Alopecia areata (AA) is the most common chronic-relapsing inflammatory hair disease [1]. It represents one of the most common autoimmune diseases [[2], [3], [4]], even though it can also be induced by innate, non-antigen-specific immune mechanisms [[5], [6], [7], [8]]. There is no curative therapy available for this psychologically often highly disruptive hair loss disorder [[9], [10], [11], [12]], and none of the current AA therapeutics, including glucocorticosteroids and JAK inhibitors (JAKi) [[13], [14], [15], [16]], reliably suppress disease recurrence [[17], [18], [19], [20]]. Namely, the almost preprogrammed relapse of AA seen after the discontinuation of successful, hair regrowth-inducing JAKi therapy [[21], [22], [23]] raises the pertinent question of whether these therapeutics fail to restore hair follicle (HF) immune privilege (IP) long-term. This is critical since AA does not develop without IP collapse, and HFs that fail to restore their IP remain deeply vulnerable to immune attack [2,24,25]. Since JAKi suppress signaling through multiple different receptors [13,14,16], one wonders whether JAKi also inhibit known HF-IP-protective pathways, such as IL-10R-mediated signaling by JAK1i, given that JAK1-dependent IL-10R-signaling is an important guardian of human HF-IP [18,25,26].

JAKi have been introduced into AA management that, among other receptors, also or preferentially inhibit signaling through the common γ–chain used by the interleukin-15 (IL-15) receptor (IL-15R) [16,27] by blocking JAK3 [28], not the least since, IL-15 is often portrayed as an important pathogenic cytokine in AA [[29], [30], [31], [32], [33], [34]]. Therefore, clarifying the role of IL-15R-mediated signaling in human HF-IP and AA pathogenesis has become both urgent and clinically important. The current study aims to initiate this much-needed line of translational research.

Based on robust preclinical research evidence, interferon-gamma (IFNγ) is now accepted as the key cytokine that drives AA pathogenesis [3,6,[35], [36], [37], [38], [39], [40], [41]] since it potently induces human HF-IP collapse [42], premature HF regression (catagen), and HF dystrophy [26,43], both in vivo and ex vivo – i.e., three hallmarks of AA [2,3]. Instead, the concept that IL-15 is also critically involved in AA pathobiology [25,[44], [45], [46], [47]] is predominantly based on the well-documented autoimmunity-promoting functions of this pleiotropic cytokine, mainly in non-immune privileged sites [[48], [49], [50]]. In fact, IL-15 can activate IFNγ-secreting NKG2D+ NK and CD8+ T cells, which are among the recognized cellular drivers of human AA pathogenesis [25,33,[51], [52], [53], [54]], while IL-15 controls NK cell survival and function [[55], [56], [57], [58]]. Also, in mice, AA-like disease can be induced by virtual memory T cells, whose pathogenic activity is mediated by NKG2D-dependent innate-like cytotoxicity and augmented by stimulation with IL-15 stimulation [59].This argues in favor of targeting IL-15/IL-15R signaling in AA management. Also, many AA patients show increased IL-15 serum levels [30,31], even though other cytokines (namely IL-2 and IL-4) that also signal through the common γ–chain exhibit higher serum levels in AA patients than IL-15 [29]. Moreover, in the C3H/HeJ mouse model of AA, whose relevance for human AA requires critical analysis [6], antibody-mediated blockade of the IL-15Rβ chain, which is shared with the IL-2R and also transduces signals received via IL-15Rα [48,60], has been reported to prevent the development of AA-like hair loss and reduce the accumulation of CD8+/NKG2D+ T cells in alopecic skin [33]. However, this effect may also have resulted from suppression of IL-2Rβ signaling since IL-2R activation of NKG2D+ immunocytes such as CD8+ T and NK cells is critically required for experimental induction of AA lesions in healthy human scalp skin in vivo [5,6].

Importantly, pro-inflammatory, HF autoimmunity-promoting effects of IL-15 are primarily studied in mice [33,59] or in organs/tissues that do not exhibit IP [48,[61], [62], [63], [64], [65]]. Instead, physiologically, anagen HFs keep their microenvironment under stringent immunoinhibitory control [3,25], even though the HF epithelium in mice chemotactically attracts T cells and dendritic cells, e.g., to regulate its epithelial stem cells and HF microbiome [[66], [67], [68]]. In this unique, IP-dominated tissue-level signaling milieu, IL-15 may well exert functions that differ from its activities elsewhere. In addition, IL-15-induced signaling effects may differ between immune cells and the HF epithelium itself. Also, the substantial differences between murine and human immunology on multiple levels caution against extrapolating from one system to the other. Thus, we do not really know yet whether the blockade of IL-15R γ-chain-mediated signaling by JAK3i [69] and/or of IL-15R β-mediated signaling by JAK1i [14] within human HFs is beneficial or counterproductive in the treatment of hair loss disorders, namely in AA [2,3]. Moreover, it remains entirely unknown which role the IL-15 private receptor, IL-15Rα [70], plays in HF-IP maintenance or collapse, i.e., the most critical event in human AA immunopathogenesis [25,71].

Skepticism that IL-15R-mediated signaling drives AA pathogenesis is bred by our recent demonstration that high concentrations (50–100 ng/ml) of rhIL-15 significantly promote human scalp HF growth, prolong the anagen phase, and reduce apoptosis in hair matrix (HM) keratinocytes ex vivo [72] – the exact opposite of what one sees in AA, since AA is associated with massive HM keratinocyte apoptosis and premature catagen induction [2,7,43]. Moreover, clinical trials testing an IL-15 superagonist (ALT-8303/N803) have not reported any hair loss phenomena [73], while IL-15 transgenic mice reportedly even show excessive hair growth in vivo [74]. In fact, IL-15 is one of the most potent multi-system apoptosis-suppressing cytokines identified so far [72,[75], [76], [77]], and an IL-15-IgG fusion protein awards murine HFs relative protection from chemotherapy-induced apoptosis [78]. In contrast, evidence emerging from clinical trials with IL-15/IL-15Rα-antibodies/antagonists have reported (unspecified) “hair loss” in a few patients [79].

Finally, IL-15 promotes autophagy in iNKT cells and, thus, their survival and maturation [80,81]. This could be dually important in the context of AA: we have shown that immunoinhibitory IL-10-secreting iNKT cells protect human scalp skin from experimentally induced AA development in vivo [82]. Thus, IL-15Rα-mediated signaling might exert an AA-protective physiological role in human scalp skin by maintaining and stimulating perifollicular iNKT cell activity. Furthermore, autophagic flux is required for anagen maintenance in human HFs [83], but is defective in AA [52,84], while stimulating autophagy provides relative protection from HF-IP collapse induced by IFNγ [84]. Taken together, this questions whether IL-15 is indeed a key pathogenic cytokine in human AA and whether blocking IL-15/IL-15R mediated signaling might exert undesired effects during AA management.

Therefore, we have asked in this study whether IL-15/IL-15R signaling a) shows expression abnormalities in AA patients, b) induces or even prevents HF-IP collapse in organ-cultured, healthy human scalp HFs in the presence/absence of IFNγ [26,85], and c) whether rhIL-15 treatment promotes hair re-growth in experimentally induced alopecia areata lesion in a humanized AA mouse model in vivo [5,6,8,82]. This was studied by quantitative immunohistomorphometry (qIHM) in healthy scalp skin and lesional/non-lesional AA biopsies, in organ-cultured healthy human scalp HFs in the presence and absence of IFNγ [7,26,85], or IL-15Rα siRNA, and in the humanized AA mouse model in vivo [6,8,82,86,87].