Supplementary Figure 3: Additional data confirming the role of Gata6 in specifying the SD lineage.

(a) RT-qPCR of mRNA from dorsal epidermis of wild type (WT) and Gata6 knockout (cKO) mice confirming loss of Gata6 expression in the Gata6 cKO. Data are means ± s.d. from n = 4 mice. (b) Quantification of basal Itga6 + Cd34- keratinocytes by flow cytometry (see Fig. 1h). Data are means ± s.d. from n = 3 mice. c, Tail skin sections from WT, Gata6 epidermal knockout (cKO), K14ΔNLef1 and K14ΔNLef1 x Gata6 cKO (K14ΔNLef1 cKO) stained for the duct marker Plet1. Inserts are higher magnification views. (d,e) K14ΔNLEF1 cysts express Plet1 and Atp6v1c2 and expression is reduced in the absence of Gata6. RT–qPCR analysis of mRNA from WT, cKO, K14ΔNLef1 and K14ΔNLef1 cKO epidermis (d). K14ΔNLef1 and K14ΔNLef1 cKO back skin sections were stained for the SD marker Atp6v1c2 and the SG marker Scd1 (e). Data are means ± s.d. from n = 4 WT, n = 5 cKO, n = 3 K14ΔNLef1 and n = 4 K14ΔNLef1 cKO mice. ^∗P < 0.05; ^∗∗P < 0.005, by unpaired Student’s t-test. Scale bars, 50 μm. (f) Gene Ontology enrichment analysis of Gata6 direct target genes highly expressed in Gata6 + Itga6 + cells (see Fig. 2a). g, Plots of ChIP-Seq reads aligned to the Cep55, Ncaph and Spc24 (representative genes involved in mitosis) loci are shown. h, RT–qPCR of RNA from wild type (WT), Gata6 knockout (cKO), K14ΔNLef1 and K14ΔNLef1 x Gata6 cKO (K14ΔNLef1 cKO) epidermal cells. (i–k) Validation of gene expression profiles from manual microdissected tail interfollicular epidermis (IFE), hair follicles (HF) and sebaceous glands (SG). Bright field image of tail epidermal whole mount showing schematic of microdissection strategy (i). Z score heat maps representing array results for differentially expressed genes (j) validated by RT-qPCR (k). Data are means ± s.d. from n = 4 mice for IFE; n = 6 mice for SG; n = 2 for HF; n = 6 for whole tail epidermis (EPID). l, Sections of wild type (WT), Gata6 epidermal knockout (cKO), K14ΔNLef1 and K14ΔNLef1 x Gata6 cKO (K14ΔNLef1 cKO) dorsal skin stained for Gata6 and Blimp1 showing reduction in Blimp1 + cells in HF but not IFE in absence of Gata6. Inserts are higher magnification views of the JZ (red channel only). Asterisk indicates nonspecific staining. Dashed lines indicate epidermal-dermal boundary. (m) Nuclear Androgen Receptor, Ar (black arrows), normally present in WT SG (insert panel), is reduced in K14ΔNLef1 ectopic SG (top panel) but restored in K14ΔNLef1 epidermis when Gata6 is ablated (bottom panel). Back skin sections were labeled with Ar antibody. o, RT–qPCR analysis of RNA from WT, cKO, K14ΔNLef1 and K14ΔNLef1 cKO adult back skin dermis. Data are means ± s.d. from n = 4 WT, n = 5 cKO, n = 3 K14ΔNLef1 and n = 4 K14ΔNLef1 cKO mice ^∗P < 0.05 by unpaired Student’s t-test. (p) Schematic representation of Gata6 regulation of Ar and Blimp1. Representative images of n = 3 mice per genotype. Scale bars, 50 μm.