Introduction: The hair follicle bulge has recently been added to a growing list of human tissue compartments that exhibit a complex combination of immunosuppressive mechanisms, termed immune privilege (IP), which appear to restrict immune mediated injury in specific locations. As epithelial hair follicle stem cells (eHFSC) reside in the hair follicle bulge region it is conceivable that these IP mechanisms protect this vital compartment from immune-mediated damage, thereby ensuring the ongoing growth and cyclic regeneration of the hair follicle. Lichen planopilaris (and variants) are inflammatory hair disorders that result in hair follicle destruction and permanent alopecia. Growing evidence suggests that eHFSC destruction is a key factor in the permanent follicle loss seen in these conditions, and that IP collapse may predispose these cells to immune mediated injury. Aims: The overall aim of this project was to generate immunohistomorphometric, gene profiling, and limited functional evidence to probe the 'bulge immune privilege collapse' hypothesis in a carefully selected model disease for inflammation-induced epithelial stem cell death, lichen planopilaris (LPP). Methods: Adult patients with LPP (or variant frontal fibrosing alopecia) were recruited. Biopsies from lesional and non-lesional scalp skin were performed and either snap frozen in liquid nitrogen, fixed in formalin for paraffin embedding or transferred immediately for hair follicle organ culture. Both frozen and paraffin embedded tissue was processed for immunohistochemistry (IHC) analysis using various immune privilege, hair bulge (eHFSC) and immune cell markers. Cultured samples were supplemented with various chemicals know to influence hair follicle immune privilege with analysis performed using IHC. Further, additional paired lesional and non-lesional samples were sectioned horizontally for laser capture microdissection of bulge cells. Following extraction of RNA, reverse transcription and amplification of cDNA from these selected bulge cells; gene expression profiling was performed comparing lesional with non-lesional samples. Selected, differentially regulated genes were validated using IHC and quantitative real-time PCR. Results: Bulge IP collapse is present in active LPP, as suggested by increased expression of MHC class I, β2microglobulin and MHC class II, along with reduced expression of the locally produced immunosuppressant TGFβ2, at both the gene and protein level. Microarray pathway analysis supports these data with the antigen processing canonical pathway being prominently enriched. Cell mediated immune responses are prominent in active disease, suggested by significantly increased numbers of activated and cytotoxic T-cells infiltrating the bulge epithelium, along with greater numbers of mast cells and macrophages in the peri-follicular connective tissue sheath. Bulge cell eHFSC loss is suggested by loss of bulge cell markers (e.g. keratin 15) on IHC, microarray and qPCR, and supported by microarray analysis showing virtually global loss of recognised bulge eHFSC signatures. Hair follicle organ culture experiments confirm bulge IP collapse in a separate cohort of patients, and demonstrate that the pro-inflammatory cytokine, interferon-γ can further collapse IP in the bulge epithelium of cultured hair follicles. Conclusion: These data identify collapse of immune privilege in bulge cells in active LPP, and identify prominent cell-mediated immune responses and loss of eHFSC signatures in active disease. The pro-inflammatory cytokine, interferon-γ also appears to play a prominent role in IP collapse and contributes to immune cell trafficking into affected tissue. Future study is required to ascertain triggering factors of IP collapse and pursue other identified candidates from gene expression analysis.

Primary scalp alopecia is classically divided into cicatricial (scarring) and non-cicatricial (non-scarring). Challenging cases are assessed with a scalp biopsy. The use of both horizontal and vertical sections (HoVert sections) has dramatically improved the accuracy of histopathological diagnosis. In this work, we have used immunostaining to address diagnostic difficulties, which persist despite all currently available tools. We performed an immunostain panel (CD3, CD4, CD8 and CD20) in order to distinguish pattern hair loss from alopecia aerate in cases which do not have the usual peribulbar lymphocytic infiltrate and showed that CD3+ T-lymphocytes within the empty fibrous follicular tracts favor a diagnosis of alopecia areata. We performed CD123 in order to distinguish lichen planopilaris from alopecia lupus erythematosus in cases with only a superficial lymphocytic infiltrate and an uninvolved interfollicular epidermis and showed that clusters of CD123+ plasmacytoid dendritic cells favor a diagnosis of lupus erythematosus. We performed cytokeratin 15 in order to assess whether the loss of the follicular bulge stem cells has diagnostic value in cicatricial alopecia and demonstrated that the loss of cytokeratin 15+ bulge stem cells is identified in lichen planopilaris, frontal fibrosing alopecia, and lupus erythematous, so cytokeratin 15 has no diagnostic value. We have attempted to integrate the new concepts and our findings into the traditional classifications of alopecia and proposed a new diagnostic algorithm. In conclusion, immunostaining combined with HoVert grossing advances the accuracy of histopathological diagnosis of primary scalp alopecia. / L’alopécie primitive du cuir chevelu est habituellement classée en cicatricielle et non-cicatricielle. Dans les cas difficiles, la biopsie du cuir chevelu peut aider au diagnostic. L’utilisation de coupes, à la fois verticales et horizontales sur le même spécimen (technique HoVert), a radicalement amélioré le diagnostic histopathologique. Dans ce travail, nous avons utilisé l’immunohistochimie pour évaluer les difficultés diagnostiques qui persistent malgré tous les outils actuels. Nous avons utilisé les CD3, CD4, CD8 et CD20 pour différencier l’alopécie androgénique de la pelade dépourvue de l’infiltrat lymphocytaire péribulbaire habituel et nous avons démontré que la présence de lymphocytes CD3+ dans les travées folliculaires fibreuses est en faveur de la pelade. Nous avons utilisé le CD123 pour différencier le lichen plan pilaire du lupus érythémateux alopécie avec infiltrat lymphocytaire superficiel et sans atteinte de l’épiderme interfolliculaire et nous avons démontré que la présence d’amas de cellules dendritiques plasmacytoïdes CD123+ est en faveur du lupus érythémateux. Nous avons utilisé la cytokératine 15 pour évaluer si la perte des cellules souches du bulge a une valeur diagnostique dans l’alopécie cicatricielle et nous avons démontré que cette perte s’observait de manière identique dans le lichen plan pilaire, l’alopécie frontale fibrosante comme dans le lupus érythémateux et n’avait donc aucune valeur diagnostique. Nous avons tenté d’intégrer les nouveaux concepts et nos données dans les classifications traditionnelles des alopécies et nous avons élaboré un nouvel algorithme diagnostique. L’association des immunomarquages avec la technique HoVert ouvre de nouvelles perspectives dans le diagnostic histopathologique des alopécies primaires du cuir chevelu. / Doctorat en Sciences médicales (Médecine) / info:eu-repo/semantics/nonPublished