NM1 localizes in nucleoli and is required for Pol I transcription. (A) Confocal sections of isolated nucleoli immunostained with antibodies to β-actin, NM1, UBF and fibrillarin (bar, 1 μm), as well as an anti-NM1 pre-immune serum (NM1 pre-serum; scale bar, 20 or 5 μm in the inset; supplementary information online). (B) Immunostaining and confocal microscopy showing colocalization of NM1 with green fluorescent protein (GFP)–RPA53 overexpressed in HeLa cells (scale bar, 5 μm). (C) Anti-NM1 antibodies inhibit Pol I transcription in vitro. A 10 μg fraction of a DEAE-280 fraction was pre-incubated for 30 min at 30°C with 0.5 or 1 μl of native or heat-treated (marked by asterisks) anti-NM1 antibodies (lanes 1–4) or rabbit IgGs (lanes 5–8) before transcription was started. The numbers below show the quantification of the transcriptional activities. (D) The Pol I-specific autoimmune serum S57299 inhibits Pol I transcription. HeLa nuclear extracts were pre-incubated for 20 min with increasing amounts of a control human serum (lanes 1–4) or S57299 (lanes 5–8) before transcriptional activity was assayed in the presence of template pHrP₂ linearized with EcoRI. A western blot showing that S57299 recognizes the largest subunit of Pol I (RPA194) is shown in Fig 3B.

NM1 is associated with WSTF–SNF2h. (A) Resolution of NM1-containing complexes by gel filtration. Proteins were fractionated according to the scheme on the left. Fractions of the Superose 6HR column were analysed on western blots. The position of the 670 kDa marker (thyroglobulin) is indicated. (B) Co-precipitation of NM1 with anti-WSTF antibodies from Superose 6HR fractions containing ∼2–3 MDa but not ∼1 MDa protein complexes. In lanes 1 and 4, 20% of the input material was loaded. (C) A complex containing NM1, WSTF and SNF2h is precipitated from nuclear extracts with anti-WSTF antibodies. 10% of the input is shown in lane 1. IP, immunoprecipitation. (D) Co-precipitation of NM1 and SNF2h from nuclear extracts. 10% of the input is shown in lane 1.

NM1, WSTF and SNF2h are associated with rDNA and promote 45S pre-rRNA synthesis. (A) Nucleolar colocalization of NM1 (scale bar, 1 μm), WSTF (scale bar, 1 μm) and SNF2h with Pol I (scale bar, 2 μm). (B) NM1, WSTF and SNF2h are associated with the transcription-competent subpopulation of Pol I. Proteins were crosslinked in vivo with dithiobis-succinimidyl-propionate (DSP) and Pol I was precipitated from cell lysates with anti-Pol I antiserum S57299. Co-precipitated NM1, WSTF, SNF2h, actin and RPA194 were detected on immunoblots. IP, immunoprecipitation. (C) Chromatin immunoprecipitation analysis. Lysates of crosslinked cells were incubated with the indicated antibodies and co-precipitated DNA was subjected to PCR using primers that amplify the rDNA promoter, the 18S rRNA coding region, ARPP P0 and tRNA genes. Lane 1 shows PCR amplification of 0.5% input chromatin (supplementary information online).

Both NM1 and the WSTF complex stimulate Pol I transcription. (A) Antibodies against NM1 inhibit run-off but not abortive Pol I transcription. A 20 μg portion of fractionated nuclear extract (DEAE-280 fraction) was pre-incubated for 45 min with rabbit IgGs, antibodies to NM1 or WSTF (2 μg each), or 0.1 μl of the anti-Pol I serum S57299 before being assayed for run-off transcription (upper panel) and the synthesis of ACU trimers (lower panel). (B) Anti-WSTF antibodies impair rDNA transcription on pre-assembled chromatin templates. Transcription assays contained circular pMrT₂ as naked DNA (lanes 1–5) or pre-assembled into chromatin (lanes 6–10). The DEAE-280 fraction was pre-incubated with rabbit IgGs (2 μg) or anti-WSTF antibodies (1 and 2 μg) before transcription was started by adding nucleotides, circular template DNA and recombinant TTF-I (10 ng) as indicated. Read-through and terminated transcripts are indicated; the asterisk marks unspecific transcripts that originate from traces of contaminating DNA in the Drosophila extract used for chromatin assembly. (C) RNA interference-mediated depletion of WSTF inhibits pre-rRNA synthesis. Cells were transfected with the empty vector (‘pSuper'), pSUPER-shWSTF (‘WSTF-shRNA') or an unrelated shRNA (‘control-shRNA'), and the level of the indicated RNAs was determined by reverse transcription–PCR. A representative experiment is shown on the top. The bar diagram below shows the relative amount of the indicated RNAs determined in five independent experiments. Error bars depict standard deviations estimated by Student's t-test. The significance for the WSTF RNAi knockdown is P=0.001 and that for 45S pre-rRNA is P=0.032.