Identification of RNA polymerase III-transcribed genes in eukaryotic genomes.

The RNA polymerase (Pol) III transcription system is devoted to the production of short, generally abundant noncoding (nc) RNAs in all eukaryotic cells. Previously thought to be restricted to a few housekeeping genes easily detectable in genome sequences, the set of known Pol III-transcribed genes (class III genes) has been expanding in the last ten years, and the issue of their detection, annotation and actual expression has been stimulated and revived by the results of recent high-resolution genome-wide location analyses of the mammalian Pol III machinery, together with those of Pol III-centered computational studies and of ncRNA-focused transcriptomic approaches. In this article, we provide an outline of distinctive features of Pol III-transcribed genes that have allowed and currently allow for their detection in genome sequences, we critically review the currently practiced strategies for the identification of novel class III genes and transcripts, and we discuss emerging themes in Pol III transcription regulation which might orient future transcriptomic studies. This article is part of a Special Issue entitled: Transcription by Odd Pols.

Transcript levels, alternative splicing and proteolytic cleavage of TFIIIA control 5S rRNA accumulation during Arabidopsis thaliana development.

Ribosome biogenesis is critical for eukaryotic cells and requires coordinated synthesis of the protein and rRNA moieties of the ribosome, which are therefore highly regulated. 5S ribosomal RNA, an essential component of the large ribosomal subunit, is transcribed by RNA polymerase III and specifically requires transcription factor IIIA (TFIIIA). To obtain insight into the regulation of 5S rRNA transcription, we have investigated the expression of 5S rRNA and the exon-skipped (ES) and exon-including (EI) TFIIIA transcripts, two transcript isoforms that result from alternative splicing of the TFIIIA gene, and TFIIIA protein amounts with respect to requirements for 5S rRNA during development. We show that 5S rRNA quantities are regulated through distinct but complementary mechanisms operating through transcriptional and post-transcriptional control of TFIIIA transcripts as well as at the post-translational level through proteolytic cleavage of the TFIIIA protein. During the reproductive phase, high expression of the TFIIIA gene together with low proteolytic cleavage contributes to accumulation of functional, full-length TFIIIA protein, and results in 5S rRNA accumulation in the seed. In contrast, just after germination, the levels of TFIIIA-encoding transcripts are low and stable. Full-length TFIIIA protein is undetectable, and the level of 5S rRNA stored in the embryo progressively decreases. After day 4, in correlation with the reorganization of 5S rDNA chromatin to a mature state, full-length TFIIIA protein with transcriptional activity accumulates and permits de novo transcription of 5S rRNA.

Mechanisms of regulation of RNA polymerase III-dependent transcription by TORC1.

We have found earlier that Tor1 binds to 5S rDNA chromatin but the functional significance has not been established. Here, we show that association with 5S rDNA chromatin is necessary for TOR complex 1 (TORC1) to regulate the synthesis of 5S ribosomal RNA and transfer RNAs (tRNAs) by RNA polymerase (Pol) III, as well as the phosphorylation and binding to Pol III-transcribed genes of the Pol III repressor Maf1. Interestingly, TORC1 does not bind to tRNA genes, suggesting that TORC1 modulates tRNA synthesis indirectly through Maf1 phosphorylation at the rDNA loci. We also find that Maf1 cytoplasmic localization is dependent on the SSD1-v allele. In W303 cells that carry the SSD1-d allele, Maf1 is constitutively nuclear but its nucleolar localization is inhibited by TORC1, indicating that TORC1 regulates nucleoplasm-to-nucleolus transport of Maf1. Finally, we show that TORC1 interacts with Maf1 in vivo and phosphorylates Maf1 in vitro, and regulates Maf1 nucleoplasm-to-nucleolus translocation. Together, these observations provide new insights into the chromatin-dependent mechanism by which TORC1 controls transcription by Pol III.

Alcohol induces RNA polymerase III-dependent transcription through c-Jun by co-regulating TATA-binding protein (TBP) and Brf1 expression.

Chronic alcohol consumption is associated with steatohepatitis and cirrhosis, enhancing the risk for hepatocellular carcinoma. RNA polymerase (pol) III transcribes a variety of small, untranslated RNAs, including tRNAs and 5S rRNAs, which determine the biosynthetic capacity of cells. Increased RNA pol III-dependent transcription, observed in transformed cells and human tumors, is required for oncogenic transformation. Given that alcohol consumption increases risk for liver cancer, we examined whether alcohol regulates this class of genes. Ethanol induces RNA pol III-dependent transcription in both HepG2 cells and primary mouse hepatocytes in a manner that requires ethanol metabolism and the activation of JNK1. This regulatory event is mediated, at least in part, through the ability of ethanol to induce expression of the TFIIIB components, Brf1, and the TATA-binding protein (TBP). Induction of TBP, Brf1, and RNA pol III-dependent gene expression is driven by enhanced c-Jun expression. Ethanol promotes a marked increase in the direct recruitment of c-Jun to TBP, Brf1, and tRNA gene promoters. Chronic alcohol administration in mice leads to enhanced expression of TBP, Brf1, tRNA, and 5S rRNA gene transcription in the liver. These alcohol-dependent increases are more pronounced in transgenic animals that express the HCV NS5A protein that display increased incidence of liver tumors. Together, these results identify a new class of genes that are regulated by alcohol through the co-regulation of TFIIIB components and define a central role for c-Jun in this process.

Mirror-image T7 transcription of chirally inverted ribosomal and functional RNAs.

To synthesize a chirally inverted ribosome with the goal of building mirror-image biology systems requires the preparation of kilobase-long mirror-image ribosomal RNAs that make up the structural and catalytic core and about two-thirds of the molecular mass of the mirror-image ribosome. Here, we chemically synthesized a 100-kilodalton mirror-image T7 RNA polymerase, which enabled efficient and faithful transcription of the full-length mirror-image 5S, 16S, and 23S ribosomal RNAs from enzymatically assembled long mirror-image genes. We further exploited the versatile mirror-image T7 transcription system for practical applications such as biostable mirror-image riboswitch sensor, long-term storage of unprotected kilobase-long l-RNA in water, and l-ribozyme-catalyzed l-RNA polymerization to serve as a model system for basic RNA research.

Ribosomal protein L11 associates with c-Myc at 5 S rRNA and tRNA genes and regulates their expression.

The c-Myc oncoprotein promotes cell growth by enhancing ribosomal biogenesis. Overexpression of c-Myc and aberrant ribosomal biogenesis lead to deregulated cell growth and tumorigenesis. Hence, c-Myc activity and ribosomal biogenesis must be tightly coordinated during normal homeostasis. We previously found that ribosomal protein L11 inhibits c-Myc activity by blocking the recruitment of its co-activator transformation/transcription domain-associated protein (TRRAP) to the promoter regions of c-Myc target genes that are transcribed by RNA polymerases I and II. In this study, we extended the role of L11 to the regulation of c-Myc-driven transcription of the 5 S rRNA and tRNA genes by RNA polymerase III. L11 co-resided with c-Myc at the 5 S rRNA and tRNA genes and significantly inhibited the binding of TRRAP to these genes. Knocking down endogenous L11 enhanced c-Myc-dependent transcription of these genes. Interestingly, in response to ribosomal stress induced by the treatment of cells with a low dose of actinomycin D or serum starvation, L11 binding to these genes was increased, and inversely TRRAP binding to these genes was decreased. Consistently, knockdown of L11 rescued the reduction of the expression of these genes by the two treatments. These results demonstrate that L11 suppresses c-Myc-dependent and RNA polymerase III-catalyzed transcription of 5 S rRNA and tRNA genes in response to ribosomal stress, ensuring a tight coordination between c-Myc activity and ribosomal biogenesis.

DNAzyme-mediated recovery of small recombinant RNAs from a 5S rRNA-derived chimera expressed in Escherichia coli.

BACKGROUND: Manufacturing large quantities of recombinant RNAs by overexpression in a bacterial host is hampered by their instability in intracellular environment. To overcome this problem, an RNA of interest can be fused into a stable bacterial RNA for the resulting chimeric construct to accumulate in the cytoplasm to a sufficiently high level. Being supplemented with cost-effective procedures for isolation of the chimera from cells and recovery of the recombinant RNA from stabilizing scaffold, this strategy might become a viable alternative to the existing methods of chemical or enzymatic RNA synthesis. RESULTS: Sequence encoding a 71-nucleotide recombinant RNA was inserted into a plasmid-borne deletion mutant of the Vibrio proteolyticus 5S rRNA gene in place of helix III - loop C segment of the original 5S rRNA. After transformation into Escherichia coli, the chimeric RNA (3×pen aRNA) was expressed constitutively from E. coli rrnB P1 and P2 promoters. The RNA chimera accumulated to levels that exceeded those of the host's 5S rRNA. A novel method relying on liquid-solid partitioning of cellular constituents was developed for isolation of total RNA from bacterial cells. This protocol avoids toxic chemicals, and is therefore more suitable for large scale RNA purification than traditional methods. A pair of biotinylated 8-17 DNAzymes was used to bring about the quantitative excision of the 71-nt recombinant RNA from the chimera. The recombinant RNA was isolated by sequence-specific capture on beads with immobilized complementary deoxyoligonucleotide, while DNAzymes were recovered by biotin affinity chromatography for reuse. CONCLUSIONS: The feasibility of a fermentation-based approach for manufacturing large quantities of small RNAs in vivo using a "5S rRNA scaffold" strategy is demonstrated. The approach provides a route towards an economical method for the large-scale production of small RNAs including shRNAs, siRNAs and aptamers for use in clinical and biomedical research.

ncl-1 is required for the regulation of cell size and ribosomal RNA synthesis in Caenorhabditis elegans.

Regulation of ribosome synthesis is an essential aspect of growth control. Thus far, little is known about the factors that control and coordinate these processes. We show here that the Caenorhabditis elegans gene ncl-1 encodes a zinc finger protein and may be a repressor of RNA polymerase I and III transcription and an inhibitor of cell growth. Loss of function mutations in ncl-1, previously shown to result in enlarged nucleoli, result in increased rates of rRNA and 5S RNA transcription and enlarged cells. Furthermore, ncl-1 adult worms are larger, have more protein, and have twice as much rRNA as wild-type worms. Localization studies show that the level of NCL-1 protein is independently regulated in different cells of the embryo. In wild-type embryos, cells with the largest nucleoli have the lowest level of NCL-1 protein. Based on these results we propose that ncl-1 is a repressor of ribosome synthesis and cell growth.

Comparative analysis of five varieties in Perilla frutescens (L.) Britton by 45S rDNA FISH and 5S rDNA sequences.

In this study, the relationships of five varieties in Perilla frutescens (L.) Britton were analyzed. All the varieties showed the same chromosome number 2n = 40, and the banding patterns of PI-DAPI staining and the FISH results of 45S rDNA probe were on the same site of the chromosome pair. The 5S rRNA gene spacers were ranged from 413 bp in P. frutescens var. purpurascens and 408 bp in var. frutescens. The varieties analyzed here were clustered into two clades according to the phylogenetic analysis. The sequence identity of 96% between var. auriculato-dentata and var. arguta suggested they could be considered the same variety.

TFIIIB Subunit Bdp1 Participates in RNA Polymerase III Transcription in the Protozoan Parasite Leishmania major.

Leishmania major, a protozoan parasite that diverged early from the main eukaryotic lineage, exhibits unusual mechanisms of gene expression. Little is known in this organism about the transcription factors involved in the synthesis of tRNA, 5S rRNA, and snRNAs, transcribed by RNA Polymerase III (Pol III). Here we identify and characterize the TFIIIB subunit Bdp1 in L. major (LmBdp1). Bdp1 plays key roles in Pol III transcription initiation in other organisms, as it participates in Pol III recruitment and promoter opening. In silico analysis showed that LmBdp1 contains the typical extended SANT domain as well as other Bdp1 conserved regions. Nevertheless, LmBdp1 also displays distinctive features, including the presence of only one aromatic residue in the N-linker region. We were not able to produce null mutants of LmBdp1 by homologous recombination, as the obtained double replacement cell line contained an extra copy of LmBdp1, indicating that LmBdp1 is essential for the viability of L. major promastigotes. Notably, the mutant cell line showed reduced levels of the LmBdp1 protein, and its growth was significantly decreased in relation to wild-type cells. Nuclear run-on assays demonstrated that Pol III transcription was affected in the mutant cell line, and ChIP experiments showed that LmBdp1 binds to 5S rRNA, tRNA, and snRNA genes. Thus, our results indicate that LmBdp1 is an essential protein required for Pol III transcription in L. major.

A telomerase subunit homolog La protein from Trypanosoma brucei plays an essential role in ribosomal biogenesis.

The autoantigen La protein is an important component of telomerase and a predominantly nuclear phosphoprotein. As a telomerase subunit, La protein associates with the telomerase ribonucleoprotein and influences telomere length. In the reverse transcription, La protein stimulates enzymatic activity and increases repeated addition processivity of telomerase. As nuclear phosphoprotein, La protein binds the 3' poly(U)-rich elements of nascent RNA polymerase III transcripts to facilitate its correct folding and maturation. In this work, we identified a La protein homolog (TbLa) from Trypanosoma brucei (T. brucei). We revealed that TbLa interacts with ribosome-associated protein P34/P37, 40S ribosomal protein SA, and 60S ribosomal subunit L5 in T. brucei. In the interactions between TbLa protein and (P34/P37)/L5/SA, RNA recognition motif (RRM) domain of TbLa was indicated to make the major contribution to the processes. We determined the solution structure of TbLa RRM domain. NMR chemical shift perturbations revealed that the positively charged RNA-binding pocket of TbLa RRM domain is responsible for its interaction with ribosomal and ribosome-associated proteins P37/L5/SA. Furthermore, depletion of TbLa affected the maturation process of 5S rRNA and ribosomal assembly, suggesting TbLa protein might play a significant role in the ribosomal biogenesis pathway in T. brucei. Taken together, our results provide a novel insight and structural basis for better understanding the roles of TbLa and RRM domain in ribosomal biogenesis in T. brucei. DATABASE: Structural data are available in the PDB under the accession number 5ZUH.

A test of the model that RNA polymerase III transcription is regulated by selective induction of the 110 kDa subunit of TFIIIC.

TFIIIC is a RNA polymerase (pol) III-specific DNA-binding factor that is required for transcription of tRNA and 5S rRNA genes. Active human TFIIIC consists of five subunits. However, an inactive form has also been isolated that lacks one of the five subunits, called TFIIIC110. A model was proposed in which pol III transcription might be regulated by the specific induction of TFIIIC110, allowing formation of active TFIIIC from the inactive form. We have tested this model by transient transfection of HeLa and HEK293 cells with a vector expressing TFIIIC110. We have also made stably transfected HeLa cell lines that carry a doxycycline-inducible version of the cDNA for TFIIIC110. We show that the induced TFIIIC110 enters the nucleus, binds other TFIIIC subunits and is recruited to tRNA and 5S rRNA genes in vivo. However, little or no effect is seen on the expression of pol III transcripts. The data argue against the model that pol III transcription can be effectively modulated through the specific induction of TFIIIC110.

Structural requirements of 5S rRNA for nuclear transport, 7S ribonucleoprotein particle assembly, and 60S ribosomal subunit assembly in Xenopus oocytes.

Structural requirements of 5S rRNA for nuclear transport and RNA-protein interactions have been studied by analyzing the behavior of oocyte-type 5S rRNA and of 31 different in vitro-generated mutant transcripts after microinjection into the cytoplasm of Xenopus oocytes. Experiments reveal that the sequence and secondary and/or tertiary structure requirements of 5S rRNA for nuclear transport, storage in the cytoplasm as 7S ribonucleoprotein particles, and assembly into 60S ribosomal subunits are complex and nonidentical. Elements of loops A, C, and E, helices II and V, and bulged and hinge nucleotides in the central domain of 5S rRNA carry the essential information for these functional activities. Assembly of microinjected 5S rRNA into 60S ribosomal subunits was shown to occur in the nucleus; thus, the first requirement for subunit assembly is nuclear targeting. The inhibitory effects of ATP depletion, wheat germ agglutinin, and chilling on the nuclear import of 5S rRNA indicate that it crosses the nuclear envelope through the nuclear pore complex by a pathway similar to that used by karyophilic proteins.

Yeast ribosomal protein L1 is required for the stability of newly synthesized 5S rRNA and the assembly of 60S ribosomal subunits.

Ribosomal protein L1 from Saccharomyces cerevisiae binds 5S rRNA and can be released from intact 60S ribosomal subunits as an L1-5S ribonucleoprotein (RNP) particle. To understand the nature of the interaction between L1 and 5S rRNA and to assess the role of L1 in ribosome assembly and function, we cloned the RPL1 gene encoding L1. We have shown that RPL1 is an essential single-copy gene. A conditional null mutant in which the only copy of RPL1 is under control of the repressible GAL1 promoter was constructed. Depletion of L1 causes instability of newly synthesized 5S rRNA in vivo. Cells depleted of L1 no longer assemble 60S ribosomal subunits, indicating that L1 is required for assembly of stable 60S ribosomal subunits but not 40S ribosomal subunits. An L1-5S RNP particle not associated with ribosomal particles was detected by coimmunoprecipitation of L1 and 5S rRNA. This pool of L1-5S RNP remained stable even upon cessation of 60S ribosomal subunit assembly by depletion of another ribosomal protein, L16. Preliminary results suggest that transcription of RPL1 is not autogenously regulated by L1.

Role of maturation-promoting factor (p34cdc2-cyclin B) in differential expression of the Xenopus oocyte and somatic-type 5S RNA genes.

Transcription of 5S rRNA and tRNA genes by RNA polymerase III (pol III) in cytosolic extracts of unfertilized Xenopus eggs and in a reconstituted system derived from Xenopus oocytes is repressed by the action of one or more mitotic protein kinases. Repression is due to the phosphorylation of a component of the pol III transcription apparatus. We find that the maturation/mitosis-promoting factor kinase (MPF, p34cdc2-cyclin B) can directly mediate this repression in vitro. Affinity-purified MPF and immune complexes formed with antibodies to the protein subunits of MPF (p34cdc2 and cyclin B) retain both histone H1 kinase activity and the capacity to repress transcription in the reconstituted transcription system. Transcription complexes of oocyte-type 5S RNA genes and tRNA genes are quantitatively more sensitive to MPF repression than the corresponding transcription complexes of the somatic-type 5S RNA gene. The differential transcription of oocyte- and somatic-type genes observed during early Xenopus embryogenesis has been reproduced with the reconstituted transcription system and affinity-purified MPF. This differential transcription may be due to the instability of transcription complexes on the oocyte-type genes and the heightened sensitivity of soluble transcription factors to inactivation by mitotic phosphorylation. Our results suggest that MPF may play a role in vivo in the establishment of the embryonic pattern of pol III gene expression.

A conserved sequence in tRNA and rRNA promoters of Lactococcus lactis.

A tRNA operon (trnA) from Lactococcus lactis consisting of seven tRNA genes and a 5S rRNA gene was cloned and sequenced. Promoter-fusion of the trnA promoter to a promoter-less beta-galactosidase gene of Leuconostoc mesenteroides resulted in high levels of beta-galactosidase activity in L. lactis. Searching for sequences with similarity to the sequence of the promoter region revealed a consensus sequence of promoters preceeding rRNA operons and tRNA operons from Lactococcus species including a not previously described conserved sequence (AGTT).

Structural requirements for the processing of Escherichia coli 5 S ribosomal RNA by RNase E in vitro.

Processing of 9 S precursor RNA in Escherichia coli requires the endoribonuclease RNase E, which makes two cleavages to liberate p5, the immature form of 5 S rRNA. The contributions of primary and secondary structure to RNase E-mediated cleavage of 9 S RNA were investigated. The structure of the 5' domain of 9 S RNA was probed by partial ribonuclease digestion and chemical modification. Our structural analysis of 9 S RNA supports a model in which the 5' spacer domain folds into tandem hairpins so that the first processing cleavage site 5' to the 5 S moiety resides in a stretch of single-stranded residues. Site-directed mutagenesis of a cloned 9 S RNA sequence was performed and synthetic transcripts derived from a variety of such mutant templates were assayed as substrates for RNase E-dependent endonuclease activity in fractionated extracts. Partial or complete deletion of the 5 S sequence did not eliminate site-specific processing of 9 S RNA. Mutations affecting the 5' domain revealed that secondary structure upstream from the first cleavage site is important in maintaining efficient processing. However, secondary structure downstream from either cleavage site is dispensable. Our results suggest that RNase E specifically recognizes and cleaves single-stranded RNA sequences only when presented in a proper conformational context. Adjacent secondary structures appear to play a direct and critical role in the enzyme's recognition of its substrate. Additionally, it may serve to anchor single-stranded regions to ensure the availability of the RNase E cleavage sites.

Relative contributions of the zinc fingers of transcription factor IIIA to the energetics of DNA binding.

We have expressed and purified a series of recombinant zinc finger polypeptides derived from the cDNA for the Xenopus 5 S gene-specific transcription factor TFIIIA. Dissociation constants for the interaction of each of the truncated polypeptides with the 5 S gene promoter have been measured using gel mobility shift assays. DNase I footprinting and proteolysis experiments provide additional insights into the interactions of individual fingers within complexes of the truncated proteins. These results are discussed in terms of recently proposed models for the TFIIIA-DNA interaction. The effects of mutations in two of the strongly binding proteins, zf1-3 and zf1-7, on DNA binding affinity have been investigated. Mutations have been made both in putative DNA-contact residues and in the linker regions between zinc fingers. The observed decreases in binding affinity cannot be explained simply in terms of loss of protein-DNA contacts. Our results support a model in which DNA binding is accomplished through sets of interacting zinc fingers that make different energetic contributions to the overall binding of the protein and different contacts with the DNA.

An immuno-electron microscopical analysis of transcribing multinucleosomal templates: what happens to the histones?

Immuno-electron microscopy was used to visualize the structure of reconstituted chromatin after in vitro transcription by purified T7 RNA polymerase. T7 RNA polymerase disrupts the nucleosomal structure in the transcribed region. This disruption is not influenced by the template, linear or supercoiled, and the presence or absence of nucleosomal positioning sequences in the transcribed region. In this study, we used monoclonal autoantibodies reacting with the nucleosome core particles and epitopes within several regions of the four different core histones. Some of the residues recognized by the autoantibodies are accessible on the surface of the nucleosomes and some are more internal and therefore less exposed at the surface. We show that the loss of the nucleosomal configuration during transcription is due to the loss of histone/DNA binding and that at least part of the histones are transferred to the nascent RNA chains. Consequently, after in vitro transcription by T7 RNA polymerase, the nucleosomal template does not conserve its original configuration, and no interaction of antigen/antibodies is observed anymore in the region that has been transcribed. Therefore, we conclude that in our in vitro transcription assay, nucleosomes are detached from the template, and not simply unfolded with histones remaining attached to the DNA.

The 5'-flanking sequence of the loach oocyte 5S rRNA gene contains a signal for effective transcription.

Deletion mutants of loach oocyte 5S rRNA genes were injected and transcribed in vivo in the nuclei of loach (Misgurnus fossilis) and Xenopus laevis. A control region was found in the 5'-flanking sequence, the elimination of which greatly decreases in vivo transcription of 5S rRNA genes. This cis-acting element is located in the region between nt-18 and the transcription start point. We propose that the oocyte nucleus contains (a) specific transcriptional factor(s), NTFO, which interacts with the cis-acting element we described. We also propose that NTFO is inactivated in maturing oocytes when nucleoplasm interacts with oocyte cytoplasm after germinal vesicle breakdown. The residual activity of this factor(s) may be responsible for low-level synthesis of oocyte 5S rRNA at the beginning of embryogenesis. We consider the disappearance of NTFO during gastrulation to be responsible for the total inactivation of oocyte 5S rRNA genes in embryonic and somatic tissue.