Fig 2: Mutant p53 associates with MCM2, MCM4 and MCM7 on chromatin. a Protein levels of MCM2, MCM4, MCM7, mtp53 and fibrillarin in the chromatin fraction were determined by Western blot analysis in MDA-468.shp53 cells grown in the presence or absence of 8 µg/ml of doxycycline for 12 days, and HT-29 colon cancer cells transfected with p53-siRNA (p53) or control siRNA (Con). b Analysis of p53/MCM2 complexes (red) by immunofluorescence microscopy in combination with in situ proximity ligation assay (PLA) in MDA-468 vector and MDA-468.shp53 cells grown in the presence of 8 µg/ml of doxycycline for 12 days. DNA was counterstained with DAPI (blue) and GFP (green) was an indicator of doxycyline-mediated induction. The z stack confocal maximum intensity projection images of p53/MCM2 and DAPI, p53/MCM2 and GFP are shown. c Co-immunoprecipitation (co-IP) of MCM2 and MCM4 with exogenously expressed mtp53 (R175H) and wtp53 in H1299 cells. Whole cell lystates of H1299 cells transfected with wtp53 or mtp53 were incubated with anti-p53 antibody followed by immunoblot with anti-MCM2 or anti-MCM4 antibodies. d Co-IP of MCM4 and mtp53 in thymic lymphomas from mtp53 (Trp53R172H/R172H) mice. Thymic lymphomas from mtp53 mice and p53-/- mice as well as normal thymic tissue from mtp53 mice were subjected to co-IP assays using anti-p53 antibody followed by immunoblot with anti-MCM4 antibody

Fig 3: Systematic screening of human r-proteins reveals that uL5 (RPL11) and uL18 (RPL5) are the strongest contributors to nucleolar structure maintenance.(a) Experimental strategy: all 80 r-proteins were depleted one by one in human cells by use of specific siRNAs. The nucleolar structure (fluorescence microscopy), the accumulation of mature 18S and 28S rRNAs (electropherograms), pre-rRNA processing (high-resolution northern blotting), and steady-state accumulation of p53 (fluorescent western blotting) were monitored. (b) PCA showing a classification of r-proteins according to their requirement for nucleolar structure maintenance. Each r-protein was depleted in three knockdown experiments, each performed with a different siRNA. The image-processing algorithm that we designed for this analysis involves selecting five discriminant shape and textural features, computing five dk values, and reducing the five dimensions to two by PCA. In the resulting plot, each coloured dot represents one population of cells treated with one siRNA. Dot colour is indicative of the targeted protein: green for SSU r-proteins and magenta for LSU r-proteins. The mean of three populations of cells treated with a non-targeting control siRNA (SCR) is shown in red. Blue symbols represent the six calibration controls (FBL, GFP, nucleolin, nucleophosmin, MOCK and TIF1A, see Supplementary Fig. 1). Insets show images of the nuclei of cells depleted of representative proteins with the DNA stained in blue and the nucleoli appearing in green (FBL). For a few representative examples, a specific symbol is used (for example, a diamond for uL5). RPL, r-proteins of the LSU; RPS, r-proteins of the SSU. (c) r-proteins and calibration controls classified according to the severity of nucleolar disruption caused by their absence. The iNo was defined as the sum of the dk values of the five most discriminant shape and textural features identified in this work (Methods section). Higher iNo correspond to more severe disruption. Colour-coding as in b. The coloured dots are the means of three individual experiments (shown in grey). Note: the r-proteins are named according to a recently revised nomenclature24 where the ‘e' prefix stands for eukaryote-specific and ‘u' for universal (present in bacteria, archaea and eukaryotes).

Fig 4: Involvement of human r-proteins in p53 homeostasis.(a) Steady-state level of p53 determined by quantitative fluorescent western blotting. Western blots analysis are shown for representative r-proteins, with the p53 level indicated underneath as a mean of biological triplicates obtained after treatment of cells with the same siRNA (i, ii and iii). The siRNA used was selected on the basis of its proven efficacy in the processing and nucleolar screens (Figs 1 and 4). The p53 signal corrected for loading (using ß-actin as reference) was expressed with respect to the level observed in cells treated with a non-targeting siRNA control (p53+/+). Red signal, p53; green signal, ß-actin. A complete data set for all 80 r-proteins is available at www.RibosomalProteins.com and in Supplementary Fig. 11. As loading control we used HCT116 p53+/+ cells transfected with a non-targeting siRNA (p53+/+) providing the basal level of p53 or with an antisense oligonucleotide suppressing the activity of the box C/D snoRNA U8 (#U8), thereby stimulating p53 accumulation up to sixfold (D.L.J.L. submitted). As background control, we used a matched isogenic HCT116 cell line that does not express p53 (HCT116 p53-/-, (ref. 39)) treated with a non-targeting siRNA (p53-/-). (b) r-proteins classified according to their impact on the p53 steady-state level. The non-targeting (SCR) control is shown in red, the SSU r-proteins in green, and the LSU r-proteins in magenta. The histogram bars are the means of triplicates with s.d. r-proteins whose depletion leads to a fivefold increase in p53 level are highlighted in a grey box.

Fig 5: Distribution of mutant p53 association index (mPAI) scores in the cytosol (top) and chromatin (bottom) fractionations. Histograms of mPAI values (for equation see statistics in the methods) show close fit to normal curves (in dashed lines) obtained with the same mean (µ = 0.0060 for cytosol, µ= 0.085 for chromatin) and standard deviation (s = 0.92 for cytosol, s = 0.96 for chromatin). As expected (and as a negative control of the experiments and the mPAI statistic), most proteins are not significantly impacted by mtp53 depletion, showing -1 < mPAI < 1 (shaded in light gray). As a positive control, the mTP53 protein level (in red) shows significantly high positive mPAI values. As another positive control, the PARP1 levels (in blue) show contrasting mPAI values between the two fractionations, consistent with previous experimental results.6 The six components of the MCM2–7 complex (in green) show significant positive association with mtp53 in both cytosol and nucleus