CDK12 and CDK13 suppress distinct intronic polyadenylation sites
Hulver, M.; Polevoda, A.; Huang, K.-L.; Wagner, E. J.; Boutz, P. L.
Show abstract
Cyclin-dependent kinases CDK12 and CDK13 are RNA polymerase II C-terminal domain Ser2 kinases that regulate transcriptional elongation and co-transcriptional RNA processing. Loss-of-function (LOF) mutations in these kinases are observed in multiple cancers and are characterized by increased intronic polyadenylation (IPA), which generates prematurely terminated transcripts and can give rise to novel intronic peptides. To investigate the distinct contributions of CDK12 and CDK13 to suppressing early transcriptional termination, we modeled selective kinase inhibition in OVCAR4 ovarian cancer cells using THZ531, a dual CDK12/13 inhibitor. A CRISPR-engineered CDK12 C1039S mutation conferred THZ531 resistance, enabling functional distinction between CDK12 and CDK13 LOF conditions. Using Poly(A) Click-seq (PACseq) across a six-point inhibitor dose curve, we identified thousands of IPA sites with distinct dose-response behaviors, allowing us to classify CDK12- and CDK13-dependent subsets. Oxford Nanopore long-read sequencing confirmed premature termination events and revealed full-length isoform switches, demonstrating how IPA alters coding potential by truncating protein domains and producing novel, intronically encoded peptide sequences. In parallel, integration of long-read transcriptomes with mass spectrometry data identified peptides derived from IPA isoforms, confirming that intronic sequences can be translated. Together, these findings establish distinct roles for CDK12 and CDK13 in maintaining transcript integrity, demonstrate that IPA transcripts are not only abundant but also translationally competent, and provide evidence for a mechanistic link between transcriptional dysregulation and the generation of novel peptide isoforms in CDK12/13-deficient cancers. These results highlight a potential therapeutic vulnerability through the exploitation of IPA-derived peptides as tumor-specific neoantigens.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Cytoplasmic Switch of ARS2 Isoforms Promotes Nonsense-Mediated mRNA Decay and Arsenic Sensitivity 95%
- Single-nucleotide-resolution genomic maps of O6-methylguanine from the glioblastoma drug temozolomide 95%
- METTL3 drives telomere targeting of TERRA lncRNA through m6A-dependent R-loop formation: a therapeutic target for ALT-positive neuroblastoma 95%
Similar papers in this journal
- A high-content RNAi screen reveals multiple roles for long noncoding RNAs in cell division 96%
- The RING Finger E3 Ligase RNF25 Protects DNA Replication Forks Independently of its Canonical Roles in Ubiquitin Signaling 95%
- A multi-omic dissection of super-enhancer driven oncogenic gene expression programs in ovarian cancer 95%
Similar papers in this journal
- E4F1 and ZNF148 are transcriptional activators of the A57C and wildtype TERT promoter 95%
- Mutational scanning of CRX classifies clinical variants and reveals biochemical properties of the transcriptional effector domain 94%
- Functional Annotation of Human Long Non-Coding RNAs via Molecular Phenotyping 94%
Similar papers in this journal
- Metabolic reprogramming of cancer cells by JMJD6-mediated pre-mRNA splicing is associated with therapeutic response to splicing inhibitor 95%
- Co-regulation and functional cooperativity of FOXM1 and RHNO1 bidirectional genes in ovarian cancer 95%
- Oncogenic PKA signaling stabilizes MYC oncoproteins via an aurora kinase A-dependent mechanism 95%
Similar papers in this journal
- Long read proteogenomics to connect disease-associated sQTLs to the protein isoform effectors of disease 94%
- Loss-of-function of the Zinc Finger Homeobox 4 ( ZFHX4 ) gene underlies a neurodevelopmental disorder 93%
- Biallelic pathogenic variants in TRMT1 disrupt tRNA modification and induce a syndromic neurodevelopmental disorder 93%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.