Summary
Chromatin domains delimited by CTCF can restrict the range of enhancer action. However, disruption of some domain boundaries results in mild gene dysregulation and phenotypes. We tested whether perturbating a domain with multiple developmental regulators would lead to more severe outcomes. We chose a domain with three FGF ligand genes—Fgf3, Fgf4, and Fgf15—that control different murine developmental processes. Deletion of a 23.9kb boundary defined by four CTCF sites led to ectopic interactions of the FGF genes with enhancers active in the brain and induced FGF expression. This caused orofacial clefts, encephalocele, and fully penetrant perinatal lethality. Loss of the single CTCF motif oriented towards the enhancers—but not the three towards the FGF genes—recapitulated these phenotypes. Our works shows that small sequence variants at particular domain boundaries can have a surprisingly outsized effect and must be considered as potential sources of gene dysregulation in development and disease.
Graphical Abstract
eTOC blurb
Chakraborty et al, studied the role of chromatin structure in regulation of three FGF genes. Mice with a single CTCF motif deletion died perinatally with severe phenotypes, showing that even small sequence variants at domain boundaries must be considered as potential sources of gene dysregulation during development and disease.
Introduction
Successful embryonic development requires dynamic spatial and temporal expression of genes in a cell type specific manner. This is particularly important for genes that encode developmental regulators of processes such as embryonic patterning or cell specification. Gene expression is regulated by enhancers that can act across large distances from their target genes1–4. The spatial range of enhancer action has been proposed to be delimited by topologically associated domains (TADs), which are characterized by higher levels of intra-domain contact frequency compared to neighboring chromatin regions5–10. TADs can prevent enhancer off-target gene activation by reducing the frequency of contacts across their boundaries. In vertebrates, these chromatin domains are mostly formed through the process of loop extrusion by the cohesin complex, where domain boundaries are frequently delimited by CTCF, which retains extruding cohesin complexes in an orientation-dependent manner determined by the sequence of its DNA binding motif11–16. The boundaries of these chromatin domains are often composed of clusters of multiple CTCF binding motifs in both orientations, which can increase the chances of cohesin retention and thus generate better insulation17,18.
The loop extrusion model predicts that perturbation of the extrusion machinery, or sequence variants that disrupt CTCF binding, can result in loss of insulation, ectopic enhancer-promoter interactions, and off-target gene activation with potential pathogenic outcomes. Indeed, such effects have been observed in multiple studies19–33. However, disruption of chromatin domains does not always strongly impact gene regulation or affect physiological processes. For example, in vitro, rapid degradation of cohesin or CTCF leads to the dissolution of TADs but the impact on transcription is minimal13,14. Disruption of some chromatin domains has a negligible effect on gene regulation and animal development19,34–39. Our own previous work showed that some Sox2 enhancers can bypass strong ectopic CTCF-mediated boundaries, which contributes to phenotypic robustness40. In sum, we do not yet fully understand the physiological significance of CTCF-mediated boundaries in orchestration of gene regulation and ensuring successful embryonic development.
Here, we asked whether perturbation of CTCF-mediated insulation in a chromatin domain with multiple developmental regulators can result in more acute gene dysregulation and severe developmental phenotypes. To test this, we targeted a chromatin domain that harbors three FGF ligand genes—Fgf3, Fgf4, and Fgf15—separated from each other by clusters of CTCF motifs. Like other members of this large family of signaling ligands41, these genes have distinct expression patterns and developmental functions (Figure S1). We show that disruption of CTCF-mediated insulation can result in remarkable and unanticipated developmental phenotypes with full penetrance even in heterozygosity. In fact, loss of CTCF binding at a single motif within a large cluster was enough to severely dysregulate gene expression and impair mouse development. Importantly, we did not observe such a strong phenotypic impact from perturbing other CTCF clusters of this chromatin domain, and many developmental processes controlled by these FGF ligands were not affected in any of our mutants. Our work emphasizes how development and gene regulation can be completely intolerant of sequence variants at specific chromatin boundaries.
Results
Deletion of CTCF clusters in a chromatin domain with three FGF genes disrupts mouse development
Fgf3, Fgf4 and Fgf15 reside within a 63kb region of the mouse chromosome 7 and are separated from each other by clusters of CTCF motifs (Figure 1A). Upstream of Fgf3, there is a 23.9kb cluster with four CTCF motifs that we named C1-C4. The 6kb C5-C6 cluster is located between Fgf3 and Fgf4, and immediately upstream of Fgf15, the C7-C10 cluster occupies 8.5kb. The C11-C14 cluster separates Fgf15 from Lto1 and Ccnd1. Each cluster has CTCF motifs in both forward and reverse orientations and ChIP-seq in mouse embryonic stem (ES) cells confirmed co-occupancy of the cohesin subunit RAD21 at all CTCF sites except C8, C11 and C13. Capture-HiC (CHi-C) in ES cells showed that the three FGF genes reside within a highly interacting domain delimited by C1-C4 on its centromeric end. (Figure 1A and S2A). The C5-C6 cluster overlaps a weak boundary between Fgf3 and Fgf4, and C7-C10 also does not completely insulate Fgf15 from Fgf3 and Fgf4 as some interactions can be observed across this cluster. In contrast, Fgf15 is well insulated from downstream genes including Lto1 and Ccnd1, in a boundary delimited by the C11-C14 cluster that marks the telomeric end of this domain.
Figure 1. Deletion of CTCF clusters in a chromatin domain with three FGF genes causes embryonic lethality.
A CHi-C heatmap at 2kb resolution and ChIP-seq of RAD21 and CTCF in wt mES cells with CTCF motif orientation and insulation score. B Expression of Fgf3 and Fgf4 in single cells labelled as Epi or PrE in E4.5 blastocysts (left) as quantified in endoderm-explorer.com. Representative IF of E4.5 blastocysts: Trophectoderm-CDX2; PrE-GATA6; Epi-NANOG. C Expression of the 3 FGF genes in E9.5 mouse embryos labeled by HCR (n=3/3). AF-Anterior forebrain, BA-Branchial Arches, CE-Caudal end, Mb-Midbrain, OC-Otic cup, Tr-Trunk D Viability of the three mouse lines with CTCF cluster deletions. All founders with at least one copy of the C1-C4Δ died perinatally. Pups generated from heterozygous crosses of the C5-C6Δ and C7-C10Δ mouse lines were assessed at weaning E Representative image of C1-C4Δ founder pups with encephalocele and orofacial cleft (left) (n=6/6). H&E staining on a paraffin frontal section of a C1-C4Δ pup. F Brain and heart defects seen in C7-C10Δ/Δ homozygotes using Micro-CT scans. Blue arrow points to fused brain ventricles with edema and yellow arrow to enlarged heart ventricles (n=6/7).
These FGF genes exhibit diverse tissue-specific expression patterns and functions. Fgf4 is essential for cell fate specification in the blastocyst, and successful post-implantation development42,43. At embryonic day 4.5 (E4.5), publicly available single-cell RNA-seq data44 show that Fgf3 is expressed in cells classified as primitive endoderm (PrE) of blastocysts, while Fgf4 is specifically found in epiblast (Epi) cells (Figure 1B). At mid-gestation (E9.5), in situ hybridization chain reaction (HCR) analysis demonstrated that the FGF genes have not only tissue-specific expression patterns but also overlapping ones. For example, the branchial arches expressed all three FGF ligands but higher levels of Fgf3 and Fgf4. Combined loss of Fgf3 and Fgf4 in the branchial arches causes lethality45. Fgf3 is the only gene out of the three that is highly expressed in the otic cup and contributes to patterning of the inner ear46 (Figure 1C). Fgf15 is strongly expressed in the midbrain and anterior forebrain, where it is required for normal development, as well as the branchial arches where it acts on the cardiac neural crest47,48. In the trunk, both Fgf3 and Fgf4 are highly expressed towards the caudal end, while Fgf15 is found more proximally in a non-overlapping domain.
Since the three FGF genes achieve different patterns of tissue-specific expression (Figure S1) despite their genomic proximity, we hypothesized that the CTCF clusters separating them are required to insulate each other’s regulatory elements. To test this, we generated mouse lines where we deleted the CTCF clusters that are found between FGF genes (C5-C6 and C7–10). Furthermore, to test if perturbing the structure of a domain with multiple developmental regulators would cause severe phenotypes, we decided to target one of the clusters that marks a boundary of this domain. We targeted C1-C4 because we were more confident this would fully disrupt a domain boundary as all CTCF loops in this domain use motifs on this cluster as the centromeric anchor, while on the telomeric end loops are formed with motifs in C7-C10, C11-C14, and over Ccnd1. These three mouse lines were generated by injecting zygotes with Cas9 and gRNAs fully surrounding each cluster to ensure complete loss of cohesin retention and disruption of insulation (Figure S2B). All founder mice with an allele where the C1-C4 cluster was deleted (C1-C4Δ) were found dead at birth, despite repeated injection attempts. In contrast, we successfully established mouse lines with deletions of the C5-C6 and C7-C10 clusters (C5-C6Δ and C7-C10Δ, respectively). Crosses between F1 C5-C6Δ/+ heterozygous animals resulted in recovery of homozygotes at the expected Mendelian ratio (Figure 1D). Strikingly, almost all C7-C10Δ/Δ homozygotes died perinatally. To understand how individual CTCF cluster deletions affected viability and development, we investigated the cause of lethality. C1-C4Δ founder mice exhibited craniofacial abnormalities with orofacial clefts (Figure 1E). H&E staining of frontal sections of the head demonstrated the presence of encephalocele, a neural tube closure defect where neural tissues protrude outside the skull and are covered with an epithelial membrane. Postnatal lethality of C7-C10Δ/Δ homozygotes could not be attributed to specific developmental abnormalities but Micro-CT scan of E18.5 fetuses revealed that homozygous mutants had a range of phenotypes including occluded and reduced brain vesicles as well as enlarged heart ventricles (Figure 1F). In sum, unlike disruption of CTCF-mediated chromatin boundaries at other loci19,35,37,40, deletion of CTCF clusters in a chromatin domain with multiple developmental regulators severely impacted animal development.
Cell type-specific expression of FGF genes in blastocysts does not rely on CTCF-mediated insulation
Although Fgf3 and Fgf4 are close genomic neighbors, they are divergently expressed in mouse blastocysts. Fgf3 is found exclusively in the PrE, while Fgf4 is an Epi marker (Figure 1B). The chromatin environment of these genes also displays contrasting patterns. ES cells, which are an Epi in vitro model, are enriched with H3K27ac at Fgf4—including on the 3’ UTR where sequences that can drive Epi expression have been identified49—while Fgf3 is decorated with the repressive mark H3K27me3. In stark contrast, in vitro models of the PrE (XEN cells) have the exact opposite pattern of chromatin marks at these two genes (Fig S3A). Because tight regulation of FGF4 levels is essential for post-implantation development, we expected that deletion of the CTCF clusters on either side of Fgf4—in the C5-C6Δ and C7-C10Δ lines—would affect expression of the FGF ligands and perturb implantation. However, all mutant mice successfully implanted (Figure 1D). We therefore decided to focus on this early developmental stage to better understand the mechanisms underlying phenotypic resilience to perturbation of CTCF boundaries in some developmental contexts. We first assessed if deletion of the C5-C6 and C7-C10 clusters causes ectopic activation of the FGF genes located across each deleted boundary using RNA-seq in single blastocysts. Surprisingly, we did not detect any of the three FGF ligand genes in either C5-C6Δ/Δ or C7-C10Δ/Δ homozygotes to be differentially expressed (false discovery rate (FDR) <0.1 and log2FC>1) (Figure 2A). This suggests that the enhancers that control expression of Fgf3 and Fgf4 in blastocysts, do not rely on the CTCF clusters to achieve specificity of target gene induction.
Figure 2. Tissue-specific expression of Fgf3 and Fgf4 in blastocysts does not require CTCF-mediated insulation.
A Expression of FGF genes measured by RNA-seq in C5-C6Δ/Δ and C7-C10Δ/Δ in blastocysts compared to WT littermates. Each circle represents an embryo, and the bar represents the median value of each genotype. B CHi-C heatmap of C5-C6Δ/Δ and C7-C10Δ/Δ homozygous mES cells, compared to WT at 2kb resolution. Arrowheads – increased interactions in CTCF cluster mutants C Differential CHi-C interaction frequency heatmap. Red signal represents interactions that occur at higher frequency in mutant cells compared to control, and blue are interactions at lower frequency. D RCMC heatmap in C5-C6Δ/Δ homozygous ES cells compared to WT at 400bp resolution. Virtual viewpoints centered on the Fgf4 promoter and the C10 CTCF motif are shown at 50bp resolution. Arrow - interaction between C5 and the C7-C10 cluster. Arrowheads - increased interactions in C5-C6Δ/Δ. Bracket - increased interactions from the Fgf4 viewpoint. E CUT&RUN H3K27ac data in mES cells of three different genotypes.
Since we did not observe changes in gene expression in these three genes, we first verified whether the C5-C6 and C7-C10 clusters can function as CTCF loop anchors capable of insulating interactions. This was especially relevant for the C5-C6 cluster, as wild-type ES cells do not display strong insulation at this region (Figure 1A). Since the low cell numbers of blastocysts impair high-resolution chromosome conformation capture techniques, we used ES cells as proxies for the Epi state. We derived ES cell lines from C5-C6Δ/Δ and C7-C10Δ/Δ homozygous blastocysts in parallel with wild-type littermates and performed CHi-C (Figure 2B). Loss of CTCF motifs in C5-C6Δ/Δ cells led to increased focal interactions between the upstream cluster C1-C4 with C7-C10 (black arrowhead in Figure 2B). Similarly, deletion of motifs in C7-C10Δ/Δ cells enhanced contacts between C1-C4 and the CTCF cluster downstream of Fgf15 (white arrowhead in Figure 2B). Differential interaction frequency heatmaps highlight the decrease in insulation and increase in contacts across the deleted clusters (Figure 2C). In a complementary approach, we differentiated C5-C6Δ/Δ ES cells and matched wild-type controls into XEN cells that express Fgf3. XEN cells also lost insulation between chromatin domains and increased contacts between neighboring CTCF clusters (Figure S3B; S3C). Together, these data demonstrate that both C5-C6 and C7-C10 CTCF clusters function as loop anchors that can insulate interactions between chromatin domains.
To identify potential ectopic enhancer–promoter interactions caused by loss of CTCF-mediated insulation we generated Region Capture Micro-C (RCMC) contact maps that have higher resolution than CHi-C and allow better identification of regulatory interactions (Figure 2D)50. RCMC again confirmed that C5-C6 functions as a chromatin domain boundary. As observed with CHi-C, C5-C6Δ/Δ ES cells had slightly elevated contact frequency between CTCF motifs of the C1-C4 and the C7-C10 cluster (black arrowheads Figure 2D). Furthermore, these high resolution data—using 50bp bins—demonstrate that the C5 motif serves as an anchor in loops with motifs in the C7-C10 cluster (black arrow in WT heatmap Figure 2D). To characterize changes in enhancer activity we complemented RCMC with H3K27ac Cut&Run (Figure 2E). Despite the loss of insulation, contacts between the active Fgf4 promoter and enhancers across this chromatin domain were only slightly increased in C5-C6Δ/Δ cells (bracket in Fgf4 virtual viewpoints in Figure 2D; S3D) and no ectopic enhancer-promoter contacts were found. Furthermore, the pattern of H3K27ac in C5-C6Δ/Δ and C7-C10Δ/Δ cells showed no spreading of active chromatin marks and mostly accumulated proximal to Fgf4. (Figure 2E). In summary, these data suggest that blastocyst expression of Fgf3 and Fgf4 is mostly driven by extremely target-specific proximal enhancers independently of CTCF, which may contribute to the resilience of blastocysts to perturbation of chromatin structure during mammalian implantation.
C7-C10Δ/Δ phenotypes likely arise from perturbation of chromatin structure and deletion of enhancers
In contrast to the phenotypic robustness of blastocysts to perturbation of chromatin structure, loss of CTCF binding in the C1-C4 and C7-C10 clusters was remarkably impactful later in development. We first asked whether cluster deletions affected only CTCF binding or if phenotypes could be caused by loss of other types of regulatory elements. We first tried to understand the cause of the heart and brain defects of C7-C10Δ/Δ homozygotes (Figure 1F). For this, we assessed expression of Fgf3, Fgf4 and Fgf15 at E9.5 using HCR and saw that expression of Fgf15 in the brain, and of all three FGF genes in the branchial arches was severely reduced (Figure S4A). Due to the physical proximity of the C7-C10Δ deletion to Fgf15, we wondered if putative enhancer elements could have inadvertently been deleted in this allele. Analysis of publicly available ChIP-seq and ATAC-seq in wild-type developing brain revealed enrichment of H3K27ac and accessible chromatin regions that partially overlapped with the C7-C10 deletion (Figure S4B). Furthermore, this region contains sequences that recapitulate some of the expression patterns of Fgf1551. Along with the loss of Fgf3, Fgf4, and Fgf15 expression in the brain and branchial arches, this strongly suggests that the phenotypes seen in C7-C10Δ/Δ homozygotes may not solely be caused by loss of CTCF binding and could also be due to deletion of enhancers. Importantly, Fgf15 expression was not disrupted in all tissues. In fact, C7-C10Δ/Δ homozygotes showed ectopic expression of Fgf15 in the apical ectodermal ridge of forelimbs (Figure S4C), which confirmed that the Fgf15 promoter was not disrupted in the C7-C10Δ deletion. This ectopic expression could be driven by an Fgf4 enhancer located in its 3’UTR known to drive AER expression52, and that may gain access to Fgf15 in C7-C10Δ mutants. Unlike C7-C10Δ/Δ pups that die at birth, animals without Fgf15 survive until weaning47,48. Since the defects we report here are more severe than those described for the Fgf15 knock out, it is likely that the developmental defects of C7-C10Δ/Δ mice are caused by a combination of insulation loss and deletion of enhancers within the C7-C10 cluster. Because distinguishing between these two effects would be very challenging, we switched our focus to the CTCF cluster deletion at the centromeric end of this domain (C1-C4Δ).
Loss of CTCF binding at the C1-C4 cluster causes perinatal heterozygous lethality
On the centromeric end of the domain, the heterozygous lethality of pups with the C1-C4Δ allele, created an obstacle to study the function of this cluster in animal development. To overcome this and assess whether the phenotypes observed in mice with the C1-C4Δ allele resulted from loss of CTCF binding, we attempted a genetic rescue of the embryonic defects (Figure 3A). We used the same gRNAs that generated the C1-C4Δ allele (Figure S2C) along with a 672bp repair template containing the four CTCF motifs found in the C1-C4 boundary, flanked by loxP sites—an allele we named C1-C4flx. Strikingly, despite deleting 23.9kb, reintroduction of these four CTCF motifs generated C1-C4flx/flx F0 homozygotes that were viable and fertile (Figure 3B). We then bred C1-C4flx/+ heterozygous mice with homozygous E2A-Cret/t animals that ubiquitously express Cre recombinase to generate embryos with heterozygous loss of the C1-C4 cluster (C1-C4flx/+; E2A-Cret)53. These embryos recapitulated the encephalocele and orofacial cleft defects in a fully penetrant manner and died perinatally (Figure 3C; 3D; S4D). The rescue of the severe developmental phenotypes in mice with the floxed CTCF cassette, and the recapitulation of these phenotypes upon Cre recombination provides direct evidence that the defects observed in C1-C4Δ mice are caused by loss of CTCF-binding and not by deletion of other regulatory elements. It also highlights the sensitivity of mouse development to loss of CTCF binding at this domain boundary, since heterozygous deletion of the CTCF cluster is sufficient to cause such strong developmental phenotypes and fully penetrant perinatal lethality.
Figure 3. Heterozygous deletion of the C1-C4 cluster causes lethality because of loss of CTCF binding.
A Strategy used to generate the C1-C4flx allele, by Cre-mediated recombination. Green lines represent regions of homology in the repair template with 120bp from each of the four CTCF motifs of the C1-C4 boundary flanked by loxP sites (gray triangles). Upon Cre-mediated recombination the four CTCF motifs are lost and a single loxP motif is left behind B Pups from crosses of C1-C4flx/+ heterozygotes were recovered at expected mendelian ratio. C Embryos and fetuses from the breeding of heterozygous C1-C4flx/+ with E2A-Cret/t homozygotes (n=6/6). D Skeletal preparations of the same crosses show a defect in parietal bone development (n=3/3).
Encephalocele is a severe birth defect found in one out of 10,500 human births54, but its etiology is poorly understood in part because existing mouse models recapitulate the features that characterize the human disease with low penetrance55. To improve our understanding of this phenotype, we examined C1-C4flx/+; E2A-Cret heterozygous embryos at E12.5 and E16.5. This revealed an accumulation of fluid in mutant heads, over-expansion of neural tissue (Figure 3C), and abnormal parietal bone development (Figure 3D). It was unclear if the bone pathology was caused by secondary effects of brain herniation, ossification defects or anomaly in skull bone fusion. Because craniofacial phenotypes are often caused by defects in neural crest cells, we next investigated if defects in this cell type contributed to the phenotypes. To discern the origin of the phenotypes, we crossed the C1-C4flx/+ heterozygotes with either Wnt1-Cre or Sox10-Cre mice. The Wnt1-Cre line expresses Cre throughout the neural tube including neural crest cells, while the Sox10-Cre line is restricted to migratory neural crest cells56. All double heterozygotes, C1-C4flx/+; Wnt1- Cret showed perinatal lethality and recapitulated both encephalocele and cleft defects (Figure S4E). Conversely, none of these phenotypes were observed in C1-C4flx/+; Sox10-Cret progeny (Figure S4F). These data strongly suggest that the craniofacial phenotypes do not originate from migratory neural crest cells.
Disruption of the C1-C4 boundary leads to ectopic expression of FGF genes in the brain
To understand the molecular mechanisms that caused pronounced phenotypes when CTCF binding was lost at the C1-C4 cluster, we bred C1-C4flx mice with ROSA26-CreERT2 mice. This line ubiquitously expresses the Cre recombinase with its activity controlled by tamoxifen treatment. C1-C4flx/flx homozygous females with one copy of the CreERT2 allele were then mated with C1-C4flx/flx homozygous males and tamoxifen was orally administered to the females at E6.5 (see breeding scheme in Figure S5A). This generated embryos that carried the intact C1-C4flx allele in both chromosomes (referred to as C1-C4flx/flx) as well as littermates where the C1-C4 rescue cassette was deleted in homozygosity (referred to as C1-C4flx/flx; Cre). As the midbrain was visibly enlarged at E12.5 (Figure 3B), we collected this brain region at E11.5 when defects are not yet obvious. Subsequently, we performed CHi-C on each genotype and reads were aligned to a custom mm10 genome where the 23.9kb C1-C4 cluster was replaced by the 672bp C1-C4flx cassette. This confirmed that the C1-C4flx/flx rescue cassette can function as an anchor capable of establishing loops with both upstream and downstream CTCF clusters (white arrowheads in Figure 4A). Littermates with Cre-induced C1-C4 deletion showed complete loss of upstream and downstream loops, with increased interactions from flanking CTCF sites on both ends in the absence of the C1-C4 anchor (black arrowheads in Figure 4A). Importantly, the loss of insulation caused fusion of the domain with the three FGF genes with the upstream domain containing the Ano1 gene.
Figure 4. Disruption of the C1-C4 boundary leads to ectopic expression of FGF genes in the brain.
A Differential CHi-C heatmap at 4kb resolution between C1-C4flx/flx;CRE and C1-C4flx/flx E11.5 midbrains. Red signal represents higher frequency of contacts in mutants compared to control and blue shows lower frequency (top). CHi-C 1D heatmap in C1-C4flx/flx;CRE and C1-C4flx/flx midbrains (bottom). Data were mapped to a custom genome. Arrowheads - increased focal interactions. B E9.5 HCRs in the midbrain and anterior forebrain (n=4). C Expression of the three FGF genes and Ano1 in micro-dissected midbrain and anterior forebrain at E11.5 measured by RNA-seq. Each circle is one embryo, and bar represents the median value of each genotype. Scalebars represent 200μm.
Next, we examined the expression of genes flanking the C1-C4 cluster at E9.5 using HCR. This stage was chosen to facilitate imaging. Loss of the C1-C4 rescue cassette strongly increased Fgf3, Fgf4 and Fgf15 expression in the midbrain and anterior forebrain (Figure 4B). Interestingly, all three FGF genes in mutant brains recapitulated the expression pattern of Ano1, a gene that encodes a chloride ion channel located in the upstream domain. We then bred homozygous C1-C4flx/flx males with homozygous E2A-Cre females to generate embryos with heterozygous loss of the C1-C4 boundary and control littermates. Importantly, these heterozygous embryos also showed the striking upregulation of the FGF genes (Figure S5B). To better quantify changes in expression of the FGF genes and match expression data with CHi-C, we used RNA-seq on micro-dissected E11.5 midbrains and anterior forebrains. As indicated by HCR, Fgf4 exhibited the most dramatic upregulation (14.9-fold in the forebrain and 34.3-fold in the midbrain). Fgf3 was also strongly upregulated in both brain regions (above four-fold), while Fgf15 was only slightly upregulated in the midbrain (Figure 4C). Interestingly, despite massive induction of Fgf3 and Fgf4 levels in the anterior forebrain, very few other genes were dysregulated at this stage (Figure S5C). In contrast, the midbrain displayed more widespread dysregulation, which may explain the early neural expansion of the midbrain. In summary, our data revealed that deletion of the C1-C4 boundary fused the Ano1 and FGF domains and led to strong ectopic expression of the FGF genes in the brain, which may result in expanded neural tissue and encephalocele. Supporting our hypothesis that increased FGF activity affects skull development, a previous study revealed that induction of Fgf3 and Fgf4 driven by random insertion of viral elements can cause craniofacial dysmorphology57.
Loss of CTCF-mediated insulation exposes FGF genes to distal brain enhancers of Ano1
The strong upregulation of the FGF genes and recapitulation of the brain Ano1 expression pattern suggest that loss of the C1-C4 boundary causes ectopic contacts with enhancers located in the domain that harbors the Ano1 gene. To identify candidate regulatory elements, we analyzed histone modification patterns in mouse wild-type midbrains at E11.5. The intronic region of Ano1 showed two distinct H3K27ac peaks suggestive of putative active enhancers that may drive its expression in the brain (Figure 5A, bottom panel). In line with its expression, Fgf15 also had H3K27ac enrichment in the midbrain. Furthermore, we saw strong enrichment of the polycomb-deposited repressive mark H3K27me3 at Fgf3 and Fgf4 in wild-type brain, which likely contributes to their transcriptional silencing. We hypothesized that loss of insulation in embryos without the C1-C4 cluster could induce ectopic contacts of the FGF genes with these distal Ano1 brain enhancers. To test this, we micro-dissected midbrains at E11.5 and performed RCMC to map interactions at finer resolution than CHi-C and increase the chances of detecting enhancer–promoter interactions. As with CHi-C data, we observed loss of both upstream and downstream CTCF-anchored loops (white arrowheads in Figure 5A) due to deletion of the C1-C4flx rescue cassette, and an increase in contacts between CTCF motifs of the clusters surrounding C1-C4. Surprisingly, we detected an interaction between Fgf3 and the most centromeric of the Ano1 intronic enhancers (black arrowhead in Figure 5A) both in embryos where the C1-C4flx/flx cassette was deleted, as well as littermates where the CTCF motifs were intact. Fgf15, which has a very similar brain expression pattern to Ano1, also interacted with the Ano1 brain enhancers. To better quantify changes in interaction frequency between the FGF promoters and the Ano1 enhancers, we performed promoter capture Micro-C using the three FGF promoters as viewpoints (Figure 5B; S6A). This strategy confirmed that deletion of the C1-C4 cluster resulted in loss of interactions at the CTCF boundary generated by the C1-C4flx rescue cassette (gray highlighted region on the right). There was also a sharp increase in long-range interactions of the Fgf3 and Fgf15 promoters specifically with the most centromeric of the Ano1 intronic brain enhancers (highlighted region on the left). Rather than increased interactions specifically at this enhancer, Fgf4 showed higher contact frequency across the entire Ano1 domain (see subtraction track in Figure 5B where the signal from C1-C4flx/flx cells was subtracted from C1-C4flx/flx;CRE). In summary, loss of CTCF-mediated insulation at the C1-C4 boundary promotes physical interaction of FGF promoters with putative Ano1 brain enhancers.
Figure 5. Loss of CTCF-mediated insulation exposes FGF genes to distal brain enhancers of Ano1.
A Region capture micro-c (RCMC) heatmap at 400bp resolution of C1-C4flx/flx;CRE and C1-C4flx/flx micro E11.5 midbrains. White arrowhead - loops from C1-C4flx rescue cassette. Black arrowheads - interactions between the most centromeric Ano1 brain enhancer with Fgf3 and Fgf15. Data were mapped to custom genome. Gray highlight shows putative Ano1 brain enhancer B Promoter Capture Micro-C shown at 50bp resolution for the Fgf3 and Fgf4 promoters. Gray highlight shows interactions between the FGF promoters and the C1-C4flx rescue cassette or with the putative Ano1 brain enhancer. C Alleles to analyze phenotypes and gene expression at E13.5. D Representative images of embryos (n= 21/21, 3/3, 7/7, 3/3). Asterisk indicates encephalocele. Scalebars represent 1mm. E qPCR measurement of Fgf3 and Fgf4 in E13.5 dissected midbrains using the ΔΔCT method and Gapdh as a reference. Each circle is one embryo and the bar the median value for each genotype. Significance assessed with Wilcoxon two-sided test.
We then did another genetic rescue experiment to prove that induction of the FGF genes—triggered by ectopic contacts with the distal putative Ano1 brain enhancers (ABEs)—results in encephalocele. Specifically, we hypothesized that extending the 23.9kb deletion of the C1-C4 cluster to include these candidate intronic regulatory elements, would prevent the ectopic activation of the FGF genes—despite boundary deletion—and rescue the encephalocele phenotype. To test this, we electroporated zygotes with Cas9 and gRNAs designed to generate such deletion and analyzed phenotypes and gene expression directly in E13.5 founder embryos (Figure 5C). As a positive control, we also re-targeted the CTCF cluster C1-C4 which, as expected, resulted in encephalocele in embryos carrying the C1-C4Δ allele (Figure 5D). Strikingly, all embryos with the larger 175.6kb deletion allele—that we named ABEs-C4Δ—displayed no overt phenotype at E13.5. We also screened for inversions where the targeted region in ABEs-C4Δ was inverted and repositioned the ABEs closer to FGF genes, putting the CTCF boundary further away towards the centromeric side. In contrast to embryos where the enhancers were deleted, the encephalocele phenotype was clearly visible in all ABEs-C4inv embryos (Figure 5D). In line with the rescue of the encephalocele phenotype, Fgf3 and Fgf4 levels were significantly reduced in the midbrain of ABEs-C4Δ mutants as compared to wildtype littermates or C1-C4Δ mutants (Figure 5E). ABEs-C4inv mutants showed higher levels of Fgf3, likely because the inversion allele places the ABEs closer to the FGF genes, as observed for other loci58. Together, these data reveal that the loss of the C1-C4 boundary exposes the three FGF genes to brain enhancers of Ano1, which results in strong ectopic brain expression and encephalocele.
Deletion of a single CTCF motif within a large multi-motif cluster can compromise its insulator function
Our data show that the chromatin domain boundary established by the C1-C4 cluster has a remarkable ability to completely insulate the Ano1 brain enhancers from activating the FGF genes in wild-type embryos. This is especially clear for Fgf4 whose expression is almost undetectable before recombination but becomes highly upregulated once the C1-C4 cluster is deleted. To elucidate the contribution of the different types of CTCF motifs within the C1-C4 cluster to insulate the FGF genes, we targeted the boundary by injecting zygotes with Cas9 and gRNAs in different combinations (Figure 6A). The C1 motif is in reverse orientation, pointing to the domain of the Ano1 gene, whereas the C2-C4 motifs are in the forward orientation, towards the domain with the three FGF genes. To understand if all four CTCF motifs are required to segregate the FGF genes from the ABEs, and if their orientation determines different insulation strengths, we generated two deletion mutants—C1Δ and C2-C4Δ (Figure 6B). Surprisingly, all E13.5 embryos with just the C1Δ deletion recapitulated the encephalocele phenotype observed in C1-C4Δ, while C2-C4Δ mutants did not display any phenotype. Notably, the expression levels of Fgf3 and Fgf4 in C1Δ mutants were significantly upregulated and comparable to the full cluster deletion of C1-C4Δ mutants (Figure 6C). In contrast, C2-C4Δ mutant embryos had no statistically significant differences in Fgf3 and Fgf4 midbrain expression compared to wild-type littermates. In summary, these data highlight the importance of insulation by the C1-C4 cluster to ensure that Fgf3 and Fgf4 are not expressed in the midbrain and how loss of a single motif within a multi-motif cluster can compromise its insulator function and result in ectopic expression of these genes. Our data also suggest that CTCF motifs oriented towards the side of active enhancers may be better transcriptional insulators than those oriented to inactive chromatin regions.
Figure 6. Loss of a single motif within C1-C4 recapitulates deletion of entire C1-C4 cluster.
A Alleles to analyze phenotypes and gene expression at E13.5. B Representative images of embryos with each of the alleles (n= 21/21, 3/3, 6/7, 3/3). Scalebars represent 1mm. C qPCR measurement of Fgf3 and Fgf4 in E13.5 dissected midbrains using the ΔΔCT method and Gapdh as a reference. Each circle represents one embryo and the bar the median value for each genotype. Significance assessed with Wilcoxon two-sided test.
Discussion
The discovery that vertebrate genomes fold into chromatin domains delimited by CTCF binding presents an attractive and simple model to understand how the spatial range of enhancers can be restricted to genes within the same domain. However, disruption of these structures can result in minor changes in gene expression, and modest phenotypes36,59,60. This raises doubts on the physiological impact of chromatin domains and their roles in gene regulation. Here, we show that in some developmental contexts, such impact can be dramatic. In the developing murine brain, although Fgf3 and Fgf4 are polycomb targets marked for silencing by H3K27me3 enrichment, loss of their centromeric domain boundary induces strong ectopic upregulation by the intronic enhancers of Ano1, leading to orofacial clefts, encephalocele, and fully penetrant perinatal lethality (Figure S6B). Strikingly, loss of just one CTCF motif, pointing towards the Ano1 enhancers is sufficient to recapitulate these severe phenotypes.
We hypothesized that disruption of domains with multiple developmental regulators may induce particularly striking changes in gene expression and animal development. While this was true when we targeted the centromeric domain boundary and observed prominent defects in brain development, this phenomenon was not manifested in other developmental processes controlled by the FGF genes nor when we targeted the CTCF clusters separating them. For example, we showed that although blastocyst development requires precise control of FGF4 levels, CTCF boundaries around this gene are not necessary for early development. Our data suggest that these different phenotypic outcomes are caused by different enhancer-specificities of the FGF promoters. In the brain, all three FGF genes have strong compatibility with a distal enhancer of Ano1 and therefore CTCF insulation is essential to prevent their ectopic activation. In contrast, in the blastocyst these same genes are remarkably specific to their own proximal enhancers and gene regulation is independent of CTCF-mediated boundaries (Figure S7) As explained below, we propose that different strategies to achieve specificity in enhancer-promoter interactions and gene expression may be driven by how insulation and domains are built during development.
Mammalian fertilization is characterized by sequential re-establishment of chromatin structure. Domain boundaries are initially very permeable and gain insulation strength with each cleavage stage61–64. ES cells, which mimic the epiblast of the blastocyst, also show weaker boundaries than more differentiated cells such as neurons or even neural progenitors65. Perhaps early in development, gene regulation must rely on more stringent promoter specificity that may be better conferred by proximal enhancers. In contrast, once domain boundaries provide better insulation, cell-type specific expression can be driven by more distal enhancers, whose spatial range of gene activation is controlled by CTCF-mediated boundaries. In line with this, depletion of maternal and zygotic CTCF does not cause dramatic dysregulation of gene expression in early blastocysts, providing further support to the hypotheses that early in development, CTCF is less important to ensure enhancer–promoter specificity66.
The chromatin domain that contains Ano1, the three FGF genes, and Ccnd1 is in a syntenic region found across vertebrates. Importantly, clusters of CTCF motifs between these genes are also conserved in most vertebrate species. Interestingly, duplication of this region in the genome of Ridgeback dogs leads to dermoid sinus and causes dysregulation of FGF genes67. In humans, duplication of the chromosomal 11q13 region, which includes FGF3 and FGF4 causes intellectual disability, craniosynostosis and microcephaly68–71. Furthermore, global DNA hypermethylation seen in gastrointestinal stromal tumor (GIST) patients leads to recurrent loss of CTCF binding in the boundary between ANO1 and the FGF genes, and higher levels of FGF activity72. Although it isn’t yet clear how much the structure of the chromatin domain is affected by these different chromosomal rearrangements, these observations suggest that spatial organization and regulatory interactions within this region may be preserved throughout evolution to fulfill essential developmental functions. In contrast to mouse development where loss of a single CTCF motif—if oriented towards an active enhancer—was sufficient to activate ectopic FGF expression, in human GIST cell lines, disruption of all four CTCF motifs found in its cluster is required for induction of FGF3 by ANO1 enhancers73. The difference in sensitivity between the two models may be related to the boundary, or to the enhancers that drive expression of Ano1 in each of these two tissues, i.e. the mouse midbrain, and the human intestine. Nonetheless, our work, together with these observations across multiple vertebrates, suggests that structural perturbation of highly conserved chromosomal domains may be more likely to result in strong gene dysregulation and more severely affect physiological processes74,75.
Recent sequencing efforts have highlighted the high prevalence of sequence variants in the non-coding genome. Our work further emphasizes the importance of whole genome sequencing to consider how even small changes in CTCF binding motifs can mis-wire enhancer–promoter interactions, induce ectopic gene expression, and initiate disease states or perturb development. Our data also provide important insight into potential congenital origins of encephalocele, and it will be important to investigate whether human encephalocele patients present sequence variants that may affect CTCF binding at the boundary between ANO1 and the FGF genes.
Limitations of the study
We have analyzed the impact of perturbing a single chromatin domain. Therefore, we do not know whether disruption of other highly conserved domains with multiple developmental regulators will also result in similarly severe phenotypic consequences. In fact, even at this domain, it will be important to assess whether perturbing its telomeric end causes as severe phenotypes as those we describe in the centromeric side, especially considering the more ubiquitous role of Ccnd1 in cell-cycle regulation. Many different factors are likely to affect physiological outcomes of perturbations, and much work is still needed to understand and predict their impact. Clearly, the function of gene products potentially dysregulated after boundary disruption is an important determining factor of phenotypic impact. As we show here, overexpression of ligands of signaling pathways can easily affect a wide variety of developmental processes in a way that other genes would not, even when ectopically expressed. But the identity of the genes within a TAD, while important, is not enough to predict whether boundary disruption will have a phenotypic impact. In fact, based on the differences we describe between the brain and blastocysts, we believe that the intrinsic ability of cell-type specific enhancers to activate off-target promoters upon boundary loss, likely plays a crucial role in determining the phenotypic outcome of chromatin structure perturbations.
Resource Availability
Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Pedro P. Rocha (pedrorocha@nih.gov)
Materials Availability
Mouse transgenic lines and cell lines generated in this study are available upon request.
Data and Code Availability
This study only used available software. Software parameters used are described in Methods. FASTQ and processed data can be found in GEO under accession number GSE271760. Chromosome conformation capture data can be navigated at: resgen.io/pedrorocha/FGFs/views/. Any additional information required to reanalyze the data reported in this work paper is available from the Lead Contact upon request.
STAR Methods
Experimental model and study participant details
Mice
Transgenic mouse lines were generated at National Cancer Institute, Frederick using super ovulated C57Bl6/6NCr female mice as embryo donors. C57Bl6 and the following Cre lines were purchased from The Jackson Laboratory: E2A-CRE (strain 003724)53, ROSA26-CreERT2 (strain 008463)76, Wnt1-Cre (strain 022501)77, and Sox10-Cre (strain 025807)78. All animals were housed and bred in specific pathogen free animal facility at NIH, under a 12-hour light/dark cycle at 24°C with ad libitum access to food and water. Zygotic injection of Cas9-gRNA ribonucleoproteins was used for mouse line generation, as previously described40. sgRNAs were designed using sgRNA Scorer 2.079 and subsequently tested for editing activity in P19 cells, as previously described80. The best gRNAs were then purchased as synthetically modified RNAs (Synthego). Cas9 protein was generated in house (Protein Expression Lab, Frederick National Lab) using an E. coli expression plasmid obtained as a gift from Niels Geijsen (Addgene #62731)81. Zygotes were collected on day E0.5, microinjected and allowed to recover for 2 hours by incubation (5% CO2, 37°C), following which, viable embryos were surgically transferred to oviducts of pseudopregnant recipient females. The cocktail used for injection comprised of 75 ng/μl of in vitro synthetized gRNAs (Synthego) and 50ng/μl of Cas9 protein in 50μl total volume, kept on dry ice until just prior to microinjection. For insertion of a repair template, 75 ng/μl of single-stranded DNA was added to microinjection cocktail. Founder mice were then bred to C57Bl6 mice from The Jackson Laboratory. Repair templates, gRNAs, primers used for genotyping, and sequencing of ligation junctions of the deletion lines can be found in Table S1. Targeting for analysis of founder F0 E14.5 embryos was done by zygotic electroporation using the Bex zygote genome editing electroporator and the following conditions: 30V, 1ms Pd on, 1000ms pd off. For these experiments, 12μl of 250ng/μl of Cas9 (IDT 1081058), 100ng/μl of gRNA (ordered from IDT).
All procedures were performed according to NIH and PHS guidelines and only after approval by the Animal Care and Use Committees of the National Cancer Institute and Eunice Kennedy Shriver National Institute of Child Health and Human Development. NCI-Frederick, where transgenic mouse lines were made, is accredited by AAALAC International and follows the Public Health Service Policy for the Care and Use of Laboratory Animals. Animal care was provided in accordance with the procedures outlined in the “Guide for Care and Use of Laboratory Animals (National Research Council; 1996; National Academy Press; Washington, D.C.).
Cell lines
Mouse embryonic stem cells lines were derived from blastocysts and genotyped as previously described40. For assays with cells in the Epiblast-like state, cells were grown in serum-free 2i media constituted of Neurobasal medium (Thermo Fisher, 21103049), DMEM/F12 Nutrient mixture (Thermo Fisher,11320033), 1% penicillin—streptomycin (Thermo Fisher, #15140163), 2 mM Glutamax (Thermo Fisher, #35050079), -mercaptoethanol supplemented with N2 (Thermo Fisher, #17502001), B-27 (Thermo Fisher, #17504001), MEK/ERK pathway inhibitor (PD0325901, Reprocell, #04-0006-02), GSK3 signaling inhibitor (CHIR99021, Reprocell, #04-0004-02) and leukemia inhibition factor (LIF, Millipore Sigma, #ESG1107). Before differentiation, ES cells were grown on mouse embryonic fibroblasts (Millipore Sigma, #PMEF-CFL) and serum containing media: Knockout Dulbecco’s Modified Eagle Medium (DMEM) (Thermo Fisher, #10829018) with 15% FBS (VWR, #97068-091), 2 mM Glutamax (Thermo Fisher, #35050079), 0.1 mM 2-mercaptoethanol, 0.1 mM MEM non-essential amino acids (Thermo Fisher, #11140050), 1 mM sodium pyruvate (Thermo Fisher, #11360070), 1% penicillin-streptomycin (Thermo Fisher, #15140163) and Recombinant mouse LIF (Sigma, #ESG1107). Differentiation into the primitive endoderm-like state (XEN cells) was done as previously described82. Briefly, ES cell lines were first grown on mouse embryonic fibroblasts, enzymatically passaged with 0.05% trypsin (Thermo Fisher, #25300054) and feeder depleted for 45 mins. Cells were washed twice with standard XEN media: advanced RPMI (Thermo Fisher, #12633012) with 15% FBS (VWR, #97068-091), 1% penicillin—streptomycin (Thermo Fisher, #15140163) and 0.1 mM β-mercaptoethanol to remove residual LIF. Cells were counted, 96000 cells were plated in gelatinized 6-well plates and grown overnight in standard XEN media. After 24 hours, media was replaced with standard XEN media supplemented with 0.01μM retinoic acid (Millipore Sigma, #R2625) dissolved in DMSO, 10ng/ml activin (R&D Systems, #338-AC), 24ng/ml recombinant FGF2 (R&D Systems, # 3139-FB) and 1μg/ml heparin. Cells were grown in presence of supplements for 2–3 days depending on morphology and confluency and media was replenished every 24 hours. Cells were passaged and grown for 7–10 days until XEN like colonies with stellate and refractile morphologies emerged. XEN-like colonies were scraped and picked with a pipette under the microscope and plated in gelatinized plates to enrich their populations and cultured for an additional 2 weeks until cell morphology was homogenous.
Method Details
Skeletal stainings
Skeletal analyses were performed using Alcian blue and Alizarin red staining method as described previously83. Briefly, fetuses were obtained at E18.5 by caesarean section, eviscerated and soaked in water for 2–4 hours. After a one-minute heat shock at 65°C fetuses were skinned and fixed in 100% ethanol for 2 days. Fixed fetuses were incubated with 150mg/l Alcian blue in 80% ethanol and 20% acetic acid for 12 hours followed by overnight incubation in 100% ethanol. Fetuses were treated with 2% KOH for 6 hours and stained with 50mg/l Alizarin red in 2% KOH for 3 hours and cleared again in 2% KOH for 12– 20 hours. Stained and cleared fetuses were stored in 25% glycerol in water before imaging.
Micro-CT scans
This was done by the Mouse Biology Program, University of California, Davis, CA, United States as previously described84. E18.5 embryos were incubated in a hydrogel stabilizing solution (4% PFA, 4% acrylamide, 0.05% bis-acrylamide, 0.25% VA044 Initiator, 0.05% saponin in PBS) for three days at 4°C to preserve tissue integrity. Thereafter, the vials containing the embryos were placed in a desiccation chamber and saturated with nitrogen gas to replace the air and embryos were incubated in a 37°C water bath. Finally, embryos were removed from the encasing hydrogel, swiped clean and immersed in a Lugol solution [0.7% iodine solution (0.1N)] for at least 24 hours at room temperature while rocking and then oriented and embedded in 1% agarose and oriented for μCT imaging. Mouse embryos imaged at the Center for Molecular and Genomic Imaging (UC Davis) with high-resolution X-ray CT. Three embryos were embedded stacked in agar in a standard test tube that fit the embryos tightly and the conical vial was glued to a sample base for imaging. Embryos were imaged using a high resolution MicroXCT-200 specimen CT scanner (Carl Zeiss X-ray Microscopy). The CT scanner has a variable x-ray source capable of a voltage range of 20–90kV with 1–8W of power. Embryos were placed on the scanner’s sample stage, which has a submicron level of position adjustments. Scan parameters were adjusted based on the manufacturers recommended guidelines. The systems 0.4x detector was used for imaging. The source and detector distances were adjusted based so that the Detector-RA and Source-RA distances were 122.5mm and 25mm respectively. Once the source and detector settings were established, the optimal x-ray filtration was determined by selecting among one of 12 proprietary filters: LE3 was the filtration selected. Following this procedure, the optimal voltage and power settings were determined for optimal contrast (40kV and 200microAmp). 1,600 image projections were obtained over a 360-degree rotation. The camera pixels were binned by two to increase signal to noise in the image and the source-detector configuration resulted in a voxel size of 11.4791 microns. Images were reconstructed with a smoothing factor of 0.7 and a beam hardening of 0.2 into 16-bit values with common global minimum and maximum values for proper histogram matching.
RNA-seq
RNA from the developing brain was isolated using trizol reagent. After confirming that the RNA integrity number for each sample was above 8, libraries were prepared using TruSeq Stranded mRNA prep kit with PolyA purification and sequenced on Novaseq6000. For RNA-seq in single blastocysts, DNA and RNA was isolated from each embryo using Dynabeads mRNA DIRECT Purification Kit (Thermo Fisher, 61012) as previously described85. Blastocysts at E4.5 were flushed out of the uterine horns from super-ovulated females using M2 medium and transferred into 50 μl of pre-warmed lysis buffer (100mM Tris-HCl pH-7.5, 500mM LiCl, 10mM EDTA pH-8, 1%LiDS, 5mM DTT) in DNA low binding PCR tubes. Lysed embryos were stored in −20C and processed within a few weeks. Dynabeads Oligo(dT)25 mRNA isolation beads were warmed to room temperature for 30 mins and rinsed with 100 μl of lysis buffer by vortexing continuously for 5 mins. 10 μl of bead suspension was used per embryo lysate. The poly-A tail of mRNA was allowed to anneal to beads by mixing the embryo lysates and bead suspension in a vortexer for 5 mins at low speed, followed by 5 mins incubation without shaking. Tubes were briefly spun and placed in a magnetic stand to collect the clear supernatant, which was used for DNA isolation using 2X SPRI beads to determine the genotype of each embryo. mRNA bead complexes were washed twice using Wash Buffer A and twice with Wash Buffer B by vortexing for 5 mins each. cDNA and library prep were then processed using an adaptation of the smartseq2 protocol86–88. Buffers were removed and beads annealed with mRNA was resuspended in 12.5μl of RNA suspension mix containing 20U/μl SUPERase-In RNase inhibitor (Thermo Fisher, #AM2694), 10mM each dNTP Mix (New England Biolabs, #N0447L) and 100μM polydT oligonucleotide primer (custom sequence ‘AGACGTGTGCTCTTCCGATCTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTVN’ synthesized by Integrated DNA Technologies). Bead suspension is heated at 75°C for 5mins to denature the RNA and supernatant is transferred into a new RNase free 0.2ml PCR tube. To each sample tube 7.5μl of first strand reverse transcription mix with 100mM DTT, 5X SuperScript IV Buffer, SuperScript™ IV Reverse Transcriptase (Thermo Fisher, #18090050), 20U/μl SUPERase-In RNase inhibitor (Thermo Fisher, #AM2694), 100μM of template switching oligo with 3G’s and an adaptor sequence same as polydT oligonucleotide (custom sequence ‘AGACGTGTGCTCTTCCGATCTNNNNNrGrGrG’, Integrated DNA Technologies) was added and incubated at 50°C for 60 mins to prepare the first strand of cDNA, followed by 85°C for 5mins to inactivate the enzyme. The first strand of cDNA is bound by the template switching oligo and a complementary sequence to the template switching oligo was synthesized, where a PCR oligo is hybridized (custom sequence ‘AGACGTGTGCTCTTCCGATCT’, Integrated DNA Technologies) and amplified. PCR cDNA amplification was done for 16 cycles (98°C for 15 secs, 67°C for 20 secs and 72°C for 10 mins) in a thermal cycler with KAPA Hifi HotStart ReadyMix (Roche, #9420398001). Pre-amplification cDNA mix was purified with 0.8X SPRI beads and quality of cDNA was assessed with a D5000 Screen Tape assay (Agilent, #5067–5588) using Agilent 4150 Tapestation. cDNA mix was diluted to 0.2ng/μl and used for library preparation as described earlier88 using Nextera XT DNA Library Prep Kit (Illumina, #FC-131–1024). Briefly, in a 384 well plate, 1 μl of cDNA, 2μl of Tagment DNA Buffer and 1μl of Amplicon Tagment Mix was mixed in ice and tagmentation was done in a thermocycler at 55°C for 10 mins. Reaction was neutralized with 1 μl of Neutralize Tagment Buffer, incubated at room temperature for 5 mins and transferred to ice. Amplification PCR was set up with 3μl of Nextera PCR Mastermix and 2μl of each adaptor from Illumina DNA/RNA UD Indexes Set A (Illumina, #20027213) per sample. Tagmented DNA was incubated at 72°C for 3 mins, 95°C for 30 secs and 12 cycles of 95°C for 10 secs, 55°C for 30 secs, 72°C for 1 min and a final extension at 72°C for 5 mins. Library DNA was purified with 0.9X SPRI and quality was assessed using D1000 Screen Tape assay (Agilent, #5067–5585) before sequencing.
RNA-seq analyses
RNA-seq analysis for brain was performed on 50-bp paired-end raw sequence reads using lcdb-wf v1.9rc [github.com/lcdb/lcdb-wf]. Adapters were trimmed, along with light quality trimming, using cutadapt v3.4 with parameters “-a AGATCGGAAGAGCACACGTCTGAACTCCAGTCA -q 20 --minimum-length = 25”89. Sequencing quality was assessed with fastQC v0.11.9 (bioinformatics.babraham.ac.uk/projects/fastqc/) and evaluation for common sequencing contaminants was performed using fastqc_screen v0.14.0 with parameters “--subset 100000 --aligner bowtie2”. No significant quality issues were detected. Trimmed reads were provided to Salmon v1.10.190 for quantification. The Salmon index was built from a transcriptome fasta file created using the GENCODE vM18 basic annotation and the GENCODE vM18 genome fasta file, using the gffread package91. TPM (transcripts per million) values were imported into DESeq2 v1.22.192 using the tximport function from the tximport v1.22.0 package93, using default parameters and providing a transcript-to-gene mapping built from the GENCODE GTF to ensure complete and accurate summing to the gene level. A separate DESeqDataSet object was created for each brain region and the model included a blocking factor for sex using the model ~sex + genotype. The genotype contrast was extracted followed by log2 fold change shrinkage using the ‘normal’ method. The analysis used DESeq2’s default of considering a gene differentially expressed if the false discovery rate (FDR) of the gene was <0.1. An absolute log2 fold change threshold of one was applied. Count data were normalized by DESeq2-calculated sizeFactors for visualization purposes. RNA-seq analysis for blastocysts was similar to above except that in the upstream processing, the cutadapt parameters were changed to “-a CTGTCTCTTATACACATCT” and “-A CTGTCTCTTATACACATCT” to reflect the different library prep method. In DESeq2 a separate DESeqDataSet object was created for each deletion (along with matched control) using the model ~genotype and the genotype contrast was extracted.
Hybridization Chain Reaction (HCR)
HCR was performed as previously described94. Briefly, embryos were fixed in 4% PFA overnight at 4°C while rocking, then gradually dehydrated into 100% methanol where they can be stored indefinitely. Before beginning HCR, embryos are rehydrated and washed into PBT (PBS + 0.1% Tween), bleached with 6% hydrogen peroxide in PBS for 30 minutes, then digested with 10μg/mL ProK (Sigma 3115836001) in PBT for 12 minutes. Digestion was stopped by briefly washing with PBT then post-fixing in 4% PFA for 20 minutes at room temperature. Embryos are equilibrated in hybridization buffer at 37°C then hybridized with overnight at 37°C with probes for Fgf3, Fgf4, Fgf15, and Ano1, with the initiators B3, B4, B1, and B2 respectively. Hairpins were diluted in AMP buffer and rocked overnight at RT. Hairpins used for each initiator were: B3–546, B4–647, B1–488, B2–750. Embryos were then washed in PBT then PBTx (PBS + 0.1% TritonX-100) with DAPI for 36–48 hours at RT. Probes, hairpins, hybridization buffer, probe wash buffer, and amplification buffer were purchased from Molecular Instruments. Embryos were mounted in 1% ultra-low-melt agarose (Sigma A5030) then cleared in Ce3D++. All images of HCR-stained embryos were taken on a Nikon A1 confocal microscope using a Plan Apo 10x objective (NA: 0.45).
Capture Hi-C and Promoter Capture Hi-C
Hi-C Libraries used for Capture and Promoter Capture were generated as previously described with a few modifications40. Briefly, for mouse ES cell lines, 1 million cells per sample were trypsinized, washed in growth media and fixed for 35 minutes at room temperature while rotating in 1mg/ml DSG (Thermo A35392) in 1 ml of PBS. 1% formaldehyde (Thermo 28908) was then added, and samples fixed for 10 more minutes. CHi-C was processed separately for the two independent lines of each genotype that had been established from two independent blastocysts. E11.5 midbrains were dissected in chilled artificial cerebrospinal fluid (ACSF: 87 mM NaCl, 26 mM NaHCO3, 2.5 mM KCl, 1.25 mM NaH2PO4, 0.5 mM CaCl2, 7 mM MgCl2, 10 mM glucose, 75 mM sucrose, saturated with 95% O2, 5% CO2, pH 7.4). Cells were then incubated with 0.1% Pronase in PBS for 15 minutes (Sigma 10165921001) at room temperature. Cells were then fixed as described above. To stop fixations, Glycine was added at final concentration of 0.13M and incubated for 5 minutes at RT and 15 minutes on ice. Cells were then washed once in cold PBS, centrifuged at 2500g 4°C for 5 mins (these centrifugation conditions were used for all washes following fixation) and pellets frozen at −80°C. Thawed cell pellets were then lysed (10mM Tris-HCL pH8, 10mM NaCl, 0.2% Igepal CA-630, Roche Complete EDTA-free Sigma #11836170001) and digested with 600U of DpnII (NEB). Biotin fill-in was done by incubating cells with a mixture of dCTP, dTTP, dGTP, Klenow polymerase (NEB M0210L) and 37.5μl Biotin-14-dATP (Thermo 19524016) for 4h at RT while shaking at 900rpm for 10 seconds every 5 minutes. Ligation was done overnight at 16°C using T4 ligase (NEB cat #M0202M). Sonication was done using Covaris onetube-10 AFA strips using the following parameters for a 300bp fragment size (Duration: 10secs, repeat for 12 times, total time 120 secs, peak power-20W, duty factor 40%, CPB-50). Library-prep with material on the beads was done using the Hyper Prep kit (Roche KK8502).
For Capture Hi-C capture reactions, 1ug of Hi-C library per sample were hybridized with 120 nucleotide biotinylated RNA probes using the SureSelect kit from Agilent (G9916B) and PCR amplification using the polymerase from the Kapa Hyper Prep Kit. For promoter Capture Hi-C, 1μg of Hi-C library per sample was hybridized with 120bp biotinylated oligos using the SeqCap EZ kits (Roche). Following washes, material was amplified by PCR using the Kapa polymerase from the Hyper Prep kit (Roche KK8502). Material from different samples was then combined and 1μg of pooled libraries was recaptured. Probes used for these assays can be found in Table S1.
Region Capture Micro-C
Libraries for Micro-C were prepared as previously described with a few changes as detailed below95,96. Dissection, dissociation, and fixation of cells grown in vitro and from E11.5 midbrains was done as described for Hi-C libraries. MNase was titrated with 1 million cells at 3U, 5U and 10U for 20 min at 37 °C in 100μl. Then, the micrococcal nuclease reaction was carried out with 5 million cells per replicate, and the micrococcal nuclease step was scaled up to 500 μL with the optimal concentration of micrococcal nuclease determined with titration experiment. Phosphorylation of DNA ends was done for 30 minutes at 37 °C using T4 Polynucleotide Kinase (NEB M0201), 3’ exonuclease activity of Klenow polymerase (NEB M0210) was done also at 37 °C for 30 minutes while shaking. Biotin incorporation with dATP and dCTP (Jena Biosciences NU-835-BIO14-L and NU-809-BIOX-L) was done at 25 °C for 90 minutes. T4-mediated proximity ligation was always done overnight with T4 ligase (NEB M0202M). Proteinase K (NEB P8107S) incubation was also performed overnight at 65 °C. After phenol/chloroform/iso-amyl alcohol extraction, the sample was split into two equal aliquots and was purified on two Zymo DNA clean and concentrator kit columns, eluted with 25 μL (preheated to 70 °C) elution buffer then pooled. Samples were loaded onto 3% TBE NuSieve GTG agarose gel in four separate wells. After excision of fragments containing dinucleosomes from each lane (>220bp, <400bp), samples were purified using a Zymo Gel DNA Recovery kit to extract DNA. DNA fragments were not polished prior to streptavidin binding. For biotin purification, 50 μL of Streptavidin C1 beads (Thermo 65002) were used per sample. Libraries were prepared using the KAPA Biosystems HyperPrep kit (Roche KK8502) and Illumina primers. After running a small-scale PCR to determine the optimal number of cycles for the required yield, four 50μL PCR reactions were set up for each sample as instructed by manufacturers. 200 μL was transferred to a new tube and incubated with 0.9x SPRI beads. Samples were sequenced by 50 bp pair- end sequencing with a NovaSeq 6000 with an SP100 kit by the NICHD Molecular Genomics Core. Region Capture of Micro-C samples was done as described previously50 and as recommended by the Twist Bioscience’s Standard Hybridization Target Enrichment Protocol with a few modifications. Specifically, to increase stringency, all wash steps were done at 70 °C instead of 48 °C and room temperature. The regions covered by probes can be found in TableS1. While biotinylated probes used in cells were of 120bp, the probes used in embryonic material were of 80bp.
Region Capture Micro-C and CHI-C analysis
Capture Hi-C and Region Capture Micro-C data were mapped to the mm10 genome using BWA (0.7.17)97. BAM files were then processed to pairs format, filtered, sorted, selected for pairs that included both ends of an interaction within the captured region, and deduplicated using pairtools (1.0.2)98. Technical replicates resulting from sequencing of the same sample more than once were merged prior to duplicate removal. Replicates where libraries were obtained from different cells were merged post duplicate removal. Cool and balanced mcool files were generated using cooler (0.8.11)99. Visualization of contact data as heatmaps was done using higlass100 at resgen.io. For generation of viewpoint tracks, pairtools select was used to identify pairs where one mate is located within the viewpoint of interest. A custom genome containing the floxed C1-C4 rescue cassette instead of the C1-C4 boundary was generated using reform (github.com/gencorefacility/reform) to create a modified fasta mm10 genome file and GTF file. These files were then used instead of mm10.fa for alignment. Insulation scores were calculated on the CHi-C cool files at 1kb resolution with a 100kb window using FAN-C (v 0.9.28)101. White diagonal lines on heatmaps reflect regions without mapped reads caused by deletions or failure to map to repetitive regions.
Cut&Run
Cut&Run was done as previously described102. Briefly, ES cells were detached from culture plates using Accutase (Sigma), counted and 200,000 cells per clone were spun at 600g for 3 mins at room temperature. Supernatant was discarded and cells were resuspended in Wash Buffer with 20mM HEPES pH 7.5, 150mM NaCl, 0.5mM Spermidine and 1x Protease inhibitor cocktail at 600g for 3 mins at room temperature. BioMag Plus Concanavalin A beads (Bangs Laboratories) were equilibrated in Binding Buffer with 20 mM HEPES pH 7.5, 10 mM KCl, 1 mM CaCl2 and 1 mM MnCl2. Cell pellets were resuspended in Wash Buffer, mixed with a slurry of equilibrated Concavalin A coated magnetic beads, and rotated for 10 mins at room temperature. For each sample, 10 μl bead slurry was used and were placed on a magnetic separator to discard the supernatant. Beads were again resuspended in Wash Buffer containing 2 mM EDTA, 0.1% bovine serum albumin, 0.05% Digitonin, and 1:50 dilution of primary antibody against H3K27Ac (Abcam, ab4729). This was incubated on a nutating platform for 2 hours at room temperature. After incubation, beads were washed twice in Digitonin Buffer (20 mM HEPES pH 7.5, 150 mM NaCl, 0.5 mM Spermidine, 1x Roche Complete Protease Inhibitor no EDTA, 0.05% Digitonin and 0.1% bovine serum albumin), then incubated with lab prepared pA-MNase (600 μg/ml, 1:200) in Digitonin Buffer for 1 hour at 4 °C. After incubation, beads were washed twice, resuspended in 150 μl of Digitonin Buffer, and equilibrated to 0°C before adding 2mM CaCl2 and incubated at 0°C for 1 hour. After incubation, 150 μl of 2X Stop Buffer containing 200 mM NaCl, 20 mM EDTA, 4 mM EGTA, 50 μg/ml RNase A and 40 μg/ml glycogen was added. Beads were incubated for 30 mins at 37°C and then spun at 16,000 g for 5 mins at 4°C. Supernatant was transferred, mixed with 3 μl 10% SDS and 1.8U Proteinase K (NEB #P8107S) and incubated overnight at 55°C, shaking at 900 rpm. After incubation, 300 μl of 25:24:1 Phenol/Chloroform/Isoamyl Alcohol was added, solutions were vortexed and transferred to Maxtrack phase-lock tubes (Qiagen #129046). Tubes were centrifuged at 16,000 g for 3 mins at room temperature. 300 μl of Chloroform was added, solutions were mixed by inversion and centrifuged at 16,000 g for 3 mins at room temperature. Aqueous layers were transferred to new tubes and DNA was isolated by ethanol precipitation and resuspended in 10 mM Tris-HCl pH 8.0 (Thermo Fisher #15568025). Cut&Run libraries were prepared following the SMARTer ThruPlex TAKARA Library Prep kit with small modifications. For each sample, 10 μl of double stranded DNA, 2 μl of Template Preparation D Buffer and 1 μl of Template Preparation D Enzyme were combined, and End Repair and A-tailing was performed in a thermocycler with a heated lid at 22°C, for 25 mins and 55°C for 20 mins. 1 μl each of Library Synthesis D Buffer and Library Synthesis D Enzyme were subsequently added, and library synthesis was performed at 22 °C for 40 mins. Immediately after, 25 μl of Library Amplification D Buffer, 1 μl of Library Amplification D Enzyme, 4 μl of nuclease-free water and 5 μl of unique Illumina-compatible indexed primer were added to each sample. Library amplification was performed using the following cycling conditions: For denaturation - 72°C for 3 mins, 85°C for 2 mins and 98°C for 2 mins, addition of unique indexes - 4 cycles of 98°C for 20 secs, 67°C for 20 secs and 72°C for 10 secs, library amplification - 14 cycles of 98°C for 20 secs and 72°C for 10 secs. Post-PCR clean-up was performed on amplified libraries with SPRI beads using 0.6X left/1x right double size selection, washed twice gently in 80% ethanol and eluted in 10–12 μl 10 mM Tris pH 8.0.
Quantification and Statistical Analysis
All experiments were performed using at least 2 different biological replicates. Details of quantification and statistical tools are described in methods section and figure legends.
Additional Resources
A list of publicly available data used in this study can be found in Key Resources Table and Table S1 which include data from the following studies40,102–104.
Key resources table
| REAGENT or RESOURCE | SOURCE | IDENTIFIER |
|---|---|---|
| Antibodies | ||
| Rabbit Anti-Histone H3 (acetyl K27) | Abcam | Cat #ab4729, RRID: AB_2118291 |
| Rabbit anti Mouse IgG (H&L) | Abcam | Cat #ab46540, RRID: AB_2614925 |
| Rabbit Anti-Histone H3 (acetyl K27) | Abcam | Cat #ab4729, RRID: AB_2118291 |
| Rabbit anti Mouse IgG (H&L) | Abcam | Cat #ab46540, RRID: AB_2614925 |
| Chemicals, peptides, and recombinant proteins | ||
| Alt-R™ S.p. Cas9 Nuclease V3, 100 μg | Integrated DNA Technologies | Cat #1081058 |
| Alt-R™ S.p. Cas9 Nuclease V3, 100 μg | Integrated DNA Technologies | Cat #1081058 |
| Neurobasal™ Medium | Thermo Fisher Scientific | Cat #21103049 |
| Neurobasal™ Medium | Thermo Fisher Scientific | Cat #21103049 |
| Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 | Thermo Fisher Scientific | Cat #11320033 |
| Penicillin-Streptomycin (10,000 U/mL) | Thermo Fisher Scientific | Cat #15140163 |
| Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 | Thermo Fisher Scientific | Cat #11320033 |
| GlutaMAX™ Supplement | Thermo Fisher Scientific | Cat #35050079 |
| Penicillin-Streptomycin (10,000 U/mL) | Thermo Fisher Scientific | Cat #15140163 |
| N-2 Supplement (100X) | Gibco | Cat #17502001 |
| GlutaMAX™ Supplement | Thermo Fisher Scientific | Cat #35050079 |
| 2-Mercaptoethanol | Gibco | Cat #21985023 |
| N-2 Supplement (100X) | Gibco | Cat #17502001 |
| B-27™ Supplement (50X) | Gibco | Cat #17504001 |
| Stemolecule™ PD0325901 (MAPK/ERK inhibitor) | Reprocell | Cat #04-0006-02 |
| 2-Mercaptoethanol | Gibco | Cat #21985023 |
| Stemolecule™ CHIR99021 (GSK3β inhibitor) | Reprocell | Cat #04-0004-02 |
| B-27™ Supplement (50X) | Gibco | Cat #17504001 |
| ESGRO® Recombinant Mouse LIF Protein | Millipore Sigma | Cat #ESG1107 |
| Stemolecule™ PD0325901 (MAPK/ERK inhibitor) | Reprocell | Cat #04-0006-02 |
| KnockOut™ DMEM | Thermo Fisher Scientific | Cat #10829018 |
| Stemolecule™ CHIR99021 (GSK3β inhibitor) | Reprocell | Cat #04-0004-02 |
| Premium Grade Fetal Bovine Serum (FBS) | Avantor® Seradigm, VWR | Cat #97068-091 |
| ESGRO® Recombinant Mouse LIF Protein | Millipore Sigma | Cat #ESG1107 |
| MEM Non-Essential Amino Acids Solution (100X) | Thermo Fisher Scientific | Cat #11140050 |
| KnockOut™ DMEM | Thermo Fisher Scientific | Cat #10829018 |
| Sodium Pyruvate (100 mM) | Thermo Fisher Scientific | Cat #11360070 |
| Premium Grade Fetal Bovine Serum (FBS) | Avantor® Seradigm, VWR | Cat #97068-091 |
| Trypsin-EDTA (0.05%), phenol red | Thermo Fisher Scientific | Cat #25300054 |
| MEM Non-Essential Amino Acids Solution (100X) | Thermo Fisher Scientific | Cat #11140050 |
| Advanced RPMI 1640 Medium | Thermo Fisher Scientific | Cat #12633012 |
| Retinoic acid | Millipore Sigma | Cat #R2625 |
| Sodium Pyruvate (100 mM) | Thermo Fisher Scientific | Cat #11360070 |
| Recombinant Human/Mouse/Rat Activin A Protein | R&D Systems | Cat #338-AC |
| Trypsin-EDTA (0.05%), phenol red | Thermo Fisher Scientific | Cat #25300054 |
| Recombinant Mouse FGF basic/FGF2/bFGF Protein | R&D Systems | Cat #3139-FB |
| Advanced RPMI 1640 Medium | Thermo Fisher Scientific | Cat #12633012 |
| Heparin sodium salt from porcine intestinal mucosa | Millipore Sigma | Cat #H3393 |
| Alcian Blue 8GX | Millipore Sigma | Cat #A5268 |
| Retinoic acid | Millipore Sigma | Cat #R2625 |
| Alizarin Red S | Millipore Sigma | Cat #A5533 |
| Recombinant Human/Mouse/Rat Activin A Protein | R&D Systems | Cat #338-AC |
| TRIzol™ Reagent | Thermo Fisher Scientific | Cat #15596026 |
| Recombinant Mouse FGF basic/FGF2/bFGF Protein | R&D Systems | Cat #3139-FB |
| TruSeq® Stranded Total RNA Library Prep kit | Illumina | Cat #20020596 |
| Heparin sodium salt from porcine intestinal mucosa | Millipore Sigma | Cat #H3393 |
| Dynabeads™ mRNA DIRECT™ Purification Kit | Thermo Fisher Scientific | Cat #61012 |
| Alcian Blue 8GX | Millipore Sigma | Cat #A5268 |
| M2 medium | Millipore Sigma | Cat #M7167 |
| Alizarin Red S | Millipore Sigma | Cat #A5533 |
| SUPERase·In™ RNase Inhibitor (20 U/μL) | Thermo Fisher Scientific | Cat #AM2694 |
| Deoxynucleotide (dNTP) Solution Mix | New England Biolabs | Cat #N0447L |
| TRIzol™ Reagent | Thermo Fisher Scientific | Cat #15596026 |
| SuperScript™ IV Reverse Transcriptase | Thermo Fisher Scientific | Cat #18090050 |
| TruSeq® Stranded Total RNA Library Prep kit | Illumina | Cat #20020596 |
| KAPA HiFi HotStart ReadyMix | Roche | Cat #9420398001 |
| Dynabeads™ mRNA DIRECT™ Purification Kit | Thermo Fisher Scientific | Cat #61012 |
| SPRIselect for Size Selection | Beckman Coulter | Cat #B23318 |
| D5000 Screen Tape | Agilent | Cat #5067-5588 |
| M2 medium | Millipore Sigma | Cat #M7167 |
| Nextera XT DNA Library Preparation Kit | Illumina | Cat #FC-131-1024 |
| SUPERase·In™ RNase Inhibitor (20 U/μL) | Thermo Fisher Scientific | Cat #AM2694 |
| D1000 Screen Tape | Agilent | Cat #5067-5585 |
| Deoxynucleotide (dNTP) Solution Mix | New England Biolabs | Cat #N0447L |
| Proteinase K, recombinant, PCR Grade | Millipore Sigma | Cat #3115836001 |
| SuperScript™ IV Reverse Transcriptase | Thermo Fisher Scientific | Cat #18090050 |
| Ultra-low Gelling Temperature Agarose | Millipore Sigma | Cat #A5030 |
| KAPA HiFi HotStart ReadyMix | Roche | Cat #9420398001 |
| Pierce™ DSG, No-Weigh™ Format | Thermo Fisher Scientific | Cat #A35392 |
| SPRIselect for Size Selection | Beckman Coulter | Cat #B23318 |
| Pierce™ 16% Formaldehyde (w/v), Methanol-free | Thermo Fisher Scientific | Cat #28908 |
| D5000 Screen Tape | Agilent | Cat #5067-5588 |
| Pronase | Millipore Sigma | Cat #10165921001 |
| Nextera XT DNA Library Preparation Kit | Illumina | Cat #FC-131-1024 |
| complete™, Mini, EDTA-free Protease Inhibitor Cocktail | Millipore Sigma | Cat #11836170001 |
| D1000 Screen Tape | Agilent | Cat #5067-5585 |
| DpnII | New England Biolabs | Cat #R0543T |
| Proteinase K, recombinant, PCR Grade | Millipore Sigma | Cat #3115836001 |
| DNA Polymerase I, Large (Klenow) Fragment | New England Biolabs | Cat #M0210L |
| Ultra-low Gelling Temperature Agarose | Millipore Sigma | Cat #A5030 |
| Biotin-14-dATP | Thermo Fisher Scientific | Cat #19524016 |
| Pierce™ DSG, No-Weigh™ Format | Thermo Fisher Scientific | Cat #A35392 |
| T4 DNA Ligase | New England Biolabs | Cat #M0202M |
| Pierce™ 16% Formaldehyde (w/v), Methanol-free | Thermo Fisher Scientific | Cat #28908 |
| KAPA HyperPrep Kit | Roche | Cat #KK8502 |
| Pronase | Millipore Sigma | Cat #10165921001 |
| SureSelect XT HS Target Enrichment Kit | Agilent | Cat #G9916B |
| complete™, Mini, EDTA-free Protease Inhibitor Cocktail | Millipore Sigma | Cat #11836170001 |
| T4 Polynucleotide Kinase | New England Biolabs | Cat #M0210 |
| DpnII | New England Biolabs | Cat #R0543T |
| Biotin-14-dATP | Jena Bioscience | Cat #NU-835-BIO14-L |
| DNA Polymerase I, Large (Klenow) Fragment | New England Biolabs | Cat #M0210L |
| Biotin-11-dCTP | Jena Bioscience | Cat #NU-809-BIOX-L |
| Biotin-14-dATP | Thermo Fisher Scientific | Cat #19524016 |
| Proteinase K | New England Biolabs | Cat #P8107S |
| T4 DNA Ligase | New England Biolabs | Cat #M0202M |
| Dynabeads™ MyOne™ Streptavidin C1 | Thermo Fisher Scientific | Cat #65002 |
| KAPA HyperPrep Kit | Roche | Cat #KK8502 |
| Accutase® solution | Millipore Sigma | Cat #A6964 |
| SureSelect XT HS Target Enrichment Kit | Agilent | Cat #G9916B |
| BioMag Plus Concanavalin A beads | Bangs Laboratories | Cat #BP531 |
| T4 Polynucleotide Kinase | New England Biolabs | Cat #M0210 |
| UltraPure™ 1M Tris-HCI, pH 8.0 | Thermo Fisher Scientific | Cat #15568025 |
| Biotin-14-dATP | Jena Bioscience | Cat #NU-835-BIO14-L |
| SMARTer® ThruPLEX® DNA-Seq Kit - 96 Rxns | Takara | Cat #R400676 |
| Biotin-11-dCTP | Jena Bioscience | Cat #NU-809-BIOX-L |
| Proteinase K | New England Biolabs | Cat #P8107S |
| Dynabeads™ MyOne™ Streptavidin C1 | Thermo Fisher Scientific | Cat #65002 |
| Accutase® solution | Millipore Sigma | Cat #A6964 |
| BioMag Plus Concanavalin A beads | Bangs Laboratories | Cat #BP531 |
| UltraPure™ 1M Tris-HCI, pH 8.0 | Thermo Fisher Scientific | Cat #15568025 |
| SMARTer® ThruPLEX® DNA-Seq Kit - 96 Rxns | Takara | Cat #R400676 |
| Critical commercial assays | ||
| Micro-CT scans | Mouse Biology Program, University of California, Davis, CA, United States | Center for Molecular and Genomic Imaging (UC Davis) |
| Deposited data | ||
| Mouse CTCF ChIP-seq (P0 midbrain) | Encode | Accession #ENCSR985ZTV |
| Mouse H3K27me3 ChIP-seq (E11.5 midbrain) | Encode | Accession #ENCSR545BRW |
| Mouse H3K27ac ChIP-seq (E11.5 midbrain) | Encode | Accession #ENCSR088UKA |
| Mouse H3K4me3 ChIP-seq (E11.5 midbrain) | Encode | Accession #ENCSR283RFW |
| Mouse H3K27ac ChIP-seq (E10.5 midbrain) | Encode | Accession #ENCFF404RRQ |
| Mouse H3K27ac ChIP-seq (E10.5 forebrain) | Encode | Accession #ENCFF796FAE |
| Mouse ATAC-seq (E11.5 midbrain) | Encode | Accession #ENCSR382RUC |
| Mouse ATAC-seq (E11.5 forebrain) | Encode | Accession #ENCFF326ULQ |
| Mouse Rad21 ChIP-seq (E11.5 head) | Gene Expression Omnibus | Accession #GSM5720443 |
| Mouse CTCF ChIP-seq (ES cells) | Gene Expression Omnibus | Accession #GSM2418864 |
| Mouse Rad21 ChIP-seq (ES cells) | Gene Expression Omnibus | Accession #GSM2418864 |
| Mouse H3K27ac Cut&Tag (ES cells) | Gene Expression Omnibus | Accession #GSE181100 |
| Mouse H3K27ac Cut&Tag (XEN cells) | Gene Expression Omnibus | Accession #GSE181100 |
| Mouse H3K27me3 Cut&Tag (ES cells) | Gene Expression Omnibus | Accession #GSE181100 |
| Mouse H3K27me3 Cut&Tag (XEN cells) | Gene Expression Omnibus | Accession #GSE181100 |
| CaptureHiC in WT mouse ES cells_rep1 | This paper | Accession #GSM8372162 |
| CaptureHiC in WT mouse ES cells_rep2 | This paper | Accession #GSM8372163 |
| CaptureHiC in WT mouse ES cells_rep3 | This paper | Accession #GSM8372164 |
| CaptureHiC in WT mouse ES cells_rep4 | This paper | Accession #GSM8372165 |
| CaptureHiC in WT mouse ES cells_rep5 | This paper | Accession #GSM8372166 |
| CaptureHiC in WT mouse ES cells_rep6 | This paper | Accession #GSM8372167 |
| CaptureHiC in WT mouse ES cells_rep7 | This paper | Accession #GSM8372168 |
| CaptureHiC in ctcf_c5-c6d mouse ES cells_rep1 | This paper | Accession #GSM8372169 |
| CaptureHiC in ctcf_c5-c6d mouse ES cells_rep2 | This paper | Accession #GSM8372170 |
| CaptureHiC in ctcf_c5-c6d mouse ES cells_rep3 | This paper | Accession #GSM8372171 |
| CaptureHiC in ctcf_c5-c6d mouse ES cells_rep4 | This paper | Accession #GSM8372172 |
| CaptureHiC in ctcf_c7-c10d mouse ES cells_rep1 | This paper | Accession #GSM8372173 |
| CaptureHiC in ctcf_c7-c10d mouse ES cells_rep2 | This paper | Accession #GSM8372174 |
| CaptureHiC in ctcf_c7-c10d mouse ES cells_rep3 | This paper | Accession #GSM8372175 |
| CaptureHiC in ctcf_WT mouse XEN cells_rep1 | This paper | Accession #GSM8372176 |
| CaptureHiC in ctcf_WT mouse XEN cells_rep2 | This paper | Accession #GSM8372177 |
| CaptureHiC in ctcf_WT mouse XEN cells_rep3 | This paper | Accession #GSM8372178 |
| CaptureHiC in ctcf_WT mouse XEN cells_rep4 | This paper | Accession #GSM8372179 |
| CaptureHiC in ctcf_WT mouse XEN cells_rep5 | This paper | Accession #GSM8372180 |
| CaptureHiC in ctcf_ c5-c6d mouse XEN cells_rep1 | This paper | Accession #GSM8372181 |
| CaptureHiC in ctcf_ c5-c6d mouse XEN cells_rep2 | This paper | Accession #GSM8372182 |
| CaptureHiC in ctcf_ c5-c6d mouse XEN cells_rep3 | This paper | Accession #GSM8372183 |
| CaptureHiC in mouse midbrain C1-C4flxflx_rep1 | This paper | Accession #GSM8372184 |
| CaptureHiC in mouse midbrain C1-C4flxflx_rep2 | This paper | Accession #GSM8372185 |
| CaptureHiC in mouse midbrain C1-C4flxflx_rep3 | This paper | Accession #GSM8372186 |
| CaptureHiC in mouse midbrain C1-C4flxflx_rep4 | This paper | Accession #GSM8372187 |
| CaptureHiC in mouse midbrain C1-C4flxflx_rep5 | This paper | Accession #GSM8372188 |
| CaptureHiC in mouse midbrain C1-C4flxflx;CRE_rep1 | This paper | Accession #GSM8372189 |
| CaptureHiC in mouse midbrain C1-C4flxflx;CRE_rep2 | This paper | Accession #GSM8372190 |
| CaptureHiC in mouse midbrain C1-C4flxflx;CRE_rep3 | This paper | Accession #GSM8372191 |
| CaptureHiC in mouse midbrain C1-C4flxflx;CRE_rep4 | This paper | Accession #GSM8372192 |
| CaptureHiC in mouse midbrain C1-C4flxflx;CRE_rep5 | This paper | Accession #GSM8372193 |
| CaptureHiC in mouse midbrain C1-C4flxflx;CRE_rep6 | This paper | Accession #GSM8372194 |
| CaptureHiC in mouse midbrain C1-C4flxflx;CRE_rep7 | This paper | Accession #GSM8372195 |
| Cut&Run in mouse ES cells_h3k27ac_wt_rep1 | This paper | Accession #GSM8384292 |
| Cut&Run in mouse ES cells_h3k27ac_wt_rep2 | This paper | Accession #GSM8384293 |
| Cut&Run in mouse ES cells_h3k27ac_c7-c10d_rep1 | This paper | Accession #GSM8384294 |
| Cut&Run in mouse ES cells_h3k27ac_c7-c10d_rep2 | This paper | Accession #GSM8384295 |
| Cut&Run in mouse ES cells_h3k27ac_c5-c6d_rep1 | This paper | Accession #GSM8384296 |
| Cut&Run in mouse ES cells_h3k27ac_c5-c6d_rep2 | This paper | Accession #GSM8384297 |
| Cut&Run in mouse ES cells_IgG_wt_rep1 | This paper | Accession #GSM8384298 |
| Cut&Run in mouse ES cells_IgG_wt_rep2 | This paper | Accession #GSM8384299 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx_rep1 | This paper | Accession #GSM8384535 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx_rep2 | This paper | Accession #GSM8384536 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx_rep3 | This paper | Accession #GSM8384537 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx;CRE_rep1 | This paper | Accession #GSM8384538 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx;CRE_rep2 | This paper | Accession #GSM8384539 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx;CRE_rep3 | This paper | Accession #GSM8384540 |
| Region Capture MicroC in mouse ES_WT_rep1 | This paper | Accession #GSM8384541 |
| Region Capture MicroC in mouse ES_WT_rep2 | This paper | Accession #GSM8384542 |
| Region Capture MicroC in mouse ES_C5-C6d_rep1 | This paper | Accession #GSM8384543 |
| Region Capture MicroC in mouse ES_C5-C6d_rep2 | This paper | Accession #GSM8384544 |
| Region Capture MicroC in mouse midbrain C1-C4flxflx_rep1 | This paper | Accession #GSM8384545 |
| Region Capture MicroC in mouse midbrain C1-C4flxflx_rep2 | This paper | Accession #GSM8384546 |
| Region Capture MicroC in mouse midbrain C1-C4flxflx_rep3 | This paper | Accession #GSM8384547 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx;CRE_rep1 | This paper | Accession #GSM8384548 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx;CRE_rep2 | This paper | Accession #GSM8384549 |
| Promoter Capture MicroC in mouse midbrain C1-C4flxflx;CRE_rep3 | This paper | Accession #GSM8384550 |
| RNA seq_E4.5_blastocyst_C7-C10_WT_SC_221 | This paper | Accession #GSM8445724 |
| RNA seq_E4.5_blastocyst_C7-C10_HOM_SC_222 | This paper | Accession #GSM8445725 |
| RNA seq_E4.5_blastocyst_C7-C10_WT_SC_223 | This paper | Accession #GSM8445726 |
| RNA seq_E4.5_blastocyst_C7-C10_HOM_SC_224 | This paper | Accession #GSM8445727 |
| RNA seq_E4.5_blastocyst_C7-C10_WT_SC_225 | This paper | Accession #GSM8445728 |
| RNA seq_E4.5_blastocyst_C7-C10_HOM_SC_226 | This paper | Accession #GSM8445729 |
| RNA seq_E4.5_blastocyst_C7-C10_WT_SC_227 | This paper | Accession #GSM8445730 |
| RNA seq_E4.5_blastocyst_C7-C10_HOM_SC_228 | This paper | Accession #GSM8445731 |
| RNA seq_E4.5_blastocyst_C5-C6_HOM_SC_229 | This paper | Accession #GSM8445732 |
| RNA seq_E4.5_blastocyst_C5-C6_WT_SC_230 | This paper | Accession #GSM8445733 |
| RNA seq_E4.5_blastocyst_C5-C6_WT_SC_231 | This paper | Accession #GSM8445734 |
| RNA seq_E4.5_blastocyst_C5-C6_WT_SC_232 | This paper | Accession #GSM8445735 |
| RNA seq_E4.5_blastocyst_C5-C6_HOM_SC_233 | This paper | Accession #GSM8445736 |
| RNA seq_E4.5_blastocyst_C5-C6_HOM_SC_234 | This paper | Accession #GSM8445737 |
| RNA seq_E4.5_blastocyst_C5-C6_WT_SC_235 | This paper | Accession #GSM8445738 |
| RNA seq_E4.5_blastocyst_C5-C6_HOM_SC_236 | This paper | Accession #GSM8445739 |
| RNA seq_mouse_C1-C4flxflx_rep1_midbrain | This paper | Accession #GSM8445740 |
| RNA seq_mouse_C1-C4flxflx_rep2_midbrain | This paper | Accession #GSM8445741 |
| RNA seq_mouse_C1-C4flxflx_rep3_midbrain | This paper | Accession #GSM8445742 |
| RNA seq_mouse_C1-C4flxflx;CRE_rep1_midbrain | This paper | Accession #GSM8445743 |
| RNA seq_mouse_C1-C4flxflx;CRE_rep2_midbrain | This paper | Accession #GSM8445744 |
| RNA seq_mouse_C1-C4flxflx;CRE_rep3_midbrain | This paper | Accession #GSM8445745 |
| RNA seq_mouse_C1-C4flxflx;CRE_rep4_midbrain | This paper | Accession #GSM8445746 |
| RNA seq_mouse_C1-C4flxflx_rep1_forebrain | This paper | Accession #GSM8445747 |
| RNA seq_mouse_C1-C4flxflx_rep2_forebrain | This paper | Accession #GSM8445748 |
| RNA seq_mouse_C1-C4flxflx_rep3_forebrain | This paper | Accession #GSM8445749 |
| RNA seq_mouse_C1-C4flxflx;CRE_rep1_forebrain | This paper | Accession #GSM8445750 |
| RNA seq_mouse_C1-C4flxflx;CRE_rep2_forebrain | This paper | Accession #GSM8445751 |
| RNA seq_mouse_C1-C4flxflx;CRE_rep3_forebrain | This paper | Accession #GSM8445752 |
| RNA seq_mouse_C1-C4flxflx;CRE_rep4_forebrain | This paper | Accession #GSM8445753 |
| Experimental models: Cell lines | ||
| Mouse embryonic fibroblasts | Millipore Sigma | Cat #PMEF-CFL |
| Mouse: C57BL/6_WT_ES cells | Established from blastocysts (This paper) | N/A |
| Mouse: C57BL/6_C5-C6d_hom_ES cells | Established from blastocysts (This paper) | N/A |
| Mouse: C57BL/6_C7-C10d_hom_ES cells | Established from blastocysts (This paper) | N/A |
| Mouse: C57BL/6_C1-C4flxflx_hom_ES cells | Established from blastocysts (This paper) | N/A |
| Experimental models: Organisms/strains | ||
| Mouse: C57BL/6N | The Jackson Laboratory | Strain number: 005304 |
| Mouse: E2A-CRE | The Jackson Laboratory | Strain number: 003724 |
| Mouse: R26-CRE-ERT2 | The Jackson Laboratory | Strain number: 008463 |
| Mouse: B6 Wnt1-Cre2 | The Jackson Laboratory | Strain number: 022501 |
| Mouse: B6 Sox10-Cre | The Jackson Laboratory | Strain number: 025807 |
| Mouse: C57BL/6N, C5-C6d | Generated in house | N/A |
| Mouse: C57BL/6N, C7-C10d | Generated in house | N/A |
| Mouse: C57BL/6N, C1-C4flxflx | Generated in house | N/A |
| Oligonucleotides | ||
| See Table S1 | This paper | N/A |
| Recombinant DNA | ||
| See Table S1 for C1-C4 Repair Template with floxed CTCF motifs | This paper | N/A |
| Software and algorithms | ||
| lcdb-wf v1.9rc | Martin, 201189 | https://github.com/lcdb/lcdb-wf |
| DESeq2 v1.22.1 | Love et al., 201492 | https://bioconductor.org/packages/release/bioc/html/DESeq2.html |
| BWA (0.7.17) | Li and Durbin, 200997 | https://bio-bwa.sourceforge.net/ |
| pairtools (1.0.2) | Open2C et al., 202398 | https://pypi.org/project/pairtools/ |
| cooler (0.8.11) | Abdennur and Mirny, 202099 | https://pypi.org/project/cooler/ |
| higlass | Kerpedjiev et al., 2018100 | https://resgen.io/ |
| reform | https://github.com/gencorefacility/reform | |
| FAN-C (v 0.9.28) | Kruse et al., 2020101 | https://pypi.org/project/fanc/#files |
Supplementary Material
Table S1 – Related to Methods, describing sequence-based QC metrics, public datasets used, probes used in capture experiments, antibodies, PCR primers, gRNAs, CRISPR repair oligos, and CRISPR junction products.
Highlights.
Deleting a boundary of a TAD with three FGF genes causes encephalocele and lethality
CTCF insulation loss induces ectopic interactions of FGF genes with distal brain enhancers
CTCF motif oriented towards enhancers is sufficient and essential for insulation
Minor sequence variants at TAD borders can severely impact development and disease
Acknowledgements
We would like to thank all members of the Unit on Genome Structure and Regulation as well Karl Pfeifer, Judy Kassis, Joana Vidigal, Nestor Saiz, Andrew Copp, Todd Macfarlan, and Tom Misteli for comments and discussions on this project and manuscript. We thank Anders Hansen for help establishing RCMC, and Caroline Esnault for assistance with RNAseq analysis. We thank NICHD’s Molecular Genomics Core, specifically Fabio Faucz, Tianwei Li, and James Iben. This work utilized the computational resources of the NIH HPC Biowulf cluster (http://hpc.nih.gov). We thank the entire NICHD animal facility, and specifically Victoria Biggs for mouse husbandry, and Alex Grinberg and Jeanne Yimdjo for zygotic electroporations. For the Micro-CT analyses, we thank Louise Lanoue from the Mouse Biology Program and Douglas J. Rowland from the Center for Molecular and Genomic Imaging, both at University of California, Davis, CA, United States. For figure 1, 2, 4, 5, S2, S3, S5 and S6 icons were created with BioRender.com. This work was funded by NIH intramural projects HD008975, HD008962, BC010338, and HD008986. This project has also been funded in part with Federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261201500003I. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Declaration of interests
The authors declare no conflicts of interest.
References
- 1.Long HK, Prescott SL, and Wysocka J (2016). Ever-Changing Landscapes: Transcriptional Enhancers in Development and Evolution. Cell 167, 1170–1187. 10.1016/j.cell.2016.09.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Furlong EEM, and Levine M (2018). Developmental enhancers and chromosome topology. Science 361, 1341–1345. 10.1126/science.aau0320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Schoenfelder S, and Fraser P (2019). Long-range enhancer-promoter contacts in gene expression control. Nat Rev Genet 20, 437–455. 10.1038/s41576-019-0128-0. [DOI] [PubMed] [Google Scholar]
- 4.de Laat W, and Duboule D (2013). Topology of mammalian developmental enhancers and their regulatory landscapes. Nature 502, 499–506. 10.1038/nature12753. [DOI] [PubMed] [Google Scholar]
- 5.Misteli T (2020). The Self-Organizing Genome: Principles of Genome Architecture and Function. Cell 183, 28–45. 10.1016/j.cell.2020.09.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Mirny L, and Dekker J (2022). Mechanisms of Chromosome Folding and Nuclear Organization: Their Interplay and Open Questions. Cold Spring Harb Perspect Biol 14. 10.1101/cshperspect.a040147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Nora EP, Lajoie BR, Schulz EG, Giorgetti L, Okamoto I, Servant N, Piolot T, van Berkum NL, Meisig J, Sedat J, et al. (2012). Spatial partitioning of the regulatory landscape of the X-inactivation centre. Nature 485, 381–385. 10.1038/nature11049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dixon JR, Selvaraj S, Yue F, Kim A, Li Y, Shen Y, Hu M, Liu JS, and Ren B (2012). Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature 485, 376–380. 10.1038/nature11082. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rowley MJ, and Corces VG (2018). Organizational principles of 3D genome architecture. Nat Rev Genet 19, 789–800. 10.1038/s41576-018-0060-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Beagan JA, and Phillips-Cremins JE (2020). On the existence and functionality of topologically associating domains. Nat Genet 52, 8–16. 10.1038/s41588-019-0561-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.de Wit E, Vos ES, Holwerda SJ, Valdes-Quezada C, Verstegen MJ, Teunissen H, Splinter E, Wijchers PJ, Krijger PH, and de Laat W (2015). CTCF Binding Polarity Determines Chromatin Looping. Mol Cell 60, 676–684. 10.1016/j.molcel.2015.09.023. [DOI] [PubMed] [Google Scholar]
- 12.Fudenberg G, Imakaev M, Lu C, Goloborodko A, Abdennur N, and Mirny LA (2016). Formation of Chromosomal Domains by Loop Extrusion. Cell Rep 15, 2038–2049. 10.1016/j.celrep.2016.04.085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Nora EP, Goloborodko A, Valton AL, Gibcus JH, Uebersohn A, Abdennur N, Dekker J, Mirny LA, and Bruneau BG (2017). Targeted Degradation of CTCF Decouples Local Insulation of Chromosome Domains from Genomic Compartmentalization. Cell 169, 930–944 e922. 10.1016/j.cell.2017.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rao SSP, Huang SC, Glenn St Hilaire B, Engreitz JM, Perez EM, Kieffer-Kwon KR, Sanborn AL, Johnstone SE, Bascom GD, Bochkov ID, et al. (2017). Cohesin Loss Eliminates All Loop Domains. Cell 171, 305–320 e324. 10.1016/j.cell.2017.09.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Schwarzer W, Abdennur N, Goloborodko A, Pekowska A, Fudenberg G, Loe-Mie Y, Fonseca NA, Huber W, Haering CH, Mirny L, and Spitz F (2017). Two independent modes of chromatin organization revealed by cohesin removal. Nature 551, 51–56. 10.1038/nature24281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zhang D, Huang P, Sharma M, Keller CA, Giardine B, Zhang H, Gilgenast TG, Phillips-Cremins JE, Hardison RC, and Blobel GA (2020). Alteration of genome folding via contact domain boundary insertion. Nat Genet 52, 1076–1087. 10.1038/s41588-020-0680-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Anania C, Acemel RD, Jedamzick J, Bolondi A, Cova G, Brieske N, Kuhn R, Wittler L, Real FM, and Lupianez DG (2022). In vivo dissection of a clustered-CTCF domain boundary reveals developmental principles of regulatory insulation. Nat Genet 54, 1026–1036. 10.1038/s41588-022-01117-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chang LH, Ghosh S, Papale A, Luppino JM, Miranda M, Piras V, Degrouard J, Edouard J, Poncelet M, Lecouvreur N, et al. (2023). Multi-feature clustering of CTCF binding creates robustness for loop extrusion blocking and Topologically Associating Domain boundaries. Nat Commun 14, 5615. 10.1038/s41467-023-41265-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rajderkar S, Barozzi I, Zhu Y, Hu R, Zhang Y, Li B, Alcaina Caro A, Fukuda-Yuzawa Y, Kelman G, Akeza A, et al. (2023). Topologically associating domain boundaries are required for normal genome function. Commun Biol 6, 435. 10.1038/s42003-023-04819-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wu HJ, Landshammer A, Stamenova EK, Bolondi A, Kretzmer H, Meissner A, and Michor F (2021). Topological isolation of developmental regulators in mammalian genomes. Nat Commun 12, 4897. 10.1038/s41467-021-24951-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kraft K, Magg A, Heinrich V, Riemenschneider C, Schopflin R, Markowski J, Ibrahim DM, Acuna-Hidalgo R, Despang A, Andrey G, et al. (2019). Serial genomic inversions induce tissue-specific architectural stripes, gene misexpression and congenital malformations. Nat Cell Biol 21, 305–310. 10.1038/s41556-019-0273-x. [DOI] [PubMed] [Google Scholar]
- 22.Franke M, Ibrahim DM, Andrey G, Schwarzer W, Heinrich V, Schopflin R, Kraft K, Kempfer R, Jerkovic I, Chan WL, et al. (2016). Formation of new chromatin domains determines pathogenicity of genomic duplications. Nature 538, 265–269. 10.1038/nature19800. [DOI] [PubMed] [Google Scholar]
- 23.Narendra V, Rocha PP, An D, Raviram R, Skok JA, Mazzoni EO, and Reinberg D (2015). CTCF establishes discrete functional chromatin domains at the Hox clusters during differentiation. Science 347, 1017–1021. 10.1126/science.1262088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lupianez DG, Kraft K, Heinrich V, Krawitz P, Brancati F, Klopocki E, Horn D, Kayserili H, Opitz JM, Laxova R, et al. (2015). Disruptions of topological chromatin domains cause pathogenic rewiring of gene-enhancer interactions. Cell 161, 1012–1025. 10.1016/j.cell.2015.04.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Juan AM, Foong YH, Thorvaldsen JL, Lan Y, Leu NA, Rurik JG, Li L, Krapp C, Rosier CL, Epstein JA, and Bartolomei MS (2022). Tissue-specific Grb10/Ddc insulator drives allelic architecture for cardiac development. Mol Cell 82, 3613–3631 e3617. 10.1016/j.molcel.2022.08.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cova G, Glaser J, Schopflin R, Prada-Medina CA, Ali S, Franke M, Falcone R, Federer M, Ponzi E, Ficarella R, et al. (2023). Combinatorial effects on gene expression at the Lbx1/Fgf8 locus resolve split-hand/foot malformation type 3. Nat Commun 14, 1475. 10.1038/s41467-023-37057-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Engel N, Raval AK, Thorvaldsen JL, and Bartolomei SM (2008). Three-dimensional conformation at the H19/Igf2 locus supports a model of enhancer tracking. Hum Mol Genet 17, 3021–3029. 10.1093/hmg/ddn200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dowen JM, Fan ZP, Hnisz D, Ren G, Abraham BJ, Zhang LN, Weintraub AS, Schujiers J, Lee TI, Zhao K, and Young RA (2014). Control of cell identity genes occurs in insulated neighborhoods in mammalian chromosomes. Cell 159, 374–387. 10.1016/j.cell.2014.09.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Flavahan WA, Drier Y, Liau BB, Gillespie SM, Venteicher AS, Stemmer-Rachamimov AO, Suva ML, and Bernstein BE (2016). Insulator dysfunction and oncogene activation in IDH mutant gliomas. Nature 529, 110–114. 10.1038/nature16490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hanssen LLP, Kassouf MT, Oudelaar AM, Biggs D, Preece C, Downes DJ, Gosden M, Sharpe JA, Sloane-Stanley JA, Hughes JR, et al. (2017). Tissue-specific CTCF-cohesin-mediated chromatin architecture delimits enhancer interactions and function in vivo. Nat Cell Biol 19, 952–961. 10.1038/ncb3573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.van Bemmel JG, Galupa R, Gard C, Servant N, Picard C, Davies J, Szempruch AJ, Zhan Y, Zylicz JJ, Nora EP, et al. (2019). The bipartite TAD organization of the X-inactivation center ensures opposing developmental regulation of Tsix and Xist. Nat Genet 51, 1024–1034. 10.1038/s41588-019-0412-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Baudic M, Murata H, Bosada FM, Melo US, Aizawa T, Lindenbaum P, van der Maarel LE, Guedon A, Baron E, Fremy E, et al. (2024). TAD boundary deletion causes PITX2-related cardiac electrical and structural defects. Nat Commun 15, 3380. 10.1038/s41467-024-47739-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ealo T, Sanchez-Gaya V, Respuela P, Munoz-San Martin M, Martin-Batista E, Haro E, and Rada-Iglesias A (2024). Cooperative insulation of regulatory domains by CTCF-dependent physical insulation and promoter competition. Nat Commun 15, 7258. 10.1038/s41467-024-51602-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ghavi-Helm Y, Jankowski A, Meiers S, Viales RR, Korbel JO, and Furlong EEM (2019). Highly rearranged chromosomes reveal uncoupling between genome topology and gene expression. Nat Genet 51, 1272–1282. 10.1038/s41588-019-0462-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Williamson I, Kane L, Devenney PS, Flyamer IM, Anderson E, Kilanowski F, Hill RE, Bickmore WA, and Lettice LA (2019). Developmentally regulated Shh expression is robust to TAD perturbations. Development 146. 10.1242/dev.179523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Paliou C, Guckelberger P, Schopflin R, Heinrich V, Esposito A, Chiariello AM, Bianco S, Annunziatella C, Helmuth J, Haas S, et al. (2019). Preformed chromatin topology assists transcriptional robustness of Shh during limb development. Proc Natl Acad Sci U S A 116, 12390–12399. 10.1073/pnas.1900672116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Despang A, Schopflin R, Franke M, Ali S, Jerkovic I, Paliou C, Chan WL, Timmermann B, Wittler L, Vingron M, et al. (2019). Functional dissection of the Sox9-Kcnj2 locus identifies nonessential and instructive roles of TAD architecture. Nat Genet 51, 1263–1271. 10.1038/s41588-019-0466-z. [DOI] [PubMed] [Google Scholar]
- 38.Rodriguez-Carballo E, Lopez-Delisle L, Willemin A, Beccari L, Gitto S, Mascrez B, and Duboule D (2020). Chromatin topology and the timing of enhancer function at the HoxD locus. Proc Natl Acad Sci U S A 117, 31231–31241. 10.1073/pnas.2015083117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Amandio AR, Beccari L, Lopez-Delisle L, Mascrez B, Zakany J, Gitto S, and Duboule D (2021). Sequential in cis mutagenesis in vivo reveals various functions for CTCF sites at the mouse HoxD cluster. Genes Dev 35, 1490–1509. 10.1101/gad.348934.121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Chakraborty S, Kopitchinski N, Zuo Z, Eraso A, Awasthi P, Chari R, Mitra A, Tobias IC, Moorthy SD, Dale RK, et al. (2023). Enhancer-promoter interactions can bypass CTCF-mediated boundaries and contribute to phenotypic robustness. Nat Genet 55, 280–290. 10.1038/s41588-022-01295-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Brewer JR, Mazot P, and Soriano P (2016). Genetic insights into the mechanisms of Fgf signaling. Genes Dev 30, 751–771. 10.1101/gad.277137.115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Kang M, Piliszek A, Artus J, and Hadjantonakis AK (2013). FGF4 is required for lineage restriction and salt-and-pepper distribution of primitive endoderm factors but not their initial expression in the mouse. Development 140, 267–279. 10.1242/dev.084996. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Feldman B, Poueymirou W, Papaioannou VE, DeChiara TM, and Goldfarb M (1995). Requirement of FGF-4 for postimplantation mouse development. Science 267, 246–249. 10.1126/science.7809630. [DOI] [PubMed] [Google Scholar]
- 44.Nowotschin S, Setty M, Kuo YY, Liu V, Garg V, Sharma R, Simon CS, Saiz N, Gardner R, Boutet SC, et al. (2019). The emergent landscape of the mouse gut endoderm at single-cell resolution. Nature 569, 361–367. 10.1038/s41586-019-1127-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Zelarayan L, Vendrell V, Dominguez-Frutos E, Lopez-Hernandez I, Schimmang-Alonso K, Alonso MT, Alvarez Y, Maier H, Anderson MJ, Lewandoski M, and Schimmang T (2022). Inactivation of Fgf3 and Fgf4 within the Fgf3/Fgf4/Fgf15 gene cluster reveals their redundant requirement for mouse inner ear induction and embryonic survival. Dev Dyn 251, 877–884. 10.1002/dvdy.435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Mansour SL, Goddard JM, and Capecchi MR (1993). Mice homozygous for a targeted disruption of the proto-oncogene int-2 have developmental defects in the tail and inner ear. Development 117, 13–28. 10.1242/dev.117.1.13. [DOI] [PubMed] [Google Scholar]
- 47.Vincentz JW, McWhirter JR, Murre C, Baldini A, and Furuta Y (2005). Fgf15 is required for proper morphogenesis of the mouse cardiac outflow tract. Genesis 41, 192–201. 10.1002/gene.20114. [DOI] [PubMed] [Google Scholar]
- 48.Borello U, Cobos I, Long JE, McWhirter JR, Murre C, and Rubenstein JL (2008). FGF15 promotes neurogenesis and opposes FGF8 function during neocortical development. Neural Dev 3, 17. 10.1186/1749-8104-3-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Fraidenraich D, Lang R, and Basilico C (1998). Distinct regulatory elements govern Fgf4 gene expression in the mouse blastocyst, myotomes, and developing limb. Dev Biol 204, 197–209. 10.1006/dbio.1998.9053. [DOI] [PubMed] [Google Scholar]
- 50.Goel VY, Huseyin MK, and Hansen AS (2023). Region Capture Micro-C reveals coalescence of enhancers and promoters into nested microcompartments. Nat Genet 55, 1048–1056. 10.1038/s41588-023-01391-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Saitsu H, Shiota K, and Ishibashi M (2006). Analysis of Fibroblast growth factor 15 cis-elements reveals two conserved enhancers which are closely related to cardiac outflow tract development. Mech Dev 123, 665–673. 10.1016/j.mod.2006.07.002. [DOI] [PubMed] [Google Scholar]
- 52.Iwahori A, Fraidenraich D, and Basilico C (2004). A conserved enhancer element that drives FGF4 gene expression in the embryonic myotomes is synergistically activated by GATA and bHLH proteins. Dev Biol 270, 525–537. 10.1016/j.ydbio.2004.03.012. [DOI] [PubMed] [Google Scholar]
- 53.Lakso M, Pichel JG, Gorman JR, Sauer B, Okamoto Y, Lee E, Alt FW, and Westphal H (1996). Efficient in vivo manipulation of mouse genomic sequences at the zygote stage. Proc Natl Acad Sci U S A 93, 5860–5865. 10.1073/pnas.93.12.5860. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Mai CT, Isenburg JL, Canfield MA, Meyer RE, Correa A, Alverson CJ, Lupo PJ, Riehle-Colarusso T, Cho SJ, Aggarwal D, et al. (2019). National population-based estimates for major birth defects, 2010–2014. Birth Defects Res 111, 1420–1435. 10.1002/bdr2.1589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Rolo A, Galea GL, Savery D, Greene NDE, and Copp AJ (2019). Novel mouse model of encephalocele: post-neurulation origin and relationship to open neural tube defects. Dis Model Mech 12. 10.1242/dmm.040683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Jacques-Fricke BT, Roffers-Agarwal J, and Gammill LS (2012). DNA methyltransferase 3b is dispensable for mouse neural crest development. PLoS One 7, e47794. 10.1371/journal.pone.0047794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Carlton MB, Colledge WH, and Evans MJ (1998). Crouzon-like craniofacial dysmorphology in the mouse is caused by an insertional mutation at the Fgf3/Fgf4 locus. Dev Dyn 212, 242–249. 10.1002/(SICI)1097-0177(199806)212:2<242::AID-AJA8>3.0.CO;2-H. [DOI] [PubMed] [Google Scholar]
- 58.Zuin J, Roth G, Zhan Y, Cramard J, Redolfi J, Piskadlo E, Mach P, Kryzhanovska M, Tihanyi G, Kohler H, et al. (2022). Nonlinear control of transcription through enhancer-promoter interactions. Nature 604, 571–577. 10.1038/s41586-022-04570-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Sarro R, Kocher AA, Emera D, Uebbing S, Dutrow EV, Weatherbee SD, Nottoli T, and Noonan JP (2018). Disrupting the three-dimensional regulatory topology of the Pitx1 locus results in overtly normal development. Development 145. 10.1242/dev.158550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.Rodriguez-Carballo E, Lopez-Delisle L, Zhan Y, Fabre PJ, Beccari L, El-Idrissi I, Huynh THN, Ozadam H, Dekker J, and Duboule D (2017). The HoxD cluster is a dynamic and resilient TAD boundary controlling the segregation of antagonistic regulatory landscapes. Genes Dev 31, 2264–2281. 10.1101/gad.307769.117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Du Z, Zheng H, Huang B, Ma R, Wu J, Zhang X, He J, Xiang Y, Wang Q, Li Y, et al. (2017). Allelic reprogramming of 3D chromatin architecture during early mammalian development. Nature 547, 232–235. 10.1038/nature23263. [DOI] [PubMed] [Google Scholar]
- 62.Ke Y, Xu Y, Chen X, Feng S, Liu Z, Sun Y, Yao X, Li F, Zhu W, Gao L, et al. (2017). 3D Chromatin Structures of Mature Gametes and Structural Reprogramming during Mammalian Embryogenesis. Cell 170, 367–381 e320. 10.1016/j.cell.2017.06.029. [DOI] [PubMed] [Google Scholar]
- 63.Flyamer IM, Gassler J, Imakaev M, Brandao HB, Ulianov SV, Abdennur N, Razin SV, Mirny LA, and Tachibana-Konwalski K (2017). Single-nucleus Hi-C reveals unique chromatin reorganization at oocyte-to-zygote transition. Nature 544, 110–114. 10.1038/nature21711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Borsos M, Perricone SM, Schauer T, Pontabry J, de Luca KL, de Vries SS, Ruiz-Morales ER, Torres-Padilla ME, and Kind J (2019). Genome-lamina interactions are established de novo in the early mouse embryo. Nature 569, 729–733. 10.1038/s41586-019-1233-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65.Bonev B, Mendelson Cohen N, Szabo Q, Fritsch L, Papadopoulos GL, Lubling Y, Xu X, Lv X, Hugnot JP, Tanay A, and Cavalli G (2017). Multiscale 3D Genome Rewiring during Mouse Neural Development. Cell 171, 557–572 e524. 10.1016/j.cell.2017.09.043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 66.Andreu MJ, Alvarez-Franco A, Portela M, Gimenez-Llorente D, Cuadrado A, Badia-Careaga C, Tiana M, Losada A, and Manzanares M (2022). Establishment of 3D chromatin structure after fertilization and the metabolic switch at the morula-to-blastocyst transition require CTCF. Cell Rep 41, 111501. 10.1016/j.celrep.2022.111501. [DOI] [PubMed] [Google Scholar]
- 67.Salmon Hillbertz NH, Isaksson M, Karlsson EK, Hellmen E, Pielberg GR, Savolainen P, Wade CM, von Euler H, Gustafson U, Hedhammar A, et al. (2007). Duplication of FGF3, FGF4, FGF19 and ORAOV1 causes hair ridge and predisposition to dermoid sinus in Ridgeback dogs. Nat Genet 39, 1318–1320. 10.1038/ng.2007.4. [DOI] [PubMed] [Google Scholar]
- 68.Jehee FS, Bertola DR, Yelavarthi KK, Krepischi-Santos AC, Kim C, Vianna-Morgante AM, Vermeesch JR, and Passos-Bueno MR (2007). An 11q11-q13.3 duplication, including FGF3 and FGF4 genes, in a patient with syndromic multiple craniosynostoses. Am J Med Genet A 143A, 1912–1918. 10.1002/ajmg.a.31863. [DOI] [PubMed] [Google Scholar]
- 69.Grillo L, Greco D, Pettinato R, Avola E, Potenza N, Castiglia L, Spalletta A, Amata S, Di Benedetto D, Luciano D, et al. (2014). Increased FGF3 and FGF4 gene dosage is a risk factor for craniosynostosis. Gene 534, 435–439. 10.1016/j.gene.2013.09.120. [DOI] [PubMed] [Google Scholar]
- 70.Ziebart T, Draenert FG, Galetzka D, Babaryka G, Schmidseder R, Wagner W, and Bartsch O (2013). The original family revisited after 37 years: odontoma-dysphagia syndrome is most likely caused by a microduplication of chromosome 11q13.3, including the FGF3 and FGF4 genes. Clin Oral Investig 17, 123–130. 10.1007/s00784-012-0676-6. [DOI] [PubMed] [Google Scholar]
- 71.Martinez Anaya D, Juarez-Velazquez MDR, Reyes Ruvalcaba S, Navarrete-Meneses MDP, Salas Labadia C, Lieberman Hernandez E, and Perez-Vera P (2023). Pure Interstitial Trisomy 11q Arising from a Nonrecurrent 11q13.1q22.3 Mosaic Intrachromosomal Duplication in a Patient with Craniofacial Dysmorphism and Genital Anomalies. Mol Syndromol 14, 310–321. 10.1159/000528472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 72.Flavahan WA, Drier Y, Johnstone SE, Hemming ML, Tarjan DR, Hegazi E, Shareef SJ, Javed NM, Raut CP, Eschle BK, et al. (2019). Altered chromosomal topology drives oncogenic programs in SDH-deficient GISTs. Nature 575, 229–233. 10.1038/s41586-019-1668-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 73.Kim KL, Rahme GJ, Goel VY, El Farran CA, Hansen AS, and Bernstein BE (2024). Dissection of a CTCF topological boundary uncovers principles of enhancer-oncogene regulation. Mol Cell 84, 1365–1376 e1367. 10.1016/j.molcel.2024.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 74.Fudenberg G, and Pollard KS (2019). Chromatin features constrain structural variation across evolutionary timescales. Proc Natl Acad Sci U S A 116, 2175–2180. 10.1073/pnas.1808631116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 75.Vietri Rudan M, Barrington C, Henderson S, Ernst C, Odom DT, Tanay A, and Hadjur S (2015). Comparative Hi-C reveals that CTCF underlies evolution of chromosomal domain architecture. Cell Rep 10, 1297–1309. 10.1016/j.celrep.2015.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 76.Ventura A, Kirsch DG, McLaughlin ME, Tuveson DA, Grimm J, Lintault L, Newman J, Reczek EE, Weissleder R, and Jacks T (2007). Restoration of p53 function leads to tumour regression in vivo. Nature 445, 661–665. 10.1038/nature05541. [DOI] [PubMed] [Google Scholar]
- 77.Lewis AE, Vasudevan HN, O’Neill AK, Soriano P, and Bush JO (2013). The widely used Wnt1-Cre transgene causes developmental phenotypes by ectopic activation of Wnt signaling. Dev Biol 379, 229–234. 10.1016/j.ydbio.2013.04.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 78.Matsuoka T, Ahlberg PE, Kessaris N, Iannarelli P, Dennehy U, Richardson WD, McMahon AP, and Koentges G (2005). Neural crest origins of the neck and shoulder. Nature 436, 347–355. 10.1038/nature03837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 79.Chari R, Yeo NC, Chavez A, and Church GM (2017). sgRNA Scorer 2.0: A Species-Independent Model To Predict CRISPR/Cas9 Activity. ACS Synth Biol 6, 902–904. 10.1021/acssynbio.6b00343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 80.Gooden AA, Evans CN, Sheets TP, Clapp ME, and Chari R (2021). dbGuide: a database of functionally validated guide RNAs for genome editing in human and mouse cells. Nucleic Acids Res 49, D871–D876. 10.1093/nar/gkaa848. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 81.D’Astolfo DS, Pagliero RJ, Pras A, Karthaus WR, Clevers H, Prasad V, Lebbink RJ, Rehmann H, and Geijsen N (2015). Efficient intracellular delivery of native proteins. Cell 161, 674–690. 10.1016/j.cell.2015.03.028. [DOI] [PubMed] [Google Scholar]
- 82.Niakan KK, Schrode N, Cho LT, and Hadjantonakis AK (2013). Derivation of extraembryonic endoderm stem (XEN) cells from mouse embryos and embryonic stem cells. Nat Protoc 8, 1028–1041. 10.1038/nprot.2013.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 83.Mallo M, and Brandlin I (1997). Segmental identity can change independently in the hindbrain and rhombencephalic neural crest. Dev Dyn 210, 146–156. 10.1002/(SICI)1097-0177(199710)210:2<146::AID-AJA7>3.0.CO;2-G. [DOI] [PubMed] [Google Scholar]
- 84.Shankar SP, Grimsrud K, Lanoue L, Egense A, Willis B, Horberg J, AlAbdi L, Mayer K, Utkur K, Monaghan KG, et al. (2022). A novel DPH5-related diphthamide-deficiency syndrome causing embryonic lethality or profound neurodevelopmental disorder. Genet Med 24, 2207. 10.1016/j.gim.2022.07.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 85.Huffman SR, Almamun M, and Rivera RM (2012). Isolation of RNA and DNA from single preimplantation embryos and a small number of Mammalian oocytes for imprinting studies. Methods Mol Biol 925, 201–209. 10.1007/978-1-62703-011-3_13. [DOI] [PubMed] [Google Scholar]
- 86.Picelli S, Faridani OR, Bjorklund AK, Winberg G, Sagasser S, and Sandberg R (2014). Full-length RNA-seq from single cells using Smart-seq2. Nat Protoc 9, 171–181. 10.1038/nprot.2014.006. [DOI] [PubMed] [Google Scholar]
- 87.Satija R, Farrell JA, Gennert D, Schier AF, and Regev A (2015). Spatial reconstruction of single-cell gene expression data. Nat Biotechnol 33, 495–502. 10.1038/nbt.3192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 88.Trombetta JJ, Gennert D, Lu D, Satija R, Shalek AK, and Regev A (2014). Preparation of Single-Cell RNA-Seq Libraries for Next Generation Sequencing. Curr Protoc Mol Biol 107, 4 22 21–24 22 17. 10.1002/0471142727.mb0422s107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 89.Martin M (2011). Cutadapt removes adapter sequences from high-throughput sequencing reads. 2011 17, 3. 10.14806/ej.17.1.200. [DOI] [Google Scholar]
- 90.Patro R, Duggal G, Love MI, Irizarry RA, and Kingsford C (2017). Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods 14, 417–419. 10.1038/nmeth.4197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 91.Pertea G, and Pertea M (2020). GFF Utilities: GffRead and GffCompare. F1000Res 9. 10.12688/f1000research.23297.2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 92.Love MI, Huber W, and Anders S (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550. 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 93.Soneson C, Love MI, and Robinson MD (2015). Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences. F1000Res 4, 1521. 10.12688/f1000research.7563.2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 94.Anderson MJ, Magidson V, Kageyama R, and Lewandoski M (2020). Fgf4 maintains Hes7 levels critical for normal somite segmentation clock function. Elife 9. 10.7554/eLife.55608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 95.Slobodyanyuk E, Cattoglio C, and Hsieh TS (2022). Mapping Mammalian 3D Genomes by Micro-C. Methods Mol Biol 2532, 51–71. 10.1007/978-1-0716-2497-5_4. [DOI] [PubMed] [Google Scholar]
- 96.Brown JL, Zhang L, Rocha PP, Kassis JA, and Sun MA (2024). Polycomb protein binding and looping in the ON transcriptional state. Sci Adv 10, eadn1837. 10.1126/sciadv.adn1837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 97.Li H, and Durbin R (2009). Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–1760. 10.1093/bioinformatics/btp324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 98.Open2C, Abdennur N, Fudenberg G, Flyamer IM, Galitsyna AA, Goloborodko A, Imakaev M, and Venev SV (2024). Pairtools: From sequencing data to chromosome contacts. PLoS Comput Biol 20, e1012164. 10.1371/journal.pcbi.1012164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 99.Abdennur N, and Mirny LA (2020). Cooler: scalable storage for Hi-C data and other genomically labeled arrays. Bioinformatics 36, 311–316. 10.1093/bioinformatics/btz540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 100.Kerpedjiev P, Abdennur N, Lekschas F, McCallum C, Dinkla K, Strobelt H, Luber JM, Ouellette SB, Azhir A, Kumar N, et al. (2018). HiGlass: web-based visual exploration and analysis of genome interaction maps. Genome Biol 19, 125. 10.1186/s13059-018-1486-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 101.Kruse K, Hug CB, and Vaquerizas JM (2020). FAN-C: a feature-rich framework for the analysis and visualisation of chromosome conformation capture data. Genome Biol 21, 303. 10.1186/s13059-020-02215-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 102.Thompson JJ, Lee DJ, Mitra A, Frail S, Dale RK, and Rocha PP (2022). Extensive co-binding and rapid redistribution of NANOG and GATA6 during emergence of divergent lineages. Nat Commun 13, 4257. 10.1038/s41467-022-31938-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 103.Cattoglio C, Pustova I, Walther N, Ho JJ, Hantsche-Grininger M, Inouye CJ, Hossain MJ, Dailey GM, Ellenberg J, Darzacq X, et al. (2019). Determining cellular CTCF and cohesin abundances to constrain 3D genome models. Elife 8. 10.7554/eLife.40164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 104.Luo Y, Hitz BC, Gabdank I, Hilton JA, Kagda MS, Lam B, Myers Z, Sud P, Jou J, Lin K, et al. (2020). New developments on the Encyclopedia of DNA Elements (ENCODE) data portal. Nucleic Acids Res 48, D882–D889. 10.1093/nar/gkz1062. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1 – Related to Methods, describing sequence-based QC metrics, public datasets used, probes used in capture experiments, antibodies, PCR primers, gRNAs, CRISPR repair oligos, and CRISPR junction products.
Data Availability Statement
This study only used available software. Software parameters used are described in Methods. FASTQ and processed data can be found in GEO under accession number GSE271760. Chromosome conformation capture data can be navigated at: resgen.io/pedrorocha/FGFs/views/. Any additional information required to reanalyze the data reported in this work paper is available from the Lead Contact upon request.