Internal initiation of translation.

Although the 5' cap-dependent scanning mechanism can account for the translational initiation of most mRNAs in eukaryotic cells, several viral and cellular mRNAs contain nucleotide sequences in their 5' non-coding regions that can mediate binding of ribosomes to the mRNA, regardless of the modification state of the 5' ends. During the past year, some nuclear proteins normally involved in RNA processing have been shown also to facilitate 'internal' ribosome binding. Unexpected dual functions have, therefore, been suggested for these RNA-binding proteins, in both RNA biogenesis in the nucleus and RNA translation in the cytoplasm.

Initiation of protein synthesis by the eukaryotic translational apparatus on circular RNAs.

The ribosome scanning model predicts that eukaryotic ribosomal 40S subunits enter all messenger RNAs at their 5' ends. Here, it is reported that eukaryotic ribosomes can initiate translation on circular RNAs, but only if the RNAs contain internal ribosome entry site elements. Long-repeating polypeptide chains were synthesized from RNA circles with continuous open reading frames. These results indicate that ribosomes can translate such RNA circles for multiple consecutive rounds and that the free 5' end of a messenger RNA is not necessarily the entry point for 40S subunits.

Gene regulation: translational initiation by internal ribosome binding.

During the past year, several examples of cellular mRNAs have been described in which translational initiation occurs by internal ribosome binding, a mechanism hitherto thought to be restricted to picornaviral RNAs. New insights into the molecular mechanism of internal ribosome entry have been provided by the structural and functional analyses of both the internal ribosome entry sites and the protein factors that stimulate translation mediated by these elements.

Evidence for involvement of trans-acting factors in selection of the AUG start codon during eukaryotic translational initiation.

The molecular mechanism with which an appropriate AUG codon is selected as the start site for translational initiation by eukaryotic ribosomes is not known. By using a cell-free translation system, small RNA molecules containing single AUG codons, surrounded by various nucleotide sequences, were tested for their abilities to interfere with the translation of a reporter mRNA. RNAs containing the AUG in an ACCAUGG context (Kozak consensus sequence) were able to inhibit translation of the reporter mRNA. In contrast, RNAs containing the AUG in a less favorable context for start site selection (for example, CAGAUGG) had no effect on the translation of the reporter mRNA. The effect mediated by the ACCAUGC-containing RNAs was not due to sequestration of ribosomal subunits or to particular structural features in these RNAs. To identify potential trans-acting factors that might be preferentially bound by ACCAUGG-containing RNAs, ACCAUGG- and CAGAUGC-containing RNAs with a single 4-thiouridine residue at the AUG were incubated with partially fractionated extracts, and AUG-binding proteins were identified after irradiation of the complexes with UV light and subsequent analysis by gel electrophoresis. The analysis (of such complexes in competition experiments revealed that proteins, approximately 50 and 100 kDa in size, were found to bind directly at the AUG codon embedded in the ACCAUGG motif. One of these proteins has been identified as the La autoantigen. These findings indicate that trans-acting factors may play a role in AUG start site selection during translational initiation.

Cap-dependent and cap-independent translation by internal initiation of mRNAs in cell extracts prepared from Saccharomyces cerevisiae.

Translation extracts were prepared from various strains of Saccharomyces cerevisiae. The translation of mRNA molecules in these extracts were cooperatively enhanced by the presence of 5'-terminal cap structures and 3'-terminal poly(A) sequences. These cooperative effects could not be observed in other translation systems such as those prepared from rabbit reticulocytes, wheat germ, and human HeLa cells. Because the yeast translation system mimicked the effects of the cap structure and poly(A) tail on translational efficiency seen in vivo, this system was used to study cap-dependent and cap-independent translation of viral and cellular mRNA molecules. Both the 5' noncoding regions of hepatitis C virus and those of coxsackievirus B1 conferred cap-independent translation to a reporter coding region during translation in the yeast extracts; thus, the yeast translational apparatus is capable of initiating cap-independent translation. Although the translation of most yeast mRNAs was cap dependent, the unusually long 5' noncoding regions of mRNAs encoding cellular transcription factors TFIID and HAP4 were shown to mediate cap-independent translation in these extracts. Furthermore, both TFIID and HAP4 5' noncoding regions mediated translation of a second cistron when placed into the intercistronic spacer region of a dicistronic mRNA, indicating that these leader sequences can initiate translation by an internal ribosome binding mechanism in this in vitro translation system. This finding raises the possibility that an internal translation initiation mechanism exists in yeast cells for regulated translation of endogenous mRNAs.

Global and specific translational regulation in the genomic response of Saccharomyces cerevisiae to a rapid transfer from a fermentable to a nonfermentable carbon source.

The global gene expression program that accompanies the adaptation of Saccharomyces cerevisiae to an abrupt transfer from a fermentable to a nonfermentable carbon source was characterized by using a cDNA microarray to monitor the relative abundances and polysomal distributions of mRNAs. Features of the program included a transient reduction in global translational activity and a severe decrease in polysome size of transcripts encoding ribosomal proteins. While the overall translation initiation of newly synthesized and preexisting mRNAs was generally repressed after the carbon source shift, the mRNA encoded by YPL250C was an exception in that it selectively mobilized into polysomes, although its relative abundance remained unchanged. In addition, splicing of HAC1 transcripts, which has previously been reported to occur during accumulation of unfolded proteins in the endoplasmic reticulum, was observed after the carbon shift. This finding suggests that the nonconventional splicing complex, composed of the kinase-endonuclease Ire1p and the tRNA ligase Rlg1p, was activated. While spliced HAC1 transcripts mobilized into polysomes, the vast majority of unspliced HAC1 RNA accumulated in nonpolysomal fractions before and after the carbon source shift, indicating that translation of unspliced HAC1 RNA is blocked at the translation initiation step, in addition to the previously reported elongation step. These findings reveal that S. cerevisiae reacts to the carbon source shift with a remarkable variety of responses, including translational regulation of specific mRNAs and activation of specific enzymes involved in a nonconventional splicing mechanism.

Naturally occurring dicistronic cricket paralysis virus RNA is regulated by two internal ribosome entry sites.

Cricket paralysis virus is a member of a group of insect picorna-like viruses. Cloning and sequencing of the single plus-strand RNA genome revealed the presence of two nonoverlapping open reading frames, ORF1 and ORF2, that encode the nonstructural and structural proteins, respectively. We show that each ORF is preceded by one internal ribosome entry site (IRES). The intergenic IRES is located 6,024 nucleotides from the 5' end of the viral RNA and is more active than the IRES located at the 5' end of the RNA, providing a mechanistic explanation for the increased abundance of structural proteins relative to nonstructural proteins in infected cells. Mutational analysis of this intergenic-region IRES revealed that ORF2 begins with a noncognate CCU triplet. Complementarity of this CCU triplet with sequences in the IRES is important for IRES function, pointing to an involvement of RNA-RNA interactions in translation initiation. Thus, the cricket paralysis virus genome is an example of a naturally occurring, functionally dicistronic eukaryotic mRNA whose translation is controlled by two IRES elements located at the 5' end and in the middle of the mRNA. This finding argues that eukaryotic mRNAs can express multiple proteins not only by polyprotein processing, reinitiation and frameshifting but also by using multiple IRES elements.

Regulation of host cell translation by viruses and effects on cell function.

Viruses have evolved a remarkable variety of strategies to modulate the host cell translation apparatus with the aim of optimizing viral mRNA translation and replication. Recent studies have revealed that modulation of both host and viral mRNA translation can be accomplished by selective alteration of translation factors in virus-infected cells. These findings provide new insights into the functioning of the translational apparatus in both uninfected and infected cells.

A histone H4-specific methyltransferase. Properties, specificity and effects on nucleosomal histones.

A histone H4-specific methyltransferase was purified 80-100-fold from nuclei of calf lymphocytes and from calf thymus. Some biochemical properties of the enzyme are described. The enzyme transfers in vitro methyl groups from S-adenosylmethionine preferentially to the lysine residue 20 of histone H4. This is the major in vivo methylation site of H4. DNA-bound or nucleosomal H4 is not methylated in vitro. We have used methylated and unmodified H4 (in the presence of sufficient quantities of the other core histones) for nucleosome reconstitution in vitro and have not found significant differences in the efficiencies of assembly.

Positive and negative modulation of viral and cellular mRNAs by liver-specific microRNA miR-122.

microRNAs (miRNAs) are small RNAs that in general down-regulate the intracellular abundance and translation of target mRNAs. We noted that sequestration of liver-specific miR-122 by modified antisense oligonucleotides resulted in a dramatic loss of hepatitis C virus (HCV) RNA in cultured human liver cells. A binding site for miR-122 was predicted to reside close to the 5' end of the viral genome, and its functionality was tested by mutational analyses of the miRNA-binding site in viral RNA, resulting in reduced intracellular viral RNA abundance. Importantly, ectopic expression of miR-122 molecules that contained compensatory mutations restored viral RNA abundance, revealing a genetic interaction between miR-122 and the viral RNA genome. Studies with replication-defective viral RNAs demonstrated that miR-122 affected mRNA abundance by positively modulating RNA replication. In contrast, interaction of miR-122 with the 3'-noncoding region (3'NCR) of the cellular mRNA encoding the cationic amino acid transporter CAT-1 resulted in the down-regulation of CAT-1 protein abundance. These findings provide evidence that a specific miRNA can regulate distinct target mRNAs in both a positive and negative fashion. The positive role of miR-122 in viral replication suggests that this miRNA could be targeted for antiviral therapy.

Takeover of host ribosomes by divergent IRES elements.

The ribosome is the macromolecular machinery in the host cell for which all viruses have to compete. Early in infection, the viral mRNAs have to compete with the host for both the ribosomes and for the limited pool of eukaryotic initiation factors that are needed to facilitate the recruitment of ribosomes to both viral and cellular mRNAs. To circumvent this competition, certain viruses have evolved to recruit ribosomes to IRESs (internal ribosome entry sites), highly specialized RNA elements that are located at the 5'-end of the viral genomes. Here, we discuss how divergent IRES elements can recruit ribosomes and start protein synthesis with only a minimal set of eukaryotic translation initiation factors, and how this mode of translation initiation aids viral gene amplification during early onset of innate immune responses.

Translation of glucose-regulated protein 78/immunoglobulin heavy-chain binding protein mRNA is increased in poliovirus-infected cells at a time when cap-dependent translation of cellular mRNAs is inhibited.

All cellular cytoplasmic mRNAs carry a 7-methylguanylate cap attached to their 5' ends. This cap structure is recognized by cap-binding proteins that then direct the binding of ribosomal subunits to this 5'-end complex. Poliovirus, a plus-stranded RNA virus, interferes with this cellular translation process by proteolytically inactivating the cap-binding protein complex. Subsequently the viral mRNA can be translated by an initiation process in which ribosomes bind internally to the mRNA [Pelletier, J. & Sonenberg, N. (1988) Nature (London) 334, 320-325], obviating cap-dependent translation. At least one cellular mRNA, encoding a heat shock-like protein, glucose-regulated protein 78/immunoglobulin heavy-chain binding protein, has been discovered to be translated at an increased rate in poliovirus-infected cells at a time when the translation of other cellular mRNAs is inhibited. The glucose-regulated protein 78/immunoglobulin heavy-chain binding protein mRNA thus exemplifies a cellular mRNA that is translated at a specifically enhanced rate by an as-yet-unresolved cap-independent initiation process in cells when the cap-binding protein complex is not functional.

Role of 3'-end sequences in infectivity of poliovirus transcripts made in vitro.

It has been shown by van der Werf et al. (S. van der Werf, J. Bradley, E. Wimmer, F. W. Studier, and J. Dunn, Proc. Natl. Acad. Sci. USA 83:2330-2334, 1986) that in vitro synthesis of poliovirus RNA by T7 RNA polymerase gives rise to infectious RNA molecules; however, these molecules are only 5% as infectious as RNA isolated from virions. A plasmid, T7D-polio, was constructed that allows the in vitro synthesis of full-length RNA molecules with two additional guanine residues at the 5' end. However, T7D-polio differed from the construct of van der Werf et al. in that RNA transcribed from T7D-polio has an authentic 3' end, ending with only a polyadenine nucleotide sequence. Transfection of these RNA molecules into mammalian cells produced wild-type poliovirus with an efficiency similar to that of virion RNA. The use of this vector in the characterization of viral mutants in vivo and in vitro is discussed.

Oxidation-reduction sensitive interaction of a cellular 50-kDa protein with an RNA hairpin in the 5' noncoding region of the poliovirus genome.

Genetic and biochemical analyses of the 5' noncoding region of poliovirus have indicated the importance of this region in both translation and amplification of the viral RNA. The role of the cellular machinery required for these events is just beginning to be revealed. Using an RNA gel retention assay, we have identified a cellular 50-kDa protein that forms a specific complex with a stable stem-loop structure present in the viral 5' noncoding region. The formation of the RNA-protein complex is dependent on the availability of free sulfhydryl groups in the protein. The possible involvement of this RNA-protein complex in the regulation of viral gene expression is discussed.

Translation-competent extracts from Saccharomyces cerevisiae: effects of L-A RNA, 5' cap, and 3' poly(A) tail on translational efficiency of mRNAs.

Yeast genetics has proven fruitful in the identification of key players that are involved in translational initiation. However, the exact roles of many translation initiation factors in translation initiation remain unknown. This has been due to lack of a suitable in vitro translation system in which the mode of action of certain translation factors can be studied. This report describes the preparation of cell-free Saccharomyces cerevisiae lysates that can mediate the translation of exogenously added mRNAs. Optimal translation required the absence of viral L-A RNA in the lysate and the presence of both a 5' cap and a 3' poly(A) tail on the mRNAs. A cooperative effect of cap and poly(A) tail on translation initiation was observed, a property that has been found to operate in intact yeast cells as well. In addition, the yeast lysates mediated translational initiation through several viral internal ribosome entry sites, demonstrating that the yeast translation apparatus can perform internal initiation. Thus, these lysates may be useful in the biochemical analysis of cap-dependent and cap-independent translation events.

Location of the internal ribosome entry site in the 5' non-coding region of the immunoglobulin heavy-chain binding protein (BiP) mRNA: evidence for specific RNA-protein interactions.

The 220 nucleotide 5'non-coding region (5'NCR) of the human immunoglobulin heavy chain binding protein (BiP) mRNA contains an internal ribosome entry site (IRES) that mediates the translation of the second cistron in a dicistronic mRNA in cultured mammalian cells. In this study, experiments are presented that locate the IRES immediately upstream of the start-site AUG codon in the BiP mRNA. Furthermore, crosslinking of thiouridine-labeled BiP IRES-containing RNA to cellular proteins identified the specific binding of two proteins, p60 and p95, to the 3'half of the BiP 5'NCR. Interestingly, both p60 and p95 bound also specifically to several viral IRES elements. This correlation suggests that p60 and p95 could have roles in internal initiation of cellular and viral IRES elements.