Translational fidelity, programmed frameshifting, and nonsense-mediated decay (NMD) may require common proteins, reflecting related events in translation and ribosome function (reviewed in Czaplinski et al., 1999). This letter focuses on the possible link between programmed frameshifting and NMD, and seeks to reconcile apparently conflicting conclusions concerning a proposed overlap in the factors involved.

The complexity of the nonsense-mediated decay (NMD) system makes it difficult to study by comparing the expression of various single reporter constructs. The known effects of the NMD genes include a reduction both in mRNA stability (reviewed by Czaplinski et al., 1999) and in the efficiency of translational initiation (Muhlrad & Parker, 1999) of nonsense-containing plasmids as well as an apparent increase in the efficiency of translational termination as evidenced by increased readthrough of nonsense mutations (Bidou et al., 2000; Maderazo et al., 2000). The single reporter system can not distinguish among these effects and inference is required to determine which mechanism underlies any observed phenotypic effect on gene expression. It is particularly problematic to differentiate the effects of translation initiation accuracy from putative effects on translational frameshifting. The dual reporter system used in our work isolates the effect of translational frameshifting from effects on mRNA stability, initiation or termination. Much is made by Dinman et al. of the relative effects of various mutations, yet it remains unclear whether these are fundamental differences or simply differences in phenotypic strength of the various mutations.

Twenty-five years ago, a basic tenet of biochemistry held that enzymes are made solely of protein. Ribosomes, which are large ribonucleoprotein enzymes, were conveniently overlooked (but see Woese, 1967; Crick, 1968; Orgel, 1968). The then current orthodoxy was first challenged by the discovery that Escherichia coli RNase P had an essential RNA subunit (Stark et al., 1978) and completely demolished when the catalytic properties of RNA were described (Kruger et al., 1982; Guerrier-Takada et al., 1983; Buzayan et al., 1986; Hutchins et al., 1986). With respect to RNase P, in particular, a new orthodoxy arose: RNase P from any source would have an essential RNA subunit although not necessarily a catalytic one. This adolescent orthodoxy is now being challenged by reports of putative RNase P activities isolated from certain organelles that lack an RNA subunit (Manam & van Tuyle, 1987; Rossmanith & Karwan, 1998; Wang et al., 1998). The following questions must be asked of the aforementioned data: (1) Are the substrate specificities the same as in the canonical reaction? (2) Are the chemical mechanisms of the newly identified “RNase P” reactions identical to the canonical reaction? (3) Have completely purified (i.e., cloned, expressed, and/or purified) activities been studied? (4) Have the phenotypes of these “new” enzymes been verified by genetic means? (5) In sum, are we dealing with RNase P or some other activities that share some characteristics with RNase P? Should these be called RNase P, in any case?

Altman and colleagues (this issue) call attention to the inability of current standardized enzyme nomenclature to distinguish between enzymatic activities that reside in nonhomologous macromolecules. This issue is highlighted by the fact that the pre-tRNA 5′-maturation activities of bacteria and plant chloroplasts present the first instance (of which I am aware) of two naturally occurring enzymes that cannot be evolutionarily related, but which catalyze an identical reaction. (In the classic example of convergent evolution between the trypsin family and subtilisin, the enzymes do not have an identical substrate specificity.) Altman and colleagues propose that a single trivial name be used only for members of a family of homologous macromolecules; in other words, that different trivial names be given to enzymes that catalyze the same precursor–product conversion but do so with different catalytic mechanisms, or which are not members of a single family of homologous macromolecules.

The final stage in the formation of the two large subunit rRNA species in Saccharomyces cerevisiae is the removal of internal transcribed spacer 2 (ITS2) from the 27SB precursors. This removal is initiated by endonucleolytic cleavage approximately midway in ITS2. The resulting 7S pre-rRNA, which is easily detectable, is then converted into 5.8S rRNA by the concerted action of a number of 3′ → 5′ exonucleases, many of which are part of the exosome. So far the complementary precursor to 25S rRNA resulting from the initial cleavage in ITS2 has not been detected and the manner of its conversion into the mature species is unknown. Using various yeast strains that carry different combinations of wild-type and mutant alleles of the major 5′ → 3′ exonucleases Rat1p and Xrn1p, we now demonstrate the existence of a short-lived 25.5S pre-rRNA whose 5′ end is located closely downstream of the previously mapped 3′ end of 7S pre-rRNA. The 25.5S pre-rRNA is converted into mature 25S rRNA by rapid exonucleolytic trimming, predominantly carried out by Rat1p. In the absence of Rat1p, however, the removal of the ITS2 sequences from 25.5S pre-rRNA can also be performed by Xrn1p, albeit somewhat less efficiently.

Replacing a cassette of 31 residues from Escherichia coli release factor 1 with the equivalent residues in release factor 2 gave a protein active in codon-specific binding to the ribosome but inactive in peptidyl-tRNA hydrolysis. Such a phenotype is also found unexpectedly with release factor 2 when expressed at high concentration in bacteria. Substituting threonine with the release factor 1 equivalent serine at position 246 within the cassette restored the impaired activity of the chimeric protein, and also that of inactive recombinant release factor 2, both in vitro and in vivo. The differences in activity are not due to posttranslational modifications or a lack of it at this residue. Random mutagenesis of codon 246 suggests that this position is pivotal for the function of the release factor, being able to affect differentially both its binding to the ribosome and its peptide release activities. We propose that amino acid 246 is close to a sharp turn (GGQ motif at position 250), and is essential for transmitting the signal from cognate codon recognition by correctly positioning the peptidyl-tRNA hydrolysis domain of the release factor into the peptidyltransferase center.

The complex of ribosomal protein L1 with 23S rRNA from Escherichia coli is of great interest because of the unique structural and functional aspects of this ribonucleoprotein domain. We have minimized the binding site for protein L1 on the 23S rRNA to nt 2120–2129, 2159–2162, and 2167–2178. This RNA fragment consists of two helices as well as an interconnecting loop of unknown structure. RNA molecules corresponding to the minimized L1 binding site, in which G, A, U, or C were individually replaced by their deoxyribo- (dN) or α-thio- (rNαS) analogs have been synthesized by T7 transcription in vitro and analyzed for their ability to bind protein L1. It has been demonstrated that the substitution of rNαS at position 2122 or 2176 decreases the affinity of the RNA for the protein in the presence of magnesium five- to tenfold, whereas the same changes have little effect on binding in the presence of manganese. This suggests that Rp oxygens in the phosphates preceding positions 2122 and 2176 are coordinated with Mg2+ and may participate in L1-23S rRNA interaction via magnesium bridges. We have also shown that this interaction is impaired by the presence of dC at position 2122 coupled with the presence of deoxyribonucleotide(s) at other positions in the RNA. This study demonstrates that the ribose–phosphate backbone of the helix encompassing nt 2120–2124/2174–2178 is intimately involved in the interaction of protein L1 with the 23S rRNA. In particular, we suggest that this helix is positioned in the cleft between the two domains of protein L1.

The 18S rRNA environment of the mRNA at the decoding site of human 80S ribosomes has been studied by crosslinking with derivatives of hexaribonucleotide UUUGUU (comprising Phe and Val codons) that carried a perfluorophenylazide group either at the N7 atom of the guanine or at the C5 atom of the 5′-terminal uracil residue. The location of the codons on the ribosome at A, P, or E sites has been adjusted by the cognate tRNAs. Three types of complexes have been obtained for each type derivative, namely, (1) codon UUU and Phe-tRNAPhe at the P site (codon GUU at the A site), (2) codon UUU and tRNAPhe at the P site and PheVal-tRNAVal at the A site, and (3) codon GUU and Val-tRNAVal at the P site (codon UUU at the E site). This allowed the placement of modified nucleotides of the mRNA analog at positions −3, +1, or +4 on the ribosome. Mild UV irradiation resulted in tRNA-dependent crosslinking of the mRNA analogs to the 18S rRNA. Nucleotide G961 crosslinked to mRNA position −3, nucleotide G1207 to position +1, and A1823 together with A1824 to position +4. All of these nucleotides are located in the most strongly conserved regions of the small subunit RNA structure, and correspond to nucleotides G693, G926, G1491, and A1492 of bacterial 16S rRNA. Three of them (with the exception of G1491) had been found earlier at the 70S ribosomal decoding site. The similarities and differences between the 16S and 18S rRNA decoding sites are discussed.

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.

Although rRNA synthesis, maturation, and assembly into preribosomal particles occur within the nucleolus, the route taken by pre-rRNAs from their synthetic sites toward the cytoplasm remains largely unexplored. Here, we employed a nondestructive method for the incorporation of BrUTP into the RNA of living cells. By using pulse-chase experiments, three-dimensional image reconstructions of confocal optical sections, and electron microscopy analysis of ultrathin sections, we were able to describe topological and spatial dynamics of rRNAs within the nucleolus. We identified the precise location and the volumic organization of four typical subdomains, in which rRNAs are successively moving towards the nucleolar periphery during their synthesis and processing steps. The incorporation of BrUTP takes place simultaneously within several tiny spheres, centered on the fibrillar centers. Then, the structures containing the newly synthesized RNAs enlarge and appear as compact ringlets disposed around the fibrillar centers. Later, they form hollow spheres surrounding the latter components and begin to fuse together. Finally, these structures widen and form large rings reaching the limits of the nucleoli. These results clearly show that the transport of pre-rRNAs within the nucleolus does not occur randomly, but appears as a radial flow starting from the fibrillar centers that form concentric rings, which finally fuse together as they progress toward the nucleolar periphery.

Human TAP and Saccharomyces cerevisiae Mex67p belong to a family of proteins that mediate mRNA export. Computer searches identified previously two Caenorhabditis elegans genes, C15H11.3 and C15H11.6, that encode putative homologs of hTAP and Mex67p (Segref et al., EMBO J, 1997, 16:3256–3271). Using RNA interference experiments in C. elegans, we found that functional knockout of C15H11.3 resulted in nuclear accumulation of poly(A)-containing RNAs and was lethal for both embryos and adult nematodes. No embryonic or progeny abnormality was observed in functional knockout of C15H11.6. Taken together, these data established that the C15H11.3 gene product is an ortholog of hTAP and Mex67p; thus, it was named Ce-NXF-1. Ce-NXF-1 binds RNA directly and is a nucleocytoplasmic shuttle protein accumulating in the nucleoplasm and at the nuclear rim. The rim association is mediated via unique signals present in the C-terminal portion of all TAP/NXF and Mex67p proteins. This region was shown to interact with the FG-repeat domains of nucleoporins Nup98, Nup153, and Nup214, indicating that the rim association occurs through components of the nuclear pore complex. In summary, Ce-NXF-1 belongs together with hTAP and Mex67p to a family of proteins that participate in mRNA export and can provide a direct molecular link between mRNAs and components of the nuclear pore complex. Therefore, despite differences in mRNA metabolism between these species, they utilize a conserved mRNA transport mechanism.

The transcript of the Saccharomyces cerevisiae gene, RPL30, is subject to regulated splicing and regulated translation, due to a structure that interacts with its own product, ribosomal protein L30. We have followed the fate of the regulated RPL30 transcripts in vivo. Initially, these transcripts abortively enter the splicing pathway, forming an unusually stable association with U1 snRNP. A large proportion of the unspliced molecules, however, are found in the cytoplasm. Most of these are still bound by L30, as only a small fraction are engaged in translation. Eventually, the unspliced RPL30 transcripts escape the grasp of L30, associate with ribosomes, and fall prey to nonsense mediated decay.

In eukaryotic cells, efficient translation of most cellular mRNAs requires the synergistic interplay between the m7GpppN cap structure and the poly(A) tail during initiation. We have developed and characterized a cell-free system from human HeLa cells that recapitulates this important feature, displaying more than one order of magnitude of translational synergism between the cap structure and the poly(A) tail. The stimulation of cap-dependent translation by the poly(A) tail is length-dependent, but not mediated by changes in mRNA stability. Using this system, we investigated the effect of the poly(A) tail on the translation of picornaviral RNAs, which are naturally polyadenylated but initiate translation via internal ribosome entry sites (IRESs). We show that translation driven by the IRESs of poliovirus (PV), encephalomyocarditis virus (EMCV), and hepatitis A virus is also significantly augmented by a poly(A) tail, ranging from an approximately 3-fold stimulation for the EMCV-IRES to a more than 10-fold effect for the PV IRES. These results raise interesting questions concerning the underlying molecular mechanism(s). The cell-free system described here should prove useful in studying these questions as well as providing a general biochemical tool to examine the translation initiation pathway in a more physiological setting.

Most eukaryotic mRNAs require the cap-binding complex eIF4F for efficient initiation of translation, which occurs as a result of ribosomal scanning from the capped 5′ end of the mRNA to the initiation codon. A few cellular and viral mRNAs are translated by a cap and end-independent mechanism known as internal ribosomal entry. The internal ribosome entry site (IRES) of classical swine fever virus (CSFV) is ∼330 nt long, highly structured, and mediates internal initiation of translation with no requirement for eIF4F by recruiting a ribosomal 43S preinitiation complex directly to the initiation codon. The key interaction in this process is the direct binding of ribosomal 40S subunits to the IRES to form a stable binary complex in which the initiation codon is positioned precisely in the ribosomal P site. Here, we report the results of analyses done using enzymatic footprinting and mutagenesis of the IRES to identify structural components in it responsible for precise binding of the ribosome. Residues flanking the initiation codon and extending from nt 363–391, a distance equivalent to the length of the 40S subunit mRNA-binding cleft, were strongly protected from RNase cleavage, as were nucleotides in the adjacent pseudoknot and in the more distal subdomain IIId1. Ribosomal binding and IRES-mediated initiation were abrogated by disruption of helix 1b of the pseudoknot and very severely reduced by mutation of the protected residues in IIId1 and by disruption of domain IIIa. These observations are consistent with a model for IRES function in which binding of the region flanking the initiation codon to the decoding region of the ribosome is determined by multiple additional interactions between the 40S subunit and the IRES.

Barley yellow dwarf virus RNA lacks both a 5′ cap and a poly(A) tail, yet it is translated efficiently. It contains a cap-independent translation element (TE), located in the 3′ UTR, that confers efficient translation initiation at the AUG closest to the 5′ end of the mRNA. We propose that the TE must both recruit ribosomes and facilitate 3′-5′ communication. To dissect its function, we determined the secondary structure of the TE and roles of domains within it. Nuclease probing and structure-directed mutagenesis revealed that the 105-nt TE (TE105) forms a cruciform secondary structure containing four helices connected by single-stranded regions. TE105 can function in either UTR in wheat germ translation extracts. A longer viral sequence (at most 869 nt) is required for full cap-independent translation in plant cells. However, substantial translation of uncapped mRNAs can be obtained in plant cells with TE105 combined with a poly(A) tail. All secondary structural elements and most primary sequences that were mutated are required for cap-independent translation in the 3′ and 5′ UTR contexts. A seven-base loop sequence was needed only in the 3′ UTR context. Thus, this loop sequence may be involved only in communication between the UTRs and not directly in recruiting translational machinery. This structural and functional analysis provides a framework for understanding an emerging class of cap-independent translation elements distinguished by their location in the 3′ UTR.

The cleavage site of the Neurospora VS RNA ribozyme is located in a separate hairpin domain containing a hexanucleotide internal loop with an A-C mismatch and two adjacent G-A mismatches. The solution structure of the internal loop and helix Ia of the ribozyme substrate hairpin has been determined by nuclear magnetic resonance (NMR) spectroscopy. The 2 nt in the internal loop, flanking the cleavage site, a guanine and adenine, are involved in two sheared G.A base pairs similar to the magnesium ion-binding site of the hammerhead ribozyme. Adjacent to the tandem G.A base pairs, the adenine and cytidine, which are important for cleavage, form a noncanonical wobble A+-C base pair. The dynamic properties of the internal loop and details of the high-resolution structure support the view that the hairpin structure represents a ground state, which has to undergo a conformational change prior to cleavage. Results of chemical modification and mutagenesis data of the Neurospora VS RNA ribozyme can be explained in context with the present three-dimensional structure.

In its natural context, the hairpin ribozyme is constructed around a four-way helical junction. This presents the two loops that interact to form the active site on adjacent arms, requiring rotation into an antiparallel structure to bring them into proximity. In the present study we have compared the folding of this form of the ribozyme and subspecies lacking either the loops or the helical junction using fluorescence resonance energy transfer. The complete ribozyme as a four-way junction folds into an antiparallel structure by the cooperative binding of magnesium ions, requiring 20–40 μM for half-maximal extent of folding ([Mg2+]1/2) and a Hill coefficient n = 2. The isolated junction (lacking the loops) also folds into a corresponding antiparallel structure, but does so noncooperatively (n = 1) at a higher magnesium ion concentration ([Mg2+]1/2 = 3 mM). Introduction of a G + 1A mutation into loop A of the ribozyme results in a species with very similar folding to the simple junction, and complete loss of ribozyme activity. Removal of the junction from the ribozyme, replacing it either with a strand break (serving as a hinge) or a GC5 bulge, results in greatly impaired folding, with [Mg2+]1/2 > 20 mM. The results indicate that the natural form of the ribozyme undergoes ion-induced folding by the cooperative formation of an antiparallel junction and loop–loop interaction to generate the active form of the ribozyme. The four-way junction thus provides a scaffold in the natural RNA that facilitates the folding of the ribozyme into the active form.

The alternative splicing of the last intron (intron D) of bovine growth hormone (bGH) pre-mRNA requires a downstream exonic splicing enhancer (FP/ESE). The presence of at least one SR protein has been shown to be essential for FP/ESE function and splicing of intron D in in vitro splicing assays. However, in vitro reconstitution of splicing using individual purified SR proteins may not accurately reflect the true complexity of alternative splicing in an intact nucleus, where multiple SR proteins in varying amounts are likely to be available simultaneously. Here, a panel of recombinant baculovirus-expressed SR proteins was produced and tested for the ability to activate FP/ESE-dependent splicing. Individual recombinant SR proteins differed significantly in their activity in promoting intron D splicing. Among the recombinant SR proteins tested, SRp55 was the most active, SC35 showed very little activity, and ASF/SF2 and 9G8 individually had intermediate activity. At least one SR protein (ASF/SF2) bound to the FP/ESE with characteristics of a cooperative interaction. Most interestingly, low concentrations of ASF/SF2 and 9G8 acted synergistically to activate intron D splicing. This was due in part to synergistic binding to the FP/ESE. Splicing of bGH intron D is inherently complex, and is likely controlled by an interaction of the FP/ESE with several trans-acting protein factors acting both independently and cooperatively. This level of complexity may be required for precise control of alternative splicing by an exon sequence, which simultaneously is constrained to maintain translational integrity of the mature mRNA.

Primary transcripts made by RNA polymerase II (Pol II), but not Pol I or Pol III, are modified by addition of a 7-methylguanosine (m7G) residue to the triphosphate 5′ end shortly after it emerges from the polymerase. The m7G “caps” of small nuclear and small nucleolar RNAs, but not messenger RNAs, are subsequently hypermethylated to a 2,2,7-trimethylguanosine (TMG) residue. U6 RNA, the only small nuclear RNA synthesized by Pol III in most eukaryotes, does not receive a methylguanosine cap. However, human U6 RNA is O-methylated on the 5′-terminal (γ) phosphate by an enzyme that recognizes the 5′ stem-loop of U6. Here we show that variant yeast U6 RNAs truncated or substituted within the 5′ stem-loop are TMG capped in vivo. Accumulation of the most efficiently TMG-capped U6 RNA variant is strongly inhibited by a conditional mutation in the largest subunit of Pol III, confirming that it is indeed synthesized by Pol III. Thus, methylguanosine capping and cap hypermethylation are not exclusive to Pol II transcripts in yeast. We propose that TMG capping of variant U6 RNAs occurs posttranscriptionally due to exposure of the 5′ triphosphate by disruption of protein binding and/or γ-methyl phosphate capping. 5′ truncation and TMG capping of U6 RNA does not appear to affect its normal function in splicing, suggesting that assembly and action of the spliceosome is not very sensitive to the 5′ end structure of U6 RNA.

Previous work from this laboratory (Nurse et al., RNA, 1995, 1:102–112) established that TruB, a pseudouridine (Ψ) synthase from Escherichia coli, was able to make Ψ55 in tRNA transcripts but not in transcripts of full-length or fragmented 16S or 23S ribosomal RNAs. By deletion of the truB gene, we now show that TruB is the only protein in E. coli able to make Ψ55 in vivo. Lack of TruB and Ψ55 did not affect the exponential growth rate but did confer a strong selective disadvantage on the mutant when it was competed against wild-type. The negative selection did not appear to be acting at either the exponential or stationary phase. Transformation with a plasmid vector conferring carbenicillin resistance and growth in carbenicillin markedly increased the selective disadvantage, as did growth at 42 °C, and both together were approximately additive such that three cycles of competitive growth sufficed to reduce the mutant strain to ∼0.2% of its original value. The most striking finding was that all growth effects could be reversed by transformation with a plasmid carrying a truB gene coding for a D48C mutation in TruB. Direct analysis showed that this mutant did not make Ψ55 under the conditions of the competition experiment. Therefore, the growth defect due to the lack of TruB must be due to the lack of some other function of the protein, possibly an RNA chaperone activity, but not to the absence of Ψ55.