Shared requirement for MYC upstream super-enhancer region in tissue regeneration and cancer

A region located upstream of the MYC oncogene that carries more risk for inherited cancer than any other human genomic region is required for intestinal regeneration.

Abstract

Cancer has been characterized as a wound that does not heal. Malignant cells are morphologically distinct from normal proliferating cells but have extensive similarities to tissues undergoing wound healing and/or regeneration. The mechanistic basis of this similarity has, however, remained enigmatic. Here, we show that the genomic region upstream of Myc, which carries more cancer susceptibility in humans than any other genomic region, is required for intestinal regeneration after radiation damage. Failure to regenerate is associated with inefficient Ly6a/Sca1⁺ stem/progenitor cell mobilization, and almost complete failure to re-establish Lgr5+ cell compartment in the intestinal crypts. The Myc upstream region is also critical for growth of adult intestinal cells in 3D organoid culture. We show that culture conditions recapitulating most aspects of adult normal tissue architecture still reprogram normal cells to proliferate using a mechanism similar to that employed by cancer cells. Our results establish a function for the Myc^2-540 super-enhancer region as the genetic link between tissue regeneration and tumorigenesis, and demonstrates that normal tissue renewal and regeneration of tissues after severe damage are mechanistically distinct.

Introduction

Out of all genomic regions, the ∼500-kb super-enhancer region on 8q24 that regulates MYC expression carries the largest burden of population-level cancer susceptibility (Sur et al, 2013; Buniello et al, 2019); it contains several common alleles that increase the risk for multiple major forms of human cancer (Gudmundsson et al, 2007, 2009, 2012; Yeager et al, 2009; Buniello et al, 2019). For example, the cancer-associated G allele of SNP rs6983267 is found in 50% and >85% of people with European and African ancestry, respectively (Haiman et al, 2007; Yeager et al, 2007). This allele alone increases the risk for colorectal and prostate cancers more than 20%, causing hundreds of thousands of cancer deaths globally per year (https://www.wcrf.org/cancer-trends/global-cancer-data-by-country/). Although this highly conserved super-enhancer region is known to regulate MYC expression in specific tumors, its normal function has not been established. We have previously deleted regulatory sequences syntenic to the super-enhancer region in mice (Dave et al, 2017) and demonstrated that the mice are healthy and fertile despite greatly reduced levels of MYC expression in epithelia of the breast, prostate, and intestine. The mice, however, are highly resistant to intestinal and mammary tumors. Although deletion of the region did not cause a major effect on fitness in our assays, the enhancer elements within it are highly conserved between mammalian species (Hallikas et al, 2006; Pomerantz et al, 2009; Tuupanen et al, 2009; Ahmadiyeh et al, 2010; Sotelo et al, 2010; Hnisz et al, 2013; Yan et al, 2013). This suggests that this region would also have a beneficial biological function that would balance the negative impact of the increased cancer predisposition.

Because MYC is one of the main drivers of cell proliferation, we hypothesized that its upstream super-enhancer region would be beneficial under some environmental condition that requires rapid cell proliferation, but which the mice fed and housed in a clean laboratory environment would not encounter. For example, the super-enhancer region could be important for responses to chronic infections, wounding, or other forms of tissue damage (Schafer & Werner, 2008; Plummer et al, 2016). In this work, we demonstrate that the regulatory region Myc^2-540 functions as the genetic link between tissue regeneration and tumorigenesis and establish that mechanisms of homeostatic cell proliferation that maintain the stable number of stem cells and their progeny are distinct from mechanisms used during the growth of intestinal organoids, healing tissues, and tumors, all of which require an exponential increase in the number of stem cells.

Results

Myc^{Δ2-540/Δ2-540} mice fail to recover from radiation-induced intestinal damage

In order to investigate whether the super-enhancer region was required for tissue repair and regeneration, we subjected the Myc^{Δ2-540/Δ2-540} mice to ionizing radiation, which causes uniform and highly reproducible damage to the intestinal lining (Kim et al, 2017; Gu et al, 2020). WT mice γ-irradiated for a total dose of 12 Gy rapidly developed intestinal damage and lost 7.0 ± 0.8% of their weight in 3 d (n = 19); the weight loss stabilized at that time, and the damage was largely repaired by day 5. In contrast, the Myc^{Δ2-540/Δ2-540} mice appeared unable to repair the damage, as they continued to lose weight beyond 3 d, losing on average 17% by 5 d (n = 10) (Fig 1A and B) at which time they had to be euthanized for humane reasons. Morphological analysis of the small intestines confirmed that the damage was repaired in the WT mice but sustained in the Myc^{Δ2-540/Δ2-540} mice because of failure to regenerate the intestinal crypts (Figs 1C and D and S1A). The WT mice displayed a strong proliferative response at both 3 and 5 d after irradiation, whereas at the same time points, in the Myc^{Δ2-540/Δ2-540} mice there was almost complete absence of proliferating cells (Fig 1E and F). The proliferative response was accompanied by the induction of MYC mRNA expression. Although irradiation induced MYC expression in the WT, with peak induction of Myc transcripts occurring at 3 days post-irradiation (dpi, Figs 2A and B and S1B and D), the number of Ki-67⁺ cells remained high at 5 dpi even when the MYC transcripts returned to pre-radiation levels. In Myc^{Δ2-540/Δ2-540} mice, Myc levels were almost undetectable both before and after irradiation (Figs 2A and B and S2D). The cellular architecture and expression profile of the small intestine of the Myc^{Δ2-540/Δ2-540} mice and WT were similar before irradiation (Fig S2A–C), suggesting that loss of proliferation was specific to the repair/regenerative stimulus. To decipher changes in cellular transcriptome when MYC levels increase after tissue damage, we performed single-cell RNA-seq (scRNA-seq) analysis at day 2 after irradiation. Analysis of scRNA-seq data revealed that Myc expression was limited to stem and transit amplifying cells of WT mice. Despite the irradiation, which robustly induced MYC in WT mice, stem cells or transit amplifying cells from Myc^{Δ2-540/Δ2-540} mice expressed almost no Myc, the signal being even lower than that from irradiated mature enterocytes of WT mice (Fig 2C). The Myc^{Δ2-540/Δ2-540} mice were, however, fully capable of mounting a normal DNA damage response as seen by the induction of p53 target genes Cdkn1a (p21) and Bbc3 (PUMA), which facilitate p53-dependent cell cycle arrest and apoptosis, respectively (Figs 2D and S1C). These results show that despite a normal tissue architecture, Myc^{Δ2-540/Δ2-540} mice fail to repair tissue after damage induced by ionizing radiation.

Figure 1. Myc^{Δ2-540/Δ2-540} intestine fails to regenerate in vivo post–γ-irradiation.

(A) Timeline of the irradiation experiment. (B) Myc^{Δ2-540/Δ2-540} mice (red) are more sensitive to weight loss post-irradiation compared with WT mice (blue). Percentage weight relative to the weight before irradiation is shown. WT: n = 19 (3 dpi), n = 11 (5 dpi), Myc^{Δ2-540/Δ2-540}: n = 21 (3 dpi), n = 10 (5 dpi). (C) Hematoxylin-and-eosin (H&E)–stained sections of small intestine showing Myc^{Δ2-540/Δ2-540} mice fail to regenerate crypts after radiation-induced damage (also see Fig S1A). (D) Crypt depth and villus height were significantly decreased in Myc^{Δ2-540/Δ2-540} mice compared with the WT after radiation. Longitudinally cut crypt and villi were measured from at least two mice per category. Each dot represents an individual crypt/villus. (E, F) Failure to regenerate the crypts is associated with loss of proliferation as shown by IHC for the proliferation marker Ki-67. Ki-67⁺ cells were counted within the crypt–villus axis from at least two mice per category (WT: nonirradiated [non-IR], n = 2; 3 dpi, n = 4; 5 dpi, n = 4 mice; Myc^{Δ2-540/Δ2-540}: non-IR, n = 2; 3 dpi, n = 3; 5 dpi, n = 3). Each dot represents the number of Ki-67⁺ cells in an individual crypt–villus. Statistical significance was calculated using an unpaired t test (two-tailed), and corresponding P-values are shown. A nonsignificant statistical difference is marked by ns. Error bars denote the mean ± SD. Magnification bars in C = 200 μm (inset = 50 μm), E = 50 μm. Representative crypts in E are outlined with dashed lines and marked with arrowheads.

Figure S1. Myc^{Δ2-540/Δ2-540} intestines fail to regenerate after damage.

(A) Higher magnification images of H&E-stained sections showing the inability of Myc^{Δ2-540/Δ2-540} crypts to regenerate after damage caused by γ-radiation. (B) Expression of ubiquitin C (Ubc) transcript in the intestine of nonirradiated mice and in mice after irradiation (3 and 5 dpi). Ubc transcripts were used as a positive control for RNA integrity of the sample shown in Fig 2A. (C) scRNA-seq, violin plots showing the induction of p53-mediated DNA damage response indicators Cdkn1a (p21) and Bbc3 (PUMA) in WT intestinal cells after γ-irradiation (2 dpi). (D) qPCR showing the time course of Myc induction after irradiation in the WT intestine (n = 3 mice per condition). Magnification bars = 50 μm. Error bars in D denote the mean ± SD. SC, stem cell; TA, transit amplifying cell; EC, enterocyte.

Figure 2. Myc^{Δ2-540/Δ2-540} intestine is unable to up-regulate Myc expression after irradiation.

(A, B, C) Myc expression (A) ISH; (B) qPCR, WT: n = 2 (non-IR), n = 3 (3 dpi), n = 3 (5 dpi); Myc^{Δ2-540/Δ2-540}: n = 2 (non-IR), n = 2 (3 dpi), n = 3 (5 dpi); and (C) scRNA-seq (violin plots depicting the expression of Myc in SC, TA, and EC populations) show almost complete absence of Myc transcripts after radiation. (D) scRNA-seq, violin plots showing that unlike Myc levels, the levels of p53-mediated DNA damage response indicators Cdkn1a (p21) and Bbc3 (PUMA) were comparable between WT and Myc^{Δ2-540/Δ2-540} intestine at 2 dpi (scRNA-seq). See also Fig S1. SC: stem cell; TA: transit amplifying cell; EC: enterocyte. Magnification bars in A = 50 μm. Representative crypts in A are outlined with dashed lines and marked with arrowheads.

Figure S2. Myc^{Δ2-540/Δ2-540} intestine has a normal proliferation and differentiation program during homeostasis.

(A) Similar pattern of cell type–specific marker expression in the WT and Myc^{Δ2-540/Δ2-540} intestine. IHC staining for lysozyme (Lys), Ki-67, mucin 2 (Muc2), and chromogranin A (Chga), and ISH staining for Lgr5 are shown. Leftmost panel: schematic of different cell types with their marker genes in parenthesis and their location within the crypt–villus axis of the intestine. (B) Positively stained cells per crypt–villus axis in IHC were counted for PC, TA, GC, and EE cell types. The data used for TA (Ki-67⁺) are the same as those used in Fig 1F. Quantification of Lgr5⁺ cells was done from scRNA-seq data. (C) Heat map showing that all the major cell types of the intestine can be identified from scRNA-seq analysis of intestinal crypts in both the WT and Myc^{Δ2-540/Δ2-540} mice. Clustering was done using the top 2,000 highly variable genes, and cell type–specific markers shown on the y-axis were according to Yum et al (2021). (D) scRNA-seq confirmed a dramatic reduction in Myc expression within the SC and TA clusters. N-Myc expression was not detectable in both the WT and Myc^{Δ2-540/Δ2-540} mice. (E) Single-cell ATAC-seq analysis showing the absence of compensatory changes in accessibility of Myc enhancer elements around the 538-kb enhancer region deleted in Myc^{Δ2-540/Δ2-540} mice. (F, G, H) Increased expression of L-Myc in the intestinal stem and transit amplifying cells of Myc^{Δ2-540/Δ2-540} mice compared with WTs. (F, G) scRNA-seq (G) ISH (RNAscope). Bright-field microscope images are shown. Ubc transcripts were used as a positive control for RNA integrity of the sample. (H) Number of L-Myc transcripts were quantified as the number of punctate RNAscope signal per crypt (WT: n = 2; Myc^{Δ2-540/Δ2-540}: n = 3). Each dot represents an individual crypt. Magnification bars in A = 50 μm (inset bar = 10 μm), G = 20 μm. Error bars in (B, H) denote the mean ± SD. Statistical significance was calculated using an unpaired t test (two-tailed in (B) and one-tailed in (H)). A nonsignificant statistical difference is marked by ns. PC, Paneth cell; SC, stem cell; TA, transit amplifying cell; GC, goblet cell; EE, enteroendocrine cell; TC, tuft cell; EC, enterocyte; EP, enterocyte progenitor.

Failure to regenerate is associated with loss of induction of MYC

To determine whether the loss of regenerative potential seen in Myc^{Δ2-540/Δ2-540} intestines was due to MYC, we analyzed the expression of all genes within 5 Mb of Myc. By far, the strongest effect seen was for Myc itself, whose expression was almost completely abolished in the mutant intestine. Transcripts of some genes telomeric of Myc were also down-regulated, but to a much lesser extent. These include the noncoding RNA Pvt1 and the gasdermin C genes Gsdmc2, Gsdmc3, and Gsdmc4 (Table S1). However, only the expression of Myc and Pvt1 was specifically up-regulated in WT cells upon irradiation. Because Pvt1 lncRNA is thought to respond to the same enhancer elements and elicit its effect via interaction with Myc (Tseng et al, 2014; Cui et al, 2016), the results indicate that MYC is the primary effector of the Myc^{Δ2-540/Δ2-540} phenotype. Consistent with this, unbiased gene set enrichment analysis revealed that a set of MYC targets (V1) failed to be up-regulated by irradiation in stem cells of the Myc^{Δ2-540/Δ2-540} mice (Fig 3A, Table S2). We also specifically analyzed the expression of a set of 126 functionally conserved MYC targets from Zielke et al (2022), most of which are involved in metabolism and ribosome biogenesis. The regulation of this set of genes is conserved between Drosophila and humans with most (∼90%) of the genes being MYC-inducible. These functionally conserved MYC targets were also strongly up-regulated in the stem cells of irradiated WT mice compared to Myc^{Δ2-540/Δ2-540} mice (Fig 3B, Table S3) with 64% of the genes showing a significant (P < 0.05) decrease in Myc^{Δ2-540/Δ2-540} cells. We also checked the expression of known MYC-repressed target genes (Diamant et al, 2025); however, these did not show a marked change in expression after irradiation (Fig 3C, Table S4). Taken together, the results establish that induction of MYC drives a robust program of increased ribosome biogenesis that can facilitate growth of new tissue to repair the damage caused by the irradiation. However, among the MYC-regulated genes, we also observed induction of the negative translational regulator eukaryotic initiation factor 4E-binding protein 1 (Eif4ebp1/4EBP1); Tameire et al, 2019; Fig 3D), suggesting that induction of MYC increases translational capacity, through its effect on ribosomal biogenesis, but is not sufficient to induce translational activity. An increase in translational activity will require additional signaling, such as activation of mTOR complex (mTORC) kinase that can phosphorylate and inactivate 4EBP1 (Saxton & Sabatini, 2017). Consistent with this model, it has been previously shown that mTORC signaling is required for regeneration of the intestine after radiation damage (Sampson et al, 2016). We also examined the expression of mTOR in our scRNA-seq data after irradiation in Myc^{Δ2-540/Δ2-540} and WT cells. At 2 dpi, the WT cells express only slightly higher levels of mTOR mRNA compared with Myc^{Δ2-540/Δ2-540} cells (log₂ FC = 0.246 in stem cells) (Table S5). Such a small change in mRNA is unlikely to explain changes in levels of Eif4ebp1 between these conditions. Instead, the change is consistent with MYC induction of Eif4ebp1 mRNA combined with TOR-dependent phosphorylation of the 4EBP1 protein (Hruby et al, 2024 Preprint; Tameire et al, 2019). IHC staining of the intestines of both WT and Myc^{Δ2-540/Δ2-540} mice before radiation damage revealed a strong signal for 4EBP1 and phosphorylated 4EBP1 (Thr37/46), whereas after radiation (3 dpi), staining was more prominent in the regenerating crypts of the WT intestine (Fig 3E and F) consistent with a role of both MYC and mTORC in the regenerative response.

Figure 3. Myc^{Δ2-540/Δ2-540} intestine fails to induce MYC targets after irradiation, compromising mTORC signaling.

(A) GSEA showing the most enriched Gene Ontology and Hallmark gene sets in the intestinal SC cluster of irradiated WT mice compared with Myc^{Δ2-540/Δ2-540} mice (also see Table S2). (B) Deletion of the Myc^2-540 region causes a modest decrease in conserved MYC target (Zielke et al, 2022) expression, whereas after irradiation, strong induction of MYC targets is seen in WT compared with Myc^{Δ2-540/Δ2-540} mice. Top 5 differentially induced MYC targets in the stem cell compartment are marked (also see Table S3). (C) Targets repressed by MYC (Diamant et al, 2025) are not affected to the same extent as conserved MYC targets after radiation. (D) Altered regulation of MYC and mTORC target Eif4ebp1 (4EBP1) (D) scRNA-seq data (2 dpi). (E) IHC showing the amount of 4EBP1 and phosphorylated 4EBP1 (phospho-4EBP1) before and after radiation. (F) Quantification of the IHC staining in E using QuPath version 1. Each dot represents an individual annotated region that was measured (sections from at least two mice per category were analyzed). In (B, C), red lines denote the median expression of MYC targets and blue dots mark targets with significant (P < 0.05) log₂ fold changes between WT and Myc^{Δ2-540/Δ2-540}. (B, C) Wilcoxon’s test was used for statistical analysis in (B, C). Statistical significance in F was calculated using an unpaired t test (two-tailed), and corresponding P-values are shown. A nonsignificant statistical difference is marked by ns. Error bars in F denote the mean ± SD. Magnification bars in E = 50 μm. SC: stem cell; TA: transit amplifying cell; EC: enterocyte.

Table S1. Mean expression (normalized read counts) of genes within the 5-Mb region around Myc whose expression can be detected in scRNA-seq data. ^{(13.2KB, xlsx)}

Table S2. GSEA of irradiated WT_IR versus Myc^{Δ2-540/Δ2-540} _IR stem cells after irradiation (2 dpi). ^{(21.1KB, xlsx)}

Table S3. Mean expression (normalized read counts) of conserved Myc targets. ^{(68.7KB, xlsx)}

Table S4. Mean expression (normalized read counts) of repressed Myc targets (Diamant et al, 2025). ^{(38.3KB, xlsx)}

Table S5. Expression of mTOR in stem and transit amplifying cells of WT and Myc^{Δ2-540/Δ2-540} mice. ^{(10.9KB, xlsx)}

We next asked whether developmental or homeostatic changes that may have occurred before radiation contributed to the failure of regeneration; we also performed scRNA-seq on the crypt-enriched epithelium of nonirradiated (non-IR) WT and Myc^{Δ2-540/Δ2-540} mice. Analysis of cell types by scRNA-seq and immunohistochemistry of markers of cell differentiation and proliferation before damage revealed no major differences between WT and Myc^{Δ2-540/Δ2-540} mice (Fig S2A–C). Despite the lack of a major difference in proliferation and histological markers, there were cell type–specific differences in gene expression between WT and Myc^{Δ2-540/Δ2-540} mice (Table S6). The top genes in the differentially expressed gene list ranked based on average expression (Fig S3) were genes involved in ribosome biogenesis (Table S6). Furthermore, in the normal nonirradiated Myc^{Δ2-540/Δ2-540} mice, the functionally conserved MYC targets were also modestly down-regulated within the stem and transit amplifying cells compared with WT mice (Fig 3B, Table S3). In contrast, there was no down-regulation of the MYC target genes in mature enterocytes, which do not express MYC. Single-cell ATAC-seq analysis confirmed the absence of compensatory changes in accessibility of Myc enhancer elements around the 538-kb enhancer region deleted in Myc^{Δ2-540/Δ2-540} mice (Fig S2E). However, we observed an increase in Mycl expression within the stem and transit amplifying cell clusters of Myc^{Δ2-540/Δ2-540} mice suggesting compensation of MYC function during homeostasis (Fig S2F–H). It is noteworthy that the functionally conserved MYC target gene set that we find enriched after tissue damage was originally detected from tumor cells (Zielke et al, 2022), suggesting that increased ribosomal biogenesis and metabolic flux drive both cancer and tissue repair. These results indicate that the regenerative failure of the Myc^{Δ2-540/Δ2-540} is most likely due to failure to respond to damage, as opposed to differences in tissue structure, morphology, or gene activity before the damage.

Figure S3. Differential expression analysis reveals a limited impact of Myc^2-540 deletion during homeostasis.

MA plots are shown for the differential gene expression analysis (WT versus Myc^{Δ2-540/Δ2-540}) in various cell types of the intestine. Genes with a log₂ FC > 1 and P < 0.05 are marked in red. Ribosomal biogenesis genes are among the top genes affected by Myc^{Δ2-540/Δ2-540} deletion (see Table S6). The Wilcoxon test was used for statistical analysis.

Table S6. Differential expression of genes (WT versus Myc^{Δ2-540/Δ2-540}). ^{(10.6MB, xlsx)}

Regeneration defects of adult and embryonic Myc^{Δ2-540/Δ2-540} intestines in 3D cultures

To further characterize the role of the Myc^2-540 region in tissue repair, we decided to test a recently described ex vivo tissue regeneration assay, where mouse ileum is decellularized with sodium deoxycholate resulting in a matrix reminiscent of the 3D intestinal structure in vivo that contains the distinct extracellular matrix structures where former crypts and villi were located. Overlaying this decellularized extracellular matrix (dECM) with isolated crypts results in repopulation of the empty crypt pits and regeneration of the entire intestinal epithelium (Iqbal et al, 2025). In this assay, cultures of WT crypts repopulated the dECMs and gave rise to crypt and villus structures, whereas crypts from Myc^{Δ2-540/Δ2-540} mice failed to do so (Fig 4A). As repopulation of the bare dECM requires cellular motility, with proliferation and self-renewal restricted to the empty crypt pits, we next placed isolated crypts in standard organoid culture conditions in Matrigel. Surprisingly, adult intestinal organoids from Myc^{Δ2-540/Δ2-540} mice failed to grow also in such culture, where self-renewal and proliferation-promoting signals are provided evenly (Fig 4B).

Figure 4. Distinct MYC requirements in adult and fetal intestinal tissues during regeneration in vitro.

(A) Immunofluorescence analysis showing failure of adult Myc^{Δ2-540/Δ2-540} crypts to re-epithelialize intestinal scaffolds. Representative images from two independent experiments are shown. (B) Crypts from Myc^{Δ2-540/Δ2-540} adult intestine fail to establish 3D organoid cultures in Matrigel. Representative images from at least three independent experiments are shown. (C) Intestinal epithelium from Myc^{Δ2-540/Δ2-540} E16.5 embryos can initiate spheroid formation in Matrigel but fail to generate budding organoids with mature crypt-like structures reminiscent of adult cultures during 10 d of culture. Representative images from three independent experiments are shown. (D) Late-passage organoids showing recovery of the budding phenotype in Myc^{Δ2-540/Δ2-540} fetal intestinal 3D cultures over time (n = 1 experiment for each category). (E) qPCR analysis showing the almost complete absence of Myc expression in Myc^{Δ2-540/Δ2-540} spheroids and potential compensation by N-MYC in 8- to 9-d-old 3D cultures (WT: n = 2; Myc^{Δ2-540/Δ2-540}: n = 2) and in organoids after passage 18 (n = 1 for each category). (F) Myc^{Δ2-540/Δ2-540} E16.5 fetal intestine is morphologically similar to WT in vivo. H&E-stained sections are shown. Magnification bars in (A, D) = 20 μm, (B, F) = 50 μm, (C) = 4 μm. Error bars in (E) denote the mean ± SD. Red arrowheads in (B) denote organoids magnified in the inset.

Establishment of organoid culture from adults, like tissue repair, is associated with a change towards a more fetal-like state (Nusse et al, 2018; Yui et al, 2018). We therefore tested whether we could establish 3D cultures from intestinal epithelium/crypts of E16.5 Myc^{Δ2-540/Δ2-540} embryos. In contrast to adults, 3D cultures from Myc^{Δ2-540/Δ2-540} embryos did grow, forming spheroids. However, unlike WT cultures, the Myc^{Δ2-540/Δ2-540} spheroids failed to generate budding mature crypt-like structures during the first 10 d of culture (Fig 4C). After extended passaging of the cultures for more than 2 wk, the spheroid growth pattern was often lost, and budding organoids, with mature crypt-like structures, formed also in the Myc^{Δ2-540/Δ2-540} 3D cultures (Fig 4D). Henceforth, the organoids, with mature crypt-like structures, are referred to as budding organoids. We confirmed that Myc was dramatically reduced in Myc^{Δ2-540/Δ2-540} spheroids and late-budding organoids, both of which expressed less than 5% of the WT Myc levels (Fig 4E). These results suggest that mechanism/s independent of MYC can drive embryonic intestinal growth in culture. Because previous studies have shown that both MYC and N-MYC are expressed in the developing mouse intestine at this stage, we also measured Mycn expression from the organoid cultures, and found that Myc^{Δ2-540/Δ2-540} organoid cultures expressed ∼3.5-fold higher levels of Mycn compared with the WT (Fig 4E), potentially compensating for the loss of MYC. To confirm that the failure of the Myc^{Δ2-540/Δ2-540} embryonic organoids to bud was due to MYC loss and not other effects caused by the loss of the regulatory element, we treated WT embryonic organoids with the MYC inhibitor 10058-F4. This compound has previously been shown to induce a MYC loss–dependent diapause in early mouse embryos (Scognamiglio et al, 2016). Treatment of WT embryonic organoids during passage 1, when they are already budding, with 100 μM of 10058-F4 resulted in a clear shift from a budding to a spheroidal growth pattern (Fig S4). Despite the difference in culture, the WT and Myc^{Δ2-540/Δ2-540} intestines of E16.5 embryos were morphologically comparable in vivo (Fig 4F), which is in line with our observation from the adult intestine that WT and Myc^{Δ2-540/Δ2-540} are histologically and morphologically comparable during homeostasis. Taken together, these results establish that although the Myc upstream super-enhancer region is not required for normal tissue renewal in mouse intestine, it is required for normal in vitro maturation of embryonic organoid cultures, and for growth of adult intestinal cells under organoid culture conditions.

Figure S4. MYC inhibitor 10058-F4 results in a morphological switch in WT 3D fetal organoid cultures reminiscent of Myc^{Δ2-540/Δ2-540} fetal cultures.

MYC inhibitor was added during passage 1. Medium and drug were changed every other day. Representative images of three independent experiments are shown. Total number of organoids/spheroids counted: DMSO = 760 and 10058-F4 = 525. Magnification bar = 100 μm. Error bars denote the mean ± SD.

Myc^{Δ2-540/Δ2-540} mice fail to mobilize Ly6a (Sca1)⁺ stem/progenitor cells during regeneration

To identify the mechanism underpinning the failure of Myc^{Δ2-540/Δ2-540} intestines to regenerate, we determined whether the loss of this region affected the mobilization of reserve stem/progenitor cells, which previously have been shown to be associated with intestinal regeneration (Nusse et al, 2018; Yui et al, 2018). We performed scRNA-seq on cells from 11-d-old organoid cultures derived from E16.5 Myc^{Δ2-540/Δ2-540} and WT embryos. At this time point, the 3D cultures from WT contain organoids with distinct budding morphology, whereas cultures from Myc^{Δ2-540/Δ2-540} consist of large spheroids and are devoid of mature crypt-like buds. scRNA-seq analysis showed that although spheroids from Myc^{Δ2-540/Δ2-540} contained adult differentiated cell types including enterocytes and goblet cells (Fig S5), the stem cell compartment expressed low levels of canonical stem cell marker genes including Lgr5 (Fig S5). We performed additional clustering of the WT and Myc^{Δ2-540/Δ2-540} stem cell population to determine its substructure. This identified five stem cell subclusters (SSC1_F-SSC5_F) (Fig 5A and B) in which the expression of proliferation markers was only minimally affected by Myc^2-540 deletion (Tables S7 and S8). However, cells of clusters SSC1_F and SSC3_F were enriched in WT cultures and expressed adult stem cell markers including Lgr5. Myc^{Δ2-540/Δ2-540} cultures were instead enriched in cells belonging to subclusters SSC2_F, SSC4_F, and SSC5_F. These subclusters mainly expressed cell markers associated with fetal and regenerative states (Nusse et al, 2018; Yui et al, 2018) (Fig 5C, Tables S7 and S8).

Figure S5. Heat maps showing the altered expression of canonical intestinal stem cell markers in scRNA-seq data from 3D organoid cultures of fetal (E16.5) intestines from WT and Myc^{Δ2-540/Δ2-540} mice.

Gray color rectangles mark different intestinal epithelial cells, whereas white rectangles mark cells of mesenchymal origin. Cell type–specific markers shown on the y-axis were according to Yum et al (2021). PC, Paneth cell; SC, stem cell; TA, transit amplifying cell; GC, goblet cell; EE, enteroendocrine cell; TC, tuft cell; EC, enterocyte; EP, enterocyte progenitor.

Figure 5. Mobilization of Ly6a (Sca1)⁺ cells during intestinal regeneration in in vitro 3D organoid cultures and in vivo after irradiation damage is Myc^2-540-dependent.

(A) UMAP plots showing the distribution of five stem cell subclusters identified in WT and Myc^{Δ2-540/Δ2-540} fetal intestinal 3D cultures. The mean value of x and y for all the cells belonging to that subcluster was used to determine the center and marked with a circle on the UMAP (fixed radius of 1.5). (B) Proportion of stem cell subclusters relative to the total number of cells from each genotype in fetal intestinal 3D cultures (top) and heat map showing the expression of de novo identified marker genes for each stem cell subcluster (bottom). (C) Myc^{Δ2-540/Δ2-540} fetal intestinal cultures express high levels of fetal/regeneration-associated markers but have much reduced levels of Ly6a (Sca1) compared with WT. (D) UMAP analysis of scRNA-seq data showing the loss of Ly6a (Sca1) induction in Myc^{Δ2-540/Δ2-540} mice compared with WT after irradiation. Ly6a (Sca1)-expressing cells (red) are distributed in SC, TA, and EP populations. Clusters were marked in the same way as in (A). (E) scRNA-seq analysis showing the reduced expression of conserved MYC targets and altered cell cycle distribution of Ly6a (Sca1)⁺ cells from irradiated Myc^{Δ2-540/Δ2-540}_IR intestine compared with irradiated WT intestine (WT_IR) (2 dpi). The red line in E denotes the median expression of conserved MYC targets. SC, stem cell; TA, transit amplifying cell; EP, enterocyte progenitor.

Source data are available for this figure.

Table S7. Differential gene expression of genes in stem cell subclusters of fetal organoids (WT versus Myc^{Δ2-540/Δ2-540}). ^{(5.9MB, xlsx)}

Table S8. Expression of stem cell marker and proliferation genes in fetal organoid stem cell subclusters. ^{(16.6KB, xlsx)}

Despite expressing several fetal markers, Myc^{Δ2-540/Δ2-540} stem cells had much lower expression of the fetal/regenerative marker Ly6a (Sca1) compared with WT (Fig 5C). Because the expression of Ly6a (Sca1) is known to be induced by colitis and radiation-induced damage, as well as damage caused by helminth infections (Nusse et al, 2018; Yui et al, 2018), we checked whether the expression of Ly6a (Sca1) was also affected in vivo in Myc^{Δ2-540/Δ2-540} mice after radiation damage. Analysis of our scRNA-seq data showed that irradiation of the WT intestine mobilized a robust population of Ly6a (Sca1)-expressing cells. These cells were distributed among the stem, transit amplifying, and enterocyte progenitor cell clusters, confirming their stem cell potential (Fig 5D). In contrast, the Myc^{Δ2-540/Δ2-540} mice almost completely lacked Ly6a (Sca1) expression after irradiation (Fig 5D). Furthermore, the failure of Myc^{Δ2-540/Δ2-540} mice to effectively mobilize Ly6a (Sca1)⁺ cells after irradiation could be attributed to reduced proliferation linked to a loss of MYC target induction (Fig 5E). These results demonstrate that although fetal intestinal 3D cultures can be propagated in almost complete absence of MYC, the adult repair and regenerative state is critically coupled to a MYC-dependent proliferation and mobilization of regenerative stem/progenitor cells.

Tumors also show Myc^2-540-dependent Ly6a (Sca1)⁺ cell recruitment

The mechanism of repair/regeneration is tightly coupled to development of neoplastic lesions; we therefore investigated whether defective recruitment of Ly6a (Sca1)-expressing cells could explain the tumor resistance of Myc^{Δ2-540/Δ2-540} mice. We have earlier shown that Myc^{Δ2-540/Δ2-540} mice and mice deficient of a single enhancer at the Myc locus (Myc-335^−/−) display 98% and 70% resistance to intestinal polyps on Apc^Min background (Sur et al, 2012; Dave et al, 2017). Analysis of disease-free survival showed that Apc^Min/+; Myc-335^−/− mice survived longer than WT mice, but eventually had to be euthanized by 1 yr of age because of intestinal polyposis (median survival = 306 d [n = 7] compared with 156 d for Apc^Min/+ [n = 9]). In contrast, most (94%) of Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice lived without observable adverse health effects for more than 500 d, but developed symptoms consistent with intestinal tumorigenesis (e.g., weight loss, blood in stool, rectal prolapse, and/or abdominal bloating) before 2 yr of age. The median survival was 546 d (18.2 mo), corresponding to normal old age in the C57BL/6 strain (18–24 mo) (Fox et al, 2006) (Fig 6A). To further investigate the mechanism of tumor resistance conferred by Myc^2-540 loss, we performed scRNA-seq analysis on intestines from 4-mo-old Apc^Min/+ and Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice. At this time point, Apc^Min/+ intestines have on an average ∼50 polyps, whereas Apc^Min/+; Myc^{Δ2-540/Δ2-540} intestines are almost completely devoid of tumors (Dave et al, 2017). Similar to our results from irradiation experiments, the Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice had a lower number of Ly6a (Sca1)⁺ cells compared with the Apc^Min/+ intestines (Fig 6B). To determine the role of Ly6a (Sca1)⁺ cells in tumorigenesis, we characterized the late tumors that do develop in very old Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice. These tumors are often relatively large, but have a nonpolypoid morphology that is very distinct from polyps that form in Apc^Min/+ mice (Fig 6C). Apc^Min/+; Myc^{Δ2-540/Δ2-540} tumors often have invaded radially to encircle the entire small intestine, and display a lateral “pushing edge” growth pattern, instead of the finger-like pattern common to polyps. The tumors were proliferative and Wnt-dependent based on the expression of Ki-67 and nuclear β-catenin, respectively (Fig 6D). These results establish that although MYC upstream super-enhancer region is necessary for formation of a particularly common type of intestinal tumors, loss of this element does not completely eliminate tumorigenesis in the intestine. The phenotype of Myc^2-540 deletion is thus consistent with disruption of tissue-specific regulatory mechanisms rather than a blanket block of Myc expression. This is further corroborated by the observation that Myc^{Δ2-540/Δ2-540} mice do not show resistance to lymphoid proliferative lesions/tumors induced by PTEN loss (Fig S6A and B) as opposed to the striking resistance for intestinal tumors.

Figure 6. Tumors require Myc^2-540-dependent Ly6a (Sca1)⁺ cell mobilization but differ from wound repair with respect to Notum expression.

(A) Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice survive to old age (average = 18.2 mo, n = 18) compared with Apc^Min/+ (average = 5 mo, n = 9) because of their inherent tumor resistance. (B) Tumor resistance of Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice is correlated with a decrease in Ly6a (Sca1)-expressing cells within the stem cell (SC), transit amplifying cell (TA), and enterocyte progenitor (EP) cell populations as determined by scRNA-seq analysis. (C) Polypoid nature of Apc^Min/+ tumors (4 mo of age) is predominantly lost in the few tumors that develop in old Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice. A tumor from 1.5-yr-old Apc^Min/+; Myc^{Δ2-540/Δ2-540} is shown. The magnified region shows a well-demarcated boundary between the tumor and the normal epithelium. (D) IHC showing proliferation (Ki-67) and active nuclear β-catenin within the tumor of aging (1.5 yr old) Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice. The bottom-most panel shows Ki-67 and active nuclear β-catenin within the normal epithelium of aging Apc^Min/+; Myc^{Δ2-540/Δ2-540} intestine. (E) ISH showing the expression of Ly6a (Sca1), Myc, Notum, and Lgr5 in tumors from 4-mo-old Apc^Min/+ and 1.5-yr-old Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice. (F) Top panel: percentage of cells belonging to the regenerative stem cell subcluster (SSC_reg), common to both WT irradiated (WT_IR; 2 dpi) and Apc^Min/+ intestines, is shown. Percentage is relative to the total stem cell number. Bottom panel: de novo identified marker genes of SSC_reg population. (G) Expression of Myc, Notum, Lgr5, and Ly6a (Sca1) in the stem cell subclusters across different genotypes and conditions. (H) Heat map showing the expression of (top panel) SSC_reg marker genes (from (F)) and (bottom panel) revival stem cell (SC_rev) markers (Ayyaz et al, 2019) across the stem cell subclusters identified in fetal (E16.5) organoid cultures from WT and Myc^{Δ2-540/Δ2-540} mice (SSC1_F-SSC5_F). The Gapdh gene from the SSC_reg marker list was not detected in scRNA-seq data from fetal organoid cultures. Magnification bars = 100 μm except in (C) where the magnification bar for an overview image of the Apc^Min/+; Myc^{Δ2-540/Δ2-540} tumor = 500 μm. T marks the tumor region, N* denotes normal epithelium bordering the tumor, and N denotes normal epithelium away from the tumor. The arrowhead in (B) marks the cell with nuclear β-catenin staining in normal epithelium. The arrowhead in (E) marks one Lgr5-expressing crypt across panels. SC, stem cell; TA, transit amplifying cell; EP, enterocyte progenitor.

Source data are available for this figure.

Figure S6. Myc^2-540 deletion does not inhibit lymphoproliferative lesions/tumors induced by PTEN loss.

(A) Kaplan–Meier survival curves for Pten^+/− and Pten^+/−; Myc^{Δ2-540/Δ2-540} are shown. Mice were euthanized when they reached the humane endpoint. (B) H&E-stained sections of lymphoproliferative lesions/tumors in the inguinal mammary gland of Pten^+/−; Myc^{Δ2-540/Δ2-540} mouse (bottom panel) and a normal inguinal mammary gland with a single well-defined lymph node from a WT mouse (top panel) are shown. Magnification bar = 500 μm.

Source data are available for this figure.

To further characterize the intestinal tumors in aged Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice, we performed ISH on tumors from old Apc^Min/+; Myc^{Δ2-540/Δ2-540} and from 4-mo-old Apc^Min/+ mice.

We used ISH probes against Myc and Lgr5 to determine the contribution of Myc and adult stem cells to tumorigenesis in Apc^Min/+ and Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice. Because Ly6a (Sca1) was altered in Myc^{Δ2-540/Δ2-540} intestine both during in vivo and in vitro regeneration, we also studied its expression in the tumors. Finally, Notum, which provides a growth advantage to Apc^−/− stem cells over the neighboring WT counterparts, was selected to probe Apc mutant cells in and around the tumors (Flanagan et al, 2021; Yan et al, 2021; Yum et al, 2021). Most cells of the normal epithelium of both Apc^Min/+ and Apc^Min/+; Myc^{Δ2-540/Δ2-540} did not express Notum or Ly6a (Sca1). However, unlike most of the normal epithelium, cells both within and close to the tumor from both Apc^Min/+ and Apc^Min/+; Myc^{Δ2-540/Δ2-540} commonly expressed Ly6a (Sca1) (Fig 6E). Based on morphology, the Ly6a (Sca1)⁺ cells within tumors appeared to represent both tumor cells and residual normal cells trapped within the tumor. The tumor also appeared to induce Ly6a (Sca1)⁺ expression in normal cells that were located at the boundary between malignant and normal epithelium, suggesting that the tumor is capable of damaging surrounding normal tissue and inducing a wound-healing response. Normal-looking crypts expressing Ly6a (Sca1) at the tumor boundary did not express detectable transcripts of Myc, or Notum, which is a specific marker of Apc^−/− cells (Figs 6E and S7), suggesting that these cells are at a growth disadvantage relative to the Myc-expressing tumor cells (Johnston, 2014; Di Giacomo et al, 2017). The expression pattern also indicates that the expression of Ly6a (Sca1) is MYC-independent. Unlike the normal epithelium, tumors in Apc^Min/+; Myc^{Δ2-540/Δ2-540} expressed Myc potentially via additional regulatory regions that are still intact in the Myc^{Δ2-540/Δ2-540} mice. Single-cell analysis of the stem cell population from Apc^Min/+ tumor-bearing epithelium and WT irradiated intestine showed that both contained a population of stem cells (SSC_reg) enriched in fetal/regenerative markers like Clusterin (Clu) (Fig 6F).

Figure S7. Ly6a/Sca1 is expressed in normal epithelium bordering the tumor.

ISH showing the expression of Ly6a (Sca1) and Notum in the tumor (T) and adjacent normal epithelium (N*) from 1.5-yr-old Apc^Min/+; Myc^{Δ2-540/Δ2-540} mice. The arrowhead points to crypt–villus of normal epithelium adjacent to the tumor, expressing Ly6a (Sca1) but not Notum. Magnification bar = 200 μm.

To determine how the wound and cancer-associated SSC_reg population of stem cells in the adult relate to the fetal stem cells described earlier, we checked the expression of SSC_reg marker genes (Fig 6F) in stem cell subclusters from fetal 3D organoid cultures. The adult regenerative gene set was more highly expressed in the fetal SSC2_F, SSC4_F, and SSC5_F subclusters, which are enriched in Myc^{Δ2-540/Δ2-540} cultures compared with the same cells of the WT mice (Fig 6H). This likely reflects the fetal reversion of adult stem cells during regeneration and cancer. However, unlike the fetal state that can support growth even with absent or reduced Myc and Ly6a (Sca1) expression, the adult regenerative state appears to critically depend on Myc- and Ly6a (Sca1)-expressing cells. In addition, SSC_reg population from Apc^Min/+ tumor–bearing epithelium and WT irradiated intestine could be distinguished from each other based on Notum expression. Notum was expressed exclusively in an Apc^Min/+ background, and its expression was not affected by Myc^2-540 deletion (Fig 6G). Because Notum in Apc^−/− cells is required for inhibition of neighboring WT stem cells and expansion of tumor cells, it would explain why wound-associated proliferation stops, whereas tumors continue to proliferate and invade into the surrounding tissue.

Discussion

Our results presented here establish the long-sought mechanistic link between wound healing and cancer, and show that rapid proliferation during tissue regeneration and cancer uses MYC regulatory mechanisms that are genetically distinct from those used during normal homeostasis. Although the role of MYC in regeneration and cancer is well established (Ashton et al, 2010; Kim et al, 2018; Sodir et al, 2022; Mule et al, 2024), our work expands these studies further by providing a function to a large upstream super-enhancer region of MYC. This region is highly conserved and carries the highest cancer susceptibility burden in our entire genome. Our prior study (Dave et al, 2017) showed that the entire region can be deleted from mice without severe phenotypic consequence, making the existence and conservation of such a cancer-predisposition region enigmatic. Here, we solve this mystery, showing that the MYC super-enhancer region is normally required for regeneration of intestinal tissues after damage. The conservation of this region can thus be explained by the critical role that intestinal health has on survival of infants and young mammals.

We find here also that organoids derived from mice lacking the MYC super-enhancer region completely fail to grow in culture. This indicates that cell culture models that are often considered to resemble normal in vivo tissue renewal actually drive cells to adopt an exponential stem cell proliferation program associated with regeneration and tumorigenesis. In striking contrast to the adult organoids, embryonic intestinal cells derived from MYC super-enhancer region–deficient mice could be cultured as organoids, suggesting that some fetal regenerative mechanisms are shut down in the adult state, making adult tissue regeneration completely dependent on MYC.

Given that up-regulation of ribosome biogenesis is required for MYC-driven proliferation (van Riggelen et al, 2010; Pihlajamaa et al, 2022; Zielke et al, 2022), the simplest model consistent with our findings is that different levels of protein synthesis are needed for distinct proliferative cell fate choices. In this model, relatively low levels of protein synthesis and MYC are sufficient for proliferation that maintains stem cell number in tissue renewal, whereas processes that require an exponential increase in the number of somatic stem cells, such as cell proliferation in vitro, regeneration, and tumorigenesis (Fig 7), require high levels of MYC and protein synthesis. Interestingly, the model has similarities to growth control in unicellular species, where nutrient availability specifies the protein synthetic rate and controls the balance between nonproliferative and proliferative states (Jorgensen & Tyers, 2004). Multicellularity could have evolved simply by the addition of a hierarchy of stem cells with new proliferative fates, which remain controlled by the same input variable, the total metabolic activity per genome. Consistently with this model, ES cells lacking all forms of MYC enter into a diapause-like nonproliferative state in vitro (Scognamiglio et al, 2016), and we also observe here that adult intestinal cells from Myc^{Δ2-540/Δ2-540} mice enter a nonproliferative state under full mitogen stimulation (organoid culture).

Figure 7. Model describing how cancer susceptibility-associated super-enhancer region Myc^2-540 drives regenerative growth in the fetal and adult intestine and in cancer.

Because most of the MYC targets are proteins involved in ribosomal biogenesis, the model further describes how different levels of protein synthesis can result in different proliferative cell fate choices ranging from normal homeostasis to regeneration and cancer. Schematic representation of the cancer risk distribution within the gene desert 1 Mb upstream and 2 Mb downstream of the MYC oncogene on human chromosome 8q24 is shown. In the topmost is shown human chromosome 8, with the 3-Mb 8q24 region outlined in red. The magnified view of the 3-MB region is shown below, with colored boxes showing risk regions for prostate cancer (PrCa), chronic lymphocytic leukemia (CLL), breast cancer (BrCa), colorectal cancer (CRC), bladder cancer (BlCa), lymphoma, lung cancer (LuCa), ovarian cancer (OvCa), endometrial cancer (EnCa), leiomyoma, acute lymphoblastic leukemia (B-ALL), and glioma. A red vertical line crossing the human genomic coordinate axis marks the enhancer element MYC-335 that carries the CRC risk–associated SNP rs6983267 (Sur et al, 2012). GRCh37/hg19 human and GRCm38/mm10 genome coordinates are shown. The schematic of the cancer susceptibility region is based on GWAS catalog data (https://www.ebi.ac.uk/gwas/). The syntenic Myc2-540 region located upstream of Myc on mouse chromosome 15 and deleted from the mouse genome in this study is shown in blue on the schematic for mouse chromosome 15.

Our work has important consequences for drug development. In particular, compounds that inhibit proliferation of normal cells in culture have commonly been discarded during the development process because of toxicity, even when a subset of them may actually target the very mechanism that is distinct between normal proliferative cells and cancer cells in vivo. Our finding that the Myc^2-540 element and MYC activity are required for adult but not fetal organoid growth provides an effective means to identify drugs that specifically target MYC, paving a way for development of a novel class of effective chemopreventive and antineoplastic agents. Furthermore, given that the first successful phase 1 clinical trial for MYC inhibitor (OMO-103) was recently published (Garralda et al, 2024), our results suggest a potential impact for combined therapies for cancer that include radiation and MYC inhibition.

Materials and Methods

Animals

Mouse lines used in this study have been described previously (Dave et al, 2017). The conditional knockout mice were generated on a C57BL/6N genetic background. After Cre-mediated deletion of the 538-kb region, mice were backcrossed to C57BL/6J for two generations and subsequently intercrossed. All mice were kept on a 12:12-h light cycle in standard housing conditions and provided with food (Teklad Global Diet 2918; Envigo) and water ad libitum. Animal experiments were done in accordance with the regulations and with approval of the local ethical committee (Stockholms djurförsöksetiska nämnd). Mice were irradiated with Gammacell 40 Exactor that uses cesium-137 as a source for γ-rays. Mice were given two doses of 6 Gy with a 6-h interval in between the doses. After irradiation, mice were provided with soft food. For measuring survival of mice with Apc^Min mutation (strain 002020 from The Jackson Laboratory), the general health of the animal was monitored and mice were euthanized when they reached the humane endpoint according to the ethical permit. Mice that had to be euthanized for reasons unrelated to intestinal tumorigenesis were not included in the analysis. For PTEN mutation (strain 01XH3 from NCI mouse repository), mice were euthanized at the humane endpoint, predominantly because of lymphoproliferative lesions that manifested as nodules/lumps on the body of mice.

Immunohistochemistry and RNA ISH

PFA-fixed and paraffin-embedded (FFPE) tissues were sectioned at 5 μm thickness and subjected to immunohistochemistry using standard methods. For antigen retrieval, the sections were boiled in microwave for 20 min in 10 mM citrate buffer, pH 6.0, and subsequently processed for respective antibody stainings. The following antibodies and dilutions were used: anti-Ki-67 (1:200; Abcam), anti-lysozyme (1:500; Abcam), anti-Muc2 (1:500; Abcam), anti-Chga (1:1,000; Abcam), anti-4EBP1 (1:100; Abcam), and anti-phospho-4EBP1 (1:2,500; Cell Signaling). ISH was also performed on 5-μm-thick sections using the RNAscope technology and according to the manufacturer’s protocol. Probes against mouse Myc (catalog nr: 413451), L-Myc (catalog nr: 552711), Lgr5 (catalog nr: 312171), Clu (catalog nr: 427891), Ly6a (Sca1) (catalog nr: 427571), and Notum (catalog nr: 428981) were used. RNA integrity was confirmed using the control probe for Ubc (catalog nr: 310771).

Quantification of microscopy data

Longitudinally sectioned crypt–villus structures were used for quantification of crypt depth, villus height, Ki-67, and cell type–specific marker stainings (CHGA, MUC2, LYZ). For crypt depth, the distance from the bottom of the crypt to villus junction was measured in microns using QuPath. Similarly, villus height was measured from the crypt–villus junction to the villus tip. Ki-67⁺, CHGA⁺, MUC2⁺, and LYZ⁺ cells were manually counted. For quantifying 4EBP1 and phospho-4EBP1 stainings, sections were manually annotated in QuPath, with each annotated region containing several crypt–villus structures. Subsequently, the % DAB-positive pixels were quantified per annotated region with QuPath. For quantification of Mycl ISH stainings, positive punctate stainings (dots) within longitudinally cut crypts were manually counted under a microscope.

qRT-PCR

Approximately 1 cm of the mouse distal ileum was cut into small pieces and stored in RNAlater/RNAProtect (QIAGEN). For 3D cultures, plates were placed on ice to melt the Matrigel and organoids/spheroids were collected by centrifugation. RNA was prepared using a combination of TRIzol (Ambion) and RNeasy Mini Kit (QIAGEN). cDNA was synthesized from either 1 μg (tissue) or 100 ng (organoid) of total RNA using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). qPCR was performed using SYBR Green Master Mix (Thermo Fisher Scientific) on a LightCycler 480 instrument (Roche). Primers for Myc and Actin were as described previously (Dave et al, 2017). For Mycn PrimeTime qPCR, primers from Integrated DNA Technologies (IDT) were used (Mm.PT.58.29713369).

scRNA-seq

Mouse intestinal crypts were isolated according to Sato and Clevers (Sato & Clevers, 2013) and dissociated with TrypLE Express (37°C/20 min; Invitrogen), and single-cell suspensions were run on a 10x Chromium controller using the Chromium Next GEM Single Cell 3′ kit v3.1 (catalog nr: PN-1000268) and according to the manufacturer’s protocol (10x Genomics). The cell number targeted was 6000. For nonirradiated samples, crypts were isolated from ileum and single-cell suspensions of FACS-sorted 7AAD⁻ (Life Technologies), CD45⁻ (eBioscience), CD31⁻ (eBioscience), TER-119⁻ (eBioscience), and EpCAM⁺ (eBioscience) cells were used. For irradiated samples, crypts from the whole small intestine were isolated and 7AAD⁻ FACS-sorted cells were used. For single-cell suspensions from 3D cultures, organoids were treated with TrypLE for 20 min and cells were fixed with PFA according to 10x Next GEM Single Cell Fixed RNA Sample Preparation Kit (catalog nr: 1000414). Libraries from fixed cells were prepared using Next GEM Single Cell Fixed RNA Hybridization and Library Kit (catalog nr: 1000415) and according to the manufacturer’s protocol (10x Genomics). All libraries were sequenced on Illumina sequencing platform NovaSeq 6000 or NextSeq 2000. Sequencing data were processed using either Cell Ranger (v6.0.1) (10x Genomics) or Cell Ranger (v7.2.0) (for fixed cells with Cell Ranger multi-pipeline) (10x Genomics). Quality control (QC) analysis of the Cell Ranger output was done in Scanpy (v1.9.1) (Wolf et al, 2018) as follows: mean and SD of genes in all cells (Mean [n_genes], SD [n_genes]) and mean, median, median absolute deviation (MAD), and SD of total counts in all cells were calculated. Cells were kept with the number of expressed genes larger than (median [n_genes]-MAD [n_genes]) total read counts in the range of (Mean [n_counts]-1.5*MAD [n_counts], Mean [n_counts]+3*SD [n_counts] and percentage of expressed mitochondrial genes larger than its median+3*MAD). In addition, genes expressed in less than 0.1% of cells were filtered out. The final number of cells used for further analysis was as follows: WT: 8509 cells, Apc^Min/+; Myc ^Δ2-540/^Δ2-540: 13763 cells, Apc^Min/+: 7468 cells, Myc ^Δ2-540/^Δ2-540: 5809 cells, Myc ^Δ2-540/^Δ2-540 irradiation day 2: 4,863 cells and 3,102 cells from WT irradiation day 2. The number of cells analyzed from fixed samples (3D organoid cultures) was as follows: 6,699 cells from WT and 3,879 cells from Myc ^Δ2-540/^Δ2-540. Subsequently for dimensional reduction, the top 2,000 highly variable genes were selected based on scRNA raw read count matrix (scanpy.pp.highly_variable_genes; flavor = ’seurat_v3′). There were no mitochondrial genes in the top 2,000 highly variable genes. The raw read count matrix with the top 2,000 highly variable genes was subjected to the Bayesian variational inference model scVI (v0.16.4) (Lopez et al, 2018) to infer a latent space for all single cells. The batch effect among different samples was automatically removed for all samples within the scVI. After that, two-dimensional visualizations were calculated using Uniform Manifold Approximation and Projection (UMAP). Clustering analysis was performed using the Leiden algorithm. First, a k-nearest–neighbor graph from the scVI latent space was constructed, and then, the sc.tl.leiden function in Scanpy was used with the resolution of 1. Clusters were annotated into intestinal cell types based on published marker genes (Yum et al, 2021). Subcluster analysis was done within SC clusters with a resolution set to 0.3 (3D organoid cultures) or 0.2 (adult intestinal epithelium). The location of the cell-type clusters/subclusters was marked on the UMAP by calculating the mean value of x and y of all the cells belonging to that cluster/subcluster to find the center and then circling it on the UMAP (fixed radius of 1.5). For further analysis, the single-cell raw read count matrix of all the genes was input into R package Scran (v1.24.1) to calculate the size factor for each cell (Lun et al, 2016). To normalize the gene expression across cells, each gene’s read count in a cell was divided by that cell’s size factor in Scanpy. The normalized gene read counts were log₂-transformed, where 1 pseudocount was added to gene’s read count to avoid a negative infinity value in the analysis (log₁ P-value). The z-score for the log₁ P-value for each gene was calculated and used for heat map visualization of cell type–specific marker genes. Differentially expressed genes between the genotypes were determined using Scanpy’s function sc.tl.rank_genes_groups. The analysis was done specifically in the stem cell (SC), transit amplifying cell (TA), and enterocyte (EC) clusters. Within this function, the Wilcoxon test was used to calculate P-values and the false discovery rate by applying the Benjamini–Hochberg approach. The log fold changes of the 126 known MYC targets (Zielke et al, 2022) were retrieved from the differential gene expression results and plotted using a dot plot. Polr1f was not detected in our dataset and was excluded from analysis (Table S3). The ranking of the genes on the x-axis was based on the median log fold change values of each MYC target gene among stem cell (SC), transit amplifying cell (TA), and enterocyte (EC) clusters combined. Gene set enrichment analysis was done using the R package clusterProfiler (v4.4.4). Gene signatures were obtained from Gene Ontology Biological Process gene sets (Ashburner et al, 2000) and the MSigDB “Hallmark gene sets” (Liberzon et al, 2015). All the expressed genes in the SC cluster were pre-ranked by their log fold changes, which were determined as described in the previous section. GSEA was then performed using ClusterProfiler (ClusterProfiler::GSEA; pvalueCutoff = 1, pAdjustMethod = ´BH´). The list of GSEA results is provided in Table S2.

Single-cell ATAC-seq (scATAC-seq)

scATAC-seq was performed according to Cheng et al (2021) Preprint. Single-cell suspensions of mouse intestinal crypts were prepared as in Andrews et al (2021), and single cells were FACS-sorted into a 384-well plate with 3 μl lysis buffer in each well (0.03 ml 1 M Tris, pH 7.4, 0.0078 μl 5 M NaCl, 0.075 μl 10% IGEPAL, 0.075 μl RNase inhibitor, 0.075 μl 1:1.2 M ERCC, 2.7372 μl H₂O). After cell lysis, 1 μl of the lysis buffer containing the nuclei was used for scATAC library preparation, whereas the remaining 2 μl was discarded. The scATAC in situ tagmentation was performed with 2 μl of the Tn5 tagmentation mix (0.06 μl 1 M Tris, pH 8.0, 0.0405 μl 1 M MgCl₂, 0.2 μl Tn5) at 37°C for 30 min. After tagmentation, 2 μl of the supernatant was aspirated and the nucleus was then washed once with 10 μl ice-cold washing buffer (0.1 μl 1 M Tris, pH 7.4, 0.02 μl 5 M NaCl, 0.03 μl 1 M MgCl₂, 9.85 μl H₂O). The remaining Tn5 was inactivated by adding 2 μl 0.2% SDS/Triton X-100 and incubating at room temperature for 15 min followed by 55°C for 10 min. Before barcoding PCR, genomic DNA was extracted from chromatin by adding 2 μl Proteinase K (0.0107 au/ml) and incubation at 55°C for 1 h followed by heat inactivation at 70°C for 15 min. Barcoding PCR was done using KAPA HiFi PCR Kit (Roche) in a final volume of 25 μl. The PCR condition was 72°C/15 min, 95°C/45 s, (98°C/15 s, 67°C/30 s, 72°C/1 min) × 22 cycles, 72°C/5 min, and then 4°C hold. After the PCR, 2 μl reaction from each well was pooled and cleaned up twice using SPRI beads (at 1.3X volume) and the library was sequenced on Illumina NextSeq 550 as paired-end, dual index (37 + 37+8 + 8). For scATAC-seq data, raw sequence reads were trimmed based on quality (phred-scaled value of >20) and the presence of Illumina adapters, and then aligned to the mm10 genome build using BWA (Li & Durbin, 2010). Reads that were not mapped, not with primary alignment, missing a mate, mapq <10, or overlapping ENCODE blacklist regions (Amemiya et al, 2019) were removed. A custom script was used to summarize the paired-end reads into a de-duplicated fragment file suitable for downstream analysis. scATAC-seq data were analyzed using ArchR. Cells with fewer than 1,000 ATAC-seq reads mapping to the transcription start site, or with TSS enrichment <4, were excluded from analysis. The “iterative LSI” function with maxClusters = 6 was used for cluster analysis. Clusters with the high expression of “Muc2” were annotated as goblet cells and discarded. Clusters high in “Ascl2” and “Lgr5” were annotated as stem cells, whereas remaining intermediate/low-expressing clusters were annotated as transit amplifying cells. Each of these superclusters was split by genotype and merged into in silico bulk experiments.

Organoid culture

Crypts were isolated from adult mouse small intestine (∼3 mo old) according to Sato & Clevers (2013). The crypts were seeded in Matrigel (Corning) and cultured in ENR medium (Advanced DMEM/F12 (Thermo Fisher Scientific), 1x penicillin–streptomycin (Thermo Fisher Scientific), 1x GlutaMAX (Thermo Fisher Scientific), 10 mM Hepes (Thermo Fisher Scientific), 1x B27 and 1x N2 (Thermo Fisher Scientific), 1.25 mM N-acetylcysteine (Sigma-Aldrich), 50 ng/ml of murine recombinant epidermal growth factor (R&D), 100 ng/ml recombinant murine Noggin (R&D), and 1 mg/ml recombinant murine R-spondin (R&D)). For establishing organoids from the E16.5 embryonal intestine, mating was set up and the day of the mating plug was taken as E0.5. The pregnant females were euthanized to get E16.5 embryos. For one experiment, 3–10 embryos of each genotype were used. The whole small intestine of the embryo was removed and cut into small pieces under a dissecting stereomicroscope (Leica). The pieces were rinsed with cold PBS and incubated with 20 ml of cold 2 mM EDTA at 4°C for 20 min. The epithelium was isolated by shaking, and fractions were collected as described for crypt isolation from adult mice. The epithelial fragments were seeded in Matrigel and cultured in ENR medium. MYC inhibitor 10058-F4 was purchased from MedChemExpress (HY-12702).

Re-epithelialization of intestinal scaffolds with crypts

The ileum from the small intestine of mice was decellularized as previously described by Iqbal et al (2025). Pieces of decellularized intestines (dECMs) were primed at 37°C with 50 µl of media containing Advanced DMEM/F12, 1x penicillin–streptomycin, 1x GlutaMAX (Thermo Fisher Scientific), 10 mM Hepes (Thermo Fisher Scientific), 1x B27 (Life Technologies), 1x N2 (Life Technologies), 1 mM N-acetylcysteine (Sigma-Aldrich), 50 ng/ml of murine recombinant epidermal growth factor (R&D), 100 ng/ml recombinant murine Noggin (PeproTech), 500 ng/ml recombinant murine R-spondin (R&D), and 10 μM Y-27632 (Sigma-Aldrich), for 15–60 min. This was followed by aspirating media carefully and plating around 100 crypts on each scaffold piece and incubating at 37°C for 10 min. Next, 10 µl of media was added to each scaffold piece and incubated at 37°C for 10 min, followed by an additional 20 µl per scaffold piece and incubation at 37°C for 20 min. Finally, 260 µl of media was added to each scaffold piece and put in the incubator at 37°C. The day after seeding, media were changed, and each scaffold piece was lifted to make it float. The second day after seeding, the media were changed to media without Y-27632. Media were subsequently replaced every other day. Re-epithelialized dECMs were fixed with 4% PFA at day 16 after plating. The re-epithelialized dECMs were analyzed using immunofluorescence. Primary antibody pan-laminin (1:300, 11575; Abcam) and secondary antibody Alexa 647–conjugated anti-rabbit (1:1,000; Life Technologies) were used. For nucleus detection, DAPI (1:1,000; Life Technologies) was used. Images were obtained with a Zeiss LSM980-Airy2 microscope using a 20x objective, NA = 08.

Data Availability

Sequencing data are available from GEO accession number GSE244298.