Construction method of vitiligo model and its application

Abstract

The present disclosure provides a drug target for vitiligo. The drug target is IFN-γ signaling. The present disclosure also provides a method for establishing a vitiligo animal model, or a method for inducing vitiligo in an animal, and the established vitiligo animal model. In addition, the present disclosure also provides a method for screening candidate drugs for treating vitiligo using the animal model.

Description

Background Art

[0001] Vitiligo is a chronic disease that causes white spots to appear on the skin where pigment cells (melanocytes) are lost. Vitiligo affects 0.5-1% of the population and occurs in all ethnic groups. 50% of patients begin to experience pigment loss before the age of 20, and about 80% before the age of 30. In 20% of patients, other family members also have vitiligo. Men and women are equally affected.

[0002] Vitiligo is considered a systemic autoimmune disease associated with a dysregulated innate immune response, but this is still controversial for segmental vitiligo. There are many drugs for treating vitiligo in the prior art, and the treatment mechanisms are different, but the treatment effects are not ideal. The reason is mainly attributed to unclear mechanisms.

[0003] Previously developed vitiligo mouse models rely on adoptive transfer of melanocyte-specific CD8+ T cells isolated from PMEL-specific TCR transgenic mice into Krt14-Kit1 transgenic mice (ref). The need for two transgenic alleles and transfer procedures to induce vitiligo in mouse skin limits additional genetic modifications that can be effectively introduced for in-depth functional studies in vivo. Other existing vitiligo mouse models either await the development of spontaneous vitiligo in transgenic mice, but the development is slow and inefficient (ref); or utilize ectopically expressed melanocyte antigens, but these antigens require complex viral packaging systems or specialized gene gun delivery equipment, which hinders general application. Most importantly, these vitiligo mouse models rely on artificial stimulation that does not occur in patients.

[0004] Therefore, it is necessary to establish an ideal vitiligo animal model to study the exact mechanism or exact pathogenesis of vitiligo in order to screen candidate drugs for the treatment of vitiligo. Summary of the invention

[0005] The inventors of the present invention used immunofluorescence staining to analyze the skin of the border area around the depigmented lesions and pigmented lesions of patients with vitiligo and found that the depigmented lesions contained slightly more CD45+ immune cells than the perilesional area, but most of the infiltrating immune cells were concentrated at the junction of the lesion area and the perilesional area. This interesting distribution pattern of CD45+ immune cells suggests that some recruitment mechanism is coordinating the local aggregation of immune cells, thereby promoting the spread of depigmented areas in the patient's skin.

[0006] The inventors of the present invention used single-cell RNA-seq to analyze all cell types present in the patient's skin to test which skin-resident cells are involved in mediating immune cell recruitment, distinguishing different disease states and revealing the main relevant signaling pathways. The results showed that progressive vitiligo skin contained more CD8+ cytotoxic T cells expressing large amounts of IFNG compared to patients in a quiescent state and healthy donors. The corresponding melanocytes in progressive vitiligo skin upregulated genes involved in immune responses, especially genes that responded to IFN-γ.

[0007] The inventors of the present invention used immunofluorescence staining to examine the spatial distribution patterns of CD8+T cells and IFN-γ responsive cells in the patient's skin. In order to quantify this spatial correlation, vitiligo skin was divided into three regions according to skin pigmentation and T cell infiltration: depigmented lesion area, T cell infiltration area (TIR) ​​and adjacent pigmented perilesional area. Quantification showed that the density of CD8+T cells in TIR was significantly higher than that in lesion area and perilesional area. It was found that the density of pSTAT1 (phosphorylated signal transduction and transcription activator protein 1) + IFN-γ responsive cells in TIR was significantly higher than that in lesion area and perilesional area. Importantly, the density of CD8+T cells was positively correlated with the density of pSTAT1+ cells. This result shows that the regional response to IFN-γ secreted by T cells is associated with a progressive disease state.

[0008] The inventors of the present invention developed a new vitiligo mouse model by inoculating mice with melanoma cells and using antibodies for treating melanoma for immunotherapy. The vitiligo mouse model showed that the response to IFN-γ secreted by T cells is required for the local CD8+ T cell aggregation and cytotoxic activity in the skin.

[0009] Together, the transplant and vitiligo induction experiments suggest that local CD8+ T cell recruitment and activation in the skin requires skin-resident IFN-γ-responsive cells. Specifically, skin dermal cells are able to mediate this effect.

[0010] Next, chimeric knockdown experiments in fibroblasts not only verified the results that IFN-γ-responsive dermal fibroblasts are the main cell type mediating local CD8+ T cell aggregation and activation, but also further revealed that the IFNGR1-JAK1-STAT1 signaling axis in fibroblasts is required for mediating local CD8+ T cell aggregation and activation. Most importantly, these experiments showed that in areas where fibroblasts have heterogeneous responses to IFN-γ, T cells preferentially aggregate toward areas with high IFNGR1-JAK1-STAT1 signaling.

[0011] The inventors of the present invention further demonstrated that IFN-γ-responsive fibroblasts are sufficient to mediate local CD8+ T cell aggregation in vivo and in vitro through secreted chemokines such as CXCL9, CXCL10 and CCL19.

[0012] The inventors have also shown that differences in the intrinsic IFN-γ response of anatomically distinct human fibroblasts correlate with regional disease variation.

[0013] In a first aspect, the present invention provides a drug target for vitiligo. The drug target is IFN-γ signaling. Preferably, the drug target is IFN-γ signaling in skin cells. The skin cells may be endothelial cells, dermal cells, smooth muscle cells or immune cells in the skin. Preferably, the drug target is IFN-γ signaling in skin dermal fibroblasts.

[0014] The present invention uses IFNGR1 KO (IFN-γ receptor 1 knockout) induced vitiligo mice to demonstrate that certain IFN-γ responsive cells or responses to IFN-γ are essential for local CD8+T cell aggregation and cytotoxic activity in the skin. Specifically, the present invention demonstrates that IFN-γ responsive skin dermal fibroblasts are the main cell type mediating local CD8+T cell aggregation and activation. In other words, fibroblasts are the main cell type responsible for coordinating local CD8+T cell aggregation and activation after IFN-γ stimulation in autoimmune skin. In particular, IFN-γ responsive fibroblasts alone are sufficient to coordinate local CD8+T cell aggregation.

[0015] Skin dermal fibroblasts are necessary and sufficient for IFN-γ induction of CD8+T cell local aggregation and activation in response to vitiligo skin. IFN-γ responsive fibroblasts mediate CD8+T cell aggregation by secreted factors. Therefore, preferably, the drug target is a chemokine specifically secreted by fibroblasts induced by IFN-γ signaling. The chemokines specifically secreted by fibroblasts induced by IFN-γ signaling control regional T cell recruitment. Preferably, the chemokine can be CCL5, CCL8, CCL19, CXCL3, CXCL9 and / or CXCL10. IFN-γ responsive fibroblasts are sufficient to mediate CD8+T cell aggregation in vivo and in vitro by secreted chemokines such as CXCL9, CXCL10 and CCL19.

[0016] T cells preferentially aggregate to areas with high IFNGR1-JAK1-STAT1 signaling. Preferably, the IFNGR1-JAK1-STAT1 signaling axis in fibroblasts is necessary for mediating local aggregation and activation of CD8+ T cells. Therefore, preferably, the drug target is IFNGR1-JAK1-STAT1 signaling in skin dermal fibroblasts.

[0017] Fibroblasts from anatomically different body locations show inherent differences in IFN-γ response. Intrinsic differences in anatomically different human dermal fibroblasts are associated with the incidence of vitiligo in different body locations. The present invention shows huge differences in the incidence of vitiligo in eight body regions: the back of the hands, chest and back skin areas are most susceptible to vitiligo, while the palms and arm skin are least susceptible to vitiligo. The incidence of vitiligo is positively correlated with the intrinsic IFN-γ response of skin fibroblasts. Therefore, fibroblasts from anatomically different body locations can be used as different drug targets in IFN-γ response. In particular, fibroblasts from the back of the hands, chest and back skin can be used as drug targets for IFN-γ response.

[0018] The chemokine genes CCL2, CXCL3, CXCL9, CXCL10, and CXCL11 were mainly upregulated in skin fibroblasts from the dorsum of the hand and dorsum of the foot. QPCR validation confirmed the differences in the intrinsic IFN-γ response of fibroblasts from different anatomical locations. The correlation between the IFN-γ response enrichment scores at different body locations and the incidence of vitiligo was evaluated, and the results showed that the incidence of vitiligo was positively correlated with the intrinsic IFN-γ response of skin fibroblasts. Therefore, the upregulated chemokine genes CCL2, CXCL3, CXCL9, CXCL10, and CXCL11 in skin fibroblasts from the dorsum of the hand and dorsum of the foot may serve as drug targets.

[0019] In a second aspect, the present invention provides a method for establishing an animal model of vitiligo, or a method for inducing vitiligo in an animal, the method comprising inoculating melanoma cells into the animal, injecting CD4 depleting antibodies, and removing the melanoma.

[0020] The melanoma cells may be B16F10 or B16 melanoma cells.

[0021] Preferably, the tumor is removed surgically before it spreads and metastasizes.

[0022] Preferably, the animal can be a mouse, rat, dog, pig or cat.

[0023] For mice, when the melanoma volume reached 62.5-256 mm 3 The tumor is then surgically removed to prevent the spread and metastasis of tumor cells.

[0024] Preferably, B16F10 melanoma cells were inoculated on the right side of the back skin of 9-week-old C57 mice; followed by injection of CD4 depleting antibodies on days 4 and 10. The tumor was surgically removed on day 12 to prevent the spread and metastasis of tumor cells.

[0025] Because melanocytes in mouse dorsal skin are located in hair follicles rather than in the epidermis, mouse tail skin was used for vitiligo analysis because it contains melanocytes localized in the epidermis similar to human skin. In tail skin, epidermal depigmentation became evident at 16 weeks after induction, which was mainly determined by the natural renewal rate of pigmented keratinocytes on the skin surface.

[0026] Compared with the tail skin of control mice that contained almost no CD8+ T cells, a large number of CD8+ T cells infiltrated into the skin epidermis of the tail on day 40 after induction of vitiligo, accompanied by a significant loss of melanocytes.

[0027] After induction of vitiligo, administration of CD8-depleting antibodies resulted in a complete blockade of CD8+ T cell infiltration in the epidermis of the tail skin and rescued melanocyte loss. Thus, the method of inducing vitiligo effectively triggered endogenously activated CD8+ T cells to infiltrate the skin, leading to a loss of natural melanocytes located in the epidermis, similar to patients with autoimmune vitiligo.

[0028] The major experimental advantage of our vitiligo induction method or method for establishing a vitiligo mouse model is that it uses only commercially available reagents and can also effectively induce patient-like vitiligo pathology in any mouse strain (even with genetic alterations, such as knockout, conditional knockout, or transgenic); thus, it will be possible to ask in-depth mechanistic questions and screen for effective drugs.

[0029] In a third aspect, the present invention provides an animal model of vitiligo induced by inoculating animals with melanoma cells, injecting CD4 depleting antibodies, and removing the melanoma.

[0030] The melanoma cells may be B16F10 or B16 melanoma cells.Preferably, the tumor is surgically removed before it spreads and metastasizes.

[0031] Preferably, the animal is a mouse, rat, dog, pig or cat.

[0032] For mice, when the melanoma volume reached 62.5-256 mm 3 The tumor is then surgically removed to prevent the spread and metastasis of tumor cells.

[0033] Preferably, B16F10 melanoma cells were inoculated on the right side of the back skin of 9-week-old C57 mice; followed by injection of CD4 depleting antibodies on days 4 and 10. The tumor was surgically removed on day 12 to prevent the spread and metastasis of tumor cells.

[0034] Because melanocytes in mouse dorsal skin are located in hair follicles rather than in the epidermis, mouse tail skin was used for vitiligo analysis because it contains melanocytes localized in the epidermis similar to human skin. In tail skin, epidermal depigmentation became evident at 16 weeks after induction, mainly depending on the natural renewal rate of pigmented keratinocytes on the skin surface.

[0035] Compared with the tail skin of control mice that contained almost no CD8+ T cells, a large number of CD8+ T cells infiltrated into the skin epidermis of the tail on day 40 after induction of vitiligo, accompanied by a significant loss of melanocytes.

[0036] After vitiligo induction, the use of CD8-depleting antibodies resulted in a complete blockade of CD8+ T cell infiltration in the epidermis of the tail skin and rescued melanocyte loss. Thus, the vitiligo induction method effectively triggered endogenously activated CD8+ T cells to infiltrate the skin, leading to a loss of natural melanocytes located in the epidermis, similar to patients with autoimmune vitiligo.

[0037] Preferably, the present invention provides a transgenic animal overexpressing IFNGR1-JAK1-STAT1.

[0038] Previously developed vitiligo mouse models rely on adoptive transfer of melanocyte-specific CD8+T cells isolated from PMEL-specific TCR transgenic mice into Krt14-Kit1 transgenic mice (ref). Two transgenic alleles and transfer procedures are required to induce vitiligo in mouse skin, limiting the effective introduction of additional genetic modifications for in-depth functional studies in vivo. Other existing vitiligo mouse models are either waiting for the spontaneous vitiligo development of transgenic mice, but the development is slow and inefficient; or they utilize ectopically expressed melanocyte antigens, but these antigens require complex viral packaging systems or specialized gene gun delivery devices, thereby hindering general application. Most importantly, these vitiligo mouse models all rely on artificial stimulation that does not occur in patients. There are no detailed reports on the hallmarks of human vitiligo disease (i.e., loss of epidermal melanocytes).

[0039] In a fourth aspect, the present invention provides an animal skin, wherein the melanocytes in the skin are reduced.

[0040] Preferably, the skin is skin showing depigmentation. Preferably, the skin is tail skin or back skin of an animal. Preferably, the skin is skin on the back of a hand, back of a foot, chest, leg or arm of an animal.

[0041] Preferably, the animal is a mouse, rat, dog, pig or cat.

[0042] Preferably, skin acquisition is induced by inoculating the animal with melanoma cells, injecting CD4 depleting antibodies, and removing the melanoma.

[0043] The melanoma cells may be B16F10 melanoma cells.

[0044] Preferably, the tumor is removed surgically before it spreads and metastasizes.

[0045] Preferably, the animal is a mouse, rat, dog, pig or cat.

[0046] In mice, the melanoma tumor size reached 62.5 to 256 mm 3 The tumor is then surgically removed to prevent the spread and metastasis of tumor cells.

[0047] The volume calculation formula for melanoma is: 0.5×a×b×b, where a is the long axis length of the tumor and b is the short axis length of the tumor.

[0048] Preferably, B16F10 melanoma cells were inoculated on the right side of the back skin of 9-week-old C57 mice; followed by injection of CD4 depleting antibodies on days 4 and 10. The tumor was surgically removed on day 12 to prevent the spread and metastasis of tumor cells.

[0049] Because melanocytes in mouse dorsal skin are located in hair follicles rather than in the epidermis, mouse tail skin was used for vitiligo analysis because it contains melanocytes localized in the epidermis similar to human skin. In tail skin, epidermal depigmentation became evident at 16 weeks after induction, which was mainly determined by the natural renewal rate of pigmented keratinocytes on the skin surface.

[0050] In a fifth aspect, the present invention provides an isolated skin cell, wherein the cell is IFN-γ responsive.

[0051] Cells include, but are not limited to, keratinocytes, melanocytes, fibroblasts, endothelial cells, smooth muscle cells, and dendritic cells. Preferably, the cells are fibroblasts. Preferably, the cells are dermal fibroblasts. Preferably, the cells are dermal fibroblasts of mouse tail skin.

[0052] Local CD8+ T cell recruitment and activation in the skin requires skin-resident IFN-γ responsive cells. In particular, skin dermal fibroblasts are the main cell type that mediates local CD8+ T cell aggregation and activation. IFN-γ responsive fibroblasts are sufficient to mediate CD8+ T cell aggregation in vivo and in vitro through secreted chemokines such as CXCL9, CXCL10, and CCL19. Therefore, IFN-γ responsive fibroblasts from the dermis can be used as cells to study the mechanism of vitiligo and / or screen candidate drugs for vitiligo.

[0053] IFNGR1-JAK1-STAT1 signaling in fibroblasts is required to mediate local CD8+ T cell accumulation and activation. Injection of IFNGR1 KO fibroblasts into the tail skin of IFNGR1 KO mice did not result in local CD8 T cell accumulation after vitiligo induction. However, injection of WT fibroblasts alone into the dermis of the tail skin of IFNGR1 KO mice resulted in local CD8 T cell accumulation after vitiligo induction. Therefore, IFN-γ-responsive fibroblasts from the dermis can be used to result in local CD8 T cell accumulation and activation in patients with vitiligo. Fibroblasts with upregulated IFN-γ responsiveness from the dermis can be used to result in local CD8 T cell accumulation in patients with vitiligo.

[0054] In a sixth aspect, the present invention provides a method for screening candidate drugs for treating vitiligo using the animal model.

[0055] In a seventh aspect, the present invention provides a method for evaluating the therapeutic effect of vitiligo using the animal model.

[0056] In an eighth aspect, the present invention provides a method for prognostic assessment of vitiligo using the animal model.

[0057] In a ninth aspect, the present invention provides the use of the animal model in screening candidate drugs for treating vitiligo in animals (including humans).

[0058] In a tenth aspect, the present invention provides a method for distinguishing vitiligo disease states using single-cell RNA-seq analysis.

[0059] Compared with resting patients and healthy donors, skin from patients with progressive vitiligo contained more CD8+ cytotoxic T cells expressing high levels of IFNG. Correspondingly, melanocytes in skin from patients with progressive vitiligo upregulated genes involved in immune responses, especially those in response to IFN-γ.

[0060] Classical IFN-γ signaling activation involves the Janus kinases JAK1 and JAK2, which phosphorylate STAT1 and render it transcription factor active (ref). Therefore pSTAT1 staining was used as a readout for IFN-γ responsive cells.

[0061] Vitiligo skin is divided into three zones based on skin pigmentation and T cell infiltration: depigmented lesions, T cell infiltrates (TIRs), and adjacent pigmented perilesional zones. The density of pSTAT1+IFN-γ responsive cells was significantly higher in the TIRs compared with the lesions and perilesional zones. Importantly, the density of CD8+ T cells was positively correlated with the density of pSTAT1+ cells. The spatial co-distribution pattern of CD8+ T cells and IFN-γ responsive cells in patient skin was consistent with single-cell RNA-seq analysis. Therefore, regional responses to T cell-secreted IFN-γ are associated with a progressive disease state.

[0062] Recent advances in single-cell RNA sequencing have deepened our understanding of the cell as a functional unit, providing new insights based on gene expression profiles of hundreds to hundreds of thousands of individual cells and revealing new cell populations with distinct gene expression profiles that were previously hidden from gene expression analyses of whole cell populations. However, appropriate analysis and utilization of the large amounts of data generated by single-cell RNA sequencing experiments is challenging and requires an understanding of the experimental and computational paths taken between the preparation of input cells and the output of interpretable data. In this article, we discuss the basic principles of these new technologies, focusing on important concepts in the analysis of single-cell RNA sequencing data. In particular, we summarize methods for determining which single cells should be included in quality control measures for further examination, data normalization and scaling methods to overcome the relatively inefficient mRNA capture rate per cell, and clustering and visualization algorithms for dimensionality reduction of data to two-dimensional plots.

[0063] The present invention encompasses all combinations of specific embodiments described herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 Single-cell RNA-seq analysis of vitiligo patient skin reveals distinct disease states and associated signaling signatures. Scale bar, 50 μm. Data reflect mean ± SD of 5 vitiligo patients and 5 healthy donors at each stage. **p < 0.01, ns not significant by t-test.

[0065] Figure 2 Regional responses to IFN-γ are shown to be associated with CD8+ T cell infiltration in the skin of patients with vitiligo. Data reflect mean ± SD from 6 human vitiligo biopsies. ***p<0.001, ****p<0.0001 by t-test.

[0066] Figure 3 Figure 3 A mouse model of vitiligo reveals that responses to IFN-γ signaling are required for local CD8+ T cell accumulation and cytotoxic activity in the skin. Scale bar, 500 μm. Data reflect mean ± SD of at least 6 mice. ***p < 0.001, ****p < 0.0001 by t-test; ns not significant.

[0067] Figure 4 Figure 2. Local recruitment of IFN-γ-responsive skin dermal cells and activation of cytotoxic activity of CD8+ T cells. Data reflect mean ± SD of at least 6 independent host mice per group. ***p < 0.001 by paired t-test; ns not significant.

[0068] Figure 5 The IFNGR1-Jak1-Stat1 signaling axis in dermal fibroblasts is shown to be essential for local CD8+ T cell accumulation and activation in autoimmune skin. Data reflect mean ± SD from 3 independent mice. ***p < 0.001 by t-test; ns not significant. For shRNA knockdown experiments, data for each shRNA were pooled from at least 3 independent mice, using 2 shRNAs per gene.

[0069] Figure 6 IFN-γ responsive fibroblasts are shown to be sufficient to recruit CD8+ T cells through secreted cytokines. Data reflect the mean ± SD of 3 independent experiments, each with three technical replicates.

[0070] Figure 7 showed that differences in intrinsic IFN-γ responses of anatomically distinct human fibroblasts correlated with regional disease variation.

[0071] Figure 8 Displays FACS profiles and skin sample information of vitiligo patients and healthy donors.

[0072] Fig. 9 Shown are single-cell RNA-seq analyses of different cell populations from vitiligo patients and healthy donors.

[0073] Fig.10 Analysis of a mouse model of vitiligo is shown.

[0074] Fig.11 Analysis of transplanted and vitiligo-induced mice is shown.

[0075] Fig.12 Analysis of in vivo dermal fibroblast-specific knockout and knockdown experiments is shown.

[0076] Fig.13 Shown are analysis of mouse fibroblasts after intradermal fibroblast injection in vivo, IFN-γ treatment of fibroblasts in vitro, and vitiligo-induced changes in fibroblast expression in vivo.

[0077] Fig.14 Human skin dermal fibroblasts were shown to exhibit intrinsic regional differences in response to IFN-γ. DETAILED DESCRIPTION

[0078] The descriptions of specific embodiments and examples are provided by way of illustration and not limitation. Those skilled in the art will readily recognize a variety of non-critical parameters that can be changed or modified to produce essentially similar results.

[0079] Experimental model and object details

[0080] Human specimens

[0081] Seven male and three female patients with pathological diagnosis of vitiligo participated in this study. Their ages ranged from 6 to 55 years old, with a median age of 24 years. Of these patients, 4 were diagnosed with active stage, while the other 6 were of uncertain stage. All patients were newly diagnosed with vitiligo and none had received chemotherapy or UVB treatment. A 10mm smear was taken from the junction between the depigmented lesion area and the pigmented perilesional area of ​​each patient. 2 Fusiform biopsies. Biopsies of the same size were also obtained from 5 healthy donors using other surgical procedures as controls. The clinical characteristics of these patients and healthy donors are summarized in Table 1. This study was approved by the Ethics Committee of Beijing Hospital and the Ethics Committee of Beijing Institute of Life Sciences. All patients in this study provided written informed consent for sample collection and data analysis.

[0082] Table 1

[0083]

[0084] Mouse

[0085] Mice were housed and maintained in the NIBS specific pathogen-free facility in accordance with the Guide for the Care and Use of Laboratory Animals of the Institute of Life Sciences (NIBS). All mice used in the experiments were housed in social housing under a 12-h light / dark cycle with free access to food and water.

[0086] C57BL / 6 mice were purchased from Charles River Laboratories. IFNGR1 KO mice (Stock No. 003288) were kindly provided by Dr. Shao Feng. OT-1 mice (Stock No. 003831) were kindly provided by Dr. Liang Chen. Pdgfra-CreER (Stock No. 018280) mice, IFNGR1 flox (Stock No. 025394) mice, and Rosa-stop-mTmG (Stock No. 007676) mice were from Jackson Laboratory.

[0087] Cell culture

[0088] All cell lines were cultured in a cell culture incubator at 37°C and 5% CO2. B16F10 cells (ATCC, CRL-6475) were maintained in DMEM (GIBCO) medium supplemented with 10% (v / v) FBS (GIBCO) and 1% (v / v) Pen-strep (Invitrigen). Primary mouse fibroblasts were maintained in DMEM supplemented with 10% (v / v) FBS and 1% (v / v) Pen-strep. 293FT cells (Thermo Fisher Scientific, Cat# R70007) used for virus packaging were maintained in DMEM medium supplemented with 10% (v / v) FBS, 1% (v / v) Pen-strep, 1% (v / v) L-glut (Lonza), 1% (v / v) 100 mM sodium pyruvate (Lonza), 1% (v / v) 7.5% sodium bicarbonate (Lonza) and 500 mg / mL G418 (Lonza).

[0089] Method details

[0090] Single cell collection

[0091] Single cell collection begins within 3 hours after skin biopsy collection. Carefully remove subcutaneous fat. Place the skin tissue dermis face down in 6mL 2.4U / mL disease solution and keep it at 37°C, 80rpm for 70 minutes. Carefully separate the epidermis with hair follicles from the dermis. Then place the separated epidermis inside down in 6mL 0.25% trypsin and incubate at 37°C for 10 minutes. Then add 6mL 5% FBS culture medium to neutralize trypsin. Use a 1mL pipette to repeatedly aspirate and blow 10 times to obtain an epidermal single cell suspension, which is then filtered with a filter (70mm, then 40mm). The cells are then stained with APC-CD45 antibody (1:300) and PE-CD117 antibody (1:300) for 15 minutes to detect immune cells and melanocytes, and then washed. The dermis was placed in 10 mL 2 mg / mL collagenase (Sigma-Aldrich, C2674) and incubated at 37 ° C, 80 rpm for 1 hour. Then 10 ml 5% FBS medium was added. A single cell suspension was obtained by repeated aspiration and blowing 10 times with a pipette. The cells were then filtered with a filter (70 mm, then 40 mm). The cells were then stained with APC-CD45 (1:300) for 15 minutes to detect immune cells and then washed.

[0092] Based on FACS analysis, single cells of different subtypes, including immune cells from the epidermis and dermis (CD45+), melanocytes from the dermis (CD117+), other dermal niches from the dermis (CD45-), and keratinocytes from the epidermis (CD45-, CD117-), were sorted into collection buffer containing 0.04% BSA dissolved in PBS to minimize cell loss and aggregation. All immune cells, melanocytes, dermal cells, and the same number of keratinocytes as all three were put together for the next step of single cell analysis.

[0093] Induction of vitiligo mouse model

[0094] 2 × 10 5B16F10 melanoma cells were cultured in 40% sham-treated mice. Then, on days 4 and 10, the mice were treated intraperitoneally with anti-CD4 antibodies purchased from BioXcell (West Lebanon, NH, USA) as described previously. Only mice that developed primary tumors were used for further analysis. On day 12, the primary tumor was surgically removed. No spontaneous tumor metastasis was observed using this B16F10 cell line, and mice in which primary tumors recurred after surgery were not used for further studies. White hair on the back skin appeared 2 weeks after surgery, starting from the right side where the primary tumor had been removed and developing to the entire back within 10 months. About 60% of the mice developed vitiligo within 30 days after induction, and the appearance of the remaining 40% of the mice was not affected.

[0095] Administration of CD4 and CD8 depleting antibodies

[0096] The CD4 antibody used for cell depletion was purchased from BioXcell (West Lebanon, NH, USA). The CD8 antibody used for cell depletion was a gift from Jianhua Sui's laboratory at the Beijing Institute of Life Sciences.

[0097] Anti-CD4 antibody (GK1.5) and anti-CD8 antibody (2.43) were administered intraperitoneally at a dose of 10 μg / g mouse body weight. Flow cytometry confirmed more than 99% depletion of the target population. CD4 antibody was administered on days 4 and 10 after B16F10 inoculation. CD8 antibody was administered every four days after tumor resection.

[0098] Tail to dorsal transplant

[0099] For tail-to-dorsal skin grafts, full-thickness tail skin from 4- to 6-week-old females was removed, flattened, and cut into 1-cm squares. Only the one-third near the base of the tail was used as donor skin. The donor skin piece was placed on the back of anesthetized 8-week-old female recipient mice of the indicated genotype, with each recipient receiving a WT and KO graft. The grafts were secured with sterile gauze and elastic bandages, which were removed after healing (8-10 days). Gender-mismatched donor skin will result in skin necrosis within 20 days. The skin near the tip of the tail is very narrow and is often lost during wound healing.

[0100] Gene tagging and conditional knockout

[0101] For CreER activation, tamoxifen was dissolved in sunflower oil containing 10% ethanol. For labeling, pdgfra-CreER;mTmG mice received a single intraperitoneal (ip) injection of 100 μL of a 10 mg / mL tamoxifen solution. Two days after injection, samples were collected for analysis. For conditional knockout, pdgfra-CreER;ifngr1 fl / fl mice received intraperitoneal (ip) injections of 200 μL of a 10 mg / mL tamoxifen solution for 7 consecutive days at 6 weeks. Knockout efficiency was verified by cell type-specific qPCR.

[0102] Immunofluorescence staining

[0103] For section staining, tissues were embedded in OCT compound, frozen, cryosectioned (20-30 μm), and fixed in 4% paraformaldehyde in PBS for 10 min. Sections were washed overnight in PBS, permeabilized in 0.3% H2O2 in methanol for 20 min at -20°C, and blocked in 2% normal donkey serum, 1% BSA, and 0.3% Triton in PBS for 1 h at room temperature. The following antibodies were used: anti-hCD45 antibody, anti-human Tyr antibody, anti-human KRT14 antibody, anti-human DCT antibody, anti-human CD8 antibody, anti-human pSTAT1 antibody, anti-human pdgfra antibody, anti-human CD31 antibody, anti-human aSMA antibody, anti-human CD11c antibody, anti-human Langerin antibody, anti-human CD3e antibody, anti-human CD8a antibody, anti-mouse DCT antibody, anti-mouse pdgfra antibody, anti-mouse CD45 antibody, anti-mouse CD31 antibody, anti-mouse aSMA antibody, anti-mouse pSTAT1 antibody, and anti-mouse KRT14 antibody. The signals of human pSTAT1, human CD8a, human pdgfra, human CD11c, human Langerin, human CD3e, mouse pdgfra, and mouse pSTAT1 were amplified using the ABC kit and the TSA kit.

[0104] For whole-mount staining, the entire tail skin was harvested and flattened. The tail skin was cut into 7-8 square pieces, and the middle 3-4 pieces were placed in a 20mM EDTA solution and incubated at 37°C, 80rpm for 1.5 hours. The epidermis was quickly removed from the dermis in the anterior-posterior direction with fine-tipped tweezers. This will keep most of the hair follicles in the dermis. A small number of hair follicles in the epidermis were removed with tweezers. The epidermis was flattened and fixed in PBS containing 4% paraformaldehyde for 10 minutes. The whole skin was washed in PBS overnight, then permeabilized in a methanol solution containing 0.3% H2O2 at -20°C for 20 minutes, and then blocked in a PBS solution containing 2% normal donkey serum, 1% BSA and 0.3% Triton for 1 hour at room temperature. The following antibodies were used: anti-mouse CD8a antibody, anti-mouse DCT antibody.

[0105] Imaging and image processing

[0106] Tissue and section samples were imaged on a NikonA1-R confocal microscope.

[0107] Images for representative images were acquired using a 20×0.75 objective and images for quantitative images were acquired using a 10×0.5 objective. Z stacks were acquired at a resolution of 1024×1024 or 512×512. Microscopy data were analyzed using Imaris software with a 3D visualization module (Bitplane). RBG images were collected in Adobe Photoshop CS3 and images were labeled using Adobe Illustrator CS6.

[0108] To quantify the CD8+ T cell infiltration area and melanocyte retention area, those areas containing T cell and melanocyte fluorescence signals were converted into digital information and then quantified by the “Surface” function in Imaris software.

[0109] In order to describe the distribution and density of CD8+T cells and melanocytes, the fluorescent signals of T cells and melanocytes were converted into digital information by the "spot" function in Imaris software. The position coordinates of each point were analyzed using the _smoothscatter_ package in R software to present the distribution and density of cells. In addition, the DBSCN package in R software was used to analyze the position information to present the size of T cell clones and the number of T cells in each clone.

[0110] To quantify the number of RFP+ fibroblasts, melanocytes, and CD8+ T cells in each size, the fluorescence signals of these cells were converted into digital information by the “dot” function in Imaris software. The number of cells in each size was quantified in Python.

[0111] Fibroblast transplantation assay

[0112] For fibroblast transplantation into the tail skin of adult mice, the entire dorsal skin of newborn female mice was harvested, flattened and placed face down on 6 mL of 2.4 U / ml Dispase solution and incubated at 37°C, 80 rpm for 60 minutes. The epidermis was removed from the dermis with force. This will retain most of the hair follicles in the epidermal part. The dermal part was placed in 10 mL of 2 mg / mL collagenase (Sigma-Aldrich, C2674) and incubated at 37°C, 80 rpm for 1 hour. Then 10 ml of culture medium containing 5% FBS was added. A single cell suspension was obtained by repeated aspiration and blowing 10 times with a Finnpipette pipette. The cells were then filtered with a filter (70 mm, then 40 mm).

[0113] Fibroblasts were then expanded and cultured in DMEM (GIBCO) medium supplemented with 10% (v / v) FBS (GIBCO), 1% (v / v) Pen-strep / L-glut (Lonza), and 1% antibiotics and antimycotics.

[0114] These primary neonatal mouse fibroblasts were then infected with lentivirus containing LV-H2BRFP. RFP-labeled fibroblasts were then injected into the dermis of the tail skin of 8-week-old adult female WT or IFNGR1 KO mice. A total of 3 × 10 6 10 fibroblasts 5 The concentration of / μL was intradermally injected into 3 sites / tail (10 μL per site, 1 cm between sites). The mice were kept for 3 days before vitiligo induction.

[0115] Fluorescence activated cell sorting (FACS)

[0116] The entire tail skin was harvested and flattened to isolate immune cells and melanocytes from the epidermis of the tail skin. The tail skin was cut into 7-8 square pieces, and the dermis was placed face down on 6mL of 2.4U / ml Dispase solution and incubated at 37°C, 80rpm for 45 minutes. Use fine-tipped tweezers to quickly remove the epidermis from the dermis in the front-to-back direction. This will keep most of the hair follicles in the dermis. Use tweezers to remove a small number of hair follicles in the epidermis. Float the removed epidermis with the inside facing down on 6mL of TripLE solution and incubate at 37°C for 10 minutes. Then add 6ml of culture medium containing 5% FBS.

[0117] Single cell suspension was obtained by repeated aspiration and blowing 10 times with a Finnpipette pipette. The cells were then filtered with a filter (70 mm, then 40 mm). The cells were then stained with Alex647-CD8 antibody (1:300) and FITC-CD45 (1:300) for 15 minutes to detect CD8+ T cells, or Alex647-CD117 antibody (1:300) to detect melanocytes, and then washed.

[0118] The entire tail skin was harvested and flattened to isolate fibroblasts and endothelial cells from the tail skin. The tail skin was cut into 7 to 8 square pieces, and the dermis was placed face down on 6 mL of 2.4 U / ml Dispase solution and incubated at 37°C, 80 rpm for 60 minutes. The epidermis was removed from the dermis with fine-tipped tweezers in the anterior-posterior direction. This will keep most of the hair follicles in the epidermis. A small number of hair follicles in the dermis were pulled out with tweezers. The dermis was placed in 10 mL of 2 mg / mL collagenase (Sigma-Aldrich, C2674) and incubated at 37°C, 80 rpm for 1 hour. Then 10 ml of culture medium containing 5% FBS was added. A single cell suspension was obtained by repeated aspiration and blowing 10 times with a Finnpipette. The cells were then filtered with a filter (70 mm, then 40 mm). The cells were then stained with Alex647-CD31 antibody (1:300) FITC-CD45 (1:300) for 15 min and then washed.

[0119] DAPI was used to exclude dead cells. Cell analysis and separation were performed on a BD AriaII sorter (BD bioscience) equipped with FACSDiva software. FACS analysis was performed using an LSII FACS analyzer (BD bioscience) and then analyzed using FlowJo software (FlowJo LLC).

[0120] Lentiviral vector construction, production, and injection

[0121] Lentivirus expressing short hairpin RNA was injected into the dermis of the tail skin of P1 using an insulin syringe. Vitiligo induction started at week 9, and whole-mount staining was performed 33 days later.

[0122] shRNA lentiviral constructs were obtained from the RNAi Consortium (TRC) mouse lentiviral library for knockdown of IFNGR1, JAK1, and STAT1. The shRNAs were then subcloned into LV-RFP. The sequences of the individual shRNAs used in the experiments are listed above. To knockdown the relevant genes in the dermis, high titer lentivirus was generated as described previously. 10 μL of high titer lentivirus (>5×10 8 cfg / mL) was injected intradermally into the base of the tail. Vitiligo induction started at 9 weeks after birth. On day 26 after vitiligo induction, samples were collected for whole-body staining.

[0123] In vitro Transwell T cell migration assay

[0124] Transwell migration of lymphocytes was performed using mature CTLs and concentrated fibroblast-conditioned medium. Briefly, splenocytes isolated from OT-1 mice were stimulated with OVA257-264 for 3 days in the presence of 10 ng / mL IL2. The cells were centrifuged and cultured for another day in fresh medium containing 10 ng / mL IL2, after which most of the cells in the culture were CTLs. To measure the cytotoxicity of CD8+ T cells, OVA peptides expressing B16F10 were mixed in killing medium at a ratio of 1:1. After 6 hours, cytotoxic efficiency was confirmed by B16F10 cell survival.

[0125] In order to obtain fibroblast conditioned medium, primary fibroblasts from the back skin of newborn mice were treated with DMEM containing 1000U / mL IFN-γ at 37°C for 6 hours. The concentration and duration of IFN-γ were previously verified to obtain conditions with significant IFN-γ responses. The culture medium was then concentrated (1×, 5×, 10×, 25×) for chemical attractants. Concentrated DMEM or DMEM containing IFN-γ was used as a control.

[0126] RNA isolation and real-time PCR

[0127] Total RNA was isolated from FACS-sorted cells using Trizol and then extracted using the Direct-Zol RNA mini-prep kit (Zymo research). Equal amounts of RNA were reverse transcribed with Oligo-dT (Vazyme, R222-01) and used to synthesize cDNA. Expression levels were normalized to that of PPIB. Real-time PCR was performed using a CFX96TM Real-Time System (Bio-RAD) with Power SYBRR Green PCR Master Mix (Life Technologies). All primer pairs were designed to have the same cycling conditions: 95°C for 10 min, initial denaturation; 95°C for 10 s denaturation, 62°C for 30 s annealing, and 65°C for 10 s extension for 40 cycles. Primers were designed to generate products spanning exon-intron boundaries in each target gene.

[0128] RNA-Seq

[0129] RNA from FACS purified cells was submitted to Novogene for quantification, RNA-seq library preparation and sequencing. The library was sequenced on the Illumina HiSeq platform using a Pair-End 150 bp sequencing strategy.

[0130] IFN-γ treatment of fibroblasts in vitro

[0131] Skin biopsy samples were collected from different body parts of aborted female fetuses at 20 or 23 weeks. After being treated with 2.4U / mL dispase at 37°C, 80rpm for 1 hour, the epidermis and dermis were mechanically separated. The dermis was further digested with 10mg / mL collagenase at 37°C, 80rpm for 1 hour. The fibroblasts were then expanded and cultured in DMEM supplemented with 10% FBS, 1% P / S, 1% antibiotics and antifungals. The 3rd or 4th generation fibroblasts were treated with 1U / mL recombinant IFN-γ, and then RNA sequencing was performed. For each region, the average FPKM value of 2 independent samples was used to estimate the gene expression level. The genes in the heat map were selected based on their differential expression (Log2 fold change>1, and p<0.01) in at least 1 of the 8 positions after IFN-γ treatment.

[0132] Quantification and statistical analysis

[0133] Construction and sequencing of single-cell RNA libraries

[0134] According to the manufacturer's instructions, single-cell cDNA libraries were prepared using the Chromium Single Cell 3'Library and Gel Bead kit v2. In short, the cell suspension in the chip was loaded onto the Chromium controller (Chromium Controller, 10×Genomics, Pleasanton, CA) to generate single-cell GEMs (gel beads in emulsion). The scRNA-seq library was then prepared using the Chromium Single Cell 3'Gel Bead and Library Kit (P / N#120236, 120237, 120262; 10×Genomics). Qualitative analysis of the DNA library was performed by an Agilent 2100 bioanalyzer. The concentration of the DNA library was measured by Qubit (Invitrogen). The library was sequenced on an Illumina NextSeq 500 (2×150 double-end reads).

[0135] Single-cell seq data processing

[0136] The raw sequencing reads were aligned and quantified by Cell Ranger (V1.3.1) software obtained from 10×Genomics (https: / / support.10xgenomics.com / single-cell-geneexpression / software / down-loads). The human hg38 assembly reference was used for analysis. The raw count matrix data were imported into R using the Seurat (V2.3.2) package for further data analysis. For all 15 samples, we initially set a first filter of min.cells 3 and min.genes 200 for each sample. Cells with more than 5000 genes or more than 25,000 UMIs were removed. We retained cells with mitochondrial gene expression less than 1%. The raw counts were normalized by 10,000 times and log-transformed to obtain log (T+1) values. Variant genes were identified by fitting a mean-variance relationship and satisfying the criteria of 0.0125<mean<3 for non-zero values ​​and standard deviation>0.4.

[0137] Cell type classification using t-SNE

[0138] Unsupervised cell clustering was performed using Seurat. Dimensionality reduction was performed using principal component analysis. The top 20 PCs were selected based on the PCA elbow plot and used for clustering with a resolution parameter of 0.1. Cell clusters were visualized using t-SNE plots with all significant principal components as input information. We integrated all sample data using canonical correlation analysis (CCA). A shared nearest neighbor graph was constructed on the inter-cell distance matrix from the top 30 aligned canonical correlation vectors. The shared nearest neighbor graphs at different resolutions were used as input information for the smart local move algorithm to obtain cell clusters and visualized using t-SNE. Differentially expressed genes were identified by Wilcoxon rank sum test with parameters min.pct = 0.25, thresh.use = 1, test.use = "wilcox". The final number of clusters between resolutions was chosen based on prior knowledge and consistency within different resolutions, which included all major cell types in the skin, resulting in 8 different clusters.

[0139] Delineation of differentially expressed genes within each cell cluster

[0140] To identify differentially expressed genes in the two melanocyte subtypes, we used Seurat's FindMarkers function to normalize the gene expression data using the following parameters: min.pct>0.25, thresh.use=0.25. The highly expressed gene cluster in C1 melanocytes was identified as C1 signature genes (fold change>2, p-value<0.01). GO analysis was performed on C1 signature genes using the GO web service (http: / / geneontology.org). To compare gene expression between the four T cell clusters, a threshold fold change>0.25, p-value<0.01 was used to select differentially expressed genes. The gene expression levels shown in the various graphs of the present disclosure were drawn by software packages in R.

[0141] RNA-seq alignment, analysis, and visualization

[0142] For RNA-seq analysis of mouse tail fibroblasts, raw transcriptome sequence data were mapped to the mouse genome (GRCm38 / mm10) using TopHat (v2.0.13) with default settings to generate a reference-guided transcript assembly. Expression levels for each sample were normalized to fragments per kilobase of transcript per million mapped reads (FPKM) using Cufflinks (v2.2.1).

[0143] Using Cuffdiff, quantify the changes in gene expression between control WT fibroblasts, vitiligo WT fibroblasts and vitiligo IFNGR1 KO fibroblasts. Select genes with significantly upregulated expression levels (p value < 0.01, vitiligo WT / original WT fold change > 1.5, vitiligo WT / vitiligo KO > 1.5) for further analysis. Gene ontology (GO) analysis was performed on upregulated genes using the GO web service (http: / / geneontology.org). The differentially expressed genes were displayed by the "ggplot2" package in R software.

[0144] For human fibroblast RNA-seq analysis treated with IFN-γ in vitro, the original transcriptome sequence data was mapped to the human genome (hg38) using STAR (v2.6.1a) with default settings to generate reference-guided transcript assemblies. Gene reads were counted using FeatureCount (v2.2.1), and the expression level of each sample was normalized to fragments per thousand bases of transcript per million mapped reads (FPKM). Differential expression analysis was performed using the DESeq2 software package. Genes with significantly increased expression levels (p value < 0.01, multiple change > 2) were selected for further analysis. Heat maps showing differentially expressed genes were displayed by the "pheatmap" and "ggrepel" software packages in R software.

[0145] Example

[0146] Example 1 Single-cell RNA-seq analysis of vitiligo patient skin reveals different disease states and associated signaling pathways.

[0147] We first analyzed the patient's skin using immunofluorescence staining to investigate the cellular and molecular mechanisms that lead to the development of localized autoimmune vitiligo in the patient. Biopsy samples of depigmented lesional skin and perilesional skin that remained pigmented were obtained from patients with vitiligo who had never received treatment ( Figure 1 A). Figure 1A shows representative photographs and immunofluorescence staining images of skin from patients with vitiligo. Photographs of lesional skin from patients with vitiligo show areas of depigmented skin. In the immunofluorescence staining images of skin sections, tyrosinase (Tyr) staining marks melanocytes located in the epidermal basal layer, CD45 staining marks infiltrating immune cells, and keratin 14 (K14) staining marks keratinocytes. The enlarged images show the lesion area, perilesional area, and immune cell infiltration area. Note that melanocytes are present in the perilesional area but are lost in the lesional area. Infiltrating immune cells are enriched in the connecting areas between them. Tyr+ melanocytes are evenly distributed in the epidermal basal layer of the lesion area but are lost in the lesion area. Consistent with vitiligo being an autoimmune disease, we detected a large number of CD45+ immune cells infiltrating the patient's skin. The depigmented lesion area contained slightly more CD45+ immune cells than the perilesional area, but most of the infiltrating immune cells were concentrated at the junction of the lesion area and the perilesional area. This interesting distribution pattern of CD45+ immune cells suggests that some recruitment mechanism is coordinating the local accumulation of immune cells, thereby driving the expansion of the depigmented area in the patient's skin. The local accumulation pattern of CD45+ immune cells was found mainly at the junction of the lesion area and the perilesional area.

[0148] We used single-cell RNA-seq to profile all cell types present in the patient's skin, and as a control we used cells isolated from skin biopsies of healthy donors ( Figure 2 B and Figure 8 ), and tested the involvement of certain skin-resident cells in mediating the recruitment of immune cells.

[0149] Figure 2 B shows quantification of CD8+ T cell density in T cell infiltration zone (TIR), perilesional zone and lesional zone of vitiligo patient skin samples. The lesional zone is defined as the lack of melanocytes and skin pigmentation. The perilesional zone is defined as the adjacent skin area that still has melanocytes and has not been infiltrated by T cells. The T cell infiltration zone is defined as the area enriched with T cells. Figure 8FACS profiles and skin sample information of vitiligo patients and healthy donors are shown: (A) Schematic diagram of the digestion process of skin biopsy samples from vitiligo patients and healthy donors for single cell-seq. The epidermis and dermis in the skin biopsy samples were enzymatically dissociated by dispase treatment. The skin dermis was then digested with collagenase and the epidermis was digested with trypsin to obtain a single cell suspension. After immunostaining, different types of cells were collected using FACS; (B-C) Representative FACS profiles of different cell populations in the epidermis (B) and dermis (C). In the epidermis, immune cells and melanocytes were enriched using CD45 and c-Kit; CD45- cells and c-Kit- cells were mainly keratinocytes. In the dermis, CD45 was used to distinguish immune cells from dermal cells; (D) Clinical information of vitiligo patients and healthy donors used in this study, as well as the number of cells sequenced in single cell analysis.

[0150] Depigmented lesional skin and border areas of pigmented perilesional skin from never-treated vitiligo patients or freshly obtained skin biopsy samples were harvested from healthy donors and enzymatically dissociated to obtain single-cell suspensions. After immunostaining, different cell types were separated using FACS. All isolated immune cells (CD45+, c-Kit-) and melanocytes (CD45-, c-Kit+) were combined with equal numbers of niche cells (CD45-, c-Kit-) (including keratinocytes and mesenchymal cells) for single-cell RNA-seq analysis. Unsupervised clustering of more than 50,000 cells from 10 vitiligo patients and 5 healthy donors revealed 8 cell clusters corresponding to 8 different cell types, which were defined by signature gene expression ( Figure 1 B and Fig. 9 A): melanocytes (DCT, TYRP1, and PMEL); T cells (TRAC, CD3D, and TRBC2); fibroblasts (COL1A1, DCN, and SFRP2); endothelial cells (PECAM, CLEC14A, ECSCR); smooth muscle cells (TAGLN, ACTA2, and NR2F2); keratinocytes (KRT14, KRT1, and KRT10); Langerhans cells (FCGBP, CD207, and CD1A); and mononuclear phagocytes, including macrophages and dendritic cells (LYZ, CLEC10A, and CD1C) ( Fig. 9 B~9R).

[0151] Figure 1B shows a schematic and overview of single-cell RNA-seq analysis of skin samples from vitiligo patients and healthy donors. Skin biopsy samples from vitiligo patients and healthy donors were enzymatically dissociated and then digested into single-cell suspensions for FACS sorting. Immune cells (CD45+, c-Kit-) and melanocytes (CD45-, c-Kit+) were enriched using CD45 and c-Kit. All immune cells, melanocytes, and an equal number of niche cells (CD45-, c-Kit-) (including keratinocytes and mesenchymal cells) were pooled together for single-cell RNA-seq. t-SNE projections of more than 50,000 cells from 10 vitiligo patients and 5 healthy donors showed 8 major cell type clusters with different expression profiles. Each dot represents a cell and is colored according to the annotation.

[0152] Fig. 9 Single cell RNA-seq analysis of different cell populations from vitiligo patients and healthy donors is shown:

[0153] (A) Heat map analysis of characteristic genes in 8 major cell types in more than 50,000 cells from 10 vitiligo patients and 5 healthy donors determined by t-SNE projection. Unsupervised clustering by spectral clustering method divided individual cells into 8 subclusters. Three criteria were used to determine the differentially expressed genes in each cluster: 1. The Log2 fold change of gene expression level in a specific cluster compared with other clusters was greater than 1; 2. The p-value of the differentially expressed genes in a specific cluster compared with other clusters was less than 1E10-50; 3. More than 50% of the cells in the cluster expressed the differentially expressed genes that were determined to be unique to the cluster. The 10 selective characteristic genes with the smallest p-values ​​were annotated next to them. Cell types are defined by the characteristic genes in each cluster and are marked at the top. Baseline color and Figure 1 The colors used in the t-SNE projections of different cell types in B are consistent;

[0154] (B) Dot plot analysis of the top 4 signature genes for each cell type. Genes enriched in each cell type are annotated next to them. Predicted cell types are labeled at the top: melanocytes, T cells, fibroblasts, endothelial cells, smooth muscle cells, keratinocytes, Langerhans cells, and mononuclear phagocytes. The size of each circle depicts the percentage of cells in which the marker was detected in that subtype, and its color depicts the normalized transcript level among expressing cells;

[0155] (C-J) Violin plot analysis showing the expression profiles of three characteristic genes in each cluster: (C) melanocytes, (D) T cells, (E) fibroblasts, (F) endothelial cells, (G) smooth muscle cells, (H) keratinocytes, (I) Langerhans cells, and (J) mononuclear phagocytes. The x-axis represents the different cell types marked at the bottom. The y-axis represents the log-transformed normalized gene expression level. The color of each plot describes the predicted cell type information.

[0156] (K-R) Feature plot analysis showing the expression patterns of two signature genes for each cluster in all cells: (K) melanocytes, (L) T cells, (M) fibroblasts, (N) endothelial cells, (O) smooth muscle cells, (P) keratinocytes, (Q) Langerhans cells, and (R) mononuclear phagocytes. Colors depict log-transformed normalized gene expression levels.

[0157] Of all these identified cell types in the skin, we first analyzed the transcriptional profile of melanocytes, the target cells of the autoimmune attack in vitiligo skin ( Figure 1 C). When all melanocytes from vitiligo patients and healthy donors were analyzed together, t-SNE clustering revealed 2 subclusters of melanocytes (called M1 and M2). Interestingly, cells in the M1 cluster were almost exclusively derived from skin of vitiligo patients. We then analyzed genes enriched in melanocytes in the M1 cluster (>2-fold, p<0.01) compared to all other melanocytes from vitiligo and healthy skin in the M2 cluster ( Figure 1 D). Functional annotation analysis revealed that most of these genes are involved in immune response, especially the response to interferon gamma (IFN-γ). Since a large number of melanocytes from the skin of vitiligo patients do belong to the M1 cluster, there is obvious heterogeneity between melanocytes from different patients. We used single-cell RNA-seq data to quantify the melanocyte composition of each patient and healthy donors ( Figure 1 E). Melanocytes from patients 1 to 5 were composed mainly of cells with strong immune responses in the M1 cluster. Patients 6 to 10 had a melanocyte composition similar to that of healthy donors. Based on the expression profile and composition of the melanocytes, we tentatively classified patients 1 to 5 as progressive and patients 6 to 10 as quiescent. Clinical information about the disease state is usually based on the patient's feedback on whether the depigmented area has gradually expanded before presentation. Therefore, a clear classification of the progressive or quiescent disease state usually requires immunohistochemical analysis of a skin biopsy specimen.

[0158] Figure 1C shows the t-SNE projection of the melanocyte dataset from vitiligo patients and healthy donors. Unsupervised clustering was performed by spectral clustering to divide the melanocytes into 2 subclusters. Each subcluster is colored according to the annotation: M1 (red) and M2 (blue). Melanocytes with different expression patterns in the M1 cluster are mainly derived from the skin of vitiligo patients.

[0159] Figure 1 D shows the volcano plot and GO analysis of genes enriched in melanocytes of the M1 cluster compared with all other melanocytes in vitiligo and healthy skin. Red dots in the volcano plot indicate genes that are upregulated >2-fold in the M1 cluster (p<0.01). Functional annotation analysis showed that most of these genes are involved in immune response, especially the response to interferon gamma (IFN-γ).

[0160] Figure 1 E shows the melanocyte composition in each patient and healthy donor. The color annotation corresponds to the 2 melanocyte clusters from (C). The melanocytes from patients 1 to 5 are mainly composed of cells with strong immune responses in the M1 cluster. Patients 6 to 10 have a melanocyte composition similar to that of healthy donors. Therefore, patients 1 to 5 are classified as progressive states and patients 6 to 10 are classified as quiescent states.

[0161] To validate our disease state classification based on single-cell RNA-seq analysis of melanocytes, we next analyzed the transcriptional profiles of T cells, the major cell type responsible for the autoimmune attack on melanocytes in vitiligo skin ( Figure 1 F). We identified four major T cell subsets from the single-cell data: CD8+ cytotoxic T cells specifically expressed cytotoxicity-related marker genes such as GZMA, GZMB, and CCL5; the CD8+ resident T cell cluster was characterized by high expression of TRGC, TRDC, and ZNF683 genes; the CD4+ effector T cell cluster was characterized by specific expression of CD4 and CD40LG; and the CD4+ regulatory T cell cluster showed expression of CTLA4, Foxp3, and TIGIT ( Figure 1 G). It has been shown that CD8+ cytotoxic T cells are able to eliminate melanocytes in an experimental mouse model of vitiligo (ref). So next we compared the percentage of CD8+ cytotoxic T cells in total immune cells of patients and healthy donors. In patients that we provisionally classified as progressive, there was a statistically significant increase in CD8+ cytotoxic T cells ( Figure 1 H). In addition, CD8+ cytotoxic T cells from patients with progressive disease expressed significantly more IFNG compared with patients in the resting state and healthy donors ( Figure 1I). These results are consistent with our transcriptional profiling of melanocytes and disease state classification.

[0162] Figure 1 F shows the t-SNE projection of T cells from vitiligo patients and healthy donors. Unsupervised clustering by spectral clustering method divides T cells into 4 subclusters. Each cluster is colored and annotated according to the characteristic gene expression pattern: CD8 cytotoxic T cells (red), CD8 resident T cells (green), CD4 effector T cells (blue), and CD4 regulatory T cells (purple).

[0163] Figure 1 G shows a volcano plot of differentially expressed genes in each T cell subcluster. Colored dots in the volcano plot represent genes that are upregulated >2-fold in all four clusters (p<0.01).

[0164] Figure 1 H shows the percentage of CD8 cytotoxic T cells among total immune cells in patients and healthy donors. Patients in the progressive state (P) contained statistically more CD8 cytotoxic T cells compared to patients in the resting state (Q) and healthy donors (HD).

[0165] Figure 1 I shows the average transcript levels of IFNG in CD8 cytotoxic T cells from patients and healthy donors. CD8 cytotoxic T cells from patients in the progressive state (P) expressed statistically more IFNG compared to patients in the resting state (Q) and healthy donors (HD).

[0166] Our single-cell RNA-seq analysis of patient skin not only helped us distinguish between different disease states, but also revealed major relevant signaling pathways in the relevant cell types. Compared with patients in the resting state and healthy donors, progressive vitiligo skin contained more CD8+ cytotoxic T cells expressing large amounts of IFNG. Correspondingly, melanocytes in progressive vitiligo skin upregulated genes involved in immune responses, especially the response to IFN-γ.

[0167] Example 2 Regional responses to T cell-secreted IFN-γ are associated with progressive disease states.

[0168] To validate our single-cell RNA-seq results obtained in Example 2, we used immunofluorescence staining to examine the spatial distribution patterns of CD8+ T cells and IFN-γ responsive cells in the patient's skin. Classical IFN-γ signaling activation involves Janus kinases JAK1 and JAK2, which phosphorylate STAT1 and activate its transcription factor activity (ref). Therefore, we used pSTAT1 staining as a readout of IFN-γ responsive cells. We tested skin biopsy samples from 6 patients with vitiligo who were initially clinically diagnosed with a progressive state and found that 4 of them were enriched with CD8+ T cells at the junction of the lesion area and the perilesional area. Figure 2 In the representative image shown in A, we can detect CD8+ T cells infiltrating vitiligo skin and pSTAT1+ IFN-γ responsive cells that are uniquely present in the same area. To quantify this spatial correlation, we divided vitiligo skin into three areas based on skin pigmentation and T cell infiltration: depigmented lesions, T cell infiltrates (TIRs), and adjacent pigmented perilesional areas. Quantification showed that the density of CD8+ T cells in the TIRs was ~250 / mm 2 The density of CD8+T cells in the lesion area and the perilesional area is about 30 / mm 2 ( Figure 2 B). We then found the density of pSTAT1+IFN-γ responsive cells in TIR (~250 / mm 2 ) was significantly higher than the lesion area (~50 / mm 2 ) and perilesional area (~30 / mm 2 )( Figure 2 C). Importantly, the density of CD8+ T cells was positively correlated with the density of pSTAT1+ cells (R=0.89, p=1.3×10 -11 )( Figure 2 D). The spatial co-distribution pattern of CD8+ T cells and IFN-γ-responsive cells in the patient's skin is consistent with our single-cell RNA-seq analysis. This result suggests that regional responses to IFN-γ secreted by T cells are associated with a progressive disease state.

[0169] Figure 2 A shows representative immunofluorescence staining images of CD8a, phosphorylated STAT1, and DCT in the skin of patients with vitiligo. The white dotted line indicates the basement membrane of the epidermis. DCT staining marks melanocytes located at the base of the epidermis, CD8a staining marks infiltrating CD8+ T cells, and pSTAT1 staining marks IFN-γ responsive cells. Bright field images show depigmented areas in vitiligo skin. Scale bar, 50 μm. Figure 2B shows quantification of CD8+ T cell density in T cell infiltration zone (TIR), lesion zone and perilesion zone of skin samples from vitiligo patients. Lesion zone is defined as lack of melanocytes and skin pigmentation. Perilesion zone is defined as adjacent skin area that still has melanocytes and has not been infiltrated by T cells. T cell infiltration zone is defined as area enriched with T cells. Figure 2 C shows the quantification of pSTAT1+ cell density in the T cell infiltrated area (TIR), perilesional area, and lesional skin area of ​​skin samples from patients with vitiligo. Figure 2 D shows the scatter plot and linear regression line of CD8+T cell density and pSTAT1+ cell density in skin samples of vitiligo patients. The density of infiltrating CD8+T cells was positively correlated with the density of pSTAT1+ cells (R=0.89, p=1.3e10-11).

[0170] The distribution pattern of pSTAT1+ cells included both the epidermis and dermis, indicating that multiple cell types in the skin responded to IFN-γ. Next, we performed immunofluorescence staining using antibodies against pSTAT1 and markers of eight different cell types to clarify the identity of IFN-γ-responsive cells in the patient's skin ( Figure 2 E). In summary, we detected pSTAT1 in fibroblasts (pdgfra+), melanocytes (DCT+), endothelial cells (CD31+), smooth muscle cells (a-SMA+), mononuclear phagocytes (CD11c+), and keratinocytes (K14+); but not in Langerhans cells (Langerin+) or T cells (CD3e+).

[0171] Figure 2 E shows representative immunofluorescence staining images of pSTAT1+ signals in melanocytes (DCT), fibroblasts (Pdgfra), endothelial cells (CD31), smooth muscle cells (α-SMA), keratinocytes (K14), bone marrow phagocytes (CD11c), Langerhans cells (Langerin), and T cells (CD3e) in the skin of patients with vitiligo. Co-immunofluorescence staining results showed that nuclear pSTAT1+ signals were present in melanocytes, fibroblasts, endothelial cells, smooth muscle cells, keratinocytes, and dendritic cells, but not in Langerhans cells and T cells. Scale bar, 10 μm.

[0172] Example 3 The vitiligo mouse model revealed that response to IFN-γ is required for local CD8+ T cell accumulation and activation in the skin.

[0173] To test whether mouse models relevant to cancer immunotherapy can be used to study the mechanisms of autoimmune diseases, we optimized relevant methods and developed our own protocol ( Figure 3 A): 9-week-old C57 mice were first inoculated with B16F10 melanoma cells on the right side of the dorsal skin; CD4 antibodies purchased from BioXcell (West Lebanon, NH, USA) for cell depletion were injected on days 4 and 10. The tumor was surgically removed on day 12 to prevent tumor cell spread and metastasis. Since the melanocytes of the mouse dorsal skin are located in the hair follicles rather than in the epidermis, the mouse tail skin was used for vitiligo analysis because it contains melanocytes located in the epidermis similar to human skin. We observed that the dorsal skin hair follicles close to the site of B16F10 tumor injection and surgical resection first began to show depigmentation 4 weeks after induction ( Figure 3 B). Based on the results of a large number of trial and error experiments, we found that this phenomenon is an early indicator of successful vitiligo induction. It should be noted here that this procedure only shaves the hair follicles at the tumor injection site and does not remove the rest of the fur on the back skin. The area of ​​hair follicles that first enter the new growth phase in 9-week-old C57 mice is the site of tumor injection and excision where the skin is injured. Therefore, the depigmented hair follicles in this area are the result of new hair growth. In order for the rest of the back skin fur to visually depigment, the initial pigmented hair follicles need to be completely removed, and the newly formed depigmented hair follicles need to appear at the same time, which requires several spontaneous hair cycles, spanning nearly a year ( Fig.10 A). In the tail skin, epidermal depigmentation becomes visually apparent 16 weeks after induction, which is primarily dependent on the natural turnover rate of pigmented keratinocytes on the skin surface ( Figure 3 B). To analyze the induction of vitiligo, we used FACS to quantify the infiltration of CD8+ T cells and the loss of epidermal melanocytes in the tail skin of mice ( Fig.10 B). Compared with control mouse tail skin that contained almost no CD8+ T cells, a large number of CD8+ T cells infiltrated into the tail skin epidermis at day 40 after vitiligo induction, accompanied by a large loss of melanocytes. To functionally test whether CD8+ T cells contribute to melanocyte loss, we used a CD8-depleting antibody after vitiligo induction. This completely blocked the infiltration of CD8+ T cells in the tail skin epidermis and rescued the loss of melanocytes ( Figure 3 C, Fig.10C). Thus, our vitiligo induction protocol effectively triggers endogenously activated CD8+ T cells to infiltrate the skin, leading to a loss of native melanocytes located in the epidermis, similar to patients with autoimmune vitiligo. The major experimental advantage of our vitiligo induction protocol is that it uses only commercially available reagents and can effectively induce patient-like vitiligo pathology in any mouse strain with genetic alterations (such as knockout, conditional knockout, or transgenic); thus allowing us to ask in-depth mechanistic questions.

[0174] Figure 3 A shows a schematic diagram of the strategy for inducing vitiligo in the mouse model. Adult C57 background mice were intradermally inoculated with 1.5 × 10 5 B16F10 cells. Afterwards, two doses of CD4-depleting antibodies were injected on days 4 and 10 to deplete CD4 T cells. On day 12, the tumors were surgically removed. Whole-mount staining was performed in the tail skin epidermis at the indicated time points to analyze T cell infiltration and loss of melanocytes; Figure 3 B shows representative photographs of depigmentation of the back and tail skin of the vitiligo mouse model at the indicated time points. The dorsal skin hair follicles near the site of B16F10 tumor injection and surgical resection first began to show depigmentation approximately 4 weeks after induction. In the tail skin, although T cell infiltration and melanocyte loss could be observed by whole-mount immunofluorescence staining as early as day 19, overall tail skin depigmentation did not become visually apparent until 16 weeks after induction, which was mainly due to the natural turnover rate of pigmented keratinocytes on the skin surface; Figure 3 C shows representative whole-mount immunofluorescence staining and density map images of CD8a+T cells and DCT+ melanocytes on the tail skin of the vitiligo mouse model on days 0, 19, 26, and 33 after induction. The SmoothScatter density map image represents the distribution and density of melanocytes (red) and infiltrating CD8T cells (blue) in the whole-mount image. Before vitiligo induction, melanocytes were evenly distributed in the epidermis of the tail skin, and there were no detectable CD8+T cells. From day 19 after vitiligo induction, infiltration and local aggregation of CD8a+T cells in the epidermis began to be observed, and melanocytes in the same area were lost. At later time points, T cell clusters gradually spread, and T cell density seemed to be particularly high in the border areas of the clusters.

[0175] Fig.10A shows representative photographs of mice at days 33, 120, and 300 after vitiligo induction. The depigmented area of ​​the dorsal skin hair follicles gradually expanded from the right B16F10 tumor injection and surgical resection site to the entire dorsal area. It should be noted that the procedure only shaved the hair follicles at the tumor injection site, and did not remove the rest of the fur on the dorsal skin. The original fur before vitiligo induction was pigmented, and only the newly appeared hair follicles after vitiligo induction would be depigmented. Since vitiligo induction began at 9 weeks of age, most of the dorsal skin hair follicles were in a prolonged resting phase at this time. The first area of ​​hair follicles to enter the new growth phase was the tumor injection and resection site where the skin was injured. In order for the rest of the dorsal skin fur to completely depigment, the original hair follicles need to completely fall off and the newly formed depigmented hair follicles need to appear at the same time, which requires several hair cycles, spanning nearly a year; Fig.10 B shows representative whole-mount immunofluorescence staining images and density map images of CD8+T cells and DCT+melanocytes in the epidermis of the tail skin on days 19, 26, and 33 after vitiligo induction. CD8+T cell immunofluorescence signals were digitally converted and analyzed by the dbscan package in R. The results showed that clonal T cell clusters aggregated and spread in the epidermis; Fig.10 C shows quantification of the number of CD8+ T cell clusters in the skin epidermis at days 19, 26, and 33 after vitiligo induction. The number of T cell clusters was acquired and analyzed by the DBSCAN package in R. The results showed that the number of CD8+ T cell clusters increased from days 19 to 26 and remained unchanged from days 26 to 33. The data reflect the mean ± SD of at least 6 mice.

[0176] To characterize the spatiotemporal progression pattern of vitiligo in an induced mouse model, we used whole-mount immunofluorescence staining to analyze the tail skin at different time points. Before vitiligo induction, melanocytes were evenly distributed in the epidermis of the tail skin, and almost no CD8+ T cells were detected. Starting from day 19 after vitiligo induction, sparse infiltration of CD8+ T cells in the epidermis and small areas of melanocyte loss could be observed. Interestingly, we noticed that the loss of melanocytes corresponded to areas where CD8+ T cells were locally clustered. After infiltration, CD8+ T cells were not evenly distributed in the skin epidermis, but continuously aggregated into small clone-like clusters, which continued to proliferate at later time points ( Figure 3 D, Fig.10 D). To quantify this interesting phenomenon, we analyzed the number and size of CD8+ T cell clusters in vitiligo skin at days 19, 26, and 33 after induction ( Fig.10E-10F). From day 19 to day 26, the number of clonal CD8+ T cell clusters increased steadily, indicating the continued formation of new clusters. However, from day 26 to day 33, the number of CD8+ T cell clusters remained stable. The size of the entire cluster continued to increase across the 3 time points; in particular, the frequency of large CD8+ T cell clusters that were not present on day 19 increased on day 33. We also noticed that the CD8+ T cell density was highest in the border area of ​​the T cell clusters, between the depigmented lesion skin and the pigmented perilesional skin, similar to the skin of patients with vitiligo in a progressive state. We did not detect local proliferation of skin-infiltrating CD8+ T cells ( Fig.10 G), so the local aggregation of CD8+ T cells at the border area of ​​existing T cell clusters and the continuous expansion of cluster size are caused by the active recruitment of skin-infiltrating CD8+ T cells to areas with high CD8+ T cell density. By Ki67 staining, we did not detect a higher proliferation rate of CD8+ T cells at the border than inside the T cell clusters ( Fig.10 H). Since melanocyte loss was observed only within T cell clusters, this also explains the spread of depigmented areas in vitiligo skin in patients and mouse models. Based on these observations, it is implied that there is a local positive feedback recruitment mechanism to coordinate the regional aggregation of tissue CD8+ T cells, which is essential for effective T cytotoxic activity. This phenomenon is similar to quorum sensing in bacterial populations, which describes changes in gene expression in response to population density. What we observed here suggests that skin-infiltrating CD8+ T cells can somehow sense the regional density of their own population and move to areas of high density.

[0177] Figure 3 D shows the quantification of CD8+ T cell density and melanocyte retention area in the tail skin epidermis of control, vitiligo-induced, and CD8-depleting antibody-treated vitiligo-induced mice. In the absence of vitiligo induction, the tail skin epidermis does not contain CD8+ T cells. On day 40 after vitiligo induction, a large number of CD8+ T cells infiltrated the tail skin epidermis, which could be completely blocked by CD8-depleting antibody treatment. Accordingly, CD8-depleting antibody treatment completely rescued the loss of melanocytes after vitiligo induction.

[0178] Fig.10 D shows the distribution of the number of CD8+ T cells in each T cell cluster as a function of vitiligo induction time points at days 19, 26, and 33 in WT skin. The number of cells in each T cell cluster was quantified by the DBSCAN and VCD software packages in R. Data reflect the mean ± SD of at least 6 mice; Fig.10E shows the schematic diagram, FACS profile and quantification results of melanocytes and CD8+T cells in the epidermis of the tail skin of control and vitiligo-induced mice on day 33 after induction. The epidermis and dermis of the tail skin were enzymatically dissociated by treatment with dispase. The skin epidermis was then digested with trypsin to obtain a single cell suspension. After immunostaining for c-Kit, CD45 and CD8, FACS was used to quantify the percentages of melanocytes (c-Kit+) and CD8+T cells (CD45+, CD8+). The data reflect the mean ± SD of 3 mice in the control group and 15 mice in the experimental group. ***p<0.001 by t-test; Fig.10 F shows a schematic diagram and representative whole-mount immunofluorescence staining images of the CD8 depletion antibody-treated vitiligo mouse model. Adult C57 background mice were intradermally inoculated with 1.5 × 10 5 B16F10 cells. After the 4th and 10th days, two doses of CD4-depleting antibodies were injected to deplete CD4 T cells. The tumor was surgically removed on the 12th day. From the 13th to the 40th day, CD8-depleting antibodies were injected every other day for a total of 7 injections. Whole-mount staining was performed on the 40th day to analyze T cell infiltration and melanocyte loss in the epidermis of the tail skin. DCT staining marks melanocytes; CD8a staining marks infiltrating T cells. This result shows that CD8+T cell depletion prevents the loss of melanocytes in the epidermis of the tail skin of vitiligo; Fig.10 G shows representative whole-mount immunofluorescence staining images of DCT+ melanocytes in the epidermis of the tail skin of vitiligo-induced WT and IFNGR1 KO mice at day 120 after vitiligo induction. Even at day 120 after vitiligo induction, melanocytes in the IFNGR1 KO skin remained intact ( Fig.10 I). (Scale bar, 500 μm.)

[0179] Our single-cell RNA-seq and immunohistochemical analysis of vitiligo patient skin revealed that regional responses to T cell-secreted IFN-γ were associated with a progressive disease state. To investigate whether IFN-γ-responsive cells in the skin play a functional role in mediating localized accumulation of T cells, we used IFNGR1 KO mice to induce vitiligo. Whole-mount immunofluorescence staining was shown at day 33 after vitiligo induction, and both WT and IFNGR1 KO mice showed a strong infiltration of CD8 T cells into the epidermis of the tail skin ( Figure 3 E). However, it is striking that only in the WT tail skin did CD8+ T cells locally aggregate into clusters, leading to a loss of melanocytes in the same area; in the IFNGR1 KO tail skin, infiltrating CD8+ T cells were evenly distributed throughout the skin epidermis, did not aggregate into clusters, and there was no loss of melanocytes ( Figure 3E). We quantified the number of CD8+ T cells and melanocyte retention area in vitiligo-induced WT and IFNGR1 KO tail skin at each scale. In WT vitiligo skin, the number of CD8+ T cells was negatively correlated with the number of melanocytes (R = -0.84, p = 2.2 × 10 -16 ); however, in IFNGR1 KO vitiligo skin, the number of CD8+T cells was not related to the number of melanocytes (R=-0.09, p=0.27) ( Figure 3 F). This quantitative result confirms that CD8+ T cells infiltrating the IFNGR1 KO tail skin lack cytotoxic activity. We also quantified the total number of CD8+ T cells and the area of ​​melanocyte retention in the epidermis of the tail skin of WT and IFNGR1 KO mice (with and without vitiligo induction) using whole mount staining ( Figure 3 G). Although similar numbers of CD8+ T cells infiltrated the tail skin of WT and IFNGR1 KO mice after vitiligo induction, T cell infiltration led to a loss of melanocytes only in WT skin. The inability of infiltrating CD8+ T cells to attack melanocytes was not due to delayed activity, as melanocytes were still intact in IFNGR1 KO tail skin even 6 months after vitiligo induction.

[0180] Figure 3 E shows representative whole-mount immunofluorescence staining images of CD8+T cells and DCT+melanocytes in the epidermis of the tail skin of vitiligo-induced WT and IFNGR1 KO mice. On day 33 after vitiligo induction in WT and IFNGR1 KO mice, the epidermis of the tail skin showed strong infiltration of CD8+T cells. However, only in the WT tail skin, CD8+T cells showed local aggregation and progressive cluster spread, resulting in the loss of melanocytes in the same area. In the IFNGR1 KO tail skin, CD8+T cells were evenly distributed, did not aggregate into clusters, and melanocytes remained intact; Figure 3 F shows the scatter plot and linear regression line of the number of CD8+T cells and the percentage of melanocytes in vitiligo-induced WT and IFNGR1 KO tail skin at each scale. In WT vitiligo skin, the number of CD8+T cells was negatively correlated with the number of melanocytes (R=-0.84, p=2.2×10-16), but in IFNGR1 KO vitiligo skin, the number of CD8+T cells was not correlated with the number of melanocytes (R=-0.09, p=0.27); Figure 3G shows the quantification of the number of T cells and the percentage of melanocyte retention area in WT and IFNGR1 KO skin with or without vitiligo induction on day 33. Although similar numbers of T cells infiltrated the tail skin of WT and IFNGR1 KO mice after vitiligo induction, only T cell infiltration in WT skin resulted in melanocyte loss.

[0181] We were particularly interested in the different distribution patterns of CD8+ T cells in IFNGR1 KO tail skin revealed by the global analysis. The smoothed scatter density map image highlights the clonal CD8+ T cell clusters in WT vitiligo skin, with T cell density particularly high in the border area. However, in IFNGR1 KO vitiligo skin, CD8+ T cells were almost completely uniformly distributed, with no localized aggregation behavior ( Figure 3 H). To quantify this dramatic difference, we analyzed CD8+ T cell cluster size in the tail skin of vitiligo-induced WT and IFNGR1 KO mice at different time points. In WT vitiligo skin, CD8+ T cell cluster size increased steadily from day 19 to day 33, indicating that skin-infiltrating CD8+ T cells were actively recruited into existing clusters. However, in IFNGR1 KO vitiligo skin, CD8+ T cell cluster sizes were significantly smaller at D33 post-induction than those found in WT vitiligo skin from day 19 to day 33 ( Figure 3 I) These results suggest that the response to IFN-γ is critical for the local accumulation and cytotoxic activity of CD8+ T cells in the skin.

[0182] Figure 3 H shows a SmoothScatter density map of infiltrating T cells (blue) in the epidermis of vitiligo-induced WT and IFNGR1 KO mouse tail skin. Note that CD8 T cells in WT skin are clone-like clusters compared to the uniform distribution of CD8 T cells in IFNGR1 KO skin; Figure 3 I shows the size distribution of T cell clusters as a function of vitiligo induction time point in WT skin at day 19, day 26, day 33 and IFNGR1 KO skin at day 33. T cell cluster size was quantified by the area of ​​each T cell cluster.

[0183] Although we used mice with a direct KO of IFNGR1, the IFN-γ-responsive cells that mediate this effect appear to be regional in the skin, as the primary defect is the inability of T cells to aggregate locally rather than an inability of T cells to infiltrate the skin. Since our immunofluorescence analysis of the patient's skin showed that the T cells were pSTAT1-, this ruled out the possibility that CD8+ T cells promote self-aggregation via autocrine IFN-γ signaling. At least six different cell types in the patient's skin respond to IFN-γ, including keratinocytes, melanocytes, fibroblasts, endothelial cells, smooth muscle cells, and dendritic cells.

[0184] Example 4 Skin dermal cells mediate local CD8+ T cell aggregation and activation via IFN-γ signaling

[0185] Before determining which skin-resident cells are responsible for orchestrating local T cell recruitment and activation following IFN-γ exposure, we first needed to rule out the possibility that the CD8+ T cells of IFNGR1 KO mice are intrinsically defective. Previous studies have shown that T cell cytotoxicity is unaffected in IFNGR1 KO mice. To test whether this is also true in our vitiligo mouse model, we investigated whether the cytotoxic activity of CD8+ T cells derived from IFNGR1 KO mice is normal if given a WT local environment using skin transplantation and vitiligo induction assays ( Figure 4 A, Fig.11 A). To this end, we transplanted full-thickness tail skin from WT mice onto the backs of IFNGR1 KO host mice (WT->IFNGR1 KO grafts). After induction of vitiligo in host mice, T cell infiltration and melanocyte loss in the transplanted tail skin were analyzed by whole-mount immunofluorescence staining. In the absence of vitiligo induction in host mice, graft detection did not result in epidermal melanocyte loss or CD8+ T cell infiltration in the transplanted tail skin; only after induction of vitiligo in host mice, there was a large number of CD8+ T cells infiltrating the transplanted tail skin, resulting in epidermal melanocyte loss ( Fig.11 B).

[0186] Figure 4 A shows a schematic diagram of the graft and vitiligo induction assay. Full-thickness tail skin from WT and IFNGR1 KO mice was grafted onto the backs of host mice with the indicated genotypes. After induction of vitiligo on host mice, the grafted tail skin was analyzed for T cell infiltration and melanocyte loss using whole-mount immunofluorescence staining.

[0187] Fig.11A shows the timeline of the graft and vitiligo induction assays. Full-thickness tail skin from 4-6 week old mice was grafted onto 8 week old host mice. The day of B16F10 tumor injection was designated as day 0, and the corresponding day of tail skin graft was designated as day -20.

[0188] Fig.11 B shows representative whole-mount immunofluorescence staining images of the transplanted tail skin epidermis on days 0 and 21 indicated in (A), with or without vitiligo induction. DCT staining marks melanocytes located in the basal layer of the epithelium; CD8a staining marks infiltrating T cells. Without vitiligo induction, there were very few CD8+T cells in the transplanted skin, and no obvious melanocyte loss on day 20. However, after vitiligo induction, there was strong CD8+T cell infiltration and melanocyte loss in the transplanted skin.

[0189] For the WT->IFNGR1 KO transplantation experiment, the tail skin from IFNGR1 KO mice was transplanted to the other side of the back of the same IFNGR1 KO host as an internal control (IFNGR1 KO->IFNGR1 KO transplantation). On the 21st day after vitiligo was induced in the host mice, CD8+T cells from IFNGR1 KO host mice strongly infiltrated into the transplanted WT tail skin and caused the loss of WT melanocytes; however, CD8+T cells from the same host only infiltrated a small amount into the transplanted IFNGR1 KO skin and did not cause the loss of IFNGR1 KO melanocytes ( Figure 4 B). On day 60 after induction of vitiligo in host mice, skin depigmentation can be easily observed on WT->IFNGR1 KO grafts, but IFNGR1 KO->IFNGR1 KO grafts on the same host are still pigmented ( Fig.11 C). We quantified the number of infiltrating CD8+ T cells and melanocytes in the transplanted skin of host mice with or without vitiligo induction ( Figure 4C). In the case of host mice without vitiligo induction, CD8+T cells were very few in both IFNGR1 KO->IFNGR1 KO grafts and WT->IFNGR1 KO grafts, and the number of melanocytes in both grafts was comparable. After vitiligo induction in host mice, CD8+T cells in WT->IFNGR1 KO grafts were significantly more than those in IFNGR1 KO->IFNGR1 KO grafts on the same host; in contrast, the number of melanocytes in WT->IFNGR1 KO grafts was much lower than that in IFNGR1 KO->IFNGR1 KO grafts on the same host. This result clearly indicates that: 1. CD8+T cell cytotoxic potential is normal in IFNGR1 KO mice; 2. Skin-resident IFN-γ responsive cells are required for the recruitment and activation of local CD8+T cells.

[0190] To further verify that skin-resident IFN-γ-responsive cells are required for local CD8+ T cell recruitment and activation, we performed reverse transplantation experiments on the same host: IFNGR1 KO->WT and WT->WT. The rationale is that if skin-resident IFN-γ-responsive cells are required for local CD8+ T cell recruitment and activation, then IFNGR1 KO transplanted skin would not be able to attract CD8+ T cells from WT hosts after vitiligo induction, but WT transplants on the same host would be able to attract CD8+ T cells from WT hosts. We were surprised to find completely different results. CD8+ T cells from WT host mice strongly infiltrated the transplanted WT and IFNGR1 KO tail skin and caused a comparable loss of melanocytes in both WT and IFNGR1 KO mice ( Figure 4 B). Quantification showed that without induction of vitiligo in host mice, CD8+T cells were very low in both IFNGR1KO->WT and WT->WT grafts, and the number of melanocytes in both grafts was comparable. After induction of vitiligo in host mice, comparable numbers of infiltrating CD8+T cells were present in WT->WT grafts and IFNGR1 KO->WT grafts on the same host, and the number of melanocytes was equally reduced in WT->WT grafts compared with IFNGR1 KO->WT grafts on the same host ( Figure 4C). This result seems to contradict our conclusion that skin-resident IFN-γ-responsive cells are required for local CD8+ T cell recruitment and activation. Because all cells in the IFNGR1 KO grafts are assumed to be unresponsive to IFN-γ, CD8+ T cell recruitment would be as efficient as in WT grafts. One explanation for this discrepancy is that WT cells from the host migrate into the transplanted IFNGR1 KO tail skin and that these WT cells functionally rescue the IFNGR1 KO transplanted skin in terms of CD8+ T cell recruitment.

[0191] To test this interpretation, full-thickness C57 tail skin was transplanted into host mice expressing membrane-tdTomato (mT) (Rosa-mT), in which all cells are genetically labeled as mT+ ( Figure 4 D). Immunofluorescence staining of the junction area between the transplanted donor C57 tail skin and the host dorsal skin of mT mice showed that host dermal cells did invade the dermis of the transplanted skin. However, neither keratinocytes nor melanocytes were shown to infiltrate from the host into the transplanted skin. This result supports our interpretation that WT cells from the host migrated into the transplanted IFNGR1 KO tail skin and functionally rescued the transplanted skin, and it also ruled out keratinocytes and melanocytes as candidate cell types for this rescue effect. Since the transplantation experiments severed neuronal connections in intact skin, neurons were also unlikely to be the target cells we were looking for. The strong infiltration of dermal cells from the host into the transplant suggests that the dermal cells of the skin are the cell type responsible for recruiting CD8+ T cells after receiving IFN-γ. To distinguish which WT dermal cells migrated into the transplanted IFNGR1 KO tail skin, we performed co-immunofluorescence staining for pSTAT1 and cell type markers ( Figure 4 E). In IFNGR1 KO->IFNGR1 KO grafts, as expected, there was no pSTAT1+ signal in the graft skin on the host mice after induction of vitiligo. However, in IFNGR1 KO->WT grafts, ectopic pSTAT1+ signals could be detected in the graft skin on the host mice after induction of vitiligo, indicating that the host's WT cells did migrate into the graft and respond to IFN-γ-secreting T cells. Co-immunofluorescence staining results showed that nuclear pSTAT1+ signals were present in fibroblasts (pdgfra+), endothelial cells (CD31+), smooth muscle cells (a-SMA+), and immune cells (CD45+); but not in keratinocytes (K14+) or Langerhans cells (Langerin+) ( Figure 4 E and Fig.11D). We quantified the distribution of pSTAT1+ in these different cell types and found that ~70% of pSTAT1+ cells were fibroblasts, ~20% were endothelial cells, and <10% were smooth muscle cells or immune cells ( Figure 4 F).

[0192] Figure 4 B shows representative whole-mount immunofluorescence staining of CD8+ T cells and DCT+ melanocytes in transplanted tail skin on host mice after induction of vitiligo. On day 21 after induction of vitiligo in host mice, CD8+ T cells from IFNGR1 KO host mice strongly infiltrated into transplanted WT tail skin, resulting in loss of WT melanocytes, but CD8+ T cells from the same host only infiltrated a small amount into transplanted IFNGR1 KO tail skin and did not result in loss of IFNGR1 KO melanocytes. In contrast, CD8+ T cells from WT host mice infiltrated into transplanted WT and IFNGR1 KO tail skin with similar intensity and resulted in effective loss of WT and IFNGR1 KO melanocytes; Figure 4 C shows the quantification of T cell numbers and melanocyte numbers in transplanted tail skins on host mice with or without induced vitiligo. Pairs of donor tail skins transplanted on the same host mouse are connected by lines. Genotypes of donor skins and host mice are marked; Figure 4 D shows a representative immunofluorescence staining of the junction area between the transplanted donor C57 tail skin and the host back skin of the membrane Tomato transgenic mouse. The mTomato signal labels all cells from the host mouse. K14 staining labels keratinocytes. DCT staining labels melanocytes that are only present in the tail skin epidermis and not in the back skin. Samples were taken 20 days after transplantation. Note that the host dermal cells clearly infiltrated into the transplanted skin dermis. However, no infiltration of keratinocytes and melanocytes from the host to the transplanted skin was shown. Figure 4 E shows representative immunofluorescence staining images of pSTAT1+ signals in fibroblasts (Pdgfra), immune cells (CD45), endothelial cells (CD31), and smooth muscle cells (α-SMA) in transplanted tail skin after induction of vitiligo in host mice. In IFNGR1KO->IFNGR1KO transplants, no pSTAT1+ signals were detected in the transplanted skin after induction of vitiligo in host mice. However, in IFNGR1KO->WT transplants, ectopic pSTAT1+ signals could be detected in the transplanted skin after induction of vitiligo in host mice, indicating that WT cells infiltrated from the host into the transplanted IFNGR1KO skin. Co-immunofluorescence staining results showed the presence of nuclear pSTAT1+ signals in fibroblasts, endothelial cells, and smooth muscle cells; Figure 4 F shows quantification of the distribution of pSTAT1+ cells in the indicated cell types in IFNGR1 KO->WT grafts after induction of vitiligo in host mice. Data reflect the mean ± SD from 3 independent host mice. (Scale bar, 50 μm)

[0193] Fig.11 C shows representative photographs of transplanted and vitiligo-induced mice. Full-thickness tail skin from WT and IFNGR1 KO mice was transplanted onto the back of IFNGR1 KO host mice. On day 60 after induction of vitiligo in host mice, the WT transplanted tail skin was visually depigmented, while the IFNGR1 KO transplanted tail skin was still pigmented; (scale bar, 50 μm).

[0194] Together, our transplantation and vitiligo induction experiments showed that since the majority of ectopic pSTAT1+ cells in IFNGR1 KO->WT grafts were fibroblasts, which resulted in a complete rescue of IFNGR1 KO skin in terms of CD8+ T cell recruitment, we next used genetic experiments to determine whether fibroblasts of the skin dermis were the major cell type mediating local CD8+ T cell recruitment and activation.

[0195] Example 5 The IFNGR1-JAk1-STAT1 signaling axis in dermal fibroblasts is required for local CD8+ T cell recruitment and activation in autoimmune skin.

[0196] To specifically ablate IFNGR1 in fibroblasts, we used Pdgfra-CreER::IFNGR1 fl / fl mice. After tamoxifen injection from P50 to P56, keratinocytes, melanocytes, immune cells, endothelial cells, and fibroblasts purified from the skin by FACS were used to verify the specificity and efficiency of the knockout ( Figure 5 A, Fig.12 A~12C). IFNGR1 expression was specifically lost in fibroblasts but was not affected in keratinocytes, melanocytes, immune cells, or endothelial cells in the skin. To investigate whether IFN-γ-responsive dermal fibroblasts are the main cell type mediating local CD8+ T cell recruitment and activation, we used Pdgfra-CreER::IFNGR1fl / fl cKO mice to induce vitiligo. On day 33 after vitiligo induction, both WT and IFNGR1 cKO mice showed strong infiltration of CD8+ T cells into the epidermis of the tail skin ( Figure 5B). However, CD8+ T cells only clustered in WT tail skin, resulting in the loss of melanocytes in the same area; in IFNGR1 cKO tail skin, infiltrating CD8+ T cells were evenly distributed throughout the skin, without clustering and without melanocyte loss. We quantified the number of CD8+ T cells and retained melanocytes in vitiligo-induced WT and IFNGR1 cKO tail skin for each scale. In WT mice, the number of CD8+ T cells was negatively correlated with the number of melanocytes (R = -0.93, p = 2.2 × 10 -16 ); however, in IFNGR1 cKO vitiligo skin, the number of CD8+T cells was not related to the number of melanocytes (R=-0.08, p=0.31) ( Figure 5 C). SmoothScatter density map image showing clonal clusters of CD8+ T cells in WT vitiligo skin, but completely uniform distribution in IFNGR1 cKO skin ( Figure 5 B). To quantify this phenomenon, we analyzed the size of CD8+ T cell clusters in the tail skin of vitiligo-induced WT and IFNGR1 cKO mice at day 33 after vitiligo induction. In IFNGR1 cKO vitiligo skin, the size of CD8+ T cell clusters was significantly smaller than those found in WT vitiligo skin ( Figure 5 D). Essentially, Pdgfra-CreER::IFNGR1fl / fl cKO mice phenocopy IFNGR1 KO mice and display the same defects in local CD8+ T cell accumulation and cytotoxic activity after induction of vitiligo. This striking result suggests that fibroblasts are the primary cell type responsible for orchestrating local CD8+ T cell accumulation and activation after IFN-γ stimulation in autoimmune skin.

[0197] To verify this conclusion and further dissect the downstream factors mechanistically, we performed chimeric knockdown in fibroblasts by intradermal injection of lentivirus expressing different shRNAs ( Figure 5 E). First, we characterized the infected cell types using intradermal injections: lentivirus expressing shRNA and H2BRFP was injected into the dermis of the tail skin of P1 K14H2BGFP mice, and skin samples were collected at P56 for FACS analysis ( Fig.12D). In the epidermis, no RFP+ cells were detected. In the dermis, RFP+ cells were detected in 3 cell types: CD45-, CD31- stromal fibroblasts, CD45-, CD31+ endothelial cells, and CD45+ immune cells. Quantification showed that approximately 75% of RFP+ cells were fibroblasts. For knockdown and induction experiments of vitiligo, lentivirus expressing shRNA and H2BRFP was injected into the dermis of the tail skin of P1 C57 mice; vitiligo was induced 9 weeks later. T cell infiltration and melanocyte loss in the tail skin were analyzed by whole-mount immunofluorescence staining at day 33 after induction. RFP+ cells marked the distribution pattern of dermal fibroblasts expressing shRNA. In the dermis expressing missense shRNA, the CD8+ T cell distribution pattern randomly overlapped with the RFP+ area; in the corresponding epidermis, the CD8+ T cell distribution pattern largely followed their own dermal population as a mirror image, and melanocytes were lost wherever there were high-density CD8+ T cell clusters ( Figure 5 E). We compared density map images of dermal RFP+ areas with corresponding epidermal melanocyte-retaining areas and found no discernible pattern. To quantify these results, we compared the number of CD8+ T cells per unit area of ​​the dermis to the percentage of infected fibroblasts, and the percentage of melanocyte-retaining area per unit area of ​​the epidermis to the percentage of infected fibroblasts per unit area of ​​the dermis ( Figure 5 F). In mice that were only infected but not induced with vitiligo, there were no infiltrating CD8+T cells in the dermis, so there was no correlation between the number of CD8+T cells and the percentage of infected fibroblasts (R=0.13, p=0.18); since melanocytes are intact in the epidermis, there was no correlation between the percentage of melanocyte area retained and the percentage of infected fibroblasts (R=0.013, p=0.89). This result indicates that lentiviral infection in the dermis does not trigger CD8+T cell infiltration or melanocyte loss in the skin. In skin expressing missense shRNA that induces vitiligo, the infiltrating CD8+T cells in the dermis increased significantly. However, the number of CD8+T cells in the RFP+ high and low areas was comparable, so there was no correlation between the number of CD8+T cells and the percentage of infected fibroblasts (R=0.04, p=0.33); even if melanocyte loss occurred in the epidermis, the reduction in the RFP+ high and low areas was the same, so there was no correlation between the percentage of melanocyte area remaining and the percentage of infected fibroblasts (R=0.05, p=0.29). This result indicates that chimeric expression of missense shRNA in dermal fibroblasts does not affect local CD8+T cell aggregation and cytotoxic activity after induction of vitiligo.

[0198] Therefore, we next tested whether knocking down key factors mediating IFN-γ signaling in dermal fibroblasts would inhibit local aggregation and activation of CD8+ T cells in WT mice. We selected two shRNAs with high knockdown efficiency for each of the IFNGR1, JAK1, or STAT1 genes and confirmed their role in blocking IFN-γ-induced responses in cultured fibroblasts in vitro ( Fig.12 E~12F). Figure 5 As shown in E, CD8+T cells expressing JAK1 shRNA in the dermis tended to move toward the RFP-WT fibroblast area in the chimeric skin and were essentially absent in the RFP+ fibroblast area expressing JAK1 shRNA. In the corresponding epidermis, the CD8+T cell distribution pattern reflected their own dermal population and resulted in the loss of melanocytes in the high-density CD8+T cell area. We compared the density map images of the dermal RFP+ area and the corresponding epidermal melanocyte-retaining area, which overlapped to a large extent, implying that fibroblasts expressing JAK1 shRNA protected the melanocytes above in the same area within the chimeric skin. The quantification results showed that the number of CD8+T cells in the RFP+ high area was lower than that in the RFP- area in the skin expressing shRNA against IFNGR1, JAK1 or STAT1. Overall, the number of CD8+T cells was negatively correlated with the percentage of infected fibroblasts: IFNGR1 shRNA (R = -0.43, p < 2.2 × 10 -16 ); JAK1 shRNA (R=-0.5, p<2.2×10 -16 ); STAT1 shRNA (R=-0.47, p<2.2×10 -16 ). Correspondingly, the loss of melanocytes in the epidermis was higher in RFP- areas than in RFP+ areas. Overall, the percentage of area with preserved melanocytes was positively correlated with the percentage of infected fibroblasts: IFNGR1 shRNA (R = 0.29, p < 1.9 × 10 -16 ); JAK1 shRNA (R=0.39, p<2.2×10 -16 ); STAT1 shRNA (R=0.38, p<2.2×10 -16 ).

[0199] Figure 5A shows a schematic diagram and QPCR validation of fibroblast-specific ablation of IFNGR1 using Pdgfra-CreER::IFNGR1 fl / fl mice. QPCR analysis was performed on FACS-purified keratinocytes, melanocytes, immune cells, endothelial cells, and fibroblasts after tamoxifen injection from P50 to P56; Figure 5 B shows representative density map images of CD8+T cells and whole-mount immunofluorescence staining images of CD8+T cells and DCT+melanocytes in the epidermis of the tail skin of WT and Pdgfra-CreER::IFNGR1 fl / fl mice induced with vitiligo. The SmoothScatter density map image represents the distribution and density of T cells (blue) corresponding to the whole-mount immunofluorescence staining image. On day 33 after induction of vitiligo in WT and IFNGR1 cKO mice, the epidermis of the tail skin showed strong infiltration of CD8+T cells. But only in the WT tail skin, CD8+T cells aggregated into clusters and caused the loss of melanocytes in the same area. In the IFNGR1 cKO tail skin, CD8+T cells were evenly distributed, did not aggregate into clusters, and there was no loss of melanocytes. Scale bar, 500um; Figure 5 C shows the scatter plot and linear regression line of the number of CD8+T cells and the percentage of melanocytes in the tail skin of WT and Pdgfra-CreER::IFNGR1 fl / fl mice induced with vitiligo at each scale. In WT vitiligo skin, the number of CD8+T cells was negatively correlated with the number of melanocytes (R = -0.93, p = 2.2 × 10 -16 ), but in IFNGR1 cKO vitiligo skin, the number of CD8+T cells was not related to the number of melanocytes (R=-0.08, p=0.31); Figure 5 D shows the size distribution of T cell clusters as a function of vitiligo induction time at day 33 in WT and Pdgfra-CreER::IFNGR1 fl / fl skin; Figure 5 E shows a schematic diagram of the skin knockdown experiment, representative whole-mount immunofluorescence staining images, and density maps. Lentivirus expressing shRNA and H2BRFP was injected into the dermis of the P1 tail skin. Vitiligo was induced at 9 weeks, and whole-mount staining was performed 33 days later. RFP marked lentiviral-infected dermal fibroblasts. CD8a staining marked infiltrating CD8+ T cells in the epidermis or dermis. DCT staining marked melanocytes in the epidermis. SmoothScatter density map images represent the distribution and density of shRNA-expressing fibroblasts (red), T cells (blue), and melanocytes (gray), corresponding to the whole-mount immunofluorescence staining images. Scale bar, 500 μm; Figure 5F shows box-and-line scatter plots and linear regression lines of the number of T cells versus the percentage of infected fibroblasts (upper panel) and the percentage of area with retained melanocytes versus the percentage of infected fibroblasts (lower panel) in the tail skin of vitiligo induced at each scale in the dermal knockdown experiment. R indicates correlation and p indicates the significance of the correlation.

[0200] Fig.12 A shows a schematic diagram and FACS profile of the isolation of different cell populations in the epidermis and dermis of the skin. The epidermis and dermis of the tail skin were enzymatically dissociated by treatment with dispase. The skin epidermis was then digested with trypsin and the dermis was digested with collagenase to obtain a single cell suspension. Different cell types were collected using FACS after immunostaining for c-Kit, CD45, or CD31. In the epidermis, CD45 and c-Kit were used to enrich immune cells (CD45+, c-Kit-), melanocytes (CD45-, c-Kit+), and keratinocytes (CD45-, c-Kit-). In the dermis, CD45 and CD31 were used to distinguish immune cells (CD45+, CD31-), endothelial cells (CD45-, CD31+) from stromal fibroblasts (CD45-, CD31-); Fig.12 B shows the Fig.12 QPCR analysis of cell type-specific genes in the FACS-isolated population shown in A. KRT14, DCT, CD45, PDGFRA, and CD31 are used as signature genes for keratinocytes, melanocytes, immune cells, fibroblasts, and endothelial cells, respectively. Note that each signature gene is highly expressed only in the expected cell type, indicating the specificity of the cell type isolated using FACS; Fig.12 C shows an immunofluorescence staining image of a representative section of Pdgfra-CreER::mTmG tail skin after tamoxifen injection. K14 staining marks epithelial cells; CD31 staining marks endothelial cells. mGFP signal marks Pdgfra-CreER-labeled cells. The mutually exclusive pattern of K14 and CD31 with mGFP indicates that Pdgfra-CreER marks most dermal fibroblasts of the skin, but not keratinocytes and endothelial cells; Fig.12D shows a schematic diagram and FACS graph of different cell populations in the epidermis and dermis of the skin after intradermal injection of lentivirus. Lentivirus expressing shRNA and H2BRFP was injected into the dermis of the tail skin of P1K14H2BGFP mice. Skin samples were collected at P56 for FACS analysis. The epidermis and dermis of the tail skin were enzymatically dissociated by dispase treatment. The skin epidermis was then digested with trypsin and the dermis was digested with collagenase to obtain a single cell suspension. After immunostaining for CD45 and CD31, FACS was used to analyze the percentage of different cell types in RFP+ cells. In the epidermis, RFP+ cells were not detected in GFP+ keratinocytes. In the dermis, RFP+ cells are divided into 3 main cell types present in the dermis of the tail skin: CD45+, CD31- immune cells; CD45-, CD31+ endothelial cells; and CD45-, CD31- matrix fibroblasts. Quantification results showed the composition of cell types in RFP+ cells after intradermal injection of lentivirus; Fig.12 E shows a schematic diagram and qPCR validation of shRNA knockdown efficiency. Two shRNAs targeting IFNGR1, JAK1, and STAT1, each member of the IFN-γ signaling pathway, were used. Lentivirus expressing shRNA and H2BRFP was packaged in 293ft cells and used to infect 3T3 fibroblasts in vitro. Lentivirus expressing missense shRNA was used as a control. The knockdown efficiency of each shRNA was compared with cells expressing missense shRNA; Fig.12 F shows a schematic diagram and qPCR analysis of the effect of shRNA in blocking the IFN-γ response. After lentiviral infection, 3T3 cells were treated with IFN-γ for 6 hours and then collected for qPCR analysis. The classic IFN-γ target gene GBP2 was used as a reporter gene for activation of the IFN-γ signaling pathway. Lentivirus expressing missense shRNA was used as a control. After 6 hours of IFN-γ treatment, GBP2 expression levels were significantly increased in cells expressing missense shRNA. Knockdown of IFNGR1, JAK1, and STAT1 all significantly reduced the IFN-γ-induced increase in GBP2. (Scale bar, 50 μm. Data reflect the mean ± SD from 3 independent mice or three technical replicates).

[0201] The chimeric knockdown experiments in fibroblasts not only validated our results obtained using Pdgfra-CreER::IFNGR1 fl / fl mice, but also further revealed that the IFNGR1-JAK1-STAT1 signaling axis in fibroblasts is required for mediating the local accumulation and activation of CD8+ T cells. Most importantly, these experiments showed that in areas where fibroblasts responded heterogeneously to IFN-γ, T cells preferentially accumulated toward areas with high IFNGR1-JAK1-STAT1 signaling.

[0202] Example 6 IFN-γ signaling-induced fibroblast-specific secretion of chemokines controls the regional recruitment of T cells.

[0203] To test whether purified fibroblasts alone are sufficient to induce localized accumulation of T cells in vivo, we isolated primary fibroblasts from the skin of WT and IFNGR1 KO mice, labeled them with RFP, and injected them intradermally into the tail skin of IFNGR1 KO mice, then induced vitiligo in the host mice ( Fig.13 A). The aggregation of CD8+ T cells was analyzed by whole-body immunofluorescence staining and quantification. Injection of IFNGR1 KO fibroblasts into the tail skin of IFNGR1 KO mice did not result in local CD8 T cell aggregation after induction of vitiligo. However, injection of WT fibroblasts alone into the dermis of the tail skin of IFNGR1 KO mice resulted in local CD8 T cell aggregation after induction of vitiligo. The quantification results showed that in the absence of induction of vitiligo, no CD8+ T cells could be detected in IFNGR1 KO skin, regardless of injection of WT or IFNGR1 KO fibroblasts; but after induction of vitiligo, there was an increase in CD8+ T cells only in the RFP+WT fibroblast area, while there were very few CD8+ T cells in the RFP- area of ​​the same skin, similar to the case of injection of RFP+ / -IFNGR1 KO fibroblasts ( Figure 6 A) These findings provide clear evidence that IFN-γ-responsive fibroblasts alone are sufficient to orchestrate local CD8+ T cell recruitment in a local environment that is completely devoid of IFN-γ responses.

[0204] To distinguish whether this effect is dependent on cell-cell contact or secreted factors, we performed an in vitro T cell transwell migration assay using conditioned medium from IFN-γ-treated fibroblasts ( Figure 6 B). First, the maximum activation effect of IFN-γ treatment dose and time was optimized through in vitro assay ( Fig.13B~13D). Primary fibroblasts isolated from WT and IFNGR1 KO skin were then treated in vitro with 1000U / mL IFN-γ for 6 hours, and the conditioned medium was concentrated to different degrees and then added to the lower chamber of the transwell plate. In vitro, OT1 splenocytes obtained from OT1 mice were activated with OVA peptide. The activated T cells were added to the upper chamber of the transwell plate. The transwell migration index of T cells was calculated based on the quantification of the number of cells in the upper and lower chambers by FACS. Among all the culture media tested, only the conditioned medium from IFN-γ-treated WT fibroblasts could induce the migration of CD8+ T cells in a concentration-dependent manner. In contrast, the conditioned medium from IFNGR1KO fibroblasts treated with IFN-γ had no such chemotactic effect ( Figure 6 B) These results suggest that IFN-γ-responsive fibroblasts mediate the recruitment of CD8+ T cells through secreted factors.

[0205] To further identify the secreted factors involved in this process, we first performed RNA-seq analysis on fibroblasts from WT and IFNGR1 KO mice at day 33 after vitiligo induction, as well as from WT mice without vitiligo induction ( Figure 6 C). Bioinformatics analysis showed that among the 432 genes upregulated in WT vitiligo fibroblasts (>2-fold differential expression, p<0.01) compared with WT control fibroblasts and IFNGR1 KO vitiligo fibroblasts, 118 genes encoded secreted proteins. We next compared the single-cell transcriptome data of fibroblasts from patients with progressive vitiligo and fibroblasts from healthy donors. Unsupervised clustering by spectral clustering divided fibroblasts from healthy donors into four subclusters and fibroblasts from patients with progressive vitiligo into five subclusters ( Figure 6 D). In the unique fibroblast cluster present in patients with progressive vitiligo, 59 genes were enriched (>1.5-fold up-regulated, p<0.01) encoding secreted proteins ( Figure 6 E). Comparing the gene signatures of specific secretory proteins from human and mouse vitiligo fibroblasts, we identified 29 overlapping genes. Eleven of these were conserved major histocompatibility complex genes and six were chemokines, including CCL5, CCL8, CCL19, CXCL3, CXCL9, and CXCL10 ( Figure 6 F-6G). To functionally test the ability of these secreted chemokines to promote CD8+ T cell migration, we cultured activated CD8+ OT1 T cells in transwells in the presence of various chemokines in the lower chamber ( Figure 6H). Among the chemokines tested, CXCL9, CXCL10, and CCL19 induced CD8+ T cell migration in a dose-dependent manner.

[0206] Figure 6 A shows representative whole-mount immunofluorescence staining images and quantification results of intradermal fibroblast injection experiments. WT or IFNGR1 KO fibroblasts expressing RFP were intradermally injected into the tail skin of IFNGR1 KO mice. Infiltration of CD8+ T cells was analyzed by whole-mount immunofluorescence staining and quantification after induction of vitiligo in host mice. Injection of IFNGR1KO fibroblasts into the tail skin of IFNGR1 KO mice did not result in local CD8+ T cell accumulation after induction of vitiligo. Injection of only WT fibroblasts into the tail skin of IFNGR1 KO mice resulted in local CD8+ T cell accumulation after induction of vitiligo. Scale bar, 500 μm. Data reflect mean ± SD from 3 independent host mice; Figure 6 B shows a schematic diagram and quantification results of the T cell transwell migration assay. Primary skin fibroblasts isolated from WT or IFNGR1 KO mice were treated with 1000U / mL IFNγ in vitro for 6 hours. The conditioned medium was then concentrated to varying degrees and added to the lower chamber of the transwell plate. In vitro, OT1 splenocytes obtained from OT1 mice were activated with OVA peptide. The activated T cells were added to the upper chamber. After 3 hours of incubation, cells were collected from both chambers and counted by flow cytometry. The migration index of CD8 T cells was calculated by dividing the number of migrated cells (cells in the lower chamber) by the number of migrated cells plus the number of non-migrated cells (cells in the upper chamber + cells in the lower chamber); Figure 6 C shows a heat map of differentially expressed genes in fibroblasts from the tail skin of WT control, WT-induced vitiligo, and IFNGR1KO-induced vitiligo mice. Genes were selected based on >2-fold differential expression (p<0.01) in WT vitiligo compared with WT control and WT vitiligo compared with IFNGR1KO vitiligo; Figure 6 D shows the t-SNE projection of fibroblasts from patients with progressive vitiligo and healthy donors. Unsupervised clustering by spectral clustering divided fibroblasts from healthy donors into 4 subclusters and fibroblasts from patients with progressive vitiligo into 5 subclusters. Each cluster is colored and annotated according to the characteristic gene expression pattern: F1_APCDD1 (red), F2_CXCL12 (green), F3_WIF1 (blue), F4_TNN (purple) and F5_GBP1 (orange). The F5 cluster is uniquely present in patients with progressive vitiligo; Figure 6E shows the volcano plot of differentially expressed genes in fibroblasts of patients with progressive vitiligo compared with healthy donors. The red dots in the volcano plot indicate genes that are upregulated >1.5-fold (p<0.01) in fibroblasts of patients with progressive vitiligo; Figure 6 F shows a Venn diagram of genes encoding secreted proteins that are commonly upregulated in WT-induced vitiligo mouse fibroblasts and progressive vitiligo patient-specific fibroblasts. RNA-seq analysis of fibroblasts from day 33 vitiligo mouse skin identified a total of 118 genes for secreted proteins that are specifically expressed in WT-induced vitiligo mouse fibroblasts compared to control mouse skin (fold change > 1.5 and p-value < 0.01). Single-cell RNA-seq analysis identified a total of 77 genes for secreted proteins that are specifically expressed in progressive vitiligo patient fibroblasts compared to healthy donors (fold change > 1.5 and p-value < 0.01). Overlap of the gene signatures of these two secreted proteins from mouse and human vitiligo fibroblasts revealed 29 overlapping genes, including 11 conserved major histocompatibility complex genes; Figure 6 G shows 29 commonly upregulated secretory protein genes in human or mouse fibroblasts under the indicated conditions ( Figure 6 F) Heat map analysis. Upper panel: Expression patterns of 29 genes in fibroblasts from vitiligo patients and healthy donors. Patient ID and disease status are noted next to them. Lower panel: Expression patterns of 29 genes in fibroblasts from WT or IFNGR1 KO mice with or without vitiligo induction as indicated. Mouse genotype and vitiligo induction status are noted next to them; Figure 6 H shows the quantification of T cell transwell migration assay. Activated T cells were added to the upper chamber of the transwell plate. Increasing concentrations of recombinant proteins of the indicated cytokines were added to the lower chamber. After 3 hours of incubation, the CD8+ T cell migration index was calculated using the number of migrated cells (cells in the lower chamber) divided by the number of migrated cells plus the number of non-migrated cells (cells in the upper chamber + lower chamber).

[0207] Fig.13 A shows a schematic diagram and representative cross-sectional fluorescence images of the intradermal injection experiment of fibroblasts. In vitro, primary fibroblasts isolated from the dorsal skin of WT or IFNGR1 KO newborn mice were infected with lentivirus expressing shRNA and RFP. RFP+ fibroblasts were then injected intradermally into the tail skin of P56 mice. The presence and location of the injected cells were examined 2 weeks later. Note that the injected fibroblasts were mainly located in the lower dermis area in the cross-sectional images; Fig.13B shows QPCR analysis of primary fibroblasts treated with IFN-γ, using the target gene CXCL10 as a readout. Primary mouse fibroblasts were treated with IFN-γ in vitro with increasing doses and durations. The treatment time ranged from 6 to 48 hours; the IFN-γ concentration gradient ranged from 5 U / mL to 200 U / mL. CXCL10 expression was induced by IFN-γ in a dose-dependent manner. However, long-term treatment resulted in a weakening of the induction effect; Fig.13 C shows QPCR analysis of CXCL10 gene expression levels to optimize the IFN-γ treatment dose to achieve maximum activation. In vitro, primary mouse fibroblasts were treated with increasing doses of IFN-γ for 6 hours. The IFN-γ concentration gradient ranged from 5U / mL to 1000U / mL. Based on the expression level of CXCL10, 1000U / mL IFN-γ induced the greatest activation; Fig.14 C shows a representative FACS profile of isolated tail skin dermal fibroblasts for RNA-seq analysis. The epidermis and dermis of the tail skin were enzymatically dissociated by dispase treatment. The skin dermis was then digested with collagenase to obtain a single cell suspension. After immunostaining for CD45 and CD31, fibroblasts were collected as CD45-, CD31- populations using FACS. (Scale bar, 100 μm. Data reflect the mean ± SD of 3 independent experiments with three technical replicates.)

[0208] In conclusion, IFN-γ-responsive fibroblasts are sufficient to mediate CD8+ T cell recruitment in vivo and in vitro via secreted chemokines such as CXCL9, CXCL10, and CCL19.

[0209] Example 7 Differences in intrinsic IFN-γ responses of anatomically distinct human fibroblasts are associated with regional disease variation.

[0210] Our results here indicate that skin dermal fibroblasts are necessary and sufficient to induce local accumulation and activation of CD8+ T cells in response to IFN-γ in vitiligo skin. One of the most interesting features of vitiligo is the bilaterally symmetrical pattern of depigmentation that is common in most patients with nonsegmental vitiligo. We sought to identify some of the local cues involved in the recruitment of immune cells into bilaterally symmetrical body locations.

[0211] We first quantified the frequency of vitiligo in different body locations. Although any skin area on the body may be affected by vitiligo, clinical evidence suggests that depigmentation commonly occurs in several specific body areas. We analyzed the location of lesions in 2265 patients with non-segmental vitiligo from Beijing Hospital and Xijing Hospital in China. The lesion sites were divided into eight major areas: dorsal hands, chest, back, legs, dorsal feet, head, arms, and palms. The frequency of vitiligo in each location was calculated by dividing the number of patients with pathological skin depigmentation in the specified body area by the total number of patients. The results showed that the incidence of vitiligo in the eight body areas varied greatly: the dorsal hands, chest, and back skin areas were most susceptible to vitiligo, while the palms and arms were least susceptible ( Figure 7 A).

[0212] We next sought to determine whether anatomically distinct fibroblasts have intrinsic differences in their IFN-γ signaling responses by performing RNA-seq analysis of primary human skin dermal fibroblasts from eight body locations that were treated with or without IFN-γ in vitro ( Fig.14 , Figure 7 C) HOX gene expression patterns indicate that fibroblasts retain their positional information during in vitro culture ( Figure 7 B). We identified 1195 genes that were upregulated in fibroblasts from at least one body location following IFN-γ treatment ( Figure 7 C). Heat map analysis of the combined up-regulated genes in the eight body locations showed that the up-regulated genes differed between the different body locations. Fibroblasts from the dorsal hand and dorsal foot showed similar IFN-γ response transcriptional signatures. The chest, back, legs, and arms had similar up-regulated genes. The head and palms, respectively, showed transcriptional profiles that were different from the other body locations ( Figure 7 C). These results suggest that fibroblasts from anatomically distinct body locations display intrinsic differences in IFN-γ responses. From the combined lists of upregulated genes, we identified a total of 195 secretory protein genes enriched in IFN-γ-activated fibroblasts, 34 of which overlapped with those identified in fibroblasts from patients with progressive vitiligo, including several chemokine genes ( Figure 7 D). Chemokine genes CCL2, CXCL3, CXCL9, CXCL10, and CXCL11 were upregulated mainly in skin fibroblasts of the dorsum of the hands and feet ( Figure 7 E). QPCR analysis confirmed that fibroblasts from different anatomical locations have intrinsic differences in IFN-γ response ( Figure 7F). Finally, to test whether differences in regional fibroblast-intrinsic IFN-γ responses could explain differences in regional disease incidence, we assessed the association of IFN-γ response enrichment scores with vitiligo incidence at different body locations ( Figure 7 G). The results showed that the incidence of vitiligo was positively correlated with the intrinsic IFN-γ response of skin fibroblasts (R=0.62, p=0.01).

[0213] Figure 7 A shows the quantification of the frequency of lesions in anatomically different skin areas of patients with vitiligo. The incidence of vitiligo in each location was calculated as the number of patients with lesions in that body location divided by the total number of patients, i.e., 2265 patients with non-segmental vitiligo who were registered in Beijing Hospital and Xijing Hospital. The eight anatomical locations included the palms, backs of the hands, arms, chest, back, legs, dorsum of the feet, and head; Figure 7 B shows the QPCR analysis of HOXB8, HOXC8, HOXB13, and HOXD11 in fibroblasts from eight anatomical locations. The results showed that HOXB8 and HOXC8 were enriched in fibroblasts from the chest, back, legs, and arms, while HOXB13 and HOXD11 were enriched in fibroblasts from the back of the hand, back of the foot, head, and palm. The data reflect the mean ± SD of three technical replicates from three independent individuals; Figure 7 C shows a schematic diagram and heat map analysis of fibroblasts isolated from 8 different anatomical locations and treated with in vitro IFN-γ. Responsive genes (fold change > 2 and p-value < 0.01) in fibroblasts from each anatomical location were identified by RNA-seq analysis of fibroblasts treated with 1 U / mL recombinant IFN-γ. Columns show fibroblasts from 8 body locations from 2 individuals; rows show differentially expressed genes sorted by p-value. A total of 1167 differentially expressed genes are shown. Data for each gene were normalized using R's Scale function; Figure 7 D shows a Venn diagram depicting 34 secretory protein genes commonly upregulated in fibroblasts treated with IFN-γ in vitro and in fibroblasts from patients with progressive vitiligo in vivo. A total of 195 secretory protein genes were upregulated in fibroblasts treated with IFN-γ, of which 34 overlapped with secretory protein genes upregulated in fibroblasts from patients with progressive vitiligo; Figure 7 E shows a heat map analysis of 34 commonly upregulated secretory protein genes in IFN-γ-treated fibroblasts from eight different anatomical locations. The skin regions are labeled next to them. Boxes highlight genes that are specifically upregulated in fibroblasts from certain anatomical locations. Data were calculated by calculating the mean from two individuals and normalized for each gene using the Scale function in R; Figure 7F shows QPCR analysis of characteristic genes of vitiligo fibroblasts in fibroblasts from 8 different anatomical locations after IFN-γ treatment. Data reflect the mean ± SD of three technical replicates from three independent individuals; Figure 7 G shows a scatter plot and a linear regression line showing the correlation between the intrinsic IFN-γ response differences of skin fibroblasts in 8 different anatomical locations and the incidence of vitiligo. The intrinsic IFN-γ response differences are reflected by the enrichment score, which is calculated by gene set enrichment analysis (GSEA) using 1167 upregulated genes in fibroblasts in different regions after IFN-γ activation. The results showed that the incidence of vitiligo was positively correlated with the intrinsic IFN-γ response of skin fibroblasts (R = 0.62, p = 0.01).

[0214] Fig.14 A shows QPCR analysis of the IFN-γ target gene CXCL9 in primary human fibroblasts after treatment with the indicated IFN-γ. Primary human fibroblasts were treated with increasing doses and durations of IFN-γ in vitro. The IFN-γ concentration gradient ranged from 0 to 5 U / mL. The results show that 1 U / mL IFN-γ treatment for 3 hours was within the range of log activation. The data reflect the mean ± SD of three technical replicates of three independent experiments; Fig.14 B shows a heat map of upregulated genes in human fibroblasts from different body locations after IFN-γ treatment (log2 fold change > 1, and p value < 0.01). RNA-seq analysis was performed using fibroblasts from 8 body locations of two individuals that received or did not receive IFN-γ treatment. The columns show 8 body locations from 2 individuals, and the rows show differentially expressed genes. 1195 differentially expressed genes are shown. The gene expression values ​​for each gene were normalized by subtracting the mean of all samples from each sample value. The results show that fibroblasts from each location have a specific series of IFN-γ response signature genes; Fig.14 C shows the RNA-seq results of HOX gene expression patterns in fibroblasts from 8 different body locations that received or did not receive IFN-γ treatment. The rows show the 8 body locations from two individuals, and the columns show the expression levels of different HOX genes. Each column is colored based on the HOX gene cluster.

[0215] Results showed that anatomically distinct fibroblasts harbored a specific set of HOX signature genes and that the expression pattern was not altered after IFN-γ treatment.

Claims

1. A method for screening candidate drugs for treating vitiligo using a vitiligo animal model, the method comprising: Injecting a lentivirus expressing shRNA and H2BRFP into the dermis of the tail skin of P1 mice to knock down the IFNGR1, JAK1 and STAT1 genes in the dermis, wherein the fibroblasts infected by the lentivirus are RFP+ fibroblasts; Vitiligo was induced starting from week 9: mice were inoculated with melanoma cells to generate tumors, injected with CD4-depleting antibodies, and the melanomas were removed before they spread and metastasized; On the 33rd day after induction, whole-body immunofluorescence staining was used to analyze the distribution of RFP+ fibroblasts and CD8+ T cells in the dermis of the tail skin, and the distribution of CD8+ T cells and melanocytes in the corresponding epidermis. If the number of CD8+ T cells was negatively correlated with the percentage of RFP+ fibroblasts, and the percentage of melanocyte area was positively correlated with the percentage of RFP+ fibroblasts, it was determined that shRNA that knocked down the IFNGR1, JAK1 and STAT1 genes in the dermis could be used to treat vitiligo.

2. The method according to claim 1, characterized in that The titer of the injected lentivirus was >5×10 8 cfg / mL.

3. The method of claim 1, wherein the melanoma cells are B16F10 melanoma cells.

4. The method according to claim 1, characterized in that The whole immunofluorescence staining includes the following steps: harvesting the entire tail skin and flattening it; cutting the tail skin into 7 to 8 square blocks, placing the middle 3 to 4 blocks in a 20 mM EDTA solution, incubating at 37°C, 80 rpm, for 1.5 hours; using fine-tipped tweezers to quickly remove the epidermis from the dermis in the anterior-posterior direction; this will keep most of the hair follicles in the dermis; using tweezers to remove a small number of hair follicles in the epidermis; flattening the epidermis and fixing it in PBS containing 4% paraformaldehyde for 10 minutes; washing the entire skin in PBS overnight, then permeabilizing it in a methanol solution containing 0.3% H2O2 at -20°C for 20 minutes, and then blocking it in a PBS solution containing 2% normal donkey serum, 1% BSA and 0.3% Triton for 1 hour at room temperature; staining with anti-mouse CD8a antibodies and anti-mouse DCT antibodies.

5. The method according to claim 1, characterized in that Analysis of the distribution of CD8+ T cells and melanocytes in the corresponding epidermis included the following steps: the fluorescence signal of the whole-body immunofluorescence staining was converted into digital information by the Surface function and Spot function in Imaris software, and the position coordinates of each point were analyzed using the smooth scatter package in R software to present the distribution and density of cells.

Images