Fig 2: THRAP3 and BCLAF1 double knockdown leads to impaired splicing and export of DDR transcripts. (A and B) Expression levels of (A) pre-spliced BRCA2 (Exon 18–Intron 18), FANCD2 (Exon 36–Intron 36), FANCL (Exon6–Intron 6), RAD51 (Exon 6–Intron 6), PALB2 (Exon 9–Intron 9) and GAPDH (Exon 3–Intron 3) mRNAs and (B) post-spliced BRCA2 (Exon 18-Exon 19), FANCD2 (Exon 36-Exon 37), FANCL (Exon 6-Exon 7), RAD51 (Exon 6-Exon 7), PALB2 (Exon 9-Exon 10) and GAPDH (Exon 3–Exon 4) mRNAs. Expression was assessed via qRT-PCR on cDNA generated from DNAse treated nuclear RNA extracts and normalized to ACTB levels in the same sample. Graphs represent the mean of three independent experiments ±SEM. Significance of changes was assessed using Student’s two-tailed t-test with significant changes indicated by **P < 0.005; ***P < 0.0005. (C) Nuclear/cytoplasmic ratio of FANCD2, FANCL, RAD51, PALB2 and GAPDH mRNA. Expression was assessed via qRT-PCR on cDNA generated from DNAse treated RNA samples and normalized to ACTB. Efficiency of nuclear and cytoplasmic RNA extracts was confirmed via qRT-PCR for MALAT1. Graphs represent the mean of three independent experiments ±SEM. Significant differences in the ratios were assessed using Student’s two-tailed t-test and are indicated by *P < 0.05; **P < 0.005; ***P < 0.0005. (D) Representative western blots demonstrating reduced protein expression levels of BRCA2, FANCD2, FANCL and RAD51 in control (siCtrl) and THRAP3/BCLAF1 double depleted cells either untreated or treated with IR (6 Gy 1 h). Protein levels of PALB2 and GAPDH were used as a positive control. (E and F) Clonogenic survival assays in control (siCtrl), THRAP3, BCLAF1 and double depleted 293T cells following treatment with mitomycin C (MMC) (E) and Olaparib (F). Mean cell viability of three independent experiments is plotted ±SEM. (G) Quantification of the DR-GFP reporter assay in control (siCtrl) and BRCA1, THRAP3, BCLAF1, double knockdown U2OS cells. 10 000 cells were scored per experiment using high-content imaging. Graphs represent the mean GFP signal from three independent experiments ±SEM. Significance of changes was assessed using Student’s two-tailed t-test with significant changes indicated by *P < 0.05, **P < 0.001. (H) Overrepresented motifs of lengths 8–20 nt from the 3' UTR regions of transcripts downregulated in the cytoplasm of THRAP3/BCLAF1 double depleted cells were determined using Multiple Em (expectation maximization) for Motif Elicitation (MEME). Two overrepresented 12-mer motifs in the 3' UTR of the transcripts identified are indicated.

Fig 3: Loss of hCINAP leads to a defect in error-free HR pathway. a hCINAPWT and hCINAP-/- HEK293T cells were subjected to HR assays. The data are presented as the mean ± SEM of three replicates. More than 1000 cells were counted in each group, two-tailed students’ t test, ***P < 0.001. b hCINAPWT and hCINAP-/- U2OS cells treated with or without IR (10 Gy) were subjected to immunofluorescence assays. Representative immunofluorescence images are shown (left), and statistics of BRCA1 foci numbers in each group are shown (right). ****P < 0.0001. c NPM1 knockdown U2OS cells transfected with the indicated plasmids were treated with IR (10 Gy) and released for 1 h. The accumulation of BRCA1 was then assessed. Scale bar, 10 µm. d Schematic of the role of different BRCA1 complex in end resection. e–g hCINAPWT and hCINAP-/- U2OS cells were immunostained with anti-RAP80 or anti-CtIP antibodies. Representative immunofluorescence images are shown in e, and BRCA1 foci numbers in cells were counted in f and g. Scale bar, 10 µm. ****P < 0.0001. h, i Immunofluorescence h and its quantification ishowing that hCINAP deficiency blocked the RAD51 foci recruitment. Representative images and percentages of cells with > 10 RAD51 foci were counted. The U2OS cells were treated with 4 Gy IR and released 4 h before fixing. Scale bar, 10 µm. ***P < 0.001. j HEK293T cells transfected with the indicated vectors were subjected to co-IP analysis using the indicated antibodies. k Cells were exposed to 10 Gy IR and then recovered for 1 h before being subjected to immunofluorescence. Scale bar, 10 µm. Additional cell images and quantification of RAD51 foci are shown in Supplementary Fig. 12. For b, c, f, g, the results are shown as the mean ± SEM from three independent experiments. Data were analyzed using Student’s t test. More than 100 cells were counted per group. Unprocessed scans of blots are provided in Supplementary Fig. 13

Fig 4: Defect in DNA damage repair and RNA export in THRAP3 associated mutants. (A) Representative immunofluorescent staining of 53BP1 marked DNA damage in U2OS cells depleted of THRAP3 with further transfection of THRAP3 cancer associated mutants 24 h following 2Gy IR. (B) Quantification of three independent experiments described above (=100 cells were scored/experiment). Mean fraction of cells containing =5 53BP1 foci is plotted ±SEM. Significant differences in the fraction of cells containing =5 53BP1 foci were assessed using Student’s two-tailed t-test and are indicated by **P < 0.005, ***P < 0.0005. (C) Nuclear/cytoplasmic ratio of ATM, BRCA2, FANCD2, FANCL, RAD51, PALB2 and GAPDH mRNA levels in control (siCtrl) and THRAP3 depleted cells as well as in cells expressing THRAP3 cancer associated mutations R101* and R837* was quantified by qRT-PCR in three independent experiments. Graphs represent the mean mRNA expression levels relative to control from three independent experiments ±SEM.

Fig 5: miR-10a targets Rad51 is essential for DSB repair during meiosis. (A) Immunofluorescence staining of SYCP3 and ?-H2AX with DAPI at pachytene spermatocyte spreads in WT and Ddx4-cOE mice. Scale bars = 20 µm. Arrows indicate XY chromosome and arrowheads indicate autosomes. (B) A schematic representation of predicted miR-10a binding sites in the mouse Jarid2-3’UTR, Tbx5-3’UTR, Sohlh2-3’UTR and Rad51-3’UTR, respectively. (C) The relative luciferase activity was assayed following co-transfection of miR-10a mimics with the constructs encoding the wild-type or mutant miR-10a binding site of Rad51 3’-UTR into the HEK293T cells. The constructed reporter vectors (psiCHECK2 represent empty vector, psiCHECK2-Rad51 represent Rad51-3’UTR reporter plasmid and psiCHECK2-mut-Rad51 represent Rad51-mut-3’UTR reporter plasmid) are shown on the x-axis, and the y-axis represents the normalized luciferase activity (firefly luciferase activities were normalized to Renilla luciferase activities). The experiments were performed independently in triplicate. *P < 0.05. (D) The expression of luciferase Jarid2-3’UTR, Tbx5-3’UTR, Sohlh2-3’UTR and Rad51-3’UTR in HEK293T cells, respectively. The experiments were performed independently in triplicate. *P < 0.05. (E) RT-qPCR analyses of Rad51 expression level in miR-10a mimics co-transfected HEK293T, NIH3T3 and GC-1 cell lines, respectively. Data are presented as mean ± SEM (n = 3). *P < 0.05, **P < 0.01. (F) Protein level of RAD51 in miR-10a mimics co-transfected 293T, 3T3 and GC-1 cell lines by Western blot and protein quantification analyses, respectively. ß-ACTIN as loading control. Protein quantification data are presented as mean ± SEM (n = 3). *P < 0.05, **P < 0.01. NC means negative control.