Regeneration of multilineage skin epithelia by differentiated keratinocytes

Abstract

Although homeostasis of rapidly renewing tissues like skin epithelia is maintained by stem cells, the committed progeny of stem cells in the basal layer of epidermis retain regenerative potential and are capable of forming epidermis in response to environmental cues. It is not clear, however, at what point within the epidermal lineage keratinocytes lose this regenerative potential. In this study, we examined the extent of tissue formation by post-mitotic differentiated keratinocytes. We show that cultures of mouse keratinocytes that were, by all measures, differentiated were able to reform a self-renewing, hair-bearing skin when transplanted onto suitable sites in vivo. Genetic labeling and lineage-tracing studies in combination with an involucrin-driven Cre/lox reporter system confirmed that transplanted differentiated keratinocytes were indeed the source of the regenerated skin. More importantly, analysis of early stages of skin regeneration showed hallmarks of dedifferentiation of transplanted differentiated keratinocytes. These data indicate that commitment to differentiation does not prohibit cells from re-entering the cell cycle, de-differentiating, and acquiring "stemness". These findings suggest that epidermis can use different strategies for homeostasis and tissue regeneration.

Figure 1. Regeneration of skin epithelia by keratinocytes grown in low and high Ca conditions

(a, f) Phase contrast images of epidermal cultures, (b, c) light and (e–h) surface GFP images of skin reconstituted from keratinocytes cultured in low and high Ca at 6 weeks post grafting. (d, i) Fluorescence and (e, j) H&E-stained images of reconstituted skin at 15 weeks post grafting. GFP-expressing cells (green) were present in epidermis, sebaceous glands, and hair follicles derived from keratinocytes cultured in low and high Ca. Higher exposure showing GFP in the basal cells (inset). Sections were counterstained with DAPI (blue nuclear staining). Bar=50 μm (d, i) and 100 μm (e, j).

Figure 2. Morphology, proliferation rate, and expression of differentiation markers in mouse keratinocytes cultured in high Ca

(a) Phase contrast photographs of culture in low or high Ca condition, Bar=50 μm. (b) BrdU-labeling index of keratinocytes exposed to low or high Ca after 6 hours pulse. Values are expressed as mean labeled nuclei±SD in a minimum of 10 fields (>100 cells per field) from three independent experiments. (c) Western blot analysis of INV (Inv), loricrin (Lor), p63, and actin in primary mouse keratinocytes grown for 4 days in low Ca to confluence (lane 1), and for additional 3 days either in low Ca (lane 2) or in high Ca (lane 3). (d) Flow cytometric analysis of involucrin expression in confluent cultures of keratinocytes in low Ca (INV Low) or after exposure to high Ca (INV High). The percent involucrin positive keratinocytes in each culture is indicated.

Figure 3. Skin regeneration by differentiated cultures is not attributed to a small population of Ca-resistant keratinocytes

(a) The size of reconstituted skin is directly correlated with the number of implanted proliferating keratinocytes. The calculated area of GFP-expressing skin regenerated from mixtures of GFP-labeled proliferating keratinocytes and γ-irradiated non-labeled keratinocytes at ratios of 2, 10, and 100% at 6 weeks post grafting are shown. (b–h) Retrovirus transduction of differentiated cultures showing a minimal contribution of proliferating cells in differentiated cultures to tissue formation. (c, f) Phase contrast photograph of X-gal-stained cultures of proliferative or differentiated mouse keratinocytes transduced with a retrovirus encoding LacZ and used for fate mapping analysis during skin regeneration (original magnification ×10). (d–f) Surface GFP and X-gal whole mount staining of a representative skin reconstituted from LacZ-transduced cultures are shown. To allow visualization of both X-gal and GFP, whole mounts were examined by an inverted fluorescent microscope. Arrows point to LacZ-positive cells in the graft. Bar=50 μm (c, f) and 700 μm (d–e, g–h). Values in (b) are expressed as mean percentage of labeled cells or area±SD in a log scale (n=3).

Figure 4. Genetic labeling and tracing of involucrin-expressing keratinocytes

(a) A schematic representation of the strategy used to generate triple-transgenic mice allowing lineage tracing of involucrin-expressing keratinocytes. In these mice, activation of involucrin promoter during differentiation induces irreversible Cre-mediated recombination between loxP sites (triangles) resulting in constitutive expression of YFP in all of the daughter cells. (b) Immunofluorescent staining of newborn mouse skin for K14, involucrin (INV), and α6-integrin shows restriction of YFP to suprabasal layers of epidermis. (c) The proportion of YFP-labeled cells in epidermal preparations of newborn triple-transgenic mice in freshly isolated epidermal preparations or in cultures grown in low or high levels of Ca. (d–e) Representative skin reconstituted from keratinocytes isolated from triple-transgenic mice and cultured in low (d) or high Ca conditions (e) at 6 weeks post grafting. Grafted mice were maintained in a repressed state (+Dox) to inhibit de novo expression of YFP after transplantation. Distinct patterns of YFP expression between skin reconstituted from proliferating (d) and differentiated (e) cultures. YFP expression in the latter does not coincide with INV expression and includes K14-expressing keratinocytes. To show differentiation-induced YFP expression in skin reconstituted from proliferating cultures, Dox was removed (− Dox) from drinking water 14 days before analysis. Bars=50 μm.

Figure 5. Dedifferentiation of terminally differentiated keratinocytes during early stages of skin regeneration

Tissues reconstituted from GFP-expressing keratinocytes cultured in low or high Ca were analyzed at 1 week post grafting. (a–b) H&E staining, (c–d) epifluorescent GFP, and (e–p) immunofluorescent staining for proliferating antigen Ki67, α6-integrin, intermediate differentiation marker involucrin (INV), late differentiation markers filaggrin (FIL) and loricrin (LOR), and to keratin 14 (K14). GFP+ cells appear in green, whereas markers are shown in red. Bar=50 μm.

Figure 6. Normalization of epidermal differentiation in skin epithelia regenerated from differentiated cultures

(a–c) H&E staining of regenerated tissue showing initiation of hair germ formation (arrows) by day 10 post-grafting. By 6 weeks, no difference is observed between skin regenerated from differentiated and proliferative cultures. (d–o) Immunofluorescent staining of skin grafts harvested at either 10 days or 6 weeks post grafting for markers indicated in the figure. (n) Inset shows CD34 staining in hair follicle below sebaceous gland. GFP+ cells appear in green, whereas markers appear in red. Red staining in sebaceous gland is non-specific. GFP expression is present in the basal layer but at lower levels than cornified layers in the epidermis reconstituted from Nagy-GFP cells (inset in h). Bar=75 μm for a–c and 50 μm for d–o.