Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 112 (33), 10407-12

Pleiotropic Age-Dependent Effects of Mitochondrial Dysfunction on Epidermal Stem Cells

Affiliations

Pleiotropic Age-Dependent Effects of Mitochondrial Dysfunction on Epidermal Stem Cells

Michael C Velarde et al. Proc Natl Acad Sci U S A.

Abstract

Tissue homeostasis declines with age partly because stem/progenitor cells fail to self-renew or differentiate. Because mitochondrial damage can accelerate aging, we tested the hypothesis that mitochondrial dysfunction impairs stem cell renewal or function. We developed a mouse model, Tg(KRT14-cre/Esr1) (20Efu/J) × Sod2 (tm1Smel) , that generates mitochondrial oxidative stress in keratin 14-expressing epidermal stem/progenitor cells in a temporally controlled manner owing to deletion of Sod2, a nuclear gene that encodes the mitochondrial antioxidant enzyme superoxide dismutase 2 (Sod2). Epidermal Sod2 loss induced cellular senescence, which irreversibly arrested proliferation in a fraction of keratinocytes. Surprisingly, in young mice, Sod2 deficiency accelerated wound closure, increasing epidermal differentiation and reepithelialization, despite the reduced proliferation. In contrast, at older ages, Sod2 deficiency delayed wound closure and reduced epidermal thickness, accompanied by epidermal stem cell exhaustion. In young mice, Sod2 deficiency accelerated epidermal thinning in response to the tumor promoter 12-O-tetradecanoylphorbol-13-acetate, phenocopying the reduced regeneration of older Sod2-deficient skin. Our results show a surprising beneficial effect of mitochondrial dysfunction at young ages, provide a potential mechanism for the decline in epidermal regeneration at older ages, and identify a previously unidentified age-dependent role for mitochondria in skin quality and wound closure.

Keywords: cellular senescence; oxidative stress; skin aging; stem cell proliferation; superoxide dismutase 2.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. S1.
Fig. S1.
Generation of keratinocyte-specific conditional Sod2-deficient mice. (A) Diagram of primers designed to detect deletion (Deleted) or presence (Intact) of exons 1–3 of the Sod2 gene and an unmodified control gene region (Total) by PCR. (B–D) Representative PCR gel showing amplification of DNA with deleted or intact Sod2 sequences from whole dorsal skin (B), separated epidermis and dermis (C), and tail, heart, liver, toe, intestinal jejunum, and lung (D) of K14S mice after corn oil/vehicle (veh) or tamoxifen (TAM) treatment. (E) Average percent recombination at the Sod2 locus, determined by qPCR, using genomic DNA isolated from the epidermis and dermis of 22-mo-old K14S and K14R mice, treated with corn oil (vehicle control, veh; n = 8) or tamoxifen (TAM; n = 8) at age 4 mo. (F and G) Sod2 mRNA levels determined by qPCR of RNA isolated from the epidermis (F) and dermis (G) of K14S and K14R mice, treated with veh (n = 5) or TAM (n = 5) at age 4 mo. Mean ± SEM values with asterisks indicate significant differences at P < 0.05 relative to vehicle control by Student’s t test. (H) Representative photomicrographs of skin of 8-mo-old K14S mice, treated with veh or TAM at age 4 mo, stained for cytochrome c oxidase (COX; brown) activity. (Upper) Hair follicles outlined by dashed lines. (Lower) Skeletal muscle beneath the SC layer. (I) Quantification of succinate dehydrogenase (SDH) and COX staining scored by ranking the staining intensities. The highest intensity was assigned the highest ranked value, and mean averages were calculated. Two blinded researchers independently ranked the intensities. Asterisks indicate significant differences at P < 0.05 by Student’s t test.
Fig. 1.
Fig. 1.
Keratinocyte-specific mitochondrial dysfunction in mice. (A) Coimmunostaining for SOD2 (red) and the keratinocyte marker CD49f (green) and nuclear staining by DAPI (blue) on tail whole mounts of 2-y-old K14S mice treated with (Left) corn oil (vehicle; veh) or (Right) tamoxifen (TAM) at age 4 mo. (B) Representative photomicrographs of skin of 8-mo-old K14S mice treated with veh or TAM at age 4 mo, stained for succinate dehydrogenase (SDH; blue). (Upper) Hair follicles outlined by dashed lines. (Lower) Skeletal muscle beneath the SC layer.
Fig. 2.
Fig. 2.
Wound closure in 8 mo old K14S mice treated with veh or TAM at age 4 mo. (A) Average wound size (mean ± SEM) after 8-mm punch biopsy in K14S mice (veh, n = 11; TAM, n = 8). Percent wound size refers to the wound area relative to the initial wound area × 100. Asterisks indicate differences at P < 0.05 by Student’s t test. (B) Representative photomicrographs of H&E staining of wound areas in K14S mice at 2 d and 4 d after injury (n = 3 each). Layers above the dashed lines indicate the epidermis. White arrows show area of reepithelialization. (C) Quantification of PCNA immunofluorescence in wound areas of K14S mice at 2 d (n = 3 each) and 6 d (n = 5 each) after skin injury. (D and E) mRNA levels determined by qPCR of genes associated with proliferation (D), expressed predominantly in epidermal basal cells, and differentiation (E), expressed predominantly in epidermal suprabasal cells, in wound areas of K14S mice at 2 d (veh and TAM, n = 3) or 6 d (veh, n = 5; TAM, n = 7) after injury. Mean ± SEM values with asterisks indicate differences at P < 0.05 by two-way ANOVA followed by Bonferroni post hoc analysis. (F) Representative photomicrographs of H&E staining of wound areas of K14S mice at 6 d after injury (veh and TAM, n = 5). Stratum basale (SB; arrows) cells are identified by strong nuclear staining (blue) at the wound bottom. Stratum spinosum (SS; black line) are cells with low nuclear staining above the SB layer. Stratum granulosum (SG; asterisks) are cells with granules in the cytoplasm. Stratum corneum (SC; blue line) is the outermost layer. (G) Representative photomicrographs of loricrin (LOR; red) staining by immunofluorescence and nuclear staining by DAPI (blue) in wound areas of K14S mice at 6 d after injury. Dashed lines indicate the wound site. Multiple photomicrographs of wound sections were taken under 10× magnification and tiled in Adobe Photoshop.
Fig. S2.
Fig. S2.
Collagen formation, cell proliferation and apoptosis in wounds of young K14S mice. (A) Representative photomicrographs of picrosirius red staining of the wound area of 8-mo-old veh- or TAM-treated K14S mice at 2–10 d after injury. To capture the complete wound site, multiple photomicrographs of different sections were taken at 4× magnification, then tiled in Adobe Photoshop. (B) Representative photomicrographs of PCNA (green) immunofluorescence in wound areas of 8-mo-old veh- or TAM-treated K14S mice at 6 d after skin injury (veh and TAM, n = 5). Nuclei are stained by DAPI (blue). (C) Representative photomicrographs and quantitative measurement of Ki67 (green) immunofluorescence in wound areas of 8 mo old veh- and TAM-treated K14S mice at 6 d after skin injury (veh and TAM, n = 5). Nuclei are stained by DAPI (blue). Cell layers above the dashed lines indicate the epidermis. Mean ± SEM values with asterisks indicate significant differences at P < 0.05 by Student’s t test. (D) mRNA levels of genes associated with proliferation and expressed predominantly in epidermal basal cells, determined by qPCR, in wound areas of 8-mo-old veh- or TAM-treated K14S mice, at 2 d (veh and TAM n = 3) and 6 d (veh n = 5; TAM n = 7) after injury. Means ± SEM with asterisks indicate differences at P < 0.05 by two-way ANOVA followed by Bonferroni post hoc analysis. (E) Representative photomicrographs of TUNEL- positive (green) immunofluorescence in wound areas of 8-mo-old veh- or TAM-treated K14S mice, at 2 d and 6 d after skin injury (veh and TAM, n = 5). Nuclei are stained by DAPI (blue). Tissues treated with DNase I served as a positive control (Pos Ctrl). The TUNEL assay was performed using the DeadEnd Fluorometric TUNEL System (Promega).
Fig. S3.
Fig. S3.
Gene expression and histology of healing wounds in young K14S mice. (A) mRNA levels of inflammatory cytokines, determined by qPCR, in wound areas of 8-mo-old veh- and TAM-treated K14S mice, at 2 d (veh and TAM, n = 3) and 6 d (veh, n = 5; TAM, n = 7) after injury. Mean ± SEM values with asterisks indicate differences at P < 0.05 by two-way ANOVA followed by Bonferroni post hoc analysis. (B and C) Representative photomicrographs of H&E staining of the wound area of 8-mo-old veh- and TAM-treated K14S mice at 6 d (B) and 10 d (C) after injury (veh and TAM, n = 5). Stratum basale (SB, arrows) are identified by strong nuclear staining (blue) at the bottom of the wound site. Stratum granulosum (SG, asterisks) are cells with granules in the cytoplasm. Stratum spinosum (SS) are cells with low nuclear staining above the SB layer. Stratum corneum (SC) is the outermost layer.
Fig. 3.
Fig. 3.
Wound closure in old K14S mice. (A) Average wound size (mean ± SEM) after injury by 8-mm punch biopsy in 14-mo-old K14S mice treated with veh or TAM (veh, n = 10; TAM, n = 11) at age 4 mo. Percent wound size refers to the wound area relative to the initial wound area × 100. (B) Quantification of epidermal thickness (in µm) in H&E-stained K14S mouse skin at 4, 7, and 10 mo after veh (Left) or TAM (Right) treatment (n = 5 each). (C) mRNA levels of genes expressed predominantly in K14+ stem cells, analyzed by qPCR, in skin from K14S mice at 4 or 11–20 mo after treatment with veh or TAM. Mean ± SEM values with asterisks indicate differences at P < 0.05 relative to vehicle control by two-way ANOVA, followed by Bonferroni post hoc analysis. (D) Quantification of flow cytometry histographs of epidermal cells stained for CD49f, CD34, and Sca1, isolated from the skin of 14-mo-old K14S mice treated with veh or TAM at age 4 mo (n = 5 for each). Populations of bulge stem cells (CD49fhi/CD34+), suprabasal bulge stem cells (CD49flo/CD34+), and junctional zone stem cells (CD49fhi/CD34/Sca1) were quantified as the average percent cell population (mean ± SEM) in the epidermis. Asterisks indicate differences at P < 0.05 by Student’s t test.
Fig. S4.
Fig. S4.
Wound closure in aging K14S mice. (A) Average wound size (mean ± SEM) after injury by an 8-mm dermal punch biopsy in 11-mo-old K14S mice, treated with veh or TAM (veh, n = 13; TAM, n = 11) at age 4 mo. Percent wound size refers to the wound area relative to the initial wound area × 100. (B) Representative photomicrographs of H&E staining of skin of K14S mice at 4, 7, and 10 mo after veh (Left) or TAM (Right) treatment (veh and TAM, n = 5). (C and D) Heat map (C) and mRNA levels (D) of genes expressed predominantly in basal interfollicular epidermis (IFE), suprabasal IFE, K14+ stem cells, and involucrin (Ivl) committed progenitor cells, analyzed by qPCR, in skin from K14S mice at 4 mo and 11–20 mo after treatment with veh or TAM.
Fig. S5.
Fig. S5.
Epidermal stem cells in old K14S mice. Representative histographs (A) and quantification (B) of flow cytometry analysis of epidermal cells stained for CD49f, CD34, and Sca1, isolated from the skin of 14-mo-old K14S mice treated with veh or TAM at age 4 mo (n = 5 for each). Populations of bulge stem cells (CD49fhi/CD34+), suprabasal bulge stem cells (CD49flo/CD34+), junctional zone stem cells (CD49fhi/CD34/Sca1), isthmus stem cells (CD49flo/CD34/Sca1), basal interfollicular epidermal cells (CD49fhi/CD34/Sca1+), and suprabasal interfollicular epidermal cells (CD49flo/CD34/Sca1+) are labeled.
Fig. 4.
Fig. 4.
Cellular senescence in mouse and human keratinocytes with mitochondrial dysfunction. (A) Average percentage of SA-βgal+ cells quantified from the surface with blue staining relative to the total epidermal surface. Mean ± SEM values with asterisks indicate differences at P < 0.05 by Student’s t test. (B) Average percentage of 8-mo-old K14S mice with detectable levels (Ct <40) of p16INK4a mRNA in epidermal samples at 4 mo after treatment with veh (n = 4) or TAM (n = 7). (C) Representative photomicrographs of H&E-stained skin of acetone-treated (ctrl) or 12-O-tetradecanoylphorbol-13-acetate (TPA)-treated 8-mo-old K14S mice 48 h after treatment with veh or TPA (n = 3 for both). In both cases, animals were pretreated with veh or TAM at age 4 mo. Arrows indicate epidermal thickness and asterisks indicate disrupted epidermis after TPA treatment. (D) Quantification of Ki67+ cells in the epidermis from acetone- or TPA-treated 8-mo- old K14S mice, treated with veh or TAM at age 4 mo. Mean ± SEM values with asterisks indicate significant differences at P < 0.05 relative to vehicle control by two-way ANOVA, followed by Bonferroni post hoc analysis. (E and F) Human keratinocytes treated with 100 nM rotenone (Rot) for 7 d (n = 3) and analyzed for mRNA levels of p16INK4a (E) and proliferation- and differentiation-associated genes (F). Mean ± SEM values with asterisks indicate differences at P < 0.05 by Student’s t test and two-way ANOVA, followed by Bonferroni post hoc analysis.
Fig. S6.
Fig. S6.
Cellular senescence in K14S mice and human keratinocytes. (A) Representative photomicrographs of skin of 8-mo-old K14S mice treated with veh or TAM at age 4 mo, stained for SA-βgal activity (blue) and nuclei (red). The blue line indicates surface region with SA-βgal+ cells; the red line, surface region of total epidermal cells. (B) Average percentage of 14 mo old K14S mice with detectable levels (Ct <40) of p16INK4a mRNA in isolated epidermal samples, 10 mo after veh or TAM treatment. (C) mRNA levels of wound healing-related growth factors, as determined by qRT-PCR of RNA isolated from wound areas of 8-mo-old veh- or TAM-treated K14S mice, treated at age 4 mo at 2 d (Upper) or 6 d (Lower) after skin injury. Mean ± SEM values with asterisks indicate significant differences at P < 0.05 by Student’s t test. (D) Ki67 (red) and CD49f (green) staining by immunofluorescence and nuclear staining by DAPI (blue) of skin from acetone- or TPA-treated 8-mo-old K14S mice, treated with veh or TAM at age 4 mo. (E) Human keratinocytes treated with 100 nM rotenone for 7 d (n = 3) and analyzed for morphology by bright field microscopy. (F) Targeting strategy for inactivating Sod2 by Cre-lox recombination. The Sod2 locus is shown in black, the five exons of Sod2 are shown in dark blue, the PGK/neo cassette is shown in light blue and orange, and the loxP insertion sites are marked in red.

Similar articles

See all similar articles

Cited by 17 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback