Dual role of the C34 subunit of RNA polymerase III in transcription initiation
1997; Springer Nature; Volume: 16; Issue: 18 Linguagem: Inglês
10.1093/emboj/16.18.5730
ISSN1460-2075
Autores Tópico(s)RNA Research and Splicing
ResumoArticle15 September 1997free access Dual role of the C34 subunit of RNA polymerase III in transcription initiation Isabelle Brun Isabelle Brun Service de Biochimie et Génétique Moléculaire, Bâtiment 142, CEA/Saclay, F-91191 Gif-sur-Yvette, Cedex, France Search for more papers by this author André Sentenac André Sentenac Service de Biochimie et Génétique Moléculaire, Bâtiment 142, CEA/Saclay, F-91191 Gif-sur-Yvette, Cedex, France Search for more papers by this author Michel Werner Corresponding Author Michel Werner Service de Biochimie et Génétique Moléculaire, Bâtiment 142, CEA/Saclay, F-91191 Gif-sur-Yvette, Cedex, France Search for more papers by this author Isabelle Brun Isabelle Brun Service de Biochimie et Génétique Moléculaire, Bâtiment 142, CEA/Saclay, F-91191 Gif-sur-Yvette, Cedex, France Search for more papers by this author André Sentenac André Sentenac Service de Biochimie et Génétique Moléculaire, Bâtiment 142, CEA/Saclay, F-91191 Gif-sur-Yvette, Cedex, France Search for more papers by this author Michel Werner Corresponding Author Michel Werner Service de Biochimie et Génétique Moléculaire, Bâtiment 142, CEA/Saclay, F-91191 Gif-sur-Yvette, Cedex, France Search for more papers by this author Author Information Isabelle Brun1, André Sentenac1 and Michel Werner 1 1Service de Biochimie et Génétique Moléculaire, Bâtiment 142, CEA/Saclay, F-91191 Gif-sur-Yvette, Cedex, France *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:5730-5741https://doi.org/10.1093/emboj/16.18.5730 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The C34 subunit of yeast RNA polymerase (pol) III is part of a subcomplex of three subunits which have no counterpart in the other two nuclear RNA polymerases. This subunit interacts with TFIIIB70 and is therefore thought to participate in pol III recruitment. To study the role of C34 in transcription, we have mutagenized RPC34, the gene encoding C34, and found that mutations affecting growth also altered C34 interaction with TFIIIB70. The two mutant pol III that were purified had catalytic properties indistinguishable from those of the wild-type pol III on a poly[d(A–T)] template, while specific transcription of pol III genes in the presence of general transcription factors was impaired. The defect of the C34-1124 mutant enzyme could be compensated by increasing the amount of pol III present in the reaction, suggesting that the enzyme had a lower affinity for pre-initiation complexes. In contrast, the C34-1109 mutant enzyme was defective in transcription initiation due to impaired open complex formation. These observations demonstrate that the C34 subunit is a major determinant in pol III recruitment by the pre-initiation complex and further acts at a subsequent stage that involves the configuration of an initiation-competent form of RNA polymerase. Introduction Genes in eukaryotes are transcribed by one of three RNA polymerases. Pol I transcribes the 35S precursor of large rRNAs, pol II transcribes mRNAs and some small stable RNAs, and pol III transcribes tRNAs, 5S rRNA and some other small RNAs. Transcription initiation begins with the binding of general transcription factors to the gene promoter, forming the pre-initiation complex, each polymerase having its own set of factors. The pre-initiation complex is then recognized by its cognate RNA polymerase to form the initiation complex in a process that is poorly understood. Transcription initiation by pol III, except for 5S RNA genes which require TFIIIA, an additional transcription factor, begins with the binding of TFIIIC which then recruits TFIIIB. An initiation complex is then formed by the binding of pol III (White, 1994, and references therein). TFIIIB is the general transcription factor recognized by pol III since a TFIIIB.DNA complex can direct multiple rounds of transcription in vitro (Kassavetis et al., 1990). TFIIIB is composed of three polypeptides, TATA-binding protein (TBP; Huet and Sentenac, 1992; Kassavetis et al., 1992b), a general transcription factor required for transcription by all eukaryotic and archaebacterial RNA polymerases (Struhl, 1995, and references therein), TFIIIB90, a 90 kDa subunit which has no equivalent among the other RNA polymerase general transcription factors (Kassavetis et al., 1995; Roberts et al., 1996; Rüth et al., 1996), and TFIIIB70, a 70 kDa protein which is homologous to archaeal general factor TFB (Hausner and Thomm, 1995; Qureshi et al., 1995) and to pol II factor TFIIB (Buratowski and Zhou, 1992; Colbert and Hahn, 1992; López-De-León et al., 1992). TFIIB is the last general transcription factor to enter the class II pre-initiation complex before pol II (Buratowski et al., 1989), pointing to the possibility that TFIIIB70 might similarly recruit pol III. Three pol III subunits, C82, C34 and C31, that have no counterpart in the other RNA polymerases (Mosrin et al., 1990; Chiannilkulchai et al., 1992; Stettler et al., 1992), form a subcomplex (Werner et al., 1992, 1993) which might be implicated in transcription initiation. Indeed, mutations in the gene encoding the C31 subunit affect transcription initiation but not the general catalytic properties of the enzyme (Thuillier et al., 1995). The role of the C34 subunit is not known presently, but several lines of evidence suggest that it is also implicated in transcription initiation. Of all pol III subunits, C34 is the one that cross-links the furthest upstream on the promoter DNA in initiation complexes (Bartholomew et al., 1993; Persinger and Bartholomew, 1996). Moreover, antibodies directed against the subunit inhibit in vitro transcription of a tRNA template but not non-specific transcription on poly[d(A–T)] (Huet et al., 1985). Finally, the observation that C34 interacts both in vivo and in vitro with TFIIIB70 has led us to propose that it might be implicated in the recruitment of pol III by the pre-initiation complex (Werner et al., 1993; Khoo et al., 1994). In this study, using conditional mutations affecting the C34 subunit of pol III, we demonstrate that it plays an essential role in transcription initiation, not only in the recognition of the pre-initiation complex by pol III, but also, more unexpectedly, at the level of open complex formation. Results Mutagenesis of RPC34 Three rpc34 conditional mutations have been described previously (Stettler et al., 1992). We wanted to pursue the characterization of the C34 pol III subunit, first by obtaining tighter conditional growth mutations and, second, by assaying the effect of the mutations on the interaction of C34 with its partners. For that purpose, we used an RPC34 allele, RPC34-1001, which behaves as the wild-type and allows the fusion of the RPC34 open reading frame (ORF) in-frame with the GAL4 DNA-binding domain (GDB) or GAL4 activation domain (GAD) to test the effect of mutations in the two-hybrid system (Werner et al., 1993; see below). We thus mutagenized RPC34-1001 using oligonucleotides following the 'charged cluster analysis' strategy (Wertman et al., 1992). This method targets the residues located at the surface of the protein by changing positively or negatively charged residues to alanine when at least a pair is present within a sequence of five amino acids. Thirty three such mutations were constructed to cover the entire RPC34 ORF (see Materials and methods; Figure 1). Two other mutations, rpc34-1145 and rpc34-1146, were obtained spuriously during oligonucleotide mutagenesis. Finally, eight mutations were targeted at specific residues among which two (rpc34-1139 and rpc34-1140) altered the same amino acid as the rpc34-E89K and rpc34-D171H mutations described previously (Stettler et al., 1992). The mutant genes were transformed into strain D57-12C which harbours the rpc34-Δ::HIS3 deletion complemented by a wild-type copy of the RPC34 gene borne on a URA3 plasmid pYS34 (Stettler et al., 1992). The phenotype of the mutations was tested by plasmid shuffling (Boeke et al., 1987), selecting the Ura− clones that had lost the wild-type resident plasmid. Figure 1.Charged cluster mutagenesis of the C34 subunit. The mutations which have been introduced into the C34 subunit are indicated beneath the protein sequence. Multiple substitutions are indicated by dashes joining the mutant amino acids when they are not adjacent. * indicates a nonsense mutation. The allele names of the mutations that give rise to a growth phenotype are indicated in parentheses. The phenotypes are coded in the following way: l, lethal; cs, cryosensitive growth at 16°C; sg, slow growth at 24°C. Download figure Download PowerPoint Of the 33 mutations constructed according to the 'charged cluster analysis' scheme, only two displayed a conditional phenotype. Mutant rpc34-1109 displayed reduced growth at the permissive temperature of 24°C and ceased to grow at 16°C. Mutant rpc34-1124 grew nearly as well as the wild-type at 24°C but showed only marginal growth at 16°C (Figure 2). Curiously, the RE102-103AA change was silent while the RE102-103VA change (rpc34-1146) displayed a cryosensitive phenotype. Since, of all the conditional mutations, rpc34-1109 had the most drastic effect, we separated the K135A and K138A amino acid replacements and found that only the former (K135A in rpc34-1135) had a detectable effect, though it was less pronounced than when combined with K138A. Other mutations on either side of rpc34-1135 were phenotypically silent, showing that K135 is critical for C34 function. The phenotype of rpc34-1139 and rpc34-1140 (identical to mutations rpc34-D171K and rpc34-E89K; Stettler et al., 1992) are also shown in Figure 2 for comparison. Quite strikingly, none of the mutations were thermosensitive at 37°C. Figure 2.Crysosensitive growth of mutant RPC34 strains. Wild-type (RPC34-1001) or mutant (rpc34-1109, -1124, -1135, -1139, -1140, -1146) strains were grown for 4 days at 24°C or 7 days at 16°C on YPD rich medium. Download figure Download PowerPoint Three mutations, rpc34-1014, -1138 and -1145, were lethal: rpc34-1014 resulted from a EG103-104IR double substitution, overlapping the conditional rpc34-1146 mutation, rpc34-1145 had six amino acid changes and rpc34-1138 a deletion of the 74 C-terminal amino acids. This latter mutation, which shows that the C34 C-terminus is required for the function of the subunit, was constructed because multiple mutations that changed up to five contiguous acidic residues in that region had no phenotypic effect (Figure 1). The low number of lethal mutations was unexpected since a 'charged cluster mutagenesis' of region f of C160 pol III subunit yielded 21 lethal and four conditional mutants out of a total of 27 (Thuillier et al., 1996). This double alanine scanning mutagenesis and previous attempts at obtaining conditional mutations affecting C34 through hydroxylamine mutagenesis (Stettler et al., 1992) suggest that the C34 subunit is very resistant to amino acid substitutions that lead to thermosensitivity. The C34 subunit interacts with C31 and C82 subunits of pol III and with the TFIIIB70 subunit of TFIIIB, as evidenced by two-hybrid experiments (Werner et al., 1993). To explore the physiological relevance of these interactions, we tested whether the phenotype of the lethal or conditional mutations could be due to weakened protein–protein contacts between C34 and its partners. Each C34 mutant allele was cloned in the pACT2 vector to yield GAD::C34-1### (where ### represents three digits) fusions. These were tested in the two-hybrid system against GDB::C31, GDB::C82 and GDB::TFIIIB70 fusions cloned in vector pAS2 after growth at 30°C (see Materials and methods). In each case, the correct expression of the GAD::C34-1### fusion was tested using antibodies directed against C34 (Huet et al., 1985) in order to eliminate the possibility that some of the rpc34 mutations might reduce the level of expression and/or the stability of the fusion protein. Strikingly, all the mutations that showed a growth defect affected the interaction with TFIIIB70 (Table I). The two lethal mutations (rpc34-1138 and rpc34-1145) that were tested were also severely affected in their interaction with the C31 and C82 subunits of pol III. The altered interactions were observed even though the cells were grown at 30°C, the permissive temperature for the mutations. Two possibilities might explain this observation. First, we observed that the rpc34-1124 and -1139 mutants, though growing normally at 30°C, showed reduced in vivo transcription of tRNAs, indicating that the mutation already exerted its effect even at the permissive temperature (see below). Second, in pol III, the mutant C34 subunit is part of a multiprotein complex that might stabilize its conformation and allow its interaction with TFIIIB70. This is probably not the case in the two-hybrid assay that reflects direct interactions between protein pairs overproduced in the cells. Whatever the case, our observations strongly suggested that the decreased interaction between C34 mutant subunits and TFIIIB70 were responsible for the growth phenotype of the mutant strains. Table 1. Two-hybrid interations between mutant C34 proteins and C31, C82 and TFIIIB70 RPC34 allele Growth phenotypeb Two-hybrid interactiona C31c C82d TFIIIB70c RPC34-1001 wild-type ++ ++ ++ rpc34-1109 Sg, Cs ++ + − rpc34-1124 Cs ++ ++ − rpc34-1135 Cs ++ + − rpc34-1138 lethal + − − rpc34-1139 Cs ++ ++ − rpc34-1140 Cs ++ ++ − rpc34-1145 lethal + +/− − rpc34-1146 Cs ++ ++ + a ++ represented the wild-type level of lacZ activation as determined by the β-galactosidase overlay assay on patches of cells growing at 30°C on minimal medium. This assay is linear in response to the β-galactosidase activity (Werner et al., 1993). + represented intermediate levels of coloration, +/− very light blue colour and − white colour. For comparison, interaction between wild-type C34 and C31 led to the production of 123 U of β-galactosidase, to 145 U with C82 and to 246 U with TFIIIB70 (Werner et al., 1993); the background level was 84% of the 17 nucleotide RNA transcript formed by the wild-type pol III. The rate of synthesis of full-length RNA was then followed by analysing the RNA chain pattern at different time points after the addition of GTP and heparin, which prevents pol III recycling. As shown in Figure 5C, the pause pattern was similar for the wild-type and mutant enzymes. Unexpectedly, the mutant pol III synthesized full-length transcripts faster than the wild-type enzyme since full-length SUP4 pre-tRNA appeared after 2 s in the first case and 6 s in the second (Figure 5C). This observation at least proved that an elongation defect could not account for the slower rate of RNA synthesis by the mutant pol III. Recently, Dieci and Sentenac (1996) observed that transcription initiation by a recycling pol III was faster than the first initiation step, suggesting that there is a facilitated recycling pathway for the enzyme. We thus asked whether recycling might also be affected in pol III with mutant C34 subunit. This was tested by performing multiple round transcription assays on SUP4 DNA starting with ternary complexes stalled at position 17 in the absence of GTP. The number of transcription cycles during a short incubation period was then determined after addition of GTP with heparin (single round transcription) or without heparin (multiple rounds; Thuillier et al., 1995, 1996). The average time needed by the C34-1109 enzyme to complete one cycle was 45 s at early time points, which was significantly longer than the 30 s required by the wild-type enzyme (Figure 6). Furthermore, the measured cycling time increased to >60 s for the mutant pol III, while it remained constant for the wild-type enzyme (Figure 6B). The behaviour of the mutant showed that the C34 subunit plays a role in both the first initiation event and the facilitated recycling pathway. Figure 6.Kinetics of transcript accumulation by re-initiating wild-type and C34-1109 pol III. (A) Wild-type or mutant pol III were incubated at room temperature for 15 min in the presence of the SUP4 tRNA gene, 40 ng of recombinant TBP, 50 ng of recombinant TFIIIB70, 400 ng of a B″ fraction (containing TFIIIB90), 50 ng of affinity-purified TFIIIC, 500 μM ATP, 500 μM CTP and 30 μM 32P-labelled UTP to form a halted ternary complex complex at position 17. Elongation of the 17mer RNA was allowed to resume for 1.5–8 min by addition of 600 μM GTP and 300 μg/ml heparin, which enabled only a single round of transcription (SRT), or GTP alone, which allowed multiple rounds of transcription (MRT). Lanes 1–8, wild-type pol III; lanes 9–16, C34-1109 pol III. (B) PhosphorImager quantification of the data shown in (A). □ Wild-type pol III; ▴ C34-1109 pol III. Download figure Download PowerPoint Altogether, these experiments suggested that C34-1109 pol III was impaired at a step subsequent to pre-initiation complex recognition and prior to elongation, i.e. open complex formation and/or promoter clearance. Since open complex formation is strongly dependent on temperature (Kassavetis et al., 1992a), we assayed transcription initiation on a supercoiled SUP4 template, as a function of temperature, again allowing the reaction to proceed for 15 min. As shown in Figure 7, contrary to what was observed for the wild-type pol III, transcription by the C34-1109 enzyme was very inefficient at low temperatures (4–10°C). The transition temperature (to reach 50% of total transcripts) was 15°C for the mutant pol III as compared with 5°C for the wild-type enzyme. The same experiment was done using a linear template and also revealed a higher transition temperature for the mutant pol III as compared with the wild-type enzyme (data not shown). Figure 7.Transcription initiation by wild-type and C34-1109 pol III as a function of temperature. Stable pre-initiation complexes were formed on the SUP4 tRNA gene at 24°C as described in Materials and methods, and then incubated further at the temperature indicated for 10 min. Wild-type or mutant pol III were then added together with 500 μM ATP, 500 μM CTP and 10 μM 32P-labelled UTP equilibrated at the same temperature. Synthesis of the 17mer was allowed to proceed for 15 min at the temperature indicated. The reaction products were separated on a 15% polyacrylamide–7 M urea gel. (A) Autoradiographs of the RNA products. The position of the 17 nucleotide transcript is indicated. (B) PhosphorImager quantification of the data shown in (A). □ Wild-type pol III; ▵ C34-1109 pol III. Download figure Download PowerPoint Open and closed complexes exhibit a different sensitivity to heparin (Kassavetis et al., 1992a; Dieci and Sentenac, 1996). We therefore compared the heparin sensitivity of pre-formed wild-type and mutant initiation complexes. Mutant or wild-type pol III was pre-incubated for 15 min with TFIIIB.TFIIIC.DNA complexes at 24°C then assayed for 17mer synthesis by the addition of ATP, CTP and labelled UTP in the presence of various heparin concentrations (this second incubation was performed for 20 min). As shown in Figure 8, mutant initiation complexes were clearly more sensitive to heparin than wild-type complexes. The concentration of heparin required to achieve 50% inhibition of mutant complexes was intermediate between that required to inhibit free pol III (0.25–0.5 μg/ml; Kassavetis et al., 1992a) and open complexes (5 μg/ml; Dieci and Sentenac, 1996). Figure 8.Heparin sensitivity of initiation complexes. Pre-initiation complexes were formed on the SUP4 tRNA gene as described in Materials and methods. Then 100 ng of wild-type or mutant pol III were added and the incubation continued for 15 min at 24°C to form the initiation complexes. Then 500 μM each of ATP and CTP and 3 μM 32P-labelled UTP were added, together with varying concentrations of heparin, as indicated. Synthesis of the 17mer was allowed to proceed for 20 min at 24°C. The 17mer, the 20mer slippage product and the 15mer cleavage product (Dieci et al., 1995) were separated on a 15% polyacrylamide–7 M urea gel. (A) Autoradiographs of the RNA products. (B) PhosphorImager quantification of the trancripts (sum of 15, 17 and 20mer intensities) observed on the autoradiogram shown in (A). □ Wild-type pol III; ▵ C34-1109 pol III. Download figure Download PowerPoint The cold and heparin sensitivities of the mutant enzyme were strongly suggestive of a defect at the level of open complex formation but could also stem from an uncharacterized effect of the rpc34-1109 mutation. To eliminate this latter possibility, the formation of the open complex was investigated directly using KMnO4 footprinting to probe the accessibility of T residues in the transcription bubble (Kassavetis et al., 1992a). Pre-initiation complexes were first formed by incubating end-labelled SUP4 template with TFIIIC and recons
Referência(s)