RNA polymerase III (RNA pol III) transcribes structural RNAs involved in RNA processing (U6 snRNA) and translation (tRNA), thereby regulating the growth rate of cells. Proper initiation by RNA pol III requires the transcription factor TFIIIB. Gene-external U6 snRNA transcription requires TFIIIB consisting of Bdp1, TBP, and Brf2. Transcription from the gene-internal tRNA promoter requires TFIIIB composed of Bdp1, TBP, and Brf1. TFIIIB is a target of tumor suppressors, including PTEN, ARF, p53, and RB, and RB-related pocket proteins. Breast cancer susceptibility gene 1 (BRCA1) tumor suppressor plays a role in DNA repair, cell cycle regulation, apoptosis, genome integrity, and ubiquitination. BRCA1 has a conserved amino-terminal RING domain, an activation domain 1 (AD1), and an acidic carboxyl-terminal domain (BRCA1 C-terminal region). In Saccharomyces cerevisiae, TFIIB interacts with the BRCA1 C-terminal region domain of Fcp1p, an RNA polymerase II phosphatase. The TFIIIB subunits Brf1 and Brf2 are structurally similar to TFIIB. Hence, we hypothesize that RNA pol III may be regulated by BRCA1 via the TFIIB family members Brf1 and Brf2. Here we report that:(1) BRCA1 inhibits both VAI (tRNA) and U6 snRNA RNA pol III transcription;(2) the AD1 of BRCA1 is responsible for inhibition of U6 snRNA transcription, whereas the RING domain and AD1 of BRCA1 are required for VAI transcription inhibition; and (3) overexpression of Brf1 and Brf2 alleviates inhibition of U6 snRNA and VAI transcription by BRCA1. Taken together, these data suggest that BRCA1 is a general repressor of RNA pol III transcription.

The higher-order assembly of chromatin imposes structural organization on the genetic information of eukaryotes and is thought to be largely determined by posttranslational modification of histone tails. Here, we study a 20-kilobase silent domain at the mating-type region of fission yeast as a model for heterochromatin formation. We find that, although histone H3 methylated at lysine 9 (H3 Lys9) directly recruits heterochromatin protein Swi6/HP1, the critical determinant for H3 Lys9 methylation to spread in cis and to be inherited through mitosis and meiosis is Swi6 itself. We demonstrate that a centromere-homologous repeat (cenH) present at the silent mating-type region is sufficient for heterochromatin formation at an ectopic site, and that its repressive capacity is mediated by components of the RNA interference (RNAi) machinery. Moreover, cenH and the RNAi machinery cooperate to nucleate heterochromatin assembly at the endogenous mat locus but are dispensable for its subsequent inheritance. This work defines sequential requirements for the initiation and propagation of regional heterochromatic domains.

Indirect evidence suggests that the 5' end of U1 snRNA recognizes the 5' splice site in mRNA precursors by complementary base pairing. To test this hypothesis, we asked whether point mutations in the alternative 12S and 13S 5' splice sites of the adenovirus E1A gene can be suppressed by compensatory base changes in human U1 snRNA. When the mutant E1A and U1 genes are contransfected into HeLa cells, we observe efficient suppression of one mutation at position +5 in the 12S splice site, but exceedingly weak suppression of another mutation at position +3 in the 13S splice site. These and other results suggest that base pairing between U1 and the 5' splice site is necessary but not sufficient for the splicing of mRNA precursors.

Significant advances have been made in elucidating the biogenesis pathway and three-dimensional structure of the UsnRNPs, the building blocks of the spliceosome. U2 and U4/U6*U5 tri-snRNPs functionally associate with the pre-mRNA at an earlier stage of spliceosome assembly than previously thought, and additional evidence supporting UsnRNA-mediated catalysis of pre-mRNA splicing has been presented.

The U2 snRNP binds to the site of branch formation during splicing of mammalian pre-mRNA in vitro. In Saccharomyces cerevisiae the branch site is within the so-called TACTAAC box (UACUAAC box), an absolutely conserved intron sequence required for splicing. Based on the identification and sequence of a U2 analogue in yeast, a specific base pairing interaction between the UACUAAC box and a highly conserved region of this snRNA can be proposed. To test this hypothesis, we have taken advantage of two mutations constructed previously in the UACUAAC box of an actin-HIS4 fusion. These mutant strains were transformed with stable plasmids bearing U2-like snRNAs into which changes predicted to restore base pairing had been introduced. Allele-specific suppression of biological and biochemical phenotypes was observed in both cases. Recognition of the UACUAAC box thus relies, at least in part, on Watson-Crick base pairing with the yeast U2 analogue.

Multiple processing events are required to convert a single eukaryotic pre-ribosomal RNA (pre-rRNA) into mature 18S (small subunit), 5.8S and 25-28S (large subunit) rRNAs. We have asked whether U3 small nucleolar RNA is required for pre-rRNA processing in vivo by depleting Saccharomyces cerevisiae of U3 by conditional repression of U3 synthesis. The resulting pattern of accumulation and depletion of specific pre-rRNAs indicates that U3 is required for multiple events leading to the maturation of 18S rRNA. These include an initial cleavage within the 5' external transcribed spacer, resembling the U3 dependent initial processing event of mammalian pre-rRNA. Formation of large subunit rRNAs is unaffected by U3 depletion. The similarity between the effects of U3 depletion and depletion of U14 small nucleolar RNA and the nucleolar protein fibrillarin (NOP1) suggests that these could be components of a single highly conserved processing complex.

RNA export from the nucleus has been analyzed in Xenopus oocytes. U1 snRNAs made by RNA polymerase II were exported into the cytoplasm, while U1 snRNAs synthesized by RNA polymerase III, and therefore with a different cap structure, remained in the nucleus. Export of the polymerase II-transcribed RNAs was inhibited by the cap analog m7GpppG. Spliced mRNAs carrying monomethylguanosine cap structures were rapidly exported, while hypermethylated cap structures delayed mRNA export. The export of a mutant precursor mRNA unable to form detectable splicing complexes was also significantly delayed by incorporation of a hypermethylated cap structure. The results suggest that the m7GpppN cap structure is likely to be a signal for RNA export from the nucleus.

The yeast snoRNP protein, NOP1, is structurally and functionally homologous to vertebrate fibrillarin and is essential for viability. A conditionally lethal allele was constructed by placing NOP1 expression under the control of a GAL promoter. Growth on glucose medium results in the depletion of NOP1 over several generations, during which cell growth is progressively impaired. Pulse labelling of proteins shows that NOP1 depleted strains are greatly impaired in the production of cytoplasmic ribosomes, and they have a reduced level of rRNA. Northern hybridization and pulse-chase labelling of pre-rRNA show a progressive impairment of all pre-rRNA processing steps. The pathway leading to 18S rRNA is particularly affected. Methylation of pre-rRNA is concomitantly impaired and unmethylated pre-rRNA accumulates and is not processed over long periods. NOP1 depletion does not prevent the accumulation of seven snoRNAs tested including U3; the levels of two species, U14 and snR190, decline. The snoRNAs synthesized in the absence of NOP1 retain TMG cap structures. Subnuclear fractionation and immunocytochemistry indicate that they continue to be localized in the nucleolus.

The interaction between the U1 snRNP-specific U1 A protein and U1 snRNA has been analysed. The binding site for the protein on the RNA is shown to be in hairpin II, which extends from positions 48 to 91 in the RNA. Within this hairpin the evolutionarily conserved loop sequence is crucial for interaction with U1 A protein. U1 A protein can also bind the loop sequence when it is part of an artificial RNA which cannot form a stable hairpin structure. The region of the protein required to bind to U1 snRNA consists of a conserved 80 amino acid motif, previously identified in many ribonucleoprotein (RNP) proteins, together with (maximally) 11 N-terminal and 10 C-terminal flanking amino acids. Point mutations introduced into two of the most highly conserved regions of this motif abolish RNA binding. U1 snRNA mutants from which the U1 A binding site has been deleted are shown to be capable of assembly into RNP particles which are immunoprecipitable by patient antisera which recognize U1 A protein. The role of RNA-protein and protein-protein interactions in U snRNP assembly are discussed.

The small nuclear RNA U6 and its gene have been isolated from yeast. In striking contrast to other yeast spliceosomal RNAs, U6 is very similar in size, sequence and structure to its mammalian homologue. The single-copy gene is essential. These properties suggest a central role in pre-mRNA processing. An extensive base-pairing interaction with U4 snRNA is described; the destabilization of the U4/U6 complex seen during splicing thus requires a large conformational change.

Eukaryotic cells contain a large number of small nucleolar RNAs (snoRNAs). A major family of snoRNAs features a consensus ACA motif positioned 3 nucleotides from the 3' end of the RNA. In this study we have characterized nine novel human ACA snoRNAs (U64-U72). Structural probing of U64 RNA followed by systematic computer modeling of all known box ACA snoRNAs revealed that this class of snoRNAs is defined by a phylogenetically conserved secondary structure. The ACA snoRNAs fold into two hairpin structures connected by a single-stranded hinge region and followed by a short 3' tail. The hinge region carries an evolutionarily conserved sequence motif, called box H (consensus, AnAnnA). The H box, probably in concert with the flanking helix structures and the ACA box characterized previously, plays an essential role in the accumulation of human U64 intronic snoRNA. The correct processing of a yeast ACA snoRNA, snR36, in mammalian cells demonstrated that the cis- and trans-acting elements required for processing and accumulation of ACA snoRNAs are evolutionarily conserved. The notion that ACA snoRNAs share a common secondary structure and conserved box elements that likely function as binding sites for common proteins (e.g., GAR1) suggests that these RNAs possess closely related nucleolar functions.