iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata

Journal of Autoimmunity xxx (2018) 1e12

Contents lists available at ScienceDirect

Journal of Autoimmunity journal homepage: www.elsevier.com/locate/jautimm

iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata Amal Ghraieb a, Aviad Keren a, Alex Ginzburg a, Yehuda Ullmann a, Adam G. Schrum b, Ralf Paus c, d, Amos Gilhar a, * a

Skin Research Laboratory, Rappaport Faculty of Medicine, Technion e Institute of Technology, Haifa, Israel Departments of Molecular Microbiology & Immunology, Surgery, and Bioengineering, Schools of Medicine and Engineering, University of Missouri, Columbia, MO, USA c Dermatology Research Centre, University of Manchester, MAHSC and NIHR Manchester Biomedical Research Centre, Manchester, UK d Department of Dermatology and Cutaneous Surgery, University of Miami Miller School of Medicine, Miami, FL, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 8 August 2017 Received in revised form 31 March 2018 Accepted 5 April 2018 Available online xxx

Alopecia areata (AA) is understood to be a CD8þ/NKG2Dþ T cell-dependent autoimmune disease. Here, we demonstrate that human AA pathogenesis of is also affected by iNKT10 cells, an unconventional T cell subtype whose number is significantly increased in AA compared to healthy human skin. AA lesions can be rapidly induced in healthy human scalp skin xenotransplants on Beige-SCID mice by intradermal injections of autologous healthy-donor PBMCs pre-activated with IL-2. We show that in this in vivo model, the development of AA lesions is prevented by recognized the iNKT cell activator, a-galactosylceramide (a-GalCer), which stimulates iNKT cells to expand and produce IL-10. Moreover, in preestablished humanized mouse AA lesions, hair regrowth is promoted by a-GalCer treatment through a process requiring both effector-memory iNKT cells, which can interact directly with CD8þ/NKG2Dþ T cells, and IL-10. This provides the first in vivo evidence in a humanized model of autoimmune disease that iNKT10 cells are key disease-protective lymphocytes. Since these regulatory NKT cells can both prevent the development of AA lesions and promote hair re-growth in established AA lesions, targeting iNKT10 cells may have preventive and therapeutic potential also in other autoimmune disorders related to AA. © 2018 Elsevier Ltd. All rights reserved.

Keywords: Alopecia areata Animal model iNKT10 IL-10 a-GalCer

1. Introduction Alopecia areata (AA) is a common, T cell-dependent, organspecific autoimmune disease that predominantly attacks growing hair follicles (HF) that have lost their relative immune privilege [1e4]. Lesions can contain many types of lymphocytes, including CD4þ and CD8þ T cells, invariant (i)NKT cells and NK cells [5,6]. While a central AA-promoting role of NKG2Dþ CD8þ T cells is now well-appreciated [3,7e12], other cells such as NK cells, unconventional T cells and mast cells likely contribute to the pathobiology of AA [5,6,9,12,13]. Since iNKT cells have become a recent focus of interest in autoimmune disease research [14e19], we used the best currently available humanized mouse model for AA [20e23] to assess

* Corresponding author. E-mail address: [email protected] (A. Gilhar).

whether iNKT cells may also play a role in human AA. Animal models are routinely used for the study of human autoimmune diseases, including psoriasis, atopic dermatitis, and multiple sclerosis [24,25]. Unfortunately, differences between human and animal immune systems often severely limit the ability of animal models to mimic the human autoimmune condition (24). However, some humanized mouse models (including those for psoriasis and AA) that employ human skin xenotransplants, which are intracutaneously injected with defined, autologous human immunocytes present characteristic lesions of these two skin diseases that very closely mimic the human disease phenotype [20,21,23,26]. For example, we have previously demonstrated that the humanized mouse model of AA leads to rapid and predictable development of focal hair loss, which demonstrates all the clinical, histological, and immunohistochemical features used for the clinical diagnosis of AA [21,22]. Therefore, this humanized mouse model ideally suited to delineate intralesional immune responses and interactions of human tissue (skin) with defined human

https://doi.org/10.1016/j.jaut.2018.04.001 0896-8411/© 2018 Elsevier Ltd. All rights reserved.

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

2

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

immunocyte populations in vivo in an autologous setting. Furthermore, they can be used to dissect common immunopathology mechanisms shared between mammalian species, to test novel therapeutic strategies, and to understand mechanisms of drug action within the human target tissue in question [20e23,26,27]. We previously demonstrated that the AA humanized mouse model leads to rapid and predictable development of focal hair loss lesions that display all the characteristic clinical, histological and immunohistochemical features used for the clinical diagnosis of AA [21]. Recently, impressive regulatory effects of iNKT10 were described by Sag et al. [28]. These authors demonstrated the expansion of iNKT10 by repeated stimulation with the potent glycolipid antigen, alpha-galactosylceramide (a-GalCer). Therefore, we were particularly interested in probing the effects of a-GalCer on experimentally induced human AA in vivo, using the humanized mouse model of AA [21,22,27]. Our data generated in this model strongly support the hypothesis that iNKT10 cells with regulatory function play an important, previously unappreciated role in the pathobiology of human AA, and that targeting their function pharmacologically may represent a promising novel management strategy for both preventing and treating human AA. To the best of our knowledge, these translationally highly relevant preclinical data also provide the first evidence available in the literature that iNKT10 cells can exert an immunoregulatory effect in a human disease model [29]. Since autoimmunity in AA has tended to serve as a model for other T celldependent autoimmune diseases [1,5,8,30], our study also encourages a more systematic dissection of the potential role for iNKT10 cells and their therapeutic targeting in other autoimmune diseases. 2. Materials and Methods 2.1. Animals In this study, 90 C.B-17/IcrHsd-scid-bg (Beige-SCID) mice (Harlan Laboratories Ltd., Jerusalem, Israel) were used at 2e3 months of age. NOD-scid IL2rgnull (NSG) mice (Jackson laboratories) were also used. Both mouse strains demonstrated good integrity following autologous human skin and PBMC engraftment. The mice were housed in the pathogen-free animal facility of the Rappaport Faculty of Medicine, Technion e Israel Institute of Technology. Animal care and research protocols were in accordance with institutional guidelines and were approved by the Institutional Committee on Animal Use. 2.2. Donors Thirteen healthy donors aged 43 ± 10 years were included in this study. Full thickness healthy scalp skin (diameter of 3 mm) was transplanted onto Beige-SCID or NSG mice. Twenty ml of venous blood was collected from the same donors. For immunohistochemical analysis, biopsies were obtained from lesional areas of six patients with alopecia areata (age 37 ± 8). The study was approved by our institute's Institutional Ethics Committee. 2.3. Transplantation procedure Full thickness biopsies were taken from healthy donors undergoing plastic surgery in the scalp. Biopsies from each donor were dissected horizontally to generate pieces with a diameter of 3 mm. Three 3 mm pieces were grafted orthotopically into the subcutaneous layer of each mouse. The transplantation of the skin biopsies was performed as previously described [20e23]. Seven days

following surgery, mice were treated with Minoxi-5 (hair regrowth treatment for men containing 5% Minoxidil active ingredient) by spreading it on the grafts twice a day until we received optimal expedited hair growth (period of two months). 2.4. Culture of peripheral blood mononuclear cells Peripheral blood mononuclear cells (PBMC) were isolated from healthy donors by centrifugation of Ficoll/Hypaque (Pharmacia, Amershem Pharmacia Biotech, Uppsala, Sweden). The PBMCs were cultured for 14 days with 100 U/ml human IL-2 (PROSPEC, protein specialists) alone or in combination with a-GalCer 100 ng/ml (Abcam) in filtered medium composed of RPMI 1640, 10% human AB serum (Sigma-Aldrich Co. LLC), 1% L-glutamine and 1% Penicillin-Streptomycin antibiotics (media components; biological industries, Kibbutz Beit Ha’Emek, Israel). The medium was changed as needed. After culture, PBMCs were injected intradermally in the skin grafts. In our previous work, we used a high concentration (7  106) of injected cells [20e23]. In this work, we investigated the use of lower cell concentration (3  106). The latter demonstrated the same phenotypic characterization as was observed in the previously published grafts where 7  106 cells had been injected. Photo-documentation of the grafts was made. The grafts were harvested after 75 days, and then the skin tissues were fixed and embedded in paraffin. The mice were humanely sacrificed. 2.5. Research groups Mice were divided randomly into the following 11 treatment groups: Group 1 (18 mice) e PBMCs were incubated with a high dose of IL-2 (100 U/ml). After 14 days of incubation, cells were injected into the grafts. Group 2 (8 mice) e PBMCs were activated nonspecifically with phytohemagglutinin (PHA) (10 mg/1 ml). After 14 days of incubation, cells were injected into the grafts. Group 3 (8 mice) e PBMCs were incubated with a high dose of IL-2 (100 U/ml) combined with a-GalCer (100 ng/ml) for 14 days. Cells were injected into the grafts. Group 4 (8 mice) e PBMCs were incubated with a high dose of IL-2 (100 U/ml). After 14 days of incubation, cells were injected into the grafts. Subsequent to PBMC injection, intragraft injection of aGalCer (2 mg, 2Xweek) was performed. Group 5 (6 mice) e PBMCs were incubated with a high dose of IL-2 (100 U/ml) in combination with a-GalCer (100 ng/ml) for 14 days. Cells were then bound with anti-iNKT cell (6B11) antibody conjugated to ferromagnetic microbeads (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany) and directed through a cell separation column containing a magnetic field (Miltenyi Biotec). Depleted cells were collected for FACS analysis and were injected into the implanted grafts. Group 6 (6 mice) e PBMCs were incubated with a high dose of IL-2 (100 U/ml) in combination with a-GalCer (100 ng/ml) for 11 days. Then, cells were incubated with 6B11 anti-TCR Va24 neutralization antibody (IgG1, Beckman Coulter Inc., 10 mg/ml) for an additional three days. Cells were subsequently injected into the grafts. Group 7 (6 mice) e (Comparator for Group 6) PBMCs were incubated with a high dose of IL-2 (100 U/ml) in combination with a-GalCer (100 ng/ml) for 11 days. Then, instead of 6B11 anti-TCR Va24 neutralization antibody, cells were incubated with isotype control (nonspecific IgG1, Beckman Coulter Inc., 10 mg/ml( for an additional three days. Cells were subsequently injected into the grafts. Group 8 (6 mice) e PBMCs were incubated with a high dose of

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

IL-2 (100 U/ml) in combination with a-GalCer (100 ng/ml) for 14 days. Cells were subsequently injected into grafts. Beginning directly after PBMC injection, neutralization anti-human IL-10 monoclonal antibody (0.04 mg) was injected daily. Group 9 (6 mice) e (Comparator for Group 8) PBMCs were incubated with a high dose of IL-2 (100 U/ml) in combination with a-GalCer (100 ng/ml) for 14 days. Cells were subsequently injected into grafts. Beginning directly after PBMC injection, instead of anti-

3

human IL-10 monoclonal antibody, isotype control for IL-10 antibody (IgG2B) (0.04 mg) was injected daily. Group 10 (6 mice) e PBMCs were incubated with a high dose of IL-2 (100 U/ml). After AA-like appearance occurred, IL-10 recombinant protein solution (200 ng) was injected twice a week. Unspecific protein BSA (200 ng) was injected as negative control. Group 11 (3 mice) e PBMCs were incubated with a high dose of IL-2 (100 U/ml). After AA-like appearance occurred, a-GalCer

Fig. 1. AA lesioning assessed by vertical section histopathology of human scalp skin xenografts. (A) A graft injected with IL-2-activated PBMCs shows lymphocytic infiltration. (B) Normal terminal follicles are present in a control graft (no PBMCs injected). (C) Normal HF is present in a graft injected with a-GalCer þ IL-2-activated PBMCs. (D) A scalp graft treated with the same culture condition as (C), but depleted of NKT cells, shows hair shaft with non-pigmented bulb and an irregular shaggy border with peribulbar and intrabulbar lymphocytic infiltration, degenerate matrix cells and lymphocytic infiltration of the inferior segment. (E) Damaged HF was observed in the dermis of a graft injected with aGalCer þ IL-2-activated PBMCs that was also treated with an anti-NKT blocking antibody. (F) Lymphocytic infiltration was observed in HFs of a graft injected with a-GalCer þ IL-2activated PBMCs, which was also injected daily with IL-10 blocking antibodies. Immunohistochemical staining of HFs in grafts treated with a-GalCer þ IL-2-activated PBMCs that had been subsequently injected 1x/day with anti-IL-10 blocking antibodies revealed infiltrates of cells that were (G) CD4þ (H) CD8þ (I) IFN-gþ (J) HLA-A,B,Cþ and (K) HLA- DRþ. In contrast, (I-P) grafts treated with isotype control injections instead of IL-10 blocking antibodies exhibited an absence of the inflammatory cells and DR expression by the HFs. (Q) Quantitative data demonstrate the change in inflammatory cell number per experimental condition. (R) HLA-A,B,C and -DR expression in grafts treated with pre-incubation of aGalCer þ IL-2-activated PBMCs compared with grafts that had been injected daily with anti-IL-10 blocking antibodies. Control grafts treated with pre-incubation of a-GalCer þ IL-2activated PBMCs and injected daily with isotype control demonstrated preservation of the protective effect of a-GalCer. n ¼ eight healthy human donors, 5e13 mice per each treatment group. Bar e 200 mm, dermal papilla (DP), connective tissue sheath (CTS), hair matrix (HM), perifollicular dermis (PFD), hair bulb (HB).

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

4

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

solution (2 mg) was injected twice a week. 2.6. In vitro co-culturing of iNKT10 and CD8/NKG2D cells Enriched CD8/NKG2D cells are defined here as PBMCs that have been cultured for 14 days in high-dose IL-2 (100 U/ml) according to our previously published characterization [23]. To investigate the regulatory effect of iNKT10 cells on CD8/NKG2D cells, PBMCs from the same blood donor were divided into two wells. The first well (well 1) was cultured for 14 days with a-GalCer (100 ng/ml) and IL2 (100 U/ml), while the second (well 2) was cultured only with IL-2 (100 U/ml) to generate enriched CD8/NKG2D cells. Then, magnetically purified iNKT10 cells from well 1 (1  105) were co-cultured with enriched CD8/NKG2D cells from well 2 (1  106). For control cultures, no iNKT10 cells were added. The cells were co-cultured for three days prior to harvest for FACS analysis. 2.7. Immunohistochemistry and immunofluorescence analysis Five-micrometer paraffin sections were used. Antigen retrieval was for 20 min at 90  C in a microwave. Specimens were blocked for 30 min to prevent nonspecific binding and were incubated with 1ry antibody (Ab) overnight, followed by a wash and incubation with biotinylated 2nd Ab (Jackson ImmunoResearch, West Grove, PA), and subsequent binding with horseradish peroxidase conjugated streptavidin. Markers were revealed with AEC (red) (Aminoethyl Carbazole Substrate kit). Sections were then mounted and analyzed under a light microscope. Slides were evaluated blindly by experienced observers. The fields of immunostained sections were presented randomly to the blinded observers. The positive cells around and within the hair follicles were counted in an area of 0.66 mm2. The following primary antibodies were used: anti-CD8 (Cell MARQUE), anti-CD4 (DAKO), anti-HLA-A/B/C (Abcam), anti-HLA-DR (Abcam) and anti-IFN-g (Abcam). For immunofluorescence analysis, the same protocol was used as above (immunohistochemical analysis), except slides were blocked for 1 h to prevent nonspecific binding and then incubated with a mixture of primary antibodies overnight. The following day, slides were washed and incubated with secondary antibodies for 1 h. Slides were washed and incubated with DAPI for 10 min, then washed and mounted. Analysis was done using the Confocal LSM 700 Upright Microscope. The following primary antibodies were used: anti-TCR Vb11 (Beckman Coulter Inc.), anti-IL-10 (Santa Cruz biotechnology), anti-CD1d (Abcam), and 6B11 anti-iNKT cell (Beckman Coulter Inc.). The following secondary antibodies and reagents were used:

Alexa Flour 488 goat anti-Mouse IgG, Alexa Flour 488 goat antirabbit IgG, Alexa Flour 594 goat anti-Mouse IgG, Alexa Flour 647conjugated Streptavidin, Biotin goat anti-rabbit IgG, and Biotin goat anti-mouse IgG2a (Jackson Immune Research Laboratories). 2.8. Phenotypic characterization of cells PBMCs from healthy donors were activated with IL-2 and/or aGalCer as described above. After 14 days, cells were counted, and six hours prior to FACS staining, Brefeldin A was added to the cell culture (in a final concentration of 1:1000). Cells were then collected (1e1.5  106 cells/tube) and centrifuged at 300 g (1200 RPM) for 5 min, and washed twice in staining buffer (1 ml of 1% Bovine Serum Albumin (BSA) in 1x sterile PBS). Following Brefeldin A treatment, cells were washed once with 1 ml staining buffer, then Fixation/Permeabilization solution (250 ml) was added and cells were incubated for 20 min at 4  C. Cells were then washed twice with 1xBD Perm/Wash buffer (to keep cells permeabilized). Intracellular staining antibody mixtures were added and incubated for 30 min at RT in darkness, and the cells were then washed twice with 1xBD Perm/Wash buffer (BD Cytofix/Cytoperm™ Fixation/ Permeabilization Kit). Five-hundred ml sterile PBS was added to each tube and specimens were prepared for FORTESSA flow cytometry analysis. The following antibodies were used: FITC anti-iNKT cell (6B11), APC anti-IL-10 (BD PharmingenTM), PE/Cy7 anti-CD20, PE-Alexa 594 anti-CD49d, brilliant violet 421 anti-CD279 (PD1), brilliant violet 605 anti-CD45RA, brilliant violet 510 anti-CD304 (BioLegend), brilliant violet 421 anti-CD69 and anti-CD28 (BioLegend), PerCP/ Cy5.5 anti-perforin (BioLegend), pacific blue anti-granzyme B (BioLegend), and PE/Cy7 anti-CCR7 (BioLegend). Compensation was done using CompBeads (BDTM Biosciences). The live lymphocyte population was gated and the level of positive cells was assessed using FlowJo software (Tree Star, Ashland, OR). 2.9. Cell division analysis Cell division analysis was performed using 5, 6carboxyfluorescein diacetate succinimidyl ester (CFSE)-emitting cells (CFSE cell division kit, BioLegend). CD8/NKG2D cells from Section 2.6 (see above) were labeled with CFSE (0.2 mM, cells incubated at 37  C for 10 min). Cells were then washed immediately with RPMI 1640, 10% human AB serum (Sigma-Aldrich Co. LLC), 1% L-glutamine, and 1% Penicillin-Streptomycin antibiotics (media components; biological Industries, Kibbutz Beit Ha’Emek, Israel). Then, cells were incubated with iNKT10 cells and cultured for four days. FACS analysis was performed as described above.

Table 1 Distribution of inflammatory cells, and HLA-A,B,C and HLA-DR expression by HFs. Normal scalp skin biopsies were transplanted to Beige-SCID mice and injected intradermally with autologous IL-2-activated PBMCs (PBMCs/IL-2). For some experimental groups, PBMCs were additionally cultured in the presence of a-GalCer, and these cells were either (i) subsequently injected into grafts, (ii) depleted of iNKT cells and then injected into grafts, or (iii) cultured in the presence of iNKT-blocking or isotype control antibodies prior to injection into grafts. Other mice bearing grafts that were injected with IL-2-activated PBMCs also received 1x/day intragraft injections of anti-IL-10 blocking antibodies (see Fig. 1). Grafts were harvested at 75 days post-PBMC-injection and analyzed by immunochemistry. Marker

Group PBMC's/IL-2/aGalCer

CD8 CD4 IFNy HLA-A,B,Ca HLA-DRa Damaged hairfollicles (%) a

Depletion of iNKT cells

Isotyoe control (iNKT)

P<

20.7 ± 14 7.2 ± 6 11 ± 1.4 62 ± 4.2 71 ± 0.7 80 ± 1.2

6±2 1.7 ± 1 2.5 ± 2 18 ± 4.4 25 ± 0.7 20 ± 0.7

0.05 0.03 0.05 0.05 0.05 0.05

Percent of positive follicles.

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

5

Fig. 2. A. Flow cytometry analysis of iNKT10 cells. Representative flow cytometry dot plots visualize iNKT10 cells (red circle) after two weeks of PBMC culture. PBMCs cultured with IL-2 and a-GalCer yielded 41.6% iNKT10 cells among CD45RA-high, CD49dþ cells, whereas the equivalent gate in PBMCs cultured with IL-2 alone yielded only 1.39% iNKT10 cells. Other markers found to be increased in iNKT10 cells included CD49d, CD45RA, CD279 and CD304, while expression of CD20 was negative (20,26). B. The mean number of iNKT10 cells after culture of PBMCs including all six donors, with and without a-GalCer (p 0.04). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

6

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

Fig. 3. Immunofluorescence analysis of AA pathology in human scalp skin xenografts, and a primary AA versus control patient. (A,B) Cells in xenografts were stained with DAPI (marking nuclei, blue), anti-iNKT cell TCR (6B11 mAb, red), and anti-IL-10 cytokine (green). (A) iNKT10 cells (yellow) were more numerous in grafts injected with aGalCer þ IL-2-activated PBMCs than in grafts (B) injected with IL-2-activated PBMCs. (C) iNKT10 cell counts were analyzed as shown above in (A-B), and an inverse correlation was observed with the development of AA lesions. iNKT10 cells were more numerous in grafts injected with a-GalCer þ IL-2-activated PBMCs (associated with no AA pathology) than in grafts injected with IL-2-activated PBMCs (positive for AA pathology; p < 0.0008). When iNKT cells were either depleted from a-GalCer þ IL-2-activated PBMCs or blocked by antiiNKT antibody during PBMC culture, iNKT10 cells were then reduced in tissue sections and protection from AA lesion development was lost (iNKT depletion, p < 0.02; iNKT blockade, p < 0.001). Similarly, iNKT10 cell number was reduced and the development of AA was observed in xenografts that received a-GalCer þ IL-2-activated PBMCs followed by 1x/day injection of anti-IL-10 blocking antibodies, when compared to isotype control injections (p < 0.004). (D-E) Sections from an AA patient versus control volunteer were stained with DAPI (blue) and anti-CD1d (red). (D) Stronger CD1d expression by the follicular epithelium in the AA patient was observed compared (E) with that of the control volunteer. (F)

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

2.10. Statistical analysis Data are presented as the mean ± standard error of the mean (SEM). The study parameters of each group were compared using the Kruskal-Wallis test, followed by a Mann-Whitney U test. Statistical significance was set at p < 0.05. 3. Results and discussion 3.1. In vivo experiments 3.1.1. a-GalCer pre-treatment prevents the experimental induction of AA lesions in human scalp skin xenografts in vivo In this study, we used a humanized mouse model that recapitulates human AA pathology [20e22]. In this in vivo model, AA lesions can be rapidly induced experimentally in previously healthy human scalp skin xenotransplants by intragraft injection of autologous healthy-donor PBMCs that were pre-activated in vitro by IL-2 [23]. Briefly, full-thickness hair-bearing scalp skin from healthy human donors was transplanted onto Beige-SCID mice, while autologous human PBMCs (from the same donors) were activated by high-dose IL-2 in cell culture (see Materials and Methods). PBMCs activated in this way expand multiple NKG2Dþ lymphocyte subsets, which, when subsequently injected into the skin grafts, induce a phenocopy human AA [27] with all of its clinical hallmarks [1e4]: focal hair loss in normal appearing skin, histological hair follicle (HF) dystrophy and miniaturization, premature HF regression (catagen induction), and HF immune privilege collapse (Fig. 1 A; Supplementary Fig. 1) [21,23]. By comparison, these AA hallmarks were absent, and normal terminal anagen HFs were present in control xenografts that had been injected with PHA-activated PBMCs [21] (Fig. 1 B; Supplementary Fig. 1). The degree of hair loss (Supplementary Fig. 1) and HF pathology in the model correlated well with the inflammatory changes thought to be causally involved in AA pathogenesis [1e3]: as expected, the intragraft injection of IL-2-activated PBMCs caused lesional accumulation of CD4þ and CD8þ T cells, together with upregulation of IFN-g, HLAA,B,C and HLA-DR expression in and around HFs (Supplementary Fig. 2 AeE) as indicators of HF immune privilege collapse [31]. Interestingly, when the known iNKT cell activator, a-GalCer [28,32,33], was added to the PBMCs cell culture (100 ng/ml for 14 days), IL-2-activated PBMC injection lost the ability to induce AA lesions in the scalp skin xenografts and HFs continued to show a healthy phenotype (Fig. 1 C; Supplementary Fig. 1). Similarly, when the injection of a-GalCer into skin xenografts was performed subsequent to the IL-2-activated PBMC injection, the development of AA lesions was also prevented (Supplementary Fig. 1). Based on previous work [28,32,33], this suggested the hypothesis that the AA-like phenotype might have been prevented by an a-GalCer-induced expansion of iNKT cell regulatory activity [34,35]. 3.1.2. AA prevention by a-GalCer in human skin xenotransplants in vivo depends on NKT cells and IL-10 This hypothesis was probed by assessing whether the AAprotective effect of a-GalCer was lost when iNKT cells were depleted by magnetic beads (using the anti-NKT cell TCR antibody, 6B11) prior to PBMCs injection into the human scalp skin grafts (Fig. 1 D; Supplementary Fig. 1), or when NKT TCR-blocking

7

antibody was added to the PBMCs cell culture (Fig. 1 E; Supplementary Fig. 1). As shown in Fig. 1D and E and Supplementary Fig. 1, this was indeed the case. Protection of human skin xenotranslants from experimentally induced AA lesions in vivo was also lost if grafts that had been injected with a-GalCer þ IL-2activated PBMCs were, in addition, injected daily with IL10blocking antibodies (Fig. 1 F; Supplementary Fig. 1). After co-culture with a-GalCer, IL-2-activated PBMCs injected into grafts induced no perifollicular accumulation of CD4þ and CD8þ T cells and no upregulation of peri- and intrafollicular IFN-g, HLA-A,B,C and HLA-DR expression (Supplementary Fig. 2 FeJ). Importantly, these peri- and intrafollicular human T cell infiltrates reappeared in vivo when iNKT cells were depleted from the injected PBMCs (Table 1; Supplementary Fig. 2 KeO), or when iNKT cells had been blocked by anti-NKT TCR antibodies during cell culture prior to PBMCs injection (Supplementary Fig. 2 PeT). The AA-protective effect of co-culturing IL-2-activated PBMCs with a-GalCer was also lost when grafts were injected daily with IL-10 blocking antibodies, which subsequently produced the hair loss and inflammation phenotype characteristic for AA (Fig. 1GeK). In contrast, grafts injected with isotype control antibodies demonstrated normal anagen HFs (Fig. 1 LeP and Q,R). While post-IL-10 receptors signaling events could not be further dissected in the current study, it has already been established in several experimental systems that following engagement of the IL10 receptor distinct JAK-STAT pathways and downstream signaling events that effect nuclear transcriptional events result in the initiation of extensive anti-inflammatory pathways and cytokines [36e39]. The present study provides the first evidence that an iNKT/IL-10 axis exhibits regulatory, autoimmunity-suppressing functional properties [28] in a human organ in vivo. Additionally, a-GalCer administration can be used to prevent the development of experimentally induced AA in human skin in vivo by a mechanism that requires both NKT cells and secreted IL-10. 3.1.3. Anti-iNKT TCR or anti-IL-10 blocking antibodies prevent the regulatory effect of a-GalCer These data raised the possibility that iNKT10 cells [28,40] can play a previously unrecognized protective role in the pathobiology of human AA. If confirmed, this would be important since a functional or even therapeutic role for iNKT10 cells has not previously been demonstrated in any human autoimmune disease. Moreover, there is currently no method available for reliably preventing disease progression in AA patients [1e4]. Therefore, we further investigated the role of iNKT10 cells in this humanized AA mouse model. Flow cytometry showed that neither fresh PBMCs nor IL-2activated PBMCs contained detectable numbers of iNKT10 cells. In contrast, a-GalCer þ IL-2-activated PBMCs contained a clearly identifiable population of typical iNKT10 cells, as demarcated by increased expression of CD49d, CD45RA, CD279, CD304, the NKT TCR Va24, and the demonstration of intracellular IL-10, while these cells were negative for CD20 [28,40] (Fig. 2A). FACS analysis revealed a robust, significant and reproducible difference between the mean number of iNKT10 cells after culture of PBMCs of all the 6 donors, with and without a-GalCer (p < 0.04) (Fig. 2B). PBMCs cultured with IL-2 plus a-GalCer contained 8.3 þ 2.1% of iNKT cell compared to 0.03 ± 0.06% in PBMCs cultured with IL-2 and without

Immunohistomorphometry analysis of CD1d expression by lesional HF epithelium of AA patients compared to HFs of normal scalp skin of volunteers (p < 0.04). (G, H) Cells in xenografts were stained with DAPI (blue), anti-CD1d (red) and anti-IL-10 cytokine (green). (G) Colocalization of CD1d and IL-10 around the HF appeared stronger in grafts injected with a-GalCer þ IL-2-activated PBMCs than (H) in grafts injected with IL-2-activated PBMCs. n ¼ eight healthy donors, 5e13 mice per each treatment group. Bar e 150 mm, dermal papilla (DP), connective tissue sheath (CTS), hair matrix (HM), perifollicular dermis (PFD), hair bulb (HB). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

8

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

Fig. 4. Therapeutic effect of exogenous IL-10 or a-GalCer in pre-established AA lesions in humanized xenografts. (AeC) Testing the therapeutic potential of exogenous IL-10. (A) Prior to the induction of experimental AA, human scalp skin xenografts bore hair. (B) Injection of IL-2-activated autologous PBMCs caused autoimmune lesions and hair loss. (C) Subsequent treatment with exogenous IL-10induced hair re-growth. (DeF) Testing therapeutic potential of a-GalCer. (D) Prior to the induction of experimental AA, human scalp skin xenografts bore hair. (E) Injection of IL-2-activated autologous PBMCs caused autoimmune lesions and hair loss. (F) Subsequent treatment with a-GalCer induced hair regrowth. (G) Statistical analysis demonstrates a significant increase in the number of hairs in the pre-induced AA lesions following treatment with either IL-10 or a-GalCer compared to control BSA treatment. Three healthy donors were included in this analysis, transplanted onto 3e6 mice for each treatment group.

a-GalCer. Since they are CD45RA-, CCR7-, these iNKT10 cells represent effector memory T cells [42], rather than naive T cells (Supplementary Figs. 3A and B) [41]. Immunofluorescence microscopy and quantitative immunohistomorphometry documented significantly more iNKT10 cells in human skin xenografts grafts injected with a-GalCer þ IL-2activated PBMCs (Fig. 3 A) than in grafts injected with IL-2activated PBMCs not co-cultured with a-GalCer (p < 0.0008) (Fig. 3B and C). In contrast, significantly fewer iNKT10 cells were observed in AA-induced grafts under these conditions: (i) if aGalCer þ IL-2-activated PBMCs were depleted of iNKT cells prior to xenotransplant injection, (p < 0.02); (ii) if anti-NKT TCR blocking antibody was added during cell culture (p < 0.001) (Fig. 3C); and (iii) if grafts were injected with anti-IL-10 blocking antibodies (p < 0.004) (Fig. 3C). These results demonstrate that a-GalCer treatment induced an increased number of iNKT10 cells in human skin xenotransplants in an IL-10- dependent manner. 3.1.4. a-GalCer treatment downregulates CD1d expression by HFs in human skin xenotransplants We next asked whether the HF epithelium in hair loss lesions

that had developed spontaneously in AA patients in vivo express CD1d, which activates iNKT cells via presentation of glycolipids [28,40]. Indeed, immunofluorescence microscopy showed significantly stronger CD1d protein expression in situ in lesional HF epithelium of AA patients (p < 0.05) (Fig. 3D) compared to the HF epithelium of healthy volunteers (Fig. 3E and F). Importantly, CD1d overexpression was also seen in the humanized AA mouse model, thus underscoring how well the latter reflects the human disease [23,27]. CD1d overexpression was observed only in the epithelium of lesional HFs in the xenotransplants that were not treated with a-GalCer compared to those that were treated (p < 0.05) (Fig. 3G and H). Moreover, the perifollicular co-localization of CD1d and IL-10 in human xenografts grafts injected with a-GalCer þ IL-2-activated PBMCs (Fig. 3G) appeared stronger than in grafts injected with IL-2-activated PBMCs (3.5 ± 1.4 versus 0.9 ± 0.2, respectively, p < 0.03) (Fig. 3H). In the present study we thus demonstrate that PBMCs cultured with high-dose IL-2 and a-GalCer induce the expansion of iNKT10, which produce high levels of IL-10. The observed downregulation of CD1d expression in the a-GalCer treatment group is in line with the literature, since IL-10 is known to reduce CD1d expression

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

Fig. 5. Co-culture of CD8þNKG2Dþ cells with iNKT10 cells. CFSE-labeled PBMCs were cultured with IL-2 (100 U/ml), which induced increased numbers of CD8/NKG2D cells. In parallel, PBMCs from the same donors were treated with a-GalCer and IL-2, which generated increased numbers of iNKT10 cells. From the latter cultures, iNKT cells were purified by magnetic separation and were added to cultures of CD8/NKG2D cells. For comparison, cultured PBMCs treated only with IL-2 were used as controls [21]. (A) CD8 T cell proliferation, assessed by CFSE dilution, was significantly inhibited by the presence of iNKT10 cells compared to (B) the control culture. (C) Quantitative data illustrate the significant inhibitory effect of iNKT10 cells on CD8/NKG2D cells (P < 0.04). (D-F) A significant decrease was observed for intracellular expression of (D) IFN-g, (E) granzyme and (F) perforin by CD8 cells following coculture with iNKT10 cells compared to the control groups (G-I) (P < 0.02, P < 0.03 and P < 0.001, respectively). Similarly, decreased expression of (J) NKG2D, (K) CD25, (L) CD69 and (M) CD28 by the CD8 cells was observed following co-culture with iNKT10 cells compared to the controls (N-Q) (P < 0.02, P < 0.03, P < 0.002 and P < 0.001, respectively).

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

10

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

Fig. 6. Double IHC staining demonstrated (A) the presence of cell-cell contacts between iNKT10 cells and CD8þNKG2Dþ cells in AA-induced xenotransplants (PBMCs/IL-2) treated with a-GalCer versus (B) relative absence of these cell-cell contacts in xenotransplants that were not treated with a-GalCer and (C) relative absence of these cell-cell contacts in normal scalp xenotransplants. (D) Quantitative data illustrate the significant differences (p < 0.001 and p < 0.001, respectively). Similarly, triple IHC staining demonstrated increased cell-cell contacts between (E) iNKT10 cells, CD8þNKG2Dþ cells and dendritic cells (CD11cþ) in the AA-induced xenografts that had been treated with a-GalCer compared to (F) xenotransplants not treated with a-GalCer (PBMCs/IL-2) and (G) normal xenotransplants. (H) Quantitative data show the significant differences (p < 0.002 and p < 0.002, respectively).

[42,43]. The reduced level of IFN-g secretion by CD8/NKG2Dþ cells after could contribute to this [44]. To confirm this plausible working hypothesis, follow-up work would need to repeat the CD1d downregulation assay in the presence/absence of IL-10-or IFN-g neutralizing antibody or selective antagonists of the cognate receptors. Taken together, these data strongly suggest that iNKT10 cells are promising therapeutic targets even in established AA lesions in human skin. 3.1.5. a-GalCer and IL-10 blockade both promote hair re-growth in established AA lesions in vivo To probe this hypothesis further, AA lesions were induced experimentally as described above until the point where human scalp skin grafts displayed fully established AA lesions with hair loss [27]. Grafts were then injected daily with either a-GalCer (2 mg/ graft), IL-10 (200 ng/graft), or BSA (200 ng/graft, negative control) for 45 days. Significant, macroscopically visible hair regrowth occurred in response to a-GalCer (p < 0.01) or IL-10 (p < 0.01) injections compared to the control group (Fig. 4AeG; Supplementary Fig. 4). Interestingly, IL-10 has previously been shown to partially restore experimentally induced collapse of human HF immune privilege ex vivo [45], and intralesional upregulation of IL-10 expression might underlie the beneficial therapeutic effect of contact sensitizer therapy in AA patients that respond well to this topical treatment [46]. Conversely, downregulation of IL-10 expression by perifollicular mast cells may promote AA pathogenesis [6]. Hence, the a-GalCer-induced upregulation of IL-10 (and possibly other crucial hair follicle “immune privilege guardians” [1,31]) in and around HFs may restore the collapsed immune privilege of lesional HFs. This likely represents a key element in the pathobiology of AA, since this hair loss pattern has never been shown to develop in HFs with intact immune privilege [1,48] (Supplementary Fig. 4).

In conclusion, our study shows for the first time that Inkt10 cells exhibit regulatory, autoimmunity-suppressing functional properties in a human organ in vivo. Moreover, we demonstrate that aGalCer administration prevents the development of experimentally induced AA in human skin in vivo by a mechanism that requires both iNKT cells and secreted IL-10. Besides identifying iNKT10 cells as novel key players in AA, we show that this common autoimmune disorder is an excellent model disease for exploring the role of iNKT10 cells with regulatory functions in human autoimmune pathology in vivo. Our findings may be particularly relevant for extracutaneous autoimmune diseases that are prominently associated with immune privilege collapse, e.g., autoimmune uveitis [31,47] or in sites of dysfunctional active regulation of immunosurveillance such as in the central nervous system during multiple sclerosis [48]. Our study also underscores the utility and instructiveness of the humanized mouse model for basic and translational AA research [23]. Finally, a-GalCer could be considered as a complementary or alternative therapeutic modality to JAK-STAT inhibitors that are currently undergoing clinical trials in AA [4,49]. The present study strongly encourages systematically exploring next how iNKT10 cells can be best targeted in clinical autoimmune disease settings. It is conceivable that, for expansion of iNKT10 activity and function, a-GalCer could be considered a relatively safe lead compound that may be repositioned for AA therapy. 3.2. In vitro experiments 3.2.1. NKT10 cells suppress CD8þ/NKG2D directly and/or indirectly Next, we asked how iNKT10 cells may interact with the CD8þ/ NKG2Dþ T cells, i.e. the key pathogenic T cell population in AA pathobiology [1,50], and how NKT10 cells may reduce the accumulation of CD8þ/NKG2DþT cells and the dermal IFN response in experimentally induced human skin lesions in vivo. To this end.

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12

We co-cultured CFSE-labeled CD8þ/NKG2Dþ cells with iNKT10 cells. Upon gating CD8þ T cells (Fig. 5AeC), we found that the proliferation of CD8þ cells was significantly suppressed by the presence of iNKT10 cells (P < 0.04), as was the intracellular expression of IFN-g, granzyme, and perforin in CD8*T cells (P < 0.02, P < 0.03, P < 0.001 respectively) (Fig. 5DeI). Co-culture with iNKT10 cells also resulted in a reduced membrane expression of NKG2D, CD25, CD69, and CD28 of CD8þ T cells (P < 0.02, P < 0.03, P < 0.002, P < 0.001, respectively) (Fig. 5JeQ). Our results suggest that proliferation and cytokine production of CD8þNKG2Dþ cells can be inhibited directly by cell-cell interactions with iNKT10 cells, at least to some extent, conceivably involving the release of immunosuppressive cytokines such as IL10, given that IL-10 inhibits proliferation and production of inflammatory cytokines by T cells [51]. In addition, IL-10 can reduce CD8þ T cell activity by decreasing antigen sensitivity [52] and can prevent effector/memory T cell activity [53]. Double immunofluorescence staining revealed physical cell-cell contact between iNKT10 and CD8þ/NKG2Dþ cells (Fig. 6AeC). This further supports that direct suppressive effects of iNKT10 on CD8 cells is possible. Triple immunostaining showed that cell contacts can also be seen between iNKT10, CD8þ/NKG2Dþ and dendritic cells (CD11cþ) in the induced AA lesions in human skin xenografts treated with a-GalCer (Fig. 6DeF). Hence, iNKT10 cells may suppress CD8þ/NKG2Dþ also indirectly via interacting with the antigen presenting cells (APCs) [54,55]. In line with the literature [56] both mechanisms may also occur side-by-side. Acknowledgements The study was supported in part by the NIHR Manchester Biomedical Research Centre (Inflammatory Hair Disease Programme; Lead: R.P.). We thank Natalia Kaploon for her technical assistance. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.jaut.2018.04.001. References [1] A. Gilhar, A. Etzioni, R. Paus, Alopecia areata, N. Engl. J. Med. 366 (2012) 1515e1525. [2] K.J. McElwee, A. Gilhar, D.J. Tobin, Y. Ramot, J.P. Sundberg, M. Nakamura, et al., What causes alopecia areata? Exp. Dermatol. 22 (2013) 609e626. [3] C.H. Pratt, L.E. King Jr., A.G. Messenger, A.M. Christiano, J.P. Sundberg, Alopecia areata, Nat. Rev. Dis. Primers 3 (2017) 17011. [4] L.C. Strazzulla, E.H.C. Wang, L. Avila, K. Lo Sicco, N. Brinster, A.M. Christiano, J. Shapiro, Alopecia areata: disease characteristics, clinical evaluation, and new perspectives on pathogenesis, J. Am. Acad. Dermatol. 78 (2018) 1e12. [5] T. Ito, N. Ito, M. Saathoff, H. Hashizume, H. Fukamizu, B.J. Nickoloff, et al., Maintenance of hair follicle immune privilege is linked to prevention of NK cell attack, J. Invest. Dermatol. 128 (2008) 1196e1206. [6] M. Bertolini, F. Zilio, A. Rossi, P. Kleditzsch, V.E. Emelianov, A. Gilhar, et al., Abnormal interactions between perifollicular mast cells and CD8þ T-cells may contribute to the pathogenesis of alopecia areata, PLoS One 9 (2014), e94260. [7] A. Gilhar, Y. Ullmann, T. Berkutzki, B. Assy, R.S. Kalish, Autoimmune hair loss (alopecia areata) transferred by T lymphocytes to human scalp explants on SCID mice, J. Clin. Invest. 101 (1998) 62e67. [8] L. Xing, Z. Dai, A. Jabbari, J.E. Cerise, C.A. Higgins, W. Gong, et al., Alopecia areata is driven by cytotoxic T lymphocytes and is reversed by JAK inhibition, Nat. Med. 20 (2014) 1043e1049. [9] H. Guo, Y. Cheng, J. Shapiro, K. McElwee, The role of lymphocytes in the development and treatment of alopecia areata, Expet Rev. Clin. Immunol. 11 (2015) 1335e1351. [10] A. Jabbari, J.E. Cerise, J.C. Chen, J. Mackay-Wiggan, M. Duvic, V. Price, et al., Molecular signatures define alopecia areata subtypes and transcriptional biomarkers, EBioMedicine 7 (2016) 240e247. [11] J.C. Chen, J.E. Cerise, A. Jabbari, R. Clynes, A.M. Christiano, Master regulators of infiltrate recruitment in autoimmune disease identified through networkbased molecular deconvolution, Cell Syst. 1 (2015) 326e337.

11

[12] Z. Dai, L. Xing, J. Cerise, E.H. Wang, A. Jabbari, A. de Jong, et al., CXCR3 blockade inhibits T cell migration into the skin and prevents development of alopecia areata, J. Immunol. 197 (2016) 1089e1099. [13] G. Kaufman, R. d'Ovidio, A. Kaldawy, B. Assy, Y. Ullmann, A. Etzioni, et al., An unexpected twist in alopecia areata pathogenesis: are NK cells protective and CD49bþ T cells pathogenic? Exp. Dermatol. 19 (2010) e347e349. [14] S. Beyaz, J.H. Kim, L. Pinello, M.E. Xifaras, Y. Hu, J. Huang, et al., The histone demethylase UTX regulates the lineage-specific epigenetic program of invariant natural killer T cells, Nat. Immunol. 18 (2017) 184e195. [15] E. Bianchini, S. De Biasi, A.M. Simone, D. Ferraro, P. Sola, A. Cossarizza, et al., Invariant natural killer T cells and mucosal-associated invariant T cells in multiple sclerosis, Immunol. Lett. 183 (2017) 1e7. [16] X. Fan, A.Y. Rudensky, Hallmarks of tissue-resident lymphocytes, Cell 164 (2016) 1198e1211. [17] J. O'Keeffe, M. Podbielska, E.L. Hogan, Invariant natural killer T cells and their ligands: focus on multiple sclerosis, Immunology 145 (2015) 468e475. [18] S. De Biasi, A.M. Simone, M. Nasi, E. Bianchini, D. Ferraro, F. Vitetta, et al., iNKT cells in secondary progressive multiple sclerosis patients display proinflammatory profiles, Front. Immunol. 7 (2016) 555. [19] D.I. Godfrey, A.P. Uldrich, J. McCluskey, J. Rossjohn, D.B. Moody, The burgeoning family of unconventional T cells, Nat. Immunol. 16 (2015) 1114e1123. [20] A. Gilhar, A. Keren, A. Shemer, Y. Ullmann, R. Paus, Blocking potassium channels (Kv1.3): a new treatment option for alopecia areata? J. Invest. Dermatol. 133 (2013) 2088e2091. [21] A. Gilhar, A. Keren, A. Shemer, R. d'Ovidio, Y. Ullmann, R. Paus, Autoimmune disease induction in a healthy human organ: a humanized mouse model of alopecia areata, J. Invest. Dermatol. 133 (2013) 844e847. [22] A. Keren, A. Shemer, Y. Ullmann, R. Paus, A. Gilhar, The PDE4 inhibitor, apremilast, suppresses experimentally induced alopecia areata in human skin in vivo, J. Dermatol. Sci. 77 (2015) 74e76. [23] A. Gilhar, A.G. Schrum, A. Etzioni, H. Waldmann, R. Paus, Alopecia areata: animal models illuminate autoimmune pathogenesis and novel immunotherapeutic strategies, Autoimmun. Rev. 15 (2016) 726e735. [24] K. Kabashima, T. Nomura, Revisiting murine models for atopic dermatitis and psoriasis with multipolar cytokine axes, Curr. Opin. Immunol. 48 (2017) 99e107. [25] I. Bjelobaba, V. Begovic-Kupresanin, S. Pekovic, I. Lavrnja, Animal models of multiple sclerosis: focus on experimental autoimmune encephalomyelitis, J. Neurosci. Res. 96 (2018) 1021e1042. [26] A. Keren, A. Shemer, A. Ginzburg, Y. Ullmann, A.G. Schrum, R. Paus, A. Gilhar, Innate lymphoid cells 3 induce psoriasis in xenotransplanted healthy human skin, J. Allergy Clin. Immunol. (2018) ahead of print. [27] A. Gilhar, A.G. Schrum, A. Etzioni, H. Waldmann, R. Paus, Alopecia areata: animal models illuminate autoimmune pathogenesis and novel immunotherapeutic strategies, Autoimmun. Rev. 15 (2016) 726e735. [28] D. Sag, P. Krause, C.C. Hedrick, M. Kronenberg, G. Wingender, NKT10 cells are a distinct regulatory invariant NKT cell subset, J. Clin. Invest. 124 (2014) 3725e3740. [29] J.W. Oh, J. Kloepper, E.A. Langan, Y. Kim, J. Yeo, M.J. Kim, et al., A guide to studying human hair follicle cycling in vivo, J. Invest. Dermatol. 136 (2016) 34e44. [30] M. Kinori, J.E. Kloepper, R. Paus, Can the hair follicle become a model for studying selected aspects of human ocular immune privilege? Invest. Ophthalmol. Vis. Sci. 52 (2011) 4447e4458. [31] R. Paus, B.J. Nickoloff, T. Ito, A 'hairy' privilege, Trends Immunol. 26 (2005) 32e40. [32] C.S. Fernandez, G. Cameron, D.I. Godfrey, S.J. Kent, Ex-vivo a-galactosylceramide activation of NKT cells in humans and macaques, J. Immunol. Meth. 382 (2012) 150e159. [33] J.T. Hung, J.R. Huang, A.L. Yu, Tailored design of NKT-stimulatory glycolipids for polarization of immune responses, J. Biomed. Sci. 24 (2017) 22. [34] J. Chen, M. Wu, J. Wang, X. Li, Immunoregulation of NKT cells in systemic lupus erythematosus, J. Immunol. Res. (2015) 206731. [35] E. Clancy-Thompson, G.Z. Chen, P.M. Tyler, M.M. Servos, M. Barisa, P.J. Brennan, et al., Monoclonal invariant NKT (iNKT) cell mice reveal a role for both tissue of origin and the TCR in development of iNKT functional subsets, J. Immunol. 1 (199) (2017) 159e171. [36] J.K. Riley, K. Takeda, S. Akira, R.D. Schreiber, Interleukin-10 receptor signaling through the JAK-STAT pathway. Requirement for two distinct receptorderived signals for anti-inflammatory action, J. Biol. Chem. 274 (1999) 16513e16521. [37] X. Ma, W. Yan, H. Zheng, Q. Du, L. Zhang, Y. Ban, N. Li, F. Wei, Regulation of IL10 and IL-12 production and function in macrophages and dendritic cells, F1000Res. 17 (2015) 4. [38] S. Commins, J.W. Steinke, L. Borish, The extended IL-10 superfamily: IL-10, IL19, IL-20, IL-22, IL-24, IL-26, IL-28, and IL-29, J. Allergy Clin. Immunol. 121 (2008) 1108e1111. [39] Y. Zhu, Z. Liu, Y.P. Peng, Y.H. Qiu, Interleukin-10 inhibits neuroinflammationmediated apoptosis of ventral mesencephalic neurons via JAK-STAT3 pathway, Int. Immunopharm. 50 (2017) 353e360. [40] G. Wingender, D. Sag, M. Kronenberg, NKT10 cells: a novel iNKT cell subset, Oncotarget 6 (2015) 26552e26553. ~ a, M. Wolodarski, I. Poschke, Y. Yoshimoto, Y. Yang, M. Nystro € m, [41] Y.P. de Coan et al., Ipilimumab treatment decreases monocytic MDSCs and increases CD8

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001

12

[42]

[43]

[44]

[45]

[46]

[47] [48] [49]

A. Ghraieb et al. / Journal of Autoimmunity xxx (2018) 1e12 effector memory T cells in long-term survivors with advanced melanoma, Oncotarget 8 (2017) 21539e21553. K. Venken, T. Decruy, S. Aspeslagh, S. Van Calenbergh, B.N. Lambrecht, D. Elewaut, Bacterial CD1d-restricted glycolipids induce IL-10 production by human regulatory T cells upon cross-talk with invariant NKT cells, J. Immunol. 191 (2013) 2174e2183. G. Gerlini, A. Tun-Kyi, C. Dudli, G. Burg, N. Pimpinelli, F.O. Nestle, Metastatic melanoma secreted IL-10 down-regulates CD1 molecules on dendritic cells in metastatic tumor lesions, Am. J. Pathol. 156 (2004) 1853e1863. T. Ito, N. Ito, M. Saathoff, A. Bettermann, M. Takigawa, R. Paus, Interferongamma is a potent inducer of catagen-like changes in cultured human anagen hair follicles, Br. J. Dermatol. 152 (2005) 623e631. T. Ito, N. Ito, A. Bettermann, Y. Tokura, M. Takigawa, R. Paus, Collapse and restoration of MHC class-I-dependent immune privilege: exploiting the human hair follicle as a model, Am. J. Pathol. 164 (2004) 623e634. €nig, et R. Hoffmann, E. Wenzel, A. Huth, P. van der Steen, M. Sch€ aufele, A. Ko al., Cytokine mRNA levels in alopecia areata before and after treatment with the contact allergen diphenylcyclopropenone, J. Invest. Dermatol. 103 (1994) 530e533. A.W. Taylor, T.F. Ng, Negative regulators that mediate ocular immune privilege, J. Leukoc. Biol. (2018) ahead of print. N. Negi, B.K. Das, CNS: not an immunoprivilaged site anymore but a virtual secondary lymphoid organ, Int. Rev. Immunol. 37 (2018) 57e68. M. Kennedy Crispin, J.M. Ko, B.G. Craiglow, S. Li, G. Shankar, J.R. Urban, et al.,

[50]

[51] [52]

[53]

[54]

[55]

[56]

Safety and efficacy of the JAK inhibitor tofacitinib citrate in patients with alopecia areata, JCI Insight 1 (2016), e89776. R. Paus, S. Bulfone-Paus, M. Bertolini, Hair follicle immune privilege revisited: the key to alopecia areata management, J. Invest. Dermatol. Symp. Proc. 19 (2018) S12eS17. K.N. Couper, D.G. Blount, E.M. Riley, IL-10: the master regulator of immunity to infection, J. Immunol. 180 (2008) 5771e5777. L.K. Smith, G.M. Boukhaled, S.A. Condotta, S. Mazouz, J.J. Guthmiller, R. Vijay, N.S. Butler, et al., Interleukin-10 directly inhibits CD8þ T cell function by enhancing N-glycan branching to decrease antigen sensitivity, Immunity 40 (2018) 299e312. M. Kamanaka, S. Huber, L.A. Zenewicz, N. Gagliani, C. Rathinam, W. O'Connor Jr., et al., Memory/effector (CD45RB(lo)) CD4 T cells are controlled directly by IL-10 and cause IL-22-dependent intestinal pathology, J. Exp. Med. 208 (2011) 1027e1040. J.E. Jang, C.H. Hajdu, C. Liot, G. Miller, M.L. Dustin, D. Bar-Sagi, Crosstalk between regulatory T cells and tumor-associated dendritic cells negates antitumor immunity in pancreatic cancer, Cell Rep. 20 (2017) 558e571. A. Schildknecht, S. Brauer, C. Brenner, K. Lahl, H. Schild, T. Sparwasser, et al., FoxP3þ regulatory T cells essentially contribute to peripheral CD8þ T-cell tolerance induced by steady-state dendritic cells, Proc. Natl. Acad. Sci. U. S. A. 107 (1) (2010) 199e203. A. Schmidt, N. Oberle, P.H. Krammer, Molecular mechanisms of Treg-mediated T cell suppression, Front. Immunol. 21 (2012) 51.

Please cite this article in press as: A. Ghraieb, et al., iNKT cells ameliorate human autoimmunity: Lessons from alopecia areata, Journal of Autoimmunity (2018), https://doi.org/10.1016/j.jaut.2018.04.001