Fig 2: Downstream effects of the SF3B1 mutation.a PRL secretion in SF3B1 wild type and SF3B1 mutant primary human prolactinomas cells. b Representative western blot for SF3B1 expression levels in primary human prolactinoma cells transfected with control or specific SF3B1 siRNA. β-actin was used as internal control. c Suppression of the PRL secretion in primary human prolactinoma cells after the SF3B1 knockdown using SF3B1 siRNA. d PRL secretion in primary human prolactinoma cells infected with Ad-null, Ad-SF3B1WT and Ad-SF3B1R625H are shown, respectively. e Focus formation was assessed in GH3 cells infected with Ad-null, Ad-SF3B1WT, and Ad-SF3B1R625H at the indicated MOI. f Results of CCK-8 cell proliferation assay showing in GH3 cells infected with Ad-null, Ad-SF3B1WT, and Ad-SF3B1R625H at the indicated MOI. g Annexin V/PI staining and flow cytometry showed the percentages of apoptosis of the GH3 cells infected with Ad-null, Ad-SF3B1WT, and Ad-SF3B1R625H. h Focus formations were assessed in MMQ cells infected with Ad-null, Ad-SF3B1WT, and Ad-SF3B1R625H at the indicated MOI. i Results of CCK-8 cell proliferation assay showing in MMQ cells infected with Ad-null, Ad-SF3B1WT, and Ad-SF3B1R625H at the indicated MOI. j Annexin V/PI staining and flow cytometry showed the percentages of apoptosis of the MMQ cells infected with Ad-null, Ad-SF3B1WT, and Ad-SF3B1R625H. c–i n = 9 per group. k Focus formations were assessed in stable GH3-control, GH3-SF3B1WT, GH3-SF3B1R625H cells (n = 6 for GH3-control and n = 3 for two other cells). l Results of CCK-8 cell proliferation assay showing in stable GH3-control, GH3-SF3B1WT, GH3-SF3B1R625H cells (n = 5 per group). m Annexin V/PI staining and flow cytometry showed the percentages of apoptosis of stable GH3-control, GH3-SF3B1WT, GH3-SF3B1R625H cells. Data are represented as mean ± SD. The p value by two-tailed unpaired t test is indicated in a. The p values by one-way ANOVA followed by Dunnett’s multiple comparisons test in c, d, k, l, m and followed by Tukey’s multiple comparisons post hoc test in e, f, g, h, i, j are indicated. Source data are provided as a Source Data file.

Fig 3: Downstream effects on aberrant splicing on ESRRG.a The qRT-PCR results display ESRRG mRNA expression levels in primary human prolactinoma cells transfected with control or specific ESRRG siRNA. GAPDH was used as an internal control (n = 3 per group). b Suppression of PRL secretion in primary human prolactinomas cells after ESRRG knockdown using ESRRG siRNA (n = 9 per group). c Results of CCK-8 cell proliferation assay in GH3 cells infected with Ad-null, Ad-canonical ESRRG, and Ad-cryptic ESRRG (n = 5 per group). d Results of CCK-8 cell proliferation assay in MMQ cells infected with Ad-null, Ad-canonical ESRRG, and Ad-cryptic ESRRG (n = 5 per group). e Focus formation was assessed in GH3 cells infected with Ad-null, Ad-canonical ESRRG and Ad-cryptic ESRRG (n = 9 per group). f Focus formations were assessed in MMQ cells infected with Ad-null, Ad-canonical ESRRG, and Ad-cryptic ESRRG (n = 9 per group). g Annexin V/PI staining and flow cytometry showed the percentages of apoptosis of the GH3 cells infected with Ad-null, Ad-canonical ESRRG, and Ad-cryptic ESRRG (n = 9 per group). h Annexin V/PI staining and flow cytometry showed the percentages of apoptosis of the MMQ cells infected with Ad-null, Ad-canonical ESRRG, and Ad-cryptic ESRRG (n = 9 per group). Results are expressed as mean ± SD. The p values by one-way ANOVA followed by Dunnett’s multiple comparisons test in a, b and followed by Tukey’s multiple comparisons post hoc test in c–h are indicated. Source data are provided as a Source Data file.

Fig 4: The effects of aberrant splicing on ESRRG.a ESRRG mRNA levels in immunoprecipitates were determined through qRT-PCR analysis (n = 3 per group). ESRRG mRNA expression levels were presented as fold enrichment ratios compared with IgG. b Cryptic ESRRG mRNA levels in immunoprecipitates of wild-type and mutant primary human prolactinoma cells were determined by qRT-PCR (n = 3 per group). Expression levels of ESRRG mRNA were presented as fold enrichment ratios compared with IgG. c–e CLIP of SF3B1-bound ESRRG mRNA in MCF7 cells c, SF3B1WT prolactinomas d, or SF3B1R625H prolactinomas e (n = 3 per group). qPCR was used to identify the region in ESRRG bound by SF3B1 protein. The amount of immunoprecipitated RNAs in each sample is represented as a signal relative to the negative (IgG) sample. Schematic representation of human ESRRG segments amplified by primer pairs for CLIP-qPCR. f GST-Pit-1 fusion protein immobilized on glutathione beads and incubated with HIS-tagged cryptic or canonical ESRRG proteins. Bound ESRRG proteins were detected by anti-HIS immunoblotting. g Immunoblotting with the indicated antibodies of FLAG immunoprecipitated from lysates of HEK293 cells co-transfected with FLAG-tagged Pit-1, HA-tagged cryptic, or canonical ESRRG. h Relative luciferase activity of PRL promoter in HEK293 cells transfected with pCDNA3.1-Pit-1 (Pit-1), pGL3-basic-PRL promoter (PRL-p), or pCDH-ESRRG-canonical/cryptic and their corresponding empty vectors including pCDNA3.1, pGL3-basic, and pCDH. pRL-TK was used as a control (n = 4 per group). Results are expressed as mean ± SD. The p values by one-way ANOVA followed by Dunnett’s multiple comparisons test in a and followed by Bonferroni’s multiple comparisons test in h are indicated. The p value by two-tailed unpaired t test is indicated in b. Source data are provided as a Source Data file.