Fig 2: Global analysis of Etv4 and Etv5 binding regions in the genome. (A) Expression of Etv4 and Etv5 by RNA-seq of EpH4 and EpRas cells treated with 1 ng/ml TGF-ß for 48 h. FPKM, fragments per kilobase of exons per million mapped fragments. The experiment was performed in two biological replicates. Data are a subset of the findings shown in Supplementary Fig. S2C. (B) Immunoblot analysis of endogenous Etv4 protein expression in EpH4 and EpRas cells. pSmad2, phosphorylated Smad2. (C) Relationship between the anti-Etv4 and anti-Etv5 ChIP-seq data and FAIRE-seq data obtained from TGF-ß-treated EpRas cells. (D) De novo motif prediction identifies ETS family (ELK4) binding motif, and shown as in Fig. 3A. Anti-Etv4 ChIP-seq data were used for enriched motif calculation. q value (minimal false discovery rate required to include the motif) is shown. (E) Etv4 and Etv5 binding regions at the Cdh1 and Fn1 gene loci. HBB cluster region is shown as a control. Black bars represent the significant Etv4 and Etv5 binding regions. Significant FAIRE-positive regions in EpH4 and EpRas cells, without (first row) and with (second row) TGF-ß stimulation, are shown as blue and red bars, respectively, as a reference (shown in Fig. 2A).

Fig 3: Effect of Etv4 and Etv5 siRNAs on stress fiber formation, cell invasiveness and global gene expression in EpRas cells. (A) Effect of the siRNAs on the expression of Etv4 protein. EpRas cells were transfected with the siRNAs as indicated. After 24 h of incubation, cells were treated with TGF-ß for additional 48 h and lysed for immunoblotting analysis. (B) Effect of Etv4/5 siRNAs on the TGF-ß-induced down-regulation of E-cadherin protein. siRNA-transfected cells were treated with TGF-ß for 8 days. Transfection of siRNAs were repeated at day 2 and 5. (C) Effect of Etv4/5 siRNAs on the stress fiber formation induced by TGF-ß. F-actin formation of EpRas cells was evaluated by phalloidin staining. Cells were transfected with siRNAs for Etv4/5 and stimulated with TGF-ß for 48 h. (D) Matrigel invasion assay in EpRas cells. Cells were transfected with the indicated siRNAs and then seeded on the Matrigel-coated plate, incubated for 48 h with TGF-ß, and fixed. (upper panels) Representative images of the cells that migrated through the Matrigel-coated membrane. (bottom graph) Quantified data representing the means of four independent experiments. Error bars, standard deviations. (E and F) EpRas cells were transfected with siRNAs for Etv4/5 and treated with 1 ng/ml TGF-ß for 48 h. Then, RNA-seq and ontology analysis using GSEA were performed. Genes with maximum FPKM (fragments per kilobase of exon per million mapped sequence reads) values =5 among the samples were selected for evaluation. (E) A list of top-enriched c5 (gene ontology) gene sets (MSigDB) downregulated by siEtv4/5 in the absence of TGF-ß. The averaged gene expression data from the two distinct sets of siEtv4/5-transfected cells were compared to the data from the siNC-transfected cells. SIZE: number of genes in the phenotype, NES: normalized enrichment score, FDR: false discovery rate?. (F) An enrichment plot showing the up-regulation of a gene set “EXTRACELLULAR REGION” by Etv4 and Etv5, which was identified as the top-enriched phenotype by GSEA.

Fig 4: Adult human nerve- and skin-derived SCs are phenotypically indistinguishable in vitro and express proteins associated with precursor and immature stages of SC development. A–D, Representative immunocytochemical confocal images (46-year-old male) of adult human nerve (hN-SC)- and skin (hSk-SC)-derived SCs (NES+, green) and adult human skin-derived dermal fibroblasts (hSk-Fibro, FN1+, green) cultured in identical conditions. Immunocytochemistry revealed that SCs isolated from both human skin and nerve express several precursor and immature SC antigens (red, arrowheads), including (A) PAX3, (B) CDH19, (C) ETV5, and (D) SOX2. E, Quantification of the percentage of Hoechst+ cells expressing SOX2 in SC cultures. F, hSk-SCs also expressed the proliferative marker, ki67, in a similar percentage of cells to hN-SCs. Each dot represents the mean percentage within a patient culture (27-, 28-, 37-, and 46-year-old male) across three to four replicates (mean ± SEM, unpaired t test; scale bar, 25 µm).

Fig 5: Regulation of Mmp13 expression by Etv4 and Etv5. (A) FAIRE-seq and ChIP-seq data at the matrix metalloproteinase 13 (Mmp13) gene locus as a target of Etv4 and 5. Data are shown as in Figs 2A and 4E. Arrow heads show the positions evaluated by FAIRE-qPCR in (D). (B) RNA-seq data of siEtv4/5-transfected EpRas cells treated with TGF-ß. siNC; negative control siRNA. Error bars, 95% confidence intervals. (C) Matrigel invasion assay in EpRas cells. Cells were transfected with the indicated siRNAs and then seeded on the Matrigel-coated plate, incubated for 48 h with TGF-ß, and fixed. (upper panels) Representative images of the cells that migrated through the Matrigel-coated membrane. (bottom graph) Quantified data representing the means of four independent experiments. Error bars, standard deviations. (D) Changes in the FAIRE-seq signal intensities by Etv4/5 siRNAs at FAIRE-positive regions of Mmp13 gene locus shown in (A). EpRas cells transfected with the indicated siRNAs were treated with TGF-ß for 4 days and fixed for FAIRE-seq data acquisition. The data represent the result of two biological replicates. Error bars, standard deviations.