Fig 2: BMI1 inhibition delays CRPC progression in vivo. a, b Representation of combination index plot of enzalutamide combined with PTC209 in a AR-positive cells and b AR-negative cells. Doses below the dotted line represent the synergistic effect and doses above the dotted line represent the antagonistic effect. c Castration-resistant VCaP xenograft mouse models were generated as described in “Methods” section. Castrated mice carrying CRPC xenograft received vehicle or PTC209 (60 mg kg−1 per day) 5 days per week (n = 12 per group). Caliper measurements were taken every 4 days to obtain tumor volume. Mean tumor volume ± SEM, *P < 0.05 vs. Vehicle. d Tumor tissues were lysed and blotted for BMI1, AR, and GAPDH. The upper panel shows the representative western blot. Protein levels were quantified and normalized against GAPDH (lower panel), *P < 0.05 vs. Vehicle (mean ± SEM, n = 6). e Castrated mice carrying CRPC xenograft received enzalutamide (10 mg kg−1 per day), PTC209 (60 mg kg−1 per day), or PTC209 (60 mg kg−1 per day) + enzalutamide (10 mg kg−1 per day) 5 days per week (n = 6 per group). Caliper measurements were taken every 4 days to obtain tumor volume (e, left panel). Mean tumor volume ± SEM; e, right panel, Kaplan–Meier survival plot compares the progression-free survival; *P < 0.05, PTC209 + enzalutamide vs. enzalutamide; #P < 0.05, PTC209 + enzalutamide vs. PTC209. f Castrated mice carrying LuCaP 35CR, an enzalutamide-resistant and abiraterone-resistant PDX model, received enzalutamide (10 mg kg−1 per day) or PTC209 (60 mg kg−1 per day) + enzalutamide (10 mg kg−1 per day), 5 days per week (n = 9 per group). Caliper measurements were taken every 4 days to obtain tumor volume. f, left panel, mean tumor volume ± SEM; *P < 0.05. f, right panel, Kaplan–Meier survival plot compared with the progression-free survival; *P = 0.048

Fig 3: Phospho-mutant BMI1 is intrinsically unstable. a Determining the half-life of Wild-type (WT) and S110A mutant BMI1 (Mut-BMI1). The half-life of wild-type and mutant BMI1 in CP20 cells expressing the respective protein was determined by blocking synthesis of new proteins using 100 μg/ml CHX for different time points (0–45 min) and the rate of BMI1 degradation was determined by western blot analysis of FLAG(BMI1) with HSP60 serving as a loading control. Residual BMI1 protein after each time point was normalized to HSP60 using densitometry measurements. The time-dependent decrease of BMI1 was quantified to determine the half-life and is graphically represented in the right panel. The densitometry analysis was performed using Image J software (NIH, Bethesda, MD), and the graph was generated by plotting % residual protein vs time of CHX treatment. b Effect of proteosomal inhibition on Wild-type (WT) and S110A mutant BMI1 (Mut-BMI1). 30 h post transfection, CP20 cells were treated with 10 μM MG132, for different time points (0–120 min) to determine the accumulation of WT or mutant BMI1. Left panel represents the immunoblots for FLAG and HSP60 and quantification of the accumulated proteins by densitometry analysis of signal present in respective lanes (normalized to the individual HSP60) is graphically represented in the right panel. c Effect of proteosomal inhibition on BMI1 protein accumulation in CK2α silenced cells. CP20 cells were transfected with scrambled siRNA (siCTL) or siRNA against CK2α (siCK2α) and 48 h post transfection, cells were treated with 10 μM MG132, for different time points (0–120 min) to determine the effect of silencing CK2α on MG132 induced accumulation of endogenous BMI1. siCTL or siCK2α transfected cells were harvested after each time point and analyzed for accumulation of the endogenous BMI1 protein by western blot analysis and using HSP60 as a loading control (bottom left panel). Efficient silencing by CK2α siRNA was also determined (top left panel). The accumulated proteins were quantified by densitometry analysis of signal present in respective lanes and by normalizing it to the individual HSP60 and graphically represented in the right panel

Fig 4: BMI1 is a substrate of CK2α. a Co-precipitation of BMI1 with CK2α. Reciprocal immunoprecipitation (IP) assay was performed with BMI1 and CK2α antibodies crosslinked with the agarose resin and immunoprobed with CK2α and BMI1 antibodies respectively. b In vitro kinase assay with CK2α and BMI1. In vitro kinase assay was performed with 400nM CK2α, 200nM or 400nM BMI1-GST and radioactive ATP and representative autoradiograph image is presented.. Reaction mixture without substrate (lane 1), only GST protein (lane2), without enzyme (lane 3), and enzyme with GST only (lane 4) served as negative controls. c Kinase assay with immunoprecipitated (IP) endogenous CK2α and purified BMI1. CK2α was immunoprecipitated from CP20 cells using agarose A/G beads, the beads was washed and incubated in a kinase assay buffer supplemented with purified BMI1-GST and radiolabeled ATP, in presence or absence of the specific CK2 inhibitor TBB. A representative autoradiograph is provided in the right panel. Efficient Immunoprecipitation is demonstrated in the left panel by immunoblotting a small fraction of the IPed beads with CK2α antibody

Fig 5: Expression and regulation of BMI1 and CK2α in ovarian cancer cells. a Slow migrating phospho-BMI1 bands were reduced upon treatment with λ-Phosphatase but not with AKT inhibitor. OV90 and CP20 cells were treated with 5 μM AKT inhibitor (AKTi) or vehicle (C) for 15 h. Protein extracted from the vehicle treated cells (C) were further subjected to λ-phosphatase (λ-P) treatment or left untreated. Expression of BMI1 and Phospho-AKT (Serine 473) were determined in the C, λ-P and AKTi treated cells. α-Tubulin and AKT were used as a loading control. Slow migrating BMI1 bands that were removed on phosphatase treatment is considered as the phospho species of BMI1 and is indicated by the open arrow head whereas the basal BMI1 is indicated by the closed arrow head. b Expression of BMI1, CK2α, and AKT (S473) in a panel of ovarian cancer cells. Protein expression was determined in 4 cell lines representing high grade serous ovarian cancer (TYKNU, OVSAHO and OV90) or cisplatin resistant CP20 and compared to the normal Fallopian Tube Epithelial Cells (FTE188). α-Tubulin and AKT served as a loading. Open and closed arrow heads represent phospho species of BMI1 and basal BMI1 respectively. c-d Effect of CK2α knockdown on BMI1. c OV90 and CP20 cells were either transfected with scrambled siRNA (siCTL) or siRNA against the CK2α gene (siCK2α) and expression of BMI1, RING1A and CK2α were determined by immunoblotting. d Relative mRNA expression of BMI1 and CK2α were determined by RT-qPCR. Gene expression of 18 s rRNA was used as an endogenous control. Values represent mean fold change (± standard deviation) over control (siCTL). Open and closed arrow heads represent phospho species of BMI1 and basal BMI1 respectively. e Effect of CK2α overexpression on BMI1. FTE188 cells were transfected with CK2α plasmid and 24 h post transfection the expression of BMI1 and CK2α were determined by immunoblotting. Open and closed arrow heads represent phospho species of BMI1 and basal BMI1 respectively