Rac1 is required for epithelial stem cell function during dermal and oral mucosal wound healing but not for tissue homeostasis in mice

Abstract

Background: The regenerative capacity of the skin, including the continuous replacement of exfoliated cells and healing of injuries relies on the epidermal stem cells and their immediate cell descendants. The relative contribution of the hair follicle stem cells and the interfollicular stem cells to dermal wound healing is an area of active investigation. Recent studies have revealed that the small GTPase Rac1, which regulates cell migration and nuclear gene expression, is required for hair follicle stem function but not for the normal homeostasis of the interfollicular skin.

Methodology/principal findings: Here we explored whether Rac1 contributes to wound healing in the skin and in the oral mucosa, the latter an anatomical site that presents similar architecture to that of the skin but is devoid of any hair follicle structures, and hence lacks hair follicle stem cells. Epidermal Rac1 gene excision led to the clearly delayed closure of cutaneous wounds. Remarkably, genetic ablation of Rac1 from the oral mucosa resulted in the complete inability of oral wounds to heal. We present evidence that the lack of oral mucosal re-epithelization may result from the reduced migratory capacity of cells lacking Rac1 together with altered expression of injury-induced proliferative and cellular stress-related expression programs.

Conclusions/significance: Together, these observations support that while the normal development and homeostasis of the interfollicular skin and oral mucosa do not require Rac1 function, the interfollicular and oral epithelial stem cells may require a Rac1-dependent program to orchestrate the tissue response to injury and ultimate for wound closure. Ultimately, these findings may enable the molecular characterization of the acute tissue regenerative response of these stem cell populations, thus facilitating the identification of novel molecular-targeted strategies aimed at accelerating wound closure.

Figure 1. Defective epidermal wound-healing in Rac1 deficient epidermis.

(A) K14CreRac1 ^F/F conditional knockout mice show delayed wound-healing, as shown on day 6 after injury compared with control littermates. There was a significant decrease in the rate of wound closure (p<0.0001) when comparing the percentage of mice exhibiting open wounds after surgical incision in K14CreRac1 ^F/F mice (red squares, n  =  10) and control mice (blue circle, n  =  10) at the indicated days. (B) Left panels, representative H&E stained histological sections of wounded epidermis at day 6 after wounding showing the normal extension of the epithelial tongue of control (uppers panel) but defective epithelial tongue in K14CreRac1 ^F/F mice (lower panel) showing limited migration at day 6. High magnification (20x) of the epithelial tongue of control and K14CreRac1 ^F/F mice (dashed rectangles) are shown in the corresponding left panels. Epithelial tongues are delimitated in yellow dashes.

Figure 2. Deficient epithelial tongue migration in K14CreRac1^F/F mice.

(A) Control mice. Immunofluorescence for cytokeratins 10 and 14 (K10/14, red) reveals the epidermal layer of the skin and the epithelial tongue migrating underneath the fibrin clot at day 6 (Green). DNA staining (Hoechst 33342-blue) delineates cellular nuclei. Note the extensive migration of the epithelial tongue in merged figure (high Magnification-20x). (B) K14CreRac1 ^F/F conditional knockout mice stained for cytokeratins 10 and 14 (K10/14, red) show a more limited migration of the epithelial tongue by day 6 after injury. The deficient migration of the epithelial tongue is visualized in the merged figure (High magnification-20x).

Figure 3. Absence of oral wound-healing in K14CreRac1 ^F/F mice.

(A) K14CreRac1 ^F/F mice show a complete absence of oral wound closure when compared to control littermates (p<0.0001). The percentage of mice exhibiting open oral wounds after surgical incision in K14CreRac1 ^F/F mice (red squares, n  =  17) and control mice (blue circle, n  =  10) are depicted at the indicated days. (B) left panel, representative control mouse showing the intraoral wound site stained with surgical dye (black spot) 3.5 days after injury. Middle panels, H&E stained sections of control mice 2 and 3.5 days after oral injury showing active epithelial tongue migration over the fibrin clot by day 2, and complete closure of the wound site by day 3.5. Right panels, higher magnification (20x) of the corresponding area depicted with dashed rectangles in the middle panels. (C) Left panel, K14CreRac1 ^F/F mouse showing the intraoral surgical anatomical site stained in black (surgical dye) at day 3.5 after surgery. Middle panel, H&E staining of K14CreRac1 ^F/F mice show lack of healing capacity of the oral mucosa with a complete absence of an epithelial tongue. Right panel, high magnification (20x) of the open wound depicted by a dashed rectangle in the middle panel.

Figure 4. Lack of oral epithelial tongue migration in K14CreRac1^F/F mice.

(A) Immunofluorescence for cytokeratins 10 and 14 (K10/14, red), fibrinogen (green) and DNA staining (Hoechst 33342-blue) delineating cellular nuclei of control tissue sections staining the oral mucosa. Two days after the oral wound the epithelial tongue is visible migrating over the fibrin(ogen) clot (green) and the oral mucosa is healed by day 3.5. (B) Tissue sections from K14CreRac1 ^F/F mice stained as above show complete impairment of oral mucosa migration and maintenance of the punch biopsy trajectory evident by the fibrin(ogen) staining by day 3.5 (green). Note that high magnification (20x) of K10/14 staining show progressive loss of the well defined basal layer of the oral mucosa and absence of a epithelial tongue (yellow arrows).

Figure 5. Impairment of human oral keratinocytes migration and proliferation after knock down of Rac1.

(A) The human oral keratinocyte cell line (NOK-SI) was transfected with siRNAs and knockdown of Rac1 was confirmed by Western blot analyses of Rac1 in cellular lysates 72 h after transfection of two different targeting siRNAs (siRNA#1 and siRNA#2). A non-targeting siRNA oligonucleotide (siControl) and untransfected cells (Control) were used as controls. Scratch wound assay in NOK-SI cell line after control and Rac1 siRNA#2. Scratches were generated after cell confluence. In vitro cell migration and wound closure were assessed every 12 h. Representative pictures of the control and Rac1 siRNA#2 transfected cell cultures at the indicated time after the initial scratch are depicted. (B) Top graphic, areas of migration were measured (dotted line from figure A) in multiple wells and represented at the indicated time. The lower graphic shows cell proliferation of NOK-SI transfected with siRNA#2 against Rac1. Incorporation of [³H]thymidine in cells stimulated with EGF (+) (30 ng/ml) or left untreated (-) was measured by the accumulation of radioactivity in cellular DNA, and represented as the average of CPM± s.e.m. in triplicate samples from a representative experiment that was repeated three times (***p<0.001; *p<0.05). (C) GST pull-down of active Rac1 was performed using NOK-SI stimulated with EGF. Cell lysates were incubated with GST-PAK-N for 30 min to affinity precipitate active Rac1. PAK-bound Rac1 and total Rac1 in the corresponding total lysates were analyzed by Western blotting with a monoclonal antibody against Rac1. Total and phospho-specific antibodies were used to detect two downstream targets of Rac1, JNK and PAK1 and their corresponding phosphorylated species. (D) NOK-SI cells were transfected with Rac1 siRNA (siRNA#2) or control siRNA (siControl) and stimulated with EGF (30 ng/ml). Western blot analysis of the Rac1 downstream targets JNK and PAK1 show absence of basal p-JNK and 3 min after EGF stimulation compared to control that show detectable basal levels of p-JNK that increase after EGF exposure. p-PAK1 also show reduced basal levels and a delayed accumulation after activation by EGF as compared to control cell lysates. GAPDH was used as loading controls. In C and D.

Figure 6. Rac1 excision leads to decreased cell proliferation and limited expression of cytokeratin 6, a stress and growth-related marker, in the skin and oral mucosa after wounding.

(A) Impairment of epithelial cell proliferation in the wounded area of the skin and oral mucosa of K14Cre Rac1^F/F mice. Sections of epidermis adjacent to wounds healing skin (6 days after injury) and oral mucosa wounds (3.5 days after wounding) from of K14Cre Rac1^F/F and control mice were subjected to Ki-67 staining as a marker of cell proliferation. Both anatomical sites of the K14Cre Rac1^F/F mice show clear reduction in the proliferation capacity. Bar chart represents the total number of Ki-67 positive cells in the skin and oral cavity of K14Cre Rac1^F/F and control mice (**p<0.01, * p<0.05). (B) Normal interfollicular skin does not express detectable levels of cytokeratin 6 (K6). The epithelial tongue adjacent to the wound area from normal skin (control) show remarkable up regulation of K6 in all the epithelial layers after 6 days of injury. K14Cre Rac1^F/F mice fail to express K6 in the basal layer of the epithelial tongue from the skin (yellow arrows in the high magnification insert). (C) Normal oral mucosa does not express K6 during homeostasis (Normal). Expression of K6 in the oral mucosa epithelial tongue of control mice is evident at day 2 and 3.5 after injury. Expression pattern of K6 in the oral mucosa wound healing follows the skin pattern. K14Cre Rac1^F/F mice show lack of K6 staining in the basal layer of the oral mucosa adjacent to the wound site as depicted (yellow arrowhead).