An essential role for senescent cells in optimal wound healing through secretion of PDGF-AA

Abstract

Cellular senescence suppresses cancer by halting the growth of premalignant cells, yet the accumulation of senescent cells is thought to drive age-related pathology through a senescence-associated secretory phenotype (SASP), the function of which is unclear. To understand the physiological role(s) of the complex senescent phenotype, we generated a mouse model in which senescent cells can be visualized and eliminated in living animals. We show that senescent fibroblasts and endothelial cells appear very early in response to a cutaneous wound, where they accelerate wound closure by inducing myofibroblast differentiation through the secretion of platelet-derived growth factor AA (PDGF-AA). In two mouse models, topical treatment of senescence-free wounds with recombinant PDGF-AA rescued the delayed wound closure and lack of myofibroblast differentiation. These findings define a beneficial role for the SASP in tissue repair and help to explain why the SASP evolved.

Figure 1. Characterization of p16-3MR Cells

(A) Schematic of the p16-3MR transgene. See Results and Supplemental Experimental Procedures for details. (B) qRT-PCR analysis of RNA isolated from nonsenescent (NS) and senescent (SEN; induced by 10 Gy IR) MEFs derived from p16-3MR embryos. RNA was analyzed for mRNA levels of the indicated genes relative to actin mRNA (control for cDNA quantity) (n = 3 independent experiments; A.U., arbitrary units). (C) Immunofluorescence of cells described in (B). Blue shows DAPI-stained nuclei; red shows mRFP immunostaining. (D) Luminescence measurements of cells described in (B). Cells were incubated with coelentarazine and lysed, and luminescence intensity was quantified using a luminometer (n = 3; A.U., arbitrary units). (E) WT or p16-3MR MEFs were treated with GCV (10 μg/ml) for 7 days and evaluated for the percentage of surviving cells using the MTS assay (n = 4). Data shown are the mean ± SD. **p < 0.01, ***p < 0.001.

Figure 2. Characterization of p16-3MR Mice

(A) Mock or IR p16-3MR mice, treated with vehicle (PBS) or 25 mg/kg of GCV for 5 days (daily i.p. injections; GCV), were injected with coelentarazine, and luminescence was quantified using a Xenogen Imaging system. (B) RNA was extracted from the fat of mice described in (A) and quantified by qRT-PCR for mRNA levels of endogenous p16^INK4a, mRFP, IL-6, MMP-3, and IL-5. Tubulin mRNA was used as a control (n = 4). (C) Representative image of a 24-month-old p16-3MR mouse before (left) and after (right) GCV treatment. Luminescence was quantified using a Xenogen Imaging system. (D–F) Fat biopsies from old (20–24 months) or young (3–4 months) p16-3MR mice, treated with PBS or GCV as described in (A). (D) Biopsies were incubated with a coelentarazine solution and luminescence quantified using a Xenogen Imaging system. (E) Biopsies were fixed in formalin, stained at pH 6 with X-Gal solution to measure SA-β-gal activity, and recorded using a photoscanner. (F) RNA was extracted from the biopsies and quantified by qRT-PCR for mRNA levels of endogenous p16^INK4a. Actin mRNA was used as a control (n = 4). Data show are the mean ± SD. **p < 0.01; ***p < 0.001.

Figure 3. Senescent Cells Are Induced and Necessary for Optimal Cutaneous Wound Healing

In all cases, mice were wounded using a 6 mm punch to dorsal skin and treated with PBS (vehicle control) or GCV (five daily i.p. injections) from 1 to 6 days after injury (n = at least 4 mice per group). (A and B) p16-3MR mice were wounded, injected i.p. with coelentarazine, and imaged with a Xenogen imaging system at the indicated times after injury. (A) Quantification of the luminescence. (B) Typical images, both at the indicated time (days) after injury. (C) Skin biopsies of p16-3MR mice were collected 6 days after injury, fixed, and stained for nuclei (DAPI; blue) or mRFP (immunostaining; red). White asterisks define the wound edges. (D–F) p16^INK4a (D), mRFP (E), and p21 (F) mRNA levels were quantified by qRT-PCR from skin biopsies excised from PBS or GCV-treated wounds to p16-3MR mice at the indicated intervals after injury. Actin was used to control for cDNA quantity. (G) Wound sizes of WT or p16-3MR mice were measured at the indicated days after wounding. (H) Representative image of wounds at the indicated days after injury of p16-3MR mice. (I) Wound sizes of WT, p16 KO, p21 KO, and p16/p21 DKO mice were measured at the indicated days after wounding. Data shown are the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.0001.

Figure 4. Characterization of Senescent Cells Induced during Wound Healing

(A) Longitudinal sections of wounds from PBS-treated (vehicle, left panels) and GCV-treated mice (right panels) (n = 8), collected 6 days after wounding, were stained with H&E or specific antibodies and scored. Dermis is defined as D and epidermis as E. In the H&E staining, black and green arrows indicate the original wound edges and re-epithelialized edges, respectively. Scale bars represent 500 μm. Sections were additionally stained with Masson’s trichrome (for collagen fibers) or antibodies against SMA (marker for myofibroblast), vimentin (marker for mesenchymal cells), or factor VIII (marker for endothelial cells). Higher magnification of the boxed areas is shown on the bottom to the right. Positive staining is indicated by arrows. Scale bars represent 100 μm for the main panel and 20 μm for the insets. (B) Table shows the percentage of epithelialization after injury and histological scores for angiogenesis, granulation, and inflammation, based both on the staining represented in (A). (C and D) Cells from wounds collected 6 days after injury were isolated and sorted for RFP. mRNA levels encoding the indicated proteins were quantified by qRT-PCR. Actin was used as a control for RNA quantity (n = 3 independent experiments). (E) Sections described in (A) were immunostained for p21. Black arrows indicate positive fibroblasts, and green arrows indicate positive endothelium. Insets show higher magnification of the section. Scale bars represent 200 μm for the main panel and 40 μm for the insets. Data shown are the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.0001.

Figure 5. PDGF-A Is Expressed and Secreted by Senescent Cells

(A) mRNA levels of the indicated proteins were quantified by qRT-PCR in cells sorted from wounds as described in Figure 4D. Actin was used to control for RNA quantity (n = 3). (B) Skin biopsies excised from PBS- or GCV-treated wounds in p16-3MR mice were collected at the indicated intervals after injury, and PDGFA mRNA levels were quantified by qRT-PCR. Actin was used to control for RNA quantity (n = 5). (C) Skin biopsies of p16-3MR mice were collected 6 days after injury, fixed, stained with DAPI (blue; indicating nuclei), and immunostained for mRFP (red) and PDGF-A (green). (D) Cells were derived from wounds 6 days after injury and sorted for RFP. RFP⁺ cells were plated, fixed 24 hr later, and then stained with DAPI (blue) and immunostained for PDGF-A (red) and vimentin (green, left panels) or factor VIII (green, right panels). (E and F) Murine skin fibroblasts (SF) or endothelial cells (EC) were mock irradiated (NS) or made senescent by irradiation (SEN; 10 Gy X-ray). At 4 days after irradiation, RNA and conditioned media were collected and analyzed for PDGF-A mRNA and secreted protein by qRT-PCR and ELISA, respectively (n = 4). Data shown are the mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.0001.

Figure 6. Senescence-Associated PDGF-A Drives Myofibroblast Differentiation

(A and B) Skin fibroblasts were incubated with conditioned media from IR-induced senescent (CM SEN) or nonsenescent (CM NS) cells containing nonspecific rabbit IgG or a PDGF-A blocking antibody; after 48 hr, the cells were immunostained for SMA (green), and nuclei were stained with DAPI (blue). Ten ng/ml recombinant PDGF-AA was used as a positive control. (A) Representative image. (B) Percentage of SMA-positive cells relative to the total (DAPI positive) number of cells (n = 6). (C and D) Wound healing was performed as described in Figure 3G, except PDGF-AA (20 ng) or vehicle (PBS) were topically applied daily from 1 to 6 d after wounding (n = 5). (C) Biopsies from wounds 6 days after injury were embedded in paraffin and stained for SMA. The graph shows the percentage of positive cells present in the wound gap. (D) Wound sizes were measured at the indicated times after injury. In addition to the groups described, a cohort of control p16-3MR mice was treated with PDGF-RA (daily topical application of 20 ng from 1 to 6 days after injury). (E) RNA was isolated from wounded areas 6 days after injury of WT and p16/p21 DKO mice. PDGFA mRNA levels were quantified by qRT-PCR. Actin was used to control for RNA quantity (n = 3). (F) Wound healing was performed and measured as described in (B). PDGF-AA (20 ng) or vehicle (PBS) were topically applied daily from 1 to 6 days after wounding (n = 3). Data shown are the mean ± SD. **p < 0.01, ***p < 0.0001.