Mutations that attenuate DNA repair by homologous recombination (HR) promote tumorigenesis and sensitize cells to chemotherapeutic agents that cause replication fork collapse, a phenotype known as “BRCAness.” BRCAness tumors arise from loss-of-function mutations in 22 genes. Of these genes, all but one (Cdk12) directly function in the HR repair pathway. Cdk12 phosphorylates Serine 2 of the RNA Polymerase II (RNAPII) C-terminal domain (CTD) heptapeptide repeat, a modification that regulates transcription elongation, splicing, and cleavage/polyadenylation. Genome-wide expression studies suggest that Cdk12 depletion abrogates the expression of several HR genes relatively specifically, blunting HR repair. This observation suggests that Cdk12 mutational status may predict tumor sensitivity to targeted treatments against BRCAness, such as Parp 1 inhibitors, and that small-molecule inhibitors of Cdk12 may induce sensitization of otherwise HR-competent tumors to these treatments. Despite this growing clinical interest, the mechanism behind the apparent specificity of Cdk12 in regulating HR genes remains unknown. Here we find that Cdk12 globally suppresses intronic polyadenylation events, enabling the production of full-length gene products. Many HR genes harbor significantly more intronic polyadenylation sites compared to all expressed genes, and the cumulative effect of these sites accounts for the increased sensitivity of HR gene expression to Cdk12 loss. Finally, we find evidence that Cdk12 loss-of-function mutations cause increased intronic polyadenylation within HR genes in human tumors, suggesting that this mechanism is conserved. This work clarifies the biological function of CDK12 and underscores its potential both as a chemotherapeutic target and as a tumor biomarker.
Overall design: V6.5 mouse embryonic stem cells with endogenous Cdk12 knock out complemented by doxycycline (Dox)-inducible Flag- HA- epitope tagged Cdk12 transgene (Cdk12△) were grown in the presence of Cdk12 (+Dox) or were depleted of Cdk12 (-Dox). RNA sequencing experimental design: PolyA-selected RNA was isolated from two independently derived Cdk12△ clones both in biological duplicate for cells mainained in Dox (+Dox), withdrawn from Dox for 24 hours (-Dox 24hrs), or withdrawn from Dox for 48 hours (-Dox 48hrs). In parallel, polyA-selected RNA was isolated in biological duplicate from parental Cdk12 Floxed cells (V6.5 mouse embryonic stem cells with homoqygous LoxP sites flanking Cdk12 exon 4 without introduction of the Dox-inducbile transgene) that were pretreated with Dox for 17 days and then either harvested in Dox or harvested after withdrawal of Dox for 24 or 48 hours to serve as a control for gene expression changes caused by long-term Dox treatment followed by short term withdrawal. A total of 18 libraries were prepared and sequenced. ChIP Sequencing Experimental Design: We performed chromatin immunoprecipitation (ChIP) sequencing on Cdk12∆ cells in +Dox or –Dox 48 hour conditions for total RNA Polymerase II (RNAPII), RNAPII phosphorylated on Serine 2 of the CTD (Ser2p), and RNAPII phosphoryalted on Serine 5 of the CTD (Ser5p). Two independent antibodies were used for each antigen as follows: 8WG16 (Abcam ab817) and Rpb3 (Bethyl A303-771A) for total RNAPII density; H5 Clone (Abcam ab24758) and 3E10 Clone (Millipore 04-1571) for Ser2p RNAPII density; and H14 clone (Abcam Ab24759) and 3E8 clone (Millipore 04-1572) for Ser5p RNAPII density. Two biological replicates were processed for each antibody and dox condition. For each dox condition, we also processed 4 negative control libraries: one whole-cell extract (WCE) sample and 3 mock IP samples with the following antibodies: goat anti-mouse IgM (Thermo Fisher 31172), goat anti-rat IgG (Thermo Fisher 31226), mouse IgG2a [MOPC-173] Isotype Control (Abcam ab18413). A total of 32 libraries were prepared and sequenced. For detailed descriptions of the experimental design, please refer to the referenced manuscript.
Less...