CPSF1

CPSF1 is a sequence-specific recognition factor for the AAUAAA polyadenylation signal that governs pre-mRNA 3' end formation and, through this activity, shapes alternative splicing and isoform choice (PMID:23151878, PMID:35402071). By binding intronic AAUAAA signals positioned near 5' splice sites, CPSF1 competes with spliceosome assembly to drive exon skipping (IL7R exon 6) and cryptic-exon polyadenylation (AR-v7 in prostate cancer), coupling 3' end processing to splice-site selection (PMID:23151878, PMID:35402071). Genome-wide, CPSF1 acts as a suppressor of intergenic poly(A) site usage; its loss causes use of distal poly(A) sites, 3' UTR lengthening, and reduced mRNA levels for glycolysis genes, lowering glycolytic output (PMID:39847481). Beyond RNA processing, CPSF1 participates in protein degradation: it is itself ubiquitinated and degraded by the SIAH1 E3 ligase (PMID:35402071), and it functions as a CUL4A-associated E3 ligase component that targets HIF-1α and MYC for proteasomal degradation, an activity antagonized by ABL-mediated phosphorylation of CUL4A (PMID:37040401). In vivo, loss of cpsf1 in zebrafish causes defective polyadenylation of snrnp70, apoptotic loss of hematopoietic stem cells, and aberrant retinal ganglion cell axon projection with reduced eye size (PMID:21330472, PMID:30689892).

1. 2011 High

   Whether CPSF1-dependent 3' end processing is required for a specific developmental program was unknown; a zebrafish loss-of-function mutant linked cpsf1 to defined substrate processing and cell survival.

   Evidence ENU mutagenesis, positional cloning, TUNEL and reporter assays, and a snrnp70 polyadenylation assay in the grechetto mutant

   PMID:21330472

   Open questions at the time

   • Did not define how broadly cpsf1 processes substrates beyond snrnp70
   • Mechanism of substrate selectivity not established
2. 2012 High

   It was unclear how polyadenylation machinery could affect splicing; CPSF1 was shown to bind an intronic AAUAAA near a 5' splice site and block spliceosome binding, establishing competition between the two machineries.

   Evidence RNA affinity chromatography with MS, siRNA knockdown, AAUAAA mutagenesis, and minigene splicing assays on IL7R exon 6

   PMID:23151878

   Open questions at the time

   • Did not resolve structural basis of spliceosome exclusion
   • Generality of splice-competition across transcripts not tested
3. 2019 Medium

   Whether cpsf1 contributes to neural development was unaddressed; knockdown linked it to retinal ganglion cell axon guidance and eye size.

   Evidence Morpholino knockdown with mRNA rescue and live imaging of RGC axons in zebrafish

   PMID:30689892

   Open questions at the time

   • Molecular substrate underlying the axon phenotype not identified
   • Morpholino specificity supported only by mRNA rescue
4. 2020 Low

   A possible signaling role was raised by data placing CPSF1 upstream of MAPK/ERK in retinal endothelial cells under high glucose.

   Evidence AAV overexpression in an STZ diabetic rat model plus knockdown/overexpression with p-ERK western blotting in HRVECs

   PMID:32046510

   Open questions at the time

   • Pathway placement rests on phospho-western without direct binding or epistasis
   • Single lab, no mechanistic link to RNA processing
5. 2022 Medium

   How CPSF1 abundance is controlled and how it drives an oncogenic isoform was addressed by identifying SIAH1-mediated degradation and CPSF1 binding to the AR-v7 cryptic-exon AAUAAA.

   Evidence Co-IP, ubiquitination assay, m6A analysis, and CPSF1 manipulation with RT-PCR readout of AR CE3 in prostate cancer

   PMID:35402071

   Open questions at the time

   • SIAH1-CPSF1 interaction from Co-IP without reciprocal structural validation
   • Direct CPSF1 occupancy on CE3 inferred from functional readouts
6. 2022 Low

   A pro-tumorigenic APA function in gastric cancer was proposed via CPSF1-driven 3' UTR shortening of NSDHL.

   Evidence Knockdown with RNA-seq 3' UTR length analysis and NSDHL rescue in proliferation/migration assays

   PMID:36381317

   Open questions at the time

   • No direct binding shown between CPSF1 and the NSDHL 3' UTR
   • Mechanism inferred from RNA-seq rather than direct measurement
7. 2023 Medium

   An unexpected protein-degradation activity was uncovered: CPSF1 acts as a CUL4A-associated E3 ligase targeting HIF-1α and MYC, with ABL kinase antagonism providing hypoxic regulation.

   Evidence FACS-based CRISPR screen, Co-IP of ABL with CUL4A, CPSF1/ABL competition binding, proteasome rescue, and protein-level readouts

   PMID:37040401

   Open questions at the time

   • Direct ubiquitin transfer by a CPSF1-CUL4A complex not biochemically reconstituted
   • Relationship between this ligase role and CPSF1's RNA-processing function unresolved
8. 2025 Medium

   The genome-wide consequence of CPSF1 loss was defined as derepression of intergenic poly(A) sites, linking 3' UTR lengthening to suppressed glycolysis.

   Evidence siRNA knockdown with 3' end sequencing, RNA-seq, and Seahorse glycolysis assays in prostate cancer cells

   PMID:39847481

   Open questions at the time

   • Direct CPSF1 occupancy at suppressed intergenic sites not mapped
   • Single cancer model
9. 2025 Low

   Transcriptional and metabolic context was extended by placing CPSF1 downstream of HSF1 and upstream of SREBP1 and AKT/mTOR in milk fat synthesis.

   Evidence Promoter-binding/ChIP for HSF1 at CPSF1, CPSF1 manipulation, and AKT/mTOR westerns in MAC-T cells

   PMID:39932399

   Open questions at the time

   • ChIP rigor unspecified and pathway placement indirect
   • No direct CPSF1 RNA-processing target tied to SREBP1

• How CPSF1's canonical role in poly(A) signal recognition is mechanistically reconciled with its reported CUL4A-dependent E3 ligase activity, and what determines substrate selection in each role, remains unresolved.
• No structural model integrating RNA-binding and ubiquitin-ligase functions
• Substrate selectivity rules for both activities undefined