Genes & Development

Boys experience an overall increased incidence of several childhood cancers, including medulloblastoma, a clinically heterogeneous cerebellar tumor. In subtypes of Group 3 and Group 4 medulloblastoma, males are three times more prevalent than females. As medulloblastoma is suspected to initiate during fetal development, we hypothesized that this sex bias reflects a combination of prenatal, sex-specific developmental processes and somatic alterations. To test these hypotheses, we compiled a large multi-omics dataset from children with medulloblastoma, which revealed sex-specific alterations, including frequent loss of the inactive X chromosome in females with Group 4. Generation of a sex-matched single-cell transcriptome atlas of the developing murine cerebellum enabled investigation of putative developmental factors underlying sex bias. Progenitors giving rise to Group 3/4 subgroups were more abundant, more proliferative, and harbored more open chromatin for recruitment of LMX1A and OTX2, master transcription factors defining Group 3/4 identity. Advanced genetically engineered mouse models and human cerebellar organoids were leveraged to determine whether sexual dimorphism arises from intrinsic or extrinsic factors. These models showed that the XY genotype contributed to the phenotype, but the predominant effect was driven by presence of the male gonadal hormone testosterone. Our findings provide a sex-specific genetic and neurodevelopmental explanation for male bias in an aggressive pediatric brain tumor. Outcomes from this study may inform novel treatment strategies delivered according to sex and are likely to be broadly applicable to other sex-biased malignancies arising in early life. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=79 SRC="FIGDIR/small/714163v1_ufig1.gif" ALT="Figure 1"> View larger version (18K): org.highwire.dtl.DTLVardef@3a06faorg.highwire.dtl.DTLVardef@1a01bb7org.highwire.dtl.DTLVardef@7bc9c2org.highwire.dtl.DTLVardef@fb206d_HPS_FORMAT_FIGEXP M_FIG C_FIG

MicroRNAs (miRNAs) are key post-transcriptional regulators of cell state transitions, yet their function in early human brain development is largely unknown. Here, we present a longitudinal analysis of miRNA function in developing human forebrain organoids. We show that mRNAs and miRNAs expression mirrors known developmental gene programs and that miRNA biogenesis peaks at neural commitment. To test the function of miRNAs in regulating commitment, we impaired their biogenesis at defined stages. miRNA disruption during pre-neuronal commitment caused severe patterning defects, whereas post-commitment perturbation had minimal impact on forebrain identity. We show that miRNA loss during pre-commitment increased WNT and BMP signaling, thus shifting cell fates towards non-forebrain identity such as midbrain/hindbrain. These effects could be partially rescued by expressing five miRNAs. Our findings uncover a critical time window where miRNAs regulate morphogen signaling in early human neurodevelopment, establishing them as essential temporal determinants of cell fate and brain regional identity.

DNA-RNA hybrids (R-loops) form transiently on the genome and regulate cellular homeostasis. They also influence genome editing outcomes, highlighting their therapeutic potential in vivo. This protocol enables high-resolution mapping of DNA-RNA hybrids directly from frozen mouse tissues. Following tissue homogenisation and lysis, genomic DNA is extracted, digested and DNA-RNA hybrids are isolated using the hybrid-specific S9.6 monoclonal antibody. The purified hybrids are then processed for whole-genome sequencing to generate R-loop profiles. For complete details on the use and execution of this protocol, please refer to Puzzo, Crossley et al1. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=188 SRC="FIGDIR/small/716701v1_ufig1.gif" ALT="Figure 1"> View larger version (39K): org.highwire.dtl.DTLVardef@1c0e72borg.highwire.dtl.DTLVardef@188cd77org.highwire.dtl.DTLVardef@695c1corg.highwire.dtl.DTLVardef@11e6eff_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOgraphical abstractC_FLOATNO C_FIG

Proper polyadenylation site (PAS) selection is critical for RNA isoform determination. Core spliceosomal components, including U1 snRNP, regulate PAS choice, but whether they work with other splicing factors in this role remains unclear. Here, we establish that the splicing factor SRSF1 regulates PAS selection independently of and through interactions with U1 snRNP. Independent of U1 snRNP, SRSF1 binds RNA near proximal PASs within 3 UTRs to promote their usage, and, in line with this observation, breast cancer tumors with altered SRSF1 levels display shifted 3'-end selection. In conjunction with U1 snRNP, SRSF1 acts on PASs through U1 snRNP-mediated SRSF1-Pol II interactions. Consistent with co-transcriptional regulation, SRSF1 reduces the Pol II elongation index and limits transcription readthrough. Together, our results reveal that SRSF1 shapes RNA isoform determination beyond its canonical role in splicing, through a combination of direct RNA binding and U1 snRNP-dependent coordination with Pol II.

Transcription factors often act within defined developmental windows, yet how naive pluripotent cells acquire competence to execute specific transcription factor-driven fate programmes remains unclear. Pioneer transcription factors that engage target sites in closed chromatin to initiate gene expression programmes often act at the top of hierarchies in cell identity transitions. However, we show that the ability of ASCL1 to induce a coherent neuronal programme emerges only after exit from pluripotency, coincident with progressive chromatin remodelling and accumulation of permissive histone marks at neuronal ASCL1 target sites. Binding analysis reveals that although ASCL1 can access a subset of neuronal loci in mESCs and EpiLCs, ASCL1 is preferentially diverted to non-neuronal sites, resulting in divergent transcriptional responses. Increasing global histone acetylation enhances activation of individual neuronal genes but is insufficient to drive full neuronal differentiation. In contrast, co-expression of the homeodomain transcription factor PHOX2B redirects ASCL1 towards neuronal targets while suppressing inappropriate programmes in mESCs. These findings demonstrate that ASCL1 pioneer activity is highly context-dependent and that developmental priming of chromatin is essential for appropriate lineage specification. HIGHLIGHTSO_LIEctopic ASCL1 drives non-neuronal transcriptional responses in naive and formative pluripotent cells C_LIO_LIASCL1 occupies distinct, predominantly non-neuronal genomic targets in pluripotent cells due to differential chromatin accessibility C_LIO_LIASCL1 pioneer activity is locus- and cell type-specific and predicted by histone acetylation status C_LIO_LICo-expression of ASCL1 with Phox2 homeodomain cofactors potentiates neuronal lineage acquisition in pluripotent cells C_LI

In mammals, neural stem cells (NSCs) generate the neurons and glia essential to brain development, while defective NSCs drive brain cancer (glioma). NSC fate is not only controlled cell intrinsically, but also relies on signalling from the cellular microenvironment, or stem cell niche. Defective communication between NSCs and their niche can, therefore, drive stem cell renewal over differentiation to promote glioma. Here, we demonstrate that the orthologue of glioma-driver mutation FUBP1, Drosophila Psi, is essential in the cortex glia, the NSC niche, for preventing NSC overproliferation. We further demonstrate that Psi controls NSC fate through direct transcriptional repression of EGFR ligands, spitz (spi) and gurken (grk); with spi required cell intrinsically to enable proliferative growth of cortex glia, while grk functions cell extrinsically to control NSC fate. These observations highlight the complexity of EGFR function in modulating communication between NSCs and their niche. Our findings demonstrate the importance of understanding the nuances of intrinsic and extrinsic control of NSC fate, mechanisms of cell-cell communication critical to animal development that will provide insight into FUBP1/EGFR-driven glioma.

Small RNA pathways provide a robust and dynamic regulatory network that enables spatiotemporal regulation of the germline genome in response to environmental cues. The flexibility of RNA interference (RNAi)-mediated gene regulation and network architecture of these pathways requires molecular mechanisms that can fine-tune their regulatory potential and function to ensure proper execution of physiological processes, such as fertility. In C. elegans, we previously discovered a set of small RNA sensors that modulate the production of one class of small RNAs to adjust amplification resources based on cellular needs. These sensors maintain homeostatic levels of 22G-RNAs for the distinct RNAi branches that compete for resources in the mutator complex. Here we show this molecular feedback is essential for restricting expression of spermatogenic transcripts to an appropriate threshold during development and preventing spermiogenesis defects. Furthermore, we demonstrate 22G-RNA homeostasis is critical for proper meiotic progression in the germline and piRNA pathway function within pachytene germ cells. Together, our work reveals that RNAi homeostasis is critical for developmental and physiological processes, such as sperm-based fertility. Further, our findings show that small RNA pathway function is more than the sum of its parts and disrupting the ability to maintain homeostasis within the regulatory pathway itself leads to deleterious physiological consequences.

Primordial follicle oocyte activation (PFA) and zygotic genome activation (ZGA) represent two major waves of transcription activation respectively required for oocyte growth and preimplantation embryo development. Although many shared molecular hallmarks between PFA and ZGA suggest potential common factors and mechanisms driving both waves of transcriptional activation, such factors are yet to be identified. Here we demonstrate that the pioneer factor NFYA belongs to such regulators. Oocyte-specific Nfya deletion impairs open chromatin establishment and transcriptional activation during PFA, which triggers non-canonical ferroptosis leading to early folliculogenesis failure. Moreover, acute NFYA depletion in zygotes causes defective ZGA and predominantly two-cell embryo arrest. Mechanistically, although NFYA exhibits distinct chromatin-binding preferences predominantly targeting promoters during PFA and enhancers during ZGA, pre-occupied NFYA regulates chaperones and histone genes in both PFA and ZGA through conserved promoter binding. Together, our studies establish NFYA as a multifaceted regulator of genome activation during both PFA and ZGA. HighlightsO_LINFYA deficiency impairs primordial follicle oocyte activation (PFA) and triggers non-canonical ferroptosis resulting in early folliculogenesis failure C_LIO_LINFYA depletion impairs zygotic genome activation (ZGA) and causes predominantly 2-cell embryo arrest C_LIO_LIConserved and distinct NFYA-chromatin interactions drive both PFA and ZGA C_LIO_LIChaperones are pre-occupied and regulated by NFYA and their inhibition impairs both PFA and ZGA. C_LI

Poly(ADP-ribose) polymerase inhibitors (PARPi) exploit homologous recombination deficiency (HRD) in BRCA1/2-mutated cancers to induce synthetic lethality1-3. PARPi also kill cancer cells lacking the DNA damage-responsive kinase ATM4-6, however, inconsistent evidence of HRD7-9 and variable clinical responses9-11 have obscured the underlying mechanism. Here we define how PARPi induce cytotoxicity in ATM-deficient cells and reveal a critical role for ATM in regulating DNA replication. In the absence of ATM, unrestrained PRIMPOL-dependent repriming at spontaneous oxidative base adducts generates discontinuous daughter strands containing DNA gaps that activate PARP. This defect is sustained by aberrant suppression of replication fork slowing - presumably via fork reversal - by the BRCA1-A complex, whose recruitment to stalled forks is normally counteracted by ATM. The resulting gaps require homologous recombination (HR) for post-replicative repair and underlie the synthetic lethal interaction with PARPi. Suppressing repriming, reducing oxidative stress, or blocking base excision repair alleviates these defects. Collectively, our findings reveal how spontaneous base damage cooperates with replication dysfunction to drive PARPi sensitivity and establish a paradigm of post-replicative repair addiction in ATM-deficient cells. Together, our findings define a mechanistic link between oxidative DNA damage and ATM-dependent replication control, illuminating how oxidative stress may exacerbate genome instability in Ataxia Telangiectasia.

Many regions of heterochromatin associate with the nuclear periphery and are known as Lamin-associated domains (LADs). Histone acetyltransferase 1 (Hat1) is a highly conserved enzyme which acetylates newly synthesized histones H4 on lysines 5 and 12 prior to their deposition on chromatin. Hat1 is required to preserve chromatin accessibility within a subset of LADs called Hat1-dependent accessibility domains (HADs). Here we profile a diverse set of histone modifications in Hat1 KO and WT immortalized mouse embryonic fibroblasts (iMEFs) and find that Hat1 regulates diverse aspects of the structure of HADs and non-HAD LADs (nhLADS). In HADs, these changes include the conversion of H3K9me2 to H3K9me3. Analysis of H3K9-specific histone methyltransferases (HMTs) shows that that Suv39h1 and Suv39h2 have distinct localization patterns, where only Suv39h2 localizes to LADs. G9a only localizes to LADs in regions enriched for H3K9me2. We find that Hat1 loss results in a redistribution of these HMTs in both HADs and nh LADs. There is a decrease in the levels of G9a with a concomitant increase in Suv39h2. These results suggest Hat1 functions to restrain the formation of a more strongly heterochromatic state and highlight a role for Hat1 as an essential regulator of heterochromatin inheritance. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=97 SRC="FIGDIR/small/713225v1_ufig1.gif" ALT="Figure 1"> View larger version (23K): org.highwire.dtl.DTLVardef@9f6238org.highwire.dtl.DTLVardef@1e95415org.highwire.dtl.DTLVardef@18f3e0aorg.highwire.dtl.DTLVardef@1322426_HPS_FORMAT_FIGEXP M_FIG C_FIG

The cyclin-dependent kinase CDK11 functions in transcription, mitotic progression, and mRNA splicing. Specifically, spliceosome activation during the B to Bact transition depends on phosphorylation of the U2 snRNP component SF3B1 by the CDK11-cyclin L-SAP30BP complex. Here, we present the structure of this spliceosome-activating CDK-cyclin complex, determined by cryogenic electron microscopy at 2.3 [A] resolution. Our structure and biochemical experiments show that SAP30BP forms extensive interactions with cyclin L2, thereby stabilising it, and forms critical interactions with the C-terminal kinase lobe of CDK11 that promote complex assembly. Destabilisation of cyclin L2 in the absence of SAP30BP suggests that these principles are applicable to all CDK11-cyclin L complexes. Furthermore, we identify a pseudo-substrate sequence near the CDK11 C-terminus and provide evidence for a role of this segment in CDK11 auto-regulation. Finally, the structure of the CDK11-cyclin L2-SAP30BP complex bound to the clinical high-affinity CDK11 inhibitor OTS964 and a comparison to OTS964-bound off-target complexes provide insight into the mechanism of OTS964 selectivity and specificity.

Epigenetic memory ensures stable inheritance of gene expression patterns critical for embryonic development. Environmental exposures can disrupt this memory, yet the mechanisms remain unclear. Here we demonstrate that chronic morphine exposure induces a persistent transcriptomic and epigenetic memory by repressing Smchd1, a key chromatin regulator, in mouse embryonic stem cells, preimplantation embryos, and human induced pluripotent stem cells. This repression compromises maintenance of X-chromosome inactivation and genomic imprinting, leading to sustained dysregulation of developmentally important gene clusters. Morphine-induced epigenetic alterations also involve changes in DNA methylation and histone modifications along the X chromosome and notably increased H3K27me3 at the Smchd1 locus. These findings reveal a conserved mechanism by which opioid exposure disrupts higher-order chromatin architecture and epigenetic memory during early development, potentially contributing to long-term developmental and clinical outcomes. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=97 SRC="FIGDIR/small/713629v1_ufig1.gif" ALT="Figure 1"> View larger version (37K): org.highwire.dtl.DTLVardef@16986cborg.highwire.dtl.DTLVardef@1108eaaorg.highwire.dtl.DTLVardef@66373org.highwire.dtl.DTLVardef@16b37f6_HPS_FORMAT_FIGEXP M_FIG C_FIG

When and how lineage competence first emerges in the epiblast remains a central question in mammalian development. During the transition from naive to formative pluripotency, epiblast cells acquire responsiveness to lineage-inducing cues, yet whether transcriptional heterogeneity in this window reflects a regulated programme of lineage emergence or stochastic variation, and which molecular regulators shape developmental competence and potential lineage biases, remain poorly defined. Here, using a targeted CRISPRa screen, we identify the developmental regulator Wilms tumor 1 (WT1) as an unexpectedly early regulator of formative pluripotency. WT1 is transiently induced during the transition to formative pluripotency in vitro and in vivo, with peak expression coinciding with the emergence of lineage-associated transcriptional biases. Precocious Wt1 induction overrides the naive transcriptional network and advances cells toward a post-implantation epiblast identity, even under naive-stabilizing conditions. Genome-wide binding analyses show that WT1 engages active regulatory elements of the emerging post-implantation gene regulatory network together with core formative transcription factors, including Otx2 and Oct4. Alternative WT1 splice isoforms encode distinct lineage-biased transcriptional programmes associated with anterior and posterior fates. In the E5.5 epiblast, WT1 expression and splice composition align with lineage-biased transcriptional states, linking isoform usage to anterior-posterior transcriptional tendencies in vivo. Isoform-dependent gene expression modules are conserved in human pluripotent cells, indicating that this regulatory logic is preserved across species. Together, our findings indicate that lineage-associated transcriptional programmes begin to diversify during formative pluripotency and identify WT1 as an isoform-tuned regulator that biases these transcriptional outputs.

Germline development and successful embryogenesis depend upon the post-transcriptional regulation of maternal mRNAs. In Caenorhabditis elegans, the Notch-like receptor glp-1 is necessary for germline progenitor cell proliferation in adults and anterior cell fate determination in embryos. The spatiotemporal patterning of GLP-1 protein has long served as a paradigm of maternal mRNA regulation in metazoans. The glp-1 3'UTR has been shown to be sufficient to pattern the expression of reporter genes, and multiple regulatory regions and RNA-binding protein interaction sites have been mapped. The RNA-binding proteins POS-1 and GLD-1 directly regulate glp-1 mRNA via sequence specific interactions with motifs found in the glp-1 3'UTR. The impact of mutating the endogenous glp-1 3'UTR has not been studied, and the mechanism by which POS-1 and GLD-1 mediate repression is not understood. Here, we investigate the post-transcriptional mechanisms that govern glp-1 expression, revealing that GLD-1 and POS-1 regulate this pattern through different pathways requiring different co-factors. Remarkably, mutations in the endogenous locus that disrupt either POS-1 or GLD-1 binding to the glp-1 3'UTR have minimal impact on reproductive fecundity. By contrast, a larger deletion that eliminates the binding of both has a strong effect on brood size, hatch rate, and displays an increase in the length of the germline mitotic region that corresponds with enhanced mitotic activity. Together, our results show that multiple post-transcriptional mechanisms work in concert to ensure robust GLP-1 patterning and thus maximize reproductive outcomes.

DNA breaks activate PARP1/2 to synthesize poly(ADP-ribose) (PAR), which relaxes chromatin and recruits DNA repair factors. Normally, PAR is short-lived, rapidly degraded by poly(ADP-ribose) glycohydrolase (PARG). While PARP1/2 inhibitors are established therapies for homologous recombination (HR)-deficient cancers, predictive biomarkers for PARG inhibition (PARGi) remain undefined. Using parallel genome-wide CRISPR screens with PARP and PARG inhibitors, we show that PARGi is synthetically lethal with loss of several PAR-binding factors, including XRCC1-LIG3, POLB, ALC1/CHD1L, ARH3, and PARG itself, but notably not with HR deficiency. Conversely, loss of PARP1, NMNAT1 (required for nuclear NAD synthesis), or UNG (upstream of APE1 cleavage and PARP1 activation), confers PARGi resistance. Mechanistically, PARGi induces time- and dose-dependent formation of PARP1-and PAR-dependent nuclear condensates containing XRCC1 and associated repair factors in otherwise undamaged cells. These condensates do not harbor active DNA breaks but instead sequester PAR-binding repair proteins, depleting their available nuclear pool and impairing their recruitment to genuine DNA breaks. While our analysis focused on XRCC1, PARG inhibition likely sequesters additional PAR- and PARP1-binding proteins. Thus, we propose that PARGi sequesters PAR-binding proteins to elicit toxicity, explaining the essentiality of PARG (but not PARP1) and identifying the loss of PAR-binding factors as candidate predictive biomarkers for PARG-targeted therapy.

A hallmark of Parkinsons disease (PD) is the degeneration of midbrain dopaminergic neurons (mDANs). Genome-wide association studies (GWAS) have identified single nucleotide polymorphisms (SNPs) associated with PD, but causal variants and mechanisms remain unknown. Many PD-associated SNPs reside in regulatory regions, where they may disrupt transcription factor binding sites (TFBS) and alter gene expression. To assess how non-coding PD SNPs affect gene regulation in mDANs, we identify variants predicted to alter TF binding and functionally validate their effects in a cell type-specific context. We integrate time-series transcriptome and chromatin accessibility data from iPSC-derived neurons with chromatin topology and genetic variants. We profile 3D chromatin conformation in neuronal progenitors (smNPCs) and mDANs using LowC, identifying changes in A/B compartments and topologically associated domains. PD SNPs are enriched near genes expressed in mDANs, and we predict 254 regulatory variants that create or disrupt TFBS. Using chromatin conformation data, we link variants to target genes. At the BAG3 and SCARB2 loci, reporter assays in mDANs show reduced transcription driven by PD-associated alleles. Knock-down of NR2C2, a putative SCARB2 regulator, increases SCARB2 expression in differentiating neurons. The PD-associated SCARB2 allele shows reduced chromatin accessibility in mDANs and is associated with decreased expression in brain eQTL data. Insertion of PD-associated BAG3 allele by prime editing reduces chromatin accessibility across cell types, consistent with altered binding of LIM-homeodomain transcription factors. Together, these results prioritize functional PD SNPs and show that variants at SCARB2 and BAG3 modulate gene expression in mDANs, providing mechanistic insight into PD.

Olfactory sensory neurons (OSNs) project a single axon from the olfactory epithelium to the olfactory bulb. OSNs initially target large, distinct, individually identifiable neuropils called protoglomeruli in the zebrafish embryo. Here we examine the contributions Robo axonal guidance receptors make to OSN axon targeting of protoglomeruli. We show that OSNs that project to the DZ protoglomerulus express higher levels of robo2 than those that project to the CZ protoglomerulus, and concordant with this observation, DZ-projecting axons are more often misrouted by loss of robo2 than are CZ-projecting axons. Further, we demonstrate that in the absence of robo2, robo1 contributes to DZ-targeting but not to CZ-targeting. The loss of either robo1 or robo3 by themselves do not affect targeting to either the CZ or DZ protoglomeruli. These findings identify OSN subtype-dependent contributions of Robo receptors to vertebrate olfactory circuit assembly. In the absence of repellent Slit/Robo signaling, we propose that Netrin1b steers OSN axons to ectopic ventral midline locations where Slit1a and Netrin1b are both expressed.

PRMT5 is a type II arginine methyltransferase that forms an active complex with methylosome protein WDR77 (MEP50) to catalyze the symmetric dimethylation (SDMA) of arginine residues in proteins that regulate biological roles including apoptosis, DNA damage response and RNA processing. Some of the best characterized PRMT5 substrates are the small nuclear ribonucleoproteins SNRPB, SNRPD1 and SNRPD3, which are critical for spliceosome assembly and RNA splicing fidelity. MTAP-deleted cancers exhibit increased sensitivity to PRMT5 inhibition due to elevated levels of methylthioadenosine (MTA), a natural inhibitor of PRMT5. This vulnerability is exploited by MTA-cooperative PRMT5 inhibitors, exemplified by TNG908 and TNG462 which selectively target PRMT5 in MTAP-deleted cells while sparing MTAP-wildtype (WT) cells. Consistent with this mechanism, treatment with TNG908 in preclinical studies induces widespread splicing alterations in MTAP-deleted cancer models, with minimal effects in MTAP-WT cells. These splicing changes are consistent across diverse MTAP-deleted tumor types, including glioblastoma, pancreatic, and non-small cell lung cancer, indicating a histology-agnostic response to PRMT5 inhibition. Moreover, treatment of MTAP-WT cells with exogenous MTA mimics the splicing alterations observed with PRMT5 inhibition, as does pharmacologic inhibition of MTAP further supporting a mechanistic link between MTA accumulation, PRMT5 modulation, and aberrant splicing. Given that MTAP deletions occur in approximately 10-15% of human cancers, the identification of a robust RNA splicing signature offers a valuable pharmacodynamic biomarker for monitoring the activity of PRMT5 inhibitors. This splicing-based readout may also serve as a predictive biomarker of therapeutic response, offering greater specificity than global SDMA levels. Collectively these data suggest that a PRMT5-dependent RNA splicing signature can monitor the pharmacodynamic activity of MTA-cooperative PRMT5 inhibitors in MTAP-deleted cells.

Gene architecture in higher eukaryotes exhibits substantial heterogeneity. While most genes follow a canonical pattern of GC-rich exons and AT-rich introns, a subset displays GC-leveled architecture, characterized by uniformly high GC content across both exons and introns. These genes are often associated with nuclear speckles, membraneless compartments enriched in RNA-processing factors, yet the mechanistic basis of this spatial and functional relationship remains unclear. Here, we identify the nuclear speckle protein SON as a critical factor that safeguards the splicing of GC-rich genes. Acute depletion of SON in mouse embryonic stem cells selectively impairs the splicing of short, GC-rich introns. These SON-dependent introns are enriched in highly expressed and functionally essential genes, whose GC-rich architecture contributes to efficient RNA processing and expression. Mechanistically, these introns harbor atypical C-rich, U-poor polypyrimidine tracts at their 3 splice sites, which exhibit reduced affinity for core splicing factors. SON is recruited to these sites via U2 snRNP and further interacts with SR proteins to stabilize the association of U2 snRNP and U2AFs at these C-rich weak splice sites. Notably, the evolutionary expansion of SONs intrinsically disordered region is required to promote efficient splicing of GC-rich genes that emerged during evolution. Together, our study suggests that the evolutionary transition toward GC-rich gene architecture enhances gene expression efficiency, with SON acting to safeguard the splicing of this gene class.

The RUNX1 transcription factor mediates cell-type specific gene expression. RUNX1 suppression and perturbations are recurrently associated with breast tumor initiation and progression. However, the mechanisms governing the dual roles of RUNX1 in sustaining the mammary epithelial phenotype while epigenetically suppressing initiation of cancer-compromised gene expression are poorly understood. To address this, we used the power of degron-mediated acute, selective, and complete RUNX1 ablation in human mammary epithelial cells. RUNX1 mediates promoter and distal enhancer-driven expression of a gene cohort. Dynamic epigenomic responsiveness upon RUNX1 ablation reveals a rapid and selective decrease in chromatin accessibility and H3K27ac at RUNX1-bound enhancers, but not promoters. While differentially initiated and expressed genes contacted by RUNX1-bound enhancers are enriched in pathways involved in epithelial maintenance and stemness, genes with RUNX1-promoter occupancy support DNA damage responsiveness. Modified cell morphology, metabolic control, increased breast cancer stemness, plasticity, anchorage-independent survival, chemoresistance, and perturbed DNA damage reactivity are observed upon RUNX1 ablation. Together, these findings define RUNX1 as an epigenetic tumor suppressor that maintains epithelial cell state by preserving enhancer activity and preventing gene expression associated with hallmarks of cancer.