Fig 2: Working model of RBBP6 in regulating GSCs.RBBP6 regulates the maintenance of GSCs. Depletion of RBBP6 leads to the elimination of K63-linked ubiquitination of CPSF3, resulting in the destabilization of CPSF3 and the subsequent shortening of the 3’ UTRs of MYC ceRNAs. This shortening promotes the release of miR-590-3p from the truncated UTRs, thereby reducing MYC expression and hindering the survival and progression of GSCs.

Fig 3: APA interacted with pre-mRNA 3’ end processing factors.a GO enrichment analysis of the RBBP6-interacting proteins identified by IP-MS. b Heatmap showing that RBBP6 interacted with pre-mRNA 3’ end processing factors identified by IP-MS. c HEK293T cells were transfected with HA-RBBP6 plasmids for 48 h and were collected for IP/IB. IB, immunoblotting. d The number of the pre-mRNA 3’ end processing factors shortening genes that overlap with RBBP6 shortening genes follows a hypergeometric distribution. The y-axis is the signed Log10 P-values of the hypergeometric test P-values. e qPCR results of dPAS usage for APA analysis in GSC468 and GSC3565 cells transduced with two separate sgRNAs targeting CPSF3 or sgNT. Statistical significance was assessed using an ordinary one-way ANOVA with Dunnett’s multiple comparisons test, **P < 0.01; *** P < 0.001; **** P < 0.0001, n = 3. f The protein level of MYC was measured in GSC468 and GSC3565 cells transduced with two separate sgRNAs targeting CPSF3 or sgNT. g HEK293T cells transfected with Flag-CPSF3, Myc-UB, HA-RBBP6 and HA-RBBP6-ΔRING plasmids were collected for IP/IB. h HEK293T cells transfected with Flag-CPSF3, Myc-RBBP6, and HA-UB mutant plasmids were collected for IP/IB. i HEK293T cells transfected with Flag-CPSF3, Myc-RBBP6, and HA-UB mutant plasmids were collected for IP/IB. j HEK293T cells transfected with HA-RBBP6, Myc-UB, and Flag-CPSF3 mutant plasmids were collected for IP/IB. k MS spectrum of an identified polyubiquitinated CPSF3 peptide. l The protein level of CPSF3 was measured in GSC468 and GSC3565 cells transduced with two separate sgRNAs targeting RBBP6 or sgNT. m The GSC468 cells transduced with a sgRNA targeting RBBP6 or sgNT followed by CHX (50 µg/mL) treatment for the indicated times were collected for IB. n Quantitation of the results shown in m. Two-way ANOVA was used for statistical analysis with Sidak’s multiple comparisons test. *P < 0.05, n = 3.

Fig 4: CRISPR/Cas9 screens revealed novel regulators of GSCs.a Workflow of the CRISPR screens to identify potential targets for glioblastoma therapy. b Dot plot represents the CRISPR screen results. c Venn diagram showing the overlap between the in vivo CRISPR screen and in vitro CRISPR screen results in GSCs. d RBBP6 mRNA expression levels in the TCGA GBM dataset (U133A). Statistical significance was assessed using a t-test. e The mRNA level of RBBP6 in GSCs and matched DGCs. Statistical significance was assessed using a t-test, ***P < 0.001, n = 3. f RBBP6 mRNA expression level in a panel of 44 GSCs and 9 NSCs. Statistical significance was assessed using a t-test. g H3K27ac signal at the RBBP6 locus in three GSCs, three DGCs and four NSCs.

Fig 5: Targeting RBBP6 and CPSF3 inhibitd GSC-derived xenograft growth in vivo.a Kaplan‒Meier survival curves of immunocompromised NSG mice bearing intracranial GSC468 cells transduced with sgNT, sgRBBP6#1, or sgRBBP6#7 (Top). Representative images of H&E staining of mouse brains. Brains were isolated after the presentation of the first neurological sign in any cohort (Bottom). b Kaplan‒Meier survival curves of immunocompromised NSG mice bearing intracranial GSC3565 cells transduced with sgNT, sgRBBP6#1, or sgRBBP6#7 (Top). Representative images of H&E staining of mouse brains (Bottom). Brains were isolated after the presentation of the first neurological sign in any cohort. c Kaplan‒Meier survival curves of immunocompromised NSG mice bearing intracranial GSC468 cells transduced with sgNT, sgCPSF3#1, or sgCPSF3#2 (Top). Representative images of H&E staining of mouse brains (Bottom). Brains were isolated after the presentation of the first neurological sign in any cohort. d Kaplan‒Meier survival curves of immunocompromised NSG mice bearing intracranial GSC3565 cells transduced with sgNT, sgCPSF3#1, or sgCPSF3#2 (Top). Representative images of H&E staining of mouse brains (Bottom). Brains were isolated after the presentation of the first neurological sign in any cohort. Scale bars: 5 mm. Log-rank test was used for statistical analysis (a–d). e The bioluminescence image of the vehicle or JTE-607 treated mice. f Kaplan‒Meier curve showing survival of mice following implantation with GSC3565 and treatment with vehicle or JTE-607. Log-rank test was used for statistical analysis. g Immunofluorescent staining of MYC in the vehicle or JTE-607 treated tumors (left). Quantification of MYC positive cells (right, t-test, **P < 0.01, n = 4). h Immunofluorescent staining of Ki-67 in vehicle or JTE-607 treated tumors (left). Quantification of Ki-67 positive cells (right t-test, ***P < 0.001, n = 4). i Immunofluorescent staining of TUNEL in vehicle or JTE-607 treated tumors (left). Quantification of TUNEL positive cells (right, t-test, ***P < 0.001, n = 4). Scale bars: 75 μm (g–i). j Kaplan‒Meier curves showing survival based on RBBP6 mRNA expression in glioma patients from the CGGA dataset. Log-rank test was used for statistical analysis. k Kaplan‒Meier curves showing survival based on CPSF3 mRNA expression in glioma patients from the CGGA dataset. Log-rank test was used for statistical analysis.