Extended Data Fig. 2: Generation and validation of Hsd3b6∆skin mice by CRISPR/Cas9 genomic tar-geting and baseline analysis of the Hsd3b6∆skin mice compared to control.
From: Skin androgens regulate Staphylococcus aureus pathogenicity via quorum sensing

a, Schematic diagram of CRISPR/Cas9-mediated gene insertion of loxP sites flanking the first coding exon of the Hsd3b6 locus. Hsd3b6fl/fl mice crossed with K14-Cre+/- mice to generate Hsd3b6∆skin mice (K14-Cre+/-; Hsd3b6fl/fl). b, Immunofluorescence staining of HSD3B6 expression in Hsd3b6fl/fl and Hsd3b6∆skin mice skin. Scale bar 50 µM. c, Genomic confirmation of loxP mediated deletion of Hsd3b6 in the K14Cre+/- mice in the skin, with PCR of testicular samplings demonstrating tissue specificity. d, qRT-PCR for Hsd3b6 expression Hsd3b6fl/fl (n = 4) and Hsd3b6∆skin (n = 4) in the female skin and Hsd3b6fl/fl (n = 9) and Hsd3b6∆skin (n = 7) in the male skin, Hsd3b6fl/fl (n = 3) and Hsd3b6∆skin (n = 3) in testis, and Hsd3b6fl/fl (n = 3) and Hsd3b6∆skin (n = 3) in the adrenal gland normalized to Rplp0 expression. e, Serum progesterone in male Hsd3b6fl/fl (n = 5) and Hsd3b6∆skin (n = 5) and DHT in male Hsd3b6fl/fl (n = 6) and Hsd3b6∆skin (n = 5) quantified by hormone immunoassay. f, Transepidermal water loss (TEWL) of male Hsd3b6fl/fl (n = 9) and Hsd3b6∆skin (n = 7) and female Hsd3b6fl/fl (n = 5) and Hsd3b6∆skin (n = 10) mice measured by Vapometer device as a metric of barrier integrity. g, Hematoxylin and eosin staining of Hsd3b6fl/fl and Hsd3b6∆skin skin. Scale bar 220 µM. h, Weight measurements of male Hsd3b6fl/fl (n = 9) and Hsd3b6∆skin (n = 7) and female Hsd3b6fl/fl (n = 6) and Hsd3b6∆skin (n = 10) mice. i, Progesterone quantified from the skin secretions of male Hsd3b6fl/fl (n = 11) and Hsd3b6∆skin (n = 8) and female Hsd3b6fl/fl (n = 7) and Hsd3b6∆skin (n = 11) age-matched mice at 7 weeks by hormone immunoassay. Means ± SEM (error bars) are plotted. (d) *p < 0.05, ***p < 0.001, ns, not significant by one-way ANOVA. (e-f,h) ns by Mann-Whitney U-test (two-sided). (i) **p < 0.01, ****p < 0.0001 by Kruskal-Wallis test with correction for false discovery (two-sided).