Active Center Control of Termination by RNA Polymerase III and tRNA Gene Transcription Levels In Vivo

Abstract

The ability of RNA polymerase (RNAP) III to efficiently recycle from termination to reinitiation is critical for abundant tRNA production during cellular proliferation, development and cancer. Yet understanding of the unique termination mechanisms used by RNAP III is incomplete, as is its link to high transcription output. We used two tRNA-mediated suppression systems to screen for Rpc1 mutants with gain- and loss- of termination phenotypes in S. pombe. 122 point mutation mutants were mapped to a recently solved 3.9 Å structure of yeast RNAP III elongation complex (EC); they cluster in the active center bridge helix and trigger loop, as well as the pore and funnel, the latter of which indicate involvement of the RNA cleavage domain of the C11 subunit in termination. Purified RNAP III from a readthrough (RT) mutant exhibits increased elongation rate. The data strongly support a kinetic coupling model in which elongation rate is inversely related to termination efficiency. The mutants exhibit good correlations of terminator RT in vitro and in vivo, and surprisingly, amounts of transcription in vivo. Because assessing in vivo transcription can be confounded by various parameters, we used a tRNA reporter with a processing defect and a strong terminator. By ruling out differences in RNA decay rates, the data indicate that mutants with the RT phenotype synthesize more RNA than wild type cells, and than can be accounted for by their increased elongation rate. Finally, increased activity by the mutants appears unrelated to the RNAP III repressor, Maf1. The results show that the mobile elements of the RNAP III active center, including C11, are key determinants of termination, and that some of the mutations activate RNAP III for overall transcription. Similar mutations in spontaneous cancer suggest this as an unforeseen mechanism of RNAP III activation in disease.

Fig 1. Distribution of point mutations in S. pombe RNAP III C1 termination mutants.

A) Unselected randomly chosen bacterial colonies. B) Linear schematic representation of S. pombe Rpc1 (C1), based on sequence homology to S. cerevisiae C160 [17]. C) Mutations in LOF termination (readthrough) mutants selected in yKR1. D) Mutations in GOF termination mutants selected in yAS70.

Fig 2. C1 mutants cluster map to highly conserved RNAP motifs.

A-D) Sequence alignments of homologous regions of RNAP II and III largest subunits of three species, Homo sapiens (hs), Schizosaccharomyces pombe (sp) and Saccharomyces cerevisiae (sc); the bridge helix (BH, A), trigger loop (TL, B), funnel (C) and cleft-pore entrance loop (CPEL, D); black and grey asterisks indicate positions of strong and weaker phenotypes of LOF mutants respectively; # indicates GOF mutations. Upward triangles were placed above residues which when mutated in S. cerevisiae RNAP II caused increased elongation rate, and downward triangles were placed above residues which when mutated in S. cerevisiae RNAP II caused decreased elongation rate as summarized in figure 3 of Kaplan et al [42].

Fig 3. Mapping S. pombe C1 active center LOF mutants onto S. cerevisiae RNAP III structure.

A) Active center bridge helix (BH, cyan) and trigger loop (TL, green) as well as the cleft-pore entrance loop (CPEL, brown) [17]; numbers indicate S. cerevisiae positions (with S. pombe positions in parentheses). B) The holoenzyme EC structure with mutations in the funnel and CPEL highlighted as red spheres. α20 and α21 refer to RNAP II motifs; CPEL: cleft-pore entrance loop (see text); other subunits and structural features are indicated. The PDB ID used is 5FJ8 [17]. C) RNAP III EC structure (PDB 5FJ8) into which the S. cerevisiae TFIIS CTD and linker was placed (shown in sticks and ribbon backbone mode). D) A high resolution view of the acidic hairpin CTD of TFIIS with its linker placed into the RNAP III cryo-EM structure shown in C. The inset shows a sequence alignment of the acidic hairpin regions of the CTDs of C11 and TFIIS from S. cerevisiae (sc) and S. pombe (sp), as indicated. The acidic DE residues at the tip of the hairpin are colored cyan to match their stick representation in the structure placement model, which come within 4.9 Å of the RNA 3' end (magenta). Another close contact of 3.8 Å between a C1 residue found mutated in mutant D854N and an invariant Q in TFIIS (the second Q in the sequence alignment) is also indicated. The CPEL, funnel helix loop and α21 are also indicated.

Fig 4. Suppression phenotype strengths of C1 LOF readthrough mutants.

A) schematic of tRNA reporter allele used for LOF screen and to subsequently examine individual C1 mutants in panels B and C which differ in the number of Ts at the test terminator. B-D) Phenotypes of LOF mutants after transformation of recovered plasmids into yKR1 (T5 terminator, B), yJI1 (6T terminator, C) and yJI1 in the presence of thiamin which represses the plasmid promoter, D). Brightest = strongest, darker = weaker. Top line: Rep4X and Rpc1-Wt are negative controls; C2-T455I is a positive control (see text).

Fig 5. Promoter-dependent in vitro transcription reveals oligo(T) length-dependent terminator readthrough by RNAP III C1 mutants.

A) Schematic of tRNA gene arrangement used for in vitro transcription by S100 extracts; the plasmids used for transcription differed only in the number of Ts at the test terminator, T1, as listed above the lanes of B & C. B-C) S100 extracts from yKR1-C1 mutants or -C1-Wt control were used as a source of initiation factors and RNAP III as indicated above the lanes. Plasmids containing tRNA genes that differ only in the number of Ts in the oligo(T) terminator as indicated above the lanes were used as templates to program the transcription reactions. RT = readthrough; termination at the failsafe terminator, T = termination at the test terminator, IC = internal control. D) Quantitation of % readthrough as defined on the Y-axis; reactions contained 32P-αGTP.

Fig 6. Representative RNAP III C1-E850K mutant exhibits increased elongation.

A) Schematic of 3'-tailed templates used for promoter-independent transcription by FLAG-purified RNAPs III (see S2 Fig for establishment of assay). The template design on the left was used for panel B and the design on the right was used for panel C. FL = full length 300 nt transcript resulting from termination at 12T, RO = run-off transcript. B) Time course of elongation by purified RNAPs III C1-Wt and C1-E850K as indicated above the lanes. C) RNAP III was stalled at position +22 followed by a synchronized chase time course of transcription elongation (see text). (The blemish between lanes 5 & 6 is not a transcription signal.)

Fig 7. Validation of terminator readthrough in vivo.

A-D) Northern blots of total RNA probed for the RNA species indicated to the right of the panels, produced by C1 mutant alleles in strain yAS76 (A & D) whose tRNA reporter gene has a 5T terminator (see Fig 5A) or in strain yYH1 (B) which has a 7T terminator; RT = readthrough transcript. Lanes 1 and 2 represent control strains yAS68 and yAS99 (see text). Error bars in (C) reflect RT/U5 from duplicate experiments plotted as % of yAS68 (right axis, see text).

Fig 8. RNAP III C1 mutants recognize and read through normal terminators.

A) Schematic of the tRNA reporter allele used for RT detection in strain yAS76, illustrating the transcripts produced from the 5T terminator and the 8T failsafe terminator. B-D) Northern blots. B) sup-tRNA intron-specific probe ("tRNA-Mser-int" see DNA oligos, Table 4); the RT transcript product resulting from termination at the failsafe 8T terminator, and pre-tRNA products resulting from termination at the 5T terminator, are indicated to the left. The intron-containing pre-tRNA processing species are indicated to the right. C) A readthrough-specific probe (equimolar mixture of three tRNA-LysCUU-RT oligo-DNAs, Methods) complementary to the region downstream of the natural terminator of tRNA-LysCUU gene, detects readthrough termination of endogenous tRNA gene (see text). D) A tRNALysCUU intron-specific probe (tRNA-LysCUU-int, Methods) of the same blot as in B (see text).

Fig 9. RNAP III C1 mutants increase RNA production from a tRNA gene.

A) Schematic of the tRNA reporter gene in the strain yAS68 and the RT transcript produced from the 8T terminator. B) tRNA reporter gene readthrough (RT) transcript-specific probing of RNA from C1-mutants and controls as indicated above the lanes. Transcription of this tRNA reporter gene makes only RT transcripts and not the normal pre-tRNAs (upper panel, i, see text). The second panel (ii) shows probing of the same blot as in upper panel, for U5 snRNA. The next lower panel (iii) shows probing for the 3'-trailer of a nascent pre-tRNASerGCU whose endogenous single copy gene bears an 8T terminator (black arrow); the small grey arrow points to an intron-containing, 3'-trailer-containing intermediate. The lowest panel (iv) shows probing for the unique 3' trailer of a single copy nascent pre-tRNAValUAC whose gene does not contain an intron and ends with a T8 terminator (Black arrow); the small grey arrow points to the more abundant mature tRNAValUAC product of two genes that overlap with this probe. C) Quantification of data as in (A) from duplicate experiments calibrated by U5 RNA on the same blots; note that C1-F1069L, lane 8, is from a GOF mutant. D) RT transcript (Y-axis) from 5T terminator RT data plotted against amount of tRNA reporter gene output (X-axis) as in panel A; each represented by duplicate data sets reflected by error bars. The Y-axis represents a proxy for relative elongation rate (see text). E) Cells transformed with the pRep4X vector, C1-mutant or C1-Wt alleles carrying a Flag tag were subjected to immunoblotting for their Rpc1 protein.

Fig 10. Synthesis rather than posttranscriptional RNA stability is increased by C1-mutants and unlinked to Maf1.

C1-Wt and the multiple isolate, C1-E850K cells whose tRNA reporter gene (in yAS68 as in Fig 9B) makes only readthrough transcripts were treated with the transcriptional inhibitor, 1,10-phenanthroline (110-P), and RNA was purified from the cultures at the times indicated above the lanes. A) Ethidium bromide-stained gel that was blotted for sequential probing below as described in the text. B & C) Readthrough (cRT)-specific probe (tRNA-Mser-cRT; mix, Methods), long and short exposures; reveals nearly identical half-life of RT RNA in C1-Wt and C1-E850K, of about 20 minutes. D) Mature U5 snRNA probe. E) tRNALysCUU intron-specific probe reveals half-life of nascent pre-tRNALysCUU of about 5 minutes in C1-Wt and C1-E850K; note that C1-Wt and C1-E850K produce nearly equal amounts of this pre-tRNALysCUU (see also lanes 5 and 13 of Fig 7D, see text). F) A maf1-deletion strain, yKR101 carrying the same reporter tRNA gene as in Fig 9A and the C1-mutant or -Wt alleles indicated above the lanes were examined for the RT transcript (upper panel) and the U5 snRNA (lower).