The Treacher Collins syndrome (TCOF1) gene product is involved in ribosomal DNA gene transcription by interacting with upstream binding factor

Abstract

Treacher Collins syndrome (TCS) is an autosomal dominant disorder characterized by an abnormality of craniofacial development that arises during early embryogenesis. TCS is caused by mutations in the gene TCOF1, which encodes the nucleolar phosphoprotein treacle. Even though the genetic alterations causing TCS have been uncovered, the mechanism underlying its pathogenesis and the function of treacle remain unknown. Here, we show that treacle is involved in ribosomal DNA gene transcription by interacting with upstream binding factor (UBF). Immunofluorescence labeling shows treacle and UBF colocalize to specific nucleolar organizer regions and cosegregate within nucleolar caps of actinomycin d-treated HeLa cells. Biochemical analysis shows the association of treacle and UBF with chromatin. Immunoprecipitation and the yeast two-hybrid system both suggest physical interaction of the two nucleolar phosphoproteins. Down-regulation of treacle expression using specific short interfering RNA results in inhibition of ribosomal DNA transcription and cell growth. A similar correlation is observed in Tcof(+/-) mouse embryos that exhibit craniofacial defects and growth retardation. Thus, treacle haploinsufficiency in TCS patients might result in abnormal development caused by inadequate ribosomal RNA production in the prefusion neural folds during the early stages of embryogenesis. The elucidation of a physiological function of treacle provides important information of relevance to the molecular dissection of the biochemical pathology of TCS.

Fig. 1.

Cellular colocalization of treacle and UBF. (A) HeLa cells were indirectly stained with antibodies against treacle (Ab 014, red) and other antinucleolar proteins (green). Nuclear DNA was visualized by using Hoechst stain. Mitotic and interphase cells are marked M and I, respectively. (B) Treacle colocalizes with UBF at various stages of mitosis. HeLa cells were treated with 0.1 μg/ml nocodazole for 4 h, incubated in fresh medium overnight, then analyzed by indirect immunofluorescence. (C) HeLa cells were blocked in mitosis with 0.1 μg/ml colchicine for 6 h and analyzed by immunofluorescence labeling of chromosome spreads by using Hoechst stain (blue), anti-treacle (green), and anti-UBF (red) antibodies. (D) In a separate experiment HeLa cells were treated with 0.1 μg/ml colchicine for 14 h, and mitotic cells were harvested by mechanical shock. Chromatin (Chr) and cytoplasmic (Cyt) extracts were prepared as described (27) and used for Western blot analysis. Cell extract (CE) refers to whole mitotic cells boiled in Laemmli buffer.

Fig. 2.

Effects of actinomycin d on the localization of nucleolar proteins. HeLa cells were treated with 50 ng/ml actinomycin d for 2 h and stained by indirect immunofluorescence. Nuclei with dark-phase nucleoli are shown. (A) Treacle (red) localization in untreated and treated cells is compared with RNA helicase II/Guα (green). (B) The localizations of full-length treacle and its C terminus both fused to GFP (green) are compared with endogenous UBF (red) in actinomycin d-treated cells. (C) Enlarged view of a superimposed phase and GFP fluorescence of actinomycin d-treated HeLa cell transfected with full-length treacle–GFP construct.

Fig. 3.

Direct interaction between treacle and UBF. (A) HeLa cells were transfected with double-FLAG-tagged treacle (amino acids 717-1488) expression construct and analyzed by immunofluorescence staining using anti-FLAG and anti-Guα antibodies. Similar transfected cells were used for immunoprecipitation analysis by tumbling nuclear extract with anti-FLAG antibody-agarose. Protein samples were analyzed by Western blot enhanced chemiluminescence using anti-UBF antibody. The membrane was stripped and analyzed by using anti-Guα antibody. (B) Analysis of the treacle–UBF interaction by the yeast two-hybrid system in a double drop-out medium (-Trp, -Leu) that permits growth of cells transfected with pGBKT7 and pGADT7 constructs, and in a triple drop-out medium (-Trp, -Leu, -His) that screens for protein–protein interaction. The interaction of p53 and simian virus 40 large T antigen was used as a positive control.

Fig. 4.

Effects of siRNA-mediated down-regulation of treacle on rRNA production. (A) HeLa cells were treated with 40 nM siRNA or mock-transfected. si1011 targets the TCOF1 mRNA and si934Scr is a control siRNA. After 48 h of transfection, total RNA was isolated and analyzed by RT-PCR using TCOF1-specific primers and U1C primers as an internal control. In a separate 4-day transfection, cells were boiled in Laemmli buffer and analyzed by Western blot enhanced chemiluminescence using anti-treacle antibody. The blot was stripped and reprobed with anti-Guα antibody. Treacle comigrates at ≈220 kDa near the top of the gel, whereas RNA helicase II/Guα comigrates near the bottom of the gel at ≈90 kDa. (B) An RNase protection assay shows a 47% decrease in the level of pre-rRNA in cells treated with si1011 compared to mock-transfected cells. Lane 1 contains 1/10 of the riboprobe used in the other lanes. Lanes 2–5 contain 8 μg yeast RNA or total RNA isolated from cells mock-treated or treated with the indicated siRNA. (C) Total RNAs (1.0 μg) from ³²P-metabolically labeled transfected cells were analyzed by gel electrophoresis. The chase times with nonradioactive phosphate are labeled 0, 0.5, and 1 h. The numbers below correspond to the amount of specific RNA band relative to mock-transfected samples (set at 100) obtained with image-quant software. (D) The same blot was stained with 0.2% methylene blue to visualize total RNA. (E) BrUTP incorporation assay. HeLa cells grown on a chamber slide were transfected twice with 40 nM si1011 (24 h apart) and analyzed for BrUTP incorporation 72 h after the last transfection. Mock- and si1011-transfected cells were indirectly stained with anti-treacle (green) and anti-BrdUrd (red) antibodies. All pictures were taken at a preset brightness, contrast, and gamma level of 50, 50, and 1.0, respectively, and image scale range of 100–3,000. (F) Fluorescence intensity profiles of the images shown in E.

Fig. 5.

Analysis of RNA from mouse embryos. RT-PCR was used to determine the levels of expression of Tcof1 in embryonic day 10.5 heterozygous (+/-) embryo relative to its WT (+/+) littermate. (Left) Tcof1 signals were normalized by using U1C as an internal control. (Right) The level of pre-rRNA was determined by RNase protection assay using a 5′ pre-rRNA probe and β-actin as an internal control. Similar results were obtained with another set of embryos.