Snail1 and Prrx1 complementary expression in development and disease. a Lateral view of 20-somite zebrafish embryos showing snail1a, snail1b, prrx1a and prrx1b expression in whole-mount (top) and transverse sections (1), showing complementary patterns in somites. b Transverse section of a zebrafish embryo in the cranial neural crest region showing complementary expression of snail1a/b (green) and prrx1a/b (red) taken at the level indicated by (2) in (a) with or without DAPI staining (nuclei). c Heatmap showing hierarchical clustering of scRNA-seq data from 18 hpf zebrafish embryos, from public database GEO: GSM3067194, with significant negative correlations between gene pairs (detailed in Supplementary Fig. 1a). d Dorsal view of HH10 chicken embryos showing PRRX1 and SNAIL1 expression in whole-mount and transverse sections at the level indicated by dashed lines, showing complementary patterns for SNAIL1 and PRRX1 In the somites (arrow) and in the LPM (splanchnopleura and somatopleura, respectively; arrowheads). e Expression of Snail1 and Prrx1 in dorsal views of E8.5 mouse embryos. Transverse sections of E8.5 embryos from the regions indicated by dashed lines (anterior and posterior, 1 and 2, respectively), showing complementary expression of Snail1 and Prrx1 in premigratory (1, arrow), and migratory (1, arrowheads) neural crest (PNC and MNC, respectively) and mesodermal populations including presomitic mesoderm (2, arrow) and lateral plate mesoderm (2, arrowhead). f Hierarchical clustering heatmaps of scRNA-seq data from E9,5 mouse embryos for Snail1 and Prrx1 expression (GEO: GSE87038), showing a significant negative correlation. This dataset contains cells from different embryonic tissues, shown in different colors. g, h Heatmaps of hierarchical clusterings of single-cell RNA sequencing data from public datasets of breast (GEO: GSE75688) and head and neck carcinoma (GEO: GSE103322), showing mutually exclusive expression of Snail1 and Prrx1. The color scale shown in (f) is also valid for (g) and (h), representing the normalized values for the number of reads. Scale bars: 250 μm for whole mounts and 100 μm for sections. Statistical analyses for scRNA-seq data has been done using Spearman r correlation test, scRNA-seq single-cell RNA sequencing, vs. versus, hpf hours post fertilization. Source data are provided as a Source Data file

SNAIL1 and PRRX1 act antagonistically on PRRX1 promoter. a Snail1 enrichment in the human PRRX1 promoter shown by ChIP assay in BT549 cells, using anti Myc antibody (for Snail1-Myc overexpression). A schematic map is shown; red diamonds represent SNAIL1 potential binding sites (E-boxes; CANNTG) on PRRX1 promoter. (BS1: –1689, BS2: −4102, BS3: −6753, BS4: −7277, BS5: −7318 and NC: +104560). Ex1: PRRX1 exon 1 (n = 4). b Activity of different regions of the human PRRX1 promoter after SNAIL1 transfection assessed by luciferase assays in HEK293 cells (n = 4, except for BS1–5 for which n = 8). c qPCR assay showing downregulation of PRRX1 transcription upon SNAIL1 transfection in BT549 cells (n = 4). d PRRX1 directly binds to its own promoter, as assessed by ChIP assays in BT549 cells using a PRRX1 specific antibody. A schematic map is shown; cyan diamonds represent PRRX1 potential binding sites (TAATKDS) on its own promoter. (BS7: −6875, BS6: −5579, BS5: −5202, BS4: −4147, BS3: −4096, BS2: −1653, BS1: −1197 and NC: +104560). Ex1: PRRX1 exon 1 (n = 6, except for BS1–2 for which n = 4). e Activity of different regions of the human PRRX1 promoter after PRRX1 overexpression in luciferase assays in HEK293 cells (n = 4). f Lack of enrichment for PRRX1 binding to human SNAIL1 promoter assessed by ChIP assays in BT549 cells (BS4–1: −1753, −1621, −954, −198, and NC: +11799). Ex1: SNAIL1 exon 1 (n = 4). Locations of red and cyan diamonds represent distances between BSs and the promoter. Arrows represent primers used for qPCR amplification. BS biding site, NC negative control region, which does not contain potential BSs. Symbols in binding sites are as follows. K: T/G, D: G/A/T, S: G/C and N: G/A/T/C. Bars represent mean plus standard error of the mean (SEM), indicated (n) represent number of independent experiments as biological replicates and asterisks indicate significant p value in t test compared to the control in each test for (a), (e), (d), (f) and ANOVA with Bonferroni’s multiple comparison test for (b), (e) (*p < 0.05, **p < 0.01 and ***p < 0.001). Source data are provided as a Source Data file

Prrx1 directly induces the expression of miR-15 family. a Heatmap showing Robust Multi-array Average (RMA) normalized probe intensity values of PRRX1, SNAIL1, and selected miRNAs in MDA231 cells, with three control samples (MDA_C1 to C3) and three samples in which PRRX1 was overexpressed (MDA_P1 to P3). The intensities mapped as color scale show normalized fold change with respect to the average (n = 3). b Validation by Taq-Man qPCR of Mir-15-P1d and Mir-15-P2d upregulation after PRRX1 stable overexpression in MDA231 cells lines, using specific probes to detect the mature miRNAs (n = 4). cMir-15-P1d_pre and Mir-15-P2d_pre upregulation upon conditional (Dox-mediated) Prrx1 overexpression in MDA231 cells (n = 6). d qPCR assay showing downregulation of PRRX1 and premiRNAs transcription upon transient knockdown (KD) of PRRX1 in BT549 cells. PRRX1 shRNA plus YFP transfected cells were sorted after 4 days (n = 4). e PRRX1 directly binds to the human Mir-15-P1/2d promoter as shown by ChIP assay in BT549 cells using a PRRX1 specific antibody. Cyan diamonds in the schematic map represent distances between PRRX1 potential BSs and the promoter potential motifs, TAATKDS, on Mir-15-P1/2d promoter. (BS1: −293, BS2: −949, BS3: −1451 and NC: +1893). Ex1–3 represent exons of the host long noncoding RNA MIR503HG. Arrows represent primer sets used for ChIP detection. (n = 3 except for BS2–3 for which n = 4). f Activation of human Mir-15-P1/2d promoter by PRRX1 overexpression shown by dual luciferase assay in HEK293 cells. This activation is abolished upon deletion of the PRRX1 binding sites in Mir-15-P1/2d promoter (del1/2 + PRRX1) (n = 4 except for deletions for which n = 4). g Transverse sections of the cranial region of E8.5 embryos showing the expression of Prrx1, Mir-15-P1d, Mir-15-P2d, Mir-15-P1b, and Mir-15-P2a/b in similar regions (arrowheads). Scale bar: 100 μm. Dox doxycycline, sh short hairpin RNA (shRNA), BS binding site, del deletion. Bars represent mean plus SEM, indicated (n) represent number of independent experiments as biological replicates and asterisks indicate significant p value in t test for (b–e) and one-way ANOVA with Bonferroni’s multiple comparison test for (f). (*p < 0.05, **p < 0.01 and ***p < 0.001). Source data are provided as a Source Data file

Prrx1 attenuates Snail1 expression through the activation of miR-15 family members. a qPCR assay showing the downregulation of SNAIL1 upon conditional overexpression of Prrx1 in MDA231, induced by doxycycline after 48 h (n = 7). b qPCR assay showing downregulation of SNAIL1 upon transient overexpression of Mir-15-P1/2d in MDA231 cells, using seed mutated miRNAs as control (Mir-15-P1/2d-mut) (n = 6). c qPCR assay showing upregulation of SNAIL1 upon transient KD of in Mir-15-P1/2d in BT549 cells using miRNA LNA inhibitors (n = 5). d Repression of human SNAIL1 3′ UTR after Mir-15-P1/2d overexpression, using seed mutated miRNAs (Mir-15-P1/2d-mut) or MRE mutant SNAIL1 3-UTR (MRE-mut) as control (n = 5). e qPCR analysis shows upregulation of snail1a/b expression in the sponge-injected (miR-15 family knocked-down) zebrafish embryos compared to controls (n = 6). f 20-somite control or miR-15 family sponge-injected zebrafish embryos showing combined snail1a/b expression by in situ hybridization. Transverse sections taken at the levels indicated by the dashed lines. Scale bar: 250 μm. g Fluorescent miRNA in situ hybridization to detect mature miRNAs in transverse sections of 20-somite zebrafish embryo, showing the expression of Dre-miR-15-P1/2a/b in neural crest derived cell populations. Scale bar: 250 μm. h Western blot for Snail1 and β-actin from E11.5 control and Prrx1 mutant embryos, including quantification of intensity of Snail1 signal (n = 5 WT, 4 KO). i Snail1 IF in the nasal pit regions of E11.5 WT and mutant embryos. Scale bar: 250 μm for NP sections and 100 μm for insets (boxes 1 and 2). MRE miRNA responsive element, LNA locked nucleic acid, WT wild type, Mut mutant, NT neural tube, NP nasal pit. Quantifications are performed for one section of WT or mutant embryos, and the increase and expansion is observed in n = 2/2 mutant embryos compared to n = 3/3 different E11.5 WT. Bars represent mean plus SEM, indicated (n) represent number of independent experiments as biological replicates and asterisks indicate significant p value in t test for a, c, e, h and one-way ANOVA with Bonferroni’s multiple comparison test for (b), (d) (*p < 0.05, **p < 0.01 and ***p < 0.001). Source data are provided as a Source Data file

Relationship between the expression of SNAIL1, PRRX1, and miR-15 family members and overall survival in breast cancer patients. Kaplan–Meier overall survival (OS) plots from breast cancer patients with lymph-node positive status, showing that high expression of SNAIL1 correlates with low survival, while PRRX1 high expression correlates with a better survival. The expression of miR-15 family members follow a similar trend as that of PRRX1. Hazard ratio (HR) and logarithmic ranked p Value (longrank P) were analyzed to infer the significance of the differences. Numbers below each graph represent number of patients at risk in any given time (months), black for low expression and red for high expression of each gene/miRNA. The cut-off is automatically calculated based on the best performing threshold

Snail1 and Prrx1 are sequentially expressed during EMT. a qPCR assay showing the expression of prrx1a/b and snail1a/b (left) and premiRNAs (right) in zebrafish embryos after different treatments with BMP (n = 8). Asterisks indicate significant p value in one-way ANOVA with Bonferroni’s multiple comparison test compared to the control. (*p < 0.05, **p < 0.01 and ***p < 0.001). Source data are provided as a Source Data file. b Dorsal view of HH9 chicken embryos showing SNAIL1 and PRRX1 expression in control embryos and at different times after BMP-soaked bead implantation. c Transverse sections of E8.5 DES-EGFP reporter mouse embryos showing IFs for GFP and Snail1 (upper panel) or Prrx1 (lower panel). Insets show higher magnification pictures for GFP and Snail1 (1 and 2), or for GFP and Prrx1 (3 and 4). d Schematic model of the gene regulatory network. DES downstream enhancer of Snail1, PNC premigratory neural crest shown by arrows, MNC migratory neural crest. Scale bars: 100 μm. Scale bars: 100 μm for sections and 10 μm for insets (boxes 3–4)