Skip to main content
. 2019 Nov 20;8:e49921. doi: 10.7554/eLife.49921

Figure 2. Cardiopharyngeal accessibility profiles are established in multipotent progenitors.

(A–B) Correlations between differential gene expression (DE) and differential chromatin accessibility (DA) in response to FGF-MAPK perturbation in the multipotent progenitors (A) and between chromatin accessibility in response to FGF-MAPK perturbation and in multipotent progenitors (10 hpf) versus founder cells (6 hpf). (B). Colored dots are DA peaks associated with cell type-specific DE genes. ρ is the Spearman correlation of expression and accessibility for DA regions associated with DE genes (A) or of region response to MAPK perturbation with accessibility in founder cells versus multipotent progenitors (B). (C) Relationship between expression and accessibility of DE genes associated with DA regions for genes in the bottom 0.75% quantile of fold change between expression in FgfrDN and control (log2(FC) < −1.32). Microarray log2(fold change (FC)) values are shown on the left. The fold change for all time points is versus the average. (D) A 24 kb region on chromosome eight displaying expression (RNA-seq) and chromatin accessibility (ATAC-seq; normalized by total sequencing depth). Gray shaded boxes show validated ATM-specific promoters and a newly identified TVC-specific enhancer in Nk4/Nkx2-5 intron. (E) Enhancer-driven in vivo reporter expression (green) of tested ATAC-seq regions (KhC8.2200 and .2201). TVCs marked with Mesp>H2B::mCherry (red). Numbers indicate observed/total of half-embryos scored. (F) Endogenous expression of Nk4/Nkx2-5 visualized by in situ (green) in TyrosinaseCRISPR and upon CRISPR/Cas9-induced deletions of TVC-specific region. Nuclei of B7.5 lineage cells are labelled by Mesp>nls::LacZ and revealed with an anti beta-galactosidase antibody (red). Nk4/Nkx2-5 expression was not affected in the epidermis (open arrowhead). Experiment performed in biological replicates. Scale bar, 20 μm. Fisher exact test, total numbers of individual halves scored per condition are shown in 'n='. Gene expression data for 6 hpf and ‘FGF-MAPK perturbation 10 hpf’ (Christiaen et al., 2008) and 8 to 20 hpf (Razy-Krajka et al., 2014) were previously published.

Figure 2.

Figure 2—figure supplement 1. Inhibition of FGF signaling (Mesp>FgfrDN−10 hpf) induces opening of ATM-specific elements.

Figure 2—figure supplement 1.

(A) Relation between expression and accessibility for DE genes in the top 99.25% quantile of fold change between expression in FgfrDN and control at 10 hpf (log2(FC) > 1.31). (B, C) A 10 kb region on chromosome 6 and 10 including Synpo1/2 (B) and KH.C10.125 (C) displaying chromatin accessibility profiles from ATAC-seq (normalized by total sequencing depth). The newly identified enhancers are in boxed regions. ATAC-seq peaks were tested in vivo by reporter gene assay (see Supplementary file 4). Reporter gene assay in embryos at stage 23 (Hotta et al., 2007). GFP expression was detected specifically in ATM cells and not in TVCs. Nuclei of B7.5 lineage cells are labelled by Mesp>H2B::mCherry and Mesp>hCD4::mCherry (red). Numbers indicate observed/total of half-embryos scored. Scale bar = 30 μm.
Figure 2—figure supplement 2. General characterization of differential accessibility.

Figure 2—figure supplement 2.

(A) Spearman correlation of expression of bulk RNA-seq or microarray (for B7.5 6 vs. 10 hpf and Mesp>FgfrDN vs. Mesp>LacZ - 10 hpf) pairwise comparisons with ATAC-seq pairwise comparisons. Correlation was calculated based on log2(FC) of differentially expressed genes associated with differentially accessible peaks for each comparison. Differentially expressed genes in Mesp>FgfrDN vs. Mesp>LacZ at 10 hpf derived from Christiaen et al. (2008), Hand-r>FgfrDN vs. Hand-r>LacZ and Foxf>M RasCA vs. Foxf>LacZ at 18 hpf from Wang et al. (2019), FoxfCRISPR vs. ControlCRISPR at 10 hpf from the present study. (B) GSEA for ATAC-seq pairwise comparisons. A negative normalized enrichment score (NES) indicates elements annotated to that gene set are less accessible in the comparison. A positive NES indicates elements annotated to that gene set are more accessible in the comparison. (C) One-tailed hypergeometric test for enrichment of accessible elements overlapping a genomic feature. The leftmost column shows the expected distribution of accessible elements from the accessome.
Figure 2—figure supplement 3. Peakshift validation of sgRNA efficiency.

Figure 2—figure supplement 3.

(A) Chromatin accessibility profiles from ATAC-seq (normalized by total sequencing depth). Transcript model is indicated as black bar. The newly identified TVC-specific enhancer is in boxed regions; in the zoom region, blue arrows indicate primers used to amplify the region between the target sites. In wild-type embryos, the resulting PCR product is ~1.1 kilobase pairs. The two single guide RNAs used to target Nk4/Nkx2-5 intronic element are in red. (B) Alignment of cloned PCR products amplified using the primers indicated in (A), from wild-type (wt) embryos, and from embryos electroporated with 25 µg Ef1α>nls::Cas9-gem::nls and 40 µg each of U6>sgRNA2 and U6>sgRNA4. Clone '01' contains a ~0.6 kilobase deletion between the approximate sites targeted by the two sgRNAs. (C) Quantification of indel-shifted electrophoresis chromatogram peaks (‘Peakshift’ assay; Gandhi et al., 2017) revealed sgRNA mutagenesis efficacies. (D) In situ hybridization for Nk4 (green) showing expression throughout the ventral head endoderm in embryos electroporated with Mesp>H2B::mCherry (red), Mesp>nls::Cas9-Gem::nls and U6>sgRNAs targeting Nk4 intron (Nk4_KhC8.2200-.2201CRISPR) or Tyrosinase (TyrosinaseCRISPR), used as control. In control embryos (panels on the left) Nk4 expression is essentially wild-type with a strong expression in ventral head endoderm (asterix and arrowhead) and TVCs. In Nk4_KhC8.2200-.2201CRISPR embryos (panels on the right) Nk4 expression is lost specifically in the B7.5 lineage and not in the other endogenous territories. Scale bars = 30 µm.
Figure 2—figure supplement 4. Candidate TVC-specific enhancers in vivo validation by reporter gene assay.

Figure 2—figure supplement 4.

(A–G) ATAC-seq peaks specifically accessible in TVC displaying chromatin accessibility profiles from ATAC-seq (normalized by total sequencing depth). Transcript model is indicated as black bar. The newly identified enhancers are in boxed regions. ATAC-seq peaks were tested in vivo by reporter gene assay (see Supplementary file 4). (A’–G’’) Reporter gene assay in embryos at stage 23 (Hotta et al., 2007). GFP expression was detected specifically in TVC cells but not in ATMs. Nuclei of B7.5 lineage cells are labelled by Mesp>H2B::mCherry and Mesp>hCD4::mCherry (red). Numbers indicate observed/total of half-embryos scored. Scale bar = 30 μm.
Figure 2—figure supplement 5. Candidate TVC-specific enhancers in vivo validation by CRISPR/Cas9.

Figure 2—figure supplement 5.

(A–C) TVC-specific enhancers in Fbln, Smurf and Fgf4 loci targeted by two or three single guide RNA (sgRNAs) for CRISPR/Cas9-mediated deletions. Upper panel is showing chromatin accessibility profiles from ATAC-seq (normalized by total sequencing depth). Transcript model is indicated as black bar. (D–F) Endogenous expression of Fbln (D), Smurf (E) and Fgf4 (F) visualized by in situ (green) in TyrosinaseCRISPR (right panel) and upon CRISPR/Cas9-induced deletions of TVC-specific peaks (left panel). Nuclei of B7.5 lineage cells are labelled by Mesp>nls::LacZ and revealed with an anti beta-galactosidase antibody (red). Anterior to the left, stages indicated as 'st.'. Experiment performed in biological replicates. Scale bar = 20 μm. All the treatments were significant versus TyrosinaseCRISPR (Fisher exact test, p < 0.001); ‘n’: total numbers of embryos.
Figure 2—figure supplement 6. CRISPR validation on the TVC-specific Fgf4 enhancer.

Figure 2—figure supplement 6.

(A) Chromatin accessibility profiles from ATAC-seq (normalized by total sequencing depth). Transcript model is indicated as black bar. The newly identified TVC-specific enhancer is in boxed regions; in the zoom region, blue arrows indicate primers used to amplify the region between the target sites. In wild-type embryos, the resulting PCR product is ~1.2 kilobase pairs. The two single guide RNAs used to target Fgf4 intronic element are in red. (B) Quantification of indel-shifted electrophoresis chromatogram peaks (‘Peakshift’ assay; Gandhi et al., 2017) revealed sgRNA mutagenesis efficacies. (C–D) 1% Agarose gel and alignment of cloned PCR products showing the result of Fgf4 intron-I amplification with the oligos indicated in (A) from embryos electroporated with 25 µg Ef1α>nls::Cas9-gem::nls and 40 µg each of U6>sgRNA6 and U6>sgRNA8. (C) Alignment of cloned PCR products amplified using the primers indicated in (A), from the ~1.2 kilobase band that is similar to the control. (D) Alignment of cloned PCR products amplified using the primers indicated in (A), from the ~0.5 kilobase band that corresponds to the expected deletion between the approximate sites targeted by the two sgRNAs. The intronic element that is excluded from the sites targeted by the two sgRNAs remained intact (yellow box).