Retinoic acid drives hair follicle stem cell activation via Wnt/β-catenin signalling in androgenetic alopecia

Abstract

Background: Depletion or permanent quiescence of the hair follicle stem cell (HFSC) pool underlies pathogenesis in androgenetic alopecia (AGA). Reactivation of quiescent HFSCs is considered an efficient treatment strategy for hair loss. The retinoic acid (RA) is critical to ensure stem cell homeostasis and function. However, little is known about whether RA regulates HFSC homeostasis. We aimed to investigate the impact of RA on HFSC homeostasis and the underlying mechanisms, in order to provide new potential targets for medical therapies of AGA.

Methods: Microdissected hair follicles from the occipital and frontal scalp in AGA were obtained for RNA sequencing analysis and test. The C57BL/6 mice model in telogen was established to investigate the effect of exogenous RA. Miniaturized hair follicles from frontal scalp were incubated with or without RA in hair follicle organ culture to test the effects on hair shaft elongation, hair cycling and HFSC activities. A strategy to characterize the effect of RA on HFSC in primary culture was developed to identify novel mechanisms that control HFSC activation. A clinical study was performed to test the efficacy of RA treatment in AGA patients.

Results: RA signalling was inhibited in the course of AGA pathogenesis along with HFSC dysfunction. Hair regeneration was retarded in AGA miniaturized hair follicles with RA deficiency, but they tended to recover after treatment with RA. In addition, RA treatment during the telogen phase facilitated HFSC anagen entry and accelerated hair growth. Mechanistically, RA promoted hair growth by stimulating stem cells via Wnt/β-catenin signalling and accelerating the transition from a dormant to an activated state. Furthermore, a clinical study suggested that RA has obvious advantages in the early intervention of AGA by reactivating HFSCs.

Conclusions: Our study provides insights into the reactivation of HFSCs in AGA and provides potential targets for medical therapies.

FIGURE 1

Inhibition of RA signalling and dysfunction of HFSCs occurred in the course of AGA pathogenesis. (a) The volcano map showed the differentially expressed genes between AMB and ANB, and those related to retinoic acid signalling are marked. (b) The KEGG pathway enrichment analysis included the retinoic acid metabolism pathway. (c) qPCR showed the gene expression levels of ALDH1A1, ALDH1A2 and ALDH1A3 in AMB and ANB. (d) The mRNA expression of RARα, RARβ and RARγ in AMB and ANB was analysed by qPCR. (e, f) The expression of CD200, RARα and CK15 in AMB and ANB were shown by immunofluorescence. Scale bar: 50 μm. (g, h) Cell cycle in AMB and ANB was analysed by flow cytometric. (i) The expression of Ki67 in AMB and ANB is shown by immunofluorescence. Scale bar: 50 μm. Data are means ± SD, Student's t‐test, n = 3 independent experiments, *p < 0.05, **p < 0.01, RA, retinoic acid; HFSCs, hair follicle stem cells; AGA, androgenetic alopecia; AMB, AGA, miniaturized hair follicles, bulge‐containing part; ANB, AGA, normal hair follicles, bulge‐containing part; qPCR, quantitative reverse transcription polymerase chain reaction; NS, not significant.

FIGURE 2

RA induced the telogen to anagen transition of hair follicles. (a) Time‐scale for the hair cycle in C57BL/6 mice after birth and the day to harvest skin are marked with red arrows. (b) Morphology of hair follicles in different hair cycles was examined using H&E staining. Scale bars: 50 μm. (c) qPCR analysis of ALDH1A1, ALDH1A2 and ALDH1A3 mRNA expression in different hair cycles. (d) The mRNA expression of RARα, RARβ and RARγ in different hair cycles was analysed by qPCR. (e, f) Western blotting analysis of RARα expression in different hair cycles. (g) Immunofluorescence showed the expression of CK15 and RARα in the telogen, early anagen and middle anagen phases. Scale bars: 50 μm. (h) Gross observation of the trunk skin of C57BL/6 mice at different times treated with different doses of RA, minoxidil and DMSO. (i) Hair follicle cycle scores in different treatment groups based on the relative pigmentation of murine skin. (j) H&E staining showed the morphology of hair follicles after 6 days and immunofluorescence showed the proliferation of different treatment groups. Scale bars: 50 μm. Data are means ± SD, Student's t‐test, n = 4 independent experiments, **p < 0.01; ***p < 0.001; NS, not significant; qPCR, quantitative reverse transcription polymerase chain reaction; RA, retinoic acid.

FIGURE 3

RA promoted AGA hair follicle growth in organ culture. (a) Morphology of hair follicles at baseline and after 6 days of treatment with different doses of RA, minoxidil and DMSO. (b) Hair shaft elongation after treatment with different doses of RA, minoxidil and DMSO. (c, d) The morphology and hair cycle of hair follicles after 6 days of treatment in 10⁻¹² M RA, minoxidil and DMSO were analysed by H&E staining. Scale bars: 50 μm. (e, f) The proliferation in the hair bulb and bulge after treatment with 10⁻¹² M RA, minoxidil and DMSO was shown by immunofluorescence. Scale bars: 50 μm. (g) Immunofluorescence showed the expression of CK15 and CD200 in the bulge zone after 3 days of treatment with 10⁻¹² M RA, minoxidil and DMSO. Data are means ± SD, Student's t‐test, n = 6 independent experiments. Scale bars: 50 μm. *compared with the DMSO group, *p < 0.05, **p < 0.01, ***p < 0.001. #minoxidil group versus 10⁻¹² M RA, ^# p < 0.05, ^## p < 0.01, ^### p < 0.001; AGA, androgenetic alopecia; NS, not significant; RA, retinoic acid.

FIGURE 4

RA signalling is required for RA‐augmented HFSC activation. (a, b) An EdU labeling assay was performed to test the proliferation of HFSCs treated with RA or BMS493. Scale bars: 50 μm. (c) The expression of the hair follicle stem cell marker CK15 and hair follicle progenitor cell marker CD200 in HFSCs treated with DMSO, 10⁻¹⁴ M RA or BMS493 was analysed by immunofluorescence. Scale bars: 50 μm. (d, e) The ratio of CD200‐positive HFSCs was analysed by flow cytometry. (f) The gene expression of the stemness markers NFATc1, NANOG and OCT4 and the gene expression of the differentiation markers GATA3 and Lef‐1 in HFSCs treated with or without RA were analysed by qPCR. Data are means ± SD, Student's t‐test, n = 5 independent experiments. *compared with the DMSO group, *p < 0.05, **p < 0.01, ***p < 0.001. RA, retinoic acid; HFSC, hair follicle stem cell; EdU, 5‐ethynyl‐20‐deoxyuridine; NS, not significant; qPCR, quantitative reverse transcription polymerase chain reaction.

FIGURE 5

RA activated HFSCs through Wnt/β‐catenin signalling. (a) Western blotting analysis showed the protein expression of the upstream and the downstream molecules of Wnt signalling in HFSCs stimulated by RA. (b) Western blotting results showed that DKK1 significantly downregulated the protein expression of β‐catenin and Lef‐1 induced by RA. (c, d) The proportion of EdU‐positive cells induced by RA significantly decreased upon treatment with the Wnt/β‐catenin signalling inhibitor DDK1. Scale bars: 50 μm. (e) Immunofluorescence showed that DKK1 attenuated the expression of CD200 in HFSCs induced by RA. Scale bars: 50 μm. (f) The gene expression of the differentiation markers GATA3 and Lef‐1 in HFSCs induced by RA decreased when adding DKK1, and the gene expression of the stemness markers NFATc1, NANOG and OCT4 showed an opposite trend. Data are means ± SD, Student's t‐test, n = 5 independent experiments. *compared with the RA group, *p < 0.05, **p < 0.01, ***p < 0.001. # DKK1 group versus DMSO group, ^# p < 0.05, ^## p < 0.01, RA, retinoic acid; HFSCs, hair follicle stem cells; EdU, 5‐ethynyl‐20‐deoxyuridine; NS, not significant.

FIGURE 6

RA enhanced hair regrowth in AGA patients. (a) Global photography of representative AGA patients at baseline and 1 month. (b) Medical tattoo was used to mark the target area, and phototrichogram of tretinoin and minoxidil groups at baseline and 1 month. Red arrows represented hair regeneration, and green arrows represented bigger diameter in hair. (c) Hair count, hair diameter and terminal hair ratio in the tretinoin and minoxidil groups at baseline and 1, 3 and 6 months. RA, retinoic acid; AGA, androgenetic alopecia.

FIGURE 7

Retinoic acid regulated Wnt/β‐catenin signalling to activate hair follicle stem cells.