Restoration of hair follicle inductive properties by depletion of senescent cells

Abstract

Senescent cells secrete a senescence-associated secretory phenotype (SASP), which can induce senescence in neighboring cells. Human dermal papilla (DP) cells lose their original hair inductive properties when expanded in vitro, and rapidly accumulate senescent cells in culture. Protein and RNA-seq analysis revealed an accumulation of DP-specific SASP factors including IL-6, IL-8, MCP-1, and TIMP-2. We found that combined senolytic treatment of dasatinib and quercetin depleted senescent cells, and reversed SASP accumulation and SASP-mediated repressive interactions in human DP culture, resulting in an increased Wnt-active cell population. In hair reconstitution assays, senolytic-depleted DP cells exhibited restored hair inductive properties by regenerating de novo hair follicles (HFs) compared to untreated DP cells. In 3D skin constructs, senolytic-depleted DP cells enhanced inductive potential and hair lineage specific differentiation of keratinocytes. These data revealed that senolytic treatment of cultured human DP cells markedly increased their inductive potency in HF regeneration, providing a new rationale for clinical applications of senolytic treatment in combination with cell-based therapies.

Keywords: cellular senescence; dasatinib; dermal papilla; hair follicle; quercetin; regeneration; senescence‐associated secretory phenotype; senolytic.

FIGURE 1

Senescent DP cells emerge as early as passage 1 in human DP culture. (a) Experimental scheme to characterize the cellular senescence of human DP cells after in vitro expansion. (b) Cell morphology and SA‐β‐Gal activity in human DF culture showing that SA‐β‐Gal⁺ senescent DF cells were increased over passage. (c) Cell morphology and SA‐β‐Gal activity in human DP culture showing that SA‐β‐Gal⁺ senescent DP cells were increased over passage. (d, f) Immunofluorescence and quantification of human DP culture (n = 64 cells from 3 biological replicates) showing that senescent DP cells exhibited markedly expanded cytoplasm and enlarged nuclear size showing a strong positive correlation between the size of cytoplasm and the size of nucleus. (e) Quantification of number of SA‐β‐Gal⁺ senescent cells in human DF and DP cultures (n = 6 DF replicates and 8–9 DP replicates from 2 independent donors) showing that SA‐β‐Gal⁺ cells senescent were more prevalent in DP culture than in DF culture as early as P1 to P5. (g, i) Immunofluorescence and quantification of senescent DP cells (n = 114 [P1], 99 [P3], 81 [P5] cells from 2 independent donors) showing that nuclear p16 expression was increased in senescent DP cells over subsequent passage. (h, j) Immunofluorescence and quantification of senescent DP cells (n = 129 [P1], 107 [P3], 101 [P5] cells from 2 independent donors) showing that nuclear p21 expression was increased in senescent DP cells over subsequent passage and that p21⁺ cells showed no p27 expression in the nucleus, which indicates senescence, not quiescence. (k) Quantification of Ki67⁺ cells showing that the percentage of proliferating cells decreased over subsequent passages (n = 4 biological replicates/group). (l) Quantification of senescent cells showing that the percentage of proliferating cells decreased over subsequent passages (n = 8 biological replicates/group). Data are presented as the mean ± s.e.m. *p < 0.05, **p < 0.01, ***p < 0.001 (two‐way analysis of variance with Sidak's multiple comparisons test) in (e); (linear regression test) in (f); (Welch's t test) in (i–l).

FIGURE 2

SASP from senescent DP cells have repressive effects on nearby non‐senescent DP cells. (a, b) Protein arrays and quantification of the conditioned medium of human DP culture (n = 3 biological replicates) showing that a unique combination of SASP factors, including IL‐6, IL‐8, MCP‐1, TIMP‐2, osteopontin, and osteoprotegerin, accumulated in the conditioned medium over passage. (c) Western blot of cultured DP cells showing that marked increase of p21 and p16 with IL‐6 and p‐Stat3 in P5 DP culture compared to P1 DP culture in contrast to marked decrease of Gli1 and Lef1 in P5 DP culture compared to P1 DP culture. (d) Immunofluorescence showing that enlarged senescent cells exhibited cytoplasmic IL‐6 expression with nuclear GATA4 expression. (e, f) Immunofluorescence showing that p‐Stat3 activation (IL‐6 signaling) and p‐Smad2 activation (TGF‐β signaling) with decreased Gli1 and nuclear β‐catenin expression in non‐senescent DP cells nearby enlarged senescent DP cells in P3 DP culture compared to P1 DP culture. (g) Volcano plot of differentially expressed genes showing that RELN, GAS6, THBS2, CDKN2A, IL6, CDKN2B, and TP53INP1 were upregulated and GNG2, APCDD1, STMN2, LEF1, MKI67 were downregulated in cultured DP cells over passage (n = 3 biological replicates). (h) Gene ontology analysis showing that biological pathways including cellular senescence, cell–cell communication, intracellular signaling by second messengers, signaling by TGF‐beta family members, DNA repair, regulation of TP53 activity, DNA double‐strand break repair, and interleukin‐1 family signaling were upregulated in cultured DP cells. (i–k) Representative gene expression showing that the senescence drivers CDKN2A and CDKN2B were significantly upregulated along with SASP factors IL6 and CCL2, whereas MKI67 and LEF1 were significantly downregulated over passage in human DP culture (n = 3 biological replicates). Data are presented as the mean ± s.e.m. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 (two‐way analysis of variance with Dunnett's multiple comparisons test) in (b); (repeated measure one‐way analysis of variance with Sidak's multiple comparisons test) in (i–k).

FIGURE 3

Senolytic treatment reverses senescence‐driven repressive effects in human DP culture. (a) Experimental scheme to determine the effect of senolytic treatment on human DP culture. (b) Cell morphology and SA‐β‐Gal activity of P3 DP culture after vehicle or dasatinib plus quercetin (D + Q) treatment. (c) Quantification of senescent cells (n = 4 biological replicates/group) showing that the percentage of senescent DP cells decreased significantly after senolytic treatment. (d) Quantification of Ki67⁺ cells (n = 4 biological replicates/group) showing that almost no proliferation was observed in the remaining non‐senescent DP cells after senolytic treatment. (e, f) Immunofluorescence and quantification of human DP culture after vehicle or senolytic treatment (n = 159 [vehicle] 109 [D + Q] cells in [e] and 148 [vehicle] 141 [D + Q] cells in [f] from 2 independent donors) showing remaining DP cells showed high p16, p21, and p27 expression in the nucleus, indicating that remaining DP cells at the end of senolytic treatment were in quiescence status in contrast to irreversible proliferation arrest in senescent DP cells. (g, h) Protein arrays and quantification of the conditioned medium of human DP culture after vehicle (dotted baseline) or senolytic treatment (columns) showing that the DP‐specific SASP factors were overall significantly reduced in the conditioned medium after senolytic treatment (two‐way ANOVA with Šídák's multiple comparisons test). (i) Western blot of human DP culture after vehicle or senolytic treatment at P3 showing that marked decrease of p21 with IL‐6 and p‐Stat3 in senolytic‐depleted P3 DP culture compared to control P3 DP culture as well as sustained expression of Gli1 and Lef1 in senolytic‐depleted P3 DP culture compared to control P3 DP culture. (j, k) Immunofluorescence showing that activation of IL‐6 signaling and TGF‐β signaling in non‐senescent DP cells was reversed after senolytic treatment, with recovered Gli1 and nuclear β‐catenin expression. (l, m) Immunofluorescence and quantification showing that the percentage of Lef1⁺ cells in cultured DP cells significantly increased in senolytic‐depleted P3 DP culture compared to control P3 DP culture (n = 5 biological replicates/group). Data are presented as the mean ± s.e.m. *p < 0.05, ***p < 0.001, ****p < 0.0001 (Welch's t test) in (c, d); (unpaired t test) in (e, f, m); (two‐way analysis of variance) in (h).

FIGURE 4

Senolytic depletion improves hair inductivity of human DP culture. (a) Patch assay and quantification (n = 4 [P2], 5 [P3], 4 [P4] biological replicates) showing that senolytic‐depleted DP cells showed their hair inductivity at early passage (P2), but not as significantly at later passage (P4), showing an inverse correlation between the recovery potential of hair inductivity and culture passage. (b) Immunofluorescence showing that Gli1⁺ and Lef1⁺ mesenchymal structures were observed in the proximal region of regenerated HFs in senolytic‐depleted DP group, whereas no DP structures or aggregates were detected in control DP group. (c) Experimental scheme and gross morphology of 3D human skin constructs by seeding 1 million human DP cells with and without senolytic treatment, together with 2 million human keratinocytes. (d) Whole mount immunofluorescence showing that versican⁺ DP aggregates were observed in the bottom part of microwells in senolytic‐depleted DP group, whereas no DP aggregates or structures were detected in control DP group. (e) Whole mount immunofluorescence showing that distinct HF epithelial layers including the K71⁺ inner root sheath layer and the K14⁺ outer root sheath layer were observed in senolytic‐depleted DP group, but not in control DP group. Data are presented as the mean ± s.e.m. *p < 0.05, **p < 0.01, (paired t test).