HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 278, Issue 12, 10450-10457, March 21, 2003

This Article Abstract Full Text (PDF) All Versions of this Article: 278/12/10450    most recent M213310200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jourdain, S. Articles by Lefebvre, O. Search for Related Content PubMed PubMed Citation Articles by Jourdain, S. Articles by Lefebvre, O. Social Bookmarking                      What's this?

The 95 Subunit of Yeast TFIIIC Influences Upstream and Downstream Functions of TFIIIC·DNA Complexes^*

Sabine Jourdain, Joël Acker, Cécile Ducrot, André Sentenac, and Olivier Lefebvre^§

From the Service de Biochimie et de Génétique Moléculaire, CEA/Saclay, F-91191 Gif-sur-Yvette Cedex, France

Received for publication, December 9, 2002, and in revised form, January 16, 2003

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The yeast transcription factor IIIC (TFIIIC) is organized in two distinct multisubunit domains, A and B, that are respectively responsible for TFIIIB assembly and stable anchoring of TFIIIC on the B block of tRNA genes. Surprisingly, we found that the removal of A by mild proteolysis stabilizes the residual B·DNA complexes at high temperatures. Focusing on the well conserved 95 subunit that belongs to the A domain, we found that the 95-E447K mutation has long distance effects on the stability of TFIIIC·DNA complexes and start site selection. Mutant TFIIIC·DNA complexes presented a shift in their 5' border, generated slow-migrating TFIIIB·DNA complexes upon stripping TFIIIC by heparin or heat treatment, and allowed initiation at downstream sites. In addition, mutant TFIIIC·DNA complexes were highly unstable at high temperatures. Coimmunoprecipitation experiments indicated that 95 participates in the interconnection of A with B via its contacts with 138 and 91 polypeptides. The results suggest that 95 serves as a scaffold critical for A·DNA spatial configuration and B·DNA stability.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

TFIIIC¹ is a multisubunit DNA binding factor that serves as a dynamic platform to assemble pre-initiation complexes on class III genes. Once bound to tDNA intragenic promoter elements, TFIIIC directs the assembly of TFIIIB on the DNA, which in turns recruits the RNA polymerase III (pol III) and activates multiple rounds of transcription (1).

Biochemical and genetic studies have yielded a detailed characterization of Saccharomyces cerevisiae TFIIIC factor. Yeast TFIIIC consists of six subunits (named 138, 131, 95, 91, 60, and 55) distributed between two globular domains named A and B based on their ability to bind the internal tDNA promoter elements, the A and B blocks, respectively (2, 3). B·block B binding is predominant over the low affinity A·block A interaction (4, 5). The link between the A and B domains displays a remarkable flexibility as it can stretch across a range of distances separating the A and B blocks in natural tRNA genes, regardless of their helical phasing (6). The A domain is essential for transcription activation and start site selection as it mediates TFIIIB assembly, in cooperation with the TATA-binding protein (TBP), strongly directing its DNA binding position upstream of the transcribed region (6, 7).

The B subassembly is composed of 138, 91, and 60 (8-10). Protein-DNA cross-linking analyses of TFIIIC·DNA complex have localized the 138 subunit over the B block and 91 over the transcriptional terminator further downstream (11, 12). Both subunits cooperate in DNA binding. A point mutation in the TFC3 gene encoding 138 was found to severely alter TFIIIC·DNA complex stability and to be suppressed by an amino acid substitution in 91 (9, 13). The B domain also contains 60, which does not seem to be involved in DNA binding. Instead, in vivo and in vitro analysis of a mutant form of TFIIIC provided evidence in favor of a functional interaction of 60 with TBP. This result suggested that 60 could participate in TFIIIB recruitment, despite its downstream localization, and possibly link the A and B domains (10).

The A domain contains the three other TFIIIC subunits, 131, 95, and 55 (11, 14-17). Genetic and biochemical evidence point to the central role of 131 in TFIIIB assembly on the DNA. 131 interacts with Brf1 (18, 19) and with Bdp1, previously known as B" or TFIIIB90 (20), and extends far upstream within the TFIIIB binding region (11). Several lines of evidence indicate that recruitment of TFIIIB is a complex process that involves important conformational changes of 131 (as well as TFIIIB components). First, the efficiency of 131 photocross-linking to DNA was found to vary during the TFIIIB assembly process (21). Second, two-hybrid experiments with internal deletion mutants also suggested that 131 flips between different states that expose or mask TFIIIB-binding sites (18). Third, several dominant mutations in the N-terminal part of 131 increased TFIIIB recruitment (22, 23). Finally, circular dichroism spectra analysis directly revealed a conformation change following 131·Brf1 interaction that may concern both proteins (24). In contrast, little is known about the function of 95 and 55. Both polypeptides can be found in a subcomplex potentially containing other proteins (17). In the TFIIIC·DNA complex, 95 and 55 are both located over the A block (11). The 95-kDa subunit is very conserved through evolution and is thought to be responsible for the A·block A interaction (8, 11, 25, 26).

In this work, we pursued the characterization of the A domain to get further insight into TFIIIC function. The results indicate that 95 holds a key position in TFIIIC exerting both upstream and downstream influence on the TFIIIC·DNA complex via its interactions with A and B components.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Constructions and TFC1 Random Mutagenesis-- The 2.8-kb XhoI/NotI fragment from plasmid YcpCS7 (CEN, URA, wt HA-TFC1) (27), containing the promoter region and a modified open reading frame of TFC1 allowing addition of an HA tag sequence after the initiation codon, was cloned into the polylinker region of plasmid pRS314 (28) creating pYSJ1 (CEN, TRP, wt HA-TFC1).

Two oligonucleotides SJt95-5 (5'-AGTTGTTTGTTCATGATCTT) and SJt95-7 (5'-ATATTTATGTCCCCATCAGC) were used to amplify an 811-bp DNA fragment from bp 806 to 1616 of the TFC1 open reading frame under mild mutagenic PCR conditions (3 mM MgCl₂). Gapped pYSJ1 containing extremities homologous to both ends of the PCR products was produced by an NruI-NcoI digestion followed by agarose purification and was cotransformed with the PCR products into YCS7 haploid strain (ade2-101 ura3-52 lys2-801 his3-200 trp1-1 tfc1-::HIS3 YcpCS7) (27). Transformants containing recombinant gap-repaired plasmids were plated on 5-fluoro-orotic acid media at 24 °C to chase the wild type copy of TFC1 gene harbored on YcpCS7. Thermosensitive mutants were isolated by comparing growth at 30, 34, and 37 °C. One mutant showing slow growth at 34 °C and no growth at 37 °C was selected and analyzed by sequencing. The mutated tfc1 gene harbored two point mutations, one neutral (G888T) and another (G1339A) causing the substitution of Glu to Lys at position 447. The strain containing this mutated tfc1 gene was named YSJ1-E447K. The control strain, YSJ1, contains wild type HA-tagged TFC1 harbored on plasmid pYSJ1.

TFIIIC Purification-- TFIIIC was purified starting from ~15 g of YSJ1 or mutant YSJ1-E447K cells following a two-step chromatographic procedure (heparin HyperD followed by MonoQ purification) as described previously (10). Both Western blot analysis with anti-55 and anti-HA antibodies and gel retardation experiments showed that mutant TFIIIC fractions contained three times less factor than wild type fractions.

DNA Binding and in Vitro Transcription Assays-- TFIIIC·DNA interactions were monitored by gel shift assays as described previously (13). A ³²P-labeled DNA fragment (3-10 fmol; 4,000-10,000 cpm) carrying the SUP4 tRNA^Tyr (345 bp) gene was incubated for 10 min at 25 °C in a reaction mixture containing 10 mM Tris-HCl (pH 8.0), 1 mM EDTA, 10% glycerol, 2 mg/ml bovine serum albumin (Sigma), DNA competitor, TFIIIC (MonoQ fraction) at a final KCl concentration of 120 mM. To analyze B·tDNA interaction, TFIIIC·tDNA complexes were subjected to limited proteolysis for 10 min at 25 °C with 40 ng of -chymotrypsin (Sigma), and digestion was stopped by addition of 40 ng of aprotinin (Sigma).

The apparent dissociation constants (K_app) of wild type and mutant TFIIIC-DNA complexes were determined, using the above binding conditions as described (4, 29). K_app was derived from Equation 1,

(Eq. 1)

where [TFIIIC₀], [DNA], and [TFIIIC·DNA] indicate, respectively, the concentration of active TFIIIC protein and the free and complexed SUP4 tRNA^Tyr gene. Equation 1 describes a hyperbola with [TFIIIC₀] as the asymptotic value. K_app was determined using the Scatchard representation shown in Equation 2,

(Eq. 2)

The slope of the plot gives the apparent dissociation constant K_app, and the y intercept yields TFIIIC concentration [TFIIIC₀] in the assay.

To assemble TFIIIB·TFIIIC·DNA (B·C·DNA) complexes, Ultrogel-heparin TFIIIB fraction (27) was added to the TFIIIC·DNA binding reaction described above together with additional DNA competitor and further incubated for 30 min at 25 °C. The final KCl concentration was adjusted to 120 mM. To form TFIIIB·DNA complexes, TFIIIC was stripped by addition of 500 ng of heparin (5 min at 25 °C) or by raising the incubation temperature for 5 min at 50 °C. To form pol III·TFIIIB·DNA complexes, highly purified RNA polymerase III (100 ng) was added to a heat-treated TFIIIB·DNA complex and further incubated for 45 min at 25 °C. Protein·DNA complexes were analyzed by non-denaturing electrophoresis on 5% polyacrylamide gel.

Standard in vitro transcription assays were performed as described previously (10, 27) using Ultrogel-heparin-purified TFIIIB, MonoQ-purified TFIIIC, and highly purified RNA polymerase III (100 ng). The final KCl concentration was 100 mM. Plasmid templates harbored either the SUP4 tRNA^Tyr gene or a TATA-containing yeast tRNA^Asp(GTC) gene (30). Transcription reactions were allowed to proceed for 45 min at 25 °C, and transcripts were analyzed by electrophoresis on 6% polyacrylamide gel under denaturing conditions (8 M urea).

Primer Extension Analysis-- Transcripts were produced by in vitro specific transcription as described above using 0.6 mM each ATP, CTP, GTP, and UTP and 150 ng of plasmid template carrying SUP4 tRNA^Tyr gene, precipitated with ammonium acetate and 10 µg of carrier Escherichia coli tRNA (Sigma), and resuspended in 16 µl of diethyl pyrocarbonate, water. 8 µl were then mixed with 8 units of RNasin (Amersham Biosciences), 0.5 pmol of an internal 5' end-labeled oligonucleotide primer (5'-GTGATAAATTAAAGTCTTGCGCCTTAAACC, complementary to the coding region of SUP4 tRNA^Tyr gene between positions +29 and +59), and NaCl to a final concentration of 0.3 M. After DNA denaturation for 5 min at 85 °C, primer annealing was allowed to proceed for 90 min at 40 °C. Reaction mixtures were then adjusted to a final volume of 80 µl and contained 50 mM Tris-HCl (pH 8), 50 mM KCl, 10 mM MgCl₂, 5 mM spermidine, 10 mM dithiothreitol, 1 mM of each dNTP (Invitrogen), 40 units of RNasin (Amersham Biosciences). Addition of 10 units of avian myeloblastosis virus-reverse transcriptase (Promega) allowed primer extension to proceed for 90 min at 42 °C. Resulting cDNAs were precipitated with sodium acetate/ethanol. Pellets were resuspended in 10 µl of denaturing loading buffer (90% (v/v) formamide, 10 mM Tris-HCl (pH 8), 1 mM EDTA, bromphenol blue, and xylene cyanol), heated for 5 min at 90 °C, and loaded onto a 7% (w/v) polyacrylamide-urea sequencing gel in parallel with ³²P-labeled DNA kb ladder (Invitrogen).

Exonuclease Digestion-- TFIIIC·DNA complex was assembled on a NotI-XhoI DNA fragment (307 bp) from pRS316-SUP4 tRNA^Tyr, labeled by filling in the XhoI site (3' end of the non-coding strand) with 0.2 mM of each dATP, dGTP, dTTP, [-³²P]dCTP (20 µCi), and the Klenow fragment of E. coli DNA polymerase (Amersham Biosciences, 5 units) with appropriate buffer. The TFIIIC·DNA binding reaction (15 µl) was adjusted to 50 µl to reach the following conditions: 40 mM Tris-HCl (pH 8), 0.3 mM EDTA, 0.6 mg/ml of bovine serum albumin, 10% glycerol, and 20 mM MgCl₂. Digestion was started with addition of 1 unit of  exonuclease (Amersham Biosciences), followed by a 1-min incubation at 25 °C, and stopped by heating samples at 80 °C for 10 min. DNA fragments were precipitated with sodium acetate/ethanol and 20 µg of glycogen as a carrier. Pellets were resuspended in 10 µl of denaturing loading buffer and heated for 5 min at 90 °C, and samples were loaded on a 7% (w/v) polyacrylamide-urea sequencing gel. To map the 5' border site, a G + A sequence reaction was performed on the same labeled DNA fragment following Maxam and Gilbert procedure (31).

Sequence Homology Searches-- S. cerevisiae 95 protein sequence (GenBank^TM accession number A39711) was searched against several data bases using the BLAST (32) server at NCBI (similar proteins were exhaustively searched against the same data bases until all similar proteins were identified). Protein sequences were aligned using Clustal X (33) based on the BLOSUM 62 scoring matrix (34). Complete 95-like proteins were identified in Homo sapiens (GenBank^TM accession number NP_036219); Drosophila melanogaster (GenBank^TM accession AAF53767); Caenorhabditis elegans (GenBank^TM accession CAA84675); Arabidopsis thaliana (GenBank^TM accession AAG52180); Schizosaccharomyces pombe (GenBank^TM accession CAB11095); Neurospora crassa (GenBank^TM accession CAC28811); and Candida albicans (unfinished sequence obtained at the Stanford Genome Technology Center).

Immunoprecipitation Experiments-- Cell extracts were prepared as described (35). Ascitic fluid (0.5 µl) containing mouse monoclonal 12CA5 anti-HA antibody was incubated for 30 min at 10 °C with 20 µl of magnetic beads coated with human monoclonal antibodies directed against mouse immunoglobulin G and Dynabeads Pan Mouse IgG from Dynal (8 × 10⁸ beads/ml, in phosphate-buffered saline containing 0.1% bovine serum albumin). After extensive washing first in phosphate-buffered saline containing 0.1% bovine serum albumin and then in phosphate-buffered saline, 20% glycerol, 0.05% Nonidet P-40, the beads were incubated for 180 min with gentle shaking at 10 °C with cell crude extracts (50 µl, 120 µg of proteins) from cells expressing or not expressing HA-tagged 95, FLAG-tagged 138, or 91. The beads were washed twice with 200 µl of phosphate-buffered saline, 20% glycerol, 0.05% Nonidet P-40. Immunopurified proteins were eluted by heating for 3 min at 95 °C in SDS loading buffer and then analyzed by SDS-PAGE on a 8% polyacrylamide gel and revealed by Western blotting with 12CA5 anti-HA, M2 anti-FLAG (Sigma), or anti--91 antibodies, using an Amersham enhanced chemiluminescence kit.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

95 Subunit of TFIIIC Is Highly Conserved-- The strong conservation of 95 sequence from yeast to human is illustrated in Fig. 1. In addition to human TFIIIC63 (25), six additional protein sequences found in data bases presented significant homology with S. cerevisiae 95. As shown in Fig. 1, six regions of sequence similarity including an acidic tail are colinearly conserved in each of the eight protein sequences. The role of the acidic tail is unknown, but it has been shown recently (36) that this domain is not required for yeast cells viability. The central region D (bordered by amino-acids 317-392 in 95 sequence) was the most remarkably conserved. When comparing the yeast and human pol III transcription system, it appears that the most conserved components are those engaged in protein-protein interactions critical for transcription complex assembly, as is the case for 131, for example (37, 38). Thus, the conservation of 95 suggested some critical functions in addition to its postulated role in the A-block binding of TFIIIC. We therefore set out to mutagenize a DNA fragment encompassing the conserved regions D and E of 95. As described under "Experimental Procedures," PCR mutagenesis generated one thermosensitive point mutant showing a decreased growth rate at 34 °C and lethality at 37 °C. The mutation E447K (indicated by the star in Fig. 1) is mapped just upstream of the E region of sequence similarity. As seen by in vivo RNA labeling with tritiated uracil, the mutation caused a marked decrease in tRNA synthesis after 3 h of incubation of mutant cells at 37 °C (data not shown). Among all the genes of the pol III transcription system, high dosage of TFC1, BRF1, and, to a lesser extent, of TFC5 (encoding Bdp1) was able to improve the growth rate of mutant cells at 34 °C, but only TFC1 suppressed the lethality at 37 °C (data not shown).

View larger version (64K): [in this window] [in a new window]   Fig. 1.   95 subunit of TFIIIC and potential orthologs. Complete 95-like proteins were identified in H. sapiens (Hs), D. melanogaster (Dm), C. elegans (Ce), A. thaliana (At), S. pombe (Sp), N. crassa (Nc), C. albicans (Ca), and S. cerevisiae (Sc). Conserved domains are boxed and labeled by letters A-D (the most conserved) and E (the least conserved) on the upper panel. A star indicates the E447K mutation. Sequence alignments of each conserved domain (A-E) are shown on the lower panel. Amino acids conserved in at least two sequences are boxed.

95-E447K Influences Start Site Selection-- In view of the genetic interaction noted above between 95 and TFIIIB components, we examined the ability of mutant TFIIIC to recruit TFIIIB. Mutant and wild type TFIIIC were partially purified by two chromatographic steps (a heparin Hyper D column and a MonoQ column as described in Ref. 10). Based on Western blotting assays with anti-HA or anti-55 antibodies and gel shift analysis, TFIIIC preparation from mutant cells was estimated to contain one-third as much factor as wild type fractions. In the following experiments, care was taken to use similar amounts of mutant and wild type factor. TFIIIB recruitment was monitored by gel shift assay (39). TFIIIC·DNA complexes were first formed by preincubating TFIIIC with a SUP4 DNA probe for 10 min at 25 °C and then further incubated with TFIIIB to form B·C·DNA complexes that can be separated by gel electrophoresis. As shown in Fig. 2A, TFIIIB supershifted all the C·DNA complexes formed with wild type TFIIIC (lane 2). The supershifted band was probably composed of a mixture of B·C·DNA and B'·C·DNA complexes, the latter containing TBP and Brf1 but lacking Bdp1 (40, 41). Mutant TFIIIC assembled TFIIIB less efficiently (compare lanes 2 and 6) as evidenced by the large proportion of unshifted C·DNA complexes. TFIIIC can be selectively stripped from B·C·DNA complexes by heparin treatment to yield the heparin-resistant B·DNA complexes (42). Interestingly, although B·C·DNA complexes formed with wild type TFIIIC generated, as expected, B·DNA complexes of characteristic fast migration, the complexes formed with mutant TFIIIC generated, after stripping TFIIIC with heparin, the characteristic B·DNA complex plus a more slowly migrating form of complex (Fig. 2A, compare lanes 4 and 8). A similar result was obtained upon heating the B·C·DNA complexes for 5 min at 50 °C (lanes 3 and 7). The slow-migrating complex (see arrow in Fig. 2A) was presumed to be a distinct form of B·DNA complex because B'·DNA complexes are heparin-sensitive (43) and should be expected to migrate faster than B·DNA complexes. A less likely possibility could be that the slow-migrating complex had retained a fraction of TFIIIC to yield a bigger complex (that should be resistant to heparin and heat treatment). This possibility cannot be excluded even though we found that anti-55 antibodies did not supershift the two forms of complexes (data not shown).

View larger version (54K): [in this window] [in a new window]   Fig. 2.   95-E447K affects TFIIIB·DNA complex formation. Protein·DNA complex formation with wild type (WT) or mutant (E447K) TFIIIC (C), TFIIIB (B), and RNA polymerase III (Pol) was analyzed by gel retardation assays. The protein components incubated with the SUP4 tDNATyr probe are indicated above the lanes. A, lanes 4 and 8, B·C·DNA complexes were first assembled, and then heparin (500 ng) was added in order to strip TFIIIC (CBHEP). A and B, lanes 3 and 7, B·C·DNA complexes were heated for 5 min at 50 °C before loading on the gel (CB50). B, lanes 4 and 8, RNA polymerase III was added after heating the B·C·DNA complex at 50 °C, and the mixture was further incubated for 45 min at 25 °C before gel analysis. The positions of TFIIIB·DNA and RNA polymerase III·TFIIIB·DNA complexes are shown. The slow migrating complexes obtained with mutant TFIIIC are indicated by arrows.

Next, we investigated whether both forms of stripped complexes were able to recruit RNA polymerase III using gel shift assays (Fig. 2B). Control B·DNA complexes formed by heat treatment at 50 °C (lane 3) were able to recruit pol III (at 25 °C) to form a supershifted complex (lane 4) migrating approximately like B·C·DNA complexes (lane 2). Similarly, the two forms of complexes obtained by stripping mutant TFIIIC by heat treatment were both supershifted upon incubation with pol III suggesting that both complexes were able to recruit the RNA polymerase, although the slow-migrating complex appeared somewhat less efficient (lanes 7 and 8).

The above results prompted us to compare wild type and mutant TFIIIC for their ability to direct specific transcription in vitro. TFIIIC, TFIIIB, and purified pol III fractions were used to transcribe the SUP4 tDNA^Tyr or the tRNA^Asp(GTC) gene that harbors a TATA element at 30 but requires TFIIIC to be transcribed efficiently (30). As shown in Fig. 3, equal amounts of wild type or mutant TFIIIC factor yielded similar amounts of transcripts. However, mutant TFIIIC-directed transcription of both genes generated new RNA species (indicated by arrows in the autoradiograms and in the scans shown in Fig. 3) shorter than the full-size pre-tRNA. The same additional RNA species were observed when TFIIIC was stripped from the DNA before transcription (data not shown).

View larger version (37K): [in this window] [in a new window]   Fig. 3.   Mutant TFIIIC affects the transcript pattern. The SUP4 tRNATyr or the tRNAAsp(GTC) gene were transcribed as described under "Experimental Procedures" in the presence of wild type (WT) or mutant (E447K) TFIIIC. Left panel, electrophoretic analysis of the transcripts; right panel, PhosphorImager (Amersham Biosciences) quantification of radiolabeled transcripts. The continuous and dotted line profiles correspond to wild type and mutant TFIIIC, respectively. The new RNA species generated in the presence of mutant TFIIIC are shown by arrows.

A primer extension analysis of in vitro synthesized SUP4 tRNA is shown in Fig. 4. Whereas wild type TFIIIC directed essentially RNA chain initiation at position +1, chain initiation directed by mutant TFIIIC was more polydisperse, with a major start site at position +1 and additional start sites at positions +4 and +8 and to a lesser extent at +13 and +15. As shown a decade ago by Geiduschek and colleagues (42), TFIIIB is the main factor responsible for initiation by pol III. Accurate initiation depends on the proper placement of TFIIIB on the DNA, placement that depends on TFIIIC and TBP even on genes deprived of a canonical TATA box like SUP4 tDNA (7). Therefore, the present results suggested that mutant TFIIIC had misplaced a proportion of TFIIIB·DNA complexes resulting in shorter transcripts.

View larger version (49K): [in this window] [in a new window]   Fig. 4.   Mutant TFIIIC affects start site selection. The SUP4 tRNATyr gene was transcribed in the presence of wild type (WT) or mutant (E447K) TFIIIC. Start site selection was determined by primer extension of the RNA transcripts and analyzed on a sequencing gel as described under "Experimental Procedures." Sites of transcription initiation are indicated on the right.

Upstream Border of TFIIIC·DNA Complexes-- Site-specific DNA-protein photocross-linking and protein-protein interaction studies implicated 131 as the TFIIIB-assembling subunit of TFIIIC; 131 projects far upstream within the TFIIIB binding region to contact Brf1 and Bdp1 (18, 19, 35, 40, 44). To investigate whether 95-E447K affected the placement of 131, we analyzed the upstream border of TFIIIC·DNA complexes by  exonuclease digestion. Wild type or mutant TFIIIC were preincubated with SUP4 DNA probe labeled at the 3' end of the non-coding strand, and TFIIIC·DNA complexes were digested by  exonuclease for 1 min at 25 °C. Resulting DNA fragments were analyzed on a sequencing gel alongside G + A sequencing reaction products from the same probe (Fig. 5). When compared with the pattern obtained in the absence of factor, wild type TFIIIC arrested  exonuclease digestion at two positions, 53 and 47, as evidenced by the accumulation of two DNA fragments of similar labeling intensity (lanes 3 and 4). In contrast, mutant TFIIIC·DNA complexes showed only the 47 upstream border (lanes 5 and 6). As the TFIIIC factors used in these experiments were partially purified, we tested several highly purified preparations of TFIIIC to ascertain that these observed exonuclease arrests were due to TFIIIC and not to spurious contaminants. Immunopurified or DNA affinity-purified TFIIIC caused predominantly the 53 digestion arrest and the 47 arrest to a variable extent (lanes 1 and 2, and results not shown). The loss of the 53 blockage site in mutant TFIIIC complexes therefore indicated that the 95 mutation influenced 131 projection on upstream DNA.

View larger version (81K): [in this window] [in a new window]   Fig. 5.   95-E447K influences the 5' border of TFIIIC·DNA complexes. The SUP4 tDNATyr probe labeled at the 3' end of the non-coding strand was complexed with TFIIIC and digested with  exonuclease, and the digestion products were analyzed on a sequencing gel as described under "Experimental Procedures." Nuclease-resistant DNA fragments in reactions without factor (lanes 3 and 6), with wild type TFIIIC (lane 4), or mutant TFIIIC (lane 5) were separated on a sequencing gel, alongside G + A cleavage products of the same probe (G+A lane). DNA cleavage products obtained with immunopurified TFIIIC (T1) or with DNA affinity highly purified TFIIIC (T2) were analyzed, respectively, on lanes 1 and 2. Positions of the transcription start site (+1) and of the A block are indicated on the left. The arrows show the direction and extent of digestion for the main nuclease-resistant DNA fragments in the presence of bound TFIIIC.

A Affects TFIIIC·DNA Binding-- As the above observations could not account for the strong thermosensitivity and the drop of in vivo tRNA synthesis occurring in the YSJ1-E447K strain, we looked for other major defects in mutant TFIIIC. 95 is thought to be involved in block A binding (8, 11), but the overall DNA affinity of TFIIIC is predominantly governed by the interaction of B domain with the B block (5). Nevertheless, we investigated whether the 95 mutation could affect the apparent dissociation constant of TFIIIC·DNA complexes.

The apparent dissociation constant (K_app) of TFIIIC·DNA complexes was determined by titrating a constant amount (as determined by Western blot) of wild type or mutant TFIIIC with increasing concentrations of a SUP4 DNA probe at permissive temperature (25 °C), under optimal binding conditions (4). The concentration of TFIIIC·DNA complex [TFIIIC·DNA] and free DNA probe [DNA] was determined using the gel shift assay (Fig. 6). When the SUP4 tRNA^Tyr binding data were plotted for mutant TFIIIC and wild type TFIIIC (Fig. 6, A and B), good fits to the respective theoretical curves were obtained (see "Experimental Procedures"). The linearity in the Scatchard representation (Fig. 6B) suggested the presence of a single binding component in the wild type and mutant TFIIIC fractions. The apparent dissociation constant and concentration of active TFIIIC in the binding assay [TFIIIC₀] were determined by fitting the data to Equation 2 (see "Experimental Procedures") (4, 29) by linear regression. The slope of the curve (Fig. 6B) yielded a value for K_app of 0.33 × 10¹⁰ M for the wild type and 1.33 × 10¹⁰ M for mutant TFIIIC·SUP4 tDNA^Tyr complexes. The concentration of active TFIIIC in the binding assay given by the y intercept of the curve was 0.75 × 10¹⁰ M for wild type TFIIIC and 0.73 × 10¹⁰ M for mutant TFIIIC. Therefore, the weaker binding of mutant TFIIIC (at permissive temperature) was due to its 4-fold lower DNA binding affinity. Remarkably, the tsv115 mutation in 138, a B subunit, caused a similar (5-fold) reduction in DNA binding affinity at the permissive temperature of 25 °C (13). Note, however, that due to the weak binding signal, the DNA affinity of the mutant factor was estimated with lower accuracy than for the wt factor.

View larger version (22K): [in this window] [in a new window]   Fig. 6.   tDNA affinity of mutant TFIIIC. A, TFIIIC·DNA complexes formed with varying amounts of labeled SUP4 tDNATyr probe and equal amount of wild type (WT) or mutant (E447K) TFIIIC were analyzed by the gel retardation assay. The concentrations of TFIIIC complexes [TFIIIC·DNA] and free probe [DNA] formed in each reaction were determined, and the data were plotted according to Equation 1 (see "Experimental Procedures"). B, determination of TFIIIC·tDNA apparent dissociation constant by Scatchard representation. The curve slope gives the apparent dissociation constant (Kapp) of TFIIIC·tDNA complexes formed with wild type (WT) or mutant (E447K) factor. Mutant (E447K) and wild type factors are plotted as black and white circles, respectively.

Next, we investigated the stability of mutant TFIIIC·DNA complexes. TFIIIC fractions were preincubated at varying temperatures, for 10 min, before complex formation with the SUP4 DNA probe for 10 min at 25 °C. TFIIIC·DNA complexes were then analyzed by gel shift assays (Fig. 7A). Wild type TFIIIC was resistant to heat treatments at 35-40 °C and retained most of its DNA binding activity (Fig. 7A, upper panel, lanes 3 and 4, see also lane 13). In contrast, mutant TFIIIC lost much of its DNA binding activity upon incubation at 35 °C and was totally inactivated at 40 °C (Fig. 7A, lower panel, lanes 3 and 4). In other experiments (Fig. 7B), factor·DNA complexes were preformed at 25 °C then subjected to heat treatments at different temperatures before gel electrophoresis. Preformed TFIIIC·DNA complexes were more resistant to heat denaturation, but again the stability of mutant TFIIIC·DNA complexes decreased drastically at 40 °C (Fig. 7B, lower panel, lane 10), a temperature that did not affect control TFIIIC·DNA complexes that were resistant to heat treatment up to 45 °C (Fig. 7B, upper panel, lane 11).

View larger version (109K): [in this window] [in a new window]   Fig. 7.   95-E447K affects TFIIIC·DNA binding stability. A, wild type (WT) or mutant (E447K) TFIIIC was preincubated for 10 min at different temperatures as indicated. After addition of the labeled SUP4 tDNATyr probe, the mixture was further incubated for 10 min at 25 °C and then subjected to gel electrophoresis. B, temperature sensitivity of TFIIIC·DNA complexes. TFIIIC fractions were preincubated for 10 min at 25 °C in standard binding buffer with the SUP4 tDNATyr probe and competitor DNA and then incubated for 10 min at different temperatures, as indicated, before the mixtures were analyzed by gel retardation assay. C, B, the protease-resistant domain of TFIIIC, was obtained by digestion of the wild type or mutant (E447K) TFIIIC·SUP4 tDNATyr complexes with 40 ng of -chymotrypsin for 10 min at 25 °C. Digestion was stopped by addition of 40 ng of aprotinin, and B·DNA complexes were further incubated for 10 min at different temperatures, as indicated, before gel analysis. Lane 14, undigested TFIIIC·DNA complex. Lane C, control.

It was unexpected to observe such a thermolability for TFIIIC·DNA complexes mutated in a A component because B·DNA interaction is known to be dominant in TFIIIC. The same thermolability was observed in gel shift assays using a short (55 bp) DNA probe harboring only the B block sequence (data not shown). Therefore the binding of the mutant A domain on the DNA appears not to be a prerequisite for the destabilization of the B·DNA complexes. We then explored the temperature sensitivity of B·DNA complexes produced by limited proteolysis of wild type or mutant TFIIIC·DNA complexes (3). TFIIIC fractions were preincubated with the SUP4 DNA probe for 10 min at 25 °C before the -chymotrypsin treatment. The resultant protein·DNA complexes were then subjected to 10 min of heat treatment as indicated in Fig. 7C. It was remarkable that B·DNA complexes derived from wild type or mutant factor exhibited the same thermostability and were more resistant to heat denaturation at 50 °C than the native TFIIIC·DNA complexes (compare Fig. 7C, lane 19, upper and lower panels with Fig. 7B, lane 11, upper and lower panels). These results suggested that in the context of native TFIIIC, the A domain indirectly affected B·DNA interaction and that the 95 mutation exacerbated this negative influence.

95 Interacts Directly with 91 and 138-- To understand how 95, a A subunit, could influence the distal B domain, we used a coimmunoprecipitation assay to investigate possible interactions between 95 and the individual components of B, 138, 91, and 60. Epitope-tagged 95 and 138 (HA-95 and FLAG-138) were used to facilitate the identification and the purification of the protein complexes on anti-HA-coated magnetic beads. High five insect cells were coinfected with different combinations of recombinant baculovirus to express HA-tagged 95 alone or in combination with B components. SDS-PAGE analysis of crude extracts revealed that all recombinant proteins display the expected size (Fig. 8A, lanes 2, 3, and 5). The level of expression was variable and generally better when only one recombinant protein was expressed (compare Fig. 8A, lanes 2, 3, and 5 with Fig. 8A, lanes 4, 6, and 7). B components selectively retained on the beads were analyzed by SDS-PAGE and immunoblotting, using 12CA5 anti-HA, M2 anti-FLAG, or anti-91 antibodies (Fig. 8B). The anti-HA antibody revealed a single band of about 100 kDa in crude extracts from cells expressing HA-95 (Fig. 8B, lanes 2, 4, 6, and 7). This protein was absent from control extracts (Fig. 8B, IP, lanes 1, 3, and 5), was properly immunopurified on the beads (Fig. 8B, IP, lanes 2, 4, 6, and 7), and therefore corresponded to HA-95. Interestingly, FLAG-tagged 138 or 91 were found to be detectably retained by the anti-HA-coated beads only when coexpressed with HA-95 (Fig. 8B, see Co-IP of 138 and 91 for lanes 4 and 6). The coexpression of the three proteins (138, 95, and 91) markedly decreased the expression of 91 but yielded qualitatively the same coimmunopurification results (Fig. 8B, see Co-IP of 138 and 91 for lane 7). These results suggested that 95 engages multiple contacts with B domain. Due to a high background binding of recombinant 60 on magnetic beads coated with the 12CA5 antibody, it was not possible to study the interaction between 95 and 60 (data not shown).

View larger version (34K): [in this window] [in a new window]   Fig. 8.   95 interacts with 138 and 91. High five insect cells were coinfected with various combinations of recombinant baculovirus, as indicated. A, recombinant proteins. Coomassie Blue staining of an 8% polyacrylamide gel. Positions and molecular masses (in kilodaltons) of marker bands are indicated on the left. 15 µg of crude extracts from cells expressing HA-tagged 95 (lanes 2, 4, 6, and 7), FLAG-tagged 138 (lanes 3, 4, and 7), or 91 (lanes 5-7) were loaded in each lane. The positions of FLAG-tagged 138 and HA-tagged 95 and 91 are indicated by arrows on the right. B, crude extracts (120 µg of proteins) from cells expressing HA-tagged 95 (lanes 2, 4, 6, and 7), FLAG-tagged 138 (lanes 3, 4, and 7), or 91 (lanes 5-7) were mixed with magnetic beads coated with 12CA5 anti-HA antibodies. The beads were washed, and bound proteins were eluted by boiling in loading buffer and analyzed by SDS-PAGE. The input (only 15 µg of the 120 µg used for the immunoprecipitation) and the entire bound protein fraction were analyzed by Western blotting using 12CA5 anti-HA (95), M2 anti-FLAG (138), or anti-91 antibodies. For clarity, the polypeptides revealed by specific antibodies are identified on the left. IP, immunoprecipitation; Co-IP, coimmunoprecipitation. The beads coated with anti-HA antibodies retained specifically HA-95 (IP, lanes 2, 4, 6, and 7) and FLAG-138 or 91 copurified with HA-95 (Co-IP, lanes 4, 6, and 7), whereas background retention of FLAG-138 and 91 by the beads was undetectable (Co-IP, lanes 3 and 5).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this work, we show that the 95 subunit of TFIIIC that lies over the A block of tRNA genes influences start site selection and has a strong indirect effect on TFIIIC·DNA binding stability. The observations suggest that 95 influences the positioning of the TFIIIB assembling subunit (131) and participates in the interconnection of the A block and B block binding domains of TFIIIC (see Fig. 9).

View larger version (15K): [in this window] [in a new window]   Fig. 9.   Schematic representation of the upstream and downstream effects of 95-E447K mutation. The preinitiation complex B·C·DNA is represented, with the A and B blocks, the B domain, the TFIIIB-assembling subunit 131, 95-E447K, and TFIIIB. The 95 mutation is proposed to affect 131 placement resulting in a different placement of TFIIIB and consequently in downstream initiation sites (dotted lines). The TFIIIC·DNA destabilization effect of A deletion or the 95 mutation also suggests a downstream effect on B·block B interaction which is the major determinant of TFIIIC·DNA affinity.

95 is thought to participate in A block binding based on its DNA cross-linking and positioning over the A block (8, 11). The A block contributes to TFIIIC DNA binding affinity that is, however, predominantly dependent on B block binding (4), and most importantly, its location dictates the transcription start sites (6, 45) in cooperation with a preferential TBP-binding site (7). It was interesting, therefore, to find that the mutant factor assembled two distinct forms of B·DNA complexes, as seen by gel electrophoresis after removing TFIIIC by heparin or heat treatment. Both complexes were able to recruit pol III and were likely to be transcriptionally competent. The contention that both complexes had properly assembled TFIIIB was supported by their heparin resistance property which is typical of fully assembled TFIIIB·DNA complexes (43). As TFIIIB has been shown to bend DNA sharply (46-48), we presumed that the different migration rate of both complexes in gels was due to a different placement of TFIIIB on the DNA probe. Distinctly placed TFIIIB implied distinct initiation sites that were indeed evidenced by transcript analysis: mutant TFIIIC-directed transcription produced additional RNA species shorter than the full size pre-tRNA, with two different templates, and primer extension analysis mapped multiple start sites on SUP4, essentially at positions +1, +4, and +8 but also further downstream, at +13 and +15. The +4 and +8 sites were only weakly observed in the control transcription reaction with wild type TFIIIC. These alternative start sites have already been mentioned, and their relative utilization was shown to depend on distinct TFIIIB placements upon modification of the upstream promoter sequence (7).

How could a 95 point mutation influence TFIIIB placement? There is no evidence that 95 interacts with TFIIIB components but its human homolog hTFIIIC63 has been reported to interact with TBP, hBRF/TFIIIB90, and with pol III (25). As we supposed that 95 mutation could affect the upstream placement of the TFIIIB assembling subunit 131, we mapped the upstream border of TFIIIC·DNA complex. Highly purified TFIIIC blocked  exonuclease digestion of SUP4 DNA predominantly at positions 53 and also 47. Remarkably, the mutant factor lost the 53 arrest site and displayed only the 47 block. It is likely that 131 is responsible for  exonuclease blockage as it is the TFIIIC subunit that protrudes the most upstream, over the TFIIIB binding region (40). 131 was found to be cross-linked weakly to SUP4 DNA at positions 47/48 of the transcribed strand, and this cross-linking was markedly enhanced upon TFIIIB binding (44). This far upstream interaction is probably facilitated by DNA bending (46). Note that previous  exonuclease digestion experiments on a tRNA^Glu gene could not reveal such a far upstream border because the probe was truncated too close from the transcription start site (49). In contrast, photocross-linking of 95 was restricted over and closely around the A block (11). Therefore, as schematically illustrated in Fig. 9, we concluded that mutant 95 influenced the placement of 131 far upstream of the start site, thereby affecting TFIIIB placement and start site selection.

In the above experiments it should be noted that the mutant factor often imposed, or favored, particular states of C·DNA and B·C·DNA complexes or of the initiation complexes that can be weakly observed with wild type TFIIIC (like the 47 border of TFIIIC·DNA or the +4 or +8 start sites), as if the mutation favored certain conformations of the TFIIIC·DNA complex (for TFIIIC flexibility, see Ref. 7).

The second unexpected effect of the 95-E447K mutation was on TFIIIC·SUP4 DNA binding stability; a point mutation in the A block binding subunit was not expected to cause the dissociation of TFIIIC·DNA complexes. As shown by Baker et al. (4), the interaction between A block and TFIIIC contributes relatively little to the overall stability of the TFIIIC·SUP4 DNA complex. Mutations within the A block decreased the equilibrium constant (K_s) for TFIIIC·SUP4 DNA binding less than 2-fold, whereas point mutations within the B block decreased the K_s values several hundredfold. Even the total deletion of the A block reduced the K_s for TFIIIC binding at most 5-fold (4). The dissociation of mutant TFIIIC·DNA complexes at 38 °C therefore suggested a long distance effect of the mutated subunit on the DNA binding properties of B, the B-block binding module of TFIIIC (Fig. 9). Composed of three polypeptides 138, 91, and 60, two of which, 138 and 91, cooperate in DNA binding (9, 10, 13), B can be isolated by mild proteolysis of TFIIIC·DNA complexes. As shown in Fig. 7C, the mutant factor was normally susceptible to proteolysis, and B·DNA complex generated from the mutant factor displayed the same remarkable heat resistance as the control B·DNA complex. To interpret this paradoxical observation, one could imagine that the mutation causes a conformation change in 95 that is transmitted to B domain through the hinge that connects A and B. Our coimmunopurification experiments suggested that 95 participates in the interconnection of A with B via its contacts with 138 and 91. So 95 might directly influence B·DNA binding. 131 and 60 also appear as good candidates for this hinge function, 131 because this protein can be cross-linked between the A and B blocks (11, 44) and 60 because it belongs to B and participates in TFIIIB assembly (10). However, there is no evidence that 131 contact B components. Concerning 60, we want to retract the reported suppression of the ts phenotype of N-terminally tagged 60 by overexpression of 95 (10). The suppression was due to an unfortunate mishandling of plasmids and was not reproducible with bona fide TFC1.²

Our findings attribute to 95, and more globally to the A module, the ability to influence B·DNA stability. A conformation change of TFIIIC facilitating the dissociation of B may naturally occur during transcription where the RNA polymerase has to dissociate TFIIIC to transcribe the B block. Among he mechanisms that can be invoked to explain the release of TFIIIC by the advancing RNA polymerase (50), one could imagine a connection between the A block and B block binding domains of TFIIIC that would favor B block dissociation after the A block region is transcribed.

	ACKNOWLEDGEMENTS

We are grateful to Christine Conesa for suppressor verifications, to Emmanuel Favry for technical advice and protein preparations, and to Giorgio Dieci (University of Parma) for kindly providing tRNA^Asp(GTC) gene. We thank Gérald Peyroche, Christine Conesa, and Isabelle Callebaut (CNRS, UMR 7590) for helpful comments and stimulating discussions. We also thank Victor Fazakerley for improving the manuscript.

	FOOTNOTES

^* This work was supported by the Fondation pour la Recherche Médicale Fellowship FDT20000915040 (to S. J.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Present address: Neurodegeneration Research, Neurology-CEDD, GlaxoSmithKline, Third Ave., Harlow, Essex CM19 5AW, UK.

^§ To whom correspondence should be addressed: Service de Biochimie et Génétique Moléculaire, Bâtiment 144, CEA/Saclay, F-91191 Gif-sur-Yvette Cedex, France. Tel.: 33-1-69-08-59-57; Fax: 33-1-69-08-47-12; E-mail: Olivier.Lefebvre@Cea.Fr.

Published, JBC Papers in Press, January 16, 2003, DOI 10.1074/jbc.M213310200

² C. Conesa, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: TF, transcription factor; HA, hemagglutinin; wt, wild type; pol III, polymerase III; TBP, TATA-binding protein.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				H. Dumay-Odelot, C. Marck, S. Durrieu-Gaillard, O. Lefebvre, S. Jourdain, M. Prochazkova, A. Pflieger, and M. Teichmann Identification, Molecular Cloning, and Characterization of the Sixth Subunit of Human Transcription Factor TFIIIC J. Biol. Chem., June 8, 2007; 282(23): 17179 - 17189. [Abstract] [Full Text] [PDF]

				C. Ducrot, O. Lefebvre, E. Landrieux, J. Guirouilh-Barbat, A. Sentenac, and J. Acker Reconstitution of the Yeast RNA Polymerase III Transcription System with All Recombinant Factors J. Biol. Chem., April 28, 2006; 281(17): 11685 - 11692. [Abstract] [Full Text] [PDF]

				G. A. Kassavetis, E. Soragni, R. Driscoll, and E. P. Geiduschek Reconfiguring the connectivity of a multiprotein complex: Fusions of yeast TATA-binding protein with Brf1, and the function of transcription factor IIIB PNAS, October 25, 2005; 102(43): 15406 - 15411. [Abstract] [Full Text] [PDF]

				C. Conesa, R. Ruotolo, P. Soularue, T. A. Simms, D. Donze, A. Sentenac, and G. Dieci Modulation of Yeast Genome Expression in Response to Defective RNA Polymerase III-Dependent Transcription Mol. Cell. Biol., October 1, 2005; 25(19): 8631 - 8642. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 278/12/10450    most recent M213310200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jourdain, S. Articles by Lefebvre, O. Search for Related Content PubMed PubMed Citation Articles by Jourdain, S. Articles by Lefebvre, O. Social Bookmarking                      What's this?