The carboxyl-terminal domain (CTD) of the largest subunit of eukaryotic RNA polymerase II (pol II) plays an important role in promoting steps of pre-mRNA processing. To identify proteins in human cells that bind to the CTD and that could mediate its functions in pre-mRNA processing, we used the mouse CTD expressed in bacterial cells in affinity chromatography experiments. Two proteins present in HeLa cell extract, the splicing and transcription-associated factors, PSF and p54nrb/NonO, bound specifically and could be purified to virtual homogeneity by chromatography on immobilized CTD matrices. Both hypo- and hyperphosphorylated CTD matrices bound these proteins with similar selectivity. PSF and p54nrb/NonO also copurified with a holoenzyme form of pol II containing hypophosphorylated CTD and could be coimmunoprecipitated with antibodies specific for this and the hyperphosphorylated form of pol II. That PSF and p54nrb/NonO promoted the binding of RNA to immobilized CTD matrices suggested these proteins can interact with the CTD and RNA simultaneously. PSF and p54nrb/NonO may therefore provide a direct physical link between the pol II CTD and pre-mRNA processing components, at both the initiation and elongation phases of transcription.

RF3 was initially characterized as a factor that stimulates translational termination in an in vitro assay. The factor has a GTP binding site and shows sequence similarity to elongation factors EF-Tu and EF-G. Paradoxically, addition of GTP abolishes RF3 stimulation in the classical termination assay, using stop triplets.

We here show GTP hydrolysis, which is only dependent on the simultaneous presence of RF3 and ribosomes. Applying a new termination assay, which uses a minimessenger RNA instead of separate triplets, we show that GTP in the presence of RF3 stimulates termination at rate-limiting concentrations of RF1. We show that RF3 can substitute for EF-G in RRF-dependent ribosome recycling reactions in vitro. This activity is GTP-dependent. In addition, excess RF3 and RRF in the presence of GTP caused release of nonhydrolyzed fmet-tRNA. This supports previous genetic experiments, showing that RF3 might be involved in ribosomal drop off of peptidyl-tRNA. In contrast to GTP involvement of the above reactions, stimulation of termination with RF2 by RF3 was independent of the presence of GTP. This is consistent with previous studies, indicating that RF3 enhances the affinity of RF2 for the termination complex without GTP hydrolysis. Based on our results, we propose a model of how RF3 might function in translational termination and ribosome recycling.

The Upf1 protein in yeast has been implicated in the modulation of efficient translation termination as well as in the accelerated turnover of mRNAs containing premature stop codons, a phenomenon called nonsense-mediated mRNA decay (NMD). A human homolog of the yeast UPF1, termed HUpf1/RENT1, has also been identified. The HUpf1 has also been shown to play a role in NMD in mammalian cells. Comparison of the yeast and human UPF1 proteins demonstrated that the amino terminal cysteine/histidine-rich region and the region comprising the domains that define this protein as a superfamily group I helicase have been conserved. The yeast Upf1p demonstrates RNA-dependent ATPase and 5′ → 3′ helicase activities. In this paper, we report the expression, purification, and characterization of the activities of the human Upf1 protein. We demonstrate that human Upf1 protein displays a nucleic-acid-dependent ATPase activity and a 5′ → 3′ helicase activity. Furthermore, human Upf1 is an RNA-binding protein whose RNA-binding activity is modulated by ATP. Taken together, these results indicate that the activities of the Upf1 protein are conserved across species, reflecting the conservation of function of this protein throughout evolution.

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.

RNA export from the nucleus is thought to be linked to proper processing and packaging into ribonucleoprotein protein complexes. A system to observe mRNA nuclear export in living yeast cells was developed by fusing the U1A RNA-binding protein to the green fluorescent protein to follow specific mRNAs with U1A hairpins engineered into them. RNAs encoding Rpl25, Pgk1, and Ssa4 were examined for the effects of 3′ UTRs, introns, RNA processing factors, nucleoporins, and transport factors on their export. All accumulated in the nucleus in mutants affecting components of the nuclear export machinery and certain nucleoporins. However, under conditions of stress, PGK1 and RPL25 transcripts accumulate in the nucleus whereas SSA4 RNA is exported. Moreover, when export is blocked, only RNAs containing the ASH1 3′ UTR accumulated in the nucleolus. Mutations in the splicing machinery selectively blocked export of only intron-containing RNAs. Mutations in RNA14, RNA15, and PAP1, which encode factors important for 3′ processing, also blocked export of all RNAs, including SSA4, thereby linking export to the process of polyadenlyation. Taken together, these data graphically display the connections between mRNA processing and nuclear export.

The use of double-stranded RNA (dsRNA) to disrupt gene expression has become a powerful method of achieving RNA interference (RNAi) in a wide variety of organisms. However, in Trypanosoma brucei this tool is restricted to transient interference, because the dsRNA is not stably maintained and its effects are diminished and eventually lost during cellular division. Here, we show that genetic interference by dsRNA can be achieved in a heritable and inducible fashion. To show this, we established stable cell lines expressing dsRNA in the form of stem-loop structures under the control of a tetracycline-inducible promoter. Targeting α-tubulin and actin mRNA resulted in potent and specific mRNA degradation as previously observed in transient interference. Surprisingly, 10-fold down regulation of actin mRNA was not fatal to trypanosomes. This type of approach could be applied to study RNAi in other organisms that are difficult to microinject or electroporate. Furthermore, to quickly probe the consequences of RNAi for a given gene we established a highly efficient in vivo T7 RNA polymerase system for expression of dsRNA. Using the α-tubulin test system we obtained greater than 98% transfection efficiency and the RNAi response lasted at least two to three cell generations. These new developments make it possible to initiate the molecular dissection of RNAi both biochemically and genetically.

Cis-acting RNA signals are required for replication of positive-strand viruses such as the picornaviruses. Although these generally have been mapped to the 5′ and/or 3′ termini of the viral genome, RNAs derived from human rhinovirus type 14 are unable to replicate unless they contain an internal cis-acting replication element (cre) located within the genome segment encoding the capsid proteins. Here, we show that the essential cre sequence is 83–96 nt in length and located between nt 2318–2413 of the genome. Using dicistronic RNAs in which translation of the P1 and P2–P3 segments of the polyprotein were functionally dissociated, we further demonstrate that translation of the cre sequence is not required for RNA replication. Thus, although it is located within a protein-coding segment of the genome, the cre functions as an RNA entity. Computer folds suggested that cre sequences could form a stable structure in either positive- or minus-strand RNA. However, an analysis of mutant RNAs containing multiple covariant and non-covariant nucleotide substitutions within these putative structures demonstrated that only the predicted positive-strand structure is essential for efficient RNA replication. The absence of detectable minus-strand synthesis from RNAs that lack the cre suggests that the cre is required for initiation of minus-strand RNA synthesis. Since a lethal 3′ noncoding region mutation could be partially rescued by a compensating mutation within the cre, the cre appears to participate in a long-range RNA–RNA interaction required for this process. These data provide novel insight into the mechanisms of replication of a positive-strand RNA virus, as they define the involvement of an internally located RNA structure in the recognition of viral RNA by the viral replicase complex. Since internally located RNA replication signals have been shown to exist in several other positive-strand RNA virus families, these observations are potentially relevant to a wide array of related viruses.

A number of heuristic descriptors have been developed previously in conjunction with the mfold package that describe the propensity of individual bases to participate in base pairs and whether or not a predicted helix is “well-determined.” They were developed for the “energy dot plot” output of mfold. Two descriptors, P-num and H-num, are used to measure the level of promiscuity in the association of any given nucleotide or helix with alternative complementary pairs. The third descriptor, S-num, measures the propensity of bases to be single-stranded. In the current work, we describe a series of programs that were developed in order to annotate individual structures with “well-definedness” information. We use color annotation to present the information. The programs can annotate PostScript files that are created by the mfold package or the PostScript secondary structure plots produced by the Weiser and Noller program XRNA (Weiser B, Noller HF, 1995, XRNA: Auto-interactive program for modeling RNA, The Center for Molecular Biology of RNA, Santa Cruz, California: University of California; Internet: ftp://fangio.ucsc.edu/pub/XRNA). In addition, these programs can annotate ss files that serve as input to XRNA. The annotation package can also handle structure comparison with a reference structure. This feature can be used to compare predicted structure with a phylogenetically deduced model, to compare two different predicted foldings, and to identify conformational changes that are predicted between wild-type and mutant RNAs.

We provide several examples of application. Predicted structures of two RNase P RNAs were colored with P-num information and further annotated with comparative information. The comparative model of a 16S rRNA was annotated with P-num information from mfold and with base pair probabilities obtained from the Vienna RNA folding package. Further annotation adds comparisons with the optimal foldings obtained from mfold and the Vienna package, respectively. The results of all of these analyses are discussed in the context of the reliability of structure prediction.

Eukaryote ribosomal translation is terminated when release factor eRF1, in a complex with eRF3, binds to one of the three stop codons. The tertiary structure and dimensions of eRF1 are similar to that of a tRNA, supporting the hypothesis that release factors may act as molecular mimics of tRNAs. To identify the yeast eRF1 stop codon recognition domain (analogous to a tRNA anticodon), a genetic screen was performed to select for mutants with disabled recognition of only one of the three stop codons. Nine out of ten mutations isolated map to conserved residues within the eRF1 N-terminal domain 1. A subset of these mutants, although wild-type for ribosome and eRF3 interaction, differ in their respective abilities to recognize each of the three stop codons, indicating codon-specific discrimination defects. Five of six of these stop codon-specific mutants define yeast domain 1 residues (I32, M48, V68, L123, and H129) that locate at three pockets on the eRF1 domain 1 molecular surface into which a stop codon can be modeled. The genetic screen results and the mutant phenotypes are therefore consistent with a role for domain 1 in stop codon recognition; the topology of this eRF1 domain, together with eRF1-stop codon complex modeling further supports the proposal that this domain may represent the site of stop codon binding itself.

RNAIII, a 514-nt RNA molecule, regulates the expression of many Staphylococcus aureus genes encoding exoproteins and cell-wall-associated proteins. We have studied the structure of RNAIII in solution, using a combination of chemical and enzymatic probes. A model of the secondary structure was derived from experimental data with the help of computer simulation of RNA folding. The model contains 14 hairpin structures connected by unpaired nucleotides. The data also point to three helices formed by distant nucleotides that close off structural domains. This model was generally compatible with the results of in vivo probing experiments with dimethylsulfate in late exponential-phase cultures. Toe-printing experiments revealed that the ribosome binding site of hld, which is encoded by RNAIII, was accessible to the Escherichia coli 30S ribosomal subunit, suggesting that the in vitro structure represented a translatable form of RNAIII. We also found that, within the 3′ end of RNAIII, the conserved hairpin 13 and the terminator form an intrinsic structural domain that exerts specific regulatory activity on protein A gene expression.

The hammerhead ribozyme is a small RNA motif that self cleaves at a specific phosphodiester bond to produce 2′,3′ cyclic phosphate and 5′ hydroxyl termini (Hutchins et al., 1986; Forster & Symons, 1987a). The secondary structure of the hammerhead consists of three helices of arbitrary sequence and length (designated I, II, and III) that intersect at 15 nucleotides termed the catalytic core (Fig. 1A) (Forster & Symons, 1987b; Hertel et al., 1992). The X-ray crystal structures of two hammerhead ribozyme–inhibitor complexes revealed that the core residues fold into two separate domains and the helices are arranged in a Y-shape conformation with helix I and helix II forming the upper portion of the Y (Pley et al., 1994; Scott et al., 1995). Although the hammerhead is found as an intramolecular motif embedded in several RNAs in vivo (Symons, 1989), it can be assembled from two separate oligonucleotides (Fig. 1B) in three different arrangements (Uhlenbeck, 1987; Haseloff & Gerlach, 1988; Koizumi et al., 1988; Jeffries & Symons, 1989). In these bimolecular formats, the hammerhead effects RNA cleavage in a similar manner to a true “enzyme,” proceeding through multiple rounds of substrate binding, cleavage, and product release (Uhlenbeck, 1987).

The yeast superkiller (SKI) genes were originally identified from mutations allowing increased production of killer toxin encoded by M “killer” virus, a satellite of the dsRNA virus L-A. XRN1 (SKI1) encodes a cytoplasmic 5′-exoribonuclease responsible for the majority of cytoplasmic RNA turnover, whereas SKI2, SKI3, and SKI8 are required for normal 3′-degradation of mRNA and for repression of translation of poly(A) minus RNA. Ski2p is a putative RNA helicase, Ski3p is a tetratricopeptide repeat (TPR) protein, and Ski8p contains five WD-40 (beta-transducin) repeats. An xrn1 mutation in combination with a ski2, ski3, or ski8 mutation is lethal, suggesting redundancy of function. Using functional epitope-tagged Ski2, Ski3, and Ski8 proteins, we show that Ski2p, Ski3p, and Ski8p can be coimmunoprecipitated as an apparent heterotrimeric complex. With epitope-tagged Ski2p, there was a 1:1:1 stoichiometry of the proteins in the complex. Ski2p did not associate with Ski3p in the absence of Ski8p, nor did Ski2p associate with Ski8p in the absence of Ski3p. However, the Ski3p/Ski8p interaction did not require Ski2p. In addition, ski6-2 or ski4-1 mutations or deletion of SKI7 did not affect complex formation. The identification of a complex composed of Ski2p, Ski3p, and Ski8p explains previous results showing phenotypic similarity between mutations in SKI2, SKI3, and SKI8. Indirect immunofluorescence of Ski3p and subcellular fractionation of Ski2p and Ski3p suggest that Ski2p and Ski3p are cytoplasmic. These data support the idea that Ski2p, Ski3p, and Ski8p function in the cytoplasm in a 3′-mRNA degradation pathway.

Although the hammerhead reaction proceeds most efficiently in divalent cations, cleavage in 4 M LiCl is only ∼10-fold slower than under standard conditions of 10 mM MgCl2 (Murray et al., Chem Biol, 1998, 5:587–595; Curtis & Bartel, RNA, 2001, this issue, pp. 546–552). To determine if the catalytic mechanism with high concentrations of monovalent cations is similar to that with divalent cations, we compared the activities of a series of modified hammerhead ribozymes in the two ionic conditions. Nearly all of the modifications have similar deleterious effects under both reaction conditions, suggesting that the hammerhead adopts the same general catalytic structure with both monovalent and divalent cations. However, modification of three ligands previously implicated in the binding of a functional divalent metal ion have substantially smaller effects on the cleavage rate in Li+ than in Mg2+. This result suggests that an interaction analogous to the interaction made by this divalent metal ion is absent in the monovalent reaction. Although the contribution of this divalent metal ion to the overall reaction rate is relatively modest, its presence is needed to achieve the full catalytic rate. The role of this ion appears to be in facilitating formation of the active structure, and any direct chemical role of metal ions in hammerhead catalysis is small.

AU-rich elements (AREs) located in the 3′ UTRs of the messenger RNAs (mRNAs) of many mammalian early response genes promote rapid mRNA turnover. HuR, an RRM-containing RNA-binding protein, specifically interacts with AREs, stabilizing these mRNAs. HuR is primarily nucleoplasmic, but shuttles between the nucleus and the cytoplasm via a domain called HNS located between RRM2 and RRM3. We recently showed that HuR interacts with two protein ligands, pp32 and APRIL, which are also shuttling proteins, but rely on NES domains recognized by CRM1 for export. Here we show that heat shock induces increased association of HuR with pp32 and APRIL through protein–protein interactions and that these ligands partially colocalize with HuR in cytoplasmic foci. HuR associations with the hnRNP complex also increase, but through RNA links. CRM1 coimmunoprecipitates with HuR only after heat shock, and nuclear export of HuR becomes sensitive to leptomycin B, an inhibitor of CRM1. Export after heat shock requires the same domains of HuR (HNS and RRM3) that are essential for binding pp32 and APRIL. In situ hybridization and coimmunoprecipitation experiments show that LMB treatment blocks both hsp70 mRNA nuclear export and its cytoplasmic interaction with HuR after heat shock. Together, our results argue that upon heat shock, HuR switches its export pathway to that of its ligands pp32 and APRIL, which involves the nuclear export factor CRM1. HuR and its ligands may be instrumental in the nuclear export of heat-shock mRNAs.

Alphavirus genome replication is a multistep asymmetric process. Several lines of evidence suggest that the template preference of the RNA replicase is regulated by proteolytic cleavage of the viral nonstructural polyprotein. Cis-acting RNA elements in the viral genome also play crucial roles in regulating genome replication and subgenomic RNA transcription. In this report, a series of RNA templates were analyzed in vitro and in vivo to define functional elements in the 5′ end of the genome. The 5′ UTR was shown to contain distinct core promoter elements for both minus- and plus-strand synthesis. In addition, two conserved stem-loop structures within the nsP1 coding sequence enhanced RNA replication but were not required. Studies with chimeric templates and trans-competition experiments suggest that the 5′ determinant for minus-strand initiation can differ among alphaviruses and binds to one or more limiting replicase components. The results provide compelling evidence that the 5′ and 3′ ends of alphavirus genome RNAs must interact to initiate replication and we propose one model for how this interaction might occur. In addition to providing new insight into the initiation of alphavirus genome replication, these results have implications for the development of improved alphavirus vector systems with reduced recombination potential.

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.

The translation of picornavirus genomic RNAs occurs by a cap-independent mechanism that requires the formation of specific ribonucleoprotein complexes involving host cell factors and highly structured regions of picornavirus 5′ noncoding regions known as internal ribosome entry sites (IRES). Although a number of cellular proteins have been shown to be involved in picornavirus RNA translation, the precise role of these factors in picornavirus internal ribosome entry is not understood. In this report, we provide evidence for the existence of distinct mechanisms for the internal initiation of translation between type I and type II picornavirus IRES elements. In vitro translation reactions were conducted in HeLa cell cytoplasmic translation extracts that were depleted of the cellular protein, poly(rC) binding protein 2 (PCBP2). Upon depletion of PCBP2, these extracts possessed a significantly diminished capacity to translate reporter RNAs containing the type I IRES elements of poliovirus, coxsackievirus, or human rhinovirus linked to luciferase; however, the addition of recombinant PCBP2 could reconstitute translation. Furthermore, RNA electrophoretic mobility-shift analysis demonstrated specific interactions between PCBP2 and both type I and type II picornavirus IRES elements; however, the translation of reporter RNAs containing the type II IRES elements of encephalomyocarditis virus and foot-and-mouth disease virus was not PCBP2 dependent. These data demonstrate that PCBP2 is essential for the internal initiation of translation on picornavirus type I IRES elements but is dispensable for translation directed by the structurally distinct type II elements.

Mammalian SR proteins are currently thought to function in mRNA export as well as splicing. They contain multiple phosphorylated serine/arginine (RS/SR) dipeptides. Although SR domains can be phosphorylated by many kinases in vitro, the physiologically relevant kinase(s), and the role(s) of these modifications in vivo have remained unclear. Npl3 is a shuttling protein in budding yeast that we showed previously to be a substrate for the mammalian SR protein kinase, SRPK1, as well as the related yeast kinase, Sky1. Here we demonstrate that Sky1p phosphorylates only one of Npl3p's eight SR/RS dipeptides. Mutation of the C-terminal RS to RA, or deletion of SKY1, results in the cytoplasmic accumulation of Npl3p. The redistribution of Npl3p is accompanied by its increased association with poly(A)+ RNA and decreased association with its import receptor, Mtr10p, in vivo. We propose that phosphorylation of Npl3p by the cytoplasmically localized Sky1p is required for efficient release of mRNA upon termination of export.

The human coronavirus 229E replicase gene encodes a protein, p66HEL, that contains a putative zinc finger structure linked to a putative superfamily (SF) 1 helicase. A histidine-tagged form of this protein, HEL, was expressed using baculovirus vectors in insect cells. The purified recombinant protein had in vitro ATPase activity that was strongly stimulated by poly(U), poly(dT), poly(C), and poly(dA), but not by poly(G). The recombinant protein also had both RNA and DNA duplex-unwinding activities with 5′-to-3′ polarity. The DNA helicase activity of the enzyme preferentially unwound 5′-oligopyrimidine-tailed, partial-duplex substrates and required a tail length of at least 10 nucleotides for effective unwinding. The combined data suggest that the coronaviral SF1 helicase functionally differs from the previously characterized RNA virus SF2 helicases.

To provide a bridge between in vivo and in vitro studies of eukaryotic translation initiation, we have developed a reconstituted translation initiation system using components from the yeast Saccharomyces cerevisiae. We have purified a minimal set of initiation factors (eIFs) that, together with yeast 80S ribosomes, GTP, and initiator methionyl-tRNA, are sufficient to assemble active initiation complexes on a minimal mRNA template. The kinetics of various steps in the pathway of initiation complex assembly and the formation of the first peptide bond in vitro have been explored. The formation of active initiation complexes in this system is dependent on ribosomes, mRNA, Met-tRNAi, GTP hydrolysis, eIF1, eIF1A, eIF2, eIF5, and eIF5B. Our data indicate that eIF1 and eIF1A both facilitate the binding of the eIF2[bull]GTP[bull]Met-tRNAi complex to the 40S ribosomal subunit to form the 43S complex. eIF5 stimulates a step after 43S complex formation, consistent with its proposed role in activating GTP hydrolysis by eIF2 upon initiation codon recognition. The presence of eIF5B is required for the joining of the 40S and 60S subunits to form the 80S initiation complex. The step at which each of these factors acts in this reconstituted system is in agreement with previous data from in vivo studies and work using reconstituted mammalian systems, indicating that the system recapitulates fundamental events in translation initiation in eukaryotic cells. This system should allow us to couple powerful yeast genetic and molecular biological experiments with in vitro kinetic and biophysical experiments, yielding a better understanding of the molecular mechanics of this central, complex process.