PRC2 and MYC repression drives oncofetal reprogramming and autophagy dependence in Colorectal Cancer

PRC2 and MYC repression drives oncofetal reprogramming and autophagy dependence in Colorectal Cancer | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article PRC2 and MYC repression drives oncofetal reprogramming and autophagy dependence in Colorectal Cancer Lluís Espinosa, Laura Solé, Angela Montoto, Andrea Martínez Ruíz, and 21 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9151067/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The emergence of an oncofetal phenotype in colorectal cancer (CRC) has been linked to therapy resistance and metastatic progression. However, whether this state pre-exists in treatment-naïve tumors or is primarily acquired upon chemotherapy, and the molecular mechanisms underlying these transitions, remain incompletely understood. While transcriptional regulators such as YAP1 and AP1 have been proposed as oncofetal inducers, the epigenetic and metabolic determinants of this process are not fully defined. Here, we integrate transcriptomic analyses from multiple CRC patient cohorts with functional studies in patient-derived organoids (PDOs) to show that both intrinsic and therapy-induced fetal-like CRC states are characterized by coordinated repression of Polycomb Repressive Complex 2 (PRC2) and MYC signaling. This repression involves coordinated downregulation of core and accessory PRC2 components together with attenuation of MYC/MAX activity across patient datasets and experimental models. Functional inhibition of PRC2 is sufficient to unlock fetal and epithelial-to-mesenchymal transition (EMT) transcriptional programs, whereas MYC suppression, mediated in part by MXD1 induction, promotes growth arrest and metabolic adaptation without recapitulating the full fetal signature. Mechanistically, fetal-like CRC cells, irrespective of their origin, display suppression of MYC- and mTOR-driven programs and increased reliance on autophagy. While basal autophagy levels show limited predictive value, induction of fetal-like states by chemotherapy or PRC2 inhibition consistently sensitizes CRC cells to lysosomal blockade. Together, these findings identify PRC2 and MYC repression as convergent regulatory features of oncofetal reprogramming in CRC and reveal autophagy dependence as a context-specific and therapeutically exploitable vulnerability of this aggressive tumor state. Biological sciences/Cancer/Tumour biomarkers Biological sciences/Cancer/Gastrointestinal cancer/Colorectal cancer Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Surgical removal of tumors, with or without adjuvant chemotherapy, remains the standard treatment for colorectal cancer (CRC). However, despite recent advances in targeted therapies and immunotherapy, about 35% of patients relapse and/or develop metastases and eventually die (source: https://www.cancer.net/cancer-types/colorectal-cancer/statistics). Therefore, identifying markers that predict tumor progression and strategies to eradicate cancer cells that persist after first- and second-line treatments remains an urgent clinical need. Conversion of intestinal tumor cells into a fetal phenotype, whether associated with specific mutational backgrounds or not, has emerged as a major mechanism used by CRC cells to survive chemotherapy 1–6 . Acquisition of fetal traits is associated with quiescence , enhanced tumor-initiating potential , and increased metastatic competence . This process relies, at least in part, on the activation of YAP1 , which is central effector of the Hippo pathway 4,6,7 . Fetal conversion is frequently accompanied by induction of selected EMT-associated genes, which is consistent with a hybrid epithelial–mesenchymal transition 4,6 . This transcriptional configuration has repeatedly been linked to increased dissemination and metastatic colonization across cancer types 8–13 . Recent studies indicate that specific oncogenic pathways, notably KRAS, play a critical role in maintaining the fetal-like transcriptional state. In KRAS -mutant CRC, pharmacological inhibition of KRAS reverses fetal conversion and restores the capacity of chemotherapy to eliminate adult LGR5 ⁺ cancer stem cells 14 . KRAS -mutant CRCs also exhibit marked sensitivity to the combined inhibition of Polycomb Repression Complex 2 (PRC2) with chemotherapy, which is frequently upregulated in human CRC 15 , but not yet explored in the context of fetal-type tumors, and RAS-pathway signaling 16,17 . We recently identified an oncofetal signature, termed ColoStem, which is induced upon chemotherapy treatment but is also present in a subset of treatment-naïve CRC tumors. Importantly, ColoStem predicts poor patient prognosis independently of KRAS or BRAF mutational status 17 . Here, we demonstrate that PRC2 repression promotes activation of fetal and EMT-associated transcriptional programs in CRC cells, concomitant with attenuation of MYC signaling, and confers chemotherapy resistance in vivo. Furthermore, acquisition of fetal traits following chemotherapy or PRC2 inhibition establishes a marked dependency on autophagy for cancer cell survival. Results Fetal-type tumors exhibit reduced PRC2 and MYC activities We investigated whether treatment-naïve fetal-type CRC tumors, classified based on the previously published 8-gene fetal signature, ColoStem 17 (see also Methods), share a higher-order regulatory landscape that could reveal mechanisms underlying the oncofetal phenotype. To address this, we stratified CRC tumors from the Marisa 18 , TCGA and Jorissen 19 datasets into fetal-type or non-fetal-type groups. Transcriptomic analysis identified robust differentially expressed gene (DEG) signatures across cohorts (Figures 1A-C, S1A and Supplementary Data 1). To infer upstream regulatory programs associated with these DEGs, over-representation analysis (ORA) was performed with enrichR using the Chromatin Histone Enrichment Analysis (ChEA) transcription factor target database (ENCODE_and_ChEA_Consensus_TFs_from_ChIP-X). This analysis revealed a significant enrichment of Polycomb Repression Complex 2 (PRC2)-associated targets (SUZ12-upregulated gene sets) as well as MYC-downregulated gene sets (Figure 1D), similar to that observed in slow-cycling cancer populations associated with poor clinical outcomes 20 . In contrast, we observed no significant enrichment for AP1-dependent programs, which have previously been implicated in oncofetal inducer in the AKSP ( APC null ; KRAS G12D ; SMAD4 null ; TP53 null ) mouse model 4 and in human PDOs 21,22 . Across datasets, enrichment of SUZ12-regulated gene sets in fetal-type tumors was highly consistent with >71% overlap in at least two cohorts and >52% overlap across all three cohorts (Figure 1E and Supplementary Data 2). MYC target downregulation was similarly conserved, with >62% overlap across cohorts (Figure 1F). Gene Set Enrichment Analysis (GSEA) of SUZ12 targets identified TGFb, Wnt, MAPK and Hippo/YAP1 as putative upstream regulators in fetal-type tumors (Figure 1G), consistent with our results in chemotherapy-treated PDOs 6 . Downregulated MYC targets were enriched for pathways controlling protein synthesis (ribosome, ribosome biogenesis, cofactor biosynthesis) and proliferation (cell cycle, DNA replication, nucleotide metabolism) (Figure 1H). Kaplan-Meier analyses of a meta-cohort comprising 1,067 CRC patients 17 (Figures 1I and 1J), as well as independent TCGA validation (Figures S1B and S1C), demonstrated that stratification based on SUZ12- or MYC-target expression (high vs. low) revealed significant differences in disease-free survival. In both cases, high SUZ12-target and low MYC-target expression were associated with poorer prognosis, mirroring the adverse survival pattern previously observed with the ColoStem fetal signature 6,17 . These findings highlight the context-dependent roles of PRC2 and MYC in CRC biology. While MYC is widely recognized as an oncogenic driver, its tumor-suppressive functions, including the ability to promote apoptosis under specific conditions, are well established 23,24 . In the setting of oncofetal reprogramming, reduced PRC2 and MYC pathway activity is associated with poor prognosis, consistent with this context-dependent functional behavior 25–27 . Transcriptional repression of PRC2 in therapy-naïve fetal-type tumors and chemotherapy-treated CRC cells We investigated whether the reduced PRC2 activity observed in fetal-type tumors could be attributed to decreased expression of specific PRC2 components. PRC2 consists of a catalytic core (EZH2, SUZ12 and EED) that associates with alternative accessory modules defining PRC2.1 (PCL proteins, including MTF2/PCL2, as well as EPOP and PALI1/2) or PRC2.2 (AEBP2 and JARID2) 28,29 . Across all datasets analyzed, we found that the core PRC2 subunits SUZ12 and EZH2 were consistently and significantly downregulated in treatment-naïve fetal-type tumors. This reduction extended to the PRC2.1 subunits MTF2 and EPOP (Figure 2A). These findings were validated in our CRC metacohort, where PRC2 transcripts were similarly reduced (Figure S2A). To refine these observations, we examined RNA-seq data from PDOs previously classified as non-fetal (PDO4, PDO10 and PDO66) or fetal (PDO5, PDO27 and PDO127) based on the ColoStem signature 17 . Fetal PDOs showed only a modest reduction in core PRC2 transcripts but exhibited a marked depletion of PRC2.1 elements, specifically MTF2 and EPOP (Figure 2B). Mass spectrometry analysis confirmed a significant decrease in core PRC2 protein levels in fetal PDOs (Figure 2C and Supplementary Data 3). PRC2.1 and PRC2.2 components were not consistently detected by mass spectrometry. We next assessed whether chemotherapy exposure further impaired PRC2 expression. In both fetal (PDO5) and non-fetal (PDO66) PDOs, chemotherapy treatment caused a profound decrease in SUZ12 , EZH2 , EED , MTF2 and EPOP RNA levels, as determined by RNA-seq analysis (Figure 2D). We validated PRC2 downregulation at the protein level by Western blot (WB) analysis (Figures 2E and 2F). Similar results were obtained in CRC cell lines LS174T and HCT116 lines after chemotherapy treatment (Figure 2G and Supplementary Data 4). Consistently, heatmap analysis confirmed coordinated repression of core PRC2 components upon chemotherapy exposure (Figure S2B). In parallel, GSEA revealed significant enrichment of fetal-like and EMT transcriptional programs in chemotherapy-treated cells (Figure S2C). These findings recapitulate the transcriptional remodeling observed in PDO5 and PDO66 6 , further supporting a link between chemotherapy-induced PRC2 suppression and acquisition of a fetal-like/EMT-associated state. To further validate the relationship between PRC2 repression and fetal identity of CRC cells, we analyzed scRNA-seq data from eight PDO samples, including six untreated PDOs and two 5-FU+Iri.-treated counterparts, one originally classified as fetal (PDO5) and one as non-fetal (PDO66) (Figure 2H). Uniform Manifold Approximation and Projection (UMAP) visualization revealed that chemotherapy-treated PDOs (which induce fetal conversion) tended to cluster closer to treatment-naïve fetal PDOs, suggesting convergence toward a shared transcriptional state. Notably, PDO127 did not cluster tightly with the other fetal PDOs, consistent with our previous bulk RNA-seq analyses in which PDO127, although grouping within the fetal cluster, displayed a partially divergent transcriptional profile 17 . The indicated fetal-like gene signatures were enriched in the same populations, confirming their fetal transcriptional identity. Consistently, these treatment-naïve and chemotherapy-induced fetal PDOs displayed lower EZH2 expression compared with non-fetal untreated counterparts. Together, these findings indicate that reduced PRC2 expression is an intrinsic feature of treatment-naïve fetal-type tumors, while chemotherapy promotes a shift toward this pre-existing fetal-like state. Collectively, these findings identify transcriptional repression of PRC2 components, particularly PRC2.1, as a shared feature of treatment-naïve fetal-type tumors and chemotherapy-induced oncofetal conversion. PRC2 inactivation unlocks oncofetal and EMT signatures and imposes chemotherapy resistance in vivo We next investigated whether loss of PRC2 function alone can trigger the transcriptional programs induced by sublethal chemotherapy 6 . To address this, we performed RNA-seq on CRC PDOs and cell lines subjected either to pharmacologic EZH2 inhibition with tazemetostat (EPZ-6438) or CRISPR-Cas9–mediated knockout of EED (Figure S3A). Both perturbations induced a pronounced transcriptional response in PDO5 and LS174T cells (Figures 3A, S3B, S3C and Supplementary Data 5), significantly overlapping with multiple published fetal intestinal gene signatures 3,30,31 . Consistently, EPZ-6438 treatment reduced the expression of canonical adult intestinal stem cell markers, including PROM1 32 , CDCA7 33,34 , LGR5 or ASCL2 35,36 (Figure 3B). Next, we tested whether PRC2 inhibition functionally recapitulates the phenotype associated with chemotherapy-induced fetal conversion. In vitro, PRC2 inactivation did not fully reproduce the quiescent and therapy‑resistant phenotype across PDOs (Figure S3D). However, in the zebrafish patient-derived xenograft (zAvatar) model 37,38 , EED KO PDO5 cells produced tumors that were smaller than those generated by the CRISPR control counterparts (Figure 3C), which is consistent with a reduced proliferative phenotype. Notably, PDO5 tumors treated with chemotherapy showed increased levels of apoptosis as determined by active caspase 3 staining, which was not detected in EPZ-6438-treated and EED KO tumors (Figure 3D-F). This contrasts with the observations made in the CRISPR control tumors, where a clear activation of apoptosis was observed with the treatment (Figure 3E and 3F). In addition to activating the fetal program (or as part of this phenotype), PRC2 inactivation induced a robust EMT transcriptional response in PDO5 cells (Figure 3G). Supporting this observation, EED KO PDO5 cells displayed a trend toward increased in vivo metastatic potential upon orthotopic transplantation into immunodeficient mice (Figure 3H). To gain mechanistic insight, we performed GSEA on transcriptional changes imposed by PRC2 inactivation in the different CRC models. Pathways involved in intestinal stem cell regulation and CRC progression, including Notch 39 , Wnt 40,41 and TGFb 42 were significantly upregulated, whereas proliferative drivers such as MYC, E2F and mTOR pathways were predominantly and reproducibly downregulated (Figures S3F and S3G). Together, these findings indicate that PRC2 loss is sufficient to unlock oncofetal and EMT transcriptional programs in CRC. PRC2 repression induces sustained YAP1 activation independently of canonical Hippo feedback YAP1 has emerged as a central regulator of fetal-like reprogramming, epithelial–mesenchymal plasticity and therapy resistance in colorectal cancer 6,43,44 . Given the strong overlap between YAP1-driven transcriptional programs and those induced by PRC2 repression, we asked whether loss of PRC2 activity directly impacts YAP1 signaling. To address this, we first performed GSEA using a Hippo signaling signature in PDO5 and LS174T cells. PRC2 loss resulted in significant enrichment of YAP1-associated gene programs compared to untreated cells (Figure 4A), indicating enhanced YAP1 transcriptional output upon PRC2 repression. We directly monitored YAP1 activity using a YAP1-responsive fluorescent reporter. Pharmacological inhibition of PRC2 resulted in an increase in YAP1 reporter activity at 72 hours (Figure 4C). As a positive control for YAP1 activation, we pharmacologically activated YAP1 downstream of the Hippo pathway using the LATS kinase inhibitor TDI-011536 for 72 hours, a condition that robustly increased active YAP1 levels. Notably, YAP1 activity remained elevated for at least 24 hours following inhibitor washout, indicating that PRC2 repression induces a sustained YAP1-active state rather than a transient signaling response. We then asked whether YAP1 activation could, in turn, account for the repression of PRC2 components observed in fetal-like CRC cells. To address this, we activated YAP1 using the LATS kinase inhibitors TDI-011536 and TRULI. Although these treatments effectively enhanced YAP1 signaling, they did not decrease PRC2 protein levels; instead, core PRC2 components were modestly increased, as determined by WB analysis (Figure 4D). Finally, we evaluated PRC2 regulation by YAP1 by either using direct YAP1 inhibition or genetic ablation. Neither pharmacological inhibition of YAP1 by Verteporfin nor YAP1 knockout reduced EZH2 transcript levels, as determined by qPCR analysis in PDOs (Figure 4E). YAP1 knockout efficiency had been previously validated (Figure S4A in 6 ). Together, these findings establish a hierarchical and unidirectional relationship between PRC2 and YAP1 during oncofetal reprogramming, in which epigenetic repression of PRC2 precedes and enables sustained YAP1 activation, but YAP1 activation alone is insufficient to modulate PRC2 expression. MYC pathway inhibition promotes quiescence entrance and partial chemotherapy resistance MYC is a central regulator of the intestinal crypt homeostasis and APC-driven intestinal tumorigenesis 45–47 . However, MYC signaling is markedly suppressed in CRC cells following chemotherapy treatment 2,6 (Figure S4A) and is consistently downregulated in fetal-type tumors (Figure 1D). We aimed to investigate the bases underlying MYC pathway downregulation in naïve-treatment and chemotherapy-induced fetal-type CRC cells. Analysis of the public dataset revealed that MYC levels were reduced in fetal tumors, whereas its obligate partner MAX was significantly upregulated (Figures 5A and S4B). In the chemotherapy-induced oncofetal model, MYC mRNA levels decreased upon chemotherapy treatment in PDO66 and the LS174T cell line, two models displaying robust suppression of MYC targets after treatment, but remained unchanged or increased in PDO5 (Figures 5B and 5C). Mirroring treatment-naïve fetal-type tumors, MAX levels were upregulated following chemotherapy treatment. Notably, MXD1 , an endogenous MYC antagonist 48 , was consistently induced across all oncofetal contexts. In the chemotherapy setting (Figures 5B and 5C) as well as upon pharmacological PRC2 inhibition with EPZ-6438 (Figures 5D and 5E), whereas MXD3 and MXD4 displayed more variable behavior (Figures 5A–E and Figure S4C). However, we detected some differences between EPZ-6438 treatment and EED KO cells, suggesting that EPZ-6438 may exert effects beyond PRC2 inhibition in a context-dependent manner To further assess MYC expression at the single-cell level, we interrogated the scRNA-seq dataset from PDOs described before (Figure 2H). UMAP visualization revealed reduced MYC expression in treatment-naïve fetal PDOs compared with non-fetal counterparts (Figure 5F). Consistently, chemotherapy-treated PDOs also exhibited diminished MYC expression and localized closer to fetal PDOs in the transcriptional space. These data further support the convergence toward a fetal-like state characterized by MYC pathway suppression. To functionally interrogate this regulatory axis, we generated a doxycycline-inducible MXD1 model. WB analysis confirmed strong nuclear MXD1 induction upon 16 hours of doxycycline treatment in HEK293T (Figure S4D) and three PDO lines (Figure 5G and Figure S4E). Induced MXD1 expression markedly reduced cell growth in 3D cultures (Figures 5H and S4F), consistent with MYC pathway suppression. We then tested whether ectopic MXD1 modulates chemotherapy response in PDO cells. We found that doxycycline treatment conferred partial but reproducible resistance to 5-FU+iri in PDO cells (Figure 5I). These results indicate that MXD1-mediated downregulation of the MYC pathway contributes to growth suppression and partially promotes chemotherapy resistance. Polycomb inhibition and sublethal chemotherapy sensitizes CRC cells to inhibitors of autophagy Autophagy enables cellular adaptation and survival under stress conditions 49 and is essential for maintaining embryonic diapause under adverse environmental conditions 50–52 . Diapause-like, drug-tolerant states have been repeatedly linked to cancer cell persistence in several cancer models 2,5,53 . However, the therapeutic benefit of targeting autophagy in cancer remains controversial 54 , and its functional connection to oncofetal reprogramming has not been systematically explored. Our transcriptomic analysis of treatment-naïve fetal-type CRC cells (Figure 1D and S1A), as well as cells following sublethal chemotherapy exposure 6 or PRC2 suppression (Figures S3E and S3F) revealed a robust inhibition of the MYC, E2F and/or mTOR pathways. As these pathways are canonical regulators of autophagy across multiple systems 50,55–59 , these results pointed to autophagy as a potential shared vulnerability of the oncofetal state. Consistent with this notion, WB analysis of PDOs treated with chemotherapy or PRC2 inhibitors for 72 hours showed a marked reduction in phospho-S6 (and/or phospho-S6K) levels indicating inhibition of the mTOR–S6 axis under both oncofetal-inducing conditions (Figure 6A). These data suggest that oncofetal reprogramming is coupled to an mTOR-low state, a metabolic configuration known to be permissive for increased dependency on lysosomal and autophagy-dependent pathways. To functionally test this possibility, we screened our PDO collection, including non-fetal (PDO4, PDO10 and PDO66) and treatment-naïve fetal-type lines (PDO5, PDO27 and PDO127). By immunofluorescence (IF) of the autophagosome marker LC3B 60 , we detected heterogeneous levels of basal autophagy across PDOs, which increased after chemotherapy or EPZ-6438 treatment, ultimately converging to comparable LC3B-high levels (Figure 6B and 6C). While basal autophagy levels showed limited predictive value for sensitivity to autophagy inhibition (Figure S5A), induction of fetal-like states by chemotherapy or PRC2 inhibition (experimental design in Figure 6D) consistently increased vulnerability to lysosomal blockade (Figures 6E, 6F and S5B) in all PDOs tested. We next asked whether autophagy was also elevated in advanced human tumors following chemotherapy. Analysis of matched CRC biopsies collected before and after neoadjuvant therapy revealed heterogeneous LC3B-positive vesicle accumulation in a subset of cases. Notably, an increase in LC3B accumulation after treatment showed a trend toward association with poor outcome (Figure 6G). Post-treatment samples from patients who relapsed more often showed increased LC3B levels relative to their matched pre-treatment biopsies, whereas tumors from non-relapsing patients more frequently exhibited no change or reduced LC3B accumulation. Although not observed in all cases, these data suggest a link between therapy-induced LC3B accumulation and relapse. Together, our results support a model in which PRC2 repression and MYC attenuation cooperate with previously described YAP1-driven programs to promote fetal and therapy-resistant states 6,7,12,61 . Importantly, acquisition of the oncofetal phenotype is consistently associated with increased autophagic activity or reliance on autophagy across experimental contexts (Figure 7). Discussion The emergence of an oncofetal phenotype has been increasingly recognized as a mechanism of therapeutic resistance and metastatic progression in CRC. However, the molecular basis of this reprogramming and its potential therapeutic implications have remained unclear. Here, we show that suppression of PRC2 activity is a consistent feature of both intrinsic and therapy-induced fetal-type tumors, in apparent contradiction to the conventional view of PRC2 as an oncogenic driver. Our results indicate that, within this context, PRC2 loss supports cellular plasticity by releasing developmental constraints and promoting EMT and fetal-like transcriptional programs. This dual behavior, oncogenic in some contexts yet permissive to cellular reprogramming in others, may underlie the context‑dependent functions attributed to PRC2 across cancers. The coordinated repression of PRC2 and MYC observed in fetal-type CRC cells reflects the integration of two functionally distinct processes: PRC2 loss facilitates transcriptional plasticity and lineage regression, whereas MYC attenuation promotes quiescence and metabolic adaptation. The induction of MXD1 across all fetal-like contexts, including PRC2 inhibition, suggests the existence of a feedback circuit in which PRC2 loss promotes MYC pathway suppression. However, forced MXD1 expression alone does not activate fetal transcriptional programs, indicating that MYC suppression is necessary but not sufficient for fetal reprogramming. Importantly, this fetal-like state, which enables cell survival under cytotoxic stress, also imposes an increased dependency on autophagy for energy homeostasis and survival, thus creating a transient (but clinically relevant) therapeutic window (see Figure 7). Our results demonstrate that chemotherapy-induced and PRC2-inactivated CRC cells showed enhanced sensitivity to lysosomal inhibitors such as chloroquine and bafilomycin A1, irrespective of their initial phenotype. These data are consistent with the concept that cancer cells use autophagy as a mechanism of therapeutic resistance (reviewed in 62 ), but contrast with previous reports describing a limited impact of autophagy blockade once cells have already acquired a diapause-like phenotype 2 . Importantly, the observation that PRC2 inhibition alone can reproduce the autophagy-dependent phenotype provides a rationale for combining PRC2-targeting agents with autophagy inhibitors in future therapeutic designs. Our analysis of matched patient samples further suggests that elevated autophagy may characterize a subset of treatment-naïve tumors prior to therapy, rather than being uniformly induced by chemotherapy. In this context, neoadjuvant treatment appears to remodel or select against highly autophagic, fetal-like tumor cell states, highlighting autophagy as a dynamic feature of tumor plasticity rather than a static response to cytotoxic stress. At first glance, our findings that partial inactivation of PRC2 or MYC in CRC unlocks a highly plastic, fetal-like transcriptional program appear to contrast with studies in germinal-center B cells, where increased MYC activity and preservation of PRC2-dependent programs, such as those described by Melnick and colleagues 63,64 , promote clonal fitness and malignant transformation. We propose that these observations are not contradictory but instead highlight the profound context-dependence of PRC2–MYC signaling. In epithelial tissues, PRC2 and MYC maintain adult lineage identity, and their attenuation facilitates regression towards embryonic, stress-tolerant states. In contrast, germinal-center B cells reside in an inherently proliferative and epigenetically permissive environment, in which additional activation of MYC or PRC2 amplifies competitive expansion. Thus, both gain- and loss-of-function perturbations of the PRC2–MYC axis can converge on tumor-promoting plasticity, depending on the cellular baseline and epigenetic landscape. Clinically, these findings carry several important implications for patient stratification and therapeutic design. First, they highlight that the fetal-like transcriptional state, detectable in a subset of treatment-naïve tumors, marks a distinct biological and therapeutic subgroup. Second, PRC2 or MYC suppression, which can be evaluated by immunohistochemistry or transcriptomic profiling, could serve as actionable biomarkers to identify patients likely to benefit from autophagy-based combination therapies. Third, as PRC2 inactivation is also recurrent in other malignancies, including T-cell Leukemia 65 , pediatric glioblastoma 66,67 , Malignant Peripheral Nerve Sheath Tumors (MPNST) 68 or NFKBIA hemizygous glioblastoma 69,70 , our findings may have broader relevance beyond CRC. Overall, rather than attempting to prevent oncofetal conversion, our data support exploiting transient reprogrammed states as therapeutic windows, in which cancer cells reveal specific metabolic and survival dependencies. Our data suggest that both sublethal chemotherapy and pharmacologic PRC2 inhibition not only induce or reinforce oncofetal reprogramming, but also increase autophagy dependence, pushing CRC cells into a metabolically vulnerable hyper-dependent state. This provides a dual therapeutic opportunity: autophagy inhibitors as a rational second-line approach following standard chemotherapy or neoadjuvancy, and combination strategies pairing PRC2 inhibitors with autophagy blockade to selectively eradicate cancer cells that have adopted a survival-prone, fetal-like program. Thus, oncofetal reprogramming should not be viewed merely as an obstacle to therapy, but as an opportunity, a transient and druggable state in which tumor cells reveal vulnerabilities that can be strategically exploited. Methods Colorectal Cancer cell lines CRC cell lines HCT116 (CCL-247) and LS174T (CL-188) were obtained from the American Type Culture Collection [ATCC, USA]. Cell lines were grown in Dulbecco’s modified Eagle’s medium [Invitrogen] supplemented with 10% fetal bovine serum [Biological Industries] and were maintained in a 5% CO 2 incubator at 37 °C. Cell mutations and concentrations of chemotherapy used for each cell line are indicated in the Supplementary Table 1. PDO generation and culture conditions Human colorectal tumors were obtained from Parc de Salut MAR Biobank (MARbiobank) and IdiPAZ Biobank (PT23/00028), integrated into the Spanish Hospital Biobanks Network (RetBioH; www.redbiobancos.es). Written informed consent was obtained from all participants and protocols were approved by Hospital del Mar’ Ethics Committee (CEImPSMar_E1.2025-11968-I; and previous approval 2019/8595/I) and were subsequently ratified by the Clinical Research Ethics Committee of Hospital Universitario La Paz/IdiPAZ (approval code 2025.699), in accordance with Spanish regulations and the Declaration of Helsinki. For PDOs generation, primary or xenografted human colorectal tumors were disaggregated in 1.5 mg/mL collagenase II and 20 μg/mL hyaluronidase after 40 min of incubation at 37 °C, filtered in 100 μm cell strainer, and seeded in 50 μL Matrigel in 24-well plates. After polymerization, 450 μL of complete medium was added (DMEM/F12 plus penicillin (100 U/mL) and streptomycin (100 μg/mL), 100 μg/mL Primocin, 1× N2 and B27, 10mM Nicotinamide; 1.25 mM N-Acetyl-L-cysteine, 100 ng/mL Noggin and 100 ng/mL R-spondin-1, 10 μM Y-27632, 10nM PGE2, 3μM SB202190, 0.5μMA8301, 50 ng/mL EGF and 10nM Gastrin I). Cultures were maintained at 37°C, 5% CO2 and medium changed every week. PDOs were expanded by serial passaging and kept frozen in liquid Nitrogen for being used in subsequent experiments. PDOs were routinely tested for mycoplasma contamination. PDOs mutations and concentrations of chemotherapy used for each PDO are indicated in the Supplementary Table 1. The concentrations of the other treatments were as follows: EPZ-6438, 50 µM; TDI-011536, 10 µM; TRULI, 10 µM; Verterpofin, 0.25 µM; and doxycycline, 2 µg/mL. Lentiviral transduction of cells lentiCRISPR v2 plasmid was used for knock-out experiments. The sgRNA against EED gene were designed using Benchling (Supplementary Table 2). MXD1 inducible plasmid was constructed modifying the pLIX-hN1ICD plasmid inserting the MXD1 gene [Addgene #91897]. Lentiviral production was performed by transfecting HEK293T cells the lentiviral vectors and the plasmid of interest. One day after transfection, the medium was changed and viral particles were collected 24 h later and then concentrated using Lenti-X Concentrator. PDOs were infected by resuspending single cells in concentrated viruses diluted in complete medium, centrifuged for 1 hours at 650 rcf, and incubated for 5h at 37°C. PDOs were then washed with complete culture medium and seeded as described above. After 72 hours, puromycin at 1 mg/ml was added for selection of infected cells for one week, in PDOs infected with sgRNA. PDOs Immunofluorescence analysis Paraffin blocks were obtained from PDOs after previous fixation in 4% formaldehyde overnight at room temperature. Paraffin-embedded sections of 2.5 μm were deparaffinized, rehydrated, citrate‐based antigen retrieval was used (20 min, no pressure) and endogenous peroxidase activity was quenched (20 min, 1.5% H 2 O 2 ). All primary antibodies were diluted in PBS containing 0.05% BSA, incubated overnight at 4°C (Supplementary Table 3). Samples were then incubated with the Envision+ System HRP Labeled Polymer anti-Rabbit or anti-Mouse [Dako] for 1.5 h and then developed with the Tyramide Signal Amplification System (TSA) [PerkinElmer] and mounted in DAPI Fluoromount-G. Images were taken in an SP8 confocal microscope (Leica). Western Blot (WB) PDOs and cell lines were lysed for 20 min on ice in 150 μL of PBS plus 0.5% Triton X-100, 1 mM EDTA, 100 mM sodium orthovanadate, 0.2 mM phenyl-methylsulfonyl fluoride (PMSF), and complete protease and phosphatase inhibitor cocktails. Lysates were first cleared by centrifugation at maximum speed for 10 min at 4°C. The supernatant was transferred to a new tube, sonicated for 10 min (10 cycles of 30 s ON/30 s OFF), and mixed with 6X Laemmli buffer (soluble fraction). The remaining pellet fraction was resuspended in 1X Laemmli buffer and sonicated under the same conditions (insoluble fraction). Samples were boiled at 95°C for 10 min and analyzed by WB using standard SDS–polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were resolved on polyacrylamide gels and transferred onto polyvinylidene difluoride (PVDF) membranes. Membranes were incubated overnight at 4°C with the appropriate primary antibodies (Supplementary Table 4), followed by incubation with horseradish peroxidase–conjugated secondary antibodies. Signal detection was performed using enhanced chemiluminescence and imaged with the iBright CL750 Imaging System [Invitrogen]. RT-qPCR analysis Total RNA from cell lines and PDOs was extracted with the RNeasy Mini Kit and RNeasy Micro Kit [QIAGEN] respectively, and cDNA was produced with the RT-First Strand cDNA Synthesis Kit [Roche]. RT-qPCR was performed in LightCycler 480 system using SYBR Green I Master Kit. Samples were normalized relative to the housekeeping genes. Primers used for RT-qPCR are listed in Supplementary Table 5. PDO viability assays 600 single PDO cells were plated in p96-well plates in 10 μL of Matrigel with 100 μL of complete medium. For growth analysis of MXD1 -inducible PDOs, after 6 days in culture, growing PDOs were treated with doxycycline at 1 mg/ml, which was refreshed every 72 hours. Cell viability was measured after 3 days, 6 days and 9 days from the first treatment, using the CellTiter-Glo 3D Cell Viability Assay [Promega] following manufacturer’s instructions, in an Orion II multiplate luminometer. For dose-response curves, PDOs were plated in p96-well plates in Matrigel and after 6 days in culture were treated with combinations of 5-FU+Iri. (at the concentrations indicated in Supplementary Table 1), EPZ-6438 at 50 mM or doxycycline at 1 mg/ml for 72 hours. Following 72 hours of treatment, medium was changed and PDOs were treated with increasing concentrations of either 5-FU+Iri. or Bafilomycin A1 for 72 hours. Cell viability was determined as described above. Zebrafish patient-derived xenograft microinjection Animal care and handling In vivo experiments were performed using zebrafish ( Danio rerio ) strains - nacre, casper and Tg(Fli1:eGFP) , which were handled according to the standard protocols of the European Animal Welfare Legislation, Directive 2010/63/EU (European Commission, 2016) and the Champalimaud Fish Platform Program. PDOs preparation for microinjection On the day of injection, PDOs were thawed at 37ºC and washed with DMEM/F12 plus penicillin (100 U/mL), streptomycin (100 μg/mL) and Primocin (100 μg/mL). After centrifugation at 250 rcf for 5 min at 4ºC, the pellet was gently resuspended with 100 uL of injection mixture (described in 71 ). Organoid aggregate size and viability were assessed using a hemocytometer and trypan blue. The suspension was labelled on ice for 5min with CellTracker Deep Red [C34565, Invitrogen] at a concentration of 1 µL/mL. The suspension was then centrifuged and resuspended in injection medium, in order to achieve a final concentration of ~ 2 × 10⁴ cell equivalents per µL. Zebrafish embryos microinjection PDOs were microinjected using a pneumatic injector [World Precision Instruments, Pneumatic PicoPump PV820], into the zebrafish perivitelline space (PVS) under a fluorescent stereoscope [Zeiss AxioZoom.V16]. After injection, xenografts were kept in an incubator at 34ºC. Zebrafish xenografts: Screening, treatment and fixation At 1-day post-injection (dpi), xenografts were screened under a fluorescence stereoscope for the presence of a tumoral mass. Successfully injected xenografts were randomly divided into control (E3 medium) and treatment group (5-FU+Irinotecan), which were renewed daily. 5-FU was administered at 4.2 mM and Irinotecan at 4 µM, diluted in E3 medium. At 4dpi, xenografts were sacrificed and fixed in 4% (v/v) formaldehyde [Thermo Scientific] overnight. Zebrafish xenografts: Whole-mount immunofluorescence Primary antibodies: anti-Cleaved Caspase-3 (rabbit, [Cell signaling, code#9661]), anti-Human mitochondria (mouse, 1:100 [Merck Millipore, cat#MAB1273]). Secondary antibodies: anti-rabbit 594 DyLight [ThermoFisher Scientific, cat#35510] and anti-mouse 488 DyLight [ThermoFisher Scientific, cat#35502] were applied simultaneously with DAPI. Xenografts were mounted with Mowiol. Zebrafish xenografts: Confocal microscopy and analysis Zebrafish xenografts were imaged using BC43 Andor Benchtop Confocal Microscope. Sequential z-stack images were acquired with a 5 µm interval. Images were analyzed using ImageJ software. The tumor size was calculated as the sum of the total area of human mitochondria positive cells per slice. The percentage of activated Caspase-3-positive cells was quantified manually by counting apoptotic bodies in every slice of the tumor. Zebrafish xenografts: Statistical Analysis Statistical Analysis was performed using GraphPad Prism (v8.0.2). Results are represented as average (AVG) ± standard error of the mean (SEM). Data were analyzed using the non-parametric Mann-Whitney test. Outliers were assessed using the “GraphPad Outlier” tool (https:// www.graphpad.com/quickcalcs/Grubbs1.cfm). For all tests, p -values ( p ) are two-tailed with a 95% confidence interval. Differences were considered significant whenever p0.05, (ns)>0.05, *≤0.05, **≤0,01, ***≤0.001, ****≤0.0001. In vivo mouse studies For peritoneal implantation assays, equivalent amounts of disaggregated PDO cells were injected in the cecum of athymic nude mice (strain: Hsd:Athymic Nude-Foxn1nu; 5–7-week-old males). Mice were monitored regularly for general health status and signs of tumor development. Two months after tumor implantation, all animals were sacrificed simultaneously across experimental groups to ensure comparable end-point analysis. At necropsy, the peritoneal cavity was carefully examined, and the number of visible intraperitoneal tumor implants was systematically recorded for each animal. Macroscopic peritoneal nodules were counted, and when required, tumor implants were collected for further histological confirmation. All procedures involving living animals were conducted under specific pathogen-free conditions and in accordance with the guidelines established by the Animal Care Committee of the Generalitat de Catalunya. The study protocols were reviewed and approved by the Committee for Animal Experimentation at the Institute of Biomedical Research of Bellvitge (Barcelona). Analysis of CRC cohorts Transcriptomic data from colorectal cancer cohorts were obtained from publicly available datasets. Microarray expression data from GSE39582 18 and GSE14333 19 were downloaded from the Gene Expression Omnibus (GEO) and analyzed using the affy R package. RNA-seq data and associated clinical information from colon (COAD) and rectal (READ) adenocarcinoma samples were retrieved from The Cancer Genome Atlas (TCGA) using the TCGAbiolinks R package (v2.24.1) with the STAR-Counts workflow. COAD and READ datasets were merged, normalized, and transformed using edgeR (v3.38.1). Patients were classified as fetal or non-fetal based on the quantile expression of a fetal intestinal stem cell (ISC) gene signature comprising eight genes, yielding 142 fetal and 142 non-fetal cases in the Marisa cohort, 57 fetal and 57 non-fetal cases in the Jorissen cohort, and 83 fetal and 83 non-fetal cases in the TCGA cohort. Differential gene expression analyses between fetal and non-fetal tumors were performed using limma (v3.52.2) 72 . Functional enrichment analyses were conducted using enrichR (v3.0) and clusterProfiler (v4.4.2) 73 interrogating the ENCODE and ChEA Consensus Transcription Factors from ChIP-X and KEGG gene set collections, respectively. Optimal cutpoints for survival analyses were determined using the surv_cutpoint function from the survminer package (v0.4.9). Survival analysis of CRC cohorts Survival analysis was performed using the Kaplan-Meier curve estimates, considering either the in-house CRC patients Metacohort, which was previously generated 17 and is available at Zenodo (doi: 10.5281/zenodo.13303049), or the TCGA dataset. Patients were classified into two subgroups, ‘High’ or ‘Low’, based on the expression of specific signature genes (SUZ12 and MYC targets; see Supplementary Data 2). This classification was performed using the optimized cutpoint assessed by the surv_cutpoint function from the survminer (v.0.4.9) R package. If a gene was targeted by more than one array probe in the Metacohort, the probe showing the highest expression was selected. Left-censored patients were excluded from further analysis. A standard log-rank test was performed to determine the statistical significance between the two subgroups. A p-value<0.05 was considered statistically significant. Hazard ratios are also shown for each comparison in which there was at least one event per group. All survival analyses were performed using the survival (v.3.3-8) R package. Bulk RNA-seq data analysis Bulk RNA sequencing was performed in PDOs and in the colorectal cancer cell line LS174T. In the different experiments, libraries were simultaneously prepared and sequenced using Illumina NovaSeq6000 platform (50 bp paired-end reads). Raw sequencing reads in FASTQ format were aligned to the human reference genome GRCh38.p13 (Gencode release 41) using STAR (v2.7.8) 74 . Gene-level read counts were generated using the featureCounts function from the Subread package (v2.0.3 for PDO datasets and v2.8.2 for LS174T datasets). For PDO samples, genes with more than 10 reads in at least three samples were retained, whereas for LS174T samples, genes with more than 10 reads in at least two samples were kept, reflecting differences in sample number and experimental design. Raw library size differences were normalized using the trimmed mean of M values (TMM) method implemented in edgeR (v3.40.2 for PDOs and v3.36.0 for LS174T) 75 . Normalized counts were used for unsupervised analyses, including principal component analysis and clustering. For differential gene expression analyses, read counts were transformed to log2 counts per million (logCPM), and the mean–variance relationship was modeled using precision weights with the voom approach implemented in limma (v3.54.2 for PDOs and v3.50.3 for LS174T) 72 . All analyses were conducted using R (v4.2.1). Heatmaps were generated from variance-stabilized expression values (vst) obtained from raw count matrices using DESeq2. Selected gene sets were extracted, converted from Ensembl IDs to gene symbols, and expression values were scaled by gene (row-wise z-score). Heatmaps were visualized using the ComplexHeatmap package (v2.16.0) 76 with Pearson correlation–based hierarchical clustering. Pre-ranked Gene Set Enrichment Analysis (GSEA) was performed using the fgsea package (v1.26.0). Genes were ranked based on the metric −log10(p value) × sign (log2 fold change) derived from limma differential expression statistics. Enrichment analyses were conducted across multiple gene set collections, including Gene Ontology Biological Processes, KEGG canonical pathways, Hallmark gene sets from MSigDB (v7.5.1), and custom-curated gene signatures. GSEA was run using the fgseaMultilevel algorithm with a minimum gene set size of 10 and a maximum of 2,000 genes. Enrichment plots and normalized enrichment scores were generated from fgsea results. Single cell RNA-seq data analysis Data pre-processing Single-cell RNA-seq data were generated from eight human PDOs, including fetal-type, non-fetal-type and chemotherapy-treated samples. Libraries were prepared using the GEM-X Universal 3’ Gene Expression v4 4-plex kit (10x Genomics) and sequenced on a NovaSeq 6000 platform according to the manufacturer’s protocols. Samples were multiplexed in pairs within each sequencing batch using dual index combinations (OB1|OB2 or OB3|OB4) as follows: 15007AAG (PDO4 OB1|OB2; PDO10 OB3|OB4), 15008AAG (PDO27 OB1|OB2; PDO127B OB3|OB4), 15009AAG (PDO5 untreated OB1|OB2; PDO5 treated OB3|OB4), and 15010AAG (PDO66 untreated OB1|OB2; PDO66 treated OB3|OB4). Raw sequencing data were processed using Cell Ranger (v9.0.1, 10x Genomics) with the cellranger multi pipeline. Reads were aligned to the human reference genome GRCh38 (10x Genomics refdata-gex-GRCh38-2024-A). Filtered feature–barcode matrices were generated separately for each PDO sample. Doublet detection Doublets were identified independently for each PDO using scDblFinder (v1.20.2) 77 , combining three complementary approaches: (i) random artificial doublet generation, (ii) internal clustering–based detection, and (iii) external clustering information derived from Cell Ranger graph-based Louvain clustering. Cells classified as doublets by at least two methods were considered high-confidence doublets and removed from downstream analyses. Quality control and normalization Filtered matrices were imported into R (v4.4.2) and processed using Seurat (v5.0.1). PDO datasets were merged into a single Seurat object without applying batch correction or integration. Cells with fewer than 750 or more than 2,500 detected genes, more than 15% mitochondrial gene content, or fewer than 10,000 total counts were excluded. Genes expressed in fewer than 10 cells and ribosomal genes were removed. Normalization and dimensionality reduction Data were normalized using SCTransform, regressing out mitochondrial gene content. Principal component analysis (PCA) was performed on the SCT assay. UMAP embeddings were generated using the first 37 principal components, capturing more than 85% of the total variance. UMAP projections were used to visualize sample distribution, gene expression levels and gene signature scores. Gene signature scoring Fetal-like gene signature scores were computed using the JASMINE algorithm (script version V1_11October2021) 78 on SCTransform-normalized expression values. Scores were calculated at single-cell resolution and visualized on UMAP embeddings to assess the distribution of fetal transcriptional programs across PDO samples. Proteomic analysis Organoid Lysis for Proteomic Analysis Organoid pellets were resuspended in 50 µl lysis buffer (8 M Urea, 10 mM Tris-Base, 100 mM NaH 2 PO 4 , pH 8) with phosphatase inhibitor [Roche]. Organoids were sonicated (Ultrasonic Homogeniser SKL-150W, Syclon) at 20% power with 9 second pulse rates for 1 min. Organoid debris was pelleted by centrifugation at 14,000 rcf for 8 min at 4ºC and supernatant collected. Protein concentration was estimated using BCA assay [Pierce]. Proteomic Sample Digestion and Cleanup All samples were normalized to a protein concentration of 500 µg. Protein lysates were reduced with 8 mM 1,4-dithiotreitol (DTT) at 30ºC for 30 min at 1200 rpm on a thermomixer [Eppendorf]. Followed by alkylation with 20 mM iodoacetamide (IAA) at 30ºC for 30 min at 1200 rpm in the dark. Sample urea concentration was diluted to 2 M with Tris-HCl (50 mM). 10 µg of trypsin/Lys-C [Promega] was added to each sample for a final 1:50 enzyme to protein ratio. Samples were digested overnight at 37ºC at 1000 rpm. Digestion was terminated by adding formic acid to 1% final concentration. Sample clean-up was carried out using Sep-Pak C18 columns [Waters]. Columns were activated with 100% ethanol and equilibrated with 0.1% trifluoroacetic acid (TFA). Samples were passed through the columns followed by two washes with 0.1% TFA. Peptides were eluted from the column with elution buffer (80% acetonitrile, 0.1% TFA). 10% of each sample elute was used for whole proteome mass spectrometry analysis. Mass Spectrometry Acquisition of Proteomic Samples A pooled sample of 1 µg was analyzed using a Bruker timsTof Pro mass spectrometer connected to an Evosep One Liquid chromatography system. Tryptic peptides were resuspended in 0.1% formic acid and loaded on to an Evosep tip. The Evosep was configured to pick up each tip, elute and separate the peptides using a set chromatography method (30 samples a day) 79 . The mass spectrometer was operated in positive ion mode with a capillary voltage of 1300-1600 V, dry gas flow of 3 l/min and a dry temperature of 180ºC. All data was acquired with the instrument operating in a data dependent analysis parallel accumulation serial fragmentation mode (dda-PASEF). Trapped ions were selected for ms/ms using parallel accumulation serial fragmentation (PASEF). A scan range of (100-1700 m/z) was performed at a rate of 4 PASEF MS/MS frames to 1 MS scan with a cycle time of 0.53 s 80 . The resultant file was used to create the dia-PASEF method within Bruker timsControl software. The scan mode “dia-PASEF” was selected and the pooled sample dda-PASEF file was opened in the window editor in the MS/MS tab. Once opened, the adjustable parallelogram was used to select the area of the heat map where the identifiable peptides (central region in the heat map containing peptides with charge states from +2 to +5) could be found. Dia-PASEF settings used were: mass width 26.0 Da, mass overlap 1.0, mass steps per cycle 34 or 35, mobility overlap 0.00, mass range 338.0-1214 m/z or 334.7-1185.7 m/z. All samples were acquired using data independent analysis parallel accumulation serial fragmentation (dia-PASEF) 81 . Spectral Library Generation and DIA Data Processing Data acquired using dia-PASEF was analyzed using DIA-NN 2.1.0 and DIA-NN 2.3.0 Academia (Data-Independent Acquisition by Neural Networks 82 . The Homo sapiens subset of UniProtKB (Swiss-Prot and TrEMBL database) 83 was used to generate a spectral library within DIA-NN (library free mode). Specific search settings included cysteine carbamidomethylation as a fixed modification, protein N-terminal acetylation and methionine oxidation as variable modifications, maximum missed cleavages 1, min precursor +1, max precursor +4, Neural network (cross validated) was used, cross run normalization was set to retention time dependent and library generation was set to IDs, retention time (RT), and ion mobility (IM) profiling. Precursor FDR was set to 1%. Proteomic Data Analysis and Visualization Raw proteomic data from DIA-NN was uploaded into R (v2025.09.2+418). Proteins lacking annotated names were removed from the dataset. Proteins represented by multiple entries were aggregated by calculating the mean abundance. The data was log2 transformed. Samples were grouped by PDO origin (PDO4, PDO5, PDO10, PDO27, PDO66, PDO127). Proteins were retained if at least 70% of values were present in at least one PDO group. Missing values were imputed using the imputeLCMD (v2.1.) package in R, by sampling from the left-shifted normal distribution (width = 0.3, downshift = 1.8). For visualization, pre-defined proteins of interest were selected. To facilitate comparison of expression patterns across samples, protein expression values were Z-score normalized on a per-protein (row-wise) basis. Heatmaps were generated using the ComplexHeatmap package in R 76 . Hierarchical clustering was applied to proteins, while samples were displayed in a predefined order without supervised clustering. Sample annotations indicating sample condition (Fetal or Non-fetal) were included above the heatmap. Quantification and Statistical analysis Each experiment shown in the manuscript has been performed at least twice. The statistical parameters reported in the figures and figure legends include the number of events quantified, standard deviation, statistical significance and the test performed. GraphPad Prism 9 software was used for the statistical analysis, with p -values of <0.05 being considered significant (**** p -value<0.0001, *** p - value<0.001, ** p -value 0.05). Declarations DATA AVAILABILITY Public datasets used in this study are referenced in the text. Single-cell and bulk RNA-seq datasets generated in this study have been deposited in NCBI Gene Expression Omnibus (GEO) repository under GEO SuperSeries accession number GSE243803, composed in three SubSeries GSE243802 (bulk RNA-seq of LS174T), GSE277036 (bulk RNA-seq of PDOs) and GSE324317 (10x scRNAseq data). Sequencing data can be accessed through this token: yxupseuylvqpjcp. Proteomics data is available at PRIDE PRoteomics IDEntifications (PRIDE) with identifier PXD075164. Source data is provided with this paper. Further information on research design is available in the Nature Research Reporting Summary linked to this article. CODE AVAILABILITY Scripts used in the bulk RNA-seq preprocessing steps are available at GitHub repository: https://github.com/BigaSpinosaLab/LAB_RNAseq_Data_Analysis. The final CRC metacohort is stored, as an RData object, at Zenodo (doi: 10.5281/zenodo.13303049). Scripts that have been used to process the scRNA-seq and bulk RNA-seq datasets are deposited in Github repository: https://github.com/BigaSpinosaLab/PAPER_PRC2_MYC_repression_oncofetal_CRC. ACKNOWLEDGEMENTS We thank the members of the Espinosa and Bigas laboratories for their constructive discussions and valuable suggestions throughout this work. This research was supported by CIBER – Consorcio Centro de Investigación Biomédica en Red (CB16/12/00244), Instituto de Salud Carlos III (ISCIII), Ministerio de Ciencia e Innovación and Unión Europea – European Regional Development Fund (FEDER). This work was funded by the FIS project PI25/00006 and the EP PerMed project AC24/00006 (ColoStem-Applied, EpPermed2024-149), both funded by ISCIII and co-funded by the European Union (FEDER); and the project PRYGN246819ESPI from the Fundación Científica de la Asociación Española Contra el Cáncer (AECC). We thank MarBiobank for providing and characterizing patient samples (PT20/00023, from Instituto de Salud Carlos III, FEDER) and the Comprehensive Molecular Analytical Platform (CMAP) under The SFI Research Infrastructure Programme, (18/RI/5702). A.M. is a recipient of a grant from ISCIII, grant number FI23/00002, co-funded by the European Social Fund Plus (ESF+). A.Y. acknowledges financial support from the China Scholarship Council (CSC), grant number 202408430067. AUTHORS CONTRIBUTIONS STATEMENT LS, AMo, AM-R, AM-L, AY, JK, J-JC-R and JB prepared the reagents and performed the experiments with cells; LS, TL-J, EC and MM performed the bioinformatics analysis; AN and DM performed the proteomic analysis; BC and RF performed the in vivo experiments with zebrafish; AV, DA-V and MM-I performed the in vivo experiments with mice; ABa and AM generated PDOs included in the work and revised the manuscript; LS, LE and AB conceptualized and conducted the study, analyzed data and prepared the manuscript. All authors have read and approved the final manuscript and consent its publication. COMPETING INTERESTS STATEMENT The authors declare no competing interests related to this work. References Alvarez-Varela, A. et al. Mex3a marks drug-tolerant persister colorectal cancer cells that mediate relapse after chemotherapy. Nat. Cancer 3 , 1052–1070 (2022). Dhimolea, E. et al. An Embryonic Diapause-like Adaptation with Suppressed Myc Activity Enables Tumor Treatment Persistence. Cancer Cell 39 , 240-256.e11 (2021). Moorman, A. et al. Progressive plasticity during colorectal cancer metastasis. Nature 637 , 947–954 (2025). Mzoughi, S. et al. Oncofetal reprogramming drives phenotypic plasticity in WNT-dependent colorectal cancer. Nat. Genet. 57 , 402–412 (2025). Rehman, S. K. et al. Colorectal Cancer Cells Enter a Diapause-like DTP State to Survive Chemotherapy. Cell https://doi.org/10.1016/j.cell.2020.11.018 (2021) doi:10.1016/j.cell.2020.11.018. Solé, L. et al. p53 wild-type colorectal cancer cells that express a fetal gene signature are associated with metastasis and poor prognosis. Nat. Commun. 13 , 2866 (2022). Shao, D. D. et al. KRAS and YAP1 converge to regulate EMT and tumor survival. Cell 158 , 171–84 (2014). Aggarwal, V., Montoya, C. A., Donnenberg, V. S. & Sant, S. Interplay between tumor microenvironment and partial EMT as the driver of tumor progression. iScience 24 , 102113 (2021). Godin, L. et al. A Novel Approach for Quantifying Cancer Cells Showing Hybrid Epithelial/Mesenchymal States in Large Series of Tissue Samples: Towards a New Prognostic Marker. Cancers (Basel). 12 , 906 (2020). Kröger, C. et al. Acquisition of a hybrid E/M state is essential for tumorigenicity of basal breast cancer cells. Proc. Natl. Acad. Sci. U. S. A. 116 , 7353–7362 (2019). Lüönd, F. et al. Distinct contributions of partial and full EMT to breast cancer malignancy. Dev. Cell 56 , 3203-3221.e11 (2021). Pastushenko, I. et al. Fat1 deletion promotes hybrid EMT state, tumour stemness and metastasis. Nature 589 , 448–455 (2021). Youssef, K. K. et al. Two distinct epithelial-to-mesenchymal transition programs control invasion and inflammation in segregated tumor cell populations. Nat. Cancer 5 , 1660–1680 (2024). Centonze, A. et al. A Plastic EMP1+ to LGR5+ Cell State Conversion as a Bypass to KRASG12D Pharmacologic Inhibition in Metastatic Colorectal Cancer. Cancer Discov. 16 , 320–344 (2026). Ghate, N. B. et al. Phosphorylation and stabilization of EZH2 by DCAF1/VprBP trigger aberrant gene silencing in colon cancer. Nat. Commun. 14 , 2140 (2023). Loi, P. et al. Epigenetic and Oncogenic Inhibitors Cooperatively Drive Differentiation and Kill KRAS - Mutant Colorectal Cancers. Cancer Discov. 14 , 2430–2449 (2024). Solé, L. et al. ColoStem, a core oncofetal signature that identifies poor prognosis colorectal tumors. Br. J. Cancer https://doi.org/10.1038/s41416-025-03280-2 (2025) doi:10.1038/s41416-025-03280-2. Marisa, L. et al. Gene Expression Classification of Colon Cancer into Molecular Subtypes: Characterization, Validation, and Prognostic Value. PLoS Med. https://doi.org/10.1371/journal.pmed.1001453 (2013) doi:10.1371/journal.pmed.1001453. Jorissen, R. N. et al. Metastasis-associated gene expression changes predict poor outcomes in patients with Dukes stage B and C colorectal cancer. Clinical Cancer Research https://doi.org/10.1158/1078-0432.CCR-09-1431 (2009) doi:10.1158/1078-0432.CCR-09-1431. Malla, S. B. et al. Pathway level subtyping identifies a slow-cycling biological phenotype associated with poor clinical outcomes in colorectal cancer. Nat. Genet. https://doi.org/10.1038/s41588-024-01654-5 (2024) doi:10.1038/s41588-024-01654-5. Ogasawara, N. et al. Discovery of non-genomic drivers of YAP signaling modulating the cell plasticity in CRC tumor lines. iScience 27 , 109247 (2024). Xiong, L. et al. A patient-derived organoid model captures fetal-like plasticity in colorectal cancer. Cell Res. 35 , 642–655 (2025). McMahon, S. B. MYC and the Control of Apoptosis. Cold Spring Harb. Perspect. Med. 4 , a014407–a014407 (2014). Edwards-Hicks, J. et al. MYC sensitises cells to apoptosis by driving energetic demand. Nat. Commun. 13 , 4674 (2022). Marzochi, L. L. et al. Use of histone methyltransferase inhibitors in cancer treatment: A systematic review. Eur. J. Pharmacol. 944 , 175590 (2023). Sun, S., Yu, F., Xu, D., Zheng, H. & Li, M. EZH2, a prominent orchestrator of genetic and epigenetic regulation of solid tumor microenvironment and immunotherapy. Biochim. Biophys. Acta Rev. Cancer 1877 , 188700 (2022). Weber, L. I. & Hartl, M. Strategies to target the cancer driver MYC in tumor cells. Front. Oncol. 13 , 1142111 (2023). Glancy, E. et al. PRC2.1- and PRC2.2-specific accessory proteins drive recruitment of different forms of canonical PRC1. Mol. Cell 83 , 1393-1411.e7 (2023). Healy, E. et al. PRC2.1 and PRC2.2 Synergize to Coordinate H3K27 Trimethylation. Mol. Cell 76 , 437-452.e6 (2019). Gao, S. et al. Tracing the temporal-spatial transcriptome landscapes of the human fetal digestive tract using single-cell RNA-sequencing. Nat. Cell Biol. 20 , 721–734 (2018). Mustata, R. C. et al. Identification of Lgr5-Independent Spheroid-Generating Progenitors of the Mouse Fetal Intestinal Epithelium. Cell Rep. https://doi.org/10.1016/j.celrep.2013.09.005 (2013) doi:10.1016/j.celrep.2013.09.005. Zhu, L. et al. Prominin 1 marks intestinal stem cells that are susceptible to neoplastic transformation. Nature 457 , 603–7 (2009). Guiu, J. et al. Identification of Cdca7 as a novel Notch transcriptional target involved in hematopoietic stem cell emergence. Journal of Experimental Medicine 211 , 2411–2423 (2014). Zhao, M. et al. Identification of cancer stem cell-related biomarkers in lung adenocarcinoma by stemness index and weighted correlation network analysis. J. Cancer Res. Clin. Oncol. 146 , 1463–1472 (2020). Barker, N. et al. Crypt stem cells as the cells-of-origin of intestinal cancer. Nature 457 , 608–611 (2009). Merlos-Suárez, A. et al. The intestinal stem cell signature identifies colorectal cancer stem cells and predicts disease relapse. Cell Stem Cell 8 , 511–524 (2011). Costa, B. et al. Zebrafish Avatar-test forecasts clinical response to chemotherapy in patients with colorectal cancer. Nat. Commun. 15 , 4771 (2024). Fazio, M., Ablain, J., Chuan, Y., Langenau, D. M. & Zon, L. I. Zebrafish patient avatars in cancer biology and precision cancer therapy. Nat. Rev. Cancer 20 , 263–273 (2020). Espinosa, L., López-Arribillaga, E., Bachs, O. & Bigas, A. Notch Signaling in the Normal Intestine and Intestinal Cancer. in Targeting Notch in Cancer 333–352 (Springer New York, New York, NY, 2018). doi:10.1007/978-1-4939-8859-4_13. Dow, L. E. et al. Apc Restoration Promotes Cellular Differentiation and Reestablishes Crypt Homeostasis in Colorectal Cancer. Cell https://doi.org/10.1016/j.cell.2015.05.033 (2015) doi:10.1016/j.cell.2015.05.033. Fre, S. et al. Notch and Wnt signals cooperatively control cell proliferation and tumorigenesis in the intestine. Proceedings of the National Academy of Sciences 106 , 6309–6314 (2009). Fasano, M. et al. TGF-β Modulated Pathways in Colorectal Cancer: New Potential Therapeutic Opportunities. Int. J. Mol. Sci. 25 , (2024). Zanconato, F., Cordenonsi, M. & Piccolo, S. YAP/TAZ at the Roots of Cancer. Cancer Cell 29 , 783–803 (2016). Yui, S. et al. YAP/TAZ-Dependent Reprogramming of Colonic Epithelium Links ECM Remodeling to Tissue Regeneration. Cell Stem Cell https://doi.org/10.1016/j.stem.2017.11.001 (2018) doi:10.1016/j.stem.2017.11.001. Muncan, V. et al. Rapid loss of intestinal crypts upon conditional deletion of the Wnt/Tcf-4 target gene c-Myc. Mol. Cell. Biol. 26 , 8418–26 (2006). Sansom, O. J. et al. Myc deletion rescues Apc deficiency in the small intestine. Nature 446 , 676–9 (2007). Sur, I. K. et al. Mice lacking a Myc enhancer that includes human SNP rs6983267 are resistant to intestinal tumors. Science 338 , 1360–3 (2012). Ayer, D. E., Kretzner, L. & Eisenman, R. N. Mad: a heterodimeric partner for Max that antagonizes Myc transcriptional activity. Cell 72 , 211–22 (1993). Kroemer, G., Mariño, G. & Levine, B. Autophagy and the Integrated Stress Response. Mol. Cell 40 , 280–293 (2010). Alhasan, B. A. et al. The mTOR Pathway in Pluripotent Stem Cells: Lessons for Understanding Cancer Cell Dormancy. Membranes (Basel). 11 , (2021). Lee, J.-E. et al. Autophagy Regulates Embryonic Survival During Delayed Implantation. Endocrinology 152 , 2067–2075 (2011). Lin, C. et al. The transcription factor p8 regulates autophagy during diapause embryo formation in Artemia parthenogenetica. Cell Stress Chaperones 21 , 665–675 (2016). Collignon, E. Unveiling the role of cellular dormancy in cancer progression and recurrence. Curr. Opin. Oncol. 36 , 74–81 (2024). Niu, X. et al. Autophagy in cancer development, immune evasion, and drug resistance. Drug Resist. Updat. 78 , 101170 (2025). Dong, W. et al. MYC drives autophagy to adapt to stress in Penaeus vannamei. Fish Shellfish Immunol. 126 , 187–196 (2022). Medda, A. et al. c-MYC-dependent transcriptional inhibition of autophagy is implicated in cisplatin sensitivity in HPV-positive head and neck cancer. Cell Death Dis. 14 , 719 (2023). Panwar, V. et al. Multifaceted role of mTOR (mammalian target of rapamycin) signaling pathway in human health and disease. Signal Transduct. Target. Ther. 8 , 375 (2023). Toh, P. P. C. et al. Myc inhibition impairs autophagosome formation. Hum. Mol. Genet. 22 , 5237–5248 (2013). Tsuneoka, M. et al. c-myc induces autophagy in rat 3Y1 fibroblast cells. Cell Struct. Funct. 28 , 195–204 (2003). Tanida, I., Ueno, T. & Kominami, E. LC3 and Autophagy. Methods Mol. Biol. 445 , 77–88 (2008). Ayyaz, A. et al. Single-cell transcriptomes of the regenerating intestine reveal a revival stem cell. Nature 569 , 121–125 (2019). Tonkin-Reeves, A., Giuliani, C. M. & Price, J. T. Inhibition of autophagy; an opportunity for the treatment of cancer resistance. Front. Cell Dev. Biol. 11 , (2023). Béguelin, W. et al. Mutant EZH2 Induces a Pre-malignant Lymphoma Niche by Reprogramming the Immune Response. Cancer Cell 37 , 655-673.e11 (2020). Béguelin, W. et al. EZH2 Is Required for Germinal Center Formation and Somatic EZH2 Mutations Promote Lymphoid Transformation. Cancer Cell 23 , 677–692 (2013). Ntziachristos, P. et al. Genetic inactivation of the polycomb repressive complex 2 in T cell acute lymphoblastic leukemia. Nat. Med. 18 , 298–302 (2012). Lewis, P. W. et al. Inhibition of PRC2 Activity by a Gain-of-Function H3 Mutation Found in Pediatric Glioblastoma. Science (1979). 340 , 857–861 (2013). Stafford, J. M. et al. Multiple modes of PRC2 inhibition elicit global chromatin alterations in H3K27M pediatric glioma. Sci. Adv. 4 , (2018). Lee, W. et al. PRC2 is recurrently inactivated through EED or SUZ12 loss in malignant peripheral nerve sheath tumors. Nat. Genet. 46 , 1227–1232 (2014). Bredel, M. et al. NFKBIA Deletion in Glioblastomas . New England Journal of Medicine 364 , 627–637 (2011). Bredel, M. et al. Haploinsufficiency of NFKBIA reshapes the epigenome antipodal to the IDH mutation and imparts disease fate in diffuse gliomas. Cell Rep. Med. 4 , (2023). Costa, B., Estrada, M. F., Barroso, M. T. & Fior, R. Zebrafish Patient‐Derived Avatars from Digestive Cancers for Anti‐cancer Therapy Screening. Curr. Protoc. 2 , (2022). Ritchie, M. E. et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. https://doi.org/10.1093/nar/gkv007 (2015) doi:10.1093/nar/gkv007. Yu, G., Wang, L. G., Han, Y. & He, Q. Y. ClusterProfiler: An R package for comparing biological themes among gene clusters. OMICS 16 , 284–287 (2012). Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29 , 15–21 (2013). Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26 , 139–140 (2009). Gu, Z., Eils, R. & Schlesner, M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32 , 2847–2849 (2016). Germain, P.-L., Lun, A., Garcia Meixide, C., Macnair, W. & Robinson, M. D. Doublet identification in single-cell sequencing data using scDblFinder. F1000Res. 10 , 979 (2021). Noureen, N., Ye, Z., Chen, Y., Wang, X. & Zheng, S. Signature-scoring methods developed for bulk samples are not adequate for cancer single-cell RNA sequencing data. Elife 11 , (2022). Bache, N. et al. A Novel LC System Embeds Analytes in Pre-formed Gradients for Rapid, Ultra-robust Proteomics. Mol. Cell. Proteomics 17 , 2284–2296 (2018). Meier, F. et al. Online Parallel Accumulation–Serial Fragmentation (PASEF) with a Novel Trapped Ion Mobility Mass Spectrometer. Molecular & Cellular Proteomics 17 , 2534–2545 (2018). Meier, F. et al. diaPASEF: parallel accumulation–serial fragmentation combined with data-independent acquisition. Nat. Methods 17 , 1229–1236 (2020). Demichev, V., Messner, C. B., Vernardis, S. I., Lilley, K. S. & Ralser, M. DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. Nat. Methods 17 , 41–44 (2020). Bateman, A. et al. UniProt: the Universal Protein Knowledgebase in 2023. Nucleic Acids Res. 51 , D523–D531 (2023). Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryFiguresandLegendsV12.docx Supplementary Figures and Legends SupplementaryData1.xlsx Supplementary Data 1 SupplementaryData2.xlsx Supplementary Data 2 SupplementaryData3.xlsx Supplementary Data 3 SupplementaryData4.xlsx Supplementary Data 4 SupplementaryData5.xlsx Supplementary Data 5 FigureS1.pdf Figure S1 FigureS2.pdf Figure S2 FigureS3.pdf Figure S3 FigureS4.pdf Figure S4 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9151067","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":608839076,"identity":"a081838e-74b7-46b9-b145-d0542495fa05","order_by":0,"name":"Lluís Espinosa","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwElEQVRIiWNgGAWjYBACPhCRACLYGxsgDEKADa6F5yApWsBAgij1IC3s7c8ePNxhY88/83Hrhoc5DHLyDYS08JwxN0g8k5Y443Zi243EbQzGBgcIaZHIYZNIbDucwADVkriBoMPknz8DabGXv3kQrKV+PkGHSTCYgbQwbrjBCNaSwEDQYTw5QL+0pSVuPAN2mIThBkJa+NmPP3v4s83GXu748Wc3f26zkScYYgzIUQMEEoTVo2sZBaNgFIyCUYAJABmPQiL7cnWtAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-2897-4099","institution":"Hospital del Mar Research Institute Barcelona (HMRIB)","correspondingAuthor":true,"prefix":"","firstName":"Lluís","middleName":"","lastName":"Espinosa","suffix":""},{"id":608839077,"identity":"2f009414-46a3-43ad-b225-ba408ddda59d","order_by":1,"name":"Laura Solé","email":"","orcid":"https://orcid.org/0000-0002-3195-0565","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Laura","middleName":"","lastName":"Solé","suffix":""},{"id":608839078,"identity":"0d9a4a8d-c14c-4a67-a6f7-a46a4590c475","order_by":2,"name":"Angela Montoto","email":"","orcid":"https://orcid.org/0000-0002-1127-3632","institution":"Hospital del Mar Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Angela","middleName":"","lastName":"Montoto","suffix":""},{"id":608839079,"identity":"402897a3-3666-481a-988e-be5d75f47735","order_by":3,"name":"Andrea Martínez Ruíz","email":"","orcid":"","institution":"Hospital del Mar Research Institute Barcelona (HMRIB)","correspondingAuthor":false,"prefix":"","firstName":"Andrea","middleName":"Martínez","lastName":"Ruíz","suffix":""},{"id":608839080,"identity":"658da4e4-fe9c-4aba-93af-efc75b66e14d","order_by":4,"name":"Bruna Costa","email":"","orcid":"https://orcid.org/0000-0002-6697-7258","institution":"Champalimaud Centre for the Unknown, Champalimaud Foundation","correspondingAuthor":false,"prefix":"","firstName":"Bruna","middleName":"","lastName":"Costa","suffix":""},{"id":608839081,"identity":"cc276f43-ec13-4058-847b-55b5a1ad8dcc","order_by":5,"name":"Teresa Lobo-Jarne","email":"","orcid":"https://orcid.org/0000-0002-1632-438X","institution":"Hospital 12 De Octubre Research Institute (i+12)","correspondingAuthor":false,"prefix":"","firstName":"Teresa","middleName":"","lastName":"Lobo-Jarne","suffix":""},{"id":608839082,"identity":"79ffbe11-258b-4ca8-85bd-c3a702a1f168","order_by":6,"name":"Daniel Alvarez Villanueva","email":"","orcid":"","institution":"Institute Hospital del Mar d'investigations Mediques-IMIM","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"Alvarez","lastName":"Villanueva","suffix":""},{"id":608839083,"identity":"720a55aa-2e1e-4e60-817c-8b02602a3038","order_by":7,"name":"Aifang Yang","email":"","orcid":"","institution":"Hospital del Mar Research Institute Barcelona (HMRIB)","correspondingAuthor":false,"prefix":"","firstName":"Aifang","middleName":"","lastName":"Yang","suffix":""},{"id":608839084,"identity":"fb15afdd-168e-44b7-a4ab-000121b77bee","order_by":8,"name":"Aoife Nolan","email":"","orcid":"","institution":"Systems Biology Ireland, School of Medicine, University College Dublin","correspondingAuthor":false,"prefix":"","firstName":"Aoife","middleName":"","lastName":"Nolan","suffix":""},{"id":608839085,"identity":"3f0bfed8-c35d-48fd-935b-902ad2cbacb4","order_by":9,"name":"Kieran Wynne","email":"","orcid":"","institution":"University College Dublin","correspondingAuthor":false,"prefix":"","firstName":"Kieran","middleName":"","lastName":"Wynne","suffix":""},{"id":608839086,"identity":"a77cb95c-6b9e-45ff-bc3b-e02e3f24536a","order_by":10,"name":"eric canton","email":"","orcid":"","institution":"Institute Hospital del Mar d'investigations Mediques-IMIM","correspondingAuthor":false,"prefix":"","firstName":"eric","middleName":"","lastName":"canton","suffix":""},{"id":608839087,"identity":"b5b7f03a-45a8-4d3d-9a2f-da5331b42c16","order_by":11,"name":"Maria Maqueda","email":"","orcid":"","institution":"Hospital del Mar Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Maria","middleName":"","lastName":"Maqueda","suffix":""},{"id":608839088,"identity":"9b41382b-299f-4af2-aeca-bd8135111cb5","order_by":12,"name":"Júlia-Jié Cabré-Romans","email":"","orcid":"","institution":"Hospital del Mar Research Institute Barcelona (HMRIB)","correspondingAuthor":false,"prefix":"","firstName":"Júlia-Jié","middleName":"","lastName":"Cabré-Romans","suffix":""},{"id":608839089,"identity":"1e8fe533-03ff-48b6-8486-7fcb67c63803","order_by":13,"name":"Aleksandra Montcusí Litvinova","email":"","orcid":"https://orcid.org/0009-0002-1449-0832","institution":"Hospital del Mar Research Institute Barcelona (HMRIB)","correspondingAuthor":false,"prefix":"","firstName":"Aleksandra","middleName":"Montcusí","lastName":"Litvinova","suffix":""},{"id":608839090,"identity":"4325ae0e-3b7d-4ffa-be98-05abb7f2f44e","order_by":14,"name":"Joanna Kalafut","email":"","orcid":"","institution":"Hospital del Mar Research Institute Barcelona (HMRIB)","correspondingAuthor":false,"prefix":"","firstName":"Joanna","middleName":"","lastName":"Kalafut","suffix":""},{"id":608839091,"identity":"4da783a0-e229-4666-a863-c9e2c97b91fd","order_by":15,"name":"Maria Martínez-Iniesta","email":"","orcid":"https://orcid.org/0000-0001-6252-6671","institution":"Catalan Institute of Oncology (ICO)","correspondingAuthor":false,"prefix":"","firstName":"Maria","middleName":"","lastName":"Martínez-Iniesta","suffix":""},{"id":608839092,"identity":"7b452693-4451-4550-99c0-ea6dc0b5fdb3","order_by":16,"name":"Raquel Cavaco","email":"","orcid":"","institution":"Champalimaud Foundation","correspondingAuthor":false,"prefix":"","firstName":"Raquel","middleName":"","lastName":"Cavaco","suffix":""},{"id":608839093,"identity":"9572bd85-9fb4-4c9d-b8b2-c89a42a208ef","order_by":17,"name":"Antonio Barbáchano","email":"","orcid":"","institution":"Instituto de Investigaciones Biomédicas Sols-Morreale (IIBM), Spanish National Research Council (CSIC)-Autonomous University of Madrid (UAM) and IdiPAZ","correspondingAuthor":false,"prefix":"","firstName":"Antonio","middleName":"","lastName":"Barbáchano","suffix":""},{"id":608839094,"identity":"95b1362a-7ca6-4678-8b7f-15f423c6c252","order_by":18,"name":"Alberto Munoz","email":"","orcid":"https://orcid.org/0000-0003-3890-4251","institution":"Alberto Sols Biomedical Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Alberto","middleName":"","lastName":"Munoz","suffix":""},{"id":608839095,"identity":"7c81a8c8-21be-4fd5-a586-35cfdae22ff9","order_by":19,"name":"David Matallanas","email":"","orcid":"https://orcid.org/0000-0002-2360-3141","institution":"University College Dublin","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Matallanas","suffix":""},{"id":608839096,"identity":"d161c8c2-b57d-482a-aced-a81cd5cfc3ae","order_by":20,"name":"Joan Bertran","email":"","orcid":"https://orcid.org/0000-0002-4276-336X","institution":"Universitat de Vic - Universitat Central de Catalunya","correspondingAuthor":false,"prefix":"","firstName":"Joan","middleName":"","lastName":"Bertran","suffix":""},{"id":608839097,"identity":"0e6626bf-d763-4078-855d-a32d541ad61a","order_by":21,"name":"Alberto Villanueva","email":"","orcid":"https://orcid.org/0000-0001-5164-0006","institution":"Institut Català d'Oncologia (ICO)-Institut d'Investigació Biomèdica de Bellvitge (IDIBELL)","correspondingAuthor":false,"prefix":"","firstName":"Alberto","middleName":"","lastName":"Villanueva","suffix":""},{"id":608839098,"identity":"d5289d4a-c3a6-4928-8539-8051fa516c37","order_by":22,"name":"Mar Iglesias","email":"","orcid":"","institution":"Hospital del Mar","correspondingAuthor":false,"prefix":"","firstName":"Mar","middleName":"","lastName":"Iglesias","suffix":""},{"id":608839099,"identity":"fde39e31-90e3-4f67-970b-d809021ac8ae","order_by":23,"name":"Rita Fior","email":"","orcid":"","institution":"Champalimaud Centre for the Unknown","correspondingAuthor":false,"prefix":"","firstName":"Rita","middleName":"","lastName":"Fior","suffix":""},{"id":608839100,"identity":"b701ed57-ca28-4e94-9e20-016c710a0eae","order_by":24,"name":"Anna Bigas","email":"","orcid":"","institution":"Hospital del Mar Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Anna","middleName":"","lastName":"Bigas","suffix":""}],"badges":[],"createdAt":"2026-03-17 16:30:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9151067/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9151067/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105038019,"identity":"efa183bd-c5ef-4bce-94eb-8503a3454bc9","added_by":"auto","created_at":"2026-03-20 07:41:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1133571,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFetal-type tumors exhibit reduced PRC2 and MYC activities.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A-C)\u003c/strong\u003e Volcano plots from RNA-seq differential expression analysis of fetal and non-fetal tumors from the indicated CRC cohorts (n=142 fetal and n=142 no fetal patients for the Marisa cohort, n=57 Fetal patients and n=57 no fetal patients for the Jorissen cohort and n=83 fetal and n=83 no fetal patients for the TCGA cohort). Representative genes including those of the ColoStem signature are highlighted. In blue are shown downregulated genes whereas upregulated genes are in red (FDR adjusted p-value \u0026lt; 0.05 and absolute shrunken log2 fold change \u0026gt; 0.2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(D) \u003c/strong\u003eBarplot showing ORA results from fetal compared with non-fetal tumors in the indicated datasets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(E, F)\u003c/strong\u003e Venn diagrams showing the overlap of DEGs in the analyzed datasets classified as SUZ12 (E) and MYC (F) targets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(G, H)\u003c/strong\u003e Barplot showing GSEA results from RNA-seq differential expression of genes classified as SUZ12 (G) and MYC (H) targets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(I, J) \u003c/strong\u003eKaplan-Meier plots of disease-free survival (DFS) over time for CRC patients included in the metacohort according to DEGs classified as SUZ12 (I) and MYC (J) targets using the optimal value as cutoff for patient stratification. Kaplan-Meier curve estimates shown with 95% confidence interval.\u003c/p\u003e\n\u003cp\u003eSource data are provided as a Source Data file.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/1f5b9410e502c1290f7073de.png"},{"id":105038051,"identity":"36d951a3-2fa5-459e-8b52-03f58a927f9e","added_by":"auto","created_at":"2026-03-20 07:41:51","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1024892,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTranscriptional repression of PRC2 in therapy-naïve fetal-type tumors and chemotherapy-treated CRC cells.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Bar plots indicating the mRNA levels of the indicated genes of the PRC2 complex in the three datasets analyzed. Notice the significant downregulation of the core elements \u003cem\u003eSUZ12\u003c/em\u003e and \u003cem\u003eEZH2\u003c/em\u003e as well as the PRC2.1 subunits in all datasets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(B, C)\u003c/strong\u003e Heatmap representation of mRNA (B) and protein (C) levels of the different PRC2 elements in non-fetal and fetal PDOs (B: n = 3 biological replicates; C: n = 3 biological replicates performed in technical duplicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(D)\u003c/strong\u003e Heatmap representation of mRNA levels of the different PRC2 elements in PDO5 and PDO66 untreated or treated with 5-FU+Iri. for 72 hours (PDO5: n=2 untreated and n=4 treated; PDO66: n=4 biological replicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(E-G)\u003c/strong\u003e WB analysis (soluble fraction) of the different core PRC2 subunits in PDO5 (E), PDO66 (F) and two different CRC cells lines (G) untreated or treated with 5-FU+Iri. for 72 hours (from 1 out of 3 experimental replicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(H)\u003c/strong\u003e Uniform Manifold Approximation and Projection (UMAP) representations of scRNA-seq data from PDO cells. The upper left panel shows the distribution of the eight PDO samples (six untreated PDOs and two 5-FU+Iri.-treated counterparts). The upper right panel depicts \u003cem\u003eEZH2\u003c/em\u003e expression levels across cells. The bottom left and bottom right panels display the enrichment scores of the indicated fetal-like gene signatures.\u003c/p\u003e\n\u003cp\u003eStatistical tests; (A) Data are presented as box plots showing the median, the 25th–75th percentiles, and the minimum and maximum values, Wilcoxon test. Source data are provided as a Source Data file.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/4582f79637da592b4a97af27.png"},{"id":105039265,"identity":"61feecb6-6133-4d5b-8b2a-16146d83ab16","added_by":"auto","created_at":"2026-03-20 07:45:36","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":986357,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePRC2 inactivation unlocks oncofetal and EMT signatures and imposes chemotherapy resistance in vivo.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e GSEA results from control and EPZ-6438-treated (50 mM) plus EED KO#1 and #3 PDO5 cells using the indicated fetal intestinal signatures.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(B)\u003c/strong\u003e Heatmap representation of the indicated ColoStem (left panel) and adult intestinal stem cell genes (right panel) from PDO5 cells treated as indicated (n=3 biological replicates per condition).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(C)\u003c/strong\u003e Quantification of the basal tumor size from PDO5 control, EED KO#1, EED KO#3 and EPZ-6438-treated (50 mM) zebrafish xenografts, quantified at 4 days post-injection (4dpi). Each dot represents one zAvatar and the total number (n) of zAvatars analyzed is indicated in the images in E.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(D)\u003c/strong\u003e Experimental workflow of the in vivo PDO zebrafish xenograft assay, showing injection into the perivitelline space of zebrafish embryos and the subsequent experimental steps.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(E)\u003c/strong\u003e Representative images of tumors generated in the zebrafish embryos from control, EED KO#1, EED KO#3 and EPZ-treated PDO5 cells, untreated or treated with 5-FU+Iri, at 4 days post-injection (4dpi). A dashed white line delineates the tumor.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(F)\u003c/strong\u003e Quantification of the relative levels of active caspase 3 (apoptosis) of tumors generated from control, EED KO#1, EED KO#3 or EPZ-6438-treated, untreated or after 5-FU+Iri. treatment. Each dot represents one zAvatar and the total number (n) of zAvatars analyzed is indicated in the images.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(G)\u003c/strong\u003e GSEA results from control and EPZ-6438-treated (50 mM) plus EED KO#1 and #3 PDO5 cells using the EMT signature from the Hallmark gene set collection.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(H)\u003c/strong\u003e Quantification of the number of intraperitoneal implants generated by control and EED KO#1 and #3 PDO5 cells after orthotopic implantation in the cecum of nude mice (n=5 mice per condition).\u003c/p\u003e\n\u003cp\u003eStatistical tests; (C, F) Data are presented as mean ± SEM, Mann-Whitney test. (H) Data are presented as total numbers, Chi-square test. Source data are provided as a Source Data file.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/43bfac7c45d5ac9c97825456.png"},{"id":105038028,"identity":"cfe593f4-5ecd-4fd1-b69d-40bf0cdf987e","added_by":"auto","created_at":"2026-03-20 07:41:40","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":405270,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePRC2 repression induces sustained YAP1 activation independently of canonical Hippo feedback.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e GSEA results of the Hippo signaling pathway from control and EPZ-6438-treated (50 mM) plus EED KO#1 and #3 PDO5 (left panel) and from control and EPZ-6438-treated (50 mM) LS174T (right panel) cells.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(B) \u003c/strong\u003eFlow cytometry analysis of YAP1 reporter activity in HCT116 cells expressing an mCherry-based YAP1 reporter.\u003cstrong\u003e \u003c/strong\u003eRepresentative gating plots for untreated and EPZ-6438 for 72 hours conditions are shown (top). Cells were left untreated or treated with TDI-011536 (10 mM) or EPZ-6438 (50 mM) for 72 hours, with or without drug washout for 1 or 2 days as indicated. The lower panel shows quantification of the percentage of mCherry-positive cells across all conditions (n=3 experimental replicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(C) \u003c/strong\u003eWB analysis (soluble fraction) of PRC2 elements in PDO5 cells untreated or treated with TDI-011536 (10 mM), TRULI (10 mM) or 5-FU+Iri. for 72 hours (from 1 out of 3 experimental replicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(D) \u003c/strong\u003eqPCR analysis of \u003cem\u003eEZH2\u003c/em\u003e in PDO5 untreated or treated with 5-FU+Iri. alone or in combination with Verteporfin (0.25 mM) (left panel) and in control versus YAP1 KO cells treated with 5-FU+Iri. (right panel) (n=2 experimental replicates).\u003c/p\u003e\n\u003cp\u003eStatistical tests; (B) Data are presented as mean ± SEM, (D) Data are presented as mean ± SEM, ordinary one-way ANOVA and Tukey's multiple comparisons test. Source data are provided as a Source Data file.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/1758cde78fb5e40908f774a9.png"},{"id":105038050,"identity":"d42839e4-3ccb-414e-a997-60c5e3eab8e2","added_by":"auto","created_at":"2026-03-20 07:41:51","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":644661,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMYC pathway inhibition promotes quiescence entrance and partial chemotherapy resistance.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Bar plots indicating the mRNA levels of the indicated genes of the MYC signaling pathway in the three datasets analyzed. Notice the significant downregulation of \u003cem\u003eMYC \u003c/em\u003eand upregulation of \u003cem\u003eMAX\u003c/em\u003eand \u003cem\u003eMXD1\u003c/em\u003e in all datasets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(B, C)\u003c/strong\u003e Heatmap representation of the different MYC signaling elements in PDO cells (B) and LS174T CRC cells (C) treated with 5-FU+Iri. (PDO5: n=2 untreated and n=4 treated; PDO66: n=4; LS174T: n=2 untreated and n=3 treated; biological replicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(D, E)\u003c/strong\u003e Heatmap representation of the different MYC signaling elements in PDO cells (D) and LS174T CRC cells (E) treated with EPZ-6438 and/or EED KO#1 and #3 (LS174T: n=2 untreated and n=3 treated; PDO5: n=3; biological replicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(F)\u003c/strong\u003e UMAP representations of scRNA-seq data from PDO cells depicting \u003cem\u003eMYC\u003c/em\u003eexpression levels across cells. The corresponding UMAP showing the distribution of the individual PDO samples is presented in Figure 2H.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(G)\u003c/strong\u003e qPCR analysis of \u003cem\u003eMXD1\u003c/em\u003e, \u003cem\u003eEZH2\u003c/em\u003e and \u003cem\u003eSUZ12\u003c/em\u003e in PDO lines carrying an inducible MXD1 construct untreated or treated with doxycycline (n=3 experimental replicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(H)\u003c/strong\u003e Relative cell growth of the same MXD1-inducible PDO lines untreated or doxycycline-treated as determined by CellTiter-Glo (n=5/5/3 experimental replicates, respectively).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(I)\u003c/strong\u003e Dose-response curves of MXD1-inducible PDO cells treated with 5-FU+Iri. alone or together with doxycycline to induce MXD1 expression (n=3 experimental replicates).\u003c/p\u003e\n\u003cp\u003eStatistical tests; (A) Data are presented as box plots showing the median, the 25th–75th percentiles, and the minimum and maximum values, Wilcoxon test. (G) Data are presented as mean ± SEM, two-way ANOVA. (H, I) Data are presented as mean± SEM. Source data are provided as a Source Data file.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/7e19e70b0a9319336bd403a8.png"},{"id":105039483,"identity":"f84bb7ad-e9d0-4bd9-a9b4-eba48b60d318","added_by":"auto","created_at":"2026-03-20 07:46:24","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1275258,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePolycomb inhibition sensitizes CRC cells to inhibitors of autophagy.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A) \u003c/strong\u003eWB analysis (soluble fraction) of the downstream mTOR pathway effector p-P70 S6 kinase in PDO cells untreated or treated with 5-FU+Iri. for 72 hours (from 1 out of 2 experimental replicates).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(B, C) \u003c/strong\u003eRepresentative images of the autophagy marker LC3B in PDO cells untreated or treated with 5-FU+Iri. or EPZ-6438 as indicated (B) and quantification of the percentage of positive cells per organoid (C) (a minimum of n=12 spheres were examined over 2 biologically independent experiments).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(D)\u003c/strong\u003e Schematic representation of the experimental strategy used to generate the results shown in 6E and 6F (W/O: washout).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(E, F)\u003c/strong\u003e Dose-response curves of PDO cells treated with the autophagy inhibitor bafilomycin A1 for 72 hours as single agent or after pretreatment for 72 hours with 5-FU+Iri. (E) or EPZ-6438 (F) (n=3 experimental replicates, except for PDO4 and PDO5 in (E), n=6 and 5, respectively).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(G)\u003c/strong\u003e Representative images of the autophagy marker LC3B in paired CRC patient samples obtained at the time of diagnosis and post-treatment (left panel), and quantification of the number of tumors showing were decreased/unmodified or increased LC3B-positive dots in post-treatment compared to biopsy samples (right panel) (n=12 tumors from patients who relapsed and n=18 tumors from patients who didn’t relapse were examined).\u003c/p\u003e\n\u003cp\u003eStatistical tests; (C) Data are presented as individual data points indicating the median, two-way ANOVA test. (E, F) Data are presented as mean ± SEM, unpaired two-tailed t test was performed on Area Under the Curve (AUC) values. (G) Data are presented as percentages of the total number, Chi-square test. Source data are provided as a Source Data file.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/43752589367624a69c4d788d.png"},{"id":105038023,"identity":"53ba3123-b1f3-46c3-b34a-b5757c9c3a96","added_by":"auto","created_at":"2026-03-20 07:41:40","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":1044912,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eModel of oncofetal reprogramming and autophagy-dependent vulnerability.\u003c/strong\u003e Suppression of PRC2 and MYC, either intrinsic or therapy-induced, drives fetal/EMT programs and quiescence, converging on an autophagy-addicted state. This state is selectively targetable with autophagy inhibitors, either as second-line therapy following chemotherapy or in combination with PRC2 inhibitors.\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/863230106ad75d94d70ae387.png"},{"id":106874069,"identity":"18e67605-c748-41ce-958c-bd16d0ca0ce9","added_by":"auto","created_at":"2026-04-14 10:13:18","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8208565,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/e78086f4-9ac6-4608-ae20-3c9a2a1d9621.pdf"},{"id":105037904,"identity":"7b40a10c-862c-4d1e-8e5f-3e18f80a7bd8","added_by":"auto","created_at":"2026-03-20 07:40:56","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1108615,"visible":true,"origin":"","legend":"Supplementary Figures and Legends","description":"","filename":"SupplementaryFiguresandLegendsV12.docx","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/774bfb465994a8a5cd988a8b.docx"},{"id":105037906,"identity":"0e12bc3c-17a9-4e4d-9771-71ba80bc0bb1","added_by":"auto","created_at":"2026-03-20 07:40:57","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":9069313,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Data 1\u003c/p\u003e","description":"","filename":"SupplementaryData1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/598134efb45f456160ed51bc.xlsx"},{"id":105039911,"identity":"13269b57-f2ee-4006-8a47-6f4cc9dfe85e","added_by":"auto","created_at":"2026-03-20 07:47:12","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":24228,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Data 2\u003c/p\u003e","description":"","filename":"SupplementaryData2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/ecc49c53f13ca4db101ddcbb.xlsx"},{"id":105039201,"identity":"c1247acb-c8f4-4111-a19c-ff621b15d639","added_by":"auto","created_at":"2026-03-20 07:45:22","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":123962,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Data 3\u003c/p\u003e","description":"","filename":"SupplementaryData3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/2e94f21607a3786da37dc853.xlsx"},{"id":105038990,"identity":"5b29fbc9-0d40-4569-aab1-89d4b69c6eef","added_by":"auto","created_at":"2026-03-20 07:44:50","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":48381251,"visible":true,"origin":"","legend":"Supplementary Data 4","description":"","filename":"SupplementaryData4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/9b6171f9ec1f8b6bb6d59730.xlsx"},{"id":105039034,"identity":"804424c7-767b-4672-87af-fcc01d67df0b","added_by":"auto","created_at":"2026-03-20 07:44:56","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":28250093,"visible":true,"origin":"","legend":"Supplementary Data 5","description":"","filename":"SupplementaryData5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/b133542b8ebdee4737f91d27.xlsx"},{"id":105037912,"identity":"85064088-3bae-47ca-9226-5b99f27853ce","added_by":"auto","created_at":"2026-03-20 07:40:57","extension":"pdf","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":199199,"visible":true,"origin":"","legend":"Figure S1","description":"","filename":"FigureS1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/08aa24b6b6e3c59a42bb9a7a.pdf"},{"id":105038021,"identity":"e7a80731-a963-44d1-bef5-ded9e8629362","added_by":"auto","created_at":"2026-03-20 07:41:40","extension":"pdf","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":414914,"visible":true,"origin":"","legend":"Figure S2","description":"","filename":"FigureS2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/95d07c479f45e0bc55010a9d.pdf"},{"id":105039922,"identity":"385bf321-7049-47ba-aae2-8e0096ffbd35","added_by":"auto","created_at":"2026-03-20 07:47:16","extension":"pdf","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":196234,"visible":true,"origin":"","legend":"Figure S3","description":"","filename":"FigureS3.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/d1c347578d1b96d84b2a5ac6.pdf"},{"id":105037962,"identity":"b15fe1f3-825d-4bf1-969c-4f579afd17a6","added_by":"auto","created_at":"2026-03-20 07:41:10","extension":"pdf","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":318221,"visible":true,"origin":"","legend":"Figure S4","description":"","filename":"FigureS4.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9151067/v1/f2fbef7f27399daddff6be5e.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"PRC2 and MYC repression drives oncofetal reprogramming and autophagy dependence in Colorectal Cancer","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSurgical removal of tumors, with or without adjuvant chemotherapy, remains the standard treatment for colorectal cancer (CRC). However, despite recent advances in targeted therapies and immunotherapy, about 35% of patients relapse and/or develop metastases and eventually die (source: https://www.cancer.net/cancer-types/colorectal-cancer/statistics). Therefore, identifying markers that predict tumor progression and strategies to eradicate cancer cells that persist after first- and second-line treatments remains an urgent clinical need.\u003c/p\u003e\n\u003cp\u003eConversion of intestinal tumor cells into a fetal phenotype, whether associated with specific mutational backgrounds or not, has emerged as a major mechanism used by CRC cells to survive chemotherapy \u003csup\u003e1\u0026ndash;6\u003c/sup\u003e. Acquisition of fetal traits is associated with \u003cstrong\u003equiescence\u003c/strong\u003e\u003cstrong\u003e,\u003c/strong\u003e enhanced \u003cstrong\u003etumor-initiating potential\u003c/strong\u003e\u003cstrong\u003e,\u003c/strong\u003e and \u003cstrong\u003eincreased metastatic competence\u003c/strong\u003e. This process relies, at least in part, on the activation of \u003cstrong\u003eYAP1\u003c/strong\u003e, which is central effector of the Hippo pathway \u003csup\u003e4,6,7\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eFetal conversion is frequently accompanied by induction of selected EMT-associated genes, which is consistent with a hybrid epithelial\u0026ndash;mesenchymal transition \u003csup\u003e4,6\u003c/sup\u003e. This transcriptional configuration has repeatedly been linked to increased dissemination and metastatic colonization across cancer types \u003csup\u003e8\u0026ndash;13\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eRecent studies indicate that specific oncogenic pathways, notably KRAS,\u0026nbsp;play a critical role in maintaining the fetal-like transcriptional state. In \u003cem\u003eKRAS\u003c/em\u003e-mutant CRC, \u003cstrong\u003epharmacological inhibition of KRAS reverses fetal conversion\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eand restores the capacity of chemotherapy to eliminate adult \u003cem\u003eLGR5\u003c/em\u003e\u003cem\u003e⁺\u003c/em\u003e cancer stem cells \u003csup\u003e14\u003c/sup\u003e. \u003cem\u003eKRAS\u003c/em\u003e-mutant CRCs also exhibit marked sensitivity to the combined inhibition of Polycomb Repression Complex 2 (PRC2) with chemotherapy, which is\u0026nbsp;frequently upregulated in human CRC \u003csup\u003e15\u003c/sup\u003e,\u0026nbsp;but not yet explored in the context of fetal-type tumors, and RAS-pathway signaling \u003csup\u003e16,17\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWe recently identified an oncofetal signature, termed ColoStem, which is induced upon chemotherapy treatment but is also present in a subset of treatment-na\u0026iuml;ve CRC tumors. Importantly, ColoStem predicts poor patient prognosis independently of \u003cem\u003eKRAS\u003c/em\u003e or \u003cem\u003eBRAF\u003c/em\u003e mutational status \u003csup\u003e17\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eHere, we demonstrate that PRC2 repression promotes activation of fetal and EMT-associated transcriptional programs in CRC cells, concomitant with attenuation of MYC signaling, and confers chemotherapy resistance in vivo. Furthermore, acquisition of fetal traits following chemotherapy or PRC2 inhibition establishes a marked dependency on autophagy for cancer cell survival.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eFetal-type tumors exhibit reduced PRC2 and MYC activities\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe investigated whether treatment-na\u0026iuml;ve fetal-type CRC tumors, classified based on the previously published 8-gene fetal signature, ColoStem \u003csup\u003e17\u003c/sup\u003e (see also Methods),\u0026nbsp;share a higher-order regulatory landscape that could reveal mechanisms underlying the oncofetal phenotype.\u003c/p\u003e\n\u003cp\u003eTo address this, we stratified CRC tumors from the Marisa \u003csup\u003e18\u003c/sup\u003e, TCGA and Jorissen \u003csup\u003e19\u003c/sup\u003e datasets into fetal-type or non-fetal-type groups. Transcriptomic analysis identified robust differentially expressed gene (DEG) signatures across cohorts (Figures 1A-C, S1A and Supplementary Data 1). To infer upstream regulatory programs associated with these DEGs, over-representation analysis (ORA) was performed with enrichR using the Chromatin Histone Enrichment Analysis (ChEA) transcription factor target database (ENCODE_and_ChEA_Consensus_TFs_from_ChIP-X). This analysis revealed a significant enrichment of Polycomb Repression Complex 2 (PRC2)-associated targets (SUZ12-upregulated gene sets) as well as MYC-downregulated gene sets (Figure 1D), similar to that observed in slow-cycling cancer populations associated with poor clinical outcomes \u003csup\u003e20\u003c/sup\u003e. In contrast, we observed no significant enrichment for AP1-dependent programs, which have previously been implicated in oncofetal inducer in the AKSP (\u003cem\u003eAPC\u003csup\u003enull\u003c/sup\u003e\u003c/em\u003e; \u003cem\u003eKRAS\u003csup\u003eG12D\u003c/sup\u003e\u003c/em\u003e; \u003cem\u003eSMAD4\u003csup\u003enull\u003c/sup\u003e\u003c/em\u003e; \u003cem\u003eTP53\u003csup\u003enull\u003c/sup\u003e\u003c/em\u003e) mouse model \u003csup\u003e4\u003c/sup\u003e and in human PDOs \u003csup\u003e21,22\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eAcross datasets, enrichment of SUZ12-regulated gene sets in fetal-type tumors was highly consistent\u0026nbsp;with \u0026gt;71% overlap in at least two cohorts and \u0026gt;52% overlap across all three cohorts (Figure 1E and Supplementary Data 2).\u0026nbsp;MYC target downregulation was similarly conserved, with \u0026gt;62% overlap across cohorts (Figure 1F).\u003c/p\u003e\n\u003cp\u003eGene Set Enrichment Analysis (GSEA) of SUZ12 targets identified TGFb, Wnt, MAPK and Hippo/YAP1 as putative upstream regulators in fetal-type tumors (Figure 1G), consistent with our results in chemotherapy-treated PDOs \u003csup\u003e6\u003c/sup\u003e. Downregulated MYC targets were enriched for pathways controlling protein synthesis (ribosome, ribosome biogenesis, cofactor biosynthesis) and proliferation (cell cycle, DNA replication, nucleotide metabolism) (Figure 1H).\u003c/p\u003e\n\u003cp\u003eKaplan-Meier analyses of a meta-cohort comprising 1,067 CRC patients \u003csup\u003e17\u003c/sup\u003e (Figures 1I and 1J), as well as independent TCGA validation (Figures S1B and S1C), demonstrated that stratification based on SUZ12- or MYC-target expression (high vs. low) revealed significant differences in disease-free survival. In both cases, high SUZ12-target and low MYC-target expression were associated with poorer prognosis, mirroring the adverse survival pattern previously observed with the ColoStem fetal signature \u003csup\u003e6,17\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eThese findings highlight the context-dependent roles of PRC2 and MYC in CRC biology. While MYC is widely recognized as an oncogenic driver, its tumor-suppressive functions, including the ability to promote apoptosis under specific conditions, are well established \u003csup\u003e23,24\u003c/sup\u003e. In the setting of oncofetal reprogramming, reduced PRC2 and MYC pathway activity is associated with poor prognosis, consistent with this context-dependent functional behavior \u003csup\u003e25\u0026ndash;27\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTranscriptional repression of PRC2 in therapy-na\u0026iuml;ve fetal-type tumors and chemotherapy-treated CRC cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe investigated whether the reduced PRC2 activity observed in fetal-type tumors could be attributed to decreased expression of specific PRC2 components. PRC2 consists of a catalytic core (EZH2, SUZ12 and EED) that associates with alternative accessory modules defining PRC2.1 (PCL proteins, including MTF2/PCL2, as well as EPOP and PALI1/2) or PRC2.2 (AEBP2 and JARID2) \u003csup\u003e28,29\u003c/sup\u003e. Across all datasets analyzed, we found that the core PRC2 subunits\u003cstrong\u003e\u003cem\u003eSUZ12\u003c/em\u003e\u003c/strong\u003eand \u003cstrong\u003e\u003cem\u003eEZH2\u003c/em\u003e\u003c/strong\u003ewere consistently and significantly downregulated in treatment-na\u0026iuml;ve fetal-type tumors. This reduction extended to the PRC2.1 subunits \u003cstrong\u003e\u003cem\u003eMTF2\u003c/em\u003e\u003c/strong\u003e and \u003cstrong\u003e\u003cem\u003eEPOP\u003c/em\u003e\u003c/strong\u003e (Figure 2A). These findings were validated in our CRC metacohort, where PRC2 transcripts were similarly reduced (Figure S2A).\u003c/p\u003e\n\u003cp\u003eTo refine these observations, we examined RNA-seq data from PDOs previously classified as non-fetal (PDO4, PDO10 and PDO66) or fetal (PDO5, PDO27 and PDO127) based on the ColoStem signature \u003csup\u003e17\u003c/sup\u003e. Fetal PDOs showed only a modest reduction in core PRC2 transcripts but exhibited a marked depletion of PRC2.1 elements, specifically \u003cem\u003eMTF2\u003c/em\u003e and \u003cem\u003eEPOP\u003c/em\u003e (Figure 2B). Mass spectrometry analysis confirmed a significant decrease in core PRC2 protein levels in fetal PDOs (Figure 2C and Supplementary Data 3). PRC2.1 and PRC2.2 components were not consistently detected by mass spectrometry.\u003c/p\u003e\n\u003cp\u003eWe next assessed whether chemotherapy exposure further impaired PRC2 expression. In both fetal (PDO5) and non-fetal (PDO66) PDOs, chemotherapy treatment caused a profound decrease in \u003cem\u003eSUZ12\u003c/em\u003e, \u003cem\u003eEZH2\u003c/em\u003e, \u003cem\u003eEED\u003c/em\u003e, \u003cem\u003eMTF2\u003c/em\u003e and \u003cem\u003eEPOP\u003c/em\u003e RNA levels, as determined by RNA-seq analysis (Figure 2D). We validated PRC2 downregulation at the protein level by Western blot (WB) analysis (Figures 2E and 2F). Similar results were obtained in CRC cell lines LS174T and HCT116 lines after chemotherapy treatment (Figure 2G and Supplementary Data 4). Consistently, heatmap analysis confirmed coordinated repression of core PRC2 components upon chemotherapy exposure (Figure S2B). In parallel, GSEA revealed significant enrichment of fetal-like and EMT transcriptional programs in chemotherapy-treated cells (Figure S2C). These findings recapitulate the transcriptional remodeling observed in PDO5 and PDO66 \u003csup\u003e6\u003c/sup\u003e, further supporting a link between chemotherapy-induced PRC2 suppression and acquisition of a fetal-like/EMT-associated state.\u003c/p\u003e\n\u003cp\u003eTo further validate the relationship between PRC2 repression and fetal identity of CRC cells, we analyzed scRNA-seq data from eight PDO samples, including six untreated PDOs and two 5-FU+Iri.-treated counterparts, one originally classified as fetal (PDO5) and one as non-fetal (PDO66) (Figure 2H). Uniform Manifold Approximation and Projection (UMAP) visualization revealed that chemotherapy-treated PDOs (which induce fetal conversion) tended to cluster closer to treatment-na\u0026iuml;ve fetal PDOs, suggesting convergence toward a shared transcriptional state. Notably, PDO127 did not cluster tightly with the other fetal PDOs, consistent with our previous bulk RNA-seq analyses in which PDO127, although grouping within the fetal cluster, displayed a partially divergent transcriptional profile \u003csup\u003e17\u003c/sup\u003e. The indicated fetal-like gene signatures were enriched in the same populations, confirming their fetal transcriptional identity. Consistently, these treatment-na\u0026iuml;ve and chemotherapy-induced fetal PDOs displayed lower \u003cem\u003eEZH2\u003c/em\u003e expression compared with non-fetal untreated counterparts. Together, these findings indicate that reduced PRC2 expression is an intrinsic feature of treatment-na\u0026iuml;ve fetal-type tumors, while chemotherapy promotes a shift toward this pre-existing fetal-like state.\u003c/p\u003e\n\u003cp\u003eCollectively, these findings identify transcriptional repression of PRC2 components, particularly PRC2.1, as a shared feature of treatment-na\u0026iuml;ve fetal-type tumors and\u0026nbsp;chemotherapy-induced oncofetal conversion.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePRC2 inactivation unlocks oncofetal and EMT signatures and imposes chemotherapy resistance \u003cem\u003ein vivo\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next investigated whether loss of PRC2 function alone can trigger the transcriptional programs induced by sublethal chemotherapy \u003csup\u003e6\u003c/sup\u003e. To address this, we performed RNA-seq on CRC PDOs and cell lines subjected either to pharmacologic EZH2 inhibition with tazemetostat (EPZ-6438) or CRISPR-Cas9\u0026ndash;mediated knockout of EED (Figure S3A). Both perturbations induced a pronounced transcriptional response in PDO5 and LS174T cells (Figures 3A, S3B, S3C and Supplementary Data 5), significantly overlapping with multiple published fetal intestinal gene signatures \u003csup\u003e3,30,31\u003c/sup\u003e. Consistently, EPZ-6438 treatment reduced the expression of canonical adult intestinal stem cell markers, including \u003cem\u003ePROM1 \u003csup\u003e32\u003c/sup\u003e\u003c/em\u003e, \u003cem\u003eCDCA7 \u003csup\u003e33,34\u003c/sup\u003e\u003c/em\u003e, \u003cem\u003eLGR5\u0026nbsp;\u003c/em\u003eor\u003cem\u003e\u0026nbsp;ASCL2 \u003csup\u003e35,36\u003c/sup\u003e\u003c/em\u003e (Figure 3B).\u003c/p\u003e\n\u003cp\u003eNext, we tested whether PRC2 inhibition functionally recapitulates the phenotype associated with chemotherapy-induced fetal conversion.\u0026nbsp;In vitro, PRC2 inactivation did not fully reproduce the quiescent and therapy‑resistant phenotype across PDOs\u0026nbsp;(Figure S3D).\u0026nbsp;However, in the zebrafish patient-derived xenograft (zAvatar) model \u003csup\u003e37,38\u003c/sup\u003e, EED KO PDO5 cells produced tumors that were smaller than those generated by the CRISPR control counterparts (Figure 3C), which is consistent with a reduced proliferative phenotype. Notably, PDO5 tumors treated with chemotherapy showed increased levels of apoptosis as determined by active caspase 3 staining, which was not detected in EPZ-6438-treated and EED KO tumors (Figure 3D-F). This contrasts with the observations made in the CRISPR control tumors, where a clear activation of apoptosis was observed with the treatment (Figure 3E and 3F).\u003c/p\u003e\n\u003cp\u003eIn addition to activating the fetal program (or as part of this phenotype), PRC2 inactivation induced a robust EMT transcriptional response\u0026nbsp;in\u0026nbsp;PDO5 cells (Figure 3G). Supporting this observation, EED KO PDO5 cells displayed a trend toward increased in vivo metastatic potential upon orthotopic transplantation into immunodeficient mice (Figure 3H).\u003c/p\u003e\n\u003cp\u003eTo gain mechanistic insight, we performed GSEA on transcriptional changes imposed by PRC2 inactivation in the different CRC models. Pathways involved in intestinal stem cell regulation and CRC progression, including Notch \u003csup\u003e39\u003c/sup\u003e, Wnt \u003csup\u003e40,41\u003c/sup\u003e and TGFb\u0026nbsp;\u003csup\u003e42\u003c/sup\u003e were significantly upregulated, whereas proliferative drivers such as MYC, E2F and mTOR pathways were predominantly and reproducibly downregulated (Figures S3F and S3G).\u003c/p\u003e\n\u003cp\u003eTogether, these findings indicate that PRC2 loss is sufficient to unlock oncofetal and EMT transcriptional programs in CRC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePRC2 repression induces sustained YAP1 activation independently of canonical Hippo feedback\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYAP1 has emerged as a central regulator of fetal-like reprogramming, epithelial\u0026ndash;mesenchymal plasticity and therapy resistance in colorectal cancer \u003csup\u003e6,43,44\u003c/sup\u003e. Given the strong overlap between YAP1-driven transcriptional programs and those induced by PRC2 repression, we asked whether loss of PRC2 activity directly impacts YAP1 signaling.\u003c/p\u003e\n\u003cp\u003eTo address this, we first performed GSEA using a Hippo signaling signature in PDO5 and LS174T cells. PRC2 loss resulted in significant enrichment of YAP1-associated gene programs compared to untreated cells (Figure 4A), indicating enhanced YAP1 transcriptional output upon PRC2 repression.\u003c/p\u003e\n\u003cp\u003eWe directly monitored YAP1 activity using a YAP1-responsive fluorescent reporter. Pharmacological inhibition of PRC2 resulted in an increase in YAP1 reporter activity at 72 hours (Figure 4C). As a positive control for YAP1 activation, we pharmacologically activated YAP1 downstream of the Hippo pathway using the LATS kinase inhibitor TDI-011536 for 72 hours, a condition that robustly increased active YAP1 levels. Notably, YAP1 activity remained elevated for at least 24 hours following inhibitor washout, indicating that PRC2 repression induces a sustained YAP1-active state rather than a transient signaling response.\u003c/p\u003e\n\u003cp\u003eWe then asked whether YAP1 activation could, in turn, account for the repression of PRC2 components observed in fetal-like CRC cells. To address this, we activated YAP1 using the LATS kinase inhibitors TDI-011536 and TRULI. Although these treatments effectively enhanced YAP1 signaling, they did not decrease PRC2 protein levels; instead, core PRC2 components were modestly increased, as determined by WB analysis (Figure 4D).\u003c/p\u003e\n\u003cp\u003eFinally, we evaluated PRC2 regulation by YAP1 by either using direct YAP1 inhibition or genetic ablation. Neither pharmacological inhibition of YAP1 by Verteporfin nor YAP1 knockout reduced \u003cem\u003eEZH2\u003c/em\u003e transcript levels, as determined by qPCR analysis in PDOs (Figure 4E). YAP1 knockout efficiency had been previously validated (Figure S4A in \u003csup\u003e6\u003c/sup\u003e).\u003c/p\u003e\n\u003cp\u003eTogether, these findings establish a hierarchical and unidirectional relationship between PRC2 and YAP1 during oncofetal reprogramming, in which epigenetic repression of PRC2 precedes and enables sustained YAP1 activation, but YAP1 activation alone is insufficient to modulate PRC2 expression.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMYC pathway inhibition promotes quiescence entrance and partial chemotherapy resistance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMYC is a central regulator of the intestinal crypt homeostasis and APC-driven intestinal tumorigenesis \u003csup\u003e45\u0026ndash;47\u003c/sup\u003e. However, MYC signaling is markedly suppressed in CRC cells following chemotherapy treatment \u003csup\u003e2,6\u003c/sup\u003e (Figure S4A) and is consistently downregulated in fetal-type tumors (Figure 1D). We aimed to investigate the bases underlying MYC pathway downregulation in na\u0026iuml;ve-treatment and chemotherapy-induced fetal-type CRC cells. Analysis of the public dataset revealed that \u003cem\u003eMYC\u003c/em\u003e levels were reduced in fetal tumors, whereas its obligate partner \u003cem\u003eMAX\u0026nbsp;\u003c/em\u003ewas significantly upregulated (Figures 5A and S4B).\u003c/p\u003e\n\u003cp\u003eIn the chemotherapy-induced oncofetal model, \u003cem\u003eMYC\u003c/em\u003e mRNA levels decreased upon chemotherapy treatment in PDO66 and the LS174T cell line, two models displaying robust suppression of MYC targets after treatment, but remained unchanged or increased in PDO5 (Figures 5B and 5C). Mirroring treatment-na\u0026iuml;ve fetal-type tumors, \u003cem\u003eMAX\u003c/em\u003e levels were upregulated following chemotherapy treatment.\u0026nbsp;Notably, \u003cem\u003eMXD1\u003c/em\u003e, an endogenous MYC antagonist \u003csup\u003e48\u003c/sup\u003e, was consistently induced across all oncofetal contexts.\u0026nbsp;In the chemotherapy setting (Figures 5B and 5C) as well as upon pharmacological PRC2 inhibition with EPZ-6438 (Figures 5D and 5E), whereas \u003cem\u003eMXD3\u003c/em\u003e and \u003cem\u003eMXD4\u003c/em\u003e displayed more variable behavior (Figures 5A\u0026ndash;E and Figure S4C). However, we detected some differences between EPZ-6438 treatment and EED KO cells, suggesting that EPZ-6438 may exert effects beyond PRC2 inhibition in a context-dependent manner\u003c/p\u003e\n\u003cp\u003eTo further assess \u003cem\u003eMYC\u003c/em\u003e expression at the single-cell level, we interrogated the scRNA-seq dataset from PDOs described before (Figure 2H). UMAP visualization revealed reduced \u003cem\u003eMYC\u003c/em\u003e expression in treatment-na\u0026iuml;ve fetal PDOs compared with non-fetal counterparts (Figure 5F). Consistently, chemotherapy-treated PDOs also exhibited diminished \u003cem\u003eMYC\u003c/em\u003e expression and localized closer to fetal PDOs in the transcriptional space. These data further support the convergence toward a fetal-like state characterized by MYC pathway suppression.\u003c/p\u003e\n\u003cp\u003eTo functionally interrogate this regulatory axis, we generated a doxycycline-inducible \u003cem\u003eMXD1\u003c/em\u003e model. WB analysis confirmed strong nuclear MXD1 induction upon 16 hours of doxycycline treatment in HEK293T (Figure S4D) and three PDO lines (Figure 5G and Figure S4E). Induced \u003cem\u003eMXD1\u003c/em\u003e expression markedly reduced cell growth in 3D cultures (Figures 5H and S4F), consistent with MYC pathway suppression. We then tested whether ectopic MXD1 modulates chemotherapy response in PDO cells. We found that doxycycline treatment conferred partial but reproducible resistance to 5-FU+iri in PDO cells (Figure 5I).\u003c/p\u003e\n\u003cp\u003eThese results indicate that MXD1-mediated downregulation of the MYC pathway contributes to growth suppression and partially promotes chemotherapy resistance.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePolycomb inhibition and sublethal chemotherapy sensitizes CRC cells to inhibitors of autophagy\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAutophagy enables cellular adaptation and survival under stress conditions \u003csup\u003e49\u003c/sup\u003e and is essential for maintaining embryonic diapause under adverse environmental conditions \u003csup\u003e50\u0026ndash;52\u003c/sup\u003e. Diapause-like, drug-tolerant states have been repeatedly linked to cancer cell persistence in several cancer models \u003csup\u003e2,5,53\u003c/sup\u003e. However, the therapeutic benefit of targeting autophagy in cancer remains controversial \u003csup\u003e54\u003c/sup\u003e, and its functional connection to oncofetal reprogramming has not been systematically explored.\u003c/p\u003e\n\u003cp\u003eOur transcriptomic analysis of treatment-na\u0026iuml;ve fetal-type CRC cells (Figure 1D and S1A), as well as cells following sublethal chemotherapy exposure \u003csup\u003e6\u003c/sup\u003e or PRC2 suppression (Figures S3E and S3F) revealed a robust inhibition of the MYC, E2F and/or mTOR pathways. As these pathways are canonical regulators of autophagy across multiple systems \u003csup\u003e50,55\u0026ndash;59\u003c/sup\u003e, these results pointed to autophagy as a potential shared vulnerability of the oncofetal state. Consistent with this notion, WB analysis of PDOs treated with chemotherapy or PRC2 inhibitors for 72 hours showed a marked reduction in phospho-S6 (and/or phospho-S6K) levels indicating inhibition of the mTOR\u0026ndash;S6 axis under both oncofetal-inducing conditions (Figure 6A). These data suggest that oncofetal reprogramming is coupled to an mTOR-low state, a metabolic configuration known to be permissive for increased dependency on lysosomal and autophagy-dependent pathways.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo functionally test this possibility, we screened our PDO collection, including non-fetal (PDO4, PDO10 and PDO66) and treatment-na\u0026iuml;ve fetal-type lines (PDO5, PDO27 and PDO127). By immunofluorescence (IF) of the autophagosome marker LC3B \u003csup\u003e60\u003c/sup\u003e, we detected heterogeneous levels of basal autophagy across PDOs, which increased after chemotherapy or EPZ-6438 treatment, ultimately converging to comparable LC3B-high levels (Figure 6B and 6C).\u003c/p\u003e\n\u003cp\u003eWhile basal autophagy levels showed limited predictive value for sensitivity to autophagy inhibition (Figure S5A), induction of fetal-like states by chemotherapy or PRC2 inhibition (experimental design in Figure 6D) consistently increased vulnerability to lysosomal blockade (Figures 6E, 6F and S5B) in all PDOs tested.\u003c/p\u003e\n\u003cp\u003eWe next asked whether autophagy was also elevated in advanced human tumors following chemotherapy. Analysis of matched CRC biopsies collected before and after neoadjuvant therapy revealed heterogeneous LC3B-positive vesicle accumulation in a subset of cases. Notably, an increase in LC3B accumulation after treatment showed a trend toward association with poor outcome (Figure 6G). Post-treatment samples from patients who relapsed more often showed increased LC3B levels relative to their matched pre-treatment biopsies, whereas tumors from non-relapsing patients more frequently exhibited no change or reduced LC3B accumulation. Although not observed in all cases, these data suggest a link between therapy-induced LC3B accumulation and relapse.\u003c/p\u003e\n\u003cp\u003eTogether, our results support a model in which PRC2 repression and MYC attenuation cooperate with previously described YAP1-driven programs to promote fetal and therapy-resistant states \u003csup\u003e6,7,12,61\u003c/sup\u003e. Importantly, acquisition of the oncofetal phenotype is consistently associated with increased autophagic activity or reliance on autophagy across experimental contexts (Figure 7).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe emergence of an oncofetal phenotype has been increasingly recognized as a mechanism of therapeutic resistance and metastatic progression in CRC. However, the molecular basis of this reprogramming and its potential therapeutic implications have remained unclear. Here, we show that suppression of PRC2 activity is a consistent feature of both intrinsic and therapy-induced fetal-type tumors, in apparent contradiction to the conventional view of PRC2 as an oncogenic driver. Our results indicate that, within this context, PRC2 loss supports cellular plasticity by releasing developmental constraints and promoting EMT and fetal-like transcriptional programs. This dual behavior, oncogenic in some contexts yet permissive to cellular reprogramming in others, may underlie the context‑dependent functions attributed to PRC2 across cancers.\u003c/p\u003e\n\u003cp\u003eThe coordinated repression of PRC2 and MYC observed in fetal-type CRC cells reflects the integration of two functionally distinct processes: PRC2 loss facilitates transcriptional plasticity and lineage regression, whereas MYC attenuation promotes quiescence and metabolic adaptation. The induction of MXD1 across all fetal-like contexts, including PRC2 inhibition, suggests the existence of a feedback circuit in which PRC2 loss promotes MYC pathway suppression. However, forced MXD1 expression alone does not activate fetal transcriptional programs, indicating that MYC suppression is necessary but not sufficient for fetal reprogramming.\u003c/p\u003e\n\u003cp\u003eImportantly, this fetal-like state, which enables cell survival under cytotoxic stress, also imposes an increased dependency on autophagy for energy homeostasis and survival, thus creating a transient (but clinically relevant) therapeutic window (see Figure 7). Our results demonstrate that chemotherapy-induced and PRC2-inactivated CRC cells showed enhanced sensitivity to lysosomal inhibitors such as chloroquine and bafilomycin A1, irrespective of their initial phenotype. These data are consistent with the concept that cancer cells use autophagy as a mechanism of therapeutic resistance (reviewed in \u003csup\u003e62\u003c/sup\u003e), but contrast with previous reports describing a limited impact of autophagy blockade once cells have already acquired a diapause-like phenotype \u003csup\u003e2\u003c/sup\u003e. Importantly, the observation that PRC2 inhibition alone can reproduce the autophagy-dependent phenotype provides a rationale for combining PRC2-targeting agents with autophagy inhibitors in future therapeutic designs. Our analysis of matched patient samples further suggests that elevated autophagy may characterize a subset of treatment-na\u0026iuml;ve tumors prior to therapy, rather than being uniformly induced by chemotherapy. In this context, neoadjuvant treatment appears to remodel or select against highly autophagic, fetal-like tumor cell states, highlighting autophagy as a dynamic feature of tumor plasticity rather than a static response to cytotoxic stress.\u003c/p\u003e\n\u003cp\u003eAt first glance, our findings that partial inactivation of PRC2 or MYC in CRC unlocks a highly plastic, fetal-like transcriptional program appear to contrast with studies in germinal-center B cells, where increased MYC activity and preservation of PRC2-dependent programs, such as those described by Melnick and colleagues \u003csup\u003e63,64\u003c/sup\u003e, promote clonal fitness and malignant transformation. We propose that these observations are not contradictory but instead highlight the profound context-dependence of PRC2\u0026ndash;MYC signaling. In epithelial tissues, PRC2 and MYC maintain adult lineage identity, and their attenuation facilitates regression towards embryonic, stress-tolerant states. In contrast, germinal-center B cells reside in an inherently proliferative and epigenetically permissive environment, in which additional activation of MYC or PRC2 amplifies competitive expansion. Thus, both gain- and loss-of-function perturbations of the PRC2\u0026ndash;MYC axis can converge on tumor-promoting plasticity, depending on the cellular baseline and epigenetic landscape.\u003c/p\u003e\n\u003cp\u003eClinically, these findings carry several important implications for patient stratification and therapeutic design. First, they highlight that the fetal-like transcriptional state, detectable in a subset of treatment-na\u0026iuml;ve tumors, marks a distinct biological and therapeutic subgroup. Second, PRC2 or MYC suppression, which can be evaluated by immunohistochemistry or transcriptomic profiling, could serve as actionable biomarkers to identify patients likely to benefit from autophagy-based combination therapies. Third, as PRC2 inactivation is also recurrent in other malignancies, including T-cell Leukemia\u0026nbsp;\u003csup\u003e65\u003c/sup\u003e, pediatric glioblastoma \u003csup\u003e66,67\u003c/sup\u003e, \u003cem\u003eMalignant Peripheral Nerve Sheath Tumors (MPNST)\u0026nbsp;\u003c/em\u003e\u003cem\u003e\u003csup\u003e\u003cem\u003e68\u003c/em\u003e\u003c/sup\u003e\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003eor \u003cem\u003eNFKBIA\u003c/em\u003e hemizygous glioblastoma\u0026nbsp;\u003csup\u003e69,70\u003c/sup\u003e, our findings may have broader relevance beyond CRC.\u003c/p\u003e\n\u003cp\u003eOverall, rather than attempting to prevent oncofetal conversion, our data support exploiting transient reprogrammed states as therapeutic windows, in which cancer cells reveal specific metabolic and survival dependencies. Our data suggest that both sublethal chemotherapy and pharmacologic PRC2 inhibition not only induce or reinforce oncofetal reprogramming, but also increase autophagy dependence, pushing CRC cells into a metabolically vulnerable hyper-dependent state. This provides a dual therapeutic opportunity: autophagy inhibitors as a rational second-line approach following standard chemotherapy or neoadjuvancy, and combination strategies pairing PRC2 inhibitors with autophagy blockade to selectively eradicate cancer cells that have adopted a survival-prone, fetal-like program. Thus, oncofetal reprogramming should not be viewed merely as an obstacle to therapy, but as an opportunity, a transient and druggable state in which tumor cells reveal vulnerabilities that can be strategically exploited.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eColorectal Cancer cell lines\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCRC cell lines\u0026nbsp;HCT116 (CCL-247) and LS174T (CL-188) were obtained from the American Type Culture Collection [ATCC, USA]. Cell lines were grown in Dulbecco\u0026rsquo;s modified Eagle\u0026rsquo;s medium [Invitrogen] supplemented with 10% fetal bovine serum [Biological Industries] and were maintained in a 5% CO\u003csub\u003e2\u003c/sub\u003e incubator at 37 \u0026deg;C. Cell mutations and concentrations of chemotherapy used for each cell line are indicated in the Supplementary Table 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePDO generation and culture conditions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHuman colorectal tumors were obtained from Parc de Salut MAR Biobank (MARbiobank) and IdiPAZ Biobank (PT23/00028),\u0026nbsp;integrated into the Spanish Hospital Biobanks Network (RetBioH; www.redbiobancos.es). Written informed consent was obtained from all participants and protocols were approved by Hospital del Mar\u0026rsquo; Ethics Committee (CEImPSMar_E1.2025-11968-I; and previous approval 2019/8595/I) and were subsequently ratified by the Clinical Research Ethics Committee of Hospital Universitario La Paz/IdiPAZ (approval code 2025.699), in accordance with Spanish regulations and the Declaration of Helsinki.\u003c/p\u003e\n\u003cp\u003eFor PDOs generation, primary or xenografted human colorectal tumors were disaggregated in 1.5 mg/mL collagenase II and 20 \u0026mu;g/mL hyaluronidase after 40 min of incubation at 37 \u0026deg;C, filtered in 100 \u0026mu;m cell strainer, and seeded in 50 \u0026mu;L Matrigel in 24-well plates. After polymerization, 450 \u0026mu;L of complete medium was added (DMEM/F12 plus penicillin (100 U/mL) and streptomycin (100 \u0026mu;g/mL), 100 \u0026mu;g/mL Primocin, 1\u0026times; N2 and B27, 10mM Nicotinamide; 1.25 mM N-Acetyl-L-cysteine, 100 ng/mL Noggin and 100 ng/mL R-spondin-1, 10 \u0026mu;M Y-27632, 10nM PGE2, 3\u0026mu;M SB202190, 0.5\u0026mu;MA8301, 50 ng/mL EGF and 10nM Gastrin I). Cultures were maintained at 37\u0026deg;C, 5% CO2 and medium changed every week. PDOs were expanded by serial passaging and kept frozen in liquid Nitrogen for being used in subsequent experiments. PDOs were routinely tested for mycoplasma contamination. PDOs mutations and concentrations of chemotherapy used for each PDO are indicated in the Supplementary Table 1. The concentrations of the other treatments were as follows: EPZ-6438, 50 \u0026micro;M;\u0026nbsp;TDI-011536, 10 \u0026micro;M; TRULI, 10 \u0026micro;M; Verterpofin, 0.25 \u0026micro;M; and doxycycline, 2 \u0026micro;g/mL.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLentiviral transduction of cells\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003elentiCRISPR v2 plasmid was used for knock-out experiments. The sgRNA against EED gene were designed using Benchling (Supplementary Table 2). MXD1 inducible plasmid was constructed modifying the pLIX-hN1ICD plasmid inserting the \u003cem\u003eMXD1\u003c/em\u003e gene [Addgene #91897]. Lentiviral production was performed by transfecting HEK293T cells the lentiviral vectors and the plasmid of interest. One day after transfection, the medium was changed and viral particles were collected 24 h later and then concentrated using Lenti-X Concentrator. PDOs were infected by resuspending single cells in concentrated viruses diluted in complete medium, centrifuged for 1 hours at 650 rcf, and incubated for 5h at 37\u0026deg;C. PDOs were then washed with complete culture medium and seeded as described above. After 72 hours, puromycin at 1\u0026nbsp;mg/ml was added for selection of infected cells for one week, in PDOs infected with sgRNA.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePDOs Immunofluorescence analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParaffin blocks were obtained from PDOs after previous fixation in 4% formaldehyde overnight at room temperature. Paraffin-embedded sections of 2.5 \u0026mu;m were deparaffinized, rehydrated, citrate‐based antigen retrieval was used (20 min, no pressure) and endogenous peroxidase activity was quenched (20 min, 1.5% H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e). All primary antibodies were diluted in PBS containing 0.05% BSA, incubated overnight at 4\u0026deg;C (Supplementary Table 3). Samples were then incubated with the Envision+ System HRP Labeled Polymer anti-Rabbit or anti-Mouse [Dako] for 1.5 h and then developed with the Tyramide Signal Amplification System (TSA) [PerkinElmer] and mounted in DAPI Fluoromount-G. Images were taken in an SP8 confocal microscope (Leica).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWestern Blot (WB)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePDOs and cell lines were lysed for 20 min on ice in 150 \u0026mu;L of PBS plus 0.5% Triton X-100, 1 mM EDTA, 100 mM sodium orthovanadate, 0.2 mM phenyl-methylsulfonyl fluoride (PMSF), and complete protease and phosphatase inhibitor cocktails. Lysates were first cleared by centrifugation at maximum speed for 10 min at 4\u0026deg;C. The supernatant was transferred to a new tube, sonicated for 10 min (10 cycles of 30 s ON/30 s OFF), and mixed with 6X Laemmli buffer (soluble fraction). The remaining pellet fraction was resuspended in 1X Laemmli buffer and sonicated under the same conditions (insoluble fraction).\u003c/p\u003e\n\u003cp\u003eSamples were boiled at 95\u0026deg;C for 10 min and analyzed by WB using standard SDS\u0026ndash;polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were resolved on polyacrylamide gels and transferred onto polyvinylidene difluoride (PVDF) membranes. Membranes were incubated overnight at 4\u0026deg;C with the appropriate primary antibodies (Supplementary Table 4), followed by incubation with horseradish peroxidase\u0026ndash;conjugated secondary antibodies. Signal detection was performed using enhanced chemiluminescence and imaged with the iBright CL750 Imaging System [Invitrogen].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRT-qPCR analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTotal RNA from cell lines and PDOs was extracted with the RNeasy Mini Kit and RNeasy Micro Kit [QIAGEN] respectively, and cDNA was produced with the RT-First Strand cDNA Synthesis Kit [Roche]. RT-qPCR was performed in LightCycler 480 system using SYBR Green I Master Kit. Samples were normalized relative to the housekeeping genes. Primers used for RT-qPCR are listed in Supplementary Table 5.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePDO viability assays\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e600 single PDO cells were plated in p96-well plates in 10 \u0026mu;L of Matrigel with 100 \u0026mu;L of complete medium. For growth analysis of \u003cem\u003eMXD1\u003c/em\u003e-inducible PDOs, after 6 days in culture, growing PDOs were treated with doxycycline at 1\u0026nbsp;mg/ml, which was refreshed every 72 hours. Cell viability was measured after 3 days, 6 days and 9 days from the first treatment, using the CellTiter-Glo 3D Cell Viability Assay [Promega] following manufacturer\u0026rsquo;s instructions, in an Orion II multiplate luminometer. For dose-response curves, PDOs were plated in p96-well plates in Matrigel and after 6 days in culture were treated with combinations of 5-FU+Iri. (at the concentrations indicated in Supplementary Table 1), EPZ-6438 at 50\u0026nbsp;mM or doxycycline at 1\u0026nbsp;mg/ml for 72 hours. Following 72 hours of treatment, medium was changed and PDOs were treated with increasing concentrations of either 5-FU+Iri. or Bafilomycin A1 for 72 hours. Cell viability was determined as described above.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eZebrafish patient-derived xenograft microinjection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAnimal care and handling\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIn vivo\u003c/em\u003e experiments were performed using zebrafish (\u003cem\u003eDanio rerio\u003c/em\u003e) strains - nacre, casper and \u003cem\u003eTg(Fli1:eGFP)\u003c/em\u003e, which were handled according to the standard protocols of the European Animal Welfare Legislation, Directive 2010/63/EU (European Commission, 2016) and the Champalimaud Fish Platform Program.\u003c/p\u003e\n\u003cp\u003ePDOs preparation for microinjection\u003c/p\u003e\n\u003cp\u003eOn the day of injection, PDOs were thawed at 37\u0026ordm;C and washed with DMEM/F12 plus penicillin (100 U/mL), streptomycin (100 \u0026mu;g/mL) and Primocin (100 \u0026mu;g/mL). After centrifugation at 250 rcf for 5 min at 4\u0026ordm;C, the pellet was gently resuspended with 100 uL of injection mixture (described in \u003csup\u003e71\u003c/sup\u003e). Organoid aggregate size and viability were assessed using a hemocytometer and trypan blue. The suspension was labelled on ice for 5min with CellTracker Deep Red [C34565, Invitrogen] at a concentration of 1 \u0026micro;L/mL. The suspension was then centrifuged and resuspended in injection medium, in order to achieve a final concentration of ~ 2 \u0026times; 10⁴ cell equivalents per \u0026micro;L.\u003c/p\u003e\n\u003cp\u003eZebrafish embryos microinjection\u003c/p\u003e\n\u003cp\u003ePDOs were microinjected using a pneumatic injector [World Precision Instruments, Pneumatic PicoPump PV820], into the zebrafish perivitelline space (PVS) under a fluorescent stereoscope [Zeiss AxioZoom.V16]. After injection, xenografts were kept in an incubator at 34\u0026ordm;C.\u003c/p\u003e\n\u003cp\u003eZebrafish xenografts: Screening, treatment and fixation\u003c/p\u003e\n\u003cp\u003eAt 1-day post-injection (dpi), xenografts were screened under a fluorescence stereoscope for the presence of a tumoral mass. Successfully injected xenografts were randomly divided into control (E3 medium) and treatment group (5-FU+Irinotecan), which were renewed daily. 5-FU was administered at 4.2 mM and Irinotecan at 4 \u0026micro;M, diluted in E3 medium. At 4dpi, xenografts were sacrificed and fixed in 4% (v/v) formaldehyde [Thermo Scientific] overnight.\u003c/p\u003e\n\u003cp\u003eZebrafish xenografts: Whole-mount immunofluorescence\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePrimary antibodies: anti-Cleaved Caspase-3 (rabbit, [Cell signaling, code#9661]), anti-Human mitochondria (mouse, 1:100 [Merck Millipore, cat#MAB1273]). Secondary antibodies: anti-rabbit 594 DyLight [ThermoFisher Scientific, cat#35510] and anti-mouse 488 DyLight [ThermoFisher Scientific, cat#35502] were applied simultaneously with DAPI. Xenografts were mounted with Mowiol.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eZebrafish xenografts: Confocal microscopy and analysis\u003c/p\u003e\n\u003cp\u003eZebrafish xenografts were imaged using BC43 Andor Benchtop Confocal Microscope. Sequential z-stack images were acquired with a 5 \u0026micro;m interval. Images were analyzed using ImageJ software. The tumor size was calculated as the sum of the total area of human mitochondria positive cells per slice. The percentage of activated Caspase-3-positive cells was quantified manually by counting apoptotic bodies in every slice of the tumor.\u003c/p\u003e\n\u003cp\u003eZebrafish xenografts: Statistical Analysis\u003c/p\u003e\n\u003cp\u003eStatistical Analysis was performed using GraphPad Prism (v8.0.2). Results are represented as average (AVG) \u0026plusmn; standard error of the mean (SEM). Data were analyzed using the non-parametric Mann-Whitney test. Outliers were assessed using the \u0026ldquo;GraphPad Outlier\u0026rdquo; tool (https:// www.graphpad.com/quickcalcs/Grubbs1.cfm). \u0026nbsp; For all tests, \u003cem\u003ep\u003c/em\u003e-values (\u003cem\u003ep\u003c/em\u003e) are two-tailed with a 95% confidence interval. Differences were considered significant whenever p\u0026lt;0.05, and statistical output was represented by stars as follows: non-significant (ns)\u0026gt;0.05, (ns)\u0026gt;0.05, *\u0026le;0.05, **\u0026le;0,01, ***\u0026le;0.001, ****\u0026le;0.0001.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIn vivo mouse studies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor peritoneal implantation assays, equivalent amounts of disaggregated PDO cells were injected in the cecum of athymic nude mice (strain: Hsd:Athymic Nude-Foxn1nu; 5\u0026ndash;7-week-old males). Mice were monitored regularly for general health status and signs of tumor development. Two months after tumor implantation, all animals were sacrificed simultaneously across experimental groups to ensure comparable end-point analysis. At necropsy, the peritoneal cavity was carefully examined, and the number of visible intraperitoneal tumor implants was systematically recorded for each animal. Macroscopic peritoneal nodules were counted, and when required, tumor implants were collected for further histological confirmation.\u003c/p\u003e\n\u003cp\u003eAll procedures involving living animals were conducted under specific pathogen-free conditions and in accordance with the guidelines established by the Animal Care Committee of the Generalitat de Catalunya. The study protocols were reviewed and approved by the Committee for Animal Experimentation at the Institute of Biomedical Research of Bellvitge (Barcelona).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAnalysis of CRC cohorts\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTranscriptomic data from colorectal cancer cohorts were obtained from publicly available datasets. Microarray expression data from GSE39582\u0026nbsp;\u003csup\u003e18\u003c/sup\u003e and GSE14333\u0026nbsp;\u003csup\u003e19\u003c/sup\u003e were downloaded from the Gene Expression Omnibus (GEO) and analyzed using the affy R package. RNA-seq data and associated clinical information from colon (COAD) and rectal (READ) adenocarcinoma samples were retrieved from The Cancer Genome Atlas (TCGA) using the TCGAbiolinks R package (v2.24.1) with the STAR-Counts workflow. COAD and READ datasets were merged, normalized, and transformed using edgeR (v3.38.1).\u003c/p\u003e\n\u003cp\u003ePatients were classified as fetal or non-fetal based on the quantile expression of a fetal intestinal stem cell (ISC) gene signature comprising eight genes, yielding 142 fetal and 142 non-fetal cases in the Marisa cohort, 57 fetal and 57 non-fetal cases in the Jorissen cohort, and 83 fetal and 83 non-fetal cases in the TCGA cohort. Differential gene expression analyses between fetal and non-fetal tumors were performed using limma (v3.52.2)\u0026nbsp;\u003csup\u003e72\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eFunctional enrichment analyses were conducted using enrichR (v3.0) and clusterProfiler (v4.4.2)\u0026nbsp;\u003csup\u003e73\u003c/sup\u003e interrogating the ENCODE and ChEA Consensus Transcription Factors from ChIP-X and KEGG gene set collections, respectively. Optimal cutpoints for survival analyses were determined using the surv_cutpoint function from the survminer package (v0.4.9).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSurvival analysis of CRC cohorts\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSurvival analysis was performed using the Kaplan-Meier curve estimates, considering either the in-house CRC patients Metacohort, which was previously generated\u0026nbsp;\u003csup\u003e17\u003c/sup\u003e and is available at Zenodo (doi: 10.5281/zenodo.13303049), or the TCGA dataset. Patients were classified into two subgroups, \u0026lsquo;High\u0026rsquo; or \u0026lsquo;Low\u0026rsquo;, based on the expression of specific signature genes (SUZ12 and MYC targets; see Supplementary Data 2). This classification was performed using the optimized cutpoint assessed by the surv_cutpoint function from the survminer (v.0.4.9) R package. If a gene was targeted by more than one array probe in the Metacohort, the probe showing the highest expression was selected. Left-censored patients were excluded from further analysis. A standard log-rank test was performed to determine the statistical significance between the two subgroups. A p-value\u0026lt;0.05 was considered statistically significant. Hazard ratios are also shown for each comparison in which there was at least one event per group. All survival analyses were performed using the survival (v.3.3-8) R package.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBulk RNA-seq data analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBulk RNA sequencing was performed in PDOs and in the colorectal cancer cell line LS174T. In the different experiments, libraries were simultaneously prepared and sequenced using Illumina NovaSeq6000 platform (50 bp paired-end reads). Raw sequencing reads in FASTQ format were aligned to the human reference genome GRCh38.p13 (Gencode release 41) using STAR (v2.7.8)\u0026nbsp;\u003csup\u003e74\u003c/sup\u003e. Gene-level read counts were generated using the featureCounts function from the Subread package (v2.0.3 for PDO datasets and v2.8.2 for LS174T datasets). For PDO samples, genes with more than 10 reads in at least three samples were retained, whereas for LS174T samples, genes with more than 10 reads in at least two samples were kept, reflecting differences in sample number and experimental design. Raw library size differences were normalized using the trimmed mean of M values (TMM) method implemented in edgeR (v3.40.2 for PDOs and v3.36.0 for LS174T)\u0026nbsp;\u003csup\u003e75\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eNormalized counts were used for unsupervised analyses, including principal component analysis and clustering. For differential gene expression analyses, read counts were transformed to log2 counts per million (logCPM), and the mean\u0026ndash;variance relationship was modeled using precision weights with the voom approach implemented in limma (v3.54.2 for PDOs and v3.50.3 for LS174T)\u0026nbsp;\u003csup\u003e72\u003c/sup\u003e. All analyses were conducted using R (v4.2.1).\u003c/p\u003e\n\u003cp\u003eHeatmaps were generated from variance-stabilized expression values (vst) obtained from raw count matrices using DESeq2. Selected gene sets were extracted, converted from Ensembl IDs to gene symbols, and expression values were scaled by gene (row-wise z-score). Heatmaps were visualized using the ComplexHeatmap package (v2.16.0)\u0026nbsp;\u003csup\u003e76\u003c/sup\u003e with Pearson correlation\u0026ndash;based hierarchical clustering.\u003c/p\u003e\n\u003cp\u003ePre-ranked Gene Set Enrichment Analysis (GSEA) was performed using the fgsea package (v1.26.0). Genes were ranked based on the metric \u0026minus;log10(p value) \u0026times; sign (log2 fold change) derived from limma differential expression statistics. Enrichment analyses were conducted across multiple gene set collections, including Gene Ontology Biological Processes, KEGG canonical pathways, Hallmark gene sets from MSigDB (v7.5.1), and custom-curated gene signatures. GSEA was run using the fgseaMultilevel algorithm with a minimum gene set size of 10 and a maximum of 2,000 genes. Enrichment plots and normalized enrichment scores were generated from fgsea results.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSingle cell RNA-seq data analysis\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData pre-processing\u003c/p\u003e\n\u003cp\u003eSingle-cell RNA-seq data were generated from eight human PDOs, including fetal-type, non-fetal-type and chemotherapy-treated samples. Libraries were prepared using the GEM-X Universal 3\u0026rsquo; Gene Expression v4 4-plex kit (10x Genomics) and sequenced on a NovaSeq 6000 platform according to the manufacturer\u0026rsquo;s protocols. Samples were multiplexed in pairs within each sequencing batch using dual index combinations (OB1|OB2 or OB3|OB4) as follows: 15007AAG (PDO4 OB1|OB2; PDO10 OB3|OB4), 15008AAG (PDO27 OB1|OB2; PDO127B OB3|OB4), 15009AAG (PDO5 untreated OB1|OB2; PDO5 treated OB3|OB4), and 15010AAG (PDO66 untreated OB1|OB2; PDO66 treated OB3|OB4). Raw sequencing data were processed using Cell Ranger (v9.0.1, 10x Genomics) with the cellranger multi pipeline. Reads were aligned to the human reference genome GRCh38 (10x Genomics refdata-gex-GRCh38-2024-A). Filtered feature\u0026ndash;barcode matrices were generated separately for each PDO sample.\u003c/p\u003e\n\u003cp\u003eDoublet detection\u003c/p\u003e\n\u003cp\u003eDoublets were identified independently for each PDO using scDblFinder (v1.20.2)\u0026nbsp;\u003csup\u003e77\u003c/sup\u003e, combining three complementary approaches: (i) random artificial doublet generation, (ii) internal clustering\u0026ndash;based detection, and (iii) external clustering information derived from Cell Ranger graph-based Louvain clustering. Cells classified as doublets by at least two methods were considered high-confidence doublets and removed from downstream analyses.\u003c/p\u003e\n\u003cp\u003eQuality control and normalization\u003c/p\u003e\n\u003cp\u003eFiltered matrices were imported into R (v4.4.2) and processed using Seurat (v5.0.1). PDO datasets were merged into a single Seurat object without applying batch correction or integration. Cells with fewer than 750 or more than 2,500 detected genes, more than 15% mitochondrial gene content, or fewer than 10,000 total counts were excluded. Genes expressed in fewer than 10 cells and ribosomal genes were removed.\u003c/p\u003e\n\u003cp\u003eNormalization and dimensionality reduction\u003c/p\u003e\n\u003cp\u003eData were normalized using SCTransform, regressing out mitochondrial gene content. Principal component analysis (PCA) was performed on the SCT assay. UMAP embeddings were generated using the first 37 principal components, capturing more than 85% of the total variance. UMAP projections were used to visualize sample distribution, gene expression levels and gene signature scores.\u003c/p\u003e\n\u003cp\u003eGene signature scoring\u003c/p\u003e\n\u003cp\u003eFetal-like gene signature scores were computed using the JASMINE algorithm (script version V1_11October2021)\u0026nbsp;\u003csup\u003e78\u003c/sup\u003e on SCTransform-normalized expression values. Scores were calculated at single-cell resolution and visualized on UMAP embeddings to assess the distribution of fetal transcriptional programs across PDO samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eProteomic analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOrganoid Lysis for Proteomic Analysis\u003c/p\u003e\n\u003cp\u003eOrganoid pellets were resuspended in 50 \u0026micro;l lysis buffer (8 M Urea, 10 mM Tris-Base, 100 mM NaH\u003csub\u003e2\u003c/sub\u003ePO\u003csub\u003e4\u003c/sub\u003e, pH 8) with phosphatase inhibitor [Roche]. Organoids were sonicated (Ultrasonic Homogeniser SKL-150W, Syclon) at 20% power with 9 second pulse rates for 1 min. Organoid debris was pelleted by centrifugation at 14,000 rcf for 8 min at 4\u0026ordm;C and supernatant collected. Protein concentration was estimated using BCA assay [Pierce].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eProteomic Sample Digestion and Cleanup\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAll samples were normalized to a protein concentration of 500 \u0026micro;g. Protein lysates were reduced with 8 mM 1,4-dithiotreitol (DTT) at 30\u0026ordm;C for 30 min at 1200 rpm on a thermomixer [Eppendorf]. Followed by alkylation with 20 mM iodoacetamide (IAA) at 30\u0026ordm;C for 30 min at 1200 rpm in the dark. Sample urea concentration was diluted to 2 M with Tris-HCl (50 mM). 10 \u0026micro;g of trypsin/Lys-C [Promega] was added to each sample for a final 1:50 enzyme to protein ratio. Samples were digested overnight at 37\u0026ordm;C at 1000 rpm. Digestion was terminated by adding formic acid to 1% final concentration.\u003c/p\u003e\n\u003cp\u003eSample clean-up was carried out using Sep-Pak C18 columns [Waters]. Columns were activated with 100% ethanol and equilibrated with 0.1% trifluoroacetic acid (TFA). Samples were passed through the columns followed by two washes with 0.1% TFA. Peptides were eluted from the column with elution buffer (80% acetonitrile, 0.1% TFA). 10% of each sample elute was used for whole proteome mass spectrometry analysis.\u003c/p\u003e\n\u003cp\u003eMass Spectrometry Acquisition of Proteomic Samples\u003c/p\u003e\n\u003cp\u003eA pooled sample of 1 \u0026micro;g was analyzed using a Bruker timsTof Pro mass spectrometer connected to an Evosep One Liquid chromatography system. Tryptic peptides were resuspended in 0.1% formic acid and loaded on to an Evosep tip. The Evosep was configured to pick up each tip, elute and separate the peptides using a set chromatography method (30 samples a day) \u003csup\u003e79\u003c/sup\u003e. The mass spectrometer was operated in positive ion mode with a capillary voltage of 1300-1600 V, dry gas flow of 3 l/min and a dry temperature of 180\u0026ordm;C. All data was acquired with the instrument operating in a data dependent analysis parallel accumulation serial fragmentation mode (dda-PASEF). Trapped ions were selected for ms/ms using parallel accumulation serial fragmentation (PASEF). A scan range of (100-1700 m/z) was performed at a rate of 4 PASEF MS/MS frames to 1 MS scan with a cycle time of 0.53 s \u003csup\u003e80\u003c/sup\u003e. The resultant file was used to create the dia-PASEF method within Bruker timsControl software. The scan mode \u0026ldquo;dia-PASEF\u0026rdquo; was selected and the pooled sample dda-PASEF file was opened in the window editor in the MS/MS tab. Once opened, the adjustable parallelogram was used to select the area of the heat map where the identifiable peptides (central region in the heat map containing peptides with charge states from +2 to +5) could be found. Dia-PASEF settings used were: mass width 26.0 Da, mass overlap 1.0, mass steps per cycle 34 or 35, mobility overlap 0.00, mass range 338.0-1214 m/z or 334.7-1185.7 m/z. All samples were acquired using data independent analysis parallel accumulation serial fragmentation (dia-PASEF) \u003csup\u003e81\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eSpectral Library Generation and DIA Data Processing\u003c/p\u003e\n\u003cp\u003eData acquired using dia-PASEF was analyzed using DIA-NN 2.1.0 and DIA-NN 2.3.0 Academia (Data-Independent Acquisition by Neural Networks \u003csup\u003e82\u003c/sup\u003e. The \u003cem\u003eHomo sapiens\u003c/em\u003e subset of UniProtKB (Swiss-Prot and TrEMBL database) \u003csup\u003e83\u003c/sup\u003e was used to generate a spectral library within DIA-NN (library free mode). Specific search settings included cysteine carbamidomethylation as a fixed modification, protein N-terminal acetylation and methionine oxidation as variable modifications, maximum missed cleavages 1, min precursor +1, max precursor +4, Neural network (cross validated) was used, cross run normalization was set to retention time dependent and library generation was set to IDs, retention time (RT), and ion mobility (IM) profiling. Precursor FDR was set to 1%.\u003c/p\u003e\n\u003cp\u003eProteomic Data Analysis and Visualization\u003c/p\u003e\n\u003cp\u003eRaw proteomic data from DIA-NN was uploaded into R (v2025.09.2+418). Proteins lacking annotated names were removed from the dataset. Proteins represented by multiple entries were aggregated by calculating the mean abundance. The data was log2 transformed. Samples were grouped by PDO origin (PDO4, PDO5, PDO10, PDO27, PDO66, PDO127). Proteins were retained if at least 70% of values were present in at least one PDO group. Missing values were imputed using the \u003cem\u003eimputeLCMD\u003c/em\u003e (v2.1.) package in R, by sampling from the left-shifted normal distribution (width = 0.3, downshift = 1.8).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor visualization, pre-defined proteins of interest were selected. To facilitate comparison of expression patterns across samples, protein expression values were Z-score normalized on a per-protein (row-wise) basis. Heatmaps were generated using the \u003cem\u003eComplexHeatmap\u003c/em\u003e package in R \u003csup\u003e76\u003c/sup\u003e. Hierarchical clustering was applied to proteins, while samples were displayed in a predefined order without supervised clustering. Sample annotations indicating sample condition (Fetal or Non-fetal) were included above the heatmap.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eQuantification and Statistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEach experiment shown in the manuscript has been performed at least twice. The statistical parameters reported in the figures and figure legends include the number of events quantified, standard deviation, statistical significance and the test performed. GraphPad Prism 9 software was used for the statistical analysis, with \u003cem\u003ep\u003c/em\u003e-values of \u0026lt;0.05 being considered significant (****\u003cem\u003ep\u003c/em\u003e-value\u0026lt;0.0001, ***\u003cem\u003ep\u003c/em\u003e- value\u0026lt;0.001, **\u003cem\u003ep\u003c/em\u003e-value\u0026lt;0.01 and n.s. \u003cem\u003ep\u003c/em\u003e-value \u0026gt; 0.05).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eDATA AVAILABILITY\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePublic datasets used in this study are referenced in the text. Single-cell and bulk RNA-seq datasets generated in this study have been deposited in NCBI Gene Expression Omnibus (GEO) repository under GEO SuperSeries accession number GSE243803, composed in three SubSeries GSE243802 (bulk RNA-seq of LS174T), GSE277036 (bulk RNA-seq of PDOs) and GSE324317 (10x scRNAseq data). Sequencing data can be accessed through this token: yxupseuylvqpjcp. Proteomics data is available at PRIDE PRoteomics IDEntifications (PRIDE) with identifier PXD075164.\u003c/p\u003e\n\u003cp\u003eSource data is provided with this paper. Further information on research design is available in the Nature Research Reporting Summary linked to this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCODE AVAILABILITY\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eScripts used in the bulk RNA-seq preprocessing steps are available at GitHub repository: https://github.com/BigaSpinosaLab/LAB_RNAseq_Data_Analysis. The final CRC metacohort is stored, as an RData object, at Zenodo (doi: 10.5281/zenodo.13303049). Scripts that have been used to process the scRNA-seq and bulk RNA-seq datasets are deposited in Github repository: https://github.com/BigaSpinosaLab/PAPER_PRC2_MYC_repression_oncofetal_CRC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eACKNOWLEDGEMENTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank the members of the Espinosa and Bigas laboratories for their constructive discussions and valuable suggestions throughout this work. This research was supported by CIBER – Consorcio Centro de Investigación Biomédica en Red (CB16/12/00244), Instituto de Salud Carlos III (ISCIII), Ministerio de Ciencia e Innovación and Unión Europea – European Regional Development Fund (FEDER). This work was funded by the FIS project PI25/00006 and the EP PerMed project AC24/00006 (ColoStem-Applied, EpPermed2024-149), both funded by ISCIII and co-funded by the European Union (FEDER); and the project PRYGN246819ESPI from the Fundación Científica de la Asociación Española Contra el Cáncer (AECC). We thank MarBiobank for providing and characterizing patient samples (PT20/00023, from Instituto de Salud Carlos III, FEDER) and the Comprehensive Molecular Analytical Platform (CMAP) under The SFI Research Infrastructure Programme, (18/RI/5702). A.M.\u0026nbsp;is a recipient of a grant from ISCIII, grant number FI23/00002, co-funded by the European Social Fund Plus (ESF+). A.Y. acknowledges financial support from the China Scholarship Council (CSC), grant number 202408430067.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAUTHORS CONTRIBUTIONS STATEMENT\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLS, AMo, AM-R, AM-L, AY, JK, J-JC-R and JB prepared the reagents and performed the experiments with cells; LS, TL-J, EC and MM performed the bioinformatics analysis; AN and DM performed the proteomic analysis; BC and RF performed the in vivo\u0026nbsp;experiments with zebrafish; AV, DA-V and MM-I performed the in vivo experiments with mice; ABa and AM generated PDOs included in the work and revised the manuscript; LS, LE and AB conceptualized and conducted the study, analyzed data and prepared the manuscript. All authors have read and approved the final manuscript and consent its publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCOMPETING INTERESTS STATEMENT\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests related to this work.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAlvarez-Varela, A. \u003cem\u003eet al.\u003c/em\u003e Mex3a marks drug-tolerant persister colorectal cancer cells that mediate relapse after chemotherapy. \u003cem\u003eNat. Cancer\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 1052\u0026ndash;1070 (2022).\u003c/li\u003e\n\u003cli\u003eDhimolea, E. \u003cem\u003eet al.\u003c/em\u003e An Embryonic Diapause-like Adaptation with Suppressed Myc Activity Enables Tumor Treatment Persistence. \u003cem\u003eCancer Cell\u003c/em\u003e \u003cstrong\u003e39\u003c/strong\u003e, 240-256.e11 (2021).\u003c/li\u003e\n\u003cli\u003eMoorman, A. \u003cem\u003eet al.\u003c/em\u003e Progressive plasticity during colorectal cancer metastasis. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e637\u003c/strong\u003e, 947\u0026ndash;954 (2025).\u003c/li\u003e\n\u003cli\u003eMzoughi, S. \u003cem\u003eet al.\u003c/em\u003e Oncofetal reprogramming drives phenotypic plasticity in WNT-dependent colorectal cancer. \u003cem\u003eNat. Genet.\u003c/em\u003e \u003cstrong\u003e57\u003c/strong\u003e, 402\u0026ndash;412 (2025).\u003c/li\u003e\n\u003cli\u003eRehman, S. K. \u003cem\u003eet al.\u003c/em\u003e Colorectal Cancer Cells Enter a Diapause-like DTP State to Survive Chemotherapy. \u003cem\u003eCell\u003c/em\u003e https://doi.org/10.1016/j.cell.2020.11.018 (2021) doi:10.1016/j.cell.2020.11.018.\u003c/li\u003e\n\u003cli\u003eSol\u0026eacute;, L. \u003cem\u003eet al.\u003c/em\u003e p53 wild-type colorectal cancer cells that express a fetal gene signature are associated with metastasis and poor prognosis. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 2866 (2022).\u003c/li\u003e\n\u003cli\u003eShao, D. D. \u003cem\u003eet al.\u003c/em\u003e KRAS and YAP1 converge to regulate EMT and tumor survival. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e158\u003c/strong\u003e, 171\u0026ndash;84 (2014).\u003c/li\u003e\n\u003cli\u003eAggarwal, V., Montoya, C. A., Donnenberg, V. S. \u0026amp; Sant, S. Interplay between tumor microenvironment and partial EMT as the driver of tumor progression. \u003cem\u003eiScience\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 102113 (2021).\u003c/li\u003e\n\u003cli\u003eGodin, L. \u003cem\u003eet al.\u003c/em\u003e A Novel Approach for Quantifying Cancer Cells Showing Hybrid Epithelial/Mesenchymal States in Large Series of Tissue Samples: Towards a New Prognostic Marker. \u003cem\u003eCancers (Basel).\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 906 (2020).\u003c/li\u003e\n\u003cli\u003eKr\u0026ouml;ger, C. \u003cem\u003eet al.\u003c/em\u003e Acquisition of a hybrid E/M state is essential for tumorigenicity of basal breast cancer cells. \u003cem\u003eProc. Natl. Acad. Sci. U. S. A.\u003c/em\u003e \u003cstrong\u003e116\u003c/strong\u003e, 7353\u0026ndash;7362 (2019).\u003c/li\u003e\n\u003cli\u003eL\u0026uuml;\u0026ouml;nd, F. \u003cem\u003eet al.\u003c/em\u003e Distinct contributions of partial and full EMT to breast cancer malignancy. \u003cem\u003eDev. Cell\u003c/em\u003e \u003cstrong\u003e56\u003c/strong\u003e, 3203-3221.e11 (2021).\u003c/li\u003e\n\u003cli\u003ePastushenko, I. \u003cem\u003eet al.\u003c/em\u003e Fat1 deletion promotes hybrid EMT state, tumour stemness and metastasis. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e589\u003c/strong\u003e, 448\u0026ndash;455 (2021).\u003c/li\u003e\n\u003cli\u003eYoussef, K. K. \u003cem\u003eet al.\u003c/em\u003e Two distinct epithelial-to-mesenchymal transition programs control invasion and inflammation in segregated tumor cell populations. \u003cem\u003eNat. Cancer\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1660\u0026ndash;1680 (2024).\u003c/li\u003e\n\u003cli\u003eCentonze, A. \u003cem\u003eet al.\u003c/em\u003e A Plastic EMP1+ to LGR5+ Cell State Conversion as a Bypass to KRASG12D Pharmacologic Inhibition in Metastatic Colorectal Cancer. \u003cem\u003eCancer Discov.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 320\u0026ndash;344 (2026).\u003c/li\u003e\n\u003cli\u003eGhate, N. B. \u003cem\u003eet al.\u003c/em\u003e Phosphorylation and stabilization of EZH2 by DCAF1/VprBP trigger aberrant gene silencing in colon cancer. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 2140 (2023).\u003c/li\u003e\n\u003cli\u003eLoi, P. \u003cem\u003eet al.\u003c/em\u003e Epigenetic and Oncogenic Inhibitors Cooperatively Drive Differentiation and Kill KRAS \u003cem\u003e-\u003c/em\u003e Mutant Colorectal Cancers. \u003cem\u003eCancer Discov.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 2430\u0026ndash;2449 (2024).\u003c/li\u003e\n\u003cli\u003eSol\u0026eacute;, L. \u003cem\u003eet al.\u003c/em\u003e ColoStem, a core oncofetal signature that identifies poor prognosis colorectal tumors. \u003cem\u003eBr. J. Cancer\u003c/em\u003e https://doi.org/10.1038/s41416-025-03280-2 (2025) doi:10.1038/s41416-025-03280-2.\u003c/li\u003e\n\u003cli\u003eMarisa, L. \u003cem\u003eet al.\u003c/em\u003e Gene Expression Classification of Colon Cancer into Molecular Subtypes: Characterization, Validation, and Prognostic Value. \u003cem\u003ePLoS Med.\u003c/em\u003e https://doi.org/10.1371/journal.pmed.1001453 (2013) doi:10.1371/journal.pmed.1001453.\u003c/li\u003e\n\u003cli\u003eJorissen, R. N. \u003cem\u003eet al.\u003c/em\u003e Metastasis-associated gene expression changes predict poor outcomes in patients with Dukes stage B and C colorectal cancer. \u003cem\u003eClinical Cancer Research\u003c/em\u003e https://doi.org/10.1158/1078-0432.CCR-09-1431 (2009) doi:10.1158/1078-0432.CCR-09-1431.\u003c/li\u003e\n\u003cli\u003eMalla, S. B. \u003cem\u003eet al.\u003c/em\u003e Pathway level subtyping identifies a slow-cycling biological phenotype associated with poor clinical outcomes in colorectal cancer. \u003cem\u003eNat. Genet.\u003c/em\u003e https://doi.org/10.1038/s41588-024-01654-5 (2024) doi:10.1038/s41588-024-01654-5.\u003c/li\u003e\n\u003cli\u003eOgasawara, N. \u003cem\u003eet al.\u003c/em\u003e Discovery of non-genomic drivers of YAP signaling modulating the cell plasticity in CRC tumor lines. \u003cem\u003eiScience\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 109247 (2024).\u003c/li\u003e\n\u003cli\u003eXiong, L. \u003cem\u003eet al.\u003c/em\u003e A patient-derived organoid model captures fetal-like plasticity in colorectal cancer. \u003cem\u003eCell Res.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 642\u0026ndash;655 (2025).\u003c/li\u003e\n\u003cli\u003eMcMahon, S. B. MYC and the Control of Apoptosis. \u003cem\u003eCold Spring Harb. Perspect. Med.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, a014407\u0026ndash;a014407 (2014).\u003c/li\u003e\n\u003cli\u003eEdwards-Hicks, J. \u003cem\u003eet al.\u003c/em\u003e MYC sensitises cells to apoptosis by driving energetic demand. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 4674 (2022).\u003c/li\u003e\n\u003cli\u003eMarzochi, L. L. \u003cem\u003eet al.\u003c/em\u003e Use of histone methyltransferase inhibitors in cancer treatment: A systematic review. \u003cem\u003eEur. J. Pharmacol.\u003c/em\u003e \u003cstrong\u003e944\u003c/strong\u003e, 175590 (2023).\u003c/li\u003e\n\u003cli\u003eSun, S., Yu, F., Xu, D., Zheng, H. \u0026amp; Li, M. EZH2, a prominent orchestrator of genetic and epigenetic regulation of solid tumor microenvironment and immunotherapy. \u003cem\u003eBiochim. Biophys. Acta Rev. Cancer\u003c/em\u003e \u003cstrong\u003e1877\u003c/strong\u003e, 188700 (2022).\u003c/li\u003e\n\u003cli\u003eWeber, L. I. \u0026amp; Hartl, M. Strategies to target the cancer driver MYC in tumor cells. \u003cem\u003eFront. Oncol.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 1142111 (2023).\u003c/li\u003e\n\u003cli\u003eGlancy, E. \u003cem\u003eet al.\u003c/em\u003e PRC2.1- and PRC2.2-specific accessory proteins drive recruitment of different forms of canonical PRC1. \u003cem\u003eMol. Cell\u003c/em\u003e \u003cstrong\u003e83\u003c/strong\u003e, 1393-1411.e7 (2023).\u003c/li\u003e\n\u003cli\u003eHealy, E. \u003cem\u003eet al.\u003c/em\u003e PRC2.1 and PRC2.2 Synergize to Coordinate H3K27 Trimethylation. \u003cem\u003eMol. Cell\u003c/em\u003e \u003cstrong\u003e76\u003c/strong\u003e, 437-452.e6 (2019).\u003c/li\u003e\n\u003cli\u003eGao, S. \u003cem\u003eet al.\u003c/em\u003e Tracing the temporal-spatial transcriptome landscapes of the human fetal digestive tract using single-cell RNA-sequencing. \u003cem\u003eNat. Cell Biol.\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 721\u0026ndash;734 (2018).\u003c/li\u003e\n\u003cli\u003eMustata, R. C. \u003cem\u003eet al.\u003c/em\u003e Identification of Lgr5-Independent Spheroid-Generating Progenitors of the Mouse Fetal Intestinal Epithelium. \u003cem\u003eCell Rep.\u003c/em\u003e https://doi.org/10.1016/j.celrep.2013.09.005 (2013) doi:10.1016/j.celrep.2013.09.005.\u003c/li\u003e\n\u003cli\u003eZhu, L. \u003cem\u003eet al.\u003c/em\u003e Prominin 1 marks intestinal stem cells that are susceptible to neoplastic transformation. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e457\u003c/strong\u003e, 603\u0026ndash;7 (2009).\u003c/li\u003e\n\u003cli\u003eGuiu, J. \u003cem\u003eet al.\u003c/em\u003e Identification of Cdca7 as a novel Notch transcriptional target involved in hematopoietic stem cell emergence. \u003cem\u003eJournal of Experimental Medicine\u003c/em\u003e \u003cstrong\u003e211\u003c/strong\u003e, 2411\u0026ndash;2423 (2014).\u003c/li\u003e\n\u003cli\u003eZhao, M. \u003cem\u003eet al.\u003c/em\u003e Identification of cancer stem cell-related biomarkers in lung adenocarcinoma by stemness index and weighted correlation network analysis. \u003cem\u003eJ. Cancer Res. Clin. Oncol.\u003c/em\u003e \u003cstrong\u003e146\u003c/strong\u003e, 1463\u0026ndash;1472 (2020).\u003c/li\u003e\n\u003cli\u003eBarker, N. \u003cem\u003eet al.\u003c/em\u003e Crypt stem cells as the cells-of-origin of intestinal cancer. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e457\u003c/strong\u003e, 608\u0026ndash;611 (2009).\u003c/li\u003e\n\u003cli\u003eMerlos-Su\u0026aacute;rez, A. \u003cem\u003eet al.\u003c/em\u003e The intestinal stem cell signature identifies colorectal cancer stem cells and predicts disease relapse. \u003cem\u003eCell Stem Cell\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 511\u0026ndash;524 (2011).\u003c/li\u003e\n\u003cli\u003eCosta, B. \u003cem\u003eet al.\u003c/em\u003e Zebrafish Avatar-test forecasts clinical response to chemotherapy in patients with colorectal cancer. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 4771 (2024).\u003c/li\u003e\n\u003cli\u003eFazio, M., Ablain, J., Chuan, Y., Langenau, D. M. \u0026amp; Zon, L. I. Zebrafish patient avatars in cancer biology and precision cancer therapy. \u003cem\u003eNat. Rev. Cancer\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 263\u0026ndash;273 (2020).\u003c/li\u003e\n\u003cli\u003eEspinosa, L., L\u0026oacute;pez-Arribillaga, E., Bachs, O. \u0026amp; Bigas, A. Notch Signaling in the Normal Intestine and Intestinal Cancer. in \u003cem\u003eTargeting Notch in Cancer\u003c/em\u003e 333\u0026ndash;352 (Springer New York, New York, NY, 2018). doi:10.1007/978-1-4939-8859-4_13.\u003c/li\u003e\n\u003cli\u003eDow, L. E. \u003cem\u003eet al.\u003c/em\u003e Apc Restoration Promotes Cellular Differentiation and Reestablishes Crypt Homeostasis in Colorectal Cancer. \u003cem\u003eCell\u003c/em\u003e https://doi.org/10.1016/j.cell.2015.05.033 (2015) doi:10.1016/j.cell.2015.05.033.\u003c/li\u003e\n\u003cli\u003eFre, S. \u003cem\u003eet al.\u003c/em\u003e Notch and Wnt signals cooperatively control cell proliferation and tumorigenesis in the intestine. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e \u003cstrong\u003e106\u003c/strong\u003e, 6309\u0026ndash;6314 (2009).\u003c/li\u003e\n\u003cli\u003eFasano, M. \u003cem\u003eet al.\u003c/em\u003e TGF-\u0026beta; Modulated Pathways in Colorectal Cancer: New Potential Therapeutic Opportunities. \u003cem\u003eInt. J. Mol. Sci.\u003c/em\u003e \u003cstrong\u003e25\u003c/strong\u003e, (2024).\u003c/li\u003e\n\u003cli\u003eZanconato, F., Cordenonsi, M. \u0026amp; Piccolo, S. YAP/TAZ at the Roots of Cancer. \u003cem\u003eCancer Cell\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 783\u0026ndash;803 (2016).\u003c/li\u003e\n\u003cli\u003eYui, S. \u003cem\u003eet al.\u003c/em\u003e YAP/TAZ-Dependent Reprogramming of Colonic Epithelium Links ECM Remodeling to Tissue Regeneration. \u003cem\u003eCell Stem Cell\u003c/em\u003e https://doi.org/10.1016/j.stem.2017.11.001 (2018) doi:10.1016/j.stem.2017.11.001.\u003c/li\u003e\n\u003cli\u003eMuncan, V. \u003cem\u003eet al.\u003c/em\u003e Rapid loss of intestinal crypts upon conditional deletion of the Wnt/Tcf-4 target gene c-Myc. \u003cem\u003eMol. Cell. Biol.\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 8418\u0026ndash;26 (2006).\u003c/li\u003e\n\u003cli\u003eSansom, O. J. \u003cem\u003eet al.\u003c/em\u003e Myc deletion rescues Apc deficiency in the small intestine. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e446\u003c/strong\u003e, 676\u0026ndash;9 (2007).\u003c/li\u003e\n\u003cli\u003eSur, I. K. \u003cem\u003eet al.\u003c/em\u003e Mice lacking a Myc enhancer that includes human SNP rs6983267 are resistant to intestinal tumors. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e338\u003c/strong\u003e, 1360\u0026ndash;3 (2012).\u003c/li\u003e\n\u003cli\u003eAyer, D. E., Kretzner, L. \u0026amp; Eisenman, R. N. Mad: a heterodimeric partner for Max that antagonizes Myc transcriptional activity. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e72\u003c/strong\u003e, 211\u0026ndash;22 (1993).\u003c/li\u003e\n\u003cli\u003eKroemer, G., Mari\u0026ntilde;o, G. \u0026amp; Levine, B. Autophagy and the Integrated Stress Response. \u003cem\u003eMol. Cell\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 280\u0026ndash;293 (2010).\u003c/li\u003e\n\u003cli\u003eAlhasan, B. A. \u003cem\u003eet al.\u003c/em\u003e The mTOR Pathway in Pluripotent Stem Cells: Lessons for Understanding Cancer Cell Dormancy. \u003cem\u003eMembranes (Basel).\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, (2021).\u003c/li\u003e\n\u003cli\u003eLee, J.-E. \u003cem\u003eet al.\u003c/em\u003e Autophagy Regulates Embryonic Survival During Delayed Implantation. \u003cem\u003eEndocrinology\u003c/em\u003e \u003cstrong\u003e152\u003c/strong\u003e, 2067\u0026ndash;2075 (2011).\u003c/li\u003e\n\u003cli\u003eLin, C. \u003cem\u003eet al.\u003c/em\u003e The transcription factor p8 regulates autophagy during diapause embryo formation in Artemia parthenogenetica. \u003cem\u003eCell Stress Chaperones\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 665\u0026ndash;675 (2016).\u003c/li\u003e\n\u003cli\u003eCollignon, E. Unveiling the role of cellular dormancy in cancer progression and recurrence. \u003cem\u003eCurr. Opin. Oncol.\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 74\u0026ndash;81 (2024).\u003c/li\u003e\n\u003cli\u003eNiu, X. \u003cem\u003eet al.\u003c/em\u003e Autophagy in cancer development, immune evasion, and drug resistance. \u003cem\u003eDrug Resist. Updat.\u003c/em\u003e \u003cstrong\u003e78\u003c/strong\u003e, 101170 (2025).\u003c/li\u003e\n\u003cli\u003eDong, W. \u003cem\u003eet al.\u003c/em\u003e MYC drives autophagy to adapt to stress in Penaeus vannamei. \u003cem\u003eFish Shellfish Immunol.\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e, 187\u0026ndash;196 (2022).\u003c/li\u003e\n\u003cli\u003eMedda, A. \u003cem\u003eet al.\u003c/em\u003e c-MYC-dependent transcriptional inhibition of autophagy is implicated in cisplatin sensitivity in HPV-positive head and neck cancer. \u003cem\u003eCell Death Dis.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 719 (2023).\u003c/li\u003e\n\u003cli\u003ePanwar, V. \u003cem\u003eet al.\u003c/em\u003e Multifaceted role of mTOR (mammalian target of rapamycin) signaling pathway in human health and disease. \u003cem\u003eSignal Transduct. Target. Ther.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 375 (2023).\u003c/li\u003e\n\u003cli\u003eToh, P. P. C. \u003cem\u003eet al.\u003c/em\u003e Myc inhibition impairs autophagosome formation. \u003cem\u003eHum. Mol. Genet.\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 5237\u0026ndash;5248 (2013).\u003c/li\u003e\n\u003cli\u003eTsuneoka, M. \u003cem\u003eet al.\u003c/em\u003e c-myc induces autophagy in rat 3Y1 fibroblast cells. \u003cem\u003eCell Struct. Funct.\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 195\u0026ndash;204 (2003).\u003c/li\u003e\n\u003cli\u003eTanida, I., Ueno, T. \u0026amp; Kominami, E. LC3 and Autophagy. \u003cem\u003eMethods Mol. Biol.\u003c/em\u003e \u003cstrong\u003e445\u003c/strong\u003e, 77\u0026ndash;88 (2008).\u003c/li\u003e\n\u003cli\u003eAyyaz, A. \u003cem\u003eet al.\u003c/em\u003e Single-cell transcriptomes of the regenerating intestine reveal a revival stem cell. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e569\u003c/strong\u003e, 121\u0026ndash;125 (2019).\u003c/li\u003e\n\u003cli\u003eTonkin-Reeves, A., Giuliani, C. M. \u0026amp; Price, J. T. Inhibition of autophagy; an opportunity for the treatment of cancer resistance. \u003cem\u003eFront. Cell Dev. Biol.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eB\u0026eacute;guelin, W. \u003cem\u003eet al.\u003c/em\u003e Mutant EZH2 Induces a Pre-malignant Lymphoma Niche by Reprogramming the Immune Response. \u003cem\u003eCancer Cell\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 655-673.e11 (2020).\u003c/li\u003e\n\u003cli\u003eB\u0026eacute;guelin, W. \u003cem\u003eet al.\u003c/em\u003e EZH2 Is Required for Germinal Center Formation and Somatic EZH2 Mutations Promote Lymphoid Transformation. \u003cem\u003eCancer Cell\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 677\u0026ndash;692 (2013).\u003c/li\u003e\n\u003cli\u003eNtziachristos, P. \u003cem\u003eet al.\u003c/em\u003e Genetic inactivation of the polycomb repressive complex 2 in T cell acute lymphoblastic leukemia. \u003cem\u003eNat. Med.\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 298\u0026ndash;302 (2012).\u003c/li\u003e\n\u003cli\u003eLewis, P. W. \u003cem\u003eet al.\u003c/em\u003e Inhibition of PRC2 Activity by a Gain-of-Function H3 Mutation Found in Pediatric Glioblastoma. \u003cem\u003eScience (1979).\u003c/em\u003e \u003cstrong\u003e340\u003c/strong\u003e, 857\u0026ndash;861 (2013).\u003c/li\u003e\n\u003cli\u003eStafford, J. M. \u003cem\u003eet al.\u003c/em\u003e Multiple modes of PRC2 inhibition elicit global chromatin alterations in H3K27M pediatric glioma. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, (2018).\u003c/li\u003e\n\u003cli\u003eLee, W. \u003cem\u003eet al.\u003c/em\u003e PRC2 is recurrently inactivated through EED or SUZ12 loss in malignant peripheral nerve sheath tumors. \u003cem\u003eNat. Genet.\u003c/em\u003e \u003cstrong\u003e46\u003c/strong\u003e, 1227\u0026ndash;1232 (2014).\u003c/li\u003e\n\u003cli\u003eBredel, M. \u003cem\u003eet al.\u003c/em\u003e NFKBIA Deletion in Glioblastomas . \u003cem\u003eNew England Journal of Medicine\u003c/em\u003e \u003cstrong\u003e364\u003c/strong\u003e, 627\u0026ndash;637 (2011).\u003c/li\u003e\n\u003cli\u003eBredel, M. \u003cem\u003eet al.\u003c/em\u003e Haploinsufficiency of NFKBIA reshapes the epigenome antipodal to the IDH mutation and imparts disease fate in diffuse gliomas. \u003cem\u003eCell Rep. Med.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eCosta, B., Estrada, M. F., Barroso, M. T. \u0026amp; Fior, R. Zebrafish Patient‐Derived Avatars from Digestive Cancers for Anti‐cancer Therapy Screening. \u003cem\u003eCurr. Protoc.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eRitchie, M. E. \u003cem\u003eet al.\u003c/em\u003e Limma powers differential expression analyses for RNA-sequencing and microarray studies. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e https://doi.org/10.1093/nar/gkv007 (2015) doi:10.1093/nar/gkv007.\u003c/li\u003e\n\u003cli\u003eYu, G., Wang, L. G., Han, Y. \u0026amp; He, Q. Y. ClusterProfiler: An R package for comparing biological themes among gene clusters. \u003cem\u003eOMICS\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 284\u0026ndash;287 (2012).\u003c/li\u003e\n\u003cli\u003eDobin, A. \u003cem\u003eet al.\u003c/em\u003e STAR: ultrafast universal RNA-seq aligner. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 15\u0026ndash;21 (2013).\u003c/li\u003e\n\u003cli\u003eRobinson, M. D., McCarthy, D. J. \u0026amp; Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 139\u0026ndash;140 (2009).\u003c/li\u003e\n\u003cli\u003eGu, Z., Eils, R. \u0026amp; Schlesner, M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 2847\u0026ndash;2849 (2016).\u003c/li\u003e\n\u003cli\u003eGermain, P.-L., Lun, A., Garcia Meixide, C., Macnair, W. \u0026amp; Robinson, M. D. Doublet identification in single-cell sequencing data using scDblFinder. \u003cem\u003eF1000Res.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 979 (2021).\u003c/li\u003e\n\u003cli\u003eNoureen, N., Ye, Z., Chen, Y., Wang, X. \u0026amp; Zheng, S. Signature-scoring methods developed for bulk samples are not adequate for cancer single-cell RNA sequencing data. \u003cem\u003eElife\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, (2022).\u003c/li\u003e\n\u003cli\u003eBache, N. \u003cem\u003eet al.\u003c/em\u003e A Novel LC System Embeds Analytes in Pre-formed Gradients for Rapid, Ultra-robust Proteomics. \u003cem\u003eMol. Cell. Proteomics\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 2284\u0026ndash;2296 (2018).\u003c/li\u003e\n\u003cli\u003eMeier, F. \u003cem\u003eet al.\u003c/em\u003e Online Parallel Accumulation\u0026ndash;Serial Fragmentation (PASEF) with a Novel Trapped Ion Mobility Mass Spectrometer. \u003cem\u003eMolecular \u0026amp; Cellular Proteomics\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 2534\u0026ndash;2545 (2018).\u003c/li\u003e\n\u003cli\u003eMeier, F. \u003cem\u003eet al.\u003c/em\u003e diaPASEF: parallel accumulation\u0026ndash;serial fragmentation combined with data-independent acquisition. \u003cem\u003eNat. Methods\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 1229\u0026ndash;1236 (2020).\u003c/li\u003e\n\u003cli\u003eDemichev, V., Messner, C. B., Vernardis, S. I., Lilley, K. S. \u0026amp; Ralser, M. DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. \u003cem\u003eNat. Methods\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 41\u0026ndash;44 (2020).\u003c/li\u003e\n\u003cli\u003eBateman, A. \u003cem\u003eet al.\u003c/em\u003e UniProt: the Universal Protein Knowledgebase in 2023. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cstrong\u003e51\u003c/strong\u003e, D523\u0026ndash;D531 (2023).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9151067/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9151067/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The emergence of an oncofetal phenotype in colorectal cancer (CRC) has been linked to therapy resistance and metastatic progression. However, whether this state pre-exists in treatment-naïve tumors or is primarily acquired upon chemotherapy, and the molecular mechanisms underlying these transitions, remain incompletely understood. While transcriptional regulators such as YAP1 and AP1 have been proposed as oncofetal inducers, the epigenetic and metabolic determinants of this process are not fully defined.\r\nHere, we integrate transcriptomic analyses from multiple CRC patient cohorts with functional studies in patient-derived organoids (PDOs) to show that both intrinsic and therapy-induced fetal-like CRC states are characterized by coordinated repression of Polycomb Repressive Complex 2 (PRC2) and MYC signaling. This repression involves coordinated downregulation of core and accessory PRC2 components together with attenuation of MYC/MAX activity across patient datasets and experimental models. Functional inhibition of PRC2 is sufficient to unlock fetal and epithelial-to-mesenchymal transition (EMT) transcriptional programs, whereas MYC suppression, mediated in part by MXD1 induction, promotes growth arrest and metabolic adaptation without recapitulating the full fetal signature.\r\nMechanistically, fetal-like CRC cells, irrespective of their origin, display suppression of MYC- and mTOR-driven programs and increased reliance on autophagy. While basal autophagy levels show limited predictive value, induction of fetal-like states by chemotherapy or PRC2 inhibition consistently sensitizes CRC cells to lysosomal blockade. Together, these findings identify PRC2 and MYC repression as convergent regulatory features of oncofetal reprogramming in CRC and reveal autophagy dependence as a context-specific and therapeutically exploitable vulnerability of this aggressive tumor state.","manuscriptTitle":"PRC2 and MYC repression drives oncofetal reprogramming and autophagy dependence in Colorectal Cancer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-20 07:16:51","doi":"10.21203/rs.3.rs-9151067/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"64a8e2ed-a4d2-4bbb-a5f3-5a7790a6aab1","owner":[],"postedDate":"March 20th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":64787289,"name":"Biological sciences/Cancer/Tumour biomarkers"},{"id":64787290,"name":"Biological sciences/Cancer/Gastrointestinal cancer/Colorectal cancer"}],"tags":[],"updatedAt":"2026-04-14T10:12:49+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-20 07:16:51","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9151067","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9151067","identity":"rs-9151067","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}