*3.5. Epithelial-to-Mesenchymal Transition Markers Were Detected in Epidermal Dysplastic Areas*

Considering the major importance of E-cadherin and stable cellular adherens junctions in the pathophysiological process of epithelial-to-mesenchymal transition (EMT) [25], expression of EMT effectors and markers was analyzed in xenografts. Firstly, expression of the transcription factor ZEB1 was investigated (Figure 5A,B). In control xenografts (Figure 5A), as well as in normal regions of xenografts from irradiated keratinocyte precursors (not shown), ZEB1 expression was almost exclusively restricted to the dermis, and rarely observed in epidermal keratinocytes (Figure 5A). Quantification performed on whole-length sections indicated that ZEB1-positive cells represented about 3% or less of most sections (Figure 5B). In contrast, observation of xenografts from irradiated keratinocyte precursors showed that ZEB1-positive keratinocytes were abundantly present in DAs (Figure 5A), reaching up to 18% of keratinocytes (*p* < 0.0001 versus control xenografts, and versus whole-length 50 mGy xenograft sections) (Figure 5B). Of note, this increase was not significant when considering whole-length 50 mGy xenograft sections, compared to controls (Figure 5B), suggesting that activation of ZEB1 expression did not concern all keratinocytes. Of note, β-catenin expression on keratinocyte membranes was also impaired in DAs. Expression of α-smooth muscle actin (α-SMA), a typical mesenchymal marker, was then analyzed in association with keratin-14 (K14), used as a specific counterstaining of basal keratinocytes (Figure 5C,D). In control xenografts (Figure 5C), as well as in normal regions of xenografts from irradiated keratinocyte precursors (not shown), α-SMA was exclusively distributed within the dermis. This distribution was strongly modified in DAs, with the detection of α-SMA signaling in cells that co-expressed K14 (Figure 5C), showing an ectopic expression in keratinocytes. α-SMA signal quantification focused on DAs showed a detectable signal above control values (*p* < 0.0001 versus control xenografts, and versus whole-length 50 mGy xenograft sections) (Figure 5D). As for ZEB1, the level of α-SMA did not appear to be significantly affected when considering whole-length 50 Gy xenograft sections, compared to the background signal of controls (Figure 5D). Taken together, these results identified a significant link between the pathophysiological process of EMT and characteristics of dysplastic areas. *Cells* **2020**, *9*, x FOR PEER REVIEW 15 of 19

**Figure 5.** *Cont.*

**Figure 5.** Detection of epithelial-to-mesenchymal transition markers in dysplastic areas. (**A**) Immunofluorescence detection of ZEB1. Representative pictures were selected for the visualization of dermal ZEB1 expression in normal skin regions and its ectopic presence in epidermal DAs. Gray-tone visualization of ZEB1 signal is also shown. (**B**) Quantitative analysis showed rare ZEB1-positive keratinocytes in normal epidermis regions, and abundant ZEB1-positive keratinocytes in DAs. Dot plots cumulated the analyses of 26 normal regions from the 14 control (Ctl) xenografts, 25 random regions from the 14 xenografts of the 50 mGy condition and 10 selected regions corresponding to DAs in 50 mGy xenografts. Bars correspond to median values (NS *p* > 0.05; \*\*\*\* *p* < 0.0001). β-catenin staining, which marked keratinocyte contours, revealed impaired epidermal organization and its reduced expression. (**C**) Immunofluorescence detection of α-smooth muscle actin (α-SMA). Representative pictures were selected for the visualization of dermal α-SMA expression in normal skin regions and its abnormal presence in epidermal DAs. Sections were stained for keratin 14 to mark

**Figure 5.** Detection of epithelial-to-mesenchymal transition markers in dysplastic areas. (**A**) Immunofluorescence detection of ZEB1. Representative pictures were selected for the visualization of dermal ZEB1 expression in normal skin regions and its ectopic presence in epidermal DAs. Gray-tone visualization of ZEB1 signal is also shown. (**B**) Quantitative analysis showed rare ZEB1-positive keratinocytes in normal epidermis regions, and abundant ZEB1-positive keratinocytes in DAs. Dot plots cumulated the analyses of 26 normal regions from the 14 control (Ctl) xenografts, 25 random regions from the 14 xenografts of the 50 mGy condition and 10 selected regions corresponding to DAs in 50 mGy xenografts. Bars correspond to median values (NS *p* > 0.05; \*\*\*\* *p* < 0.0001). β-catenin staining, which marked keratinocyte contours, revealed impaired epidermal organization and its reduced expression. (**C**) Immunofluorescence detection of α-smooth muscle actin (α-SMA). Representative pictures were selected for the visualization of dermal α-SMA expression in normal skin regions and its abnormal presence in epidermal DAs. Sections were stained for keratin 14 to mark **Figure 5.** Detectionofepithelial-to-mesenchymaltransitionmarkersindysplasticareas. (**A**)Immunofluorescence detection of ZEB1. Representative pictures were selected for the visualization of dermal ZEB1 expression in normal skin regions and its ectopic presence in epidermal DAs. Gray-tone visualization of ZEB1 signal is also shown. (**B**) Quantitative analysis showed rare ZEB1-positive keratinocytes in normal epidermis regions, and abundant ZEB1-positive keratinocytes in DAs. Dot plots cumulated the analyses of 26 normal regions from the 14 control (Ctl) xenografts, 25 random regions from the 14 xenografts of the 50 mGy condition and 10 selected regions corresponding to DAs in 50 mGy xenografts. Bars correspond to median values (NS *p* > 0.05; \*\*\*\* *p* < 0.0001). β-catenin staining, which marked keratinocyte contours, revealed impaired epidermal organization and its reduced expression. (**C**) Immunofluorescence detection of α-smooth muscle actin (α-SMA). Representative pictures were selected for the visualization of dermal α-SMA expression in normal skin regions and its abnormal presence in epidermal DAs. Sections were stained for keratin 14 to mark basal keratinocytes. Gray-tone visualization of α-SMA signal is also shown. (**D**) Dot plots cumulated the analyses of 13 normal regions from control (Ctl) xenografts, 9 random regions from the 50 mGy xenografts and 10 selected regions corresponding to DAs in 50 mGy xenografts. Bars correspond to median values (NS *p* > 0.05; \*\*\*\* *p* < 0.0001).
