Fig 2: Role of FLASH in recruiting polyadenylation factors of the HCC to semi-recombinant holo U7 snRNP and its catalytic activity. (A) Amino acid sequence of the N-terminal region of FLASH. The essential LDLY motif is boxed and the start points of the Δ51 and Δ61 deletions are indicated with the arrows. The region that binds Lsm11 is underlined. (B) Proteins of mouse nuclear extract that associate with the in vitro assembled core U7 snRNP either lacking (lane 1) or containing (lane 2) recombinant Δ51 FLASH. The complexes were immobilized on SA beads via mH2a/2m-3′B pre-mRNA, as described in the legend for Figure 2A, and bound proteins were visualized by silver staining. (C) In vitro assembled core U7 snRNP was bound to either 139N FLASH or its inactive deletion mutant lacking the LDLY motif, Δ61. The resultant complexes were next bound in the presence of FLAG-tagged SLBP to radioactively labeled mH2a/2m-3′B pre-mRNA, immobilized on SA beads and incubated with mouse nuclear extract to reconstitute semi-recombinant holo U7 snRNP. The presence of the HCC components (symplekin, CPSF100, CPSF73 and CstF64), SLBP and each of the two forms of FLASH, each N-terminally tagged with GST, was monitored by Western blotting (top panel) and silver staining (bottom panel). (D) Small aliquots of immobilized processing complexes containing either 139N or Δ61 FLASH were incubated at 0°C and 32°C to evaluate their ability to support endonucleolytic cleavage of 5′-labeled mH2a/2m-3′B pre-mRNA. (E–I) In vitro assembled core SupU7 snRNP was bound to Δ61 or 139N FLASH. The complexes were purified by size exclusion chromatography, analyzed by Western blotting (panel E) and tested for the ability to support processing of H2a/Ys (panel F), ΔSL/Ys (panel G) and DCP/Ys (panels H and I) in a mouse nuclear extract.

Fig 3: Chromodomain–helicase‐DNA‐binding protein 7 (CHD7) regulates neural stem/progenitor cell (NSC) quiescence. (A): Schematic diagram of the experimental strategy to delete Chd7 from adult NSCs. Six days after tamoxifen (TAM) administration 5‐bromodeoxyuridine (BrdU) was injected and brains collected 24 hours later. (B–D): Quantification of the number of radial GFAP+SOX2+ NSCs (B), GFAP‐SOX2+ progenitor cells, and ASCL1+ stem/progenitor cells (D) in the subgranular zone (SGZ) of Chd7null and control (Ctrl) mice 1 week after TAM injection. (E): Quantification of the number of DCX+ cells in the dentate gyrus (DG) of Chd7null and control (Ctrl) mice 1 week after TAM injection. (F): Representative image of a section of the DG stained with antibodies against BrdU, SOX2, and glial fibrillary acidic protein (GFAP) with DAPI. Yellow arrow head shows a BrdU+ NSC (BrdU+SOX2+GFAP+ cell in the SGZ), pink arrow head shows a BrdU‐ NSC, light blue arrow heads show a BrdU+ progenitor (BrdU+SOX2+GFAP‐). Scale bar = 20 µm. (G): Quantification of the total number of BrdU+ cells in the SGZ of Chd7null and ctrl mice. (H, I): Quantification of the number of BrdU+ NSCs (H) and progenitors (I) in the SGZ of Chd7null and ctrl mice. (J, K) Quantification of the number of proliferating (MCM2+) radial GFAP+ SGZ NSCs (J) and MCM2+ GFAP‐ proliferating cells in the SGZ of Chd7null and ctrl mice (K). For all experiments, n = 3–4 animals per condition. (L): Representative image of a section of the DG stained with antibodies against MCM2, GFAP and DAPI. White arrow head shows a GFAP+MCM2+ NSC, blue arrow head shows a GFAP+MCM2‐ quiescent NSC. Scale bar = 20 µm. (M) RT‐qPCR quantification of Chd7 expression in cultured fetal‐derived NSCs from Nestin‐Cre;Chd7f/f (Chd7null) and Cre‐negative (control; Ctrl) embryos under growth conditions (Growth) and after incubation with media containing 50 ng/ml BMP4 (quiescence), relative to ctrl. Note Chd7 expression is completely absent from Chd7null cells. n = 4 cultures. (N): Western blot for CHD7 on total cell lysates of cultured fetal‐derived NSCs from Chd7null (NestinCre;Chd7‐/‐) and control (Chd7+/+) cells. Cleavage and polyadenylation specificity factor 100 (CPSF100) was used as a loading control. (O): Quantification of the percentage of proliferating (Ki67+) control and Chd7null cells in vitro under growth conditions and after incubation with media containing 50 ng/ml BMP4 (quiescence). Note that Chd7null cells are refractory to quiescence induction after 72 and 96 hours in quiescence media. Both Chd7null and ctrl cells are able to return to a proliferative state when returned to growth conditions (growth media). n = 4 cultures. All data represented as mean ± SEM; *, p < 0.05; ***, p < 0.001 Student's t‐test. Abbreviations: BrdU, 5‐bromodeoxyuridine; CHD7, chromodomain–helicase‐DNA‐binding protein 7; CPSF100, cleavage and polyadenylation specific factor 2, 100 kDa subunit; DAPI, 4′,6‐diamidino‐2‐phenylindole; GFAP, glial fibrillary acidic protein; MCM2, minichromosome maintenance deficient 2 mitotin; NSC, neural stem/progenitor cell; SGZ, subgranular zone; TAM, tamoxifen.

Fig 4: CDK9 and PP2A regulate mRNA cleavage and polyadenylation and alternative poly(A) site usage A Schematic of the nuclear and cytoplasmic qRT–PCR experiments.B qRT–PCR of nuclear polyadenylated mRNAs of several TNFa induced or noninduced genes with a 30-min DMSO, DRB, CA, or DRB + CA treatment. n = 6 biological replicates, mean ± SEM, P-value: *P < 0.05, **P < 0.01, ***P < 0.001. Statistical test: two-tailed unpaired t-test.C qRT–PCR of cytoplasmic polyadenylated mRNAs of several TNFa induced or noninduced genes with a 30-min DMSO, DRB, CA, or DRB + CA treatment. n = 4 biological replicates, mean ± SEM, P-value: *P < 0.05, **P < 0.01, ***P < 0.001. Statistical test: two-tailed unpaired t-test.D ChIP-qPCR of Ser2P, CPSF73, or CPSF2 ratioed to total pol II after 30-min treatment with DMSO, DRB, CA, or DRB + CA on KPNB1. n = 3 biological replicates, mean ± SEM, P-value: *P < 0.05, **P < 0.01, ***P < 0.001. Statistical test: two-tailed unpaired t-test.E Co-immunoprecipitation of total pol II from HeLa cells treated for 30 min with DMSO, DRB, CA, or DRB + CA followed by Western blot with total pol II, Ser2P, SF3B1, CPSF2, Xrn2, and GAPDH antibodies.

Fig 5: CDK9 phosphorylates several transcription and splicing factors in vivo and CDK9 inhibition promotes the loss of interaction of SF3B1 and CPA factors with pol II A Distribution of the proteins found to be phosphorylated by CDK9 in SILAC phosphoproteomics in HeLa cells treated or not with 100 µM DRB for 30 min (fold change > 1.5 in both biological replicates, number of phosphopeptides decreased shown in brackets).B Western blot of total SF3B1, SF3B1 T142P, and histone H3 as a loading control, on the chromatin fraction of CDK9as cells treated for 30 min with DMSO, NA, tautomycetin (TT), calyculin A (CA), NA + TT, or NA + CA. The histone H3 loading control (CA samples) is the same as the loading control shown in Appendix Fig S3B as the same Western blot experiment is shown in two different figs.C Co-immunoprecipitation of CPSF2 in the CDK9as cell treated for 30 min with DMSO or NA followed by Western blot with total pol II, CPSF2, CPSF3, SF3B1, SF3B3, and ß-actin (negative control) antibodies.D Co-immunoprecipitation of SF3B1 in the CDK9as cell treated for 30 min with DMSO, NA, or DRB followed by Western blot with total pol II, SF3B1, SF3B2, and GAPDH (negative control) antibodies.