequence information.

RAE1 is an evolutionarily conserved protein that associates with both mRNPs and nucleoporins, and may bridge the interaction between mRNP export cargoes and the nuclear pore complex (NPC). However, the mechanism by which RAE1 functions in mRNA export is still unknown and the time point at which RAE1 interacts with the exported RNP has not been directly investigated. Here we have addressed this question in the Balbiani ring (BR) system of Chironomus tentans using immunoelectron microscopy. The RAE1 protein of C. tentans, Ct-RAE1, is 70% identical to human RAE1/mrnp41 (hRAE1) and is recognized by antibodies raised against hRAE1. As in vertebrate cells, Ct-RAE1 is concentrated at the nuclear envelope and also dispersed throughout the nuclear interior. Here we show that Ct-RAE1 does not bind to the BR particle either cotranscriptionally or in the nucleoplasm. Instead, the interaction between Ct-RAE1 and the exported BR particle occurs at the NPC. Moreover, the localization of Ct-RAE1 at the NPC is correlated with the presence of an exported RNP in the NPC. Finally, the anti-RAE1 antibody does not label the cytoplasmic side of BR particles in transit through the central channel, which indicates that Ct-RAE1 either remains anchored at the nuclear side of the NPC during translocation of the RNP through the central channel or becomes transiently associated with the RNP but is rapidly released into the cytoplasm.

Introns in plant nuclear pre-mRNAs are highly enriched in U or U + A residues and this property is essential for efficient splicing. Moreover, 3′-untranslated regions (3′-UTRs) in plant pre-mRNAs are generally UA-rich and contain sequences that are important for the polyadenylation reaction. Here, we characterize two structurally related RNA-binding proteins (RBPs) from Nicotiana plumbaginifolia, referred to as RBP45 and RBP47, having specificity for oligouridylates. Both proteins contain three RBD-type RNA-binding domains and a glutamine-rich N-terminus, and share similarity with Nam8p, a protein associated with U1 snRNP in the yeast Saccharomyces cerevisiae. Deletion analysis of RBP45 and RBP47 indicated that the presence of at least two RBD are required for interaction with RNA and that domains other than RBD do not significantly contribute to binding. mRNAs for RBP45 and RBP47 and mRNAs encoding six related proteins in Arabidopsis thaliana are constitutively expressed in different plant organs. Indirect immunofluorescence and fractionation of cell extracts showed that RBP45 and RBP47 are localized in the nucleus. In vivo UV crosslinking experiments demonstrated their association with the nuclear poly(A)+ RNA. In contrast to UBP1, another oligouridylate-binding nuclear three-RBD protein of N. plumbaginifolia (Lambermon et al., EMBO J, 2000, 19:1638–1649), RBP45 and RBP47 do not stimulate mRNA splicing and accumulation when transiently overexpressed in protoplasts. Properties of RBP45 and RBP47 suggest they represent hnRNP-proteins participating in still undefined steps of pre-mRNA maturation in plant cell nuclei.

Nmd3p from yeast is required for the export of the large (60S) ribosomal subunit from the nucleus (Ho et al., 2000). Here, we show that Nmd3p forms a stable complex with free 60S subunits. Using an epitope-tagged Nmd3p, we show that free 60S subunits can be coimmunoprecipitated with Nmd3p. The interaction was specific for 60S subunits; 40S subunits were not coimmunoprecipitated. Using this coprecipitation technique and pulse-chase labeling of ribosomal subunit proteins we showed that Nmd3p bound nascent subunits, consistent with its role in export. However, under conditions in which ribosome biogenesis was inhibited (e.g., inhibition of transcription with thiolutin, inhibition of transcription of ribosomal protein and RNA genes in a sly1-1 mutant at nonpermissive temperature, and inhibition of translation in a conditional prt1 mutant), Nmd3p remained associated with 60S subunits. In addition, Nmd3Δ120, a truncated protein that lacked a nuclear localization signal, retained 60S binding. These results suggest that Nmd3p recruits nascent 60S subunits into the pool of free 60S subunits and exchanges on 60S subunits as they recycle during translation.

The replication-dependent histone mRNAs end in a conserved 26-nt sequence that forms a stem-loop structure. This sequence is required for histone pre-mRNA processing and plays a role in multiple aspects of histone mRNA metabolism. Two proteins that bind the 3′ end of histone mRNA are found in Xenopus oocytes. xSLBP1 is found in the nucleus, where it functions in histone pre-mRNA processing, and in the cytoplasm, where it may control histone mRNA translation and stability. xSLBP2 is a cytoplasmic protein, inactive in histone pre-mRNA processing, whose expression is restricted to oogenesis and early development. These proteins are similar only in their RNA-binding domains (RBD). A chimeric protein (1-2-1) in which the RBD of xSLBP1 has been replaced with the RBD of xSLBP2 binds the stem-loop with an affinity similar to the original protein. The 1-2-1 protein efficiently localizes to the nucleus of the frog oocyte, but is not active in processing of histone pre-mRNA in vivo. This protein does not support processing in a nuclear extract, but inhibits processing by competing with the active SLBP by binding to the substrate. The 1-2-1 protein also inhibits processing of synthetic histone pre-mRNA injected into frog oocytes, but has no effect on processing of histone pre-mRNA transcribed from an injected histone gene. This result suggests that sequences in the RBD of xSLBP1 give it preferential access to histone pre-mRNA transcribed in vivo.

Bacterial ribosomal protein S7 initiates the folding of the 3′ major domain of 16S ribosomal RNA by binding to its lower half. The X-ray structure of protein S7 from thermophilic bacteria was recently solved and found to be a modular structure, consisting of an α-helical domain with a β-ribbon extension. To gain further insights into its interaction with rRNA, we cloned the S7 gene from Escherichia coli K12 into a pET expression vector and introduced 4 deletions and 12 amino acid substitutions in the protein sequence. The binding of each mutant to the lower half of the 3′ major domain of 16S rRNA was assessed by filtration on nitrocellulose membranes. Deletion of the N-terminal 17 residues or deletion of the β hairpins (residues 72–89) severely decreased S7 affinity for the rRNA. Truncation of the C-terminal portion (residues 138–178), which includes part of the terminal α-helix, significantly affected S7 binding, whereas a shorter truncation (residues 148–178) only marginally influenced its binding. Severe effects were also observed with several strategic point mutations located throughout the protein, including Q8A and F17G in the N-terminal region, and K35Q, G54S, K113Q, and M115G in loops connecting the α-helices. Our results are consistent with the occurrence of several sites of contact between S7 and the 16S rRNA, in line with its role in the folding of the 3′ major domain.

Pre-rRNA processing in eukaryotic cells requires participation of several snoRNPs. These include the highly conserved and abundant U3 snoRNP, which is essential for synthesis of 18S rRNA. Here we report the characterization of Rrp9p, a novel yeast U3 protein, identified via its homology to the human U3-55k protein. Epitope-tagged Rrp9p specifically precipitates U3 snoRNA, but Rrp9p is not required for the stable accumulation of this snoRNA. Genetic depletion of Rrp9p inhibits the early cleavages of the primary pre-rRNA transcript at A0, A1, and A2 and, consequently, production of 18S, but not 25S and 5.8S, rRNA. The hU3-55k protein can partially complement a yeast rrp9 null mutant, indicating that the function of this protein has been conserved. Immunoprecipitation of extracts from cells that coexpress epitope-tagged Rrp9p and various mutant forms of U3 snoRNA limits the region required for association of Rrp9p to the U3-specific box B/C motif. Box C is essential, whereas box B plays a supportive role.

Lanthanide metals such as terbium have previously been shown to be useful for mapping metal-binding sites in RNA. Terbium binds to the same sites on RNA as magnesium, however, with a much higher affinity. Thus, low concentrations of terbium ions can easily displace magnesium and promote phosphodiester backbone scission. At higher concentrations, terbium cleaves RNA in a sequence-independent manner, with a preference for single-stranded, non-Watson–Crick base-paired regions. Here, we show that terbium is a sensitive probe of human tRNALys,3 tertiary structure and folding. When 1 μM tRNA is used, the optimal terbium ion concentration for detecting Mg2+-induced tertiary structural changes is 50–60 μM. Using these concentrations of RNA and terbium, a magnesium-dependent folding transition with a midpoint (KMg) of 2.6 mM is observed for unmodified human tRNALys,3. At lower Tb3+ concentrations, cleavage is restricted to nucleotides that constitute specific metal-binding pockets. This small chemical probe should also be useful for detecting protein induced structural changes in RNA.