Impaired Epidermal to Dendritic T Cell Signaling Slows Wound Repair in Aged Skin

Abstract

Aged skin heals wounds poorly, increasing susceptibility to infections. Restoring homeostasis after wounding requires the coordinated actions of epidermal and immune cells. Here we find that both intrinsic defects and communication with immune cells are impaired in aged keratinocytes, diminishing their efficiency in restoring the skin barrier after wounding. At the wound-edge, aged keratinocytes display reduced proliferation and migration. They also exhibit a dampened ability to transcriptionally activate epithelial-immune crosstalk regulators, including a failure to properly activate/maintain dendritic epithelial T cells (DETCs), which promote re-epithelialization following injury. Probing mechanism, we find that aged keratinocytes near the wound edge don't efficiently upregulate Skints or activate STAT3. Notably, when epidermal Stat3, Skints, or DETCs are silenced in young skin, re-epithelialization following wounding is perturbed. These findings underscore epithelial-immune crosstalk perturbations in general, and Skints in particular, as critical mediators in the age-related decline in wound-repair.

Keywords: Aging; DETC; STAT3; Skint; epidermal-immune cell cross-talk; re-epithelialization; wound healing.

Figure 1. Young and aged epidermis. A)

Schematic illustrating the differentiated layers of the epidermis. B) Images of semi-thin sections of young (2–4 months old) and aged (22–24 months old) skin stained with toluidine blue. Abbreviations: Epi, epidermis; Derm, dermis; HF, hair follicle; SubCu Fat, subcutaneous fat. Scale bars=100μm. C) Quantification of the thickness of epidermis and dermis of young and aged skin. N=8. Students t-test was used to measure statistical significance. D) Immunofluorescence images of young and aged skin labeled with antibodies (Abs) against keratin 14 (K14), β4-integrin (CD104), keratin 5 (K5) and keratin 10 (K10) [secondary Abs are color-coded as shown]. Sections were co-stained with DAPI (blue) to visualize nuclei. Scale bars=25μm. E) Volcano plot of in vivo RNA-seq data comparing young:aged basal keratinocyte transcripts. Vertical red colored lines denote fold changes greater ±2 fold. Horizontal red line denotes p-value > 0.05. Data are represented as mean ± SEM. See also Table S1.

Figure 2. Re-epithelialization of young and aged cutaneous wounds. A)

Immunofluorescence images of the temporal re-epithelization process that occurs following skin wounding (t=0) in young and aged mice. Sections are immunolabeled for basal epidermal keratinocytes (K14) and co-stained with DAPI. Scale bar=500μm. “S” denotes scab and yellow arrows denote wound edge. B) Schematics depicting progress of the re-epithelialization process at time-points indicated in young and aged animals. While wounds heal, the process is delayed in aged skin. C–D) Quantification of re-epithelialization in young and aged wounds (C) and of the length of the tongue of epidermal keratinocytes that migrate in from the wound edges during the wound-repair process (D). Students t-test was used to measure statistical significance, N=5. Data are represented as mean ± SEM. See also Figures S1 and S2.

Figure 3. Functional capabilities of young and aged keratinocytes. A)

Immunofluorescence images of young and aged wounds at d3 and d5 time-points with proliferating cells labeled with EdU (green). Dashed lines denote epidermal/dermal boundaries. Scale bars=100μm. Quantifications are of EdU incorporation from independent samples. N=7. B) Proliferation of young and aged basal layer keratinocytes under homeostatic conditions. Animals were pulsed with EdU and collected after 24 hours. EdU incorporation was measured by flow cytometry. Young (7.9±1.2%) and aged (7.6±0.5%), p=0.82, N=6. C) Proliferation of young and aged basal layer keratinocytes after depilation. EdU was 24 hours prior to collection. Graph shows percentage of basal layer keratinocytes with EdU incorporation post-depilation at indicated time-points. At 24hours post-depilation young (51.5±1.5%) and aged (28.6±7.0%), p=0.018, N=4. D) DIC images of explant cultures from young and aged tissue biopsies. Dashed lines denote the borders of keratinocyte outgrowth; yellow lines denote radial distances migrated. E=explant. Scale bars=10μm. N=12. Quantifications are of the area and distance of outgrowth of keratinocytes in explants during a 7 day time-course. E–I) Scratch wound assays. E) Migration into 600μm scratch wounds of aged and young skin keratinocytes was measured by time-lapse video-microscopy. Shown are DIC images of time-points indicated. Quantifications of wound closure are shown in the graph at right. F) EdU incorporation during the interval of the time-lapse imaging. G) Comparative measurements of distance migrated and velocity. H) Migration plots of individual cells during the wound closure. I) EdU incorporation during the interval of the time-lapse imaging. Students t-test was used to measure statistical significance. Data are represented as mean ± SEM. See also Figure S3 and Supplemental Movies S1 and S2.

Figure 4. Transcriptional profiling of wound response. A)

Heat map of differentially regulated transcripts between aged and young epidermal keratinocytes isolated from unwounded skin or the wound edge of skin and subjected directly to RNA-seq analyses. Yg wound, young keratinocytes isolated from the wound edge; aged wound, aged keratinocytes isolated from the wound edge; yg epi, young keratinocytes under homeostatic conditions; aged epi, aged keratinocytes under homeostatic conditions. Blue color denotes low FPKM expression, green high FPKM expression. B) Negatively and positively enriched GO terms in genes that were differentially regulated between aged wound and young wound samples. C) Volcano plot of differentially regulated genes between young wound and aged wound samples. Vertical rose-colored lines denote fold changes greater ± 2-fold. Horizontal rose line denotes p-value > 0.05. Blue dots indicate genes with a GO annotation relating to immune function (note marked failure of many of these genes to be up-regulated in aged wounds). D) Table of selected genes for indicated GO term from RNA-seq analysis (blue, down-regulated by log₂ of value; green, up-regulated by log₂ of value). Data are represented as mean ± SEM. See also Figure S4 and Tables S2–S6.

Figure 5. Immune cells and wound healing in aged skin. A)

Skins from young and aged mice either unwounded or within 1mm of a wound-edge were subjected to flow cytometry analysis using the scheme in Figure S4A. Percentages of specific immune subclasses relative to total immune cells (CD45+) are shown. DC, dendritic cells (MHCII⁺CD11c⁺); Macs, macrophages (CD64⁺CD11b⁺); Mono, monocytes (Ly6c^hiLy6g^neg); Neutro, neutrophils (Ly6c^negLy6g^hi); T-cells (γδTCR+TCRβ). N=6. B) Quantification of skin resident T-cells by flow cytometry. Data are presented as percentages of specific T-cell subclasses relative to total immune cells (CD45+). N=6. C) Quantifications of DETCs numbers from 0–700μm of the wound edge at indicated times after wounding. N=5. D) Schematic of epidermal sheet preparation (whole mount) of wound from birds-eye-view and maximum projections of z-series images of epidermal sheets are from skin adjacent to wound edge (yellow dotted lines), with immunostaining for γδ TCR to detect DETCs. Shown are images from d3 after wounding. E) At right are quantifications of DETC distribution, plotted in histograms as the distance proximal (0–400 μm) and distal (1200–1600 μm) from the wound edge. N=10. F) Sagittal imunofluorescence images of skin (wounded and unwounded), immunostained for DETCs. Dashed lines denote epidermal/dermal boundaries. Scale bars=25 μm. Insets show DETCs highlighted with arrows. G) Quantification of the numbers of dendrites per DETC. N=5. Students t-test was used to measure statistical significance. H) Whole mount DETC immunofluorescence of ear-skin. Shown are images prior to and at d1 and d3 after wounding. Scale bars = 100 μm. N=3. Yellow dotted lines denote wound edge (wd). I) Density plots of the distribution of rounded (no dendrites) of DETCs in ear-skin whole mount preparations at times post-wounding indicated. Vertical lines represent mean distance of rounded DETCs from wound edge (0 μm). J) Quantifications of DETCs at the wound site at times after injury. Data are represented as mean ± SEM. See also Figure S5.

Figure 6. Failure of keratinocytes to up-regulate Skints results in impaired wound healing in young mice. A)

Heatmap of Skint gene family expression from RNA-seq data. Asterisks denotes splice variant. Yg wound=young keratinocytes isolated from the wound edge, aged wound= aged keratinocytes isolated from the wound edge, young epi=young keratinocytes under homeostatic conditions, aged epi=aged keratinocytes under homeostatic conditions. B) qRT-PCR of Skint mRNAs from keratinocytes isolated from wound edges. C) Illustration in utero lentiviral infections into amniotic sacs of E9.5 embryos and selective transduction of mouse skin epidermis. D) Knockdown efficiency of Skint shRNAs as measured by qRT-PCR of adult epidermis prior to wounding. E) Immunofluorescence images of d3 backskins of wounded young mice whose epidermises were transduced in utero for the Skint shRNAs indicated. Scr, scrambled. Tissue sections are immunolabeled for K14 (green) and DAPI (blue). S, scab; arrows denote wound-edge. Scale bar= 500 μm. F) Wound closure at d5 of young mice transduced for the indicated shRNAs. N=4. G) Quantification of DETC number and number of dendrites per DETC (H) in sections of unwounded and wounded skins transduced as indicated. N=4. I) Immunofluorescence images of back-skins of re-epithelialization process following 5d or 3d after wounding of young mice of the strains indicated. Note that FVB_Jax lacks Skints 3-4-9; FVB_Tac lacks Vγ5Vδ1 DETCs. Tissue sections are immunostained for K14 (green) and DAPI (blue). S, scab; arrows denote wound-edge. Scale bar = 100 μm. N=2. J) Quantification of wound closure by re-epithelialization. K) Quantificaton of dendrities per DETC from tissue sections of wounds in C57BL6 and FVB_Jax at time-points indicated. L) Whole mount immunofluorescence and quantifications of DETCs in ear-skin of young FVBJax versus C57BI/6 mice. Scale bars= 100 μm. N=2. M) Density plots of the distribution of rounded (no dendrites) of DETCs in ear-skin whole mount preparations at times post-wounding indicated. Vertical lines represent mean distance of rounded DETCs from wound edge (0 μm). Data are represented as mean ± SEM. See also Figure S6.

Figure 7. STAT3 signaling regulates Skint expression. A)

Skint mRNA levels after IL-6 treatment of primary WT keratinocytes in vitro. B) Relative in vivo expression of Skint genes in keratinocytes isolated from d3 wound edges. C) Immunofluorescence images of sagittal wounded skin sections from young and aged WT mice. Labeling is for Abs against pSTAT3 (green), K5 (red), and DAPI (grey). Dashed line denotes epidermal/dermal boundaries, yellow arrows denote wound edges, and “S” denotes scab. Scale bars = 100 μm. D) Quantification of the epithelial tongue length at backskin wounds at times post-wound indicated. N=4. E) Quantifications of DETC numbers and (F) morphologies from d3 wound edges. N=4. G) Whole mount immunofluorescence of ear skins imaged at times indicated after wounding. Yellow dotted line denotes wound edge. Scale bar = 100 μm. N=2. H) Density plots of the distribution of rounded (no dendrites) DETCs in ear-skin whole mount preparations at times post-wounding indicated. Vertical lines represent mean distance of rounded DETCs from wound edge (0 μm). I) Quantifications of DETC numbers in ear skin whole mounts after wounding at time-points indicated. J) DIC images (from d5 time-point) of explant cultures from young and aged tissue biopsies treated with IL-6 at concentrations indicated. Dashed lines denote the borders of keratinocyte outgrowth; E, explant. Scale bars=10μm. N=8. K) Quantifications are of the distance of outgrowth of keratinocytes in explants during a 7 day time-course. N=8. Data are represented as mean ± SEM. See also Figure S7.