Lichen planopilaris and frontal fibrosing alopecia as model epithelial stem cell diseasesMatthew J Harries 1,2, Francisco Jimenez 3, Ander Izeta 4, Jonathan Hardman 2, Sreejith P Panicker 5, Enrique Poblet 6, Ralf Paus 2,7@

1 The Dermatology Centre, Salford Royal NHS Foundation Trust, Salford, Greater Manchester, UK

2 Centre for Dermatology Research, University of Manchester, MAHSC and NIHR Manchester Biomedical Research Centre, Manchester, UK

3 Mediteknia Dermatology Clinic and Universidad Fernando Pessoa Canarias, Las Palmas de Gran Canaria, Spain

4 Tissue Engineering Group, Instituto Biodonostia, Hospital Universitario Donostia, San Sebastián, Spain

5 Department of Zoology, University of Kerala, India

6 Department of Pathology, University General Hospital of Murcia, Murcia, Spain

7 Department of Dermatology, University of Miami, Miller School of Medicine, Miami, FL, USA

[@corresponding author]

Correspondence:

Professor Ralf Paus, MD, FRSB, Faculty of Biology, Medicine and Health, University of Manchester, Manchester, M13 9PT, UK

Tel. +44 (0) 161 2751665

E-mail. ralf.paus@manchester.ac.uk

ORCiD:

MH 0000-0002-0563-8690

FJ 0000-0003-1676-3056

AI 0000-0003-1879-7401

JH 0000-0002-1653-7908

SP No number

EP 0000-0001-6703-9020

RP 0000-0002-3492-9358

1

Funding sources for this work: Writing of this review was supported by the NIHR Manchester

Biomedical Research Centre, “Inflammatory Hair Diseases” programme.

Conflict of interests to declare: R.P. is founder and owner of a CRO (Monasterium Laboratory,

Münster/Germany) that also performs contract preclinical research on LPP/FFA. None of the

authors’ views expressed here are influenced by this.

Text word count: 4456

Number of references: 106

Number of Boxes: 3

Number of figure: 3

Key words: stem cells, hair follicle, PPARγ, EMT, immune privilege, IFN-

2

Abstract

Inflammation-associated, irreversible damage to epithelial stem cells (eSC) of the hair follicle in their immunologically privileged niche lies at the heart of scarring alopecia, which causes permanent and difficult-to-treat hair loss. We discuss why the two most common and closely related forms, lichen planopilaris (LPP) and frontal fibrosing alopecia (FFA), provide excellent model diseases for studying the biology, pathology and pharmacological manipulation of adult human eSCs in an easily accessible human mini-organ. Emphasizing the critical roles for interferon (IFN)-γ and peroxisome proliferator activated receptor (PPAR)-γ-mediated signalling in immune privilege (IP) collapse and epithelial-mesenchymal transition (EMT) of these eSCs respectively, we argue that these pathways deserve therapeutic targeting in the future management of LPP/FFA and other eSC diseases associated with IP collapse and EMT.

3

BOX: Highlights

LPP and FFA are inflammatory scalp conditions resulting in permanent alopecia. The central pathogenic processes in both LPP and FFA are eHFSC depletion and EMT on a

background of HF IP collapse predisposing to cytotoxic immune cell attack, eventually resulting in hair follicle destruction and replacement with fibrous tissue.

This central common pathogenesis “trunk” may explain the almost identical histological appearance of LPP and FFA, but doesn’t explain the distinctive phenotypic expression of each condition.

Unknown local, systemic, epigenetic and environmental factors have been proposed by many researchers as potential drivers in determining the ultimate phenotypic expression, with cosmetics exposure, trauma, neurogenic inflammation, hormonal and microbiome factors as potential candidates.

Knowledge of disease pathogenesis has already identified potential target for therapy in LPP/FFA, including the use of PPAR-γ agonists to prevent inflammation and EMT, and IP restorative treatments (e.g. FK506/tacrolimus) to protect eHFSC from immune attack.

4

Scalp hair follicles as a model to study the biology and pathology of human epithelial stem cells

Epithelial stem cells (eSCs) are increasingly being appreciated to be critically affected in several human diseases, such as lung disorders, inflammatory bowel disease and liver cirrhosis 1-3. While animal models have provided important insights, these models remain limited in their instructiveness and relevance for the comparable human pathology, and SCs are often difficult to access in the corresponding human organs. Instead human scalp hair follicles (HFs), which become frequently available during facelift and hair transplantation surgery and can be kept for several days in organ culture 4, have long been appreciated as optimally accessible human mini-organs for characterizing and interrogating the biology and pathology of human eSC and their progeny within their natural tissue habitat 5, 6.

Therefore, we argue that investigators interested in the biology, pathology and therapeutic targeting of human eSCs are well-advised to turn their attention to primary cicatricial alopecias (PCA), an important, but under-investigated group of rare inflammatory scalp disorders that result in progressive permanent hair loss; namely the two most common PCA variants in European populations 7, lichen planopilaris (LPP) and frontal fibrosing alopecia (FFA) 8 (Figure 1). PCAs are typically disfiguring and very difficult to treat, and are associated with distressing scalp symptoms and significant secondary morbidity, marked psychological impact, and significant loss of quality of life 8-15 (BOX – Clinician’s Corner). This unmet clinical need alone already justifies why these hair loss disorders deserve to attract more interest from the SC research community.

Here, however, we focus on portraying LPP/FFA as prototypic model stem cell diseases in which one can exemplarily study how human eSCs are physiologically protected from inflammation-induced damage, which eSC pathology and clinical consequences arise from a failure of these protective mechanisms and what strategies for therapeutic intervention promise to be most clinically useful 16. Simultaneously scrutinizing LPP/FFA permit new general insights into the biology and pathology of human eSCs in situ, besides enriching our, as yet, insufficient understanding of the biology and pathology of human HFs and their specific SC populations 5, 16, 17.

LPP and FFA – two branches of the same pathobiology tree?

Whilst LPP and FFA are clinically distinct entities, their histopathology is deceptively similar 11, 18, 19 (Figure 1; Box – Clinician’s Corner). This has encouraged the – still controversial - concept that FFA is a variant of LPP 11, 16, 20, 21. However, some investigators feel that FFA is a distinct entity from LPP and it remains conceptually unclear how a) both hair follicle-targeting dermatoses are related to each other and b) what mechanisms generate such different clinical phenotypes 22. We propose that LPP and FFA represent two phenotypically distinct branches of the same pathobiology “tree” (Figure 2), thus expanding upon the recent concept that both entities share clinical and histological characteristics of a “lichenoid folliculitis” 20.

While the prevalence of LPP is unknown, it is classified as an orphan disease (ORPHA: 525) and its numbers appear to have remained stable over recent years; whereas cases of FFA has substantially increased over the past two decades 7, 14, 21, 23, 24; despite it being only a relatively recently described entity 23. This raises the question to what extent environmental and/or epigenetic components

5

impact on FFA pathobiology and, intriguingly, suggest that FFA may also provide a model disease for studying pathogenic environmental influences on adult human eSC 25-27.

The central pathobiological process in both LPP and FFA is inflammation-induced permanent loss of eSC in the HF’s bulge region that are vital for HF cycling and regeneration 16, 28-30. In contrast to the often significant inflammation seen in the most common autoimmune hair disease alopecia areata, where massive inflammation can attack the proximal hair bulb (yet without destroying the HF) 31, even fairly discrete infiltrates around the bulge can suffice to damage the HF irreversibly. The mechanisms that normally protect these eSC from immune attack in a healthy HF, namely by the creation of an area of relative immune privilege (IP) in the bulge under physiological conditions, and how these eSC respond when these protective mechanisms fail 28, 32 (Figure 2 & 3), constitute key problems in general human stem cell biology. In fact, the “niche” that surrounds eSCs seems to dictate stem cell behaviour 33 and thus PCAs constitute a superb model to dissect the importance of preservation of individual constituents of the local microenvironment for stem cell function.

Pathobiology of LPP and FFA: Immune privilege collapse and epithelial-mesenchymal transition (EMT) in the bulge

In order to understand the pathobiology of LPP and FFA, a few key features of human HF physiology need to be called to mind.

The HF immune system

Immunohistologically, the human bulge displays high protein co-expression of keratin 15 (K15) and CD200 on eHFSC 5, 34, which are also negative for connexin 43 and quiescent in terms of cell cycle activity 6, 35. Healthy human HFs recruit several mechanisms to protect these eSC from potentially damaging immune responses 36, 37. Central to this is the down- regulation of MHC class Ia and II molecule protein expression not only in the proximal anagen HF epithelium but also in the bulge, thus restricting (auto-)antigen presentation 32. Further, the bulge shows reduced resident immune cell populations 38 and constitutes a special tissue niche characterized by the expression of potent, locally generated immunosuppressants, including α-melanocyte-stimulating hormone (MSH) and transforming growth factor- ß2 (TGFβ2), as well as prominent expression of the immunoinhibitory cell surface “no-danger” signal, CD200. Combined, these mechanisms restrict immune responses against the bulge and bestow a relative IP on this eSC niche, from which the melanocyte SCs that reside here also profit 28, 32.

General considerations on scarring hair loss (PCA)

In the skin of healthy mice, inflammatory attacks are regularly launched on the bulge region of isolated HFs, possibly as a mechanism by which damaged HF are selectively eliminated without causing widespread tissue damage (“programmed organ deletion”) 39. Therefore, one wonders whether PCAs represent a pathologically exaggerated and ill-controlled variant of an otherwise physiological phenomenon of selective HF elimination 16. What remains unclear are how such T lymphocyte- and macrophage-dominated inflammatory infiltrates, which are prominently accentuated in mice under conditions of perceived stress 40, 41, are attracted to the human bulge in the first place and what distinguishes HFs affected in this manner from their unaffected neighbours.

6

In mice, different regions of the HF epithelium secrete a distinct spectrum of chemokines, which guide the intracutaneous trafficking of defined immunocyte populations and selectively attract them to distinct locations within the HF epithelium 42, which may explain (in part) the major fluctuations in the number and location of intracutaneous immunocytes that are hair cycle dependent 36, 37. It is plausible that these principles also apply to the human HF. Therefore, changes in chemokine secretion from the bulge region in PCAs may well attract a lymphocytic infiltrate to the HF’s stem cell niche, similar to what has been shown in alopecia areata (AA), in which excessive CXCL10 secretion from the anagen HF bulb epithelium attracts a Th 1-type infiltrate 43, 44. Since the location of an inflammatory attack on the HF ultimately determines the clinical phenotype and outcomes 16, in PCA one needs to take a closer look at the bulge region.

The bulge region

In lesional LPP/FFA HFs the bulge undergoes major pathological changes such as reduced protein expression of the key bulge stem cell markers K15 and CD200 16, 28-30 (Figure 2 & 3). Moreover, microarray analysis of mRNA extracted from laser-capture microdissected bulge epithelium from lesional human LPP HFs shows loss in the expression of eHFSC signature genes, and K15+ cells in the bulge of lesional LPP and FFA HFs undergo increased apoptosis in situ 28. Indeed, mutant mice with targeted deletion of eHFSCs, using a K15 promoter-driven suicide gene approach, develop permanent HF loss, yet without substantial inflammation 45. Therefore, while the loss of eHFSC explain the disappearance of HFs and the resulting permanent alopecia seen in LPP/FFA, it does not explain the other characteristic clinical manifestations, such as fibrosis, epidermal atrophy or follicular plugging seen in active disease 16, 45, 46.

Notably, when healthy human scalp HFs are cultured ex-vivo with the pro-inflammatory cytokine interferon (IFN)-γ, expression of K15 and CD200 mRNA and protein are significantly reduced and major histocompatibility complex (MHC) class I is up-regulated 32, demonstrating experimentally-induced IP collapse and reduction of the eHFSC compartment 28, 47. CD200 may be particularly critical for the maintenance of bulge IP and down-regulation of CD200 may promote its collapse, since mutant mice whose K15+ cells fail to express CD200 develop a severe scarring alopecia phenotype as a result of a major lymphocytic HF inflammation that attacks the bulge 47. Therefore, IFN-γ-induced collapse of the physiological bulge IP likely lies at the heart of LPP/FFA pathobiology 28.

Common pathobiology features in both LPP and FFA – “The trunk”

Th-1 biased inflammation and IP collapse – lessons from alopecia areata

Distal HF inflammation is a key histological feature of both LPP and FFA, characterised by an activated Th1-biased CXCR3+ cytotoxic T-cell infiltrate along with increased numbers of INF-γ secreting (CD123+) plasmacytoid dendritic cells 28. Since IFN-inducible chemokines (CXCL9/10/11) are up-regulated in the bulge epithelium, this is one plausible mechanism for recruitment and propagation of the inflammatory response 28, 48, 49. Further, INF-γ driven increased expression of IP collapse indicators (MHC class I and class II; β2-microglobulin) on the transcript and protein level and down-regulation of locally generated key immunosuppressants (e.g. CD200 and TGFβ2) induce a collapse of the physiological HF IP, thus exposing usually hidden HF (auto-)antigens to immune surveillance (Figure 3) 28. However, it is unknown what initiates these inflammatory processes (e.g., HF stress; dysbiosis of HF microbiota?); whether IP collapse is the primary event preceding the

7

inflammatory response or a secondary phenomenon occurring after inflammation has become established.

Notably, this phenomenon shows striking similarities with the bulb IP collapse seen in alopecia areata (AA) 36. In AA, the innate immune system, namely natural killer cells and possibly also T cells, induce an INF-γ dependent collapse of HF bulb IP via the NKG2D receptor expressed on these (and CD8+) T cells, for which the proximal HF epithelium can express high levels of NKG2D ligands like MICA and ULBP3 under conditions of tissue stress 50, 51. Presumably, activation of their NKG2D receptors by MICA over expression on HF keratinocytes stimulates the secretion of INF-γ from the innate immunocytes, inducing IP collapse 36, 50. Moreover, perifollicular mast cells switch from their physiological immunoinhibitory phenotype to a pro-inflammatory one and show increased contact with CD8+ T cells 52. It is conceivable, but remains to be formally demonstrated, that a similar pathobiology scenario underlies the bulge IP collapse seen in LPP/FFA 28. However, despite these similarities in immunopathobiology, the clinical phenotype between LPP/FFA and AA is quite different (i.e. permanent vs. reversible alopecia) reflecting the location of the inflammatory attack on key HF structures (LPP/FFA = bulge eHFSC; AA = HF bulb).

Current therapies in LPP and FFA are directed at reducing inflammation by non-selective local or systemic immunosuppression 13. However, as in AA 36, the targeted reestablishment of bulge IP, with candidates such as FK506 (tacrolimus) or αMSH analogues, may successfully re-establish bulge IP. This concept can be probed preclinically in human HF organ culture 4 to screen for new candidate compounds already established in AA (e.g. Janus kinase (JAK) inhibitors 53).

Epithelial-mesenchymal transition (EMT)

In LPP/FFA, affected HF display progressive peri-follicular fibrosis with eventual replacement of the entire HF with fibrous tissue. One explanation for the fibrosis seen is epithelial-mesenchymal transition (EMT) 54, 55, a process by which epithelial cells lose polarity and cell-to-cell contact and acquire a mesenchymal phenotype, e.g. seen in wound healing, cancer development and various fibrotic diseases 56-58. Notably, the dermis of FFA lesions shows cells positive for the EMT marker Snail1 54, raising the possibility that EMT is somehow involved in the pathogenesis of FFA.

EMT does indeed occur within the bulge epithelium of lesional human LPP HFs, which shows increased gene and protein expression of mesenchymal markers (e.g. vimentin, fibronectin, CDH2) and down-regulation of the epithelial marker E-cadherin (CDH1) (Figure 3). Further, the transcription factors SNAI1 (Snail), SNAI2 (Slug) and TWIST1, which can be induced by TGFβ, epidermal growth factor (EGF) and Wnt/β catenin signalling, are also prominently expressed in lesional HF bulge epithelial samples 59. Moreover, co-localisation of K15 with vimentin in affected human LPP bulge tissue, and transmission electron microscopy confirming collagen fibre production by affected eSCs within the bulge, further support this claim 59.

K15+ cells in the bulge of healthy, human scalp HF ex vivo can be experimentally-induced to undergo molecular changes consistent with EMT by exposing them to a defined cocktail of four well-known EMT-promoting agents (i.e. INF-γ, TGF-β1, EGF and the E-cadherin-inhibiting peptide SWELYYPLRANL) 59. This shows that normal adult human eHFSC can easily switch-on an EMT program within their native stem cell niche when given appropriate signals, at least under conditions of tissue stress associated with HF culture 4. Taken together with the K15/vimentin co-expression evidence

8

mentioned above, this suggested that human eHFSC undergo EMT in LPP 59, further depleting the pool of eHFSCs that have survived apoptosis induction by cytotoxic CD8+ T cells and excessive IFN-γ signalling and are no longer protected by bulge IP 28. Thus, EMT of bulge eSCs may explain at least in part the prominent scarring and fibrosis of advanced LPP, and perhaps eSC EMT-derived progenitors may be more profibrotic than if the transition occurred in a non-stem cell area, where instructive local microenvironment signals would not be present.

Importantly, experimentally induced EMT in human eSCs ex vivo can be used to dissect the key mechanisms that underlie fibrosis development and to test pharmacological agents to prevent or reverse these effects. In fact, peroxisome proliferator activated receptor (PPAR)-γ agonists partially prevent induced EMT in this model 59. This is translationally relevant since PPARγ agonists can be clinically beneficial in some patients with LPP and FFA (see below) 60. Notably, PPAR-γ and its ligands are important in inhibiting fibrosis in other systems 61 with reduction in TGFβ1 signaling central to suppressing fibrogenesis. This model could also help to investigate other (non-dermatological) fibrotic conditions where eSC EMT is prominent.

Peroxisome proliferator activated receptor (PPAR) signaling

Karnik and colleagues62 pioneered the concept that the inflammatory pattern seen in active LPP may develop from an inhibition of fatty acid metabolism, cholesterol and peroxisome biogenesis; predisposing to pro-inflammatory lipid accumulation and immune cell infiltration. Pathway analysis identified PPAR-γ as the key regulator to these processes with significantly reduced PPAR-γ levels identified in both affected and unaffected LPP tissues compared to healthy controls 62. Further, deletion of PPAR-γ in K15+ eSC in mice (PPARγ-/-) results in an LPP-like phenotype with progressive hair loss, peri-follicular inflammation and scarring alopecia 62. This work has led to the therapeutic use of the PPAR-γ agonist, pioglitazone, with reported response rates of up to 50% 60. Moreover, PPAR-γ signaling down-modulates inflammatory responses 63, and up-regulates keratin 15 protein expression in the bulge of normal human scalp HFs ex vivo 63. Therefore, PPARγ stimulation may not only suppress and perhaps even partially reverse bulge EMT, but PPARγ-mediated signalling may exert functionally relevant eHFSC-protective and immunoinhibitory effects.

Less clear is why some groups of HFs develop a cicatricial alopecia phenotype whilst others with the same scalp region remain clinically unaffected. This cannot be explained by PPAR-γ-mediated effects alone, since there are no significant differences in PPAR-γ transcription between lesional and non-lesional HFs from the same LPP-affected individual 28. Thus, other intrafollicular parameters, such as the level of relative bulge IP, the secretary profile of each individual HF (e.g. for selected chemokines) and interfollicular differences in the HF microbiome may determine which HF develop LPP/FFA and which ones do not. Alternatively, one might invoke the existence of regional differences or spatiotemporal domains within the human scalp that might affect disease spreading via quorum sensing, similar to HF cycling and regeneration in mouse skin 64, 65.

Disrupted cholesterol biosynthesis

Gene expression profiling reveals that the cholesterol biosynthesis pathway is perturbed early in PCA development 66. Genes encoding the post-squalene cholesterol biosynthesis enzymes DHCR7 and EBP were significantly decreased in all PCA samples. Further, using the cholesterol biosynthesis inhibitor BM15766 (that targets DHCR7 and EBP), in both human cell culture and in vivo (mice),

9

accumulation of the cholesterol biosynthesis intermediates lathosterol and 7 - dehydrocholesterol (7-DHC) could be demonstrated. This accumulation of intermediate sterols resulted in the production of pro-inflammatory cytokines and induction of eHFSC apoptosis, two key processes in PCA development 49, 66.

Differences between LPP and FFA – “The branches”

We suggest that the pathobiology principles delineated above may represent a shared pathogenesis core or (to remain within botanical imagery) the common pathogenesis “trunk” of both LPP and FFA (Figure 2). Case reports documenting the typical LPP and FFA clinical features side-by-side in the same patient indeed underscore shared pathobiology elements 14, 15. However, the two distinct clinical manifestations “branch” out, guided by as yet unknown local, systemic and/or environmental regulatory factors that determine whether the LPP or the FFA phenotype predominates.

One of the most important open questions in this field of research, therefore, is what these regulatory factors might be: candidates are discussed below. Although this remains to be formally demonstrated, this discussion might also be of relevance for eSC pathology in other tissues, since dissection of some of the regulatory factors mentioned might also impact on human eSC niches elsewhere.

Environmental factors in FFA

Environmental factors may well drive the distinct FFA phenotype. Predominantly affecting those of higher socio-economic status 15, 67, 68 FFA has seen a rapidly rise in incidence (as opposed to the stably low incidence of LPP); this, along with the distinctive, symmetrical and predominantly frontal, band-like hair loss pattern and the clustering of FFA cases in post-menopausal women, have all encouraged research into potential environmental exposures in patients affected by FFA. For example, significantly increased regular use of dedicated sunscreens in people with FFA compared to age-matched controls is reported 69, 70, whereas this association has not been identified in LPP. Dissecting the environmental regulatory factors that impact on development of the FFA versus LPP phenotype could also serve as a human model system for exploration of the role of environmental stimuli in human eSC pathology, namely in promoting EMT and fibrosis.

Sex steroids

In contrast to LPP, sex steroids may also play a functionally important role in FFA: 1) although FFA very rarely occurs in men 15, 70, 71, post-menopausal women are by far the most frequently affected population 15; 2) co-existing androgenetic alopecia is frequently seen in FFA 14, 15; 3) there is a higher incidence of FFA in females with early menopause / hysterectomy (oophorectomy) 15, 22 and 4) FFA reportedly responds to anti-androgen therapy (i.e. 5α-reductase inhibitors), at least in some patients 13. Thus, androgen changes, age-related decline in dehydroepiandrosterone (DHEA) activity or oestrogen decline after menopause might predispose to FFA development 22, 72. However, normal or deficient blood sex hormones levels 21, 67, 73 and the involvement of non-androgen dependent HFs (e.g. posterior hairline, eyebrows) may argue against a key role of androgens in FFA pathobiology 67. Clarifying the role of sex steroids in FFA could also shed further light on the role of eSC in other, much less accessible tissues, e.g. in hepatic fibrosis.

10

Notably, DHEA (and DHEAS) are essential for nuclear PPAR-γ activation, fat metabolism and mitochondrial activity and its dysregulation can be associated with inflammatory and autoimmune changes 72, 74, 75. Further, DHEA levels are reduced in other fibrotic conditions and can prevent induced EMT/fibrosis in cultured cells 76; although, a DHEA-induced mouse model of ovarian fibrosis mediated by TGFβ signalling may suggests a more complex interplay 77. Finally, sex steroids and their receptors (e.g. androgen receptor (AR), oestrogen receptor (ER)-α) have long been implicated in the regulation of EMT in diverse extracutaneous human pathologies 78, 79.

The role of genetics?

A UK-based genome-wide association study in FFA is currently underway that may help to clarify the role of genetics in this disorder 80. Reports of familial cases 81-85 and data from case series suggest that 5 - 8% of reported FFA cases display a positive family history 15, 82, 83, with links to specific human leukocyte antigen (HLA) alleles being identified in some 86, 87 but not all cases 88. Arguments to explain the (albeit rare) family associations as well as the late age of FFA onset include autosomal dominant inheritance with incomplete penetrance or a germ line predisposition in these women 22. Epigenetic factors and microRNA signalling may also play a role 22, 89. Familial cases of LPP are also reported 85, 90. Again, once the role of genetic factors in the shared and distinct elements of eSC pathology that underlies FFA versus LPP has been clarified, this invites scrutiny of other human eSC pathologies for abnormalities in the expression/activity of the same genes and pathways.

Mitochondrial dysfunction and oxidative stress in FFA

Transition electron microscopy and global metabolomic profiling data have identified defects in mitochondrial β-oxidation of fatty acids leading to accumulation of medium and long chain fatty acids, along with decreased levels of (anti-oxidant) glutathione and elevated levels of oxidised glutathione (i.e. a marker of oxidative stress) in both lesional and non-lesional FFA scalp samples 91. Thus, accumulation of damaged and oxidized proteins may trigger inflammatory responses in FFA and raises the possibility of mitochondrial dysfunction as an early process in disease pathogenesis. We are currently investigating whether LPP shows similar indications of eSC mitochondrial dysfunction (as yet unknown). Given recent insights into the role of mitochondrial energy metabolism and reactive oxygen species production in human HF biology 92 this line of enquiry, which remains to be rigorously applied to eHFSCs, deserves to be followed up. That mouse eHFSC are exquisitely susceptible to mitochondrial damage 93 and that the transfer of mitochondria from mesenchymal SCs to lung or cornea epithelial cells can protect these from oxidative damage 94 further supports the importance of mitochondrial function for eHFSC health. Once this has been dissected in these accessible eSC populations, corresponding working hypothesis can then be pursued in a well-targeted manner in other human eSC pathologies.

Localized trauma

As members of the lichen planus family of dermatoses that characteristically show an appearance of skin lesions at sites of skin injury (Köbner phenomenon), it is not surprising that an association between skin injury and development of LPP has been reported 95. Moreover, cases of LPP following hair transplantation and FFA after face-lift surgery have been observed 96, 97. Localized skin trauma resulting in a non-specific inflammatory response triggering cytokine and chemokine production, up-

11

regulation of stress proteins and adhesion molecules promoting collapse of HF IP, or down-regulation of PPAR-γ expression, are all potential factors in PCA development. It is interesting to ask whether similar non-specific inflammatory cascades may promote eSC pathology in other human tissues, even if an initiating, localized physical tissue trauma cannot be evoked.

Neurogenic skin inflammation

Psycho-emotional stress has long been suspected as a triggering or aggravating factor for hair loss 9. Mouse studies clearly show that perceived stress can induce major perifollicular neurogenic inflammation causing increased mast cell degranulation and production of hair growth-inhibitory stress mediators such as corticotropin-releasing hormone and substance P that inhibit hair growth 98,

99. As substance P is a known growth factor for fibroblasts, and substance P+ nerve fibres are particularly dense around the murine and human bulge, these mechanisms might contribute not only to bulge inflammation, but also to fibrosis and the prominent bulge apoptosis seen in LPP/FFA. However, whether this actually happens in patients with LPP or FFA remains to be conclusively shown. Yet, given that extracutaneous eSC niches, e.g. in the lung and gastrointestinal tract, are also supplied with dense sensory innervation, it is conceivable that the similar neurogenic inflammation principles that impact on LPP and FFA could also apply to eSC pathology beyond the human HF.

Take home messages and future perspectives

If one views LPP and FFA as model eSC diseases, more systematic investigation of the “branching” factors that determine which clinical manifestation the underlying shared stem cell pathobiology “trunk” develops is not only of interest to PCA experts but also promises to reveal general principles of how local, systemic, epigenetic and/or environmental factors impact on the function and dysfunction of human eSC within their specific tissue niche.

Here we present LPP and FFA as disease entities that share the phenomenon of eHFSC depletion and pathological stem cell EMT on the basis of a compromised or collapsed bulge IP, which invites a cytotoxic inflammatory attack on the bulge. This common pathobiology core explains why it has been so difficult to distinguish LPP and FFA microscopically. However, what induces the above pathobiology scenario, how inflammatory cells are attracted to the bulge (e.g. chemokine secretion, excessive expression of NKG2D-activating “danger” signals like MICA?), and whether it is primarily damaged/“stressed” HFs (e.g. as a consequence of abnormalities in the HF microbiome) that are singled out for immune elimination in these diseases remains to be determined. Given that the HF microbiome massively impacts on the murine HF and skin immune system 101, 102, we need to understand whether and how a dysbiosis of the complex human HF microbiome 103 may predispose to eSC damage (Box – Outstanding questions).

Most importantly, we have delineated here how LPP and FFA can serve to generate important, clinically relevant insights, for example, into how inflammation-induced eSC damage occurs, which normal immune tolerance mechanisms are in place, why the IP of the eSC niche collapses and how it can be therapeutically restored, what drives EMT of these stem cells and how this can be halted or even reversed. This knowledge will not only help to develop more effective, pathobiology-tailored strategies for the management of PCA, but likely also for related human eSC diseases. In particular, current evidence suggests that it is particularly promising to pharmacologically target IFN-γ and

12

PPAR-γ-mediated signalling by inhibiting the former and stimulating the latter, thus preventing and potentially restoring both IP collapse and pathological EMT in the bulge (Box – Highlights).

One limitation of our proposal that the eSC pathology seen in LPP and FFA can teach important lessons on how human eSC in other tissue niches become dysfunctional, undergo EMT, and/or apoptosis, obviously is that comparatively little is known about the pathobiology of human eSC niches in extracutaneous tissues, precisely because they are so much more difficult to access and manipulate experimentally. However, given the many parallels that have already surfaced, for example, between gut, corneal limbal, and skin eSCs 104, 105, we believe that investigators specifically interested in the pathobiology of human eSC within their natural tissue habitat will find it most instructive to turn their attention to interrogating LPP and FFA as model eSC diseases.

13

BOX: Outstanding questions

What initiates the immune attack on the HF in LPP/FFA? Candidate mechanisms that we currently favour include excessive cytokine or chemokine secretion and/or (NKG2D-activating) “danger” signals like MICA expressed by the HF epithelium, which lead to an exaggerated physiological “immune elimination” of stressed or damaged HF via “programmed organ deletion” 16,39.

Which specific factors determine whether an individual develops either of the two scarring alopecia phenotypes (LPP versus FFA), and which of these regulatory principles also apply to the pathobiology of extracutaneous eSC disorders?

Can insights from other autoimmune hair diseases characterized by IP collapse of an otherwise immunologically protected epithelial tissue territory (i.e. alopecia areata) help identify novel pathogenic pathways and treatment strategies in LPP/FFA?

Do human eHFSCs, like their counterparts in murine epidermis, possess an “inflammatory memory” that renders them relatively sensitive to inflammatory damage and autoimmune disease 106, as the evolutionarily price paid for an advantageous augmented tissue repair capacity?

14

References :

1. Nadkarni, R.R., et al. (2017) Stem Cells in Pulmonary Disease and Regeneration. Chest, doi: 10.1016/j.chest.2017.1007.10022. Andersson-Rolf, A., et al. (2017) Stem Cells in Repair of Gastrointestinal Epithelia. Physiology (Bethesda) 32, 278-2893. Shin, S., and Kaestner, K.H. (2014) The origin, biology, and therapeutic potential of facultative adult hepatic progenitor cells. Curr Top Dev Biol 107, 269-2924. Langan, E.A., et al. (2015) Human hair follicle organ culture: theory, application and perspectives. Exp Dermatol 24, 903-9115. Purba, T.S., et al. (2014) Human epithelial hair follicle stem cells and their progeny: current state of knowledge, the widening gap in translational research and future challenges. Bioessays 36, 513-5256. Purba, T.S., et al. (2016) A primer for studying cell cycle dynamics of the human hair follicle. Exp Dermatol 25, 663-6687. Griffin, L.L., et al. (2012) Primary cicatricial alopecias: a U.K. survey. Br J Dermatol 167, 694-6978. Harries, M.J., et al. (2009) How not to get scar(r)ed: pointers to the correct diagnosis in patients with suspected primary cicatricial alopecia. Br J Dermatol 160, 482-5019. Chiang, Y.Z., et al. (2015) The role of beliefs: lessons from a pilot study on illness perception, psychological distress and quality of life in patients with primary cicatricial alopecia. Br J Dermatol 172, 130-13710. Jimenez, F., et al. (2016) Frontal Fibrosing Alopecia: A disease fascinating for the researcher, disappointing for the clinician and distressing for the patient. Exp Dermatol 25, 853-85411. Bolduc, C., et al. (2016) Primary cicatricial alopecia: Lymphocytic primary cicatricial alopecias, including chronic cutaneous lupus erythematosus, lichen planopilaris, frontal fibrosing alopecia, and Graham-Little syndrome. J Am Acad Dermatol 75, 1081-109912. Harries, M.J., et al. (2008) Management of primary cicatricial alopecias: options for treatment. Br J Dermatol 159, 1-2213. Racz, E., et al. (2013) Treatment of frontal fibrosing alopecia and lichen planopilaris: a systematic review. J Eur Acad Dermatol Venereol 27, 1461-147014. Meinhard, J., et al. (2014) Lichen planopilaris: Epidemiology and prevalence of subtypes - a retrospective analysis in 104 patients. J Dtsch Dermatol Ges 12, 229-235, 229-23615. Vano-Galvan, S., et al. (2014) Frontal fibrosing alopecia: a multicenter review of 355 patients. J Am Acad Dermatol 70, 670-67816. Harries, M.J., and Paus, R. (2010) The pathogenesis of primary cicatricial alopecias. Am J Pathol 177, 2152-216217. Oh, J.W., et al. (2016) A Guide to Studying Human Hair Follicle Cycling In Vivo. J Invest Dermatol 136, 34-4418. Poblet, E., et al. (2006) Frontal fibrosing alopecia versus lichen planopilaris: a clinicopathological study. Int J Dermatol 45, 375-38019. Mirmirani, P., et al. (2005) Primary cicatricial alopecia: histopathologic findings do not distinguish clinical variants. J Am Acad Dermatol 52, 637-64320. Turegano, M.M., and Sperling, L.C. (2017) Lichenoid folliculitis: A unifying concept. J Cutan Pathol 21. Kossard, S., et al. (1997) Postmenopausal frontal fibrosing alopecia: a frontal variant of lichen planopilaris. J Am Acad Dermatol 36, 59-6622. Tziotzios, C., et al. (2016) Frontal fibrosing alopecia: reflections and hypotheses on aetiology and pathogenesis. Exp Dermatol 25, 847-85223. Kossard, S. (1994) Postmenopausal frontal fibrosing alopecia. Scarring alopecia in a pattern distribution. Arch Dermatol 130, 770-774

15

24. Holmes, S. (2016) Frontal Fibrosing Alopecia. Skin Therapy Lett 21, 5-725. Cotsarelis, G., and Millar, S.E. (2001) Towards a molecular understanding of hair loss and its treatment. Trends Mol Med 7, 293-30126. Paus, R., and Cotsarelis, G. (1999) The biology of hair follicles. N Engl J Med 341, 491-49727. Schneider, M.R., et al. (2009) The hair follicle as a dynamic miniorgan. Curr Biol 19, R132-14228. Harries, M.J., et al. (2013) Lichen planopilaris is characterized by immune privilege collapse of the hair follicle's epithelial stem cell niche. J Pathol 231, 236-24729. Hoang, M.P., et al. (2009) Stem cell markers (cytokeratin 15, CD34 and nestin) in primary scarring and nonscarring alopecia. Br J Dermatol 160, 609-61530. Pozdnyakova, O., and Mahalingam, M. (2008) Involvement of the bulge region in primary scarring alopecia. J Cutan Pathol 35, 922-92531. Gilhar, A., et al. (2012) Alopecia areata. N Engl J Med 366, 1515-152532. Meyer, K.C., et al. (2008) Evidence that the bulge region is a site of relative immune privilege in human hair follicles. Br J Dermatol 159, 1077-108533. Gonzales, K.A.U., and Fuchs, E. (2017) Skin and Its Regenerative Powers: An Alliance between Stem Cells and Their Niche. Dev Cell 43, 387-40134. Ohyama, M. (2007) Hair follicle bulge: a fascinating reservoir of epithelial stem cells. J Dermatol Sci 46, 81-8935. Purba, T.S., et al. (2017) Divergent proliferation patterns of distinct human hair follicle epithelial progenitor niches in situ and their differential responsiveness to prostaglandin D2. Sci Rep 7, 1519736. Paus, R., et al. (2005) A 'hairy' privilege. Trends Immunol 26, 32-4037. Paus, R., et al. (1998) Generation and cyclic remodeling of the hair follicle immune system in mice. J Invest Dermatol 111, 7-1838. Christoph, T., et al. (2000) The human hair follicle immune system: cellular composition and immune privilege. Br J Dermatol 142, 862-87339. Eichmuller, S., et al. (1998) Clusters of perifollicular macrophages in normal murine skin: physiological degeneration of selected hair follicles by programmed organ deletion. J Histochem Cytochem 46, 361-37040. Arck, P.C., et al. (2005) Mast cell deficient and neurokinin-1 receptor knockout mice are protected from stress-induced hair growth inhibition. J Mol Med (Berl) 83, 386-39641. Arck, P.C., et al. (2001) Indications for a 'brain-hair follicle axis (BHA)': inhibition of keratinocyte proliferation and up-regulation of keratinocyte apoptosis in telogen hair follicles by stress and substance P. FASEB J 15, 2536-253842. Nagao, K., et al. (2012) Stress-induced production of chemokines by hair follicles regulates the trafficking of dendritic cells in skin. Nat Immunol 13, 744-75243. Ito, T., and Tokura, Y. (2014) The role of cytokines and chemokines in the T-cell-mediated autoimmune process in alopecia areata. Exp Dermatol 23, 787-79144. Ito, T., et al. (2013) CXCL10 produced from hair follicles induces Th1 and Tc1 cell infiltration in the acute phase of alopecia areata followed by sustained Tc1 accumulation in the chronic phase. J Dermatol Sci 69, 140-14745. Ito, M., et al. (2005) Stem cells in the hair follicle bulge contribute to wound repair but not to homeostasis of the epidermis. Nat Med 11, 1351-135446. Harries, M.J., and Paus, R. (2009) Scarring alopecia and the PPAR-gamma connection. J Invest Dermatol 129, 1066-107047. Rosenblum, M.D., et al. (2006) CD200, a "no danger" signal for hair follicles. J Dermatol Sci 41, 165-17448. Wenzel, J., and Tuting, T. (2008) An IFN-associated cytotoxic cellular immune response against viral, self-, or tumor antigens is a common pathogenetic feature in "interface dermatitis". J Invest Dermatol 128, 2392-2402

16

49. Panicker, S., et al. (2012) Disruption of the cholesterol biosynthesis pathway contributes to hair follicle cycling deficiency and stem cell apoptosis in a mouse model J Invest Dermatol 132, S75-84 (481)50. Ito, T., et al. (2008) Maintenance of hair follicle immune privilege is linked to prevention of NK cell attack. J Invest Dermatol 128, 1196-120651. Petukhova, L., et al. (2010) Genome-wide association study in alopecia areata implicates both innate and adaptive immunity. Nature 466, 113-11752. Bertolini, M., et al. (2014) Abnormal interactions between perifollicular mast cells and CD8+ T-cells may contribute to the pathogenesis of alopecia areata. PLoS One 9, e9426053. Kennedy Crispin, M., et al. (2016) Safety and efficacy of the JAK inhibitor tofacitinib citrate in patients with alopecia areata. JCI Insight 1, e8977654. Nakamura, M., and Tokura, Y. (2010) Expression of Snail1 in the fibrotic dermis of postmenopausal frontal fibrosing alopecia: possible involvement of an epithelial-mesenchymal transition and a review of the Japanese patients. Br J Dermatol 162, 1152-115455. Nakamura, M., and Tokura, Y. (2011) Epithelial-mesenchymal transition in the skin. J Dermatol Sci 61, 7-1356. Lamouille, S., et al. (2014) Molecular mechanisms of epithelial-mesenchymal transition. Nat Rev Mol Cell Biol 15, 178-19657. Skrypek, N., et al. (2016) Epithelial-to-Mesenchymal Transition: Epigenetic Reprogramming Driving Cellular Plasticity. Trends Genet 33, 943-95958. Singh, M., et al. (2017) EMT: Mechanisms and therapeutic implications. Pharmacol Ther 182, 80-9459. Imanishi, H., et al. (2017) Epithelial-mesenchymal stem cell transition in a human organ: lessons from lichen planopilaris. J Invest Dermatol, Accepted for publication60. Mesinkovska, N.A., et al. (2015) The use of oral pioglitazone in the treatment of lichen planopilaris. J Am Acad Dermatol 72, 355-35661. Burgess, H.A., et al. (2005) PPARgamma agonists inhibit TGF-beta induced pulmonary myofibroblast differentiation and collagen production: implications for therapy of lung fibrosis. Am J Physiol Lung Cell Mol Physiol 288, L1146-115362. Karnik, P., et al. (2009) Hair follicle stem cell-specific PPARgamma deletion causes scarring alopecia. J Invest Dermatol 129, 1243-125763. Ramot, Y., et al. (2014) The role of PPARgamma-mediated signalling in skin biology and pathology: new targets and opportunities for clinical dermatology. Exp Dermatol 24, 245-25164. Wang, Q., et al. (2017) A multi-scale model for hair follicles reveals heterogeneous domains driving rapid spatiotemporal hair growth patterning. Elife 665. Widelitz, R., and Chuong, C.M. (2016) Quorum sensing and other collective regenerative behavior in organ populations. Curr Opin Genet Dev 40, 138-14366. Panicker, S.P., et al. (2012) Sterol intermediates of cholesterol biosynthesis inhibit hair growth and trigger an innate immune response in cicatricial alopecia. PLoS One 7, e3844967. MacDonald, A., et al. (2012) Frontal fibrosing alopecia: a review of 60 cases. J Am Acad Dermatol 67, 955-96168. Holmes, S., and MacDonald, A. (2014) Frontal fibrosing alopecia. J Am Acad Dermatol 71, 593-59469. Aldoori, N., et al. (2016) Frontal fibrosing alopecia: possible association with leave-on facial skin care products and sunscreens; a questionnaire study. Br J Dermatol 175, 762-76770. Debroy-Kidambi, A., et al. (2017) Frontal fibrosing alopecia in men - an association with facial moisturisers and sunscreens. Br J Dermatol 177, 260-26171. Tolkachjov, S.N., et al. Frontal fibrosing alopecia among men: A clinicopathologic study of 7 cases. J Am Acad Dermatol 77, 683-690 e68272. Gaspar, N.K. (2016) DHEA and frontal fibrosing alopecia: molecular and physiopathological mechanisms. An Bras Dermatol 91, 776-780

17

73. Ranasinghe, G.C., et al. (2017) Prevalence of hormonal and endocrine dysfunction in patients with lichen planopilaris (LPP): A retrospective data analysis of 168 patients. J Am Acad Dermatol 76, 314-32074. Karbowska, J., and Kochan, Z. (2013) Effects of DHEA on metabolic and endocrine functions of adipose tissue. Horm Mol Biol Clin Investig 14, 65-7475. Hazeldine, J., et al. (2010) Dehydroepiandrosterone as a regulator of immune cell function. J Steroid Biochem Mol Biol 120, 127-13676. Mendoza-Milla, C., et al. (2013) Dehydroepiandrosterone has strong antifibrotic effects and is decreased in idiopathic pulmonary fibrosis. Eur Respir J 42, 1309-132177. Wang, D., et al. (2018) DHEA-induced ovarian hyperfibrosis is mediated by TGF-beta signaling pathway. J Ovarian Res 11, 678. Colditz, J., et al. (2016) Androgens induce a distinct response of epithelial-mesenchymal transition factors in human prostate cancer cells. Mol Cell Biochem 421, 139-14779. Shao, R., et al. (2014) Epithelial-to-mesenchymal transition and estrogen receptor alpha mediated epithelial dedifferentiation mark the development of benign prostatic hyperplasia. Prostate 74, 970-98280. Tziotzios, C., et al. (2015) Familial frontal fibrosing alopecia. J Am Acad Dermatol 73, e3781. Navarro-Belmonte, M.R., et al. (2015) Case series of familial frontal fibrosing alopecia and a review of the literature. J Cosmet Dermatol 14, 64-6982. Dlova, N., et al. (2013) Familial frontal fibrosing alopecia. Br J Dermatol 168, 220-22283. Junqueira Ribeiro Pereira, A.F., et al. (2010) Frontal fibrosing alopecia in two sisters. Br J Dermatol 162, 1154-115584. Miteva, M., et al. (2011) Frontal Fibrosing Alopecia occurring on scalp vitiligo: report of four cases. Br J Dermatol 165, 445-44785. Misiak-Galazka, M., et al. (2016) Lichen planopilaris in three generations: grandmother, mother, and daughter - a genetic link? Int J Dermatol 55, 913-91586. Chan, D.V., et al. (2015) HLA-DR1 in familial frontal fibrosing alopecia. J Am Acad Dermatol 73, e3987. Rivas, M.M., et al. (2015) Frontal fibrosing alopecia and lichen planopilaris in HLA-identical mother and daughter. Indian J Dermatol Venereol Leprol 81, 162-16588. Chan, D.V., et al. (2014) Absence of HLA-DR1 positivity in 2 familial cases of frontal fibrosing alopecia. J Am Acad Dermatol 71, e208-21089. Tziotzios, C., et al. (2017) Tissue and circulating microRNA co-expression analysis reveals potential involvement of miRNAs in the pathobiology of frontal fibrosing alopecia. J Invest Dermatol 137, 2440-244390. Pavlovsky, L., et al. (2015) Lichen planopilaris is associated with HLA DRB1*11 and DQB1*03 alleles. Acta Derm Venereol 95, 177-18091. Consolo, M., et al. (2012) Medium and long chain fatty acids elicit oxidative stress and stem cell death in human and animal models of cicatricial alopecia In J Invest Dermatol, S80 (474)92. Lemasters, J.J., et al. (2017) Compartmentation of Mitochondrial and Oxidative Metabolism in Growing Hair Follicles: A Ring of Fire. J Invest Dermatol 137, 1434-144493. Kloepper, J.E., et al. (2015) Mitochondrial function in murine skin epithelium is crucial for hair follicle morphogenesis and epithelial-mesenchymal interactions. J Invest Dermatol 135, 679-68994. Jiang, D., et al. (2016) Mitochondrial transfer of mesenchymal stem cells effectively protects corneal epithelial cells from mitochondrial damage. Cell Death Dis 7, e246795. Montpellier, R.A., and Donovan, J.C. (2014) Scalp trauma: a risk factor for lichen planopilaris? J Cutan Med Surg 18, 214-21696. Chiang, Y.Z., et al. (2012) Lichen planopilaris following hair transplantation and face-lift surgery. Br J Dermatol 166, 666-37097. Donovan, J. (2012) Lichen planopilaris after hair transplantation: report of 17 cases. Dermatol Surg 38, 1998-2004

18

98. Liu, N., et al. (2013) Chronic restraint stress inhibits hair growth via substance P mediated by reactive oxygen species in mice. PLoS One 8, e6157499. Peters, E.M., et al. (2006) Hair growth inhibition by psychoemotional stress: a mouse model for neural mechanisms in hair growth control. Exp Dermatol 15, 1-13100. Nilsson, J., et al. (1985) Stimulation of connective tissue cell growth by substance P and substance K. Nature 315, 61-63.101. Scharschmidt, T.C., et al. (2017) Commensal Microbes and Hair Follicle Morphogenesis Coordinately Drive Treg Migration into Neonatal Skin. Cell Host Microbe 21, 467-477 e465102. Scharschmidt, T.C., et al. (2015) A Wave of Regulatory T Cells into Neonatal Skin Mediates Tolerance to Commensal Microbes. Immunity 43, 1011-1021103. Ring, H.C., et al. (2017) The Follicular Skin Microbiome in Patients With Hidradenitis Suppurativa and Healthy Controls. JAMA Dermatol 153, 897-905104. Fuchs, E. (2016) Epithelial Skin Biology: Three Decades of Developmental Biology, a Hundred Questions Answered and a Thousand New Ones to Address. Curr Top Dev Biol 116, 357-374105. Kretzschmar, K., and Clevers, H. (2017) Wnt/beta-catenin signaling in adult mammalian epithelial stem cells. Dev Biol 428, 273-282106. Naik, S., et al. (2017) Inflammatory memory sensitizes skin epithelial stem cells to tissue damage. Nature 550, 475-480

19

BOX: Clinician’s Corner

The human hair follicle (HF) immune system (HIS) exhibits an array of characteristic properties that enable ongoing hair growth and cycling, whilst protecting the follicle from microbial pathogens.

LPP and FFA are characterised clinically by inflamed hair follicles surrounding an area of permanent scarring alopecia. Histologically, both conditions display a distal perifollicular lymphocytic infiltrate, squamatisation of the basal epithelium and progressive perifollicular fibrosis. Eventually, these affected HF’s are completely replaced with fibrous tissue.

In addition to distressing symptoms (e.g. itch and pain) that can accompany these disorders, it is important to recognise the significant psychological impact of hair loss. Associated autoimmune conditions (particularly thyroid dysfunction) are also seen more frequently.

Current treatments for LPP/FFA employ non-specific suppression of the immune system (e.g. with corticosteroids or immunosuppressants) to control symptoms and reduce inflammation to prevent further hair loss; hair regrowth is not possible. Hair transplantation is ineffective since most transplanted hair follicle grafts are destroyed by the inflammatory infiltrate.

Researchers might use our growing understanding of disease pathogenesis to identify novel targets for treatment in LPP/FFA. Potential targets for treatment include using PPAR-γ agonists to prevent EMT and IP restorative therapies, such as FK506/tacrolimus, to protect eHFSC from further immune assault.

Human HFs, both in health and disease (e.g. LPP/FFA), are easily accessible human tissue that allows researchers to better understand the normal physiological processes in a functional “mini-organ” whilst observing different pathological processes associated with inflammatory disease. Thus, these model diseases could provide insights into stem cell functioning, autoimmunity and fibrotic processes relevant to other systems away from the skin.

20

Figure 1: Clinical observations and histopathological features of lichen planopilaris (LPP) and frontal fibrosing alopecia (FFA).

(ai) LPP – an irregular patch of permanent (scarring) alopecia on the central scalp. (aii) When viewed with a trichoscope (x10 magnification) loss of follicular ostia, perifollicular scale and perifollicular erythema are evident at the margins of the patch. (aiii) LPP histology shows a dense perifollicular lymphocytic immune infiltrate (white arrow) and partial destruction of the outer root sheath (red arrow). (bi) FFA - Clinically FFA is characterised by a symmetrical band-like recession of the frontal and temporal hairline with loss of sideburns (white line shows the original hairline). Loss of eyebrows is a prominent and early feature in FFA; (bii) FFA – close up view of frontal hairline showing loss of follicular ostia and perifollicular inflammation. “Lonely hairs” (arrows) remain as the hairline recedes. (biii) Histologically (horizontal section), perifollicular fibrosis can be seen (grey arrows) with squamatisation of the basal layer suggesting EMT (green arrows). The immunopathology of both diseases shows a loss in the stem cell marker K15 (cii) when compared to healthy skin (ci), an up regulation of MHC class 1 suggesting IP collapse (di-ii) and a decline in the expression of the epithelial marker e-cadherin (ei-ii). APM= arrector pili muscle. [Fig. ci-ii and di-ii with permission from Harries et al. J Pathol 2013; Fig. ei-ii with permission from Imanishi et al. J Invest Dermatol 2017]

Figure 2: Lichen planopilaris (LPP) and frontal fibrosing alopecia (FFA) “Two branches of the same pathogenesis tree”.

Cartoon demonstrating the key shared properties between LPP and FFA, such as PPAR-γ deficiency and immune privilege collapse (IP), with other factors driving a divergence in phenotypes. LPP is predominantly CD8+ T cell driven, whereas FFA is potentially triggered by hormonal and epigenetic changes, with EMT being more prominent.

Figure 3: Lichen planopilaris (LPP) and frontal fibrosing alopecia (FFA) pathogenesis timeline and pathological features.

Summary diagram of key processes in LPP and FFA pathogenesis beginning with a healthy predisposed follicle, through to inflammation, cell death and epithelial-mesenchymal transition (EMT) before ending in fibrosis, scarring and permanent hair loss. [Red = common to both LPP and FFA; green = LPP only; blue = FFA only. White arrows denote inflammatory infiltrate, black arrows highlight apoptosis cells, grey arrows show early stage perifollicular fibrosis, thin black lines demarcate fibrosis tissue and thick black lines highlight a wedge shape loss in elastin expression].

21