Fig 2: ATRX is required for maintenance of HP1? SAHF in senescent cells. a LS8817 cells were treated as described. b–d Cells were then fixed and the number of ATRX foci (b), the accumulation of SA-ß-gal-positive cells (c), and the accumulation of four mRNAs of the SASP cytokine program (d) were measured. e Clonogenic growth was assessed after cells were replated at low density the absence of drug. Colony number is not quantified as individual colonies could not be discerned. f The accumulation of HP1?-positive cells (SAHF) was determined. Representative images are shown (left) and the percentage of SAHF-positive cells was quantified (right). g HRAS mRNA levels were analyzed by qPCR. All data are represented as mean ± SEM from at least three independent experiments. Asterisk indicates p < 0.05 using a two-sided Students t-test. All images were taken at the same magnification. The scale bar is 20 microns

Fig 3: Attenuation of p53 chromatin binding in DAXX_KO and ATRX KO U87 cells.a Distribution of p53 peaks within gene context. b Downregulated (blue) and upregulated (red) p53 ChIP-seq peaks in DAXX KO relative to WT cells (y axis) as a function of p53 ChIP-seq intensity (x axis). There is a trend of p53 being downregulated with only few upregulated peaks that are of lower ChIP-seq signal. c Mean p53 signal profile comparison in DAXX KO vs WT condition among the set of downregulated peaks. d Box plot of the distribution of p53 signal between DAXX KO and WT conditions among the set of downregulated peaks. Whiskers represent maximum and minimum values and box represent upper quartile, median and lower quartile. e ChIP-seq tracks for p53 in control (blue) or DAXX_KO (red) at gene loci for CDKN1A (top) and GADD45A (middle). f Consensus motifs for p53 ChIP-seq sites with most differential binding between DAXX_KO and WT (top) and for overlapping sites with known factors other than p53. g ChIP-qPCR for p53 in DAXX_KO and WT control cells treated with DMSO or 50 µM etoposide (red) for binding sites at the p21 promoter, GADD45A, 18q subtelomere, 13q subtelomere, TMCC1 or CCDC170 gene loci. ChIP-qPCR with IgG control is shown in lower panels. h Same as in panel g, except for ATRX_KO cells. Error bars are sdm and P valued determined by student t test for three biological replicates.

Fig 4: ATRX accumulates in nuclear foci in senescent cells that colocalize with HP1? SAHF. a LS8817 cells were treated with PD0332991 for the indicated number of days, fixed and subsequently stained with ATRX antibodies. The number of ATRX foci in each cell (>150 cells per graph) were counted and plotted. The percentage of cells with a number of foci is indicated on the y-axis. b LS8817 cells were treated with PD0332991 for 7 days and the co-localization of HP1? (left) or PML foci (right) with ATRX foci determined by immunofluorescence. Representative images are shown. The fraction of ATRX foci co-localized with either HP1? or PML foci was quantified in control and treated cells. Circles in the Venn diagrams are drawn to scale relative to the number of the indicated foci in each condition. c The indicated liposarcoma cell lines were treated with PD0332991 for 7 days as previously described35. The outcome of this treatment vis a vis quiescence or senescence is indicated. ATRX foci were detected by immunofluorescence and the average number of foci per cell is plotted (right). Representative images are shown on the left. All data are represented as mean ± SEM from at least 3 independent experiments. Asterisk indicates p < 0.05 using a two-sided Students t-test. All images were taken at the same magnification. The scale bar is 20 microns

Fig 5: ATRX is necessary for doxorubicin-induced senescence. LS8817 cells were transduced with either a scrambled (shSCR) or ATRX-specific (shATRX) lentiviral knockdown vector and subsequently treated with 100 nM doxorubicin for 7 days. a Extracts were made from the cells indicated above each lane and the expression of proteins determined by immunoblot as indicated on the left of each panel. ATRX was detected with the D-5 antibody. b Cells were treated as described on the left and stained with the antibodies indicated on top of each panel. Representative images are shown (top) and quantification of the foci is plotted (bottom). c The indicated cells were plated on day 0 and either treated with 100 nM doxorubicin (doxo) or left untreated (CTRL). Cell number was counted at the indicated days and the relative number of cells is plotted. d, e The accumulation of SA-ß-gal-positive cells (d) or SAHF-positive cells (e) for each individual treatment condition was measured. Representative SAHF images are shown (below). f The accumulation of four of the liposarcoma SASP transcripts was measured by qPCR under each indicated condition and their induction in doxorubicin treated cells relative to that in untreated cells is plotted and compared. g The cells were treated with 100 nM doxorubicin (7D doxo) or left untreated (CTRL) for 7 days. Cells were then collected, counted and replated at low density in the absence of drug. Clonogenic growth was measured by crystal violet staining 3 weeks later. Representative images from two independent experiments are shown. Colony number is not quantified as individual colonies could not be discerned. All data are plotted as mean ± SEM from at least three independent experiments. Asterisk indicates p < 0.05 using a two-sided Student’s t-test. All images were taken at the same magnification. The scale bar is 20 microns. See also Supplementary Fig. 10 for uncropped blots