Fig 2: Discovery of small molecule inhibitors targeting MAGE-A11.a 500 analogs from 9 scaffolds were tested in dose response TR-FRET assays for disruption of MAGE-A11 binding to PCF11 identified the quinoline scaffold. b Chemical structure of SJ521054 hit. c TR-FRET competition assay using indicated concentrations of quinoline series compounds for disruption of MAGE-A11:PCF11 binding with n = 2 biologically independent experiments with triplicates per compound. Data are mean ± SD. d SJ521054 blocks PCF11 degradation by MAGE-A11. HEK293FT cells stably expressing FLAG-vector or FLAG-MAGE-A11 were treated with indicated concentrations of SJ521054 for 24 h before immunoblotting. Source data are provided as a Source Data file. e SJ521054 reduces viability of MAGE-A11 expressing DAOY cancer cells. MAGE-A11-knockout DAOY cells or those reconstituted with MAGE-A11 were treated with indicated concentrations of SJ521054 and cell viability was measured by alamarBlue assay 24 h later (n = 3 biologically independent experiments per group). Data are mean ± SD. Black circle, Vector; Red square, MAGE-A11 WT #1; and Blue triangle, MAGE-A11 WT #2. f Derivatives of SJ521054 were synthesized and tested by TR-FRET competition assays with SJ1008066 showing improved activity (n = 2 biologically independent experiments with triplicates per compound). Data are mean ± SD. g Direct ligand binding of SJ521054 and SJ1008066 were evaluated using NanoDSF using the intrinsic fluorescence of His-SUMO-MAGE-A11 MHD at 350 nm and 330 nm with n = 3 biologically independent experiments per compound. Both compounds showed stabilization of MAGE-A11 in a six-point dose response confirming on-target activity of each compound. Data are mean ± SD. Blue circle, SJ521054 and red square, SJ1008066. h SJ521054 and SJ1008066 inhibit MAGE-A11 binding to PCF11 in cell lysate. FLAG-MAGE-A11 stably expressing HEK293FT cell lysates were incubated with 100 µM of each compound for 24 h and then subjected to pull-down with anti-FLAG followed by SDS-PAGE and immunoblotting for endogenous PCF11, and quantitated (n = 4 for SJ521054, n = 3 for SJ1008066, biologically independent experiments per compound, two-tailed unpaired Student’s t test, ***P < 0.001). Data are mean ± SEM. Source data are provided as a Source Data file. i SJ521054 and SJ1008066 do not disrupt interaction of MAGE-A3:AMPKα1 or MAGE-F1:MMS19 in cell lysate. HeLa cells or HEK293FT cells were transfected with the indicated constructs for 48 h and cell lysates were incubated with 100 µM of each compound for 24 h and then subjected to IP with anti-Myc followed by SDS-PAGE and immunoblotting for indicated proteins. Source data are provided as a Source Data file.

Fig 3: The SBC region mediates substrate recognition and specificity across the MAGE family.a Sequence conservation of MAGE-A genes mapped onto MAGE-A11:PCF11 structure. b Comparison of MAGE-A3 and -A11 SBC reveals key sequence differences in critical substrate binding residues. c In vitro GST-pulldown assays reveals that GST-PCF11 binds in vitro translated Myc-MAGE-A11, but no other Myc-MAGE-A protein. Source data are provided as a Source Data file. d–g Mutation of MAGE-A3 or -F1 SBC disrupts binding to their respective substrates AMPK and MMS19, not the E3 ligases TRIM28 and NSE1. Surface representations of MAGE-A3 (protomer model formed by PDB ID 4V0P for residues 1-296 and MAGE-A11 based homology model for C-terminal residues 297–311) (d) and MAGE-F1 (SWISS-MODEL based on PDB:4V0P) (f) show positioning of SBC residues mutated. Consistent with Fig. 2c, MAGE atoms within 5 Å of modeled PCF11 (based on superimposition of MAGE-A11:PCF11) are colored cyan. Residues in the SBC that are mutated are labeled and colored yellow. Co-IP of MAGEs with substrates and ligases were determined in HeLa cells (e) or HEK293FT cells (g) upon transfection of the indicated constructs for 48 h, IP with anti-Myc, SDS-PAGE, and immunoblotting for indicated proteins. Source data are provided as a Source Data file.

Fig 4: Mutation of MAGE-A11 SBC disrupts its oncogenic activity.a MAGE-A11-induced ubiquitination of PCF11 depends on its substrate recognition by MAGE-A11 SBC. Ubiquitinated proteins from FLAG-vector, FLAG-MAGE-A11 wild-type, or SBC mutants stably expressing HEK293FT cells were treated with 10 µM MG132 for 4 h and isolated with tandem ubiquitin binding entity-agarose followed by SDS-PAGE and immunoblotting for endogenous PCF11. Source data are provided as a Source Data file. b MAGE-A11 SBC mutants fail to promote proliferation of MAGE-A11 knockout DAOY cells. Wild-type, MAGE-A11-knockout DAOY cells, or those reconstituted with MAGE-A11 were counted for cell proliferation at the indicated time points (n = 3 biologically independent experiments per group). Data are mean ± SEM. p-value by two-way ANOVA followed by Tukey’s multiple comparisons test is indicated (****P < 0.0001). c–e MAGE-A11 knockout DAOY cells reconstituted with MAGE-A11 SBC mutants reduced clonogenic growth and xenograft tumor growth of MAGE-A11 knockout DAOY cells. Stable expression of wild-type MAGE-A11 or SBC mutants in MAGE-A11-knockout DAOY cells were assayed for clonogenic growth (c) (n = 3 biologically independent experiments per group, two-tailed unpaired Student’s t test, *P < 0.05; **P < 0.01), Data are mean ± SEM and xenograft tumor growth in mice (n = 6 per group). Tumor burden (d) and tumor weight (e) are shown. Data are mean ± SD. p-value by two-way ANOVA followed by Tukey’s multiple comparisons test is indicated in d (****P < 0.0001), ordinary one-way ANOVA followed by Sidak’s multiple comparisons test is indicated in e (**P < 0.01, ***P < 0.001, ****P < 0.0001). Red circle, Vector; Blue square, MAGE-A11 WT; Green triangle, MAGE-A11 F275A; and Pink triangle, MAGE-A11 M341R. f MAGE-A11 M341R mutant fails to promote mRNA 3’UTR shortening. Transcriptome analysis of tumors (n = 3 per group) from MAGE-A11 knockout DAOY cells stably expressing FLAG-MAGE-A11 wild-type or M341R mutant was determined and scatterplot of Percentage of Distal polyA site Usage Index (PDUI) is shown. False discovery rate ≤ 0.05 and ΔPDUIs ≥ 0.2 are colored. The shift towards distal polyadenylation site is significant (p = 5.9 × 10−16, binomial test). g Representative RNA-seq density plots are shown.

Fig 5: mRNA 3′-end processing factors are required for termination of transcription of human snRNA genes. (A) Western blot analysis of HeLa whole-cell extracts from control cells or cells transfected with an siRNA specific for Ssu72 and Pcf11. Antibodies used for detection were α−Ssu72 (left) and α-Pcf11 (right). The α−actin was used as a control for protein levels. (B) Schematic of the β-actin gene labelled as in Figure 2. Graphs represent results of pol II ChIP analysis before and after siRNA-mediated knockdown of Pcf11 (left) and Ssu72 (right). (C) Schematics of the U1 and U2 snRNA genes are labelled as in Figure 1. Graphs represent results of pol II ChIP analysis of the U1 (left) and U2 (right) snRNA genes before and after siRNA-mediated knockdown of Pcf11 (top) and Ssu72 (bottom). Pol II levels in (A) and (B) were normalized as in Figure 2. Error bars represent standard deviation of at least three independent experiments.