Figure 3.

PROX1 depletion increases MMP14 expression in vivo and in vitro. (a) Whole-mount immunofluorescence of 4 weeks old Prox1^flox; Cdh5-CreER^T2 mice ear skin. Note efficient depletion of PROX1 in the Prox1^flox/flox; Cdh5-CreER^T2 ear compared to the control. (b) Representative images of IHC staining of the ear sections from mice described in (a) for podoplanin and endomucin. Nuclei were counterstained with Hoechst 33342. (c) Representative images of IHC staining for MMP14 and LYVE −1, a lymphatic endothelial marker, of the ear sections from mice described in (a). Nuclei were counterstained with Hoechst 33342. Pearsons’s co-localization index (PC) was calculated using 4 images of three Prox1^flox/flox and four Prox1^flox/flox; Cdh5-CreER^T2. PC = 0.088 in control mice (N3); PC = 0.640 in Prox1-depleted mice (N = 4); p = 0.01 (t-test). (d,e) LECs were transfected with the indicated siRNAs and after 72 h whole cell extracts were analysed by RTqPCR (d) for the indicated targets with GAPDH as an internal control. Bars represent an average of three independent experiments, error bars show SD across the experiments. (e) Immunoblotting for the indicated proteins using γ-tubulin (TUBG1) as a loading control; numbers indicate the intensity of MMP14 band for each sample normalized to the corresponding loading control. Cropped membranes are shown, uncropped membranes can be found in Fig. S3a. (f,g) HEK293FT cells were transfected with the indicated siRNAs, 32 h later whole cell extracts were analysed by RTqPCR (f) for the indicated targets with ACT as an internal control. Bars indicate the average of two independent experiments, error bars show SD; and by immunoblotting (g) for the expression of the indicated proteins using γ-tubulin (TUBG1) as a loading control; numbers indicate the intensity of MMP14 band for each sample normalized to the corresponding loading control. Cropped membranes are shown, uncropped membranes can be found in Fig. S3b. **p < 0.01; ***p > 0.001.