Restoration of Immune Privilege in Human Dermal Papillae Controlling Epithelial-Mesenchymal Interactions in Hair Formation

Abstract

Background: Hair follicles are among a handful of organs that exhibit immune privilege. Dysfunction of the hair follicle immune system underlies the development of inflammatory diseases, such as alopecia areata.

Methods: Quantitative reverse transcription PCR and immunostaining was used to confirm the expression of major histocompatibility complex class I in human dermal papilla cells. Through transcriptomic analyses of human keratinocyte stem cells, major histocompatibility complex class I was identified as differentially expressed genes. Organ culture and patch assay were performed to assess the ability of WNT3a conditioned media to rescue immune privilege. Lastly, CD8+ T cells were detected near the hair bulb in alopecia areata patients through immunohistochemistry.

Results: Inflammatory factors such as tumor necrosis factor alpha and interferon gamma were verified to induce the expression of major histocompatibility complex class I proteins in dermal papilla cells. Additionally, loss of immune privilege of hair follicles was rescued following treatment with conditioned media from outer root sheath cells. Transcriptomic analyses found 58 up-regulated genes and 183 down-regulated genes related in MHC class I+ cells. Using newborn hair patch assay, we demonstrated that WNT3a conditioned media with epidermal growth factor can restore hair growth. In alopecia areata patients, CD8+ T cells were increased during the transition from mid-anagen to late catagen.

Conclusion: Identification of mechanisms governing epithelial and mesenchymal interactions of the hair follicle facilitates an improved understanding of the regulation of hair follicle immune privilege.

Keywords: Epithelial-mesenchymal interaction; Hair follicle; Immune privilege; MHC molecule.

Fig. 1

Loss of immune privilege in cultured DP. A Immunofluorescence (IF) was performed on cryosection biopsies of HFs using anti-MHC class I antibody. White arrows point in DS, and epidermal keratinocyte individually (scale bar = 100 µm, ×20). B ICC was performed on cultured DP cells for 0 passage (DP P0) or 3 passages (DP P3) with human anti-MHC class I antibody and quantified in C. Black arrows indicate stained cells with anti-MHC class I antibody. D Levels of MHC class I mRNA in cultured DP cells were measured by RT-qPCR analysis at each passage. Values were normalized to cultured DP at DP P0. Data are presented as the mean ± SD from three individuals. *p < 0.05 and **p < 0.001 compared with the control by Student’s t-test

Fig. 2

TNF-α and IFN-γ induce MHC class I expression of cultured DP cells. A Cultured DP (P3) cells were incubated with control media or media containing IGF-2, TNF-α, or IFN-γ for 48 h and then immunostained for MHC class I protein. Black arrows indicate stained cells with anti-MHC class I antibody (scale bar = 100 um, x20). B Percent of MHC class I+ cells in each media. C  Their relative mRNA expression. Data are presented as the mean ± SD from three individuals after normalization to control. *p < 0.05 and **p < 0.001 compared with the control by Student’s t-test

Fig. 3

Restoration of immune privilege. A DP P3 cells without treatment or incubated with either ORS CM, WNT3a CM, TGF-β, or EGF for 48 h were stained for MHC class I. Black arrows denote cells that stained  for MHC class I expression (scale bar = 100 µm, ×20). B, C Percent of MHC class I+ cells in each media and their RT-qPCR analysis. Values were normalized to DP cultured without treatment. Data are presented as the mean ± SD from three individuals. *p < 0.05 and **p < 0.001 compared with the control by Student’s t-test

Fig. 4

The growing of hair follicle in organ culture and patch assay. A, B Biopsied HFs were cultured with PBS, IFN-γ, IFN-γ + WNT3a CM, and IFN-γ + WNT3a + EGF individually. The morphology and stage of HFs (anagen, early catagen, mid-catagen, or late catagen) were classified and counted from different media after 6 days. C The relative number of proliferating cells was measured using Ki67 staining. D Relative number of TUNEL positive cells were counted. E The relative mRNA expression of MHC class I was analyzed by RT-qPCR in same condition. F–I The neogenesis of HFs were tested by NENF patch assay with each condition. The gross picture of dorsal skins in nude mice were taken by microscopy in each condition and the number of hairs were measured. The white arrows indicate new hairs (×20). Data are presented as the mean ± SD from three individuals. *p < 0.05 and **p < 0.001 compared with the control by Student’s t-test

Fig. 5

Bioinformatics analysis of MHC class I related genes from GEO. A Visualization of MHC class I+ related genes expression using microarray data from GSE 11089. Color was assigned to each range based on the log₂ fold change value. Up-regulated genes belong in − log₁₀ p-value > 1.3, log₂ fold change > 1, while down-regulated genes belong in − log₁₀ p-value > 1.3, log₂ fold change >  − 1. B From comparison MHC class I- versus MHC class I+ in keratinocyte stem cell, the up-regulated genes were shown in red color and down-regulated genes in green color. C Bubble chart displayed the GO biological process related to up-regulated and down-regulated genes of MHC class I+ group. Total 20 of biological process were listed based on − log₁₀ p-value. Bubble size is the count of overlap genes from DEGs. D A total of 16 genes involved in inflammation and WNT signaling pathway were selected based on Gene-Set Enrichment Analysis (GSEA) information

Fig. 6

The interaction of perifollicular factors in AA. A H&E staining (top) and IHC (bottom) were performed with human anti-CD8 antibody in two different stages (left: mid anagen, right: late catagen) of biopsied HFs from AA patients. (Scale bar = 50 µm, ×40) A total nine individuals were tested (Fig. 6A and Fig. S1). Blue arrows indicate CD8+ T cells. B Schematic model for AA patient based on our results. Under normal conditions, immune cells such as CD8+ T cell, natural killer (NK) cells, and mast cells can be detected around HFs. Patients with a certain specific genetic background are predisposed to abnormalities in the micro-environment of the follicle. When various disruptions occur during anagen (e.g., infection, trauma, or stress), the clinical phenotype of alopecia areata results through above systemic mechanism (for more detailed explanation of this process, see the discussion section)