Function and assembly of a chromatin-associated RNase P that is required for efficient transcription by RNA polymerase I

Abstract

Background: Human RNase P has been initially described as a tRNA processing enzyme, consisting of H1 RNA and at least ten distinct protein subunits. Recent findings, however, indicate that this catalytic ribonucleoprotein is also required for transcription of small noncoding RNA genes by RNA polymerase III (Pol III). Notably, subunits of human RNase P are localized in the nucleolus, thus raising the possibility that this ribonucleoprotein complex is implicated in transcription of rRNA genes by Pol I.

Methodology/principal findings: By using biochemical and reverse genetic means we show here that human RNase P is required for efficient transcription of rDNA by Pol I. Thus, inactivation of RNase P by targeting its protein subunits for destruction by RNA interference or its H1 RNA moiety for specific cleavage causes marked reduction in transcription of rDNA by Pol I. However, RNase P restores Pol I transcription in a defined reconstitution system. Nuclear run on assays reveal that inactivation of RNase P reduces the level of nascent transcription by Pol I, and more considerably that of Pol III. Moreover, RNase P copurifies and associates with components of Pol I and its transcription factors and binds to chromatin of the promoter and coding region of rDNA. Strikingly, RNase P detaches from transcriptionally inactive rDNA in mitosis and reassociates with it at G1 phase through a dynamic and stepwise assembly process that is correlated with renewal of transcription.

Conclusions/significance: Our findings reveal that RNase P activates transcription of rDNA by Pol I through a novel assembly process and that this catalytic ribonucleoprotein determines the transcription output of Pol I and Pol III, two functionally coordinated transcription machineries.

Figure 1. Knockdown of Rpp29 by siRNA causes inhibition of Pol I transcription.

A. HeLa cells were transfected with siRNA29 or luciferase siRNA (see Materials and Methods), whole extracts were then prepared from cells at 24, 48 and 72 h after transfection and examined by Western blot analysis for Rpp29 and β-actin. Positions of proteins are depicted. B. A scheme of the human mini-rDNA gene. This construct has a rDNA gene promoter (from position −500) and 5′ETS (position +382) that was subcloned in pBluescript in which a rDNA terminator region, consisting of 57-bp 3′ETS, Sal I box and 51-bp IGS, was inserted. The 5′ETS is 382 bp in length and harbors no known cleavage site, including the A′ site. Arrow points to the transcription initiation site. C. Transcription of the mini-rRNA and 5S rRNA genes in extracts described in A. Labeled RNA was separated in 8% polyacrylamide/7 M urea gel and subjected to autoradiography. Lanes 1–7 represent exposure for longer time than that of lanes 8–13. Positions of the 432-nt mini-rRNA (lanes 2–7) and 120-nt 5S rRNA (lanes 8–13) transcripts are shown. Asterisk represents endogenous but unknown labeled RNA. D. The optical density (in arbitrary units) of the mini-rRNA band seen in lanes 2–7 of panel C. E. HeLa cells were treated with siRNA29 or luciferase siRNA for 48 h and knockdown of Rpp29 was tested by Western blot analysis (left panel). GAPDH was used as internal control. Total RNA was extracted from cells and subjected to RT-PCR analysis of 5′ETS and U1 snRNA (see Materials and Methods). As negative control, RNA was treated with RNase A before reverse transcription of the 5′ETS (right panel). PCR products were separated in agarose gels.

Figure 2. Knockdown of the specific subunit Rpp21 inhibits transcription and pre-rRNA synthesis.

A. HeLa cells were transfected with a plasmid expressing EGS^Rpp21 or siRNA21 or mocked transfected for 48 h. Whole HeLa extracts were then prepared and tested for the presence of Rpp21 and β-actin by Western blot analysis. B. Transcription of the mini-rRNA and 5S rRNA genes in extracts described in A and the reaction products were analyzed as in Figure 1C. C and D. The bands of the mini-rRNA and 5S rRNA were quantitated and plotted. E. Nuclear run-on transcription analysis (see Material and Methods) of the human 5′ETS rRNA and 5S rRNA genes in mock-, EGS^Rpp21- or siRNA21-treated HeLa cells. Letter P indicates pBluescript background.

Figure 3. Pol I transcription deficiency can be restored by a recombinant Rpp25 protein.

A. HeLa cells were transfected with a plasmid expressing siRNA25 or mock transfected (see Materials and Methods), whole extracts were then prepared from cells at 48 and 72 h after transfection and examined for Rpp25, Rpp20 and β-actin by Western blot analysis. Positions of proteins are depicted. B. Same as in A but transfection was for 72 and 96 h and proteins tested were Rpp25, Rpp20 and Rpp21. C. Reconstitution assays of transcription of the mini-rRNA gene in extracts described in A were performed in the absence or presence of recombinant Rpp25 protein. The resulted 432-nt mini-rRNA was analyzed as in Figure 1C. Exposure time was for 96 h. D. Reconstitution of a human 5S rRNA gene transcription in extracts described in B in the absence or presence of recombinant Rpp25 protein. As controls, recombinant Rpp20 and Rpp21 proteins were used. The resulted 120-nt 5S rRNA was analyzed as in Figure 1C. Asterisk indicates endogenous tRNA labeled with radioactive nucleotide. E. The 432-nt mini-rRNA bands seen in C were quantitated and the optical density (in arbitrary units) was plotted. F. The 120-nt 5S rRNA bands seen in D were quantitated and the optical density (in arbitrary units) was plotted. G. Coomassie blue staining of purified, histidine-tagged rRpp20, rRpp21 and rRpp25 proteins analyzed in 12% SDS/PAGE. Positions of the protein size markers (M) are shown. H. Same as in B but transfection time points were 24, 48 and 72 h. I. Transcription assays of Pol I in the absence and presence of recombinant Rpp25 protein were done as described in C. The 432-nt mini-rRNA band was quantitated and the optical density (in arbitrary units) was plotted. Reconstitution assays described here have been repeated 3 times with different time points and produced similar results.

Figure 4. H1 RNA and protein subunits of RNase P are required for Pol I transcription in extracts.

A. Whole HeLa extracts were subjected to immunodepletion analysis using 200 µL of serum containing polyclonal rabbit antibodies against Rpp20, Rpp25 or p53. Transcription reactions in the immunodepleted extracts (grey bars) or in extracts reconstituted with their corresponding immunoprecipitates (black bars) were carried out using the mini-rDNA gene and labeled RNAs were analyzed as described in Figure 1C. The 432-nt mini-rRNA band was quantitated and the optical density (in arbitrary units) was plotted. B. A proposed secondary structure of H1 RNA and the nucleotide sequence against which the antisense H1-1 deoxyoligonucleotide was directed. The upper half of H1 RNA represents the specificity domain. Conserved domains, including the P4 pseudoknot in the lower (catalytic) domain are shown. C. Whole HeLa extracts (15 mg/ml) were incubated with 8 µg of H1-1 (lane 3) or scrambled H1-1sc (lane 4) deoxyoligonucleotide in the presence of RNase H for 45 min as described . Extracts were then assayed for RNase P activity in processing of ³²P-precursor tRNA^Tyr, and cleavage products were analyzed in an 8% sequencing gel. The 5′ leader sequence (5′) and shorter species (arrow head) generated as a result of substrate miscleavage, are indicated. A concentrated DEAE-purified RNase P preparation (Ctrl; lane 2) was used as control for the correct cleavage of the substrate. D. Whole HeLa extracts described in C were subjected to transcription of the mini-rDNA gene and 5S rRNA genes as described in Figure 1C. E. Optical density of the mini-rRNA and 5S rRNA bands seen in panel D.

Figure 5. Pol I components copurify and associate with human RNase P.

A. A DEAE-purified RNase P was loaded (L, load) into a Sephacryl S-100 gel filtration column and eluted fractions were assayed for RNase P activity in processing of ³²P-labeled precursor tRNA^Tyr (S) to mature tRNA (3′) and 5′ leader sequence (5′). Substrate cleavage was analyzed as in Figure 4C with control HeLa RNase P (Ctrl). B. Western blot analysis of Sephacryl S-100 eluted fractions using antibodies against RPA194, RPA43, TAF_I110, UBF, Rpp29 and Rpp25. Preparations of DEAE-purified RNase P (first lane) was used as control for input. Proteins were separated in 12% SDS/PAGE and the positions of the corresponding proteins and the protein size markers are shown. C. Fraction F28 described in A was subjected to immunoprecipitation analysis using antibodies against Rpp29 (lane 3), RPA43 (lane 4), UBF (lane 5), TAF_I110 (lane 6), RPB8 (lane 7) and IgG (lane 8) in the presence of 150 mM KCl (IP150). For Rpp29 and RPA43 immunoprecipitation was also performed at 300 mM KCl (lanes 9 and 10). Immunoprecipitates obtained were tested for RNase P activity in processing of a ³²P-labeled precursor tRNA^Tyr (S). RNase P in fraction F28 was tested as control for enzyme activity (lane 2) and cleavage products were analyzed as in A.

Figure 6. Protein subunits of RNase P occupy the promoter and coding region of rDNA.

A. Structure of a rRNA gene, showing 18S, 5.8S and 28S rRNA flanked by 5′ETS and 3′ETS and intervened by ITS1 and ITS2. RNase MRP cleaves at the A3 site in the ITS1. Small arrows indicate locations of primers used for PCR. B. ChIP was performed with rapidly dividing HeLa cells using affinity-purified polyclonal antibodies against Rpp20 (lane 2), Rpp21 (lane 3), Rpp29 (lane 4), Rpp38 (lane 5), RPB8 (lane 6), preimmune serum (lane 7) or p53 (lane 8). DNA of the precipitated chromatin was analyzed by PCR for the presence of 5.8S rDNA, tRNA^Tyr, U1 snRNA and ARPP P0 genes. Positive PCR products for each gene using 0.01, 0.1 and 1% of input DNA are shown (lanes 9–11). This panel is a composite of two mini-agarose gels with the same exposure time. C. HeLa cells were transfected with siRNA25 (lanes 1–6), which caused marked knockdown of Rpp25 (data not shown), or with control luciferase siRNA (lanes 7–12) for 48 h. ChIP analysis was then performed for 5.8S rDNA and U1 snRNA gene using antibodies against Rpp20, Rpp25, Rpp29, Rpp38, RPB8 and preimmune serum. Positive PCR reactions using 0.01, 0.1 and 1% of input DNA are shown (lanes 13–18). D. ChIP analysis was performed with cells as in B using antibodies against Rpp20 (lane 1), Rpp21 (lane 2), Rpp29 (lane 3), Rpp30 (lane 4), RPA194 (lane 5), preimmune serum (lane 6) or p53 (lane 7). DNA of the precipitated chromatin was analyzed by PCR for the presence of the promoter, 18S, 5.8S, 28S or IGS DNA using the corresponding primers shown in A. Positive PCR reactions for each rDNA region using 0.01, 0.1 and 1% of input DNA are shown (lanes 8–10).

Figure 7. Protein subunits of RNase P dissociate from rDNA loci in mitotic cells.

A. HeLa cells were arrested at G2/M by treatment with nocodazole for 16 h. Mitotic cells were separated from adhered G2 cells as described . ChIP was performed with G2 (G2) and mitotic (M) cells using antibodies against Rpp21 (lanes 2 and 7), Rpp29 (lanes 3 and 8), RPB8 (lanes 4 and 9), preimmune serum (lanes 5 and 10) or p53 (lanes 6 and 11). Detection of 5.8S rDNA and 18S rDNA was done as in Figure 6D. Positive PCR reactions using 0.01, 0.1 and 1% of input DNA are shown (lanes 12–17). The PCR signal of Rpp29 seen in mitotic cells (lane 8) was not seen in two other independent experiments (data not shown) and hence its presence may be related to incomplete separation of M and G2 cells. B. As in A but Rpp30, Rpp38, Rpp40, RPB8 and preimmune serum were tested in ChIP. The U1 snRNA gene that is transcribed by Pol II was tested as control. C. FACS analysis of cells described in C. Asynchronous HeLa cells were used as control. D. Western blot analysis of whole extracts prepared from G2 and M cells described in C by using affinity-purified antibodies directed against the indicated protein subunits of human RNase P. Rpp14 and hPop5 were difficult to detect (data not shown). Mitotic cyclin B1 and β-actin were analyzed as control.

Figure 8. Reassociation of protein subunits of RNase P with chromatin of target genes.

A. HeLa cells were synchronized by nocodazole for 16 h before mitotic cells were separated from adhered G2 cells. Mitotic cells were released from the inhibitor for 2, 6 and 14 h. Cells were at early G1, mid G1 and S phase as determined by flow cytometry analysis (data not shown). ChIP was then performed with synchronized cells using antibodies against Rpp20 (lanes 1, 6 and 10), Rpp25 (lanes 2, 7 and 11), Rpp29 (lanes 3, 8 and 12), RPB6 (lanes 4, 9 and 13) or control IgG antibody (lane 5). DNA extracted from the precipitated chromatin was analyzed by PCR for the presence of 5.8S rDNA, 5S rDNA and tRNA^Leu genes. Linear PCR reactions using 0.01, 0.1 and 1% of input DNA are shown (lanes 14–22). B. HeLa cells were synchronized as described in A and at the indicated time points after release from nocodazole they were subjected to FACS analysis. C. Whole extracts obtained from HeLa cells described in B were examined for transcription of a human tRNA^Tyr gene as previously described . The resulted primary transcript is 112 nt in length and has an intron that is promptly excised to release the exons with 5′ leader and 3′ trailer sequences (42 and 50 nt in length, respectively). The exons are ligated to produce a 92-nt intermediate precursor tRNA^Tyr, which is processed by RNase P to remove the 5′ leader sequence (5 nt) and by a 3′ endonuclease that eliminates the 3′ trailer sequence (11 nt) to generate a mature 76-nt tRNA^Tyr. D. Whole HeLa extracts described in B were assayed for 5S rRNA gene transcription. E. HeLa cells were synchronized to G1/S phase by treatment with hydroxyurea for 16 h. Cells were then released from inhibitor and analyzed by FACS for their DNA content at the indicated time points. Cells were at mitosis at 12 h. F. Western blot analysis of whole extracts obtained from cells described in E using antibodies against Rpp29, Rpp25, Rpp20 and Rpp40 and cyclin B1. Cell cycle progression shown on the top was deduced from the FACS data shown in E and from the levels of cyclin B1. Similar results were obtained in three independent synchronization experiments.