Eosinophilic esophagitis (EoE) is a chronic, food-driven allergic disease resulting in eosinophilic esophageal inflammation. We recently found that EoE susceptibility is associated with genetic variants in the promoter of CAPN14, a gene with reported esophagus-specific expression. CAPN14 is dynamically up-regulated as a function of EoE disease activity and after exposure of epithelial cells to interleukin-13 (IL-13). Herein, we aimed to explore molecular modulation of CAPN14 expression. We identified three putative binding sites for the IL-13-activated transcription factor STAT6 in the promoter and first intron of CAPN14 Luciferase reporter assays revealed that the two most distal STAT6 elements were required for the ∼10-fold increase in promoter activity subsequent to stimulation with IL-13 or IL-4, and also for the genotype-dependent reduction in IL-13-induced promoter activity. One of the STAT6 elements in the promoter was necessary for IL-13-mediated induction of CAPN14 promoter activity while the other STAT6 promoter element was necessary for full induction. Chromatin immunoprecipitation in IL-13 stimulated esophageal epithelial cells was used to further support STAT6 binding to the promoter of CAPN14 at these STAT6 binding sites. The highest CAPN14 and calpain-14 expression occurred with IL-13 or IL-4 stimulation of esophageal epithelial cells under culture conditions that allow the cells to differentiate into a stratified epithelium. This work corroborates a candidate molecular mechanism for EoE disease etiology in which the risk variant at 2p23 dampens CAPN14 expression in differentiated esophageal epithelial cells following IL-13/STAT6 induction of CAPN14 promoter activity.
Abstract Systemic Lupus Erythematosus (SLE) is an incurable, debilitating autoimmune disease characterized by widespread inflammation and rampant production of autoantibodies. The most prominent and highly replicated set of genes up-regulated in the immune cells of patients with SLE are the type I interferons (IFN-I) and IFN-responsive genes. IFN-I are predominantly made by plasmacytoid dendritic cells (pDCs), and their expression is directly regulated by the transcription factor interferon regulatory factor 7 (IRF7). While IRF7 is an established SLE risk locus, the variants responsible for disease pathology remain unknown. We hypothesize that an amino-acid changing SLE risk variant in IRF7 (rs1131665) alters expression of disease-relevant IFN-I in clinically-relevant cells to increase SLE risk. The functional genomic consequences of the SLE-associated variant were assessed in human cell lines and in genome-edited mice with an introduced SLE-risk variant at Irf7. Our data demonstrate greater than 2-fold genotype-dependence in IFN-stimulated response element-driven luciferase activity and inflammatory cytokine secretion detected in supernatant after toll-like receptor-7 stimulation. Gene expression differences in cells with IRF7/Irf7 risk variants are consistent with those dysregulated in SLE patients. In the present study, we demonstrate the functional consequences of an amino acid substitution in a critical type I interferon regulator. Understanding these mechanisms will enhance development of more effective clinical practices for autoimmune patients expressing the risk variant for IRF7.
The disappearance of fine motor control Manual skills are much better developed in primates than in rodents. This difference is in part due to species-specific differences in the control of motoneurons by the brain. Gu et al. used a range of approaches to evaluate potential corticospinal tract projections in neonatal mice. These projections exist immediately after birth but disappear within the first 2 postnatal weeks owing to the actions of plexin A, a member of the semaphorin receptor family. Targeted deletion of semaphorin receptors in mutant mice prevented elimination of corticospinal tract projection and loss of functional monosynaptic input to spinal motoneurons. Science , this issue p. 400
DREAM challenges are community competitions designed to advance computational methods and address fundamental questions in system biology and translational medicine. Each challenge asks participants to develop and apply computational methods to either predict unobserved outcomes or to identify unknown model parameters given a set of training data. Computational methods are evaluated using an automated scoring metric, scores are posted to a public leaderboard, and methods are published to facilitate community discussions on how to build improved methods. By engaging participants from a wide range of science and engineering backgrounds, DREAM challenges can comparatively evaluate a wide range of statistical, machine learning, and biophysical methods. Here, we describe DREAMTools, a Python package for evaluating DREAM challenge scoring metrics. DREAMTools provides a command line interface that enables researchers to test new methods on past challenges, as well as a framework for scoring new challenges. As of March 2016, DREAMTools includes more than 80% of completed DREAM challenges. DREAMTools complements the data, metadata, and software tools available at the DREAM website http://dreamchallenges.org and on the Synapse platform at https://www.synapse.org.Availability:DREAMTools is a Python package. Releases and documentation are available at http://pypi.python.org/pypi/dreamtools. The source code is available at http://github.com/dreamtools/dreamtools.
Article Figures and data Abstract Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Lineage specification is governed by gene regulatory networks (GRNs) that integrate the activity of signaling effectors and transcription factors (TFs) on enhancers. Sox17 is a key transcriptional regulator of definitive endoderm development, and yet, its genomic targets remain largely uncharacterized. Here, using genomic approaches and epistasis experiments, we define the Sox17-governed endoderm GRN in Xenopus gastrulae. We show that Sox17 functionally interacts with the canonical Wnt pathway to specify and pattern the endoderm while repressing alternative mesectoderm fates. Sox17 and β-catenin co-occupy hundreds of key enhancers. In some cases, Sox17 and β-catenin synergistically activate transcription apparently independent of Tcfs, whereas on other enhancers, Sox17 represses β-catenin/Tcf-mediated transcription to spatially restrict gene expression domains. Our findings establish Sox17 as a tissue-specific modifier of Wnt responses and point to a novel paradigm where genomic specificity of Wnt/β-catenin transcription is determined through functional interactions between lineage-specific Sox TFs and β-catenin/Tcf transcriptional complexes. Given the ubiquitous nature of Sox TFs and Wnt signaling, this mechanism has important implications across a diverse range of developmental and disease contexts. Introduction During embryogenesis, the pluripotent zygote progressively gives rise to specialized cell types expressing distinct sets of genes that, in turn, define the cell’s identity and encode proteins necessary for its function. Lineage-specific gene expression is controlled by the genomic integration of signaling pathways and transcription factors (TFs) on DNA cis-regulatory modules (CRMs), such as enhancers, that control transcription (Heinz et al., 2015; Stevens et al., 2017). An important goal of developmental biology is to elucidate how signaling effectors and TFs interact on distinct sets of CRMs within the chromatin landscape to form gene regulatory networks (GRNs) that activate lineage-specific transcriptional programs whilst repressing expression of alternative fates (Charney et al., 2017b). This will provide a deeper understanding of how transcriptional networks are established, dysregulated in disease, and how GRNs might be manipulated for therapeutic purposes. We have addressed this in the context of endoderm germ layer specification: one of the earliest cell fate decisions in vertebrate development that provides a relatively simple model to elucidate how GRNs control lineage-specific transcriptional programs (Charney et al., 2017b). In embryos and pluripotent stem cells (PSCs), Nodal/Smad2 and Wnt/ß-catenin (Ctnnb1; hereafter Bcat) signaling cooperate to initiate the mesendoderm program and subsequent development of definitive endoderm progenitors, which gives rise to the epithelia of the digestive and respiratory systems (Sumi et al., 2008; Zorn and Wells, 2009). Downstream of Nodal and Wnt, a core set of endoderm TFs: Sox17, Gata4-6, Eomes, Foxa1/2 and Mix family of homeodomain TFs execute the endoderm GRN (Arnold et al., 2008; Charney et al., 2017b; Engert et al., 2013; Sinner et al., 2006). In Xenopus where endoderm specification is well studied, the Nodal and Wnt pathways interact at several levels. First, maternal (m) Wnt/Bcat, active on the dorsal side of the blastula, is required for high levels of nodal expression at the onset of zygotic transcription (Hyde and Old, 2000). Then Nodal and mWnt cooperate to promote the expression of endoderm TFs Sox17, Foxa2 and many dorsal mesendoderm organizer genes (Xanthos et al., 2002). A few hours later, zygotic (z) wnt8 (a Nodal target) is expressed on the ventral side of the gastrula where, together with Nodal signaling, it promotes ventral and posterior mesendoderm identities (Charney et al., 2017b; Stevens et al., 2017). In mammals, early Wnt/Bcat similarly activates expression of Nodal and core endoderm TFs, with prolonged Wnt promoting posterior hindgut fate (Engert et al., 2013; Zorn and Wells, 2009). Functional interactions between the core endoderm TFs and the Wnt pathway are thought to (1) segregate the transient mesendoderm into endoderm and mesoderm, (2) pattern the nascent endoderm into spatially distinct subtypes and (3) execute the downstream differentiation program to give rise to endoderm-derived lineages. But mechanistically how the Wnt signaling machinery interacts with core endoderm TFs to execute this differentiation cascade is unresolved. The transcriptional targets of the endoderm TFs are largely unknown, and it is unclear exactly how Wnt/Bcat regulates distinct spatiotemporal transcription programs in the embryo. Indeed, how the canonical Wnt pathway elicits context-specific transcriptional responses in its multitude of different biological roles in development, homeostasis and cancer is still poorly understood. According to the dogma of canonical Wnt signaling, Wnt-activated Fzd-Lrp5/6 receptor complexes sequester the Bcat degradation complex, resulting in the stabilization and translocation of Bcat to the nucleus. There, it interacts with one of four HMG-box Tcf TFs (Tcf7, Tcf7L1, Tcf7L2 and Lef1) at Wnt-responsive CRMs (Cadigan and Waterman, 2012; Schuijers et al., 2014), ultimately activating a multiprotein ‘Wnt-enhanceosome’ with the scaffold proteins Bcl9, Pygopus and the ChiLS complex (Gammons and Bienz, 2018). In the absence of Bcat, Tcfs are coupled with corepressor proteins Tle/Groucho and histone deacetylase Hdac to inhibit transcription of Wnt target genes. Bcat displaces Tle and recruits a co-activator complex including histone acetyltransferases Ep300 or CBP to stimulate transcription (Cadigan and Waterman, 2012). How distinct context-specific Wnt target genes are selected is unclear since all Tcfs have nearly identical DNA-binding specificities (Badis et al., 2009; Ramakrishnan and Cadigan, 2017) and for the most part they are ubiquitously expressed. An emerging idea is that the Wnt-enhanceosome also interacts with other lineage-specific TFs to integrate lineage-specific inputs (Gammons and Bienz, 2018; Trompouki et al., 2011); yet, how these impact genomic specificity in vivo is largely untested and the idea of Tcf-independent Wnt-mediated transcription remains controversial. In this study, we investigated the possibility that Sox17 functionally interacts with Wnt/Bcat to regulate transcription in the Xenopus endodermal GRN. In all vertebrate embryos, Sox17 is specifically expressed in the gastrula endoderm where it is required for early gut development (Clements et al., 2003; Hudson et al., 1997; Kanai-Azuma et al., 2002; Viotti et al., 2014). Despite the critical role of Sox17 in endoderm development, only few of its direct transcriptional targets have been identified (e.g.: hnf1b, foxa1 and dhh) (Ahmed et al., 2004; Sinner et al., 2004; Yagi et al., 2008). In Xenopus, ectopic Sox17 is sufficient to initiate endoderm development in pluripotent blastula animal cap cells (Clements et al., 2003; Hudson et al., 1997), and co-injection of stabilized Bcat can enhance this activity (Sinner et al., 2004). Sox17 can physically interact with Bcat in vitro and suppress the transcriptional activity of generic Tcf/Bcat reporter constructs (TOPflash) in tissue culture experiments (Sinner et al., 2007; Zorn et al., 1999). However, the biological relevance of these interactions and whether Sox17 and Bcat functionally interact on chromatin to regulate the endoderm GRN remains unknown. Here, we define the genomic targets of Sox17 in the Xenopus gastrula. In addition to promoting expression of endoderm genes, Sox17 also represses ectoderm and mesoderm gene transcription, and acts in a negative feedback loop to restrain Nodal signaling. We demonstrate that functional interactions with canonical Wnt signaling is a key feature of the Sox17-regulated GRN. Over a third of all Bcat and Sox17 genomic binding in the gastrula occur at the same CRMs. In some instances, Sox17 suppresses Bcat-Tcf mediated transcription, while in other cases, Sox17 and Bcat synergistically activates enhancers apparently independently of Tcfs, indicating a novel mode of regulation. These results provide new insight into the GRN controlling endoderm development and have implications for how Sox TFs and Bcat might interact in diverse biological contexts from development to cancer. Results Sox17 regulates a genomic program controlling germ layer segregation and endoderm development To identify the transcriptional program regulated by Sox17, we performed RNA-sequencing (RNA-Seq) on control and Sox17-depleted Xenopus tropicalis embryos at multiple time points during blastula and gastrula stages (NF9-12) when the endoderm germ layer is being specified. In Xenopus tropicalis, there are three redundant genes: sox17a, sox17b.1 and sox17b.2 (collectively sox17) with indistinguishable activities and identical expression in presumptive vegetal endoderm cells of gastrula embryos (Figure 1A and Figure 1—figure supplement 1A; D'Souza et al., 2003; Hellsten et al., 2010). Microinjection of a combination of antisense morpholino oligos (sox17aMO and sox17bMO) targeting all three paralogs resulted in a robust knockdown of Sox17 protein as confirmed by immunostaining (Figure 1B and Figure 1—figure supplement 1B). The Sox17-MO phenotype was consistent with previous reports (Clements et al., 2003; Sinner et al., 2006) and phenocopied mouse mutants with defective gut development (Kanai-Azuma et al., 2002). Injection of mRNA encoding mouse Sox17 rescued both the anatomical and gene expression phenotypes confirming the efficacy and specificity of the MOs (Figure 1F and Figure 1—figure supplement 1C,D). Figure 1 with 1 supplement see all Download asset Open asset Sox17-regulated endoderm transcriptome. (A) Endoderm expression of sox17a and sox17b in NF10.5 Xenopus tropicalis gastrula. (B–C) Sox17 immunostaining in control-MO (B) and sox17a/b-MO (C) injected gastrula with an antibody that recognizes both Sox17a and Sox17b shows MO effective knockdown. (D) Time course heatmap of Sox17-regulated transcripts. Differentially expressed genes from a pairwise comparison of control-MO and sox17-MO (>2 fold change, FDR < 5%) at each stage showing transcripts that are downregulated or upregulated early (NF9-10) or late (NF10.5–12). Key genes are color coded based on regional expression from fate map in (E). (E) Enriched expression of Sox17-regulated transcripts. Gastrula fate map (dorsal right) colored by different tissues. Stacked bar graphs showing different patterns of enriched spatial expression based on RNA-seq of dissected gastrula tissues (Blitz et al., 2017). Distribution of expression patterns statistically different from all genes in the genome was determined by two-sided Kolmogorov-Smirnov tests *p<0.01. (F) In situ validation of Sox17-regulated transcripts. Disrupted expression in sox17-MOs is rescued by co-injection of mouse (m) Sox17 RNA. Differential expression analysis of control-MO and Sox17-MO embryos identified 1023 Sox17-regulated genes (>2 fold change, FDR< 5%) (Figure 1D and Supplementary file 1). Gene Ontology (GO) enrichment was consistent with the Sox17-regulated transcriptome being involved in ‘endoderm formation’, ‘epithelial differentiation’ and ‘digestive track morphogenesis’ (Figure 1—figure supplement 1E). In total, 493 transcripts were downregulated in Sox17-depleted embryos and 530 transcript were upregulated. The time course data revealed that >75% of the differentially expressed genes were changed by Sox17 depletion (either directly or indirectly) during the mid-late gastrula (NF10.5–12), consistent with a role in maintaining endodermal fate after initial induction by Nodal signaling in the blastula (NF9) (Figure 1D and Figure 1—figure supplement 1A). Differentially expressed genes include 73 TFs, including known Sox17 targets foxa2 and hnf1b (Sinner et al., 2004; Sinner et al., 2006), and paracrine signaling components (Supplementary file 1) such as the Hedgehog pathway (dhh, hhip, gli1); a key epithelial signal in gut organogenesis. This confirms that Sox17 sits atop of the regulatory hierarchy regulating endoderm differentiation. We next investigated how Sox17-regulated genes were spatially expressed, leveraging a previously published RNA-seq of different tissues dissected from gastrula embryos (Blitz et al., 2017). As predicted, a majority of Sox17-dependent genes were enriched in the vegetal endoderm and dorsal mesendoderm (organizer) (Figure 1E) including osr1, hnf1b, dhh and slc5a8 (Figure 1F and Figure 1—figure supplement 1D). Interestingly, over 30% of the genes upregulated early (NF9-10) in the Sox17-depleted embryos were normally enriched in the ectoderm or mesoderm tissue; examples include ectoderm-promoting TFs tfap2a and lhx5 (Houston and Wylie, 2003; Figure 1D–F) as well as the mesoderm TFs foxf1, tbx20 and hlx. This suggests that Sox17 plays an important role in repressing ectoderm and mesoderm fate in vegetal endoderm cells. Unexpectedly Sox17 also negatively regulated ~150 genes that are normally enriched in the vegetal endoderm and dorsal mesendoderm. Some of these vegetally enriched genes encoded components of the endoderm promoting Nodal pathway including nodal1, nodal6, gdf3, gdf6, foxh1, mix1 and bix1.1, all of which were upregulated in Sox17-depleted embryos (Figure 1D–F and Figure 1—figure supplement 1). Indeed, even sox17 transcripts were modestly increased in Sox17-MO embryos even though previous work has demonstrated that Sox17 directly maintains its own transcription (Howard et al., 2007) after initial induction by Nodal/Smad2. These data indicate that Sox17 acts as a negative feedback inhibitor that restricts excessive endoderm development by restraining Nodal activity after initial induction. The observation that over 10% of Sox17-regulated genes are enriched in the organizer mesendoderm while Sox17 is present throughout the endoderm, suggests that Sox17 might also functionally interact with the Wnt and/or Bmp dorso-ventral patterning pathways to control spatial expression. Intriguingly, we found that Sox17 regulates the expression of several Wnt/Bcat pathway components and targets including: dkk1, dkk2, fzd5, nodal3.1, pygo1, sia1, ror1, rsop2 and wnt11 (Figure 1D,F and Supplementary file 1). Together these data suggest that Sox17 regulates both endoderm specification and patterning while suppressing mesectoderm fate and that Sox17 participates in feedback loops with the Wnt and Nodal pathways. Sox17 ChIP-Seq reveals direct endodermal targets and a Nodal and Wnt feedback To identify direct Sox17 targets we generated and validated anti-Sox17 antibodies (Figure 2—figure supplement 1A–C) and performed Sox17 ChIP-seq in gastrula (NF10.5) embryos, identifying 8436 statistically significant Sox17-bound putative CRMs (IDR; p<0.05) (Supplementary file 2). These were associated with 4801 genes (Figure 2A), based on annotation to the nearest transcription start-site (TSS) by HOMER (Heinz et al., 2010). 88% of Sox17-bound loci were in introns or intergenic regions more than 1 kb away from the TSS, consistent with Sox17 binding at distal CRMs (Figure 2—figure supplement 2). A comparison to published ChIP-seq data of Xenopus embryos at the same stage (Hontelez et al., 2015) showed that most of the Sox17-bound genomic loci were also bound by Ep300, (Figure 2B,E) indicative of active enhancers. Motif analysis of the ChIP-seq peaks confirmed that Sox17 motifs were the most enriched, as expected (Figure 2D and Figure 2—figure supplement 2). Control ChIP-qPCR experiments of Sox17-MO embryos showed binding reduced to near background levels at 9 of 10 loci tested confirming a robust knockdown and the specificity of the Sox17 antibody (Figure 2—figure supplement 1D). Figure 2 with 2 supplements see all Download asset Open asset ChIP-seq identifies Sox17-bound enhancers and direct transcriptional targets. Sox17 ChIP-seq of gastrula embryos. (A) Venn diagram of Sox17-bound genes (and associated peaks) from ChIP-seq intersected with Sox17-regulated genes. Gene intersections are statistically significant based on hypergeometric tests* p<0.001. (B) Peak density plots showing that the Sox17-bound loci are cobound by the histone acetyltransferase Ep300, a marker of active enhancers. Average density of all peaks in top panel. (C) Sox17-bound genes are enriched in the endoderm. Stacked bar graphs showing different patterns of enriched spatial expression based on RNA-seq of dissected gastrula tissues (Blitz et al., 2017). Distribution of expression patterns statistically different from all genes in the genome was determined by two-sided KolmogorovSmirnov tests *p<0.01. (D) Top enriched DNA-binding motifs in Sox17 peaks, based on HOMER hypergeometric test. (E) Genome browser views of representative genes showing RNA-seq expression in control-MO and sox17-MO embryos and ChIPseq tracks of Sox17 and Ep300 binding. A comparison to published human ChIP-Seq data revealed that 20% of the Xenopus Sox17-bound genes were also SOX17-bound in hPSC-induced definitive endoderm (Figure 2—figure supplement 2D; Tsankov et al., 2015). GO analysis of the conserved SOX17-bound genes show an enrichment for ‘Tgfb receptor activity’ and ‘Bcat binding’ (Figure 2—figure supplement 2E), reinforcing the notion that functional interaction with the Nodal and Wnt pathways is a conserved feature of the Sox17-regulated endoderm GRN. Intersecting the RNA-seq and ChIP-seq data identified 315 genes associated with 609 Sox17-bound enhancers, which are likely to be direct transcriptional targets (Figure 2A). These putative direct targets had significantly enriched expression in the endoderm or dorsal mesendoderm (44%, 139/315 transcripts) (Figure 2C). Of the Sox17-bound and regulated genes, 197 genes (383 peaks) were positively regulated by Sox17 (down in Sox17MOs) including slc5a8, hnf1b, foxa1 and dhh (Figure 2E), all of which were previously suggested to be direct Sox17-targets. Sox17 negatively regulated 118 putative direct targets (up in Sox17MOs) including ectodermal genes (lhx5, foxi2 and tfap2a), endoderm-enriched Nodal pathway genes (nodal1, gdf3, gdf6 and mix1) and Wnt-regulated organizer genes (dkk and fst), suggesting direct transcriptional repression (Figure 2E). Interestingly, ~45% of peaks associated with Sox17-activated genes were also enriched for LIM-homeodomain binding sites, in contrast to peaks from Sox17-repressed genes which were enriched for Tbx or Pou motifs (Figure 2—figure supplement 2B). This suggests that Sox17 may coordinately engage enhancers with other core endoderm GRN TFs and mediate activation or repression of target genes depending on the interacting TFs. These analyses provide new insight into the endoderm GRN and reveal previously unappreciated roles for Sox17 in germ layer segregation and endoderm patterning involving functional interactions with the Nodal and Wnt pathways. These findings, together with our previous work demonstrating that Sox17-Bcat can physically interact in vitro (Sinner et al., 2007; Zorn et al., 1999) prompted us to their genomic interactions. β-catenin directly regulates the endodermal transcriptome To test the hypothesis that Sox17 and Bcat functionally interact in vivo, we set out to identify Bcat-regulated genes and compare these to the Sox17-regulated transcriptome. Injection of a well-characterized Bcat-MO (Heasman et al., 2000) resulted in (a) depletion of nuclear Bcat, (b) reduced expression of a transgenic Wnt-responsive reporter (Tran et al., 2010) and (c) the expected ventralized phenotype, which was rescued by co-injection of stabilized human Bcat mRNA (Figure 3—figure supplement 1). Together, this confirms the efficacy and specificity of the Bcat knockdown. RNA-seq of control-MO and Bcat-MO depleted embryos at eight time points from blastula and gastrula stages (NF7-12) identified a total of 2568 Bcat-regulated genes (>2 fold change, FDR < 5%), 1321 of which were downregulated and 1247 upregulated in the Bcat-MO embryos (Figure 3A and Supplementary file 3). Remarkably, the Bcat-dependent genes encoded 251 TFs, (~20% of all the TFs in the genome), reinforcing the notion that Wnt/Bcat initiates a transcriptional cascade in the early embryo. Figure 3 with 2 supplements see all Download asset Open asset The Bcat-regulated genomic program. (A) Time course heatmap of Bcat-regulated transcripts. Differential expressed genes from a pairwise comparison of control-MO and Bcat-MO embryos (>2 FC, FDR < 5%) at each stage identified 2568 Bcat-regulated transcripts; 1321 downregulated and 1247 upregulated in Bcat-MO embryos. Key genes are color coded based on regional expression from (D). (B) Venn diagram intersecting Bcat-regulated genes with Bcat-bound genes (and associated peaks) from published X. tropicalis gastrula Bcat ChIP-seq data (Nakamura et al., 2016) identified 898 putative direct targets associated with 2616 peaks. Statistically significant intersections based on hypergeometric tests* p<0.001.* (C) Motif enrichment analysis of Bcat peaks by HOMER hypergeometric test. (D) Stacked bar graphs showing different patterns of enriched spatial expression based on RNA-seq of dissected gastrula tissues (Blitz et al., 2017). Distribution of expression patterns statistically different from all genes in the genome was determined by twosided Kolmogorov-Smirnov tests *p<0.01. (E) Genome browser views of representative genes showing the RNA-seq expression in control-MO and Bcat-MO embryos and Bcat and Ep300 ChIPseq tracks. Intersecting the Bcat-MO RNA-seq data with previously published Bcat ChIP-seq data from Xenopus tropicalis gastrula (Nakamura et al., 2016) identified 898 putative direct target genes associated with 2616 Bcat-bound CRMs (Supplementary file 4). In the Bcat-MO, 546 genes were downregulated , and 352 genes had increased expression in Bcat-MO embryos (Figure 3B). These included almost all the previously known Bcat targets in early Xenopus embryos and had extensive overlap with other recent genomic analysis of Wnt targets in the Xenopus gastrula (Ding et al., 2017; Kjolby and Harland, 2017; Nakamura et al., 2016; Figure 3—figure supplement 2D). Control Bcat ChIP-qPCR of Bcat-MO embryos showed reduced binding to near background levels confirming a robust knockdown (Figure 2—figure supplement 1E). As expected, many of the positively regulated Bcat targets had enriched expression in the dorsal mesoderm (Figure 3D) including most known organizer genes such as cer1, chrd, dkk1, frzb, fst and gsc. In contrast, ~30% of Bcat-bound genes with increased expression in the early BcatMO embryos (NF7-9) were enriched for ectoderm-specific transcripts, consistent with the expansion of ectoderm and loss of neural tissue in ventralized embryos (Figure 3D). We also identified many direct Bcat targets that are likely to be regulated by zygotic Wnt8, (Kjolby and Harland, 2017; Nakamura et al., 2016) including sp5, axin2, cdx1/2, fzd10, msgrn1 and eight hox genes, all of which were downregulated in Bcat-MOs (Figure 3A,E). A number of zygotic Wnt-targets with enriched expression in the ventral mesoderm were components of BMP pathway including; bambi, bmp7, id2, msx1, smad7, szl, ventx1, and ventx3 (collectively known as the BMP synexpression group) (von Bubnoff et al., 2005). Most of these genes are known to be directly activated by both Bmp4/Smad1 and zWnt8/Bcat (Itasaki and Hoppler, 2010; Stevens et al., 2017) but were upregulated, rather than downregulated, in Bcat-depleted embryos, likely due to the increase in BMP/zWnt8 signaling in ventralized embryos. Most importantly, almost 200 Bcat-bound and -regulated genes had endoderm enriched expression including most of the core endoderm TFs; sox17, foxa, eomes, gata4, mix1 and mixer (Figure 3D and Supplementary file 3). Thus, contrary to the prevailing view that Bcat promotes endoderm fate primarily by activating Nodal ligand expression (Blythe et al., 2010; Hyde and Old, 2000), we find that Wnt/Bcat also directly regulates the transcription of many endodermal genes. Surprisingly, de novo motif analysis showed that the Bcat-bound peaks were more enriched for Sox than canonical Tcf motifs, 2nd versus 5th ranked, respectively (Figure 3C; Nakamura et al., 2016). Consistent with this, analysis of other previously published Bcat ChIP-Seq datsets in X. tropicalis gastrula embryos (Gentsch et al., 2019) and Flag-tagged Bcat ChIP-Seq in X. laevis gastrula (Kjolby and Harland, 2017) also report a high enrichment for Sox and Tcf motifs (Figure 3—figure supplement 2A–D). Together these data support the hypothesis that Sox17 and Bcat coregulate endodermal transcription. Sox17 and β-catenin co-occupy endoderm enhancers Intersection of the ChIP-seq datasets identified 3956 genomic loci co-occupied by both Sox17 and Bcat, and these had epigenetic signatures of active enhancers (Figure 4A,B). This represents a third of all Sox17 genomic binding in the gastrula. Comparison of the Sox17-MO and Bcat-MO RNA-seq datasets identified 415 transcripts that were regulated by both Sox17 and Bcat; comprising of ~40% of the Sox17-regulated and ~15% of the Bcat-regulated transcriptome (Figure 4A). A similar high overlap between Sox17 and Bcat genomic binding was also observed in another independent Bcat ChIP-seq dataset (Figure 4—figure supplement 1B; Gentsch et al., 2019). This suggests that co-regulation with Wnt/Bcat is a major feature of the Sox17-regulated endoderm GRN. Analysis of publicly available ChIP-seq revealed that Sox17 and Bcat-bound loci are also enriched for other endodermal TFs including VegT, Foxa4, Smad1 and Smad2 (Figure 4—figure supplement 1C; Charney et al., 2017a; Gentsch et al., 2019), consistent with recent reports that the combinatorial activity of multiple TFs at endodermal super-enhancers (Paraiso et al., 2019) coordinate endoderm specification. Figure 4 with 2 supplements see all Download asset Open asset Sox17 and Bcat co-occupy enhancers to regulate the endoderm GRN. (A) Eighty-four genes (191 Peaks) are cobound and coregulated by Sox17 and Bcat. Venn diagram show that 415 genes are regulated by both Sox17 and Bcat (left), whereas 2411 genes (3956 peaks) are bound by both Sox17 and Bcat. Gene intersections are significant based on hypergeometric tests, *p<0.001. (B) Density plots show that Sox17 and Bcat cobound peaks are enriched for enhancer associated epigenetic marks: Ep300, H3K4me1 and H3K27Ac and RNA pol II. (C) Regional enrichment of all subsets of enhancers cobound and coregulated by Sox17 and Bcat. (D) A representative ChIP-reChIP experiment showing that Sox17 and Bcat coocupy all the test CRMs (six1, dkk1, gli1 and lhx5) in the same cells, in contrast to the gtse1 locus, which is bound only by Bcat. (E) Genome browser views of representative genes showing peak overlap of Sox17, Bcat and Ep300 ChIP-Seq tracks. In total, 84 genes (191 Peaks) were bound and regulated by both Sox17 and Bcat (Figure 4A,B). This is likely to be an underestimation of the number of direct coregulated genes, as these had to pass a stringent thresholding of four independent statistical tests to be included in this list. A comprehensive motif analysis of the 191 co-bound and coregulated peaks using the CIS-BP database (Lambert et al., 2019) revealed that 53% (101/191) of the peaks contained both Sox17 and Tcf DNA-binding motifs (e.g. dkk1, lhx5 and osr1) compared to 31% in background genomic sequences (Figure 4—figure supplement 2E and Supplementary file 5). In contrast 24% (46/191) of the peaks had Sox motifs but no Tcf sites (e.g. six1), suggesting that Bcat might be recruited to these loci independent of Tcfs. Only 3/191 peaks had Tcf but no Sox motifs. Moreover, co-occupied enhancers often contained multiple Sox17 motifs (>5), which was also significantly enriched over random (Figure 4—figure supplement 2). In addition, most individual genes were associated with multiple Sox17 and Bcat co-occupied enhancers, suggesting that combinatorial binding and integration of several CRMs is crucial to control lineage-specific transcription. Since the ChIP-seq experiments were performed on whole embryos, it was possible that Sox17 and Bcat chromatin binding might occur in different cell populations. To test whether Sox17 and Bcat bound the same enhancers in the same cells we performed ChIP-reChIP experiments. First anti-Bcat ChIP was performed on 500 gastrulae. DNA eluted from the first ChIP was then re-precipitated with either anti-Sox17 or a negative control IgG antibody. All four loci tested by ChIP-reChIP qPCR (six1, dkk1, lhx5 and gli1) showed enrichment compared to IgG demonstrating that these genomic loci (~150 bp in length) were simultaneously bound by Sox17 and Bcat in endodermal cells. In contrast, a negative control locus gtse1 which is bound by Bcat but not Sox17 was not enriched in the ChIP-reChIP experiment (Figure 4D). The 84 co-bound and coregulated genes fell into four regulatory categories (Figure 4C,E); 32/84 were activated by both Sox17 and Bcat (e.g. six1 and osr1) and tended to have endoderm enriched expression, 27/84 were activated by Bcat but repressed by Sox17 (e.g. dkk1) and these tended to be enriched in the organizer mesendoderm, 19/84 were positively regulated by Sox17 and negatively regulated by Bcat (up in the Bcat MO) and finally 6/84 were negatively regulated by both Sox17 and Bcat. The observation that Bcat positively regulated most of these genes (59/84, 70%) was consistent with its known role in transcriptional activation. Sox17 positively regulated 60% of
Gene duplication results in two identical paralogs that diverge through mutation, leading to loss or gain of interactions with other biomolecules. Here, we comprehensively characterize such network rewiring for C. elegans transcription factors (TFs) within and across four newly delineated molecular networks. Remarkably, we find that even highly similar TFs often have different interaction degrees and partners. In addition, we find that most TF families have a member that is highly connected in multiple networks. Further, different TF families have opposing correlations between network connectivity and phylogenetic age, suggesting that they are subject to different evolutionary pressures. Finally, TFs that have similar partners in one network generally do not in another, indicating a lack of pressure to retain cross-network similarity. Our multiparameter analyses provide unique insights into the evolutionary dynamics that shaped TF networks.
Interleukin 7 receptor α-chain is crucial for the development and maintenance of T cells and genetically associated with autoimmune disorders including multiple sclerosis (MS). Exon 6 of IL7R encodes for its transmembrane domain and regulated by alternative splicing (AS): Inclusion or skipping of IL7R exon 6 results in membrane-bound or soluble IL7R isoforms, respectively. We previously identified SNP rs6897932 in IL7R exon 6, associated with MS risk, and showed that the risk allele (C) results in increased exon skipping and elevated sIL7R. Elevated levels of sIL7R have been shown to exacerbate the disease in the experimental autoimmune encephalomyelitis mouse model of MS. Here we report two mechanisms by which IL7R exon 6 is controlled. A competition between PTBP1 and U2AF2 at the polypyrimidine tract (PPT) of intron 5, and an unexpected U2AF2-mediated assembly of splicing factors in the exon. We noted the presence of a branchpoint sequence (BPS) (TACTAAT or TACTAAC) within exon 6, which is stronger with the C allele. Importantly, the BPS is followed by a PPT and we conjectured that silencing could be mediated by binding of U2AF2 to that tract. Here, we show that evolutionary conservation of the exonic PPT correlates well with the degree of AS of exon 6 in two nonhuman primate species and that U2AF2 binding to this PPT recruits U2 snRNP components to the exon. These observations provide the first explanation for the stronger silencing of IL7R exon 6 with the disease-associated C allele at rs6897932.
