Treg-Cell Control of a CXCL5-IL-17 Inflammatory Axis Promotes Hair-Follicle-Stem-Cell Differentiation During Skin-Barrier Repair

Abstract

Restoration of barrier-tissue integrity after injury is dependent on the function of immune cells and stem cells (SCs) residing in the tissue. In response to skin injury, hair-follicle stem cells (HFSCs), normally poised for hair generation, are recruited to the site of injury and differentiate into cells that repair damaged epithelium. We used a SC fate-mapping approach to examine the contribution of regulatory T (Treg) cells to epidermal-barrier repair after injury. Depletion of Treg cells impaired skin-barrier regeneration and was associated with a Th17 inflammatory response and failed HFSC differentiation. In this setting, damaged epithelial cells preferentially expressed the neutrophil chemoattractant CXCL5, and blockade of CXCL5 or neutrophil depletion restored barrier function and SC differentiation after epidermal injury. Thus, Treg-cell regulation of localized inflammation enables HFSC differentiation and, thereby, skin-barrier regeneration, with implications for the maintenance and repair of other barrier tissues.

Keywords: CXCL5; IL-17; Lrg5; barrier repair; epidermis; hair follicle stem cells; regulatory T cell (Treg); skin; stem cell.

Figure 1. Regulatory T cells facilitate epidermal regeneration after injury.

(a) Schematic showing diphtheria toxin (DT) administration schedule after skin barrier disruption. Cntrl animals are DT-treated WT mice or FoxP3^DTR littermates not given DT. (b) Transepidermal water loss (TEWL) on the indicated days of recovery. (c) qRT-PCR of epidermal differentiation genes normalized to β2m during barrier regeneration from the back affected back skin of Cntrl and Treg cell-depleted mice. (d) Representative histology 2 days after barrier injury and (e) quantification of epidermal thickness of affected back skin of Cntrl and Treg cell-depleted mice throughout the course of barrier repair. Results in b are representative of > 10 experiments (n=2–4 mice per group). Results in d & e are representative of 3 experiments (n = 3 mice per group) Scale bar in d is 50 μm. See also Supplementary Figures 1 and 2.

Figure 2. Regulatory T cells preferentially regulate Th17 and neutrophil accumulation in skin early during barrier repair

(a) Representative flow cytometry plots of IL-17A⁺ CD4⁺ (Th17) and IFN-γ⁺ CD4⁺ (Th1) T cells in skin following PMA/ionomycin stimulation of single cell suspensions 4 days after skin barrier disruption. (b) Quantification of Th17 and (c) Th1 cells of Treg cell-depleted mice following skin injury compared to Treg cell-sufficient cntrls (Cntrl) at the indicated times of barrier recovery. (d) Representative flow cytometry plots of CD11b⁺ Ly6G⁺ cells (neutrophils) in the skin 4 days after injury. Plots are pregated on live, CD45⁺ cells. (e) Percent of CD45⁺CD11b⁺Ly6G⁺ neutrophils in skin at the indicated times of barrier recovery. (f) Percent of IL-17⁺ CD4⁺, CD8⁺ and γδ T cells in the skin of Cntrl and Treg cell-depleted mice following PMA/ionomycin stimulation of single cell suspensions 4 days after skin barrier disruption. (g) Absolute number of the indicated cell types in the skin of Cntrl and Treg cell-depleted mice 4 days after injury. d4 – day 4; d7 – day 7; Uninj- uninjured skin; For all relevant panels, error bars are +/−SEM. n.s. – no significance; *p < 0.05; **p < 0.01; ***p < 0.001 according to a Student’s t-test. (n=3 mice per group) Data are pooled (b,c,e) or representative (f-h) of two independent experiments. (See also Supplementary Figure 3).

Figure 3. Lgr5-derived hair follicle stem cells contribute to epidermal repair

(a) Diagram of hair follicle anatomy. Following epidermal injury, cells derived from HFSCs located in the bulge region migrate into the upper hair follicle (isthmus and infundibulum) and interfollicular epidermis (IFE) to participate in epithelial repair. (b) Schematic of Lgr5-tdTom mice. Lgr5 drives expression of eGFP and a tamoxifen inducible Cre (CreERT2). These mice are crossed to R26-tdTomato mice. Injection with tamoxifen allows for inducible and permanent labeling of Lgr5⁺ stem cells and their progeny. (c) The back skin of Lgr5-tdTom mice were injured (as in Supplementary Figure 1a) and injected with tamoxifen on day 0. Mice were harvested on days 2,4 and 7 of recovery and compared to uninjured Lgr5-traced mice. (d) Representative high-power images of Lgr5-traced cells in the bulge, upper hair follicle and IFE during barrier regeneration compared to uninjured control. (e) Quantification of Lgr5-derived cells in the IFE during epidermal repair. Error bars are +/− S.E.M. n.s.- no significance; **p < 0.01; *** p<0.001. n=2–4 mice per group. IFE – Interfollicular epidermis; Infund. – Infundibulum.

Figure 4. Treg cells facilitate HFSC migration and differentiation during epidermal regeneration.

(a) Schematic showing DT and tamoxifen (Tam) administration in Lgr5-tdTom-FoxP3^DTR mice during skin barrier recovery for experiments described in panels b-f. Mice are compared to tamoxifen-labeled, non-DT treated littermates and harvested 4 days after skin barrier disruption. (b) Representative images demonstrating the patterns of stem cell progeny (tdTomato⁺ cell) localization in the skin of tamoxifen-labeled, Lgr5 stem cell lineage tracing mice after injury. Mice are either sufficient (-DT) or depleted (+DT) of Treg cells. (c) Quantification of Lgr5-derived cell (tdTomato⁺) localization in the upper hair follicle (isthmus and infundibulum; left panel) and IFE (right panel). tdTomato⁺/ HF and tdTomato⁺/hpf represent cell numbers normalized per hair follicle and per high power field respectively. See also Supplementary Figure 4. (d) FACS purified tdTom⁺ cells from DT-treated (i.e. Treg cell-depleted) or untreated Lgr5-tdTom-FoxP3^DTR (i.e. Treg cell-sufficient) mice were analyzed by RNA-sequencing. Comparison plots of normalized gene expression of tdTomato⁺ cells isolated from Treg cell-sufficient and Treg cell-depleted Lgr5-tdTom-FoxP3^DTR mice. Blue dots represent genes with a p-adjusted value < 0.05 and > 2-fold difference in gene expression between groups. (e) Volcano plot comparing the p-adjusted value versus fold change for tdTomato⁺ cells isolated from Treg cell-sufficient and Treg cell-deficient Lgr5-tdTom-FoxP3^DTR mice. Blue dots represent genes with a p-adjusted value < 0.05 and > 2 fold differences in gene expression between groups. (f) Heat map of gene transcripts of tdTomato⁺ cells isolated from Treg cell-sufficient and Treg cell-depleted Lgr5-tdTom-FoxP3^DTR mice focusing on differentially expressed genes with a padj. value < 0.05, fold change differences > 2 and predominant expression in either the bulge or IFE/upper hair follicle in adult mice.(Joost et al., 2016) (g) Cntrl and FoxP3^DTR mice were treated as diagramed in Figure 1a and compared to uninjured WT mice. Representative plots of CD34⁺ cells in the epidermis 11 days after epidermal injury. Plots are pre-gated on live, CD45-negative epidermal cells. (h) Percentage of CD34⁺ epidermal cells in uninjured cntrl mice compared to barrier-disrupted cntrl and Treg cell-depleted mice 5 and 11 days after injury. See also Supplementary Figure 5. For all relevant panels, error bars are +/− SEM. *p < 0.05; **p < 0.01; **** p<0.0001 by Student’s t-test or by one-way ANOVA for panel h. Results in b-c are representative of > 8 independent experiments (n=2 to 4 mice per group); d-f are representative of 2 independent experiments (n=2 mice per group) Results of h (right panel) are pooled data from 3 independent experiments. (n=2–4 mice per group) and representative of 2 experiments (left panel). Scale bar in b is 100 µm; Red-tdTomato⁺ Lgr5 stem cell progeny; Blue – DAPI.

Figure 5. Treg cells preferentially regulate cxcl5 expression during epidermal barrier repair.

Cntrl and DT-administered FoxP3^DTR mice (Treg cell-depleted) were treated as in Figure 1a, harvested 4 days after skin injury and compared to uninjured mice. (a) Scatter plot and (b) quantification of chemokines by qRT-PCR array of uninjured WT vs. epidermal injured WT mice. (c) Scatter plot and (d) quantification of chemokines from injured cntrl vs. injured Treg cell-depleted mice. (e) Plot and (f) chemokine quantification of uninjured cntrl and uninjured Treg cell-depleted mice. (g) cxcl5 expression from the indicated sort purified epidermal cells (See also Supplementary Figure 3a for gating strategy). Representative of 2 independent experiments (n=3 mice per group) Data are +/− SEM *** p< 0.001 comparing cxcl5 expression in panel d versus f by Student’s t-test.

Figure 6. CXCL5 neutralization restores skin barrier function and HFSC differentiation in the absence of Treg cells.

FoxP3^DTR mice were treated as in Figure 1a. Mice were co-administered α-CXCL5 mab or isotype control with DT on days 0,1 and 3 and harvested on day 4. (a) Representative flow cytometry plots of Ly6G⁺ CD11b⁺ neutrophils in skin. Plots are pregated on live CD45⁺ cells. (b) Percent and absolute number of Ly6G⁺ CD11b⁺ neutrophils in skin. (c) Transepidermal water loss (TEWL) 4 days after epidermal injury in Treg cell-depleted mice treated with either α-CXCL5 or isotype control antibody compared to injured Treg cell-sufficient cntrl mice. (d) Lgr5-tdTom-FoxP3^DTR mice were treated as in Figure 2a and co-administered α-CXCL5 or isotype control mAb with DT on days 0,1 and 3 of recovery. Mice were harvested on day 4. Representative IF images of Lgr5-derived cell (tdTomato⁺) localization in the epidermis compared to Treg cell-sufficient Lgr5-tdTom-FoxP3^DTR mice. (e) Quantification of Lgr5-derived cell localization in the epidermis of the indicated strains and conditions 4 days after skin injury. tdTomato⁺/ HF and tdTomato⁺/hpf are cell numbers normalized per hair follicle and per high power field respectively (f) Representative plots and (g) quantification of CD34⁺ HFSCs in the epidermis of Cntrl and FoxP3^DTR mice treated with α-CXCL5 or isotype control. For all relevant panels, error bars are +/−SEM. *p < 0.05. Results of panels a-f are representative of 3–5 independent experiments (n=2–4 mice per group). Results of g are pooled data from 3 independent experiments. Scale bar in d is 100µm; Red- tdTomato⁺ Lgr5 stem cell progeny; Blue – DAPI.