The plant-specific TFIIB-related protein, pBrp, is a general transcription factor for RNA polymerase I
2008; Springer Nature; Volume: 27; Issue: 17 Linguagem: Inglês
10.1038/emboj.2008.151
ISSN1460-2075
AutoresS. Imamura, Mitsumasa Hanaoka, Kan Tanaka,
Tópico(s)Plant Gene Expression Analysis
ResumoArticle31 July 2008 The plant-specific TFIIB-related protein, pBrp, is a general transcription factor for RNA polymerase I Sousuke Imamura Sousuke Imamura Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Tokyo, Japan Search for more papers by this author Mitsumasa Hanaoka Mitsumasa Hanaoka Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Tokyo, Japan Search for more papers by this author Kan Tanaka Corresponding Author Kan Tanaka Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Tokyo, Japan Graduate School of Horticulture, Chiba University, Matsudo, Chiba, Japan Search for more papers by this author Sousuke Imamura Sousuke Imamura Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Tokyo, Japan Search for more papers by this author Mitsumasa Hanaoka Mitsumasa Hanaoka Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Tokyo, Japan Search for more papers by this author Kan Tanaka Corresponding Author Kan Tanaka Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Tokyo, Japan Graduate School of Horticulture, Chiba University, Matsudo, Chiba, Japan Search for more papers by this author Author Information Sousuke Imamura1, Mitsumasa Hanaoka1 and Kan Tanaka 1,2 1Institute of Molecular and Cellular Biosciences, The University of Tokyo, Bunkyo-ku, Tokyo, Japan 2Graduate School of Horticulture, Chiba University, Matsudo, Chiba, Japan *Corresponding author. Graduate School of Horticulture, Chiba University, 648 Matsudo, Matsudo, Chiba 271-8510, Japan. Tel.: +81 47 308 8866; Fax: +81 47 308 8866; E-mail: [email protected] The EMBO Journal (2008)27:2317-2327https://doi.org/10.1038/emboj.2008.151 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info TFIIB and BRF are general transcription factors (GTFs) for eukaryotic RNA polymerases II and III, respectively, and have important functions in transcriptional initiation. In this study, the third type of TFIIB-related protein, pBrp, found in plant lineages was characterized in the red alga Cyanidioschyzon merolae. Chromatin immunoprecipitation analysis revealed that CmpBrp specifically occupied the rDNA promoter region in vivo, and the occupancy was proportional to de novo 18S rRNA synthesis. Consistently, CmpBrp and CmTBP cooperatively bound the rDNA promoter region in vitro, and the binding site was identified at a proximal downstream region of the transcription start point. α-Amanitin-resistant transcription from the rDNA promoter in crude cell lysate was severely inhibited by the CmpBrp antibody and was also inhibited when DNA template with a mutated CmpBrp–CmTBP binding site was used. CmpBrp was shown to co-immunoprecipitate and co-localize with the RNA polymerase I subunit, CmRPA190, in the cell. Thus, together with comparative studies of Arabidopsis pBrp, we concluded that pBrp is a GTF for RNA polymerase I in plant cells. Introduction The general class II transcription machinery is composed of RNA polymerase II (Pol II) and six general transcription factors (GTFs): TFIIA, TFIIB, TFIID, TFIIE, TFIIF and TFIIH (Hahn, 2004). As the first step in preinitiation complex (PIC) assembly, TFIID, which includes the TATA-binding protein (TBP) and TBP-associated factors (TAFs), binds to the core promoter, and the DNA–protein complex is subsequently stabilized by TFIIB. TFIIB is a monomeric protein containing two imperfect direct repeats in the carboxy-terminal region (Deng and Roberts, 2007). These two repeats can directly form a sequence-specific interaction with a core promoter element known as BRE (TFIIB recognition element) (Lagrange et al, 1998; Deng and Roberts, 2005, 2007). The amino-terminal region of TFIIB contains a zinc-chelating motif, which is conserved in all TFIIBs identified to date, and is indispensable for recruitment of the Pol II–TFIIF complex to form PIC (Tubon et al, 2004). The carboxy-terminal proximal to the zinc-ribbon motif is the most conserved region of TFIIB, and is termed as the B-finger. The X-ray crystallography of a yeast Pol II–TFIIB complex revealed that the B-finger domain of TFIIB is inserted into the presumed RNA exit channel of Pol II (Bushnell et al, 2004). It was also revealed that the B-finger has an important function after PIC formation and transcription start site selection (Bushnell et al, 2004; Deng and Roberts, 2007). In addition to TFIIB, the eukaryotic genome encodes another TFIIB-related protein, BRF; this is a component of the TFIIIB complex, which is the core RNA polymerase III (Pol III) initiation factor by virtue of its role in recruiting Pol III to the transcriptional start site, and is essential to form the transcription-ready open promoter complex (Kassavetis and Geiduschek, 2006). Therefore, TFIIB and BRF has an important function in Pol II- and Pol III-dependent transcription during the initiation processes. Recently, Lagrange et al (2003) reported that higher plant genomes encode a plant-specific TFIIB-related protein, pBrp, as well as TFIIB and BRF. The structure of pBrp contains two TFIIB features: an amino-terminal zinc-ribbon motif and carboxy-terminal imperfect direct repeats. Interestingly, in contrast to known GTFs, the majority of Arabidopsis thaliana pBrp (AtpBrp) and Spinacea oleracea pBrp proteins were bound to the cytoplasmic face of the plastid envelope membrane in leaves. AtpBrp was detected in the nucleus either when the Arabidopsis cell was treated with the proteasome inhibitor, MG132, or in a fus6 mutant, in which the COP9 signalosome is deficient. Thus, it was proposed that the AtpBrp protein harbours a proteolytic signal that can target it for rapid turnover by the proteasome-mediated protein degradation pathway in the nucleus. It was also shown that AtpBrp and AtTBP2 can cooperatively form a ternary complex in vitro with the adenovirus type 2 major late (Ad2ML) promoter, which harbours consensus BRE- and TBP-binding motifs. Therefore, their data suggested the possible involvement of AtpBrp in intracellular signalling between plastids and the nucleus. However, the relevant function in plant cells of the third type of TFIIB-related protein, pBrp, remained unknown. Cyanidioschyzon merolae is a thermo-acidophilic unicellular red alga isolated from an Italian volcanic hot spring (Kuroiwa, 1998). Recently, a genome project team, including ourselves, determined the 100% complete DNA sequence of this organism including the nuclear and two organelle genomes (Ohta et al, 1998, 2003; Matsuzaki et al, 2004; Nozaki et al, 2007). Because of its extremely simple cell structure and the minimally redundant small genome, this alga is considered to be a suitable model to study the origin and evolution of photosynthetic eukaryotes. It is also useful for studying the fundamental transcriptional network, as it possesses a small number of transcription factors, less than 100 for the 16.5 Mbp of the nuclear genome (Kuroiwa, 1998; Matsuzaki et al, 2004; Nozaki et al, 2007). With respect to TFIIB and related proteins, the C. merolae genome contains genes for TFIIB, BRF and pBrp, each as a single copy (see Results for details). In this study, as a first step to understand the function of pBrp for nuclear transcription, we analysed its target gene in C. merolae by chromatin immunoprecipitation (ChIP) assays. Our data indicated that pBrp specifically occupies the rDNA promoter region in vivo, and in vivo and in vitro analyses showed that C. merolae pBrp is positively involved in the RNA polymerase I (Pol I)-dependent rRNA synthesis in the nucleolus. Moreover, we also show evidence that the pBrp function is also conserved for pBrp in A. thaliana. Thus, we propose here that pBrp is a GTF for Pol I in plant cells. Results PBRP gene in C. merolae and its expression The C. merolae 100%-complete genome sequence (Matsuzaki et al, 2004; Nozaki et al, 2007) revealed 4775 open reading frames on 20 chromosomes and showed that CMI217C (gene number in http://merolae.biol.s.u-tokyo.ac.jp/), CML077C and CMA019C encode TFIIB-related proteins. To assess phylogenetic relationships among these genes, a maximum-likelihood tree was constructed with TFIIB and related proteins from yeast, human, green alga, higher plants and Archaea (Figure 1A). The resultant phylogenetic tree consists of four independent groups, each represented by eukaryotic TFIIB, BRF, pBrp and archaeal TFB, and showed that CMI217C, CML077C and CMA019C proteins were assigned into the pBrp, TFIIB and BRF groups, respectively. Thus, we designated here CMI217C, CML077C and CMA019C as CmpBrp, CmTFIIB and CmBRF, respectively. Comparison of the deduced amino-acid sequence of CmpBrp with those of higher plants' pBrps showed that CmpBrp has a higher molecular mass (981 aa, 104 kDa) and includes extensions on both amino- and carboxy-termini and several insertions without similarity (Supplementary Figure S1). However, the imperfect direct repeats exhibit a high degree of amino-acid sequence similarity to higher plants' pBrp proteins (Figure 1B and Supplementary Figure S1). The consensus TFIIB zinc-ribbon motif (Cys-X2-Cys/His-X15−17-Cys-X2-Cys) was observed in the higher plants' pBrp but not in CmpBrp, whereas a potential zinc-binding motif (Cys-X2-Cys-X19-Cys-X2-His) was found at the amino-terminal region of CmpBrp (Figure 1B and Supplementary Figure S1). Figure 1.TFIIB-related protein family in C. merolae. (A) Evolutionary relationship of TFIIB-related proteins. Numbers at each node represent the percentage of trees supporting the specific branching pattern in the bootstrap analysis. Branch lengths are proportional to the number of amino-acid substitutions, indicated by the scale bar below the tree. Designations and GenBank accession numbers for sequences of the TFIIB-related proteins are shown in Supplementary data. MmTFB, PtTFB, HlTFB and NpTFB were used as the out-group to root the tree. (B) Schematic representation of secondary structure of CmpBrp, CmTFIIB and CmBRF. White, grey, black and hatched boxes indicate predicted putative zinc ribbon, B-finger, imperfect direct repeats and BRF domain, respectively. Download figure Download PowerPoint To investigate CmpBrp expression, a rabbit polyclonal antibody against recombinant CmpBrp expressed in Escherichia coli was produced. The prepared antibody specifically recognizes endogenous CmpBrp protein (approximately 118 kDa, Supplementary Figure S2). We performed an immunoblot analysis using the CmpBrp antibody to probe C. merolae total protein extracted from cells exposed to several environmental stress conditions: high pH, high osmotic pressure, high temperature, high intensity of light, nitrogen starvation and carbon starvation. The results revealed that CmpBrp protein is constitutively expressed in the cell (data not shown). Identification of CmpBrp target gene by ChIP As pBrp is paralogous to TFIIB and BRF, we hypothesized that pBrp also functions as a GTF for nuclear RNA polymerase(s). To examine this possibility, we conducted ChIP analysis to identify the target gene in the nuclear genome, as a targeted gene disruption system for C. merolae has been not available to date. ChIP analysis was first carried out for proximal regions close to transcriptional start points (TSPs) of genes that are transcribed by the three classes of RNA polymerase; Pol I, Pol II and Pol III. Promoter regions of rDNA and 5S rDNA were representatively analysed as Pol I- and Pol III-dependent promoters, respectively. As for Pol II-dependent promoters, genes that show light-responsive expression patterns, that are light and dark responsive, that are constitutively expressed irrespective of the light conditions, or that are nitrogen-deprivation responsive were selected on the basis on microarray results (Kanesaki et al, our unpublished data; Figure 3A). Whereas TSPs of the Pol II-type genes could be predicted on the basis of the full-length EST analysis (Matsuzaki et al, 2004), those of the Pol I- and Pol III-dependent genes had not been experimentally identified. Therefore, these TSPs were first mapped at low resolution by northern blot analyses with several DNA probes (Figure 2A and D). The result indicated that the TSP of rDNA is located about 2000-bp upstream of the mature 5′-end (Figure 2B). Although this information was sufficient for ChIP analysis, the precise TSP of rDNA was further analysed by primer extension and S1 nuclease protection assays (Figure 2C) for in vitro biochemical analyses (Figures 4 and 5). It should be noted that the C. merolae nuclear genome encodes only three rRNA units at different chromosomal loci, named here as RRNa (CMQ305R), RRNb (CMQ401R) and RRNc (CMR208R) (Maruyama et al, 2004; Matsuzaki et al, 2004; Nozaki et al, 2007). Each of the three rDNA units in the coding and the upstream regions (up to ∼3.7-kb upstream from the putative 5′-end of 18S rRNA, http://merolae.biol.s.u-tokyo.ac.jp/) share an almost identical DNA sequence. Therefore, it is difficult to distinguish the major reverse transcription product and to determine which RRN gene(s) yielded the S1 nuclease-protected RNA. Providing that all rDNA genes are active, TSPs of rDNA genes were mapped to be 1949-bp upstream from the putative 5′-end of 18S rRNA for RRNa, 1949-bp for RRNb and 1946-bp for RRNc (Figure 2C and F). It is of note that the different TSP is due to some gaps between the mature 5′-end and TSP of the three rDNA genes. In any case, the TSP information is consistent with the data obtained by northern blot analysis (Figure 2B) and was sufficient for in vitro biochemical analyses shown in Figures 4 and 5. As for the 5S rRNA genes, there are three chromosomal loci; CML038R, CML060R and CMM232R. Nucleotide sequences around these three genes are almost identical, and the TSPs of 5S rRNA genes were mapped by northern blot analysis at almost identical positions in the mature 5′-end (Figure 2E). Thus, we did not analyse further to discriminate the three copies because the present information was sufficient for the ChIP analysis. Figure 2.Determination of transcription start sites of C. merolae rDNA and 5S rDNA. (A) Positions of probes (a–f) used in northern blot analysis for rDNA. (B) Mapping of TSP of rDNA by northern blot analysis. C. merolae total RNA (6 μg) prepared from logarithmic-growth cells (approximately OD750=0.5) under continuous light conditions was subjected to northern blot analysis with relevant specific probes shown in panel (A). Transcript size is indicated at the left in kilonucleotides. Putative positions of prematured full-length rRNA (full length), prematured 18S rRNA (Pre-18S) and matured 18S rRNA (Matured 18S) are indicated at the right. Lower panel shows rRNA stained with methylene blue as a loading control. (C) 5′-end mapping of rDNA TSP by primer extension and S1 nuclease protection assays. Positions of primer (arrowhead) and DNA probe (line) used in primer extension analysis and S1 nuclease protection assay are indicated at the top. Asterisks indicate 32P labelled 5′-end of DNA. Primer extension analysis (left) or S1 nuclease protection assay (right) was carried out with the same RNA used in panel B (lane 1) or yeast RNA (lane 2). Arrowhead indicates the TSP of rDNA. (D) Positions of probes (a–e) used in northern blot analysis for 5S rDNA. (E) Mapping of TSP of 5S rDNA by northern blot analysis. Others are the same as in panel B. (F) Sequence of rDNA (RRNb) promoter region. Arrow denotes the TSP. Download figure Download PowerPoint Cells were grown under continuous light conditions, and ChIP analysis with the CmpBrp antibody was performed as shown in Figure 3A. Intriguingly, the result indicated that CmpBrp occupies the promoter region of the rDNA (rDNA no. 1). The occupancy was not detected at the promoter regions of mRNA-encoding genes and 5S rDNA and was significantly reduced at approximately +2500 bp from the TSP of the rDNA (rDNA no. 2). These results implied that CmpBrp takes part in rRNA synthesis. To further understand this, we next examined the correlation between the occupancy of CmpBrp on the rDNA promoter region and the de novo 18S rRNA synthesis rate, which was determined by run-on transcription assay. By exposing to light after adaptation to darkness, the level of de novo 18S rRNA synthesis was apparently elevated approximately 2.0- to 2.3-fold that at L0 (just before the light irradiation). Upon re-exposure of the cells to darkness, the de novo 18S rRNA synthesis level dropped back to the level observed at L0. Under the same conditions as the run-on analysis, ChIP analysis was carried out and the result clearly indicated that levels of CmpBrp occupancies on the rDNA promoter region were well correlated with those of de novo synthesis of 18S rRNA (Figure 3B and D). The protein level of CmpBrp was constant irrespective of the light conditions (Figure 3C), suggesting that post-translational modification and/or other regulation of CmpBrp enhances its binding affinity to the rDNA promoter region. Figure 3.CmpBrp specifically binds to rDNA promoter region in vivo. (A) Identification of CmpBrp target gene. ChIP analysis was performed with CmpBrp antibody (CmpBrp Ab.) or rabbit IgG purified from preimmune serum (Pre) using fixed logarithmic growth cells (OD750=0.5) under continuous light conditions. Occupancy of CmpBrp was measured at the indicated promoter regions. For rDNA, the ≈2500-bp downstream region from its TSP (rDNA no. 2) as well as its promoter region (rDNA no. 1) were analysed. Expression types of Pol II-type genes are shown at the bottom. L, light-responsive; L & D, light- & dark-responsive; Cont., constant; -N, N deprivation-responsive. Values are averages of at least three independent experiments and represent percent recovery relative to the total input DNA. Error bars indicate standard deviation. (B) Level of de novo 18S rRNA synthesis under the light/dark cycle. C. merolae cells were grown under white light until OD750=0.5 and incubated in complete darkness for 18 h. Lights were turned on for 6 h and then turned off again. Cells were sequentially harvested at the time (h) shown at the top, and the newly synthesized 18S rRNA was detected by run-on analysis. Signals obtained with rDNA (18S rRNA) or pDEST-HIS (negative control) as probe DNAs are shown in the middle. Signal intensities of de novo synthesized 18S rRNA from three independent experiments were quantified and values (L0 as 100%) are presented (means±s.d.) as relative levels at the bottom. (C) Protein level of CmpBrp in the light/dark cycle. Aliquots of total protein (15 μg) from the cell lysate were subjected to immunoblot analysis with CmpBrp antibody. Total protein (around 55 kDa) stained with Coomassie Brilliant Blue (CBB) is shown as loading control (lower panel). (D) CmpBrp association on the rDNA promoter region in the light/dark cycle. ChIP analysis was carried out in the same way as in panel (A) but with sample prepared similar to panel (B). Occupancy of CmpBrp at promoter regions of rDNA (rDNA no. 1) and CMK028C (K028) was analysed. (E, F) ChIP analysis with CmTFIIB (E) and CmBRF (F) antibodies. ChIP analyses were carried out with sample prepared at L2. Occupancies of relevant factors were analysed at the indicated promoter regions. Others are the same as in panel (A). Download figure Download PowerPoint Next, we investigated a functional distinction among CmpBrp, CmTFIIB and CmBRF on the three types of RNA polymerase-dependent transcription. To achieve this, polyclonal antibodies against recombinant CmTFIIB and CmBRF were prepared (Supplementary Figure S2) and subjected to ChIP analysis with the sample prepared at L2. As in other eukaryotes, specific occupancies of CmTFIIB and CmBRF were detected at promoter regions of genes for mRNAs (CMJ289C and CMK028C) and 5S rRNA, respectively, whereas little occupancy was detected for the rDNA promoter region using both antibodies (Figure 3E and F). These results strongly suggested that CmpBrp is a specific GTF for rDNA transcription. Cooperative binding of CmpBrp and CmTBP to the rDNA promoter region in vitro To verify whether CmpBrp is actually able to bind directly to the rDNA promoter region, electrophoretic mobility shift assay (EMSA) analysis was performed. Recombinant CmpBrp and CmTBP proteins (Figure 4A) were subjected to EMSA, as one of the functional characteristics of TFIIB is the ability to bind TBP and form a stable ternary complex with promoters (Orphanides et al, 1996). Here, CmTFIIB was used as a control. Recombinant CmpBrp did not bind to the rDNA promoter region; however, further addition of CmTBP to the reaction mixture resulted in a stable ternary complex formation (Figure 4B, lane 3 versus 4). On the other hand, CmTFIIB did not form a complex with the rDNA promoter region even in the presence of CmTBP. To further examine the binding characteristic of CmpBrp, EMSA was also performed with the Ad2ML promoter, which is dependent on Pol II for transcription, and harbours canonical BRE and TATA-box consensus sequences. As shown in Figure 4C, a ternary complex was observed only with CmTFIIB but not with CmpBrp even in the presence of CmTBP (lane 4 versus 5), suggesting that the DNA recognition specificity of CmpBrp is different from that of CmTFIIB. The binding site of the CmpBrp–CmTBP complex in the rDNA promoter region was investigated by competition experiments with unlabelled probes at positions indicated in Figure 4D. Excess unlabelled cold probe harbouring the rDNA promoter region (−161 to +55, +1 as the TSP) and no. 7 probe (+2 to +28) markedly competed for protein binding (lanes 1 versus 2 and 9), whereas other probes did not. To identify the CmpBrp–CmTBP complex binding site in the no. 7 probe, we next performed EMSA competition analysis using scanning mutagenized no. 7 probes. The sequences of the probes are shown in Figure 4E. Our results clearly indicated that five of nine unlabelled probes, no. 7_Mt3 to no. 7_Mt7, all of which were mutated at position +8 to +22, lost competitive activity for ternary complex formation (Figure 4E). Consistent with this, DNase I footprinting analyses also revealed that the CmpBrp–CmTBP complex protected the rDNA promoter region at +6 to +22 of the sense strand and at +8 to +22 of the antisense strand (Figure 4F). To determine whether the identified site was critical for ternary complex formation, EMSA analysis was carried out using a DNA probe with mutagenized sequences at +8 to +22. As shown in Figure 4G, the stable ternary complex was not observed when the mutated probe was used (lane 2 versus 4). These results clearly indicated that CmpBrp and CmTBP cooperatively bind the rDNA promoter region in vitro, and the binding sequences corresponding to +8 to +22 are indispensable for the stable ternary complex. Figure 4.Cooperative binding of CmpBrp and CmTBP to the rDNA promoter region in vitro and identification of its site. (A) Purified recombinant CmTBP, CmpBrp and CmTFIIB. Proteins (0.2 μg each) were resolved on a 12.5% SDS–PAGE gel and stained with CBB. Positions of molecular size markers are indicated in kDa at left. (B) EMSA with the rDNA promoter region. Fragment containing promoter region of rDNA (−161 to +55, +1 as TSP) was incubated with (+) or without (−) indicated recombinant proteins. Arrowhead indicates specific ternary complex. (C) EMSA with the Ad2ML promoter region. Fragment containing promoter region of Ad2ML (−41 to +15) was incubated the same as in panel (B). (D) Search for CmpBrp–CmTBP binding region by EMSA competition analysis. EMSA analysis was performed under the same conditions as in panel (B), lane 4, without (−) or with indicated unlabelled probes at positions as shown at the top. Cold probe denotes unlabelled promoter region of rDNA (−161 to +55). (E) Identification of CmpBrp–CmTBP binding site by EMSA competition analysis. EMSA analysis was performed in the same way as in panel (D) without (−) or with indicated unlabelled probes that were mutagenized at 3-bp resolution, except for no. 7 probe. Sequences of mutagenized probes are shown at the top (lower characters indicate mutated nucleotides). (F) DNase I footprint analysis of rDNA promoter with CmpBrp–CmTBP binding site. The sense (top) and the antisense strands (bottom) were labelled at either end and used without (−) or with (+) CmpBrp and CmTBP in footprint reactions. Arrows and black boxes indicate the TSPs and the regions protected from DNase I digestion, respectively. (G) EMSA with the mutated rDNA promoter region within the CmpBrp–CmTBP binding site. Wild-type (Wt) or the mutated probe within the CmpBrp–CmTBP binding site (+8 to +22) (Mt) of the rDNA promoter region was incubated without (−) or with (+) CmpBrp and CmTBP. Download figure Download PowerPoint Involvement of Pol I and CmpBrp in the rDNA promoter transcription It is widely believed that the rDNA is transcribed by Pol I in eukaryotes, and it is natural to assume that this is also the case in C. merolae. To confirm this, we performed ChIP analysis with an antibody against CmRPA190 (Supplementary Figure S2), the largest subunit of Pol I, and found that the rDNA promoter region was specifically co-immunoprecipitated with CmRPA190 (Figure 5A). This result clearly indicated the involvement of Pol I in rRNA synthesis in C. merolae. Next, we examined whether CmpBrp and CmRPA190 coexisted in C. merolae cells by immunoprecipitation analysis. Our results indicated that endogenous CmRPA190 co-immunoprecipitated with CmpBrp and vice versa (Figure 5B, lanes 3 and 5). No co-immunoprecipitation of CmTFIIB and CmBRF with CmpBrp or CmRPA190 was observed, indicating that CmpBrp is a component of the general Pol I transcription machinery. Figure 5.CmpBrp positively contributes to Pol I-dependent rDNA transcription. (A) ChIP analysis with CmRPA190 antibody. Others are the same as shown in Figure 3E. (B) Co-immunoprecipitation of endogenous CmpBrp and CmRPA190. Immunoprecipitation was performed with antibodies against CmpBrp, CmRPA190 or relevant IgG (Pre), and co-immunoprecipitation was analysed by immunoblot analysis with each indicated antibody. As a control, 15% of cell lysate (input) was directly subjected to immunoblot analysis. (C) Specific in vitro transcription from the rDNA promoter. In vitro transcription was performed with C. merolae crude cell lysate and/or pUC119-RRN-Wt as a template (+, presence; −, absence). Schematic representation of pUC119-RRN-Wt is shown at the top. Arrow indicates the TSP of rDNA, whereas arrowhead indicates primer used for the primer extension analysis. Transcripts from rDNA promoter are marked at the right of the autoradiogram with an arrowhead. (D) Effect of antibody on in vitro transcription from the rDNA promoter. In vitro transcription analysis was performed under the same conditions as in panel C, lane 3, with (+) or without (−) indicated antibodies. (E) In vitro transcription analysis with the mutated rDNA promoter region within the CmpBrp–CmTBP binding site. In vitro transcription was performed with C. merolae crude cell lysate and pUC119-RRN-Wt (Wt) or pUC119-RRN-Mt (Mt). Hatched box indicates mutagenized region. Download figure Download PowerPoint Involvement of CmpBrp in rDNA transcription was further analysed by an in vitro transcription system using crude cell lysate prepared from C. merolae cells. When the rDNA promoter region was used for the assay as template DNA, an apparent transcript from the TSP of rDNA was detected (Figure 5C, lane 3). No band was observed in the absence of template DNA (lane 1) or cell lysate (lane 2), nor were any bands detected using the parental plasmid without the rDNA sequence (data not shown). These data indicated that the detected transcript was accurately derived from the input rDNA promoter. The in vitro transcription was not affected by addition of α-amanitin (Supplementary Figure S3), which is an inhibitor for Pol II (highly sensitive) and Pol III (slightly sensitive), to the reaction mixture. In vitro transcription was unaffected even at a high concentration of α-amanitin (250 μg/ml), at which concentration de novo syntheses of mRNA by Pol II (CMS045C and CMK028C) and 5S rRNA by Pol III were selectively inhibited (Supplementary Figure S3). Thus, we concluded that the rDNA in vitro transcription was dependent on Pol I. To examine whether CmpBrp is required for the Pol I-dependent transcription, we observed the effect of adding the CmpBrp antibody to the reaction. The result showed that the in vitro transcription was severely inhibited by addition of the CmpBrp antibody (Figure 5D, lane 1 versus 3). A similar inhibition was also observed by addition of CmRPA190 antibody (lane 1 versus 5). However, the reaction was not inhibited by addition of IgG purified from relevant preimmune serum, anti-CmTFIIB antibody, or anti-CmBRF antibody. Moreover, the in vitro transcripts almost disappeared when the DNA template with a mutated CmpBrp–CmTBP binding site was used (Figure 5E, lane 1 versus 2). Thus, these in vitro transcription experiments demonstrated that CmpBrp is indispensable for effective Pol I-dependent rDNA transcription, and the binding site of the CmpBrp–CmTBP complex is defined as an essential rDNA core promoter element. Intracellular localization of CmpBrp In eukaryotic cells, rRNA is synthesized in a specialized structure within the nucleus called the nucleolus. Pol I and its transcription-related factors have been consistently observed in the nucleolus area, in which rRNA transcription
Referência(s)