Fig 2: RUVBL1/2 directly promote RPB1 CTD clustering and transcription initiation in vitro.a Schematic diagram showing the RPB1 CTD and RUVBL1/2 pull-down assays. b The complexes pulled down by beads were analyzed by western blotting. c RUVBL1/2 promoted RPB1 CTD clustering in vitro. The scale bar is 20 µm. d Plot showing the cluster area described in (c), clusters larger than 100 square pixels were counted (Mock n = 139, RUVBL2 in order n = 300, 71 and 39, respectively) from no less than 8 different fields for each condition. Two-tailed Student’s t tests were performed to determine the significance of differences between groups, and the error bars indicate the mean + /- SD (****p < 0.0001). e Plot showing the condensed fraction in the different fields per condition presented in (d). Two-tailed Student’s t tests were used to determine the significance and the data are reported as the mean + /- SD (*p < 0.05, ***p < 0.001). f Schematic diagram showing the in vitro Pol II transcription initiation system. g Nascent RNA synthesis by the preinitiation complex on the HDM2 promoter. This result represents 3 replicates. h In vitro droplet assay with GFP-EWS-FLI1 and RPB1 CTD-mCherry. i Boxplot showing the droplet area of mCherry signals. The total number of droplets examined in one in vitro droplet experiment was n = 127 under the mock condition and n = 197 under the RUVBL1/2 addition condition. The bottom edge of the box represents the 25th percentile, and the top represents the 75th percentile. Most data points are covered by whiskers (1.5x interquartile range). Two-tailed Student’s t tests were performed to determine the significance of differences between groups (****p < 0.0001). j Strategy for detecting EWS-FLI1 condensates at sites of DNA and RPB1 CTD recruitment by RUVBL1/273. k Wide-field total internal reflection microscopy (TIRFM) images of EWS-FLI1 and RPB1 CTDN26-mCherry in DNA Curtain experiments. l The efficiency of RPB1 CTDN26-mCherry recruitment as described in (j). Three independent DNA curtain experiments were repeated. The error bars refer to mean + /- SD. Statistical significance was evaluated by two-tailed Student’s t test (**p < 0.01); confidence level: 95%.

Fig 3: Identification of direct target genes of the RUVBL2-Pol II axis.a Workflow of the process used to identify the genes directly targeted by RUVBL2. The FDR are extracted from DESeq2 by adjusting p-values using BH method. b Heatmap analyses of RUVBL2 direct target genes based on PRO-Seq expression. c Snapshots showing the RUVBL1 and RUVBL2 ChIP-Seq signals and nascent RNA (PRO-Seq) signals at the Bmp4 and c-Myc loci after RUVBL2 degradation. d ChIP-Seq signals in the promoter regions of 45 genes directly targeted by RUVBL2 and 45 nontargeted genes in 153 datasets obtained from the Cistrome DB (Supplementary Data 5). e The 10 factors with the largest and smallest coefficients in the elastic net were selected. The heatmap shows the relative occupancy of these 10 factors in the 45 genes directly targeted by RUVBL2 and in 45 nontarget genes. f Pie chart indicating the percentage of C-MYC or CXXC1 ChIP-Seq-determined target genes that overlap the RUVBL2-affected genes as determined by the RNA-Seq data shown in Fig. 6g. g Profiles of PRO-Seq signals at the gene regions of directly targeted genes (n = 45), all genes (n = 36821) and posttranscriptional early response genes (n = 41) after RUVBL2 degradation. The profile above 0 indicates the signal on the sense strand, the profile below 0 indicates the signal on the antisense strand. Two-tailed Wilcoxon tests was performed to calculate the significance (*p < 0.05, **p < 0.01, ***p < 0.001). The center line of boxplot represents median, the box limits represent upper and lower quartiles, and the whiskers represents 1.5x interquartile range or maximum/minimum value. h The RUVBL2 direct target gene Bmp4 was examined by single-molecule RNA FISH combined with RPB1 immunofluorescence. Representative images are shown, scale bar is 3 µm. i The signals of the Pol ll (RPB1 IF) cluster that overlap with nascent RNA foci were measured, the boxplot indicated median (middle line), 25th and 75th percentile (box) and 5th and 95th percentile (whiskers), as well as outliers (single black points). Statistically analyzed (two-tailed Student’s t test) from at least 10 different fields of each condition (Auxin 0 h n = 61, Auxin 1 h n = 38). **p < 0.01.

Fig 4: RUVBL2 occupies gene promoters with Pol II and diverse transcription factors in cancer cells.a The list of transcription factor motifs that were enriched at the RUVBL1- and RUVBL2-overlapping peaks and the corresponding p-values (which are calculated using HOMER with cumulative binomial distributions) is shown on the right. The transcription factors in blue were previously reported to function with RUVBL1 or RUVBL2. b Cluster analysis with RUVBL1, RUVBL2, Pol II, C-MYC, E2F1, P53, CTNNB1, and OCT4 in RUVBL1/2-overlapping peaks. c Distribution pie charts showing the percentage of the corresponding transcription factor peaks in each category (K1 to K6) shown in Fig. b. d The heatmap illustrating RUVBL2 globally colocalization with various transcription factors in MCF-7 cells. e Metagene analyses of RUVBL2 ChIP-Seq data for the regions of genes directly targeted by RUVBL2, all genes and posttranscriptional early response genes. f The cancer hallmark genes directly targeted by RUVBL2 (red) and the posttranscriptional early response genes after RUVBL2 depletion (black). g Snapshots showing RUVBL2 ChIP-Seq signals at C-MYC loci in GM12878, K562, U2OS, HeLa, MCF-7, HCT116 and HepG2 cells. h The survival probability changes over time for patients with breast cancer and high (red line) or low (black line) expression of RUVBL1 (upper panel) and RUVBL2 (bottom panel) as determined by the curve constructed with the Kaplan–Meier plotter. Significance was calculated by log-rank method. HR indicates the hazard ratio, and the confidence interval is shown in brackets. i Model of RUVBL2-activated transcription of genes with diverse cellular functions and Pol II clustering. RUVBL1/2 were shown to function as global transcription coactivators with various transcription factors and to directly promote Pol II cluster formation mediated by the RPB1 CTD.

Fig 5: Knock-down of RUVBL1 by siRNA attenuated growth and metastasis of the OSCC cell line, H157, by inhibiting ß-catenin signaling. (a) Knock-down of RUVBL1 by siRNA suppressed H157 cell proliferation as indicated using the colony formation assay. The box plot showed the number of cells. (b) Knock-down of RUVBL1 by siRNA inhibited migration of the H157 cells in the wound healing assay. Significant differences between groups were assessed using a Student's T test and shown by * and #, p < 0.05. (c). After transfection of the H157 cells with RUVBL1 siRNA or scramble siRNA (siNC), cell invasive ability was determined by cell invasion assay. All experiments were performed in triplicate. Significant differences between groups were assessed using a Student's T test and shown by * and #, p < 0.05. (d) Expression of ß-catenin pathway proteins was detected by western blot analysis. Intensity of the gray band was calculated by ImageJ and shown in the figure.