Loss of epidermal Par3 promotes MC proliferation and survival through direct KC–MC contacts. (A) PCNA immunostaining of primary melanoma formed in Ctrl;HGF⁺;Cdk4^R24C/R24C and Par3^eKO;HGF⁺;Cdk4^R24C/R24C mice. DAPI served as a nuclear counterstain. Bar, 25 µm. (B) Quantification of A. Relative proliferative index, determined as fraction of DAPI nuclei with high PCNA expression, is shown. Data are mean ± SD. n (control) = 5; n (KO) = 8. *, P = 0.0453; unpaired Student’s t test with Welch’s correction. Data are pooled from three independent experiments. (C) Phospho-ERK, phospho-Akt, and β-actin immunoblots of protein extracts isolated from primary melanoma that developed in Ctrl;HGF⁺;Cdk4^R24C/R24C and Par3^eKO;HGF⁺;Cdk4^R24C/R24C mice. (D) Quantification of C. Immunosignals of phosphorylated ERK and Akt were first normalized to β-actin (loading control) and then normalized to values of control tumors. Data are mean ± SD. pERK: n = 8 tumors per group; **, P = 0.0076. pAKT: n = 7 tumors per group; **, P = 0.00695. Student’s t test was used. Data are pooled from three independent experiments, with each tumor originating from a different mouse. (E) TRP2 immunostaining for MCs in neonatal (P0) back skin cross sections of control and Par3^eKO mice. Arrows mark MCs, and dashed lines underline the MC-containing basal epidermal layer. (F) Quantification of MC number in interfollicular epidermis of control (n = 4 independent biological replicates) and Par3^eKO (n = 5 biological replicates). Data are mean ± SD. P0 back skin: **, P = 0.0076; Student’s t test. Hair follicles: n = 4 biological replicates per group; *, P = 0.0129; Student’s t test. (G) TRP2 immunostaining of P58 back skin cross sections of control and Par3^eKO mice. Arrows indicate MCs. Data are representative of three independent experiments. (H) Quantification of MC number per hair follicle of control and Par3^eKO back skins of adult mice of different age. Data are pooled from independent experiments and depicted as mean ± SD. 7 wk: n = 3 biological replicates per group; **, P = 0.0012. 4 mo: n = 3 biological replicates per group; *, P = 0.0139. 12 mo: n = 4 biological replicates per group; *, P = 0.0291. Student’s t test was used. (I) Scheme of direct co-cultures (top) and quantification of direct co-cultures (bottom), cultivated 7 d in LC and HC medium (five images/condition) The graph shows pooled data from independent experiments depicted as mean ± SD. n = 3 biological replicates per group. **, P = 0.0022; Student’s t test). (J) Phase-contrast micrography of direct co-cultures. Orange asterisks mark MCs. (K) Scheme and quantification of MC number per image in indirect co-culture with control and Par3^KO KCs (greater than nine images/condition) The graph show pooled data from independent experiments as mean ± SEM. n = 3 biological replicates per group. LC: ns, P = 0.79. HC: ns, P = 0.99. Student’s t test was used. Subconfluent MCs cultured on a permeable filter support were indirectly co-cultured with control or Par3-deficient KC cultures. (L) BrdU and TRP2 immunostaining of direct co-cultures. Arrowheads mark BrdU-positive MCs. Data are representative of three independent experiments. (M) Quantification of MC number directly co-cultured with control or Par3^KO KCs for 10 d at HC (left) and percentage of MCs positive for BrdU (middle) or cleaved caspase3 (right; >10 images per condition). The graphs show pooled data from independent experiments depicted as mean ± SD. MC: n = 6 biological replicates per group; **, P = 0.0012. Proliferation: n = 4 biological replicates per group; **, P = 0.0011. Apoptosis: n = 3 biological replicates per group; ***, P = 0.0006. Student’s t test was used. cl., cleaved; Ctrl, control; IFE, interfollicular epidermis, pos., positive; rel., relative. Bars: (E and L) 10 µm; (G and J) 20 µm.