MCL‑1 safeguards activated hair follicle stem cells to enable adult hair regeneration

Abstract

Hair follicles cycle through expansion, regression and quiescence. To investigate the role of MCL‑1, a BCL‑2 family protein with anti‑apoptotic and apoptosis‑unrelated functions, we delete Mcl‑1 within the skin epithelium using constitutive and inducible systems. Constitutive Mcl‑1 deletion does not impair hair follicle organogenesis but leads to gradual hair loss and elimination of hair follicle stem cells. Acute Mcl‑1 deletion rapidly depletes activated hair follicle stem cells and completely blocks depilation‑induced hair regeneration in adult mice, while quiescent hair follicle stem cells remain unaffected. Single‑cell RNA‑seq profiling reveals the engagement of P53 and DNA mismatch repair signaling in hair follicle stem cells upon depilation‑induced activation. Trp53 deletion rescues hair regeneration defects caused by acute Mcl‑1 deletion, highlighting a critical interplay between P53 and MCL‑1 in balancing proliferation and death. The ERBB pathway plays a central role in sustaining the survival of adult activated hair follicle stem cells by promoting MCL‑1 protein expression. Remarkably, the loss of a single Bak allele, a pro‑apoptotic Bcl‑2 effector gene, rescues Mcl‑1 deletion‑induced defects in both hair follicles and mammary glands. These findings demonstrate the pivotal role of MCL‑1 in inhibiting proliferation stress‑induced apoptosis when quiescent stem cells activate to fuel tissue regeneration.

Fig. 1. Dynamic expression of BCL‑2 family members at different stages of the HF cycle.

Representative confocal images showing the expression of MCL‑1 (A), BCL‑XL (B), BIM (C) and BAK (D) at p60 (telogen), p22 (anagen II‑III), p29 (anagen V‑VI) and p40 (catagen). Basal skin layer is stained by KRT14, hair matrix containing TACs is stained by PCAD and DAPI is used as a nuclear counterstain in each image. Data shown are representative of n = 3 independent experiments. Scale bars, 50 µm.

Fig. 2. MCL‑1 is essential for murine HF during adulthood.

A, B Deletion of Mcl‑1 in the skin epithelium leads to progressive hair loss and HF destruction. Representative photographs of Mcl‑1^f/f or Mcl‑1^f/+, K5‑Cre/Mcl‑1^f/+ and K5‑Cre/Mcl‑1^f/f mice at p12, p29, p60 and p90 (A), alongside H&E sections of dorsal skin at these time points (B). Data represent n = 8 mice per genotype. Scale bars, 100 µm. C Representative FACS plots demonstrating efficient deletion of the floxed Mcl‑1 gene, using hCD4 as an indicator of efficient CRE‑mediated recombination as well as promoter activity. Expression of hCD4 in hair bulge (CD49f⁺/CD34⁺/SCA‑1^–), non‑bulge HF cells (CD49f⁺/ CD34^–/SCA‑1^–) and interfollicular epidermis (IFE) (CD49f⁺/CD34^–/SCA‑1⁺) from Mcl‑1^f/f or Mcl‑1^f/+, K5‑Cre/Mcl‑1^f/+ and K5‑Cre/Mcl‑1^f/f mice. n = 3 at p21 per genotype. D qPCR analysis reveals a significant reduction in Mcl‑1 transcript levels in HF cells (CD49f⁺/SCA‑1^–) and IFE (CD49f⁺/SCA‑1⁺), but not in non‑epithelial cells (CD49f^–) of K5‑Cre/Mcl‑1^f/f mice at p21, compared to control littermates. Data are presented as mean of three technical replicates for each point from two independent litters. E Representative confocal images show the absence of MCL‑1 protein in HFs of K5‑Cre/Mcl‑1^f/f mice compared to K5‑Cre/Mcl‑1^f/+ and Mcl‑1^f/f or Mcl‑1 ^f/+ littermate controls at p29 (mid anagen). Data represents n = 5 experiments. Scale bar, 50 µm. F Mcl‑1 deletion leads to apoptosis in HFs, as shown by CC3 staining at various stages: telogen (p60), early anagen (p22), mid anagen (p29), and catagen (p40), representative of n = 5 independent mice per genotype. Basal skin is marked by KRT14, hair matrix with TACs is stained with PCAD, and DAPI is used as nuclear counterstain. Scale bar, 50 µm. G Quantification of CC3⁺ cells per HF at p29 shows significantly increased apoptosis in K5‑Cre/Mcl‑1^f/+ mice (n = 70) vs. controls (n = 40). Data are presented as mean ± SEM for n = 3 mice per genotype. ****p < 0.0001, Student’s t‑test.

Fig. 3. The maintenance of adult CD34⁺ HF bulge stem cells is dependent on MCL‑1.

A, B Deletion of Mcl‑1 resulted in the gradual reduction of the CD34⁺ HF bulge cell subset. Representative FACS plots for skin epithelial cells fractionated based on CD34 and SCA‑1 expression from K5‑Cre/Mcl‑1^f/f mice or littermate control mice at p21, p60 and p90 (n = 4) (A). Bar graphs showing the percentages of the indicated cell sub‑populations within the CD49f⁺ epithelium isolated from the dorsal skin of K5‑Cre/Mcl‑1^f/f, K5‑Cre/Mcl‑1^f+, Mcl‑1^f/f or Mcl‑1^f/+ mice at p21, p60 and p90. Data are presented as mean ± SEM (n = 4 mice for each genotype). *p < 0.05; **p < 0.01; ***p < 0.001; n.s., non‑significant; Student’s t‑test (B). C Western blot analysis of the indicated BCL‑2 family members in the CD49f^– (non‑epithelial cells), CD49f⁺/SCA‑1^– (HFs), CD49f⁺/SCA‑1⁺ IFE populations of HFs at telogen (p60). Data shown are representative of n = 4 independent experiments. D Representative confocal images showing the expression of MCL‑1 and BIM protein in HF stem cells marked by CD34⁺ and KRT15⁺ in hair bulges at telogen phase (p60). Representative of n = 3 experiments. Scale bar, 50 µm. Hair matrix containing TACs is stained by PCAD, and DAPI is used as a nuclear counterstain in each image. E In vitro clonogenic assay of the CD34⁺ HF bulge stem cells. 100 Lin^–/CD49f⁺/CD34⁺/SCA‑1^– cellsisolated from the skin of K5‑Cre/Mcl‑1^f/f mice and littermate control mice at p21 were grown on 3T3 feeder cells for 14 days. Bar graphs show the numbers of colony‑forming cells among the seeded CD34⁺ HF bulge cells. Data are presented as mean ± SEM for n = 3 independent experiments. **p < 0.01; Student’s t‑test.

Fig. 4. Acute deletion of Mcl‑1 in the skin epithelium abrogates hair regeneration in adulthood.

A Experimental strategy for hair regeneration assay in adult K5‑Cre^ER/Mcl‑1^f/f mice. Deletion of Mcl‑1 was induced through administration of a tamoxifen containing diet between p55‑p62. A portion of the dorsal skin was shaved and waxed to induce hair regeneration. Skin samples were harvested at the indicated timepoints and analyzed. B Representative images of mice taken at 0, 2, 4, 7, 11 wpd. n = 8–10 mice per genotype for each timepoint. C Representative H&E‑stained histological sections of dorsal skin harvested at 3 dpd, 9 dpd and 11 wpd. n = 6–8 mice per genotype for each timepoint. Scale bars, 100 µm. D–F Confocal microscopy analysis of HFs from K5‑Cre^ER/Mcl‑1^f/f mice and littermate control mice at 3 dpd and 9 dpd. Representative confocal images depicting increased apoptosis indicated by CC3⁺ cells (D), proliferative cells as marked by PCNA staining (E), and the loss of CD34⁺ HFSC post‑depilation (F) in Mcl‑1 deficient mice. n = 6–8 mice per genotype. Scale bars, 50 µm. Basal skin layer is stained by KRT14, hair matrix containing TACs is stained with PCAD and DAPI is used as a nuclear counterstain in each image in (D–F). G–J FACS analysis of dorsal skin epithelial cells at 6 dpd (G, H) and 11 wpd (I, J), fractionated by CD34 and SCA‑1 expression. H, J Bar graphs show percentages of each cell sub‑population within CD49f⁺ epithelium, mean ± SEM (n = 7 mice for 6 dpd; n = 3 for 11 wpd per genotype). ****p < 0.0001; **p < 0.01; *p < 0.05, Student’s t‑test.

Fig. 5. Single‑cell transcriptome atlas of HF cells from murine dorsal skin at early depilation‑induced anagen.

A UMAP representation of all HF cells in the murine dorsal skin across six time points at the early anagen stages of hair regeneration after depilation, colored according to the main cell classes. The plot was generated by integration of 1732, 1627, 2342, 2500, 6322 and 4433 cells for 0 h, 4 h, 16 h, 2 d, 4 d and 6 d post‑depilation, respectively. B Dot plot of marker genes used for annotation of cell clusters in (A). The colour scales correspond to the averaged expression levels. The dot size corresponds to the ratio of cells expressing the gene in the cell type. C Heatmap of top 10 marker genes for each cell cluster in (A). D UMAP visualization of HF cells in each time point post‑depilation as indicated. E Number of differentially expressed (DE) genes for each cell type during hair regeneration along the time course. F A heatmap showing the upregulated and downregulated DE genes in the p53 signaling, mismatch repair and cell cycle pathways in cells from the OB1 and cycling_OB1 clusters over the course of hair regeneration.

Fig. 6. Deletion of Trp53 rescues hair regeneration defect in Mcl‑1‑deficient mice.

A, B Confocal analysis of HFs from FVB mice at P60, sampled at 0, 2, 4, and 6 days post depilation (dpd) to induce hair regeneration. Increased DNA damage (pH2A.X⁺) is evident in HF cells post‑depilation (A). Basal layer marked by KRT14, hair matrix by PCAD, DAPI as nuclear counterstain. DNA damage observed in a subset of CD34⁺ and KRT15⁺ HFSCs at 4 dpd (B). Data represent n = 3 mice per time point. Scale bars, 50 µm. C Experimental strategy for hair regeneration assay in adult K5‑Cre^ER/Mcl‑1^f/f/Trp53^flf mice. Mcl‑1 and Trp53 deletion was induced via tamoxifen‑containing diet between p55‑p62. A portion of the dorsal skin was shaved and waxed to induce hair regeneration. Dorsal skin samples were harvested at the indicated timepoints for analysis. D qPCR analysis showing reduced Mcl‑1 and Trp53 transcript levels in HF cells (CD49f⁺/SCA‑1^–) isolated from K5‑Cre^ER/Mcl‑1^f/f/Trp5^f/f mice, compared to littermate control. n = 2 mice per genotype. E Representative H&E‑stained histological sections of dorsal skin harvested at 9 dpd. n = 6 mice per genotype for each timepoint. Scale bars, 100 µm. F Representative images of mice taken at 0, 15, 30 dpd. n = 3 mice per genotype for each timepoint. G, H Confocal microscopy images of HFs from K5‑Cre^ER/Mcl‑1^f/f/Trp53^flf mice and littermate controls at 9 dpd, showing efficient CRE‑mediated recombination with upregulation of hCD4 (G) and downregulation of MCL‑1 (H). n = 6 mice per genotype for each timepoint. Scale bar, 50 µm. Basal layer is stained by KRT14, hair matrix stained by PCAD and DAPI as a nuclear counterstain.

Fig. 7. Rescue of HF defects in Mcl‑1‑deficient mice by concomitant loss of Bim or Bak.

A Schematics of experimental strategies to assess whether deletion of Bim or Bak rescues HF defects in the inducible K5‑Cre^ER/Mcl‑1^f/f (left panel) and constitutive K5‑Cre/Mcl‑1^f/f (right panel) models. K5‑Cre^ER/Mcl‑1^f/f mice with deletion of single or both alleles of Bim or Bak at p55 were administered with tamoxifen containing diet to induce Mcl‑1 deletion, followed by depilation. Skin samples were harvested at 9 dpd for analysis. K5‑Cre/Mcl‑1^f/f mice with single or both Bak alleles deleted were generated and analyzed. B, C Bim deletion partially rescues HF regeneration defects in K5‑Cre^ER/Mcl‑1^f/f mice. Representative H&E‑stained sections of dorsal skins harvested at 9 dpd. n = 6–8 mice per genotype. Scale bars, 100 µm (B). qPCR analysis of Bim expression in FACS sorted subpopulations. Data are mean ± SEM from 3 mice per genotype (C). D, E Bak deletion completely rescues HF regeneration defects in K5‑Cre^ER/Mcl‑1^f/f mice. Representative H&E‑stained histological sections of dorsal skins harvested at 9 dpd. n = 7–9 mice per genotype. Scale bars, 100 µm (D). qPCR analysis of Bak expression in FACS sorted subpopulations. Data are mean ± SEM from 3 mice per genotype (E). F Representative images of K5‑Cre/Mcl‑1^f/f/Bak^+/+, K5‑Cre/Mcl‑1^f/f/Bak^+/– and K5‑Cre/Mcl‑1^f/f/Bak^–/– mice at p90. n = 7–8 mice per genotype. Representative H&E‑stained sections and confocal images of dorsal skins harvested at p90. Scale bars, 100 µm (H&E), 50 µm (confocal). Basal layer stained by KRT14, hair matrix stained by PCAD and DAPI as nucleus counterstain. G, H FACS analysis of CD34 and SCA‑1 expression in epithelial cells from mice of indicated genotypes at p90 (G). Bar graph shows percentage of each subpopulation in CD49f⁺ epithelium (H). Data are mean ± SEM, n = 3–5 per genotype. *p < 0.05; **p < 0.01; n.s., non‑significant; Student’s t‑test. I, J In vitro clonogenic assay of CD34⁺ HF stem cells from mice of indicated genotypes at p21, grown on 3T3 feeder cells for 14 days. J Colony counts for each genotype, mean ± SEM for n = 3 independent experiments. **p < 0.01; Student’s t‑test.

Fig. 8. Inhibition of ERBB signaling blocks MCL‑1 protein expression and hair regeneration.

A Kinase inhibitor screen in primary keratinocytes identifies pathways that modulate MCL‑1 protein levels. Cells from newborn FVB/N pups were treated with 2 µM of each inhibitor from a panel of 154 kinase inhibitors for 24 h, followed by Western blotting for MCL‑1 expression. The intensities of MCL‑1 and ACTIN were analysed using Image J. The ratio of MCL‑1 vs Actin in inhibitor‑treated cells with normalization to DMSO treated control cells were plotted as a bar graph. Inhibitors were grouped based on the kinase targeted. Groups highlighted in blue show more than a 50% reduction of MCL‑1 levels on average (MCL‑1 vs Actin ≤ 0.5). B Validation of ERBB inhibitors from (A) shows significant reduction in MCL‑1 protein levels after 24 h of treatment (n = 2 independent experiments). C Western blot of keratinocytes treated with various afatinib concentrations for 24 h demonstrates MCL‑1 reduction without affecting other BCL‑2 family proteins (n = 3 independent experiments). D Afatinib inhibits phosphorylation of protein translation regulators in primary keratinocytes. Western blot of p‑EGFR, p‑ERBB2, p‑P70 S6K, p‑S6, p‑eIF4B, and p‑eEF2K (n = 2 independent experiments). E qPCR analysis of afatinib‑treated keratinocytes reveals Mcl‑1 transcript levels remain stable. Data are mean ± SEM (n = 3 independent experiments). F H&E‑stained sections of dorsal skin from wildtype FVB/N mice treated with 50 mg/kg afatinib or vehicle followed by depilation show impaired hair regeneration at 9 dpd (n = 9 mice per treatment group). Scale bars, 50 µm. G, H FACS plots and bar graphs indicate changes in CD34 and SCA‑1 populations in CD49f⁺ epithelial cells from afatinib‑ or vehicle‑treated mice at 9 dpd. Data are mean ± SEM. ****p < 0.0001; **p < 0.01; *p < 0.05; Student’s t‑test. I–K Confocal images of HFs from afatinib‑ or vehicle‑treated mice at 9 dpd. I MCL‑1 and p‑ERBB2 expression; J MCL‑1 and p‑S6 expression; K Proliferating (PCNA⁺) and apoptotic (CC3⁺) cells. Scale bars, 50 µm. Basal layer stained by KRT14 and DAPI as nuclear counterstain (n = 6 mice per treatment arm).