Loss of CFIm activates YAP/TAZ and connects mRNA cleavage and polyadenylation inhibition to BRCAness

Alternative polyadenylation, cancer, YAP, TAZ, DNA damage, BRCAness, NUDT21, CPSF6, CFIm, LATS1/2 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 4

Yes-associated protein (YAP) and transcriptional co-activator with PDZ-binding motif (TAZ) are two homologous transcriptional co-activators that do not directly bind to DNA but are crucial for normal development and tissue regeneration1. YAP/TAZ activation is seen in a wide range of human cancer, and activation of YAP/TAZ and their transcriptional program in cancer is an important mechanism for the development of therapeutic resistance and disease progression2. Therefore, it is essential to develop a thorough understanding in how cancer cells activate YAP/TAZ. YAP/TAZ is normally repressed by the mammalian Hippo pathway, which is a cascade consisting of two kinases, MST1/2 and LATS1/2, in addition to multiple adaptor proteins3. LATS1/2 directly phosphorylates YAP/TAZ to suppress their nuclear translocation and promote their degradation1. However, with the exception of certain cancer types such as malignant mesothelioma, meningioma and Schwannoma, mutations in the Hippo pathway components are not common in human cancer4. This suggests that cancer cells have alternative ways to activate YAP/TAZ to their advantages without shutting down the Hippo pathway. For example, more than 40% of uveal melanoma carry activating mutations in G-protein subunits GNAQ or GNA115,6, and it was found that these mutations drive YAP/TAZ activation to promote tumorigenesis7. In humans, the mRNA 3′-end processing complex consists of more than 20 proteins that are grouped into 4 factors: CPSF, CSTF, CFIIm and CFIm8. The human CFIm consists of two subunits9: a small subunit, CFIm25 (encoded by the NUDT21 gene), which directly binds to the UGUA motif in the polyadenylation site (PAS)9, and two alternative large subunits, CFIm59 (encoded by the CPSF7 gene) and CFIm68 (encoded by the CPSF6 gene), which activate the mRNA 3′ end processing by interacting with CPSF10. However, CFIm68 is a much stronger activator of the mRNA 3′-end processing than CFIm5910. CFIm is not essential for the cleavage reaction in mRNA 3′ end processing11, but it has a strong influence on PAS choices. We and .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 5 others reported predominantly (>90%) proximal APA shifts in hundreds of genes from loss of either NUDT21 or CPSF610,12-14. However, how these widespread APA changes from CFIm loss regulate the activities of cell signaling pathways is not fully understood, but the impact of CFIm on cell signaling starts to emerge from recent studies. For example, in HEK293 cells, CFIm promoted the activities of ERK1/2, which were correlated with the expression level of CFIm12. In contrast, in human primary fibroblasts, NUDT21 knockdown activated fibrotic pathways that promoted the expression of profibrotic genes15. CPSF3 is the endonuclease in the mRNA 3′-end processing complex16. CPSF3 inhibition suppresses mRNA CPA, which results in transcription readthrough and APA changes17,18. The first CPSF3 inhibitor, JTE-607, was recently shown to have anti-proliferation activities in some types of cancer such as acute myeloid leukemia (AML) and Ewing’s sarcoma19. The tumor suppressors BRCA1 and BRCA2 both play key roles in the homology- directed repair (HDR) pathways to maintain genome stability20. BRCA1/2 mutant cancer is deficient in HDR, and it has increased vulnerability to poly(adenosine 5′-diphosphate–ribose) polymerase (PARP) inhibitors and DNA crosslinking agents such as platinum salts and mitomycin C (MMC)21. The term “BRCAness” was used to describe the similar phenotype (sensitivity to PARP inhibitors) observed in cancer without BRCA1/2 germline mutations22. The presence of HDR defects is now considered as the mechanism for the BRCAness phenotype, and additional genes that cause BRCAness when mutated in cancer have been identified20,21. CFIm loss-of-function (LOF), either caused by low expression or inhibitory phosphorylation, has been reported in several types of cancer23-25. In this study, we used two different cancer cell models to investigate how CFIm loss affects the cell signaling activities. We discovered that CFIm loss activated YAP/TAZ and their target genes, and we identified LATS1 and NEDD4L as the key genes linking CFIm loss to YAP/TAZ activation. In search for a therapeutic strategy against CFIm loss, we found that CFIm loss increased sensitivity to mRNA CPA inhibition. Furthermore, we demonstrated that the combination of CFIm loss and CPSF3 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 6 inhibition induced BRCAness. Overall, our study identified restraining the YAP/TAZ transcriptional program and supporting DNA damage repair as previously unknown tumor suppressive roles of CFIm. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 7

CFIm loss activates YAP/TAZ in colon and breast cancer cells. To study the biological function of CFIm (Fig. 1A), we used the degradation tag (dTAG) system26, which can achieve a rapid, inducible and specific depletion of the target protein. In the presence of a chemical inducer such as dTAGV-1, the FKBP12F36V degron-tagged target protein binds to an endogenous E3 ubiqutin ligase and is subsequently degraded26. We chose the HCT116 colon cancer cells as the model because of its favorable characteristics: a near-diploid genome that is amenable to genome editing with the clustered regularly interspaced short palindromic repeats (CRISPR) technology27 and the reported success in implementing the dTAG system28. We generated a single-cell clone of HCT116 colon cancer cell in which both alleles of NUDT21 were tagged with the FKBP12F36V degron tag by CRISPR-Cas9 editing (referred to as dTAG-NUDT21 HCT116 cells hereafter) (Fig. S1A). The tagging of NUDT21 did not substantially change NUDT21 and CPSF6 protein expression when compared with wildtype HCT116 cells (Fig. S1B). Treating dTAG-NUDT21 HCT116 cells with dTAGV-1 resulted in complete depletion of NUDT21 proteins within 4 hours (Fig. S1C). We also verified that dTAGV-1 treatment in dTAG-NUDT21 HCT116 cells resulted in proximal APA shift of TIMP2, a known CFIm target13,29 (Fig. S1D). Next, we treated dTAG-NUDT21 HCT116 cells with either DMSO or dTAGV-1 to investigate how CFIm loss impacts major cell signaling pathways. We found that dTAGV-1 treatment increased YAP and TAZ protein expression without any changes in phospho-YAP (inactive form) abundance, suggesting possible YAP/TAZ activation from CFIm loss (Fig. 1B). We also performed RT-qPCR experiments in dTAGV-1 treated dTAG-NUDT21 HCT116 cells to examine the expression of YAP/TAZ direct target genes–CTGF, CYR61 and AMOTL230. dTAGV- 1 treatment increased the expression of all 3 genes, indicating that the YAP/TAZ transcription program was likely activated in response to CFIm loss (Fig. 1C). To examine the relationship .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 8 between CFIm loss and YAP/TAZ in a distinct biological context with a different gene inhibition technique, we next used an MDA-MB-231 breast cancer cell line that was previously generated in our lab and carries a doxycycline-inducible short hairpin RNA (shRNA) targeting CPSF631 (referred to as shCPSF6-MDA cells hereafter). Activation of the CPSF6 shRNA by doxycycline treatment for 72 hours in shCPSF6-MDA cells strongly reduced CPSF6 expression but did not

in its complete loss (Fig. S1E). Next, we grew shCPSF6-MDA cells in the presence or absence of doxycycline for 72 hours, and we performed RT-qPCR to examine the expression of YAP/TAZ target genes. Consistent with the results in dTAG-NUDT21 HCT116 cells, we found that the expression of CTGF and CYR61 was again increased in shCPSF6-MDA cells with low CPSF6 expression (Fig. 1D). Lastly, we stably transduced shCPSF6-MDA cells with a YAP/TAZ reporter (8xGTIIC–DsRED)32 to examine whether the upregulation of YAP/TAZ targets occurs at the transcription level. The induction of DsRED expression was similar between doxycycline and SM04690, a YAP activator33 (Fig. 1E), confirming the transcriptional activation effects on a YAP/TAZ-responsive promoter from CPSF6 knockdown. Taken together, our results show that CFIm loss activates YAP/TAZ in colon and breast cancer cells. CFIm loss promotes therapeutic resistance and PD-L1 expression. YAP/TAZ activation in multiple types of cancer cells increased resistance to different therapeutic agents2,4. Therefore, we next performed cell proliferation assay to examine whether CFIm loss changes the sensitivity of dTAG-NUDT21 HCT116 cells to different classes of therapeutic agents (verteporfin and MGH-CP1: YAP/TAZ inhibitors, Trametinib: MEK inhibitor; G007-LK: WNT pathway inhibitor; pictilisib: PI3K inhibitor). We found that dTAGV-1-treated dTAG-NUDT21 HCT116 cells were more resistant to Trametinib (Fig. 1F). This finding is consistent with a previous report that YAP promotes resistance to MEK inhibitors34. In contrast, the toxicity of G007-LK and pictilisib were similar between the DMSO and dTAGV-1 groups. Unexpectedly, dTAGV-1-treated dTAG-NUDT21 HCT116 cells were also more resistant to both YAP/TAZ inhibitors, verteporfin and MGH-CP1 (Fig. 1F), suggesting the possibility of additional .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 9 unidentified signaling changes that promote resistance to YAP/TAZ inhibitors35,36. PD-L1 is an immune checkpoint protein that cancer cells use to suppress T-cell activity and evade immunotherapy37. In breast cancer cells, TAZ activation increases PD-L1 expression and suppresses T-cell function38. Therefore, we next examined PD-L1 expression in shCPSF6- MDA cells grown in the presence or absence of doxycycline. Consistent with the previous report, we found increased expression of PD-L1 mRNA (Fig. 1G) and protein (Fig. 1H) in doxycycline-treated shCPSF6-MDA cells. Interestingly, treatment with MGH-CP1 completely abolished the increase in CTGF and CYR61 mRNA expression from doxycycline-induced CPSF6 knockdown (Fig. 1G, right panel, and Fig. S1F) but it only partially rescued PD-L1 mRNA expression (Fig. 1G, left panel). These results indicate that the increase in CTGF and CYR61 expression from CFIm loss is solely caused by higher YAP/TAZ transcriptional activities while the increase in PD-L1 expression has other contributory mechanisms. For example, post- transcriptional regulation of PD-L1 expression through 3′ UTR was previously reported37,39. Altogether, we demonstrate that cancer cells that lose CFIm become more resistant to different classes of chemical inhibitors and have stronger PD-L1 expression, which is beneficial for evading immune responses. LATS1/2 are required for activation of YAP/TAZ from CFIm loss. We next sought to investigate the mechanistic link between CFIm loss and YAP/TAZ activation. We first compiled a list of established YAP/TAZ regulators to examine their possible involvement in YAP/TAZ activation from CFIm loss in dTAG-NUDT21 HCT116 cells (Fig. 2A). We used siRNAs or chemical inhibitors to inhibit the function of each YAP/TAZ regulators and performed RT-qPCR to assay YAP/TAZ activation using CTGF mRNA expression as a surrogate. Because YAP/TAZ activities are regulated by cell density40, we first examined whether cell density affects YAP/TAZ activation from CFIm loss. We found that CFIm loss activated YAP/TAZ in both confluent and sub-confluent cells, indicating that YAP/TAZ activation from CFIm loss is independent of cell density (Fig. 2B). Next, we examined the roles of MST1/2 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 10 and NF2, both of which are YAP repressors in the Hippo pathway1. Treatment with XMU-MP-1, an MST1/2 inhibitor41, and siRNA-mediated knockdown of NF2 (Fig. S2A) both elevated the baseline CTGF expression, as was expected from the inhibition of a YAP repressor (Fig. 2C). In contrast, treatment with a CK2 inhibitor, CX4945, lowered the baseline CTGF expression as was expected from the inhibition of CK2, a YAP activator outside of the Hippo pathway42 (Fig. 2D). However, none of the 3 treatments blocked the increase in CTGF expression, indicating that MST1/2, NF2, and CK2 are not required for YAP/TAZ activation from CFIm loss. Next, we separately inhibited YAP and TAZ expression using siRNAs (Fig. S2B). Notably, the knockdown efficiency was lower for YAP compared with TAZ, which was likely due to a stronger role of YAP in HCT116 cell proliferation43. Nevertheless, both YAP and TAZ knockdown suppressed the increase in CTGF expression (Fig. 2E), supporting an essential role for both YAP and TAZ as expected. Lastly, we examined the role of LATS1/2 using TRULI, a specific LATS1/2 inhibitor44. TRULI increased the baseline CTGF expression as expected from inhibition of LATS1/2 (Fig. 2F). Furthermore, TRULI very strongly suppressed the increase in CTGF expression (Fig. 2F), indicating that LATS1/2 are necessary for YAP/TAZ activation from CFIm loss. To ensure these results are not specific to dTAG-NUDT21 HCT116 cells, we also examined the roles of cellular density, MST1/2, and LATS1/2 in YAP/TAZ activation from CFIm loss in shCPSF6-MDA cells. As in dTAG-NUDT21 HCT116 cells, the increase in CTGF expression from CFIm loss requires LATS1/2 but it is independent of cell density and MST1/2 in shCPSF6-MDA cells (Fig. 2G). Taken together, our results from two different cell models indicated that LATS1/2 are required for YAP/TAZ activation from CFIm loss. TAZ mRNA 3′ UTR shortening from CFIm loss increases its expression. In human, both YAP (encoded by the YAP1 gene) and TAZ (encoded by the WWTR1 gene) have short 3′ UTR mRNA isoforms (Fig. 3A). 3′ UTR shortening has been shown to relieve repressive regulation and increase gene expression45. To further explore the mechanistic .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 11 link between CFIm loss and YAP/TAZ activation, we next investigated the possibility that APA changes from CFIm loss contribute to higher expression of YAP and TAZ (Fig. 1B). We performed RT-qPCR experiments in both dTAG-NUDT21 HCT116 and shCPSF6-MDA cells to examine whether CFIm loss results in 3′ UTR shortening in YAP and TAZ. Indeed, CFIm loss resulted in a 90% or higher decrease in full-length (FL) 3′ UTR mRNA isoform for YAP and TAZ in both cell lines (Fig. 3B-C). We next adopted a previously published dual fluorescence reporter46 to measure how 3′ UTR shortening affects YAP and TAZ expression by flow cytometry (Fig. 3D). In this assay, we first cloned the 3′ UTR of interest into the empty reporter downstream of the mCherry coding sequence. We next transiently expressed the resulting reporter in HEK293T cells to measure the mCherry/GFP ratio, which reported the effects from the 3′ UTR of interest on gene expression. For YAP, the FL 3′ UTR reporter and the short 3′ UTR reporter generated similar mCherry/GFP ratios, suggesting that the 3′ UTR shortening from CFIm loss is less likely to alter YAP expression (Fig. 3E). In contrast, for TAZ, the FL 3′ UTR lowered mCherry expression compared with the short 3′ UTR (mCherry/GFP: FL, 0.71; short, 0.92) (Fig. 3E). In summary, we discovered that CFIm promotes the usage of full-length 3′ UTR in YAP and TAZ, and we further demonstrated that TAZ mRNA 3′ UTR shortening contributes to its increased expression with CFIm loss. Identification of a CFIm-NEDD4L-LATS1 regulatory axis The E3 ubiquitin ligases NEDD4 and NEDD4L were previously reported to activate YAP in different biological contexts47-49. Therefore, we next explored the possibility that NEDD4 and/or NEDD4L participated in YAP/TAZ activation from CFIm loss. We first examined our previously published PAPERCLIP profiling datasets13,31 to see whether CFIm regulates the APA of NEDD4 and/or NEDD4L. We found that CPSF6 knockdown resulted in strong 3′ UTR shortening in both NEDD4 and NEDD4L (Fig. 4A). We next performed RT-qPCR experiments in dTAG-NUDT21 HCT116 and shCPSF6-MDA cells to measure NEDD4 and NEDD4L mRNA isoform expression, which confirmed 3′ UTR shortening for both genes with CFIm loss (Fig. 4B- .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 12 C). Next, we performed western blots to examine NEDD4 and NEDD4L protein expression in dTAG-NUDT21 HCT116 and shCPSF6-MDA cells under both CFIm intact and CFIm loss conditions. We found that CFIm loss increased NEDD4 and NEDD4L protein expression in both cell lines (Fig. 4D-E). To examine whether 3′ UTR shortening in NEDD4 and NEDD4L mRNAs contributes to the increase in their protein expression, we generated FL and short 3′ UTR reporters for both NEDD4 and NEDD4L and performed the dual fluorescence reporter assay as described earlier (Fig. 3D). For both NEDD4 and NEDD4L, the FL 3′ UTR strongly suppressed mCherry expression compared with the short 3′ UTR (mCherry/GFP: NEDD4-FL, 0.59; NEDD4- short, 0.95; NEDD4L-FL, 0.54; NEDD4L-short, 0.81) (Fig. 4F). Altogether, our results support that CFIm loss shortens the 3′ UTR of both NEDD4 and NEDD4L mRNAs and increases their protein expression. We next sought to examine the requirement of NEDD4 and NEDD4L in YAP/TAZ activation from CFIm loss in both dTAG-NUDT21 HCT116 and shCPSF6-MDA cells by simultaneously suppressing both genes using siRNAs (Fig. S3A-B). Double knockdown of NEDD4 and NEDD4L did not affect the increase in CTGF expression from CFIm loss in dTAG- NUDT21 HCT116 cells (Fig. S3C), but in shCPSF6-MDA cells, it indeed blunted the induction of CTGF from CFIm loss (Fig. S3D), indicating that NEDD4 and/or NEDD4L participated in YAP/TAZ activation from CFIm loss. To further examine which of the two genes are necessary for YAP/TAZ activation from CFIm loss, we next knocked down NEDD4 and NEDD4L individually by siRNAs in shCPSF6-MDA cells (Fig. S3E). We found that the induction of CTGF from CFIm loss was not affected by NEDD4 knockdown but was suppressed by NEDD4L knockdown (Fig. 4G), indicating that NEDD4L is necessary for YAP/TAZ activation from CFIm loss in shCPSF6-MDA cells. We next transfected wildtype MDA-MB-231 cells with an NEDD4L-expressing plasmid, which induced CTGF and CYR61 expression (Fig. 4H), showing that NEDD4L overexpression alone is sufficient for YAP/TAZ activation. Lastly, because LATS1/2 are required for YAP/TAZ .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 13 activation from CFIm loss in shCPSF6-MDA cells (Fig. 2G), we sought to explore a possible regulation between NEDD4L and LATS1/2. Therefore, we generated a stable MDA-MB-231 cell line that expresses NEDD4L under a doxycycline-inducible promoter (referred to as TetOn- NEDD4L-MDA cells hereafter), and we compared LATS1/2 protein expression in TetOn- NEDD4L-MDA cells between the presence and absence of doxycycline. We found that NEDD4L overexpression had modest effects on LATS2 but clearly lowered LATS1 expression (Fig. 4I), which suggests that NEDD4L can activate YAP/TAZ signaling through LATS1 suppression. In summary, we demonstrated that NEDD4L is both necessary and sufficient for YAP/TAZ activation from CFIm loss, and our results altogether support that CFIm suppresses YAP/TAZ signaling through NEDD4L and LATS1 in MDA-MB-231 cells and its derivatives. CFIm loss sensitizes cells to mRNA CPA inhibition. Because CFIm loss promotes PD-L1 expression and therapeutic resistance to several chemical inhibitors (Fig. 1F~H), we next sought to find new therapeutic strategies for cancer with CFIm loss. JTE-607, which inhibits the endonuclease CPSF3 in the mRNA 3′-end processing complex, was recently shown to have anti-proliferation activities in some types of cancer such as acute myeloid leukemia (AML) and Ewing’s sarcoma19. Therefore, we treated dTAG-NUDT21 HCT116 cells with two different concentrations of JTE-607 in combination with either DMSO or dTAGV-1, and we compared cell proliferation after 48 h. Interestingly, CFIm loss enhanced the cytotoxic effects of JTE-607 at both concentrations (Fig. 5A), especially in the 10µM JTE-607 group, in which less than half the number of cells survived dTAGV-1 treatment compared to DMSO treatment (dTAGV-1: 37%, DMSO: 76%). Next, we wished to determine whether YAP activation plays a role in the increased JTE- 607 sensitivity from CFIm loss. To model YAP activation, we transduced HCT116 cells with a doxycycline-inducible YAP1S127/397A transgene, which is resistant to Hippo pathway-mediated phosphorylation and inactivation50. Doxycycline treatment of YAP1S127/397A HCT116 cells strongly activated CTGF expression (Fig. S4A) and increased Trametinib resistance (Fig. S4B), but it did .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 14 not result in higher cytotoxicity with simultaneous JTE-607 treatment (Fig. 5B), suggesting that YAP activation does not contribute to the increased JTE-607 sensitivity from CFIm loss. To address whether the increased vulnerability to JTE-607 from CFIm loss also extends to a different cell line, we next performed cell proliferation assay in shCPSF6-MDA cells. CFIm loss (Fig. S4C) again enhanced the cytotoxic effects of JTE-607 at both tested concentrations (Fig. 5C), with stronger effects in the 10µM JTE-607 group (Plus Dox: 40%, No Dox: 61%). Therefore, the increased vulnerability to JTE-607 from CFIm loss is not limited to HCT116 cells. Because JTE-607 increased DNA damage in Ewing’s sarcoma cells19, we next examined whether stronger DNA damage contributes to the higher toxicity observed in dTAG-NUDT21 HCT116 cells receiving both JTE-607 and dTAGV-1 treatment (Fig. 5A). Western blot experiments showed a strong increase of gamma-H2AX, a widely used marker for DNA damage51, in dTAGV-1 treated cells when compared to DMSO treated cells (Fig. 5D). Altogether, our results support that the combination of CFIm loss and CPSF3 inhibition exacerbates DNA damage and causes cytotoxicity. CFIm loss and CPSF3 inhibition together induce BRCAness. We hypothesized that suppression of DNA damage repair (DDR) gene expression contributes to the stronger DNA damage in JTE-607 treated dTAG-NUDT21 HCT116 cells with CFIm loss. To comprehensively compare the expression of 276 DDR genes52 in JTE-607 treated dTAG-NUDT21 HCT116 cells between the DMSO (CFIm intact) and dTAGV-1 (CFIm loss) conditions, we performed RNA sequencing (RNA-seq) analysis (Fig. 6A). We found that two DDR pathways, the Fanconi anemia (FA) and homology-directed repair (HDR) pathways, were selectively affected by CFIm loss (Fig. 6B). Moreover, the gene expression changes for both pathways between the two conditions are not random. More than 80% of genes in the FA and HDR pathways decreased their expression in dTAGV-1 treated dTAG-NUDT21 HCT116 cells (Table S1). LOF mutations in HDR and FA pathway genes can cause HDR defects and the .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 15 BRCAness phenotype21,53. Therefore, we hypothesized that CFIm loss and CPSF3 inhibition together induced the BRCAness phenotype in dTAG-NUDT21 HCT116 cells, and we performed a series of experiments to examine this hypothesis. We first performed RT-qPCR experiments to verify the suppression of 9 key genes in the FA and HDR pathways including BRCA1/2 and several other BRCAness genes such as RAD51D20,21 and FANCD221 (Fig. 6C) in dTAG- NUDT21 HCT116 cells receiving both dTAGV-1 and JTE-607. Notably, the expression of EZH2, which is not a DDR gene, remained at similar levels between DMSO and dTAGV-1 treated cells (Fig. 6C). A key feature of the BRCAness phenotype is an increased sensitivity to PARP inhibitors. Thus, we next examined dTAG-NUDT21 HCT116 cells for their sensitivity to Olaparib, a PARP inhibitor. The Olaparib sensitivity was indistinguishable between DMSO treated and dTAGV-1 treated dTAG-NUDT21 HCT116 cells without JTE-607 (Fig. 6D, left). In contrast, Olaparib selectively suppressed proliferation of dTAG-NUDT21 HCT116 cells in the presence of both JTE-607 and dTAGV-1 (Fig. 6D, right). These results provide support that CFIm loss and CPSF3 inhibition together induced the BRCAness phenotype in dTAG-NUDT21 HCT116 cells. Because the BRCAness phenotype is caused by HDR defects, we next performed DNA repair reporter assays54,55 using flow cytometry to examine whether the combination of JTE-607 and dTAGV-1 impairs HDR in dTAG-NUDT21 HCT116 cells. In these assays, successful DNA repair generates GFP expression, which reports the efficiency of the DNA repair pathway of interest. As expected, the combination of JTE-607 and dTAGV-1 treatment indeed strongly reduced HDR (Fig. 6E). In contrast, DNA repair by nonhomologous end-joining (NHEJ), a different DDR pathway that was not affected by JTE-607 and dTAGV-1 treatment in our RNA- seq analysis (Fig. 6B), remained intact (Fig. 6F). Lastly, we performed western blot analysis to examine whether Olaparib is synergistic to JTE-607 and dTAGV-1 in inducing DNA damage in dTAG-NUDT21 HCT116 cells. Addition of Olaparib indeed elevated gamma-H2AX expression in dTAG-NUDT21 HCT116 cells receiving .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 16 both JTE-607 and dTAGV-1 (Fig. 6G, left). Furthermore, mitomycin C (MMC), a DNA- crosslinking agent that BRCAness/HDR-deficient cancer is sensitive to21, also augmented DNA damage in dTAGV-1-treated dTAG-NUDT21 HCT116 cells in the presence of JTE-607 (Fig. 6G, right). Altogether, our results support that CFIm loss and CPSF3 inhibition together create a new vulnerability in cancer cells by impairing HDR and inducing the BRCAness phenotype (Fig. 6H). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 17

Activation of YAP/TAZ is common in human cancer and it often leads to therapeutic resistance and disease progression2,4. Therefore, understanding how cancer cells activate YAP/TAZ and their gene expression program is crucial to fight cancer therapy resistance. The Hippo pathway normally suppresses YAP/TAZ but the paucity of LOF mutations in the Hippo pathway components in cancer suggests the existence of alternative ways to activate YAP/TAZ. Our discovery that CFIm loss activates YAP/TAZ through the NEDD4L-LATS1 axis (Fig. 6H) has broad implications. First, it highlights the importance of post-transcriptional mechanisms in YAP/TAZ control, which is lesser known compared with the Hippo pathway. One example is the splicing factor ESRP2, which suppresses YAP activities by promoting the expression of Yap1 and Nf2 adult splicing isoforms in mouse liver56. We showed that CFIm suppresses YAP/TAZ activation by enforcing the use of full-length 3′ UTR in TAZ and NEDD4L (Fig. 3 and Fig. 4). Given the complexity of 3′ UTR regulation and the diverse types of cancer with YAP/TAZ activation4,57, future investigations might uncover additional mechanisms that are responsible for YAP/TAZ activation in cancer. Second, it shows that cancer cells activating YAP/TAZ through different mechanisms may have distinct responses to the same treatment. For example, we found that YAP mutation and CFIm loss both lead to YAP activation and Trametinib resistance in HCT116 cells (Fig. 1F and S4B). However, HCT116 cells with CFIm loss had higher resistance to YAP inhibitors but were more sensitive to JTE-607 (Fig. 1F and 5A). Therefore, separating cancer with high YAP/TAZ activities based on the YAP/TAZ activating mechanisms might be beneficial in revealing different treatment vulnerabilities. Increasing evidence suggests that CPSF3 inhibition might be a viable treatment strategy for certain types of cancer such as AML, Ewing’s sarcoma, ovarian cancer and pancreatic ductal adenocarcinoma17,19,58,59. Conversely, CPSF3 inhibition seems to be less effective in many other types of cancer. For example, 92 cancer cell lines were tested for JTE-607 sensitivity based on .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 18 cell viability in the original JTE-607 study19. Only 33 (36%) cell lines were sensitive to JTE-607 treatment (IC50 20µM), including HCT-116 cells that we used in our current study. Our results thus demonstrate that it is possible to sensitize cancer cells to CPSF3 inhibition by inducing CFIm loss (Fig. 5A), which might widen the use of CPSF3 inhibition to more cancer types. Tian and colleagues recently reported that higher mRNA CPA activities resulted in stronger JTE-607 inhibition17. In their study, the cellular CPA activities were measured with a tandem PAS reporter, and an elevated CPA activity was designated as a preference to utilize the proximal PAS over the distal PAS17. Because CFIm loss results in widespread proximal APA shifts10,12-14, our results that CFIm loss increases JTE-607 sensitivity are therefore consistent with their proposed model of JTE-607 inhibition. Inducing the BRCAness phenotype in HDR-proficient cancer cells has been shown to widen treatment options in breast and prostate cancer by increasing sensitivity to DNA damaging agents and PARP inihibitors60,61. Our study identified a new way to impair HDR and induce the BRCAness phenotype in cancer cells through the combination of CFIm loss and CPSF3 inhibition, and we provided multiple lines of evidence to support this finding (Fig. 6). Interestingly, a recent study showed that CPSF3 inhibition alone was sufficient to induce BRCAness phenotype through suppression of the HDR pathway in ovarian cancer59. Taken together, these results suggest that it is possible to induce the BRCAness phenotype in cancer cells through manipulating different components of the mRNA 3′ end processing complex, although the exact combination might vary in different types of cancer. Future studies to elucidate how different types of cells regulate their cellular CPA activities and their susceptibility to CPSF3 inhibition will provide important knowledge to improve cancer treatment. Taken together, our study reveals a link between mRNA 3′ end processing and the YAP/TAZ transcription program. Despite the success in developing potent YAP/TAZ inhibitors, it remains challenging to treat YAP/TAZ activation in cancer because single treatment with .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 19 YAP/TAZ inhibitors alone did not achieve sustained tumor suppression35,36. Our results suggest that the mRNA CPA inhibition might provide additional treatment benefits for cancer with CFIm loss. Besides cancer, the Hippo pathway and the YAP/TAZ transcription program also have important physiologic roles, especially in normal development and tissue regeneration3. Our

suggest that future studies on mRNA 3′ end processing and Hippo-YAP/TAZ in these contexts will provide important biological insights. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 20

We thank Amanda Clark for assistance in flow cytometry experiments. We thank the Health Sciences Sequencing Core at UPMC Children’s Hospital of Pittsburgh for high-throughput sequencing service, with special thanks to the assistant director, Will MacDonald. This work was supported by grants from the National Institutes of Health (R01NS113861 to H-W.H., R00CA207871 to H.U.O). Data analyses were supported by the University of Pittsburgh Center for Research Computing and the Extreme Science and Engineering Discovery Environment (XSEDE) supported by the National Science Foundation (OCI-1053575) via the Bridges2 system supported by the National Science Foundation (ACI- 1445606) at the Pittsburgh Supercomputing Center. Author Contributions: Conceptualization, H.-W.H.; Methodology, A.N. and H.-W.H.; Formal analysis, J.H. and H.-W.H.; Investigation, M.G. and H.-W.H.; Writing—original draft preparation: H.-W.H.; Writing—review and editing: M.G., J.H., A.N., H.U.O., and H.-W.H.; Supervision: H.U.O. and H.-W.H. Declaration of interests: The authors declare no competing interests. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 21 FIGURE LEGENDS Figure 1. Loss of CFIm activates YAP/TAZ signaling and promotes therapeutic resistance. (A) Illustrations showing the human CFIm. UGUA: CFIm binding motif. (B) Western blots showing higher YAP/TAZ expression and a lower Phospho-YAP (Ser127) (pYAP)/YAP ratio in dTAG-NUDT21 HCT116 cells treated with dTAGV-1 for 4 h. Tubulin: loading control. (C) Bar graphs showing the expression of YAP/TAZ targets measured by RT-qPCR in dTAG-NUDT21 HCT116 cells treated with either DMSO or dTAGV-1 for 24 h from 3 independent experiments (n=3). (D) Bar graphs showing the expression of YAP/TAZ targets measured by RT-qPCR in shCPSF6-MDA cells treated with doxycycline (Plus Dox) or remained untreated (No Dox) for 72 h from 3 independent experiments (n=3). (E) Bar graphs showing DsRED expression measured by RT-qPCR in 8xGTIIC–DsRED shCPSF6-MDA cells receiving different treatments for 72 h from 3 independent experiments (n=3). (F) Bar graphs showing the relative number of live dTAG-NUDT21 HCT116 cells treated with different reagents in combination with either DMSO or dTAGV-1 for 48 h from 5 independent experiments (n=5). (G) Bar graphs showing PD-L1 (left) or CTGF (right) expression measured by RT-qPCR in shCPSF6-MDA cells receiving different treatments for 72 h from 3 independent experiments (n=3). (H) Western blots showing higher PD-L1 expression in shCPSF6-MDA cells with doxycycline treatment for 72 h. dTAGV-1: 500nM. Doxycycline: 1 µg/mL. Verteporfin: 5µM. MGH-CP1: 10µM. Trametinib: 10nM. G007-LK: 10µM. Pictilisib: 1µM. SM04690: 100nM. Error bars indicate SEM. Statistical significance is determined by two-tailed t-test. NS: not significant, *: p<0.05, **: p<0.01. Figure 2. LATS1/2 are required for activation of YAP/TAZ signaling from CFIm loss. (A) An overview of the YAP/TAZ regulators, the chemical inhibitors or siRNAs used for inhibition, and the figure panels showing the experimental results. Orange: YAP signaling suppressors. Blue: YAP signaling activators. (B) Bar graphs showing CTGF expression measured by RT-qPCR in dTAG-NUDT21 HCT116 cells plated at different densities followed by DMSO or dTAGV-1 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 22 treatment for 24 h from 3 independent experiments (n=3). (C) Bar graphs showing CTGF expression measured by RT-qPCR in dTAG-NUDT21 HCT116 cells. (Left panel) Cells were treated with DMSO alone, dTAGV-1 alone, DMSO plus XMU-MP-1 or dTAGV-1 plus XMU-MP-1 for 24 h from 3 independent experiments (n=3). (Right panel) Cells were first transfected with control siRNAs (siCtrl) or siRNAs targeting NF2 (siNF2) for 48 h and then were treated with DMSO or dTAGV-1 for 24 h before RNA collection from 4 independent experiments (n=4). (D~F) Bar graphs showing CTGF expression measured by RT-qPCR in dTAG-NUDT21 HCT116 cells. (D) Cells were treated with DMSO alone, dTAGV-1 alone, DMSO plus CX4945 or dTAGV-1 plus CX4945 for 24 h from 4 independent experiments (n=4). (E) Cells were first transfected with control siRNAs (siCtrl), siRNAs targeting YAP (siYAP), or siRNAs targeting TAZ (siTAZ) for 48 h and then were treated with DMSO or dTAGV-1 for 24 h before RNA collection from 4 independent experiments (n=4). (F) Cells were treated with DMSO alone, dTAGV-1 alone, DMSO plus TRULI or dTAGV-1 plus TRULI for 24 h from 4 independent experiments (n=4). (G) Bar graphs showing CTGF expression measured by RT-qPCR in shCPSF6-MDA cells. (Left panel) Cells were plated at different densities followed by doxycycline treatment (Plus Dox) or remained untreated (No Dox) for 72 h from 4 independent experiments (n=4). (Middle panel) Cells were untreated, or treated with doxycycline alone, XMU-MP-1 alone, and doxycycline plus XMU-MP-1 for 72 h from 5 independent experiments (n=5). (Right panel) Cells were untreated, or treated with doxycycline alone, TRULI alone, and doxycycline plus TRULI for 72 h from 5 independent experiments (n=5). In (B) and (G), cells plated at high density reached complete confluence and cells plated at low density remained <50% confluence at RNA collection. XMU- MP-1: 2µM. CX4945: 2µM. TRULI: 10µM. Error bars indicate SEM. Statistical significance is determined by two-tailed t-test. NS: not significant, *: p<0.05, **: p<0.01. Figure 3. TAZ mRNA 3′ UTR shortening from CFIm loss contributes to its increased expression. (A) Illustrations showing the full-length 3′ UTR and short 3′ UTR mRNA isoforms of .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 23 YAP and TAZ as supported by Ensembl annotations and PAPERCLIP experiments. (B) Bar graphs showing expression of YAP and TAZ mRNA isoforms measured by RT-qPCR in dTAG- NUDT21 HCT116 cells treated with DMSO or dTAGV-1 for 24 h from 3 independent experiments (n=3). (C) Bar graphs showing expression of YAP and TAZ mRNA isoforms measured by RT- qPCR in shCPSF6-MDA cells treated with doxycycline (Plus Dox) or remained untreated (No Dox) for 72 h from 4 independent experiments (n=4). (D) Illustration showing the dual fluorescence reporter assay used to measure the effects from the 3′ UTR of interest on gene expression by flow cytometry. The mCherry/GFP ratio from the empty vector is set to 1. Arrows indicate the bi-directional promoter. (E) Bar graphs showing the mCherry/GFP ratio from YAP and TAZ 3′ UTR reporters in 293T cells from 4 independent experiments (n=4). Error bars indicate SEM. Statistical significance is determined by two-tailed t-test. NS: not significant, **: p<0.01. Figure 4. The identification of a CFIm-NEDD4L-LATS1 regulatory axis. (A) Results from our previously published PAPERCLIP experiments in HeLa cells and BE2C human neuroblastoma cells showing 3′ UTR shortening in NEDD4 and NEDD4L. Arrowheads: poly(A) sites identified by PAPERCLIP. P: Proximal. D: Distal. (B) Bar graphs showing expression of NEDD4 and NEDD4L mRNA isoforms measured by RT-qPCR in dTAG-NUDT21 HCT116 cells treated with DMSO or dTAGV-1 for 24 h from 3 independent experiments (n=3). (C) Bar graphs showing expression of NEDD4 and NEDD4L mRNA isoforms measured by RT-qPCR in shCPSF6-MDA cells treated with doxycycline (Plus Dox) or remained untreated (No Dox) for 72 h from 4 independent experiments (n=4). (D~E) Western blots showing increased NEDD4 and NEDD4L expression in (D) dTAG-NUDT21 HCT116 cells treated with dTAGV-1 for 24 h, and (E) shCPSF6-MDA cells with doxycycline treatment for 72 h. Dox, doxycycline. Tubulin: loading control. (F) Bar graphs showing the mCherry/GFP ratio from NEDD4 and NEDD4L 3′ UTR reporters in 293T cells from 4 independent experiments (n=4). (G) Bar graphs showing CTGF .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 24 expression measured by RT-qPCR in shCPSF6-MDA cells from 4 independent experiments (n=4). The cells were first transfected with control siRNAs (siCtrl), siRNAs targeting NEDD4 (siNEDD4), or siRNAs targeting NEDD4L (siNEDD4L) for 24 h and then were treated with doxycycline (Plus Dox) or remained untreated (No Dox) for 72 h before RNA collection. (H) Bar graphs showing CYR61 and CTGF expression measured by RT-qPCR in MDA-MB-231 cells transfected with an empty vector or an NEDD4L expression plasmid from 4 independent experiments (n=4). (I) Western blots showing decreased LATS1 expression in TetOn-NEDD4L- MDA cells with doxycycline treatment for 24 h. Dox, doxycycline. Tubulin: loading control. Error bars indicate SEM. Statistical significance is determined by one-tailed t-test (CTGF in panel H) and two-tailed t-test (all others). NS: not significant, *: p<0.05, **: p<0.01. Figure 5. CFIm loss sensitizes cells to CPSF3 inhibition by exacerbating DNA damage. (A) Bar graphs showing the relative number of live dTAG-NUDT21 HCT116 cells treated with different concentrations of JTE-607 in combination with DMSO or dTAGV-1 for 48 h from 4 independent experiments (n=4). (B) Bar graphs showing the relative number of live YAP1S127/397A HCT116 cells treated with 10µM JTE-607 alone (No Dox) or 10µM JTE-607 in combination with doxycycline (Plus Dox) for 48 h from 5 independent experiments (n=5). (C) Bar graphs showing the relative number of live shCPSF6-MDA cells treated with different concentrations of JTE-607 either alone or in combination with doxycycline for 72 h from 4 independent experiments (n=4). (D) Western blots showing increased gH2AX abundance in dTAG-NUDT21 HCT116 cells treated with 10µM JTE-607 and dTAGV-1 for 24 h. Tubulin: loading control. Error bars indicate SEM. Statistical significance is determined by two-tailed t- test. NS: not significant, *: p<0.05, **: p<0.01. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 25 Figure 6. CFIm loss and CPSF3 inhibition together induce HDR defects and the BRCAness phenotype. (A) Heatmaps showing relative expression of core DDR genes from different DDR pathways between DMSO treated (CFIm intact) and dTAGV-1 treated (CFIm loss) dTAG-NUDT21 HCT116 cells in the presence of 10µM JTE-607 measured by RNA-seq from 4 independent experiments (n=4). (B) Radar plots showing relative enrichment of different DDR pathways in dTAGV-1 treated (CFIm loss) dTAG-NUDT21 HCT116 cells in the presence of 10µM JTE-607. (C) Bar graphs showing the expression of genes in the FA and HDR pathways measured by RT-qPCR from the same experiments in (A) (n=4). EZH2: negative control. (D) Bar graphs showing the relative number of live dTAG-NUDT21 HCT116 cells treated with 10µM Olaparib in combination with DMSO, dTAGV-1, DMSO plus 1µM JTE-607 or dTAGV-1 pllus 1µM JTE-607 for 24 h from 4 independent experiments (n=4). (E and F) Bar graphs showing the DNA repair efficiency of (E) the HDR pathway and (F) the NHEJ pathway measured using reporters in dTAG-NUDT21 HCT116 cells receiving either DMSO or dTAGV-1 plus 10µM JTE- 607 from 4 independent experiments (n=4). VE-821, an ATR inhibitor that inhibits HDR, serves as a positive control in (E) and a negative control in (F). (G) Western blots showing elevated gH2AX abundance in dTAG-NUDT21 HCT116 cells treated with increasing concentrations of Olaparib (left panel) or with 100nM mitomycin C (MMC) (right panel) in the presence of 1µM JTE-607 and dTAGV-1 for 24 h. Tubulin: loading control. (H) Illustrations summarizing the cellular effects from CFIm loss. Error bars indicate SEM. Statistical significance is determined by two-tailed t-test. NS: not significant, *: p<0.05, **: p<0.01. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 26 STAR METHODS Key Resources Table Reagent type (species) or resource Designation Source or

Identifiers Additional information Cell line (H. sapiens) HEK293T/17 ATCC Cat#CRL-11268, RRID:CVCL_1926 Cell line (H. sapiens) MDA-MB-231 ATCC Cat#HTB-26, RRID:CVCL_0062 Cell line (H. sapiens) HCT-116 ATCC Cat#CCL-247, RRID:CVCL_0291 Cell line (H. sapiens) shCPSF6-MDA Herron et al. Elife, 2023 Cell line (H. sapiens) dTAG-NUDT21- HCT-116 This paper Cell line (H. sapiens) YAP1S127A/S397A- HCT-116 This paper Cell line (H. sapiens) 8xGTIIC-DsRED shCPSF6-MDA This paper Cell line (H. sapiens) TetOn-NEDD4L- MDA This paper Antibody Mouse monoclonal anti- alpha tubulin Millipore Cat#CP06; RRID:AB_2617116 1:3000 Antibody Mouse monoclonal anti- HA, clone 16B12 BioLegend Cat#901501; RRID:AB_2565006 1:3000 Antibody Mouse monoclonal anti- NUDT21 Proteintech Cat#66335-1-Ig; RRID:AB_2881715 1:2000 Antibody Rabbit polyclonal anti-CPSF6 Bethyl Laboratories Cat#A301-356A; RRID:AB_937781 1:2000 Antibody YAP/TAZ (D24E4) Rabbit mAb Cell Signaling Technology Cat#8418; RRID:AB_10950494 1:1000 Antibody YAP (D8H1X) Cell Cat#14074; 1:2000 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 27 Rabbit mAb Signaling Technology RRID:AB_2650491 Antibody Phospho-YAP (Ser127) (D9W2I) Rabbit mAb Cell Signaling Technology Cat#13008; RRID:AB_2650553 1:1000 Antibody Rabbit polyclonal anti-PD-L1/CD274 Proteintech Cat#17952-1-ap; RRID:AB_10597552 1:1000 Antibody Rabbit polyclonal anti-NEDD4 Proteintech Cat#21698-1-ap; RRID:AB_10858626 1:2000 Antibody Rabbit polyclonal anti-NEDD4L Proteintech Cat#13690-1-ap; RRID:AB_2149326 1:1000 Antibody LATS1 (C66B5) Rabbit mAb Cell Signaling Technology Cat#3477; RRID:AB_2133513 1:500 Antibody LATS2 (D83D6) Rabbit mAb Cell Signaling Technology Cat#5888; RRID:AB_10835233 1:500 Antibody Phospho-Histone H2A.X (Ser139) (20E3) Rabbity mAb Cell Signaling Technology Cat#9718; RRID:AB_2118009 1:2000 Strain, strain

NEB 5-alpha New England Biolabs Cat#C2987 Strain, strain

NEB Stable New England Biolabs Cat#C3040 Commercial assay or kit ProtoScript II First Strand cDNA Synthesis Kit New England Biolabs Cat#E6560 Commercial assay or kit Q5 High Fidelity DNA Polymerase New England Biolabs Cat#M0491 Commercial assay or kit NEBuilder HiFi DNA Assembly Master Mix New England Biolabs Cat#E2621 Commercial assay or kit 10% Bis-Tris NuPAGE gels Invitrogen Cat#NP0301BOX .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 28 Commercial assay or kit DNase I Invitrogen Cat#18068015 Commercial assay or kit Trizol reagent Invitrogen Cat#15596018 Commercial assay or kit PerfeCTa SYBR Green SuperMix QuantaBio Cat#95054 Commercial assay or kit PerfeCTa SYBR Green FastMix QuantaBio Cat#95072 Commercial assay or kit Cell Counting Kit 8 Abcam Cat#ab228554 Chemical compound, drug DMSO Sigma Cat#D2650 Chemical compound, drug Doxycycline Sigma Cat#D9891 Chemical compound, drug dTAGV-1 Tocris Bioscience Cat#6914 Chemical compound, drug SM04690 Selleckchem Cat#S8761 Chemical compound, drug Verteporfin Selleckchem Cat#S1786 Chemical compound, drug MGH-CP1 Selleckchem Cat#S9735 Chemical compound, drug Trametinib Selleckchem Cat#S2673 Chemical compound, drug G007-LK Selleckchem Cat#S7239 Chemical compound, drug Pictilisib Selleckchem Cat#S1065 Chemical compound, drug XMU-MP-1 Selleckchem Cat#S8334 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 29 Chemical compound, drug TRULI Selleckchem Cat#E1061 Chemical compound, drug CX4945 Selleckchem Cat#S2248 Chemical compound, drug JTE-607 Selleckchem Cat#E0314 Chemical compound, drug Olaparib MedChemEx press Cat#HY-10162 Chemical compound, drug VE-821 MedChemEx press Cat#HY-14731 Chemical compound, drug Mitomycin C Cayman Chemical Cat#11435 Chemical compound, drug DharmaFECT4 Horizon Discovery Cat#T-2004-02 Chemical compound, drug X-tremeGENE 9 Sigma Cat#XTG9-RO Chemical compound, drug FuGENE4K Promega Cat#E5911 Chemical compound, drug Lipofectamine 3000 Invitrogen Cat#L3000001 Chemical compound, drug TransIT-BrCa Mirus Bio Cat#MIR5500 Sequence- based reagent Table S2 This paper Primers used Sequence- based reagent Table S3 This paper siRNAs used Recombinant DNA reagent Table S4 Addgene or this paper Plasmids used or generated .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 30 Software, algorithm Prism 9.5 Graphpad Software https://www.graphpa d.com/; RRID:SCR_002798 Software, algorithm FlowJo v10.10.0 BD https://www.flowjo .com/solutions/flo wjo; RRID:SCR_008520 RESOURCE AVAILABILITY

availability Plasmids and cell lines generated for this study will be shared by the corresponding author upon request. Data availability RNA-seq data have been deposited at GEO under the accession GSE306466. EXPERIMENTAL MODEL AND SUBJECT DETAILS Cell culture HEK293T and MDA-MB-231 cells were grown in Dulbecco’s modified Eagle’s medium (DMEM). HCT116 cells were grown in McCoy’s 5A medium. All media were supplemented with 10% FBS and penicillin-streptomycin. Mycoplasma contamination in cell culture is screened with a commercial detection kit. HEK293T, MDA-MB-231, and HCT116 cells were obtained from ATCC with authentication and free of mycoplasma contamination. shCPSF6-MDA cells were previously generated in our lab31. 8xGTIIC-DsRED shCPSF6-MDA cells were generated by transducing shCPSF6-MDA cells with lentiviruses encoding an 8xGTIIC-DsRED YAP/TAZ reporter32 followed by hygromycin selection. TetOn-NEDD4L-MDA cells were generated by transducing rtTA3-expressing MDA-MB-231 cells with lentiviruses encoding doxycycline- inducible NEDD4L followed by puromycin selection. YAP1S127/397A HCT116 cells were generated by transducing HCT116 cells with lentiviruses encoding doxycycline inducible YAP1S127/397A .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 31 followed by puromycin selection. dTAG-NUDT21 HCT116 cells were generated by CRISPR- mediated insertion of dTAG degron-containing cassettes into both NUDT21 loci separately as previously described26. The transfected HCT116 cells were selected by Puromycin and Blasticidin, followed by single clone selection, DNA genotyping and Sanger sequencing to pick correctly inserted single clone cells. SiRNA transfection was performed using DharmaFECT4 (Horizon Discovery) with Silencer Select siRNAs (Invitrogen) at the final concentration of 25 nM following manufacturer’s instructions. All siRNAs used are listed in Table S2. Plasmid transfection was performed using X-tremeGENE9 (MilliporeSigma), FuGENE4K (Promega) and Lipofectamine 3000 (Invitrogen) following manufacturer’s instructions. Cell proliferation assays were performed in 96-well plates with Cell Counting Kit 8 (Abcam) following manufacturer’s instructions. The following chemicals were used at the indicated concentrations: Doxycycline (Sigma): 1 µg/mL. dTAGV-1 (Tocris Bioscience): 500nM. Verteporfin (Selleckchem): 5µM. MGH- CP1 (Selleckchem): 10µM. Trametinib (Selleckchem): 10nM. G007-LK (Selleckchem): 10µM. Pictilisib (Selleckchem): 1µM. SM04690 (Selleckchem): 100nM. XMU-MP-1 (Selleckchem): 2µM. CX4945 (Selleckchem): 2µM. TRULI (Selleckchem): 10µM. Olaparib (MedChemExpress): 20µM. VE-821 (MedChemExpress0: 10µM. Mitomycin C (Cayman Chemical): 100nM.

DETAILS RNA-seq Library preparation of total RNA was carried out by using the Lexogen 3′ mRNA-Seq QuantSeq FWD V2 Kit according to manufacturer’s instructions. Library preparation, quality control, and sequencing were carried out Health Sciences Sequencing Core at UPMC Children’s Hospital of Pittsburgh. cDNA libraries were sequenced on an Illumina NextSeq machine (1 × 100 nt). .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 32 Gene expression analysis RNA-seq data were generated for four DMSO-treated (CFIm-intact) and four dTAGV-1–treated (CFIm-loss) dTAG-NUDT21 HCT116 cell samples. Raw FASTQ files underwent quality assessment with FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/), evaluating per-base/tile sequence quality, adapter content, and duplication levels. Reads were preprocessed using fastp62 to trim auto-detected adapters and remove duplicates; sequences shorter than 25 bases after trimming were discarded. Clean reads were aligned to the Ensembl human reference genome (GRCh38, release 113) using STAR63. Gene-level counts were obtained with HTSeq64 by counting uniquely mapped reads overlapping annotated exonic regions. Protein-coding genes with ≥10 counts in at least three samples were retained for downstream analysis. Count data were normalized using size factors estimated by DESeq265, and a negative binomial generalized linear model was applied to identify differentially expressed genes (DEGs) between conditions. Genes with false discovery rate (FDR)-adjusted p < 0.05 were considered significant. Pathway enrichment analysis Gene Set Enrichment Analysis (GSEA) was performed with clusterProfiler66 to evaluate enrichment of DNA damage response (DDR) pathways. Ten curated DDR gene sets were obtained from Knijnenburg et al.52. Gene sets with Bonferroni-adjusted p < 0.05 were considered significantly enriched. SDS-PAGE and western blots 30~60 μg lysates from culture cells were separated on 10% Bis-Tris Novex NuPAGE gels (Invitrogen) and transferred to nitrocellulose membrane following standard procedures. The .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 33 following antibodies are used for western blotting: mouse monoclonal anti-alpha tubulin (Millipore, CP06), mouse monoclonal anti-HA (clone 16B12, BioLegend, 901501), rabbit polyclonal anti-CPSF6 (Bethyl Labs, A301-356A), YAP/TAZ (D24E4) Rabbit mAb (Cell Signaling Technology, 8418), YAP (D8H1X) Rabbit mAb (Cell Signaling Technology, 14074), Phospho- YAP (Ser127) (D9W2I) Rabbit mAb (Cell Signaling Technology, 13008), Phospho-Histone H2A.X (Ser139) (20E3) Rabbity mAb (Cell Signaling Technology, 9718), mouse monoclonal anti-NUDT21 (Proteintech, 66335-1-Ig), rabbit polyclonal anti-PD-L1/CD274 (Proteintech, 17952-1-ap), rabbit polyclonal anti-NEDD4 (Proteintech, 21698-1-ap), rabbit polyclonal anti- NEDD4L (Proteintech, 13690-1-ap), LATS1 (C66B5) Rabbit mAb (Cell Signaling Technology, 3477), LATS2 (D83D6) Rabbit mAb (Cell Signaling Technology, 5888). Reverse transcription and quantitative PCR (RT-qPCR) Quantitative PCR was performed using PerfeCTa SYBR Green SuperMix or PerfeCTa SYBR Green FastMix (QuantaBio) in triplicates. All primer sequences are listed in Table S2. For mRNA quantification, reverse transcription was performed using ProtoScript II First Strand cDNA Synthesis Kit (NEB) using d(T)23VN primer with DNase I (Invitrogen) digestion on 1 µg total RNA generated from Trizol (Invitrogen) extraction. The cycling parameters for qPCR were: 95ºC for 10 min. followed by 40 cycles of 95ºC for 15 sec., 58ºC for 30 sec., 72ºC for 20 sec. Quantification was calculated using the ΔΔCt method with ACTB as the endogenous control. Normalization between replicate experiments was calculated by measuring ACTB expression with GAPDH as the endogenous control using the ΔΔCt method. Cloning and constructs NEB HiFi DNA assembly following manufacturer’s instructions and standard cloning procedure (restriction digest, ligation, and transformation) was performed to generate desired constructs. All insert sequences were verified by Sanger sequencing. PITCh dTAG donor vectors for .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 34 NUDT21 targeting were generated by insertion of custom guide RNA sequences using the following plasmids: pX330-BbsI-PITCh (Addgene, 127875), pCRIS-PITChv2-BSD-dTAG (BRD4) (Addgene, 91792), pCRIS-PITChv2-Puro-dTAG (BRD4) (Addgene, 91793). The following plasmids were obtained from Addgene and were used without modification: a lentiviral vector encoding doxycycline-inducible YAP1-S127/397A (Addgene, 213585), NEDD4 and NEDD4L expression plasmids (Addgene, 27002 & 27000). The YAP/TAZ reporter, pTRE-8xGTIIC- DsRED-DD-Hygro, was generated by replacing Neomycin in pTRE-8xGTIIC-DsRED-DD (Addgene, 115798) with Hygromycin. The dual fluorescence reporter for 3′ UTR assay (pcDNA5-FRT-GFP-mCherry-NoGW) was generated by removing the Gateway cloning fragment from pcDNA5-FRT-GFP-mCherry-3pGW (Addgene, 53965). The following 3′ UTR reporters were generated by insertion of 3′ UTR sequences amplified from human genomic DNA between the multiple cloning sites in pcDNA5-FRT-GFP-mCherry-NoGW: hTAZ-FL, hTAZ-S, hYAP-FL, hYAP-S, hNEDD4-FL, hNEDD4-S, hNEDD4L-FL, hNEDD4L-S. Dual fluorescence 3′ UTR reporter assay For dual fluorescence 3′ UTR reporter assay, 293T cells were run on a FACSCanto II flow cytometer (BD) 24 h after transfection. The raw data were analyzed with FlowJo and then exported for calculation as previously described67. First, auto-fluorescence in both FITC and PE channels was calculated using untransfected cells, and the background is then defined as the mean plus 2 standard deviations of the auto-fluorescence signal. Next, for transfected cells, the measurements in both FITC and PE channels were subtracted with the background value, and cells with FITC and PE fluorescence levels below 0 after background subtraction were excluded from subsequent analyses. The data were next log-transformed and binned according to FITC levels, and the mean PE signal was calculated for each FITC bin. The average mCherry/GFP ratio was calculated using FITC bins between log2 to log4 with the empty vector set as 1. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 35 DNA repair reporter assays For HDR reporter assay, dTAG-NUDT21 HCT116 cells were transfected with pDRGFP (Addgene, #26475) and pCBASceI (Addgene, #26477) together for 6 hours. The transfected cells then received different treatments for 18 hours before running through a FACSCanto II flow cytometer (BD) to determine the percentage of GFP-positive cells. For NHEJ reporter assay, dTAG-NUDT21 HCT116 cells were transfected with pimEJ5GFP (Addgene, #44026) and pCBASceI (Addgene, #26477) together for 24 hours. The transfected cells then received different treatments for 24 hours before running through a FACSCanto II flow cytometer (BD) to determine the percentage of GFP-positive cells. The raw data were analyzed with FlowJo (BD). QUANTIFICATION AND STATISTICAL ANALYSIS Details of statistical tests are indicated below and in the Figure Legends. Statistical analyses were performed using R. For Figures 1C-G, 2B-G, 3B-C, 3E, 4B-C, 4F-G, 4H (CYR61), 5A-C, 6C-F, S1D-F, S2, S3, S4, statistical significance is determined by two-tailed Welch Two Sample t-test. For Figures 4H (CTGF), statistical significance is determined by one-tailed Welch Two Sample t- test. For all figures: *: p < 0.05, **: p < 0.01. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 36 SUPPLEMENTARY FIGURE LEGENDS Figure S1. Characterization of dTAG-NUDT21 HCT116 cells, related to Figure 1. (A) An illustration showing the NUDT21 loci in dTAG-NUDT21 HCT116 cells after CRISPR-mediated insertion of the dTAG degron-containing cassettes. Arrow: the endogenous NUDT21 promoter. PuroR: puromycin resistance gene. BlastR: Blasticidin resistance gene. (B) Western blots showing the expression of N-terminal tagged NUDT21 and endogenous CPSF6 expression in dTAG-NUDT21 HCT116 cells, which are derived from a single cell clone (2B4). α-NUDT21: NUDT21 antibody; α-HA: HA antibody. WT: wildtype HCT116 cells. (C) Western blots showing a complete loss of NUDT21 in 4 h dTAGV-1 treated dTAG-NUDT21 HCT116 cells. Tubulin: loading control. (D) Bar graphs showing TIMP2 isoform expression measured by RT-qPCR in dTAG- NUDT21 HCT116 cells treated with DMSO or dTAGV-1 for 24 h from 3 independent experiments (n=3). (E) Bar graphs showing CPSF6 expression measured by RT-qPCR in shCPSF6-MDA cells treated with doxycycline (Plus Dox) or remained untreated (No Dox) for 72 h from the same 3 independent experiments shown in Fig. 1D (n=3). (F) Bar graphs showing CYR61 expression measured by RT-qPCR in shCPSF6-MDA cells receiving different treatments for 72 h from the same 3 independent experiments shown in Fig. 1G (n=3). dTAGV-1: 500nM. Doxycycline: 1 µg/mL. MGH-CP1: 10µM. Error bars indicate SEM. Statistical significance is determined by two-tailed t-test. NS: not significant, *: p<0.05, **: p<0.01. Figure S2. siRNA knockdown efficiency in dTAG-NUDT21 HCT116 cells, related to Figure 2. (A) Bar graphs showing NF2 expression measured by RT-qPCR in dTAG-NUDT21 HCT116 cells from the same 4 independent experiments shown in the right panel of Fig. 2C (n=4). Cells were first transfected with control siRNAs (siCtrl) or siRNAs targeting NF2 (siNF2) for 48 h and then were treated with DMSO or dTAGV-1 for 24 h before RNA collection. (B) Bar graphs showing YAP and TAZ expression measured by RT-qPCR in dTAG-NUDT21 HCT116 cells from the .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 37 same 4 independent experiments shown in Fig. 2E (n=4). Cells were first transfected with control siRNAs (siCtrl), siRNAs targeting YAP (siYAP), or siRNAs targeting TAZ (siTAZ) for 48 h and then were treated with DMSO or dTAGV-1 for 24 h before RNA collection. dTAGV-1: 500nM. Doxycycline: 1 µg/mL. Error bars indicate SEM. Statistical significance is determined by two- tailed t-test. **: p<0.01. Figure S3. Examining YAP/TAZ activation in shCPSF6-MDA and dTAG-NUDT21 HCT116 cells with NEDD4 and NEDD4L double knockdown, related to Figure 4. (A) Bar graphs showing NEDD4 and NEDD4L expression measured by RT-qPCR in dTAG-NUDT21 HCT116 cells from 4 independent experiments (n=4). The cells were first transfected with control siRNAs (siCtrl) or siRNAs targeting both NEDD4 (siNEDD4) and NEDD4L (siNEDD4L) for 48 h and then were treated with DMSO or dTAGV-1 for 24 h before RNA collection from 4 independent experiments (n=4). (B) Bar graphs showing NEDD4 and NEDD4L expression measured by RT-qPCR in shCPSF6-MDA cells from 4 independent experiments (n=4). The cells were first transfected with control siRNAs (siCtrl) or siRNAs targeting both NEDD4 (siNEDD4) and NEDD4L (siNEDD4L) for 24 h and then were treated with doxycycline (Plus Dox) or remained untreated (No Dox) for 72 h before RNA collection. (C) Bar graphs showing CTGF expression measured by RT-qPCR in dTAG-NUDT21 HCT116 cells from the same 4 independent experiments shown in panel A (n=4). (D) Bar graphs showing CTGF expression measured by RT-qPCR in shCPSF6-MDA cells from the same 4 independent experiments shown in panel B (n=4). (E) Bar graphs showing NEDD4 or NEDD4L expression measured by RT-qPCR in shCPSF6-MDA cells from the same 3 independent experiments shown in Fig. 4G (n=3). Error bars indicate SEM. Statistical significance is determined by two-tailed t-test. NS: not significant, *: p<0.05, **: p<0.01. Figure S4. Doxycycline treatment of YAP1S127/397A HCT116 cells activates YAP signaling and promotes Trametinib resistance, related to Figure 5. (A) Bar graphs showing CTGF expression .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 38 measured by RT-qPCR in YAP1S127/397A HCT116 cells treated with treated with doxycycline (+Dox) or remained untreated (No Dox) for 24 h from 3 independent experiments (n=3). (B) Bar graphs showing the relative number of live YAP1S127/397A HCT116 cells treated with different chemical inhibitors, either alone (No Dox) or in combination with doxycycline (Plus Dox) for 48 h from 5 independent experiments (n=5). (C) Bar graphs showing CPSF6 expression measured by RT-qPCR in shCPSF6-MDA cells treated with doxycycline (+Dox) or remained untreated (No Dox) for 72 h from the same 3 independent experiments shown in Fig. 5G (n=4). Doxycycline: 1 µg/mL. Verteporfin: 5µM. MGH-CP1: 10µM. Trametinib: 10nM. Error bars indicate SEM. Statistical significance is determined by two-tailed t-test. NS: not significant, *: p<0.05, **: p<0.01. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 39 LIST OF NON-PDF SUPPLEMENTAL ITEMS Table S1. DDR gene expression changes from RNA-seq profiling, related to Figure 6. Table S2. List of primers, related to Methods. Table S3. List of siRNAs, related to Methods. Table S4. List of plasmids, related to Methods. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 40

1. Zhong, Z., Jiao, Z., and Yu, F.X. (2024). The Hippo signaling pathway in development and regeneration. Cell Rep 43, 113926. 10.1016/j.celrep.2024.113926. 2. Franklin, J.M., Wu, Z., and Guan, K.L. (2023). Insights into recent findings and clinical application of YAP and TAZ in cancer. Nat Rev Cancer 23, 512-525. 10.1038/s41568- 023-00579-1. 3. Zheng, Y., and Pan, D. (2019). The Hippo Signaling Pathway in Development and Disease. Dev Cell 50, 264-282. 10.1016/j.devcel.2019.06.003. 4. Piccolo, S., Panciera, T., Contessotto, P., and Cordenonsi, M. (2023). YAP/TAZ as master regulators in cancer: modulation, function and therapeutic approaches. Nat Cancer 4, 9-26. 10.1038/s43018-022-00473-z. 5. Van Raamsdonk, C.D., Bezrookove, V., Green, G., Bauer, J., Gaugler, L., O'Brien, J.M., Simpson, E.M., Barsh, G.S., and Bastian, B.C. (2009). Frequent somatic mutations of GNAQ in uveal melanoma and blue naevi. Nature 457, 599-602. 10.1038/nature07586. 6. Van Raamsdonk, C.D., Griewank, K.G., Crosby, M.B., Garrido, M.C., Vemula, S., Wiesner, T., Obenauf, A.C., Wackernagel, W., Green, G., Bouvier, N., et al. (2010). Mutations in GNA11 in uveal melanoma. N Engl J Med 363, 2191-2199. 10.1056/NEJMoa1000584. 7. Yu, F.X., Luo, J., Mo, J.S., Liu, G., Kim, Y.C., Meng, Z., Zhao, L., Peyman, G., Ouyang, H., Jiang, W., et al. (2014). Mutant Gq/11 promote uveal melanoma tumorigenesis by activating YAP. Cancer Cell 25, 822-830. 10.1016/j.ccr.2014.04.017. 8. Shi, Y., Di Giammartino, D.C., Taylor, D., Sarkeshik, A., Rice, W.J., Yates, J.R., 3rd, Frank, J., and Manley, J.L. (2009). Molecular architecture of the human pre-mRNA 3' processing complex. Mol Cell 33, 365-376. 10.1016/j.molcel.2008.12.028. 9. Yang, Q., Coseno, M., Gilmartin, G.M., and Doublie, S. (2011). Crystal structure of a human cleavage factor CFI(m)25/CFI(m)68/RNA complex provides an insight into poly(A) site recognition and RNA looping. Structure 19, 368-377. 10.1016/j.str.2010.12.021. 10. Zhu, Y., Wang, X., Forouzmand, E., Jeong, J., Qiao, F., Sowd, G.A., Engelman, A.N., Xie, X., Hertel, K.J., and Shi, Y. (2018). Molecular Mechanisms for CFIm-Mediated Regulation of mRNA Alternative Polyadenylation. Mol Cell 69, 62-74 e64. 10.1016/j.molcel.2017.11.031. 11. Yoon, Y., and Shi, Y. (2022). Human pre-mRNA 3' end processing: reconstituting is believing. Genes Dev 36, 106-107. 10.1101/gad.349453.122. 12. Ghosh, S., Ataman, M., Bak, M., Borsch, A., Schmidt, A., Buczak, K., Martin, G., Dimitriades, B., Herrmann, C.J., Kanitz, A., and Zavolan, M. (2022). CFIm-mediated alternative polyadenylation remodels cellular signaling and miRNA biogenesis. Nucleic Acids Res 50, 3096-3114. 10.1093/nar/gkac114. 13. Hwang, H.W., Park, C.Y., Goodarzi, H., Fak, J.J., Mele, A., Moore, M.J., Saito, Y., and Darnell, R.B. (2016). PAPERCLIP Identifies MicroRNA Targets and a Role of CstF64/64tau in Promoting Non-canonical poly(A) Site Usage. Cell Rep 15, 423-435. 10.1016/j.celrep.2016.03.023. 14. Li, W., You, B., Hoque, M., Zheng, D., Luo, W., Ji, Z., Park, J.Y., Gunderson, S.I., Kalsotra, A., Manley, J.L., and Tian, B. (2015). Systematic profiling of poly(A)+ transcripts modulated by core 3' end processing and splicing factors reveals regulatory rules of alternative cleavage and polyadenylation. PLoS Genet 11, e1005166. 10.1371/journal.pgen.1005166. 15. Weng, T., Ko, J., Masamha, C.P., Xia, Z., Xiang, Y., Chen, N.Y., Molina, J.G., Collum, S., Mertens, T.C., Luo, F., et al. (2019). Cleavage factor 25 deregulation contributes to .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 41 pulmonary fibrosis through alternative polyadenylation. J Clin Invest 129, 1984-1999. 10.1172/JCI122106. 16. Boreikaite, V., and Passmore, L.A. (2023). 3'-End Processing of Eukaryotic mRNA: Machinery, Regulation, and Impact on Gene Expression. Annu Rev Biochem 92, 199- 225. 10.1146/annurev-biochem-052521-012445. 17. Cui, Y., Wang, L., Ding, Q., Shin, J., Cassel, J., Liu, Q., Salvino, J.M., and Tian, B. (2023). Elevated pre-mRNA 3' end processing activity in cancer cells renders vulnerability to inhibition of cleavage and polyadenylation. Nat Commun 14, 4480. 10.1038/s41467-023-39793-8. 18. Tao, Y., Budhipramono, A., Huang, J., Fang, M., Xie, S., Kim, J., Khivansara, V., Dominski, Z., Tong, L., De Brabander, J.K., and Nijhawan, D. (2024). Anticancer benzoxaboroles block pre-mRNA processing by directly inhibiting CPSF3. Cell Chem Biol 31, 139-149 e114. 10.1016/j.chembiol.2023.10.019. 19. Ross, N.T., Lohmann, F., Carbonneau, S., Fazal, A., Weihofen, W.A., Gleim, S., Salcius, M., Sigoillot, F., Henault, M., Carl, S.H., et al. (2020). CPSF3-dependent pre-mRNA processing as a druggable node in AML and Ewing's sarcoma. Nat Chem Biol 16, 50-59. 10.1038/s41589-019-0424-1. 20. Ghanem, A., and Domchek, S.M. (2025). New Therapeutic Options for BRCA Mutant Patients. Annu Rev Med 76, 175-187. 10.1146/annurev-med-082523-083843. 21. Lord, C.J., and Ashworth, A. (2016). BRCAness revisited. Nat Rev Cancer 16, 110-120. 10.1038/nrc.2015.21. 22. Turner, N., Tutt, A., and Ashworth, A. (2004). Hallmarks of 'BRCAness' in sporadic cancers. Nat Rev Cancer 4, 814-819. 10.1038/nrc1457. 23. Liu, S., Wu, R., Chen, L., Deng, K., Ou, X., Lu, X., Li, M., Liu, C., Chen, S., Fu, Y., and Xu, A. (2023). CPSF6 regulates alternative polyadenylation and proliferation of cancer cells through phase separation. Cell Rep 42, 113197. 10.1016/j.celrep.2023.113197. 24. Masamha, C.P. (2023). The emerging roles of CFIm25 (NUDT21/CPSF5) in human biology and disease. Wiley Interdiscip Rev RNA 14, e1757. 10.1002/wrna.1757. 25. Tan, Y., Zheng, T., Su, Z., Chen, M., Chen, S., Zhang, R., Wang, R., Li, K., and Na, N. (2023). Alternative polyadenylation reprogramming of MORC2 induced by NUDT21 loss promotes KIRC carcinogenesis. JCI Insight 8. 10.1172/jci.insight.162893. 26. Nabet, B., Roberts, J.M., Buckley, D.L., Paulk, J., Dastjerdi, S., Yang, A., Leggett, A.L., Erb, M.A., Lawlor, M.A., Souza, A., et al. (2018). The dTAG system for immediate and target-specific protein degradation. Nat Chem Biol 14, 431-441. 10.1038/s41589-018- 0021-8. 27. Golden, R.J., Chen, B., Li, T., Braun, J., Manjunath, H., Chen, X., Wu, J., Schmid, V., Chang, T.C., Kopp, F., et al. (2017). An Argonaute phosphorylation cycle promotes microRNA-mediated silencing. Nature 542, 197-202. 10.1038/nature21025. 28. Jaeger, M.G., Schwalb, B., Mackowiak, S.D., Velychko, T., Hanzl, A., Imrichova, H., Brand, M., Agerer, B., Chorn, S., Nabet, B., et al. (2020). Selective Mediator dependence of cell-type-specifying transcription. Nat Genet 52, 719-727. 10.1038/s41588-020-0635- 0. 29. Masamha, C.P., Xia, Z., Yang, J., Albrecht, T.R., Li, M., Shyu, A.B., Li, W., and Wagner, E.J. (2014). CFIm25 links alternative polyadenylation to glioblastoma tumour suppression. Nature 510, 412-416. 10.1038/nature13261. 30. Zanconato, F., Forcato, M., Battilana, G., Azzolin, L., Quaranta, E., Bodega, B., Rosato, A., Bicciato, S., Cordenonsi, M., and Piccolo, S. (2015). Genome-wide association between YAP/TAZ/TEAD and AP-1 at enhancers drives oncogenic growth. Nat Cell Biol 17, 1218-1227. 10.1038/ncb3216. 31. Herron, R.S., Kunisky, A.K., Madden, J.R., Anyaeche, V.I., Maung, M.Z., and Hwang, H.W. (2023). A twin UGUA motif directs the balance between gene isoforms through .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 42 CFIm and the mTORC1 signaling pathway. Elife 12. 10.7554/eLife.85036. 32. Er, E.E., Valiente, M., Ganesh, K., Zou, Y., Agrawal, S., Hu, J., Griscom, B., Rosenblum, M., Boire, A., Brogi, E., et al. (2018). Pericyte-like spreading by disseminated cancer cells activates YAP and MRTF for metastatic colonization. Nat Cell Biol 20, 966-978. 10.1038/s41556-018-0138-8. 33. Bulos, M.L., Grzelak, E.M., Li-Ma, C., Chen, E., Hull, M., Johnson, K.A., and Bollong, M.J. (2023). Pharmacological inhibition of CLK2 activates YAP by promoting alternative splicing of AMOTL2. Elife 12. 10.7554/eLife.88508. 34. Lin, L., Sabnis, A.J., Chan, E., Olivas, V., Cade, L., Pazarentzos, E., Asthana, S., Neel, D., Yan, J.J., Lu, X., et al. (2015). The Hippo effector YAP promotes resistance to RAF- and MEK-targeted cancer therapies. Nat Genet 47, 250-256. 10.1038/ng.3218. 35. Paul, S., Hagenbeek, T.J., Tremblay, J., Kameswaran, V., Ong, C., Liu, C., Guarnaccia, A.D., Mondo, J.A., Hsu, P.L., Kljavin, N.M., et al. (2025). Cooperation between the Hippo and MAPK pathway activation drives acquired resistance to TEAD inhibition. Nat Commun 16, 1743. 10.1038/s41467-025-56634-y. 36. Sun, Y., Hu, L., Tao, Z., Jarugumilli, G.K., Erb, H., Singh, A., Li, Q., Cotton, J.L., Greninger, P., Egan, R.K., et al. (2022). Pharmacological blockade of TEAD-YAP reveals its therapeutic limitation in cancer cells. Nat Commun 13, 6744. 10.1038/s41467-022- 34559-0. 37. Arthur, A., Nejmi, S., Franchini, D.M., Espinos, E., and Millevoi, S. (2024). PD-L1 at the crossroad between RNA metabolism and immunosuppression. Trends Mol Med 30, 620- 632. 10.1016/j.molmed.2024.04.008. 38. Janse van Rensburg, H.J., Azad, T., Ling, M., Hao, Y., Snetsinger, B., Khanal, P., Minassian, L.M., Graham, C.H., Rauh, M.J., and Yang, X. (2018). The Hippo Pathway Component TAZ Promotes Immune Evasion in Human Cancer through PD-L1. Cancer Res 78, 1457-1470. 10.1158/0008-5472.CAN-17-3139. 39. Kataoka, K., Shiraishi, Y., Takeda, Y., Sakata, S., Matsumoto, M., Nagano, S., Maeda, T., Nagata, Y., Kitanaka, A., Mizuno, S., et al. (2016). Aberrant PD-L1 expression through 3'- UTR disruption in multiple cancers. Nature 534, 402-406. 10.1038/nature18294. 40. Zhao, B., Wei, X., Li, W., Udan, R.S., Yang, Q., Kim, J., Xie, J., Ikenoue, T., Yu, J., Li, L., et al. (2007). Inactivation of YAP oncoprotein by the Hippo pathway is involved in cell contact inhibition and tissue growth control. Genes Dev 21, 2747-2761. 10.1101/gad.1602907. 41. Fan, F., He, Z., Kong, L.L., Chen, Q., Yuan, Q., Zhang, S., Ye, J., Liu, H., Sun, X., Geng, J., et al. (2016). Pharmacological targeting of kinases MST1 and MST2 augments tissue repair and regeneration. Sci Transl Med 8, 352ra108. 10.1126/scitranslmed.aaf2304. 42. Guo, Y., Zhu, Z., Huang, Z., Cui, L., Yu, W., Hong, W., Zhou, Z., Du, P., and Liu, C.Y. (2022). CK2-induced cooperation of HHEX with the YAP-TEAD4 complex promotes colorectal tumorigenesis. Nat Commun 13, 4995. 10.1038/s41467-022-32674-6. 43. Corvaisier, M., Bauzone, M., Corfiotti, F., Renaud, F., El Amrani, M., Monte, D., Truant, S., Leteurtre, E., Formstecher, P., Van Seuningen, I., et al. (2016). Regulation of cellular quiescence by YAP/TAZ and Cyclin E1 in colon cancer cells: Implication in chemoresistance and cancer relapse. Oncotarget 7, 56699-56712. 10.18632/oncotarget.11057. 44. Kastan, N., Gnedeva, K., Alisch, T., Petelski, A.A., Huggins, D.J., Chiaravalli, J., Aharanov, A., Shakked, A., Tzahor, E., Nagiel, A., et al. (2021). Small-molecule inhibition of Lats kinases may promote Yap-dependent proliferation in postmitotic mammalian tissues. Nat Commun 12, 3100. 10.1038/s41467-021-23395-3. 45. Mayr, C., and Bartel, D.P. (2009). Widespread shortening of 3'UTRs by alternative cleavage and polyadenylation activates oncogenes in cancer cells. Cell 138, 673-684. 10.1016/j.cell.2009.06.016. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 43 46. Oikonomou, P., Goodarzi, H., and Tavazoie, S. (2014). Systematic identification of regulatory elements in conserved 3' UTRs of human transcripts. Cell Rep 7, 281-292. 10.1016/j.celrep.2014.03.001. 47. Bae, S.J., Kim, M., Kim, S.H., Kwon, Y.E., Lee, J.H., Kim, J., Chung, C.H., Lee, W.J., and Seol, J.H. (2015). NEDD4 controls intestinal stem cell homeostasis by regulating the Hippo signalling pathway. Nat Commun 6, 6314. 10.1038/ncomms7314. 48. Wei, Y., Yee, P.P., Liu, Z., Zhang, L., Guo, H., Zheng, H., Anderson, B., Gulley, M., and Li, W. (2020). NEDD4L-mediated Merlin ubiquitination facilitates Hippo pathway activation. EMBO Rep 21, e50642. 10.15252/embr.202050642. 49. Zhuge, R., Wang, C., Wang, J., Yu, S., Liao, L., and Zheng, X. (2023). hCINAP regulates the differentiation of embryonic stem cells by regulating NEDD4 liquid-liquid phase- separation-mediated YAP1 activation. Cell Rep 42, 111935. 10.1016/j.celrep.2022.111935. 50. Edwards, A.C., Stalnecker, C.A., Jean Morales, A., Taylor, K.E., Klomp, J.E., Klomp, J.A., Waters, A.M., Sudhakar, N., Hallin, J., Tang, T.T., et al. (2023). TEAD Inhibition Overcomes YAP1/TAZ-Driven Primary and Acquired Resistance to KRASG12C Inhibitors. Cancer Res 83, 4112-4129. 10.1158/0008-5472.CAN-23-2994. 51. Mah, L.J., El-Osta, A., and Karagiannis, T.C. (2010). gammaH2AX: a sensitive molecular marker of DNA damage and repair. Leukemia 24, 679-686. 10.1038/leu.2010.6. 52. Knijnenburg, T.A., Wang, L., Zimmermann, M.T., Chambwe, N., Gao, G.F., Cherniack, A.D., Fan, H., Shen, H., Way, G.P., Greene, C.S., et al. (2018). Genomic and Molecular Landscape of DNA Damage Repair Deficiency across The Cancer Genome Atlas. Cell Rep 23, 239-254 e236. 10.1016/j.celrep.2018.03.076. 53. Murai, J., and Pommier, Y. (2023). BRCAness, Homologous Recombination Deficiencies, and Synthetic Lethality. Cancer Res 83, 1173-1174. 10.1158/0008- 5472.CAN-23-0628. 54. Pierce, A.J., Johnson, R.D., Thompson, L.H., and Jasin, M. (1999). XRCC3 promotes homology-directed repair of DNA damage in mammalian cells. Genes Dev 13, 2633- 2638. 10.1101/gad.13.20.2633. 55. Bennardo, N., Cheng, A., Huang, N., and Stark, J.M. (2008). Alternative-NHEJ is a mechanistically distinct pathway of mammalian chromosome break repair. PLoS Genet 4, e1000110. 10.1371/journal.pgen.1000110. 56. Bangru, S., Arif, W., Seimetz, J., Bhate, A., Chen, J., Rashan, E.H., Carstens, R.P., Anakk, S., and Kalsotra, A. (2018). Alternative splicing rewires Hippo signaling pathway in hepatocytes to promote liver regeneration. Nat Struct Mol Biol 25, 928-939. 10.1038/s41594-018-0129-2. 57. Mitschka, S., and Mayr, C. (2022). Context-specific regulation and function of mRNA alternative polyadenylation. Nat Rev Mol Cell Biol 23, 779-796. 10.1038/s41580-022- 00507-5. 58. Alahmari, A.A., Chaubey, A.H., Jonnakuti, V.S., Tisdale, A.A., Schwarz, C.D., Cornwell, A.C., Maraszek, K.E., Paterson, E.J., Kim, M., Venkat, S., et al. (2024). CPSF3 inhibition blocks pancreatic cancer cell proliferation through disruption of core histone mRNA processing. RNA 30, 281-297. 10.1261/rna.079931.123. 59. Shen, P., Ye, K., Xiang, H., Zhang, Z., He, Q., Zhang, X., Cai, M.C., Chen, J., Sun, Y., Lin, L., et al. (2023). Therapeutic targeting of CPSF3-dependent transcriptional termination in ovarian cancer. Sci Adv 9, eadj0123. 10.1126/sciadv.adj0123. 60. Gui, F., Jiang, B., Jiang, J., He, Z., Tsujino, T., Takai, T., Arai, S., Pana, C., Kollermann, J., Bradshaw, G.A., et al. (2025). Acute BRCAness induction and AR pathway blockage through CDK12/7/9 degradation enhances PARP inhibitor sensitivity in prostate cancer. Sci Adv 11, eadu0847. 10.1126/sciadv.adu0847. 61. Quereda, V., Bayle, S., Vena, F., Frydman, S.M., Monastyrskyi, A., Roush, W.R., and .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint 44 Duckett, D.R. (2019). Therapeutic Targeting of CDK12/CDK13 in Triple-Negative Breast Cancer. Cancer Cell 36, 545-558 e547. 10.1016/j.ccell.2019.09.004. 62. Chen, S., Zhou, Y., Chen, Y., and Gu, J. (2018). fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884-i890. 10.1093/bioinformatics/bty560. 63. Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson, M., and Gingeras, T.R. (2013). STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15-21. 10.1093/bioinformatics/bts635. 64. Putri, G.H., Anders, S., Pyl, P.T., Pimanda, J.E., and Zanini, F. (2022). Analysing high- throughput sequencing data in Python with HTSeq 2.0. Bioinformatics 38, 2943-2945. 10.1093/bioinformatics/btac166. 65. Love, M.I., Huber, W., and Anders, S. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550. 10.1186/s13059-014- 0550-8. 66. Yu, G., Wang, L.G., Han, Y., and He, Q.Y. (2012). clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16, 284-287. 10.1089/omi.2011.0118. 67. Luna, J.M., Scheel, T.K., Danino, T., Shaw, K.S., Mele, A., Fak, J.J., Nishiuchi, E., Takacs, C.N., Catanese, M.T., de Jong, Y.P., et al. (2015). Hepatitis C virus RNA functionally sequesters miR-122. Cell 160, 1099-1110. 10.1016/j.cell.2015.02.025. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint dTAG-NUDT21 HCT116 DMSOdTAG HA-NUDT21 TAZ Tubulin YAP pYAP 1.0 2.0 1.0 0.8pYAP/YAP: 1.0 1.3 CTGF CYR61 Spacer1 Spacer2 0.0 0.5 1.0 1.5 2.0 Relative Expression 152MDA YAP targets 3Rep No Dox Plus Dox shCPSF6-MDA No Dox Plus Dox CTGF Relative expression CYR61 * * UGUA5′ 3′ NUDT21 CPSF6 A Fig. 1 C FE G B shCPSF6-MDA - +Dox: 1.0 0.1 CPSF6 1.0 1.6 Tubulin PD-L1 - +Dox: CPSF6 1.0 1.6 Tubulin PD-L1 H D shCPSF6-MDA No Dox + DMSO Plus Dox + DMSO Plus Dox + MGH-CP1 PD-L1 expression * ** C X Y 0 1 2 3 4 5 6 7 Relative Expression 152MDA PDL1 3Rep edited PDL1 * CTGF expression C X Y 0.0 0.5 1.0 1.5 2.0 2.5 Relative Expression 152MDA CTGF 3Rep CTGF ** ** NS NoDoxPlusDoxSM04690 0.0 0.5 1.0 1.5 2.0 2.5 Relative Expression A152379MDA dsRED 3Rep dsRED 8xGTIIC–DsRED shCPSF6-MDA DsRED expression * Untreated Dox SM04690 ** CTGF CYR61 AMOTL2 Spacer2 0 1 2 3 4 5 6 Relative Expression 2B4 YAP targets 3Rep DMSO dTAG dTAG-NUDT21 HCT116 CTGF Relative expression CYR61 AMOTL2 * * ** DMSO dTAGV-1 dTAG-NUDT21 HCT116 ** Verteporfin MGHCP1 TrametinibSpaceHolder 0 20 40 60 80 100 120 Cell number (% of untreated) 2B4_CCK_5Rep_Part1 DMSO dTAG G007LK Pictilisib Cisplatin SpaceHolder 0 20 40 60 80 100 120 Cell number (% of untreated) 2B4_CCK_5Rep_Part2 DMSO dTAG VerteporfinMGH-CP1Trametinib Cell number (% of untreated) * * G007-LK Pictilisib NS NS DMSO dTAGV-1 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint MST1/2 LATS1/2 NF2 YAP/TAZ CK2 siNF2XMU-MP-1 (Fig. 2C) (Fig. 2C) CX4945 siYAP/TAZ TRULI (Fig. 2D) (Fig. 2E) (Fig. 2F) YAP suppressor YAP activator The Hippo Pathway shCPSF6-MDA DMSO TRULI Spacer2 Spacer2 0 1 2 3 Relative Expression 152MDA inhibitor TRULI 5Rep V2 No Dox Plus Dox 5X 0.5X Spacer2 Spacer2 0 1 2 3 Relative Expression 152MDA YAP density 3Rep V2 No Dox Plus Dox High density CTGF expression ** Low density * No inhibitor XMU- MP-1 No Dox Plus Dox * No inhibitor TRULI * DMSO XMU-MP-1 Spacer2 Spacer2 0 1 2 3 4 5 6 7 Relative Expression 152MDA inhibitor XMU 5Rep V2 No Dox Plus Dox * NS dTAG-NUDT21 HCT116 DMSO dTAGV-1 siCtrl siNF2 Spacer1 Spacer2 0 1 2 3 4 5 6 7 8 Relative Expression 2B4 siNF2 CTGF 4Rep edited DMSO dTAG No inhibitorXMU-MP-1 Spacer2 Spacer2 0 1 2 3 4 5 6 7 8 Relative Expression 2B4 inhibitor XMU 2uM 3Rep edited DMSO dTAG No inhibitor CTGF expression * ** XMU- MP-1 ** ** siCtrl siNF2 dTAG-NUDT21 HCT116 DMSO dTAGV-1 3X 0.3X Spacer2 Spacer2 0 1 2 3 4 5 6 7 8 Relative Expression 2B4 density 3Rep edited DMSO dTAG High density CTGF expression * ** Low density dTAG-NUDT21 HCT116 DMSO dTAGV-1 No inhibitor CX4945 Spacer2 Spacer2 0 1 2 3 Relative Expression 2B4 inhibitorII 4Rep edited DMSO dTAG No inhibitor CTGF expression * ** CX4945 dTAG-NUDT21 HCT116 DMSO dTAGV-1 siCtrl siYAP siTAZ Spacer2 0 1 2 3 4 5 6 Relative Expression 2B4 siYAPTAZ CTGF 4Rep edited DMSO dTAG siCtrl CTGF expression siYAP siTAZ * NS NS dTAG-NUDT21 HCT116 DMSO dTAGV-1 No inhibitor TRULI Spacer2 Spacer2 0 1 2 3 Relative Expression 2B4 inhibitor 3Rep DMSO dTAG No inhibitor CTGF expression * TRULI NS A Fig. 2 ED F B G C .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint Dual fluorescence reporter assay mCherryGFP 3′ UTR mCherryGFP mCherry/GFP 1 ? DMSO dTAG Spacer2 Spacer2 0.0 0.5 1.0 1.5 Relative Expression YAP isoform, 3Rep YAP1-common YAP1-distal dTAG-NUDT21 HCT116 Both isoforms Full-length isoform YAP expression DMSO ** dTAG NS TAZ expression DMSO ** dTAG NS DMSO dTAG Spacer2 Spacer2 0.0 0.5 1.0 1.5 Relative Expression TAZ isoform, 3Rep TAZ-common TAZ-distal shCPSF6-MDA Both isoforms Full-length isoform YAP expression NoDox ** PlusDox ** TAZ expression NoDox ** PlusDox ** NoDox PlusDox Spacer2 Spacer2 0 1 2 3 4 5 6 Relative Expression 152MDA YAP isoform 4Rep YAP-common YAP-distal NoDox PlusDox Spacer2 Spacer2 0 1 2 3 Relative Expression 152MDA TAZ isoform 4Rep TAZ-common TAZ-distal YAP1 1kb Full-length isoform Short isoform WWTR1 (TAZ) 1kb Full-length isoform Short isoform A Fig. 3 E B C D YAP_B TAZ_A KMT2A Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression A405_A408 4Rep Flipped Full-length UTR Short UTR HEK293T Full-length 3′ UTR Short 3′ UTR mCherry/GFP YAP TAZ ** ** ** NS ** NS .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint Relative expression NEDD4L Empty vector MDA-MB-231 * CYR61 CTGF ** CYR61 CTGF Spacer2 Spacer2 0.0 0.5 1.0 1.5 2.0 Relative Expression MDA A373 4Rep Empty vector +NEDD4L shCPSF6-MDA No Dox Plus Dox siCtrl siNEDD4 siNEDD4L Spacer2 0.0 0.5 1.0 1.5 2.0 2.5 Relative Expression 152MDA siNEDD4_4L CTGF 3Rep V2 NoDox PlusDox siCtrl CTGF expression * NS siNEDD4 * siNEDD4L HEK293T Full-length 3′ UTR Short 3′ UTR NEDD4 NEDD4L Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression A329_A337 3Rep edited Alt Full-length UTR Short UTR mCherry/GFP NEDD4 NEDD4L * **** NS **** dTAG-NUDT21 HCT116 Both isoforms Distal isoform NEDD4 expression DMSO dTAG Spacer2 Spacer2 0.0 0.5 1.0 1.5 2.0 Relative Expression NEDD4 isoform, 3Rep, New NEDD4-common NEDD4-distal DMSO * dTAG NS NEDD4L expression DMSO * dTAG NS DMSO dTAG Spacer2 Spacer2 0.0 0.5 1.0 1.5 2.0 Relative Expression NEDD4L isoform, 3Rep, New NEDD4L-common NEDD4L-distal Last exon HeLa, siCPSF6 HeLa, siCtrl NEDD4 P D shCPSF6-BE2C, NoDox shCPSF6-BE2C, +Dox NEDD4L P D 1kb 1kb PAPERCLIP shCPSF6-MDA Both isoforms Distal isoform NoDox PlusDox Spacer2 Spacer2 0 1 2 3 4 Relative Expression 152MDA NEDD4 isoform 4Rep V2 NEDD4-common NEDD4-distal NEDD4 expression NoDox ** PlusDox ** NEDD4L expression NoDox ** PlusDox ** NoDox PlusDox Spacer2 Spacer2 0 1 2 3 4 Relative Expression 152MDA NEDD4L isoform 4Rep V2 NEDD4L-common NEDD4L-distal A Fig. 4 C E B F D G H shCPSF6-MDA - + CPSF6 NEDD4 1.0 1.3 NEDD4L Tubulin 1.0 1.2 Dox: dTAG-NUDT21 HCT116 HA-NUDT21 DMSO dTAG NEDD4 NEDD4L Tubulin 1.0 1.6 1.0 1.9 shCPSF6-MDA - + CPSF6 NEDD4 1.0 1.3 NEDD4L Tubulin 1.0 1.2 Dox: dTAG-NUDT21 HCT116 HA-NUDT21 DMSO dTAG NEDD4 NEDD4L Tubulin 1.0 1.6 1.0 1.9 I TetOn-NEDD4L-MDA C13p99, WB, Version 2 199-399MDA - + NEDD4L TAZ LATS1 1.0 0.9 LATS2 YAP 1.0 1.2 1.0 0.6 Dox: Tubulin 1.0 1.1 1.0 1.4 199-399MDA - + NEDD4L LATS1 1.0 0.9 LATS2 1.0 0.6 Dox: Tubulin 1.0 1.4 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint JTE25 JTE10 Holder Holder 0 20 40 60 80 100 120 Cell number (% of untreated) A152MDA_JTE_4Rep edited No Dox Plus Dox ** Cell number (% of untreated) shCPSF6-MDA JTE-607 25µM JTE-607 10µM ** No Dox Plus Dox YAP1S127/397A-HCT116 Cell number (% of untreated) No Dox Plus Dox JTE-607 10µM NS JTE607 TrametinibPlaceHolderPlaceHolder 0 20 40 60 80 100 120 Cell number (% of untreated) A386HCT_CCKIV_5Rep_V2 edited NoDox PlusDox JTE10 JTE1 MGH-CP1 Trametinib 0 20 40 60 80 100 120 Cell number (% of untreated) 2B4_CCKIII_4Rep_Part1_V2 edited DMSO dTAG** Cell number (% of untreated) * dTAG-NUDT21 HCT116 DMSO dTAGV-1 JTE-607 10µM JTE-607 1µM dTAG-NUDT21 HCT116 JTE-607 DMSO dTAG HA-NUDT21 Tubulin γH2AX A Fig. 5 C D B .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint BER DR DS FA HDR MMR NER NHEJ TLS APEX2 FEN1 P ARP1 POLB TDG UNG MGMT A TM A TR A TRIP CHEK1 FANCA FANCB FANCC FANCD2 FANCI FANCL FANCM UBE2T BARD1 BLM BRCA1 BRCA2 BRIP1 EME1 GEN1 NBN P ALB2 RAD51 RAD52 RBBP8 XRCC2 XRCC3 EXO1 MLH1 MSH2 MSH3 MSH6 ERCC6 POLE POLE3 XP A NHEJ1 PRKDC XRCC4 XRCC6 POLQ REV3L SHPRH dT AG Log2FC −2 −1 0 1 − Log10(FDR − p) 2 10 25 BER DR DS FA HDR MMR NER NHEJ TLS APEX2 FEN1 P ARP1 POLB TDG UNG MGMT A TM A TR A TRIP CHEK1 FANCA FANCB FANCC FANCD2 FANCI FANCL FANCM UBE2T BARD1 BLM BRCA1 BRCA2 BRIP1 EME1 GEN1 NBN P ALB2 RAD51 RAD52 RBBP8 XRCC2 XRCC3 EXO1 MLH1 MSH2 MSH3 MSH6 ERCC6 POLE POLE3 XP A NHEJ1 PRKDC XRCC4 XRCC6 POLQ REV3L SHPRH dT AG Log2FC −2 −1 0 1 − Log10(FDR − p) 2 10 25 Olaparib onlyOlaparib + JTE PlaceHolderPlaceHolder 0 20 40 60 80 100 120 Cell number (% of untreated) 2B4_CCKVIII_10uMCombined_4Rep DMSO dTAG Cell number (% of untreated) NS *NS * Olaparib DMSO dTAGV-1 DMSO+JTE-607 dTAGV-1+JTE-607 dTAG-NUDT21 HCT116 DMSO+JTE-607 dTAGV-1+JTE-607 FANCB FANCL FANCD2 EZH2FANCI JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE BRCA1 4Rep DMSO dTAG JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE BRCA2 4Rep DMSO dTAG Relative expression BRCA1 ** BRCA2 ** JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE BLM 4Rep DMSO dTAG BLM ** RAD51D ** BARD1 ** JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE BARD1 4Rep DMSO dTAG Relative expression ** ** ** JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE EZH2 4Rep DMSO dTAG NS** JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE RAD51D 4Rep DMSO dTAG JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE FANCB 4Rep DMSO dTAG JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE FANCL 4Rep DMSO dTAG JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE FANCD2 4Rep DMSO dTAG JTE607 Spacer2 Spacer2 Spacer2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 JTE FANCI 4Rep DMSO dTAG 0 5 10 15 20 HDR FA MMR BER NP NER Others NHEJ TLS DR − log2(Bonferroni − p) − log2(0.05) HDR efficiency (normalized to DMSO) DMSO dTAGV1+JTE-607 VE-821 ** ** MM TJ VE 0.0 0.5 1.0 1.5 Cell number (% of untreated) 2B4_A413_4Rep Normalized GFP+ MM TJ VE 0.0 0.5 1.0 1.5 Cell number (% of untreated) 2B4_A462_4Rep_V5 Normalized GFP+ NHEJ efficiency (normalized to DMSO) DMSO dTAGV1+JTE-607 VE-821 NS NS dTAGDMSO Olaparib(µM): γH2AX Tubulin JTE-607 - - 10 20 1.0 2.2 2.5 3.0 dTAGDMSO JTE-607 + MMC 1.0 1.7 Therapeutic resistance CFIm loss YAP/TAZ activation DNA damage TAZ NEDD4L FA & HDR pathway genes CPSF3 inhibition HDR defects BRCAness Fig. 6A B E C G H D F .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint CPSF6 expression No Dox Plus Dox shCPSF6-MDA ** NoDox PlusDox 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression A152MDA CPSF6 3Rep edited Mono CPSF6 C X Y 0.0 0.5 1.0 1.5 2.0 2.5 Relative Expression 152MDA CYR61 3Rep CYR61 shCPSF6-MDA No Dox + DMSO Plus Dox + DMSO Plus Dox + MGH-CP1CYR61 expression ** * NS PuroR P2A FKBP12F36V2x HA NUDT21 BlastR P2A FKBP12F36V2x HA NUDT21 A Fig. S1 B C D E F Tubulin WT 2B4 CPSF6 α-NUDT21 25 37 α-HA 25 37 Relative expression DMSO dTAG Both TIMP2 isoforms TIMP2 distal isoform DMSO dTAG 0.0 0.5 1.0 1.5 2.0 Relative Expression TIMP2 isoform, 3Rep, edited TIMP2-common TIMP2-long * * dTAG-NUDT21 HCT116 Tubulin HA-NUDT21 2h 4h0h 1h 1 0.7 0.1 0 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint A Fig. S2B dTAG-NUDT21 HCT116 siCtrl siYAP 0.0 0.5 1.0 1.5 Relative Expression 2B4 siYAP YAP 4Rep edited Mono YAP YAP expression siCtrl siYAP ** TAZ expression siCtrl siTAZ ** siCtrl siTAZ 0.0 0.5 1.0 1.5 Relative Expression 2B4 siTAZ TAZ 4Rep edited Mono TAZ siCtrl siNF2 0.0 0.5 1.0 1.5 Relative Expression 2B4 siNF2 NF2 4Rep edited Mono NF2 NF2 expression siCtrl siNF2 ** dTAG-NUDT21 HCT116 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint dTAG-NUDT21 HCT116 siCtrl siNEDD4&4L 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 siNEDD4&4L NEDD4L 4Rep edited Mono NEDD4L siCtrl siNEDD4&4L 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 2B4 siNEDD4&4L NEDD4 4Rep edited Mono NEDD4 NEDD4 expression siCtrl siNEDD4 +siNEDD4L ** NEDD4L expression ** siCtrl siNEDD4 +siNEDD4L shCPSF6-MDA NEDD4 expression siCtrl siNEDD4 +siNEDD4L ** siCtrl siNEDD4&4L 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 152MDA siNEDD4&4L NEDD4 4Rep edited Mono NEDD4 NEDD4L expression ** siCtrl siNEDD4&4L 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 152MDA siNEDD4&4L NEDD4L 4Rep Mono NEDD4L siCtrl siNEDD4 +siNEDD4L A Fig. S3 C D E B shCPSF6-MDA No Dox Plus Dox siCtrl siNEDD4&4L Spacer2 Spacer2 0.0 0.5 1.0 1.5 2.0 2.5 Relative Expression 152MDA siNEDD4&4L CTGF 4Rep V2 NoDox PlusDox siCtrl CTGF expression siNEDD4+ siNEDD4L * NS dTAG-NUDT21 HCT116CTGF expression siCtrl siNEDD4+ siNEDD4L siCtrl siNEDD4&4L Spacer2 Spacer2 0.0 0.5 1.0 1.5 2.0 Relative Expression 2B4 siNEDD4&4L CTGF 4Rep DMSO dTAG ** ** DMSO dTAGV-1 shCPSF6-MDA NEDD4 expression si- Ctrl si- NEDD4 ** NEDD4L expression si- Ctrl si- NEDD4L ** siCtrl siNEDD4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 152MDA siNEDD4 NEDD4 3Rep edited Mono NEDD4 siCtrlsiNEDD4L 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Relative Expression 152MDA siNEDD4L NEDD4L 3Rep edited Mono NEDD4L .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint YAP1S127/397A-HCT116 Verteporfin MGH-CP1 TrametinibPlaceHolder 0 20 40 60 80 100 120 Cell number (% of untreated) A386HCT_CCKIV_5Rep_Other NoDox PlusDox Cell number (% of untreated) No Dox Plus Dox NS NS ** VerteporfinMGH-CP1Trametinib NoDox PlusDox 0 2 4 6 8 10 Relative Expression A386HCT CTGF 3Rep V2 CTGF Relative CTGF expression No Dox +Dox * YAP1S127/397A-HCT116 NoDox PlusDox 0.0 0.5 1.0 1.5 Relative Expression 152MDA siNEDD4&4L CPSF6 4Rep Mono edited CPSF6 CPSF6 expression No Dox +Dox ** shCPSF6-MDA A Fig. S4B C .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted October 21, 2025. ; https://doi.org/10.1101/2025.10.21.683728doi: bioRxiv preprint