Journal: bioRxiv

Article Title: Hierarchical TBX6-FOXC Regulatory Logic Controls Human Trunk Mesoderm Diversification

doi: 10.64898/2026.03.17.712295

Figure Lengend Snippet: (A) UMAP of 8,088 mesodermal subset cells, with the somite progenitor cluster (as in ) highlighted. (B) Volcano plot of genes differentially expressed in the somite progenitor cluster compared with the remaining mesodermal subset. Genes with padj < 0.05 and |log2FoldChange| > 0.58 (±1.5-fold) are highlighted (red, upregulated; blue, downregulated). The top 20 upregulated transcription factors according to log2FC are shown in darker colour. Box highlights FOXC1 and FOXC2. (C) Inducible dual FOXC1 and FOXC2 CRISPRi strategy. Two piggyBac donor constructs each encoding a Dox-inducible dCas9-KRAB-IRES-GFP cassette and a single gRNA targeting either FOXC1 (with PuroR), or FOXC2 (with NeoR) were co-integrated into iPSCs using piggyBac transposase. Dual antibiotic selection yields a single cell line harbouring both cassettes. Dox treatment induces dCas9-KRAB expression, enabling simultaneous knockdown of FOXC1 and FOXC2 during somitic mesoderm differentiation. (D) Experimental design to test dual knockdown of FOXC1 and FOXC2 in a 2D somitic mesoderm differentiation protocol (adapted from Loh et al., ), followed by bulk RNA-seq. (E) Volcano plot of differential gene expression following dual knockdown of FOXC1 and FOXC2 in 2D somitic mesoderm differentiation. Genes with padj < 0.05 and |log2FoldChange| > 0.58 (±1.5-fold) are highlighted (red, upregulated; blue, downregulated). Selected genes associated with somitic, kidney, and muscle identities are highlighted in darker colours. (Somitic: RIPPLY1, SIX1, FOXC1, FOXC2, MEOX1, MEOX2, TCF15, PAX3 ; kidney: OSR1, NPHS1, NPHS2, EYA2, PAX2, EPCAM, CDH1, CDH6, TWIST1, JAG1, HEYL, HES5 ; muscle MYMX, MYH11, MYBPC2, MYLK3, MYH7B, MYOF ). (F) Dot plot showing enrichment of transcription factor motif families identified in regulatory regions of FOXC1/2-dependent genes versus expressed non-dependent genes (defined in (E)), using early somite ATAC-seq data . Dot position indicates the number of significantly enriched motifs per family (BH-adjusted q < 0.01); dot size reflects total motifs representation in HOMER; colour denotes the significance (-log 10 q) of the most enriched motif in each family. (G) Experimental design to test dual knockdown of FOXC1 and FOXC2 in hTLS. (H-J) Quantification of the number of nuclei per confocal image that are (G) SOX2 + , (H) PAX8 + , or (I) PAX3 + SOX2 - , in FOXC1/2-dCas9-KRAB hTLS either untreated (control, white bars) or treated with Dox from day −1 (Dox, coloured bars) (n=16 hTLS across 2 independent batches per group). Dots represent individual hTLS; bars represent median ± IQR; Mann-Whitney test (*** p < 0.001). (K) Immunofluorescence of hTLS generated from FOXC1/2-dCas9-KRAB iPSCs and harvested at day 5, either untreated (control, top) or treated with Dox from day −1 (1 µM, bottom), showing GFP (green), SOX2 (yellow), PAX3 (magenta) and PAX8 (grey). Dotted white lines outline PAX3 + SOX2 + neural structures, white arrowheads indicate PAX3 + SOX2 - somitic tissue depleted upon Dox treatment. Scale bars, 100 µm. ITR, inverted terminal repeat; TRE3G, third-generation tetracycline-responsive promoter; cHS4, chicken hypersensitive site 4 insulator; pA, polyadenylation signal; PuroR, puromycin resistance gene; NeoR, neomycin resistance gene. See also .

Article Snippet: FOXC1 and TWIST1 transgenes were synthesised by Twist Biosciences. mCherry was synthesised by Vectorbuilder (VB240515-1632gka).

Techniques: Construct, Selection, Single Cell, Expressing, Knockdown, RNA Sequencing, Gene Expression, Control, MANN-WHITNEY, Immunofluorescence, Generated

Journal: bioRxiv

Article Title: Hierarchical TBX6-FOXC Regulatory Logic Controls Human Trunk Mesoderm Diversification

doi: 10.64898/2026.03.17.712295

Figure Lengend Snippet: (A) Experimental design to test dual FOXC1 and FOXC2 knockdown in a 2D sclerotome differentiation protocol (adapted from Loh et al., ), followed by RT-qPCR. (B-D) RT-qPCR for B) early somite markers, C) FOXC1 and FOXC2, and D) dermomyotome and sclerotome markers and canonical Hedgehog target genes, in FOXC1/FOXC2-dCas9-KRAB iPSCs differentiated to early somite and then either untreated (control) or treated from day 3 with doxycycline (Dox, B-D), or also with smoothened agonist (SAG), or combined Dox and SAG treatment (C,D). Two-way ANOVA on log-transformed values, with treatment as the column factor and differentiation batch as the row factor. p values correspond to the main effect of treatment (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001). Data are plotted as 2 -ΔΔCt ; bars represent mean ± SEM; dots represent mean of technical replicates from n=5 independent differentiation batches per group. (E) Schematic of the FOXC1 overexpression strategy (F) Schematic of experimental design to overexpress FOXC1 in 2D-cultured hTLS for subsequent RT-qPCR. (G) Schematic of FOXC1 protein showing domain architecture, together with RT-qPCR analysis of FOXC1 transgene, TWIST1 , MEOX1 , PAX9, TWIST2 and SOX9 expression in hTLS derived from PB-Dox-FOXC1 iPSCs at day 5, either untreated (control) or treated with Dox (0.1 µM) from day 2. Statistical significance was assessed using two-way ANOVA on log-transformed values, with treatment as the column factor and differentiation batch as the row factor. The p values shown correspond to the main effect of treatment (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001). Data are plotted as 2 -ΔΔCt for visualisation; bars represent mean ± SEM; dots represent mean of technical replicates from n=4 independent differentiation batches per group. N-TAD, N-terminal transactivation domain; C-TAD, C-terminal transactivation domain. (H) Schematic of FOXC1 [ΔFKH] protein showing domain architecture with FKH DNA-binding domain replaced with a flexible linker; plots show RT-qPCR analysis for transgene and TWIST1 in day 5 2D-hTLS generated from iPSCs harbouring Dox-inducible FOXC1 [ΔFKH] and treated with varying concentrations of Dox (0, 0.01, 0.1, 1 µM) from day 2:. N-TAD, N-terminal activation domain; FKH, forkhead domain; C-TAD, C-terminal activation domain; grey box, putative inhibitory domain. Statistical significance was assessed using two-way ANOVA on log-transformed values (ΔCt), with treatment as the column factor and differentiation batch as the row factor. The p values shown correspond to the main effect of treatment (** p < 0.01, **** p < 0.0001). Data are plotted as 2 -ΔΔCt ; bars represent mean ± SEM; dots represent mean of technical replicates from n=2-3 independent differentiation batches per group. (I) Experimental design to overexpress FOXC1 in hTLS (J) Brightfield and corresponding red fluorescence images of hTLS generated from PB-Dox-FOXC1 iPSCs at days 3, 4, and 5. Scale bars, 300 µm. (K) Immunofluorescence of hTLS generated from PB-Dox-FOXC1 iPSCs following FOXC1 overexpression from day 2, stained for mCherry (magenta) and F-actin (phalloidin, grey). White dotted box indicates region shown at higher magnification, in which individual epithelial cysts are outlined with dotted lines. Arrows indicate FOXC1-overexpressing cells occupying a subepithelial position. Scale bars: main, 100 µm; insets, 20 µm. (L) Immunofluorescence of hTLS generated from PB-Dox-FOXC1 iPSCs following FOXC1 overexpression from day 2, stained for mCherry (magenta), PAX3 (cyan), and SOX2 (yellow). Arrows indicate FOXC1-overexpressing cells lacking both PAX3 and SOX2. Scale bars: main, 100 µm; insets, 20 µm. (M) Immunofluorescence of hTLS generated from PB-Dox-FOXC1 iPSCs following FOXC1 overexpression from day 2, stained for mCherry (magenta), PAX8 (grey), and TWIST1 (green). Arrows indicate FOXC1-overexpressing cells expressing TWIST1. Scale bars: main, 100 µm; insets, 20 µm. (N) Immunofluorescence of hTLS generated from PB-Dox-FOXC1[ΔFKH] iPSCs following overexpression from day 2, stained for mCherry (magenta), PAX3 (cyan), and F-actin (phalloidin, grey). White dotted box indicates region shown at higher magnification, in which individual epithelial cysts are outlined with dotted lines. Arrows indicate FOXC1[ΔFKH]-overexpressing cells within epithelial cysts. Scale bars: main, 100 µm; insets, 20 µm. See also and .

Article Snippet: FOXC1 and TWIST1 transgenes were synthesised by Twist Biosciences. mCherry was synthesised by Vectorbuilder (VB240515-1632gka).

Techniques: Knockdown, Quantitative RT-PCR, Control, Transformation Assay, Over Expression, Cell Culture, Expressing, Derivative Assay, Binding Assay, Generated, Activation Assay, Fluorescence, Immunofluorescence, Staining

Journal: bioRxiv

Article Title: Hierarchical TBX6-FOXC Regulatory Logic Controls Human Trunk Mesoderm Diversification

doi: 10.64898/2026.03.17.712295

Figure Lengend Snippet: (A) Schematic of sclerotome differentiation in vivo . Hedgehog ligands secreted by the notochord signal to the ventral domain of epithelial somites to activate sclerotome-associated genes including FOXC1, FOXC2, TWIST1, PAX1, PAX9, and SOX9, to downregulate PAX3, and to initiate epithelial-to-mesenchymal transition (EMT), while dorsal somites remain PAX3 + . (B) UMAP plots of 13,618 hTLS snRNA-seq cells (as in ) showing minimal expression of Hedgehog ligands ( SHH , IHH , DHH ), and expression of downstream Hedgehog pathway mediators ( GLI1 , GLI2, GLI3 ). (C) Experimental design to test the effect of Hedgehog pathway activation using smoothened agonist (SAG; 500 nM) in hTLS from day 5 to day 8 of the protocol. (D) Immunofluorescence of SOX2 (yellow), FOXC1 (grey), and PAX9 (cyan) expression in day 8 hTLS cultured as in (C), either untreated (control, top row) or treated with SAG from day 5 (500 nM, bottom row). Scale bars, 100 µm. (E-G) Quantification of the number of nuclei per confocal image positive for (E) SOX2, (F) FOXC2, or (G) PAX3 (whilst also SOX2 - ) in day 8 hTLS either untreated (control, white bars) or treated with SAG from day 5 (500 nM, coloured bars). Dots represent individual hTLS (n = 12 hTLS across 2 independent differentiation batches); bars represent median ± IQR; Mann-Whitney test (** p < 0.01, *** p < 0.001). (H) Immunofluorescence of mCherry (magenta), FOXC2 (grey), and SOX9 (green) expression in day 5 hTLS generated from PB-Dox-FOXC1 iPSCs treated with Dox from day 2 (0.1 µM). White box indicates magnified region; white arrows indicate mCherry + FOXC2 + cells expressing SOX9; arrowheads indicate a rare mCherry - FOXC2 + SOX9 + cell. Scale bars: main, 100 µm; insets, 20 µm. (I) Schematic of the Dox-inducible FOXC2-overexpression strategy and corresponding immunofluorescence of SOX2 (blue), mCherry (magenta), PAX8 (grey), and PAX3 (green) expression in a day 5 hTLS generated from PB-Dox-FOXC2 iPSCs and treated with Dox from day 2 (0.1 µM). White box indicates magnified region; the dotted lines outline an mCherry + , PAX8 - , PAX3 - cell cluster positioned between PAX8 + and PAX3 + tissue. Scale bars: main, 100 µm; insets, 20 µm.

Article Snippet: FOXC1 and TWIST1 transgenes were synthesised by Twist Biosciences. mCherry was synthesised by Vectorbuilder (VB240515-1632gka).

Techniques: In Vivo, Expressing, Activation Assay, Immunofluorescence, Cell Culture, Control, MANN-WHITNEY, Generated, Over Expression

Journal: bioRxiv

Article Title: Hierarchical TBX6-FOXC Regulatory Logic Controls Human Trunk Mesoderm Diversification

doi: 10.64898/2026.03.17.712295

Figure Lengend Snippet: (A) Schematic of a piggyBac-insertable construct for Dox-inducible overexpression of TWIST1. (B) Experimental design to test TWIST1 overexpression in hTLS. (C) Immunofluorescence of hTLS generated from PB-Dox-TWIST1 iPSCs showing mCherry (magenta), FOXC2 (cyan), F-actin (phalloidin, yellow) and SOX2 (blue) expression at day 5 following Dox treatment from day 2. The white box indicates magnified region. Dotted white lines in magnified panels indicate FOXC2 + epithelial structures, and white arrows indicate mCherry + TWIST1-overexpressing cells occupying a subepithelial position. Scale bars: main, 100 µm; insets, 20 µm. (D) Experimental design to test dual TWIST1 and TWIST2 knockdown in a 2D sclerotome differentiation protocol (adapted from Loh et al., ), followed by RT-qPCR. (E) RT-qPCR analysis of FOXC1, FOXC2, TWIST1, TWIST2, PAX9, PAX1 and PTCH1 expression in TWIST1- and TWIST2-dCas9-KRAB iPSCs differentiated to sclerotome and either untreated (control), or treated from day 3 with Dox, SAG, or combined Dox and SAG treatment. Dots represent mean of technical replicates from n = 3 independent differentiation batches per condition. Two-way ANOVA on log-transformed values (ΔCt), with treatment as the column factor and differentiation batch as the row factor. P values correspond to main treatment effect (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001). Data are plotted as 2 -ΔΔCt for visualisation, normalised to the untreated control condition within each iPSC line; bars represent mean ± SEM.

Article Snippet: FOXC1 and TWIST1 transgenes were synthesised by Twist Biosciences. mCherry was synthesised by Vectorbuilder (VB240515-1632gka).

Techniques: Construct, Over Expression, Immunofluorescence, Generated, Expressing, Knockdown, Quantitative RT-PCR, Control, Transformation Assay