Professional Documents
Culture Documents
Anatomically Distinct Fibroblast Subsets Determine Skin Autoimmune Patterns
Anatomically Distinct Fibroblast Subsets Determine Skin Autoimmune Patterns
National Institute of Biological Sciences, Beijing, China. 2Peking University–Tsinghua University–National Institute of Biological Sciences Joint Graduate Program, School of Life Sciences,
1
Peking University, Beijing, China. 3Academy for Multidisciplinary Studies, Beijing National Center for Applied Mathematics, Beijing Advanced Innovation Center for Imaging Theory and
Technology, Capital Normal University, Beijing, China. 4Peking University Third Hospital, Beijing, China. 5Department of Dermatology, Xijing Hospital, Xi’an, China. 6Institute of Dermatology,
Chinese Academy of Medical Sciences and Peking Union Medical College, Nanjing, China. 7Department of Dermatology, Beijing Hospital, National Center of Gerontology, Institute of Geriatric
Medicine, Chinese Academy of Medical Sciences, Beijing, China. 8Tsinghua Institute of Multidisciplinary Biomedical Research, Tsinghua University, Beijing, China. 9These authors contributed
equally: Zijian Xu, Daoming Chen. ✉e-mail: changjm0417@126.com; chenting@nibs.ac.cn
80
Melanocytes (%)
20
60
t-SNE 2
0
40
−20 20
g M1
M2 0
−40 1 2 3 4 5 6 7 8 9101 2 3 4 5
Perilesion T cell infiltration region Lesion −50 −25 0 25
t-SNE 1 Patient HD
ll
l
e f g h
el
ce
Tcyt in immune cells (%)
ce a te
K14 PDGFRA
Ke cl e i al c
ng go e
T erh cy
20
En obl yte
40
ll ns
La pha cyt
**
M oth t
d as
us el
br c
. o
Fi ano
M tin
15
DAPI pSTAT1
ra
el
NS
M
10
30
in immune cells (%)
Antigen presentation
5
T cell subtypes
IFNγ signalling
0
20 T cell migration
Per cent of pSTAT1+ cells
Vasculature development 0 20 40 60 80 100
3 **
IFNG normalized
g
ef
cy
re
re
T
T
T
Langerhans cell
4 8 12 >14 1 10 20
Progressive Quiescent Healthy Enrichment −log10(P value) T cell
Fig. 1 | scRNA-seq analysis reveals distinct IFNγ-responsive cell types in the T cells. f, Percentage of CD8+ cytotoxic T cells and expression of IFNG in CD8+
skin of patients with vitiligo. a, Schematic illustration of bilateral symmetric cytotoxic T cells from the skin of patients with vitiligo and healthy donors.
lesions in non-segmental vitiligo. b, Representative bright-field and g, GO categories of genes enriched in each cell type from the skin of patients
immunofluorescent images of vitiligo lesion skin. c, t-distributed stochastic in the progressive state. M. phagocyte; mononuclear phagocyte.
neighbour embedding (t-SNE) plot of melanocytes from the skin of patients h, Representative immunofluorescent images and quantification of pSTAT1
with vitiligo and healthy donor skin. d, Melanocyte composition in each patient signal in different cell types in the skin of patients with vitiligo. Scale bars,
and healthy donor (HD). e, Analysis of T cell subtype composition. Tcyt, 50 μm (b, h). For exact P values, see Source Data. For statistics, P summary and
cytotoxic T cells; Tres, resident T cells; Teff, effector T cells; Treg, regulatory sample sizes, see Methods.
(termed M1 and M2) (Fig. 1c). Upregulated genes (more than twofold; T cells (Extended Data Fig. 1p), T cells in the skin of patients in the
P < 0.01) in melanocytes of the M1 cluster compared to the M2 cluster progressive state did not show an increased IFNγ response compared
showed strong immune responses, especially in terms of the response to healthy skin. This could be related to the low expression level of
to IFNγ (Extended Data Fig. 1i, Supplementary Table 2). Melanocytes IFNGR1 and IFNGR2 in T cells compared to other cell types (Extended
from patients 1–5 were predominantly composed of M1 cluster cells, Data Fig. 1q).
whereas patients 6–10 had a similar composition of melanocytes to that To further validate the IFNγ-responsive cell types, we analysed phos-
of healthy donors (Fig. 1d). On the basis of the expression profile and phorylated (p)STAT1 in different cell types in skin with vitiligo lesions.
composition of melanocytes, we classified patients 1–5 as in a progres- Spatially, the density of CD8+ T cells positively correlated with the den-
sive state, and patients 6–10 as in a quiescent state. sity of pSTAT1+ cells (Extended Data Fig. 1r). Quantification showed
To validate the disease state classifications, we next analysed T cells. that among eight different cell types examined, dermal fibroblasts
Four major sub-clusters of T cells were identified by their expression accounted for most (around 80%) of the pSTAT1+ cells in patient skin
pattern: CD8+ cytotoxic; CD8+ resident; CD4+ effector; and CD4+ reg- (Fig. 1h, Extended Data Fig. 1s).
ulatory T cells (Extended Data Fig. 1j, k, Supplementary Table 3), all
of which were enriched in skin from patients compared to skin from
healthy donors (Fig. 1e, Extended Data Fig. 1l–m). Skin depigmenta- Skin stromal IFNγ response drives vitiligo
tion in vitiligo is mediated by antigen-specific CD8+ cytotoxic T cells To study the functional relevance of our analysis of patient samples, we
targeting melanocytes13,33,34. In patients in the progressive state, there developed a mouse model in which vitiligo was induced by inoculation
were significantly more CD8+ cytotoxic T cells, which also expressed with B16F10 melanoma coupled with depletion of regulatory T (Treg)
significantly higher levels of IFNG (Fig. 1f). cells35, which recapitulated the pathologies of patients with vitiligo
Previous studies showed that IFNγ signalling is important for the pro- (Extended Data Figs. 2, 3). After the same vitiligo induction procedure,
gression of vitiligo in mouse models19–21. Gene ontology (GO) analysis wild-type mice developed vitiligo—as shown by tail skin depigmenta-
of enriched genes (more than 1.5-fold; P < 0.01) in the skin of patients tion, significant loss of melanocytes and abundant infiltration and
in the progressive state showed that melanocytes and fibroblasts were aggregation of CD8+ T cells—but Ifngr1 knockout (KO) mice did not
overrepresented in categories related to the IFNγ signalling pathway develop vitiligo (Fig. 2a, b, Extended Data Fig. 4a–h). This phenomenon
(Fig. 1g, Extended Data Fig. 1n), which was further confirmed by analy- was not due to delayed cytotoxic activity, because even at four months
sis of signature IFNγ response genes in all cell types (Extended Data after vitiligo induction, melanocytes in Ifngr1 KO tail skin were still
Fig. 1o). Although IFNG was predominantly expressed by cytotoxic intact (Extended Data Fig. 4i). RNA-seq and quantitative PCR (qPCR)
T
KO
KO
KO CD8+ T cell FPKM
KO
KO
T
W
T
irradiated
W
Vit. induc.
W
Vit. induc. – – + + – – + +
NS f KO KO KO WT
e 2.4 ***
Melanocytes
KO skin WT skin PDGFRA CD31
1.8
1.2
KO T cell
pSTAT1
0.6
DAPI
0
NS
12 *** Per cent of pSTAT1+ cells
Fibroblast
4 Endothelial cell
Smooth muscle
0 Immune cell
WT KO T cell
Skin: Melanocyte
KO
T
KO
T
W
W
Melanocyte DAPI DCT CD8 Keratinocyte
T cell: KO WT Langerhans cell
Fig. 2 | Paracrine IFNγ signalling mediated by skin stromal cells drives CD8+ adoptive-transfer-based vitiligo model in wild-type and Ifngr1 KO tail skin
T cell cytotoxicity. a, b, Representative density plot images (a) and FACS epidermis. e, Schematic diagram, representative whole-mount images and
quantification of CD8+ T cells (b) in tail skin epidermis of wild-type (WT) and quantifications of melanocytes and CD8+ T cells in skin graft experiment.
Ifngr1 KO mice at day 33 after vitiligo induction (vit. induc.). c, Scatter plot of f, Representative immunofluorescent images and quantifications of pSTAT1
differentially expressed genes in skin CD8+ T cells from WT and Ifngr1 KO mice signal in different cell types in grafted skin after vitiligo induction in the host
after vitiligo induction. FPKM, fragments per kilobase of transcript per million mouse. Scale bars, 500 µm (e), 50 µm (f). For exact P values, see Source Data.
mapped reads. d, Scheme and FACS quantification of CD8+ T cells in For statistics, P summary and sample sizes, see Methods.
analysis revealed that the cytotoxic genes Gzma, Gzmb, Prf1 and Ccl5 target melanocytes were activated using the melanoma–Treg-induced
were significantly downregulated in CD8+ T cells isolated from Ifngr1 vitiligo method (Extended Data Fig. 5d–g). When wild-type and Ifngr1
KO skin compared to wild-type skin (Fig. 2c, Extended Data Fig. 4j). KO full-thickness tail skin were grafted onto the same Ifngr1 KO host
These results indicated that IFNγ signalling is essential for the skin mice, the activated Ifngr1 KO CD8+ T cells were only recruited into the
recruitment, local aggregation and cytotoxicity of CD8+ T cells that wild-type skin, leading to melanocyte loss, but not into the Ifngr1 KO
target epidermal melanocytes. skin (Fig. 2e). Thus, given a local environment of wild-type skin, the
To further confirm this finding, we used an intervention-free, sponta- Ifngr1 KO auto-reactive CD8+ T cells can be efficiently recruited and
neous model of vitiligo using Pmel TCR transgenic mice; this model gen- activated, leading to vitiligo development.
erates antigen-specific CD8+ T cells that recognize melanocytes36,37. The When we carried out the reverse graft experiment, the results were
Pmel;wild-type (Pmel;WT) mice spontaneously developed vitiligo-like unexpected. When wild-type and Ifngr1 KO full-thickness tail skin were
epidermal depigmentation, melanocyte loss and CD8+ T cell infiltration, grafted onto the same wild-type mice, after induction the host-derived
whereas the Pmel;Ifngr1 KO mice did not (Extended Data Fig. 4k–p). wild-type CD8+ T cells were robustly recruited into both grafted
To distinguish paracrine versus autocrine IFNγ signalling in orches- wild-type and grafted Ifngr1 KO tail skin equivalently, causing a loss of
trating the local activity of T cells in vitiligo-affected skin, we used a both wild-type and Ifngr1 KO melanocytes (Fig. 2e). This result directly
model of vitiligo based on the adoptive transfer of T cells20,38, which contradicted the findings of the adoptive T cell transfer experiment, in
relies on sub-lethally irradiated host mice receiving wild-type Pmel which the transferred wild-type Pmel T cells did not infiltrate the Ifngr1
CD8+ T cells and then hPMEL-AAV for the activation of Pmel T cells KO tail skin or lead to melanocyte loss. One possible explanation for
(Extended Data Fig. 4q–s). After receiving the same number of trans- this discrepancy is that wild-type cells from the host skin migrated into
ferred wild-type Pmel T cells (Extended Data Fig. 4t), wild-type recipient the grafted Ifngr1 KO tail skin, and functionally rescued the KO skin in
mice developed vitiligo, as shown by skin depigmentation, significant terms of driving vitiligo pathogenesis.
melanocyte loss and CD8+ T cell infiltration, whereas Ifngr1 KO recipi- A lineage-tracing experiment using C57 tail skin grafted onto
ent mice did not develop vitiligo (Fig. 2d, Extended Data Fig. 4u–x). In Rosa-mT-host mice showed that host dermal cells (mT+) did indeed
all three different models of vitiligo induction, only significantly more migrate into the grafted skin dermis, but neither host keratinocytes
CD8+ T cells were detected in wild-type skin compared to Ifngr1 KO skin, nor melanocytes did (Extended Data Fig. 5h). After induction of viti-
whereas the other five main immune cell types showed no difference ligo in the host mice, whereas no pSTAT1 signal was present in the
(Extended Data Fig. 5a–c). grafted skin of the KO-to-KO graft, in the KO-to-wild-type graft there
The adoptive T cell transfer experiment revealed that parac- were abundant nuclear pSTAT1 signals in the grafted Ifngr1 KO skin;
rine IFNγ-signalling is essential for recruiting and orchestrating quantification showed that most of the pSTAT1+ cells were host-derived
auto-reactive CD8+ T cells in skin. To further narrow down whether fibroblasts (around 70%) (Fig. 2f, Extended Data Fig. 5i). On the basis
IFNγ-responsive skin stromal cells are required for vitiligo progression, of these results we concluded that IFNγ-responsive skin stromal cells
we used a skin graft experiment: full-thickness tail skin was grafted onto are required to drive auto-reactive CD8+ T cell-mediated vitiligo in a
the back of the host mouse, and then host-derived CD8+ T cells that paracrine manner.
+ Vitiligo induction
Keratinocyte Melanocyte
Endothelial cell Fibroblast
Myeloid immune cell T cell
DAPI CD8 RFP
No No. of CD8+ T cells per scale
induction + Vitiligo induction No. of CD8+ T cells per mm2
0 100 200 300 400 Fibroblast 0 20 40 60 80 100 120
WT WT TekCre K14Cre 120
cKO cKO WT
Vit.
No. of melanocytes
100 cKO WT
Melanocyte
induc.
80
per scale
+ KO
60 R = –0.85
Csf1rCre TyrCreER Cd4Cre PdgfraCreER P = 2.1 × 10–44
cKO cKO cKO cKO 40 R = –0.14 WT
20 P = 0.06 RFP+ region
–
KO RFP– region
0
C CL 0
G A
P
X 3
C CL1
G CL9
LY BP
M 15
IS 5R
G P2
G P4
IL P5
C L5
C L8
C 74
PB
TA 6E
C M
IFNγ
2
0 100 200 300
1
C
C
B
D
B
B
B2
X
3
X
2
C
C
2 250
Progressive
80
Migration ratio (%)
Quiescent 0 200
T cells per mm
60 1× 5× 10× 25× Healthy donor –2 150
40
x 0
p
Ly bp
100
C cl3
Is 5ra
M 15
G cl9
C 74
C cl1
Il1 p5
G p2
G p4
Ta e
pb
C m
C l5
C l8
6
2
Vit. induc.
g
D
B2
b
b
b
c
2
3
x
x
C
C
20 1
– WT 50
+ WT 0
0
IFNγ – + – + – + – + – + – + – + – + + Ifngr1 KO 0
–1
Fibroblast KO WT R = 0.86 R = 0.94
P = 9.4 × 10–7 P = 9.9 × 10–10
g Per cent of CXCL9+ h No. of dermal CD8+ T cells per scale
or CXCL10+ cells
0 20 40 60 80 100 Vit. 0 100 200 300
shRNA H2BRFP induc.
Melanocyte P1 Scrambled – NS
Fibroblast Scrambled NS
+
Endothelial
shCXCL9_1 *
+
Muscle P56
shCXCL9_2 ****
+
Keratinocyte
Vitiligo
M. phagocyte shCXCL10_1 ***
+
CXCL9 induction
Langerhans CXCL10 shCXCL10_2 ****
+
P89
T cell Whole mount RFP+ region RFP– region
Fig. 3 | IFNγ-responsive dermal fibroblasts are uniquely required to fibroblast-conditioned medium. e, Heat map analysis of commonly
orchestrate autoimmune CD8+ T cells through secreted chemokines. a, upregulated secreted factors from vitiligo fibroblasts in human and mice.
Skin illustration and representative density plot images of melanocytes in tail f, Correlation analysis of densities of CD3+ T cells and CXCL9+ and CXCL10+ cells
skin epidermis of mice with the indicated genotypes at day 33 after vitiligo from the skin of patients with vitiligo. g, Quantification of the CXCL9+ and
induction. b, Representative whole-mount images and correlation analysis of CXCL10+ signal in different cell types in the skin of patients with vitiligo.
melanocytes versus CD8+ T cells in tail skin epidermis of wild-type and h, Experimental workflow and quantification of dermal CD8+ T cells from
Pdgfra CreER;Ifngr1 fl/fl cKO mice at day 33 after vitiligo induction. c, Experimental in vivo fibroblast knockdown experiment. P, postnatal day. See Extended Data
workflow, representative whole-mount images and quantification of CD8+ Fig. 4g for scale definition. Scale bars, 500 µm (b, c). For exact P values, see
T cells in the fibroblast transplant experiment. d, Schematic diagram and Source Data. For statistics, P summary and sample sizes, see Methods.
quantification of the T cell transwell migration assay using
relative to GAPDH
Hand back 558 24.64% Hand back
Chest 465 20.53%
Expression
Foot back 1.0
Back 460 20.31% Chest
Leg 405 17.88% Back
Foot back 298 13.16% 0.5
Leg
Head 296 13.07%
Arm 201 8.87% Arm
Palm 107 4.72% Head 0
Palm
Total 2,265
ba k
C ck
Ba st
Le k
Ar g
H m
Pa d
lm
ot ac
ea
he
Fo d b
2 0 –2
an
H
T cells (%) per hair follicle
d e NS f
Melanocytes
DAPI DCT P-cad 8 **** 8 Cxcl9 Cxcl10
0.905 Foot back 6
Control
R = 0.62 Chest
Enrichment score
realtive to Ppib
0.900 2 6
Back
Expression
0
0.895 Hand back 4
Arm 15 ** ***
Day 300
Leg 10
CD8+
0.890 Head Vitiligo 2
5
0.885 Palm 0 0
5 10 15 20 25 Paw Dorsal Vit. induc. – + – +
d o aw
ve Paw
al
l
ra
Vitiligo incidence (%)
al
Do l
Dorsal
rs
ra
nt
P
rs
nt
Ve
g h DAPI CD8 RFP i j
60 * Cxcl9 Cxcl10
1× 2× 5× 20 **** T0
Ifngr1 KO
25 6
Enrichment score
Migration ratio
Dorsal
15
realtive to Ppib
40 20 5
T200
Expression
(%)
10 15 4
20 3
5 10 T350
Paw dorsal
2
Fibroblast
5 1
0 0 Normal chemotactic fibroblast
0 0
IFNγ – + – + – + – + – + – + Weak chemotactic fibroblast
Paw dorsal Dorsal Activated fibroblast
Fibrolast Paw ventral Dorsal CD8+ T cell Melanocyte
+ Vitiligo induction
Vitiligo
induction shRNA
+ Vitiligo induction
No. of dermal
200 – Scrambled
+ Scrambled
100 + Ifngr1
+ Jak1
+ Stat1
0
0 25 50 75 100
DAPI CD8 CD8+ T cell DAPI CD8 CD8+ T cell Infected fibroblasts per scale (%)
RFP RFP+ fibroblast RFP RFP+ fibroblast
Fig. 4 | Anatomically distinct fibroblast subsets with differential IFNγ g, Quantification of migration ratio from T cell transwell migration assay using
responses determine organ-level skin autoimmune patterns. IFNγ-treated fibroblast-conditioned medium. h, Representative whole-mount
a, Quantification of lesion frequencies at anatomically distinct skin regions in images and quantification of enrichment score of CD8+ T cells in fibroblast-
patients with non-segmental vitiligo. b, Heat map of 92 upregulated genes in injected area of Ifngr1 KO mice after vitiligo induction. i, qPCR of Cxcl9 and Cxcl10
IFNγ-treated human dermal fibroblasts from eight anatomical regions. c, qPCR of in re-isolated fibroblasts from h. j, Mathematical model predicts the regional
CXCL9 and CXCL10 from IFNγ-treated human dermal fibroblasts in culture. distribution pattern of T cells with mosaic IFNγ-responsive fibroblasts.
d, Correlation analysis of fibroblasts IFNγ response enrichment score versus k, Representative whole-mount and density plot images and correlation analysis
vitiligo incidence frequency. e, Representative immunofluorescent images and of the number of T cells versus the percentage of infected fibroblasts in the in vivo
quantifications of melanocytes and CD8+ T cells in different skin regions in mosaic fibroblast knockdown experiment. See Extended Data Fig. 4g for scale
control and vitiligo-affected mice at day 300. f, qPCR of Cxcl9 and Cxcl10 in definition. Scale bars, 50 µm (e), 500 µm (h, k). For exact P values, see Source
IFNγ-treated-region-specific mouse dermal fibroblasts in culture. Data. For statistics, P summary and sample sizes, see Methods.
a twofold increase in the number of T cells could result in significant To further test whether regional variation in fibroblasts could result
differences in melanocyte loss (Extended Data Fig. 9k). in patterned T cell activity, we built a mathematical model to study the
Next to directly compare the region-specific fibroblasts in CD8+ T cell effect of mosaic fibroblasts on T cell activity (Extended Data Fig. 10a–c,
recruitment in vivo, we intradermally injected RFP-labelled dorsal and paw Supplementary Videos 1–6). In a field of fibroblasts composed of nor-
dorsal wild-type fibroblasts into Ifngr1 KO mice tail skin, then conducted mal and weak chemotactic regions (Fig. 4j, Extended Data Fig. 10d),
vitiligo model induction in the host mice. Quantification showed that there T cells were competitively recruited into the normal chemotactic
was a significantly stronger enrichment of CD8+ T cells in dorsal fibro- region, leading to preferential expansion of T cell clusters in those
blast area than in paw fibroblast area (Fig. 4h). In addition, the re-isolated regions and corresponding melanocyte loss (Supplementary Video 7).
dorsal fibroblasts showed 18 times higher expression of Cxcl9 and 4 times To functionally test this model, we used the fibroblast knockdown
higher expression of Cxcl10 than did paw dorsal fibroblasts (Fig. 4i). approach to create a mosaic field containing fibroblasts with different
Extended Data Fig. 3 | The melanoma–Treg-induced vitiligo mouse model present in mouse. h, i, Quantification and comparison of immune cell types
recapitulates hallmarks of human vitiligo. a, UMAP projection of skin between vitiligo (at Day 33) and control mice. Data in f-g showed the
immune cells from patients with vitiligo and healthy donors. We identified 5 predominantly enriched immune cell type is CD8+ T cell in both patients with
major clusters in immune cells, including two adaptive immune cell clusters vitiligo and the melanoma/Treg-induced vitiligo mouse model, whereas the
(CD4+ T cell and CD8+ T cell) and three innate immune cell clusters other immune cell types did not show significant difference. j, Representative
(Langerhans cell, macrophage, dendritic cell 1/2). Other immune cells of very whole-mount immunofluorescent staining images of CD4+ T cells in WT skin,
low abundance in human skin, such as mast cells and γδ T cells, could also be and Tregs in Foxp3 Cre;Ai14 skin, indicating at Day 33 of vitiligo analysis, the CD4+
observed in our single-cell RNA-seq data, but owing to their limited number T cell population have recovered from systematical CD4+ T cell depletion.
these cell types were not analysed individually. b–e, Heat map (b), dot plot k, Heat map analysis of cytotoxicity-related genes expressed in skin CD8+
(c), violin plot (d) and feature plot (e) analysis of signature genes in each T cells of patients with vitiligo and vitiligo-induced mice, indicating that in
immune cell subtype. f, Quantification and comparison of five immune cell molecular level, this melanoma/Treg-induced vitiligo mouse model elicits
types between progressive-state patients with vitiligo and healthy donors by endogenous auto-reactive CD8+ cytotoxic T cells similar to human vitiligo. Ctl:
scRNA-seq analysis. g, Representative FACS profiles analysing immune cell control, Vit: vitiligo. Scale bars, 200 µm (k). For exact p values, see Source Data.
types in mice vitiligo skin, including epidermal TCRγδhigh and TCRγδlow γδ T cell For statistics, p summary and sample sizes, see Methods.
Extended Data Fig. 4 | See next page for caption.
Article
Extended Data Fig. 4 | Analysis of wild-type and Ifngr1 KO mice with three Ifngr1 KO mice with adoptive T cell transfer-based vitiligo mouse model.
different vitiligo induction methods. a–j, Analysis of WT and Ifngr1 KO q, Schematic diagram. r, Quantification of CD117+ melanocytes and
mice with melanoma/Treg-induced vitiligo model. a, Schematic diagram. CD45+CD3+CD8+ T cells in tail skin epidermis of WT mice after γ irradiation,
b, Representative tail skin images of WT and Ifngr1 KO mice at Day 60 after with or without WT Pmel T cell transfer or AVV-hPMEL intraperitoneal
vitiligo induction. c, d, Representative whole-mount images (c) and injection. These data indicate sub-lethal irradiation alone, irradiation with
quantification (d) of tail skin epidermal melanocytes and CD8+ T cells at Day 33 hPMEL-AAV alone, or irradiation with Pmel CD8+ T cell transfer alone did not
after vitiligo induction. e, f, Representative FACS profile of CD117+ result in melanocyte loss or CD8+ T cell infiltration in WT tail skin epidermis.
melanocytes and CD3+CD8+ T cells (e), and quantification of CD117+ s, Representative immunofluorescent staining images and FACS quantification
melanocytes (f) in the tail skin epidermis at Day 33 after vitiligo induction. of epidermal melanocyte of mice post γ-irradiation 26 days with controls,
g, Whole-mount views of scales in tail skin epidermis (dotted line) in control indicating after 5 Gy γ irradiation leads to ~2-fold increase of melanocyte
and vitiligo mice. This is the unit area we used to quantify melanocyte or CD8+ number in tail skin epidermis compared to untreated mice. t, Representative
T cell density in wholemount analysis throughout this paper. h. Scatter plots FACS profiles and quantification of CD3+CD8+VB13+ Pmel T cells in spleen or
and correlation analysis of melanocyte number versus CD8+ T cell number in skin of WT and Ifngr1 KO mice at Day 26 with or without adoptive transfer-based
each scale of WT and Ifngr1 KO tail skin at Day 33 after vitiligo induction. vitiligo model induction. These data showed although both WT and Ifngr1 KO
i, Representative whole-mount images of DCT+ melanocytes in tail skin host mice contained the same number of transferred WT TCR VB13+ Pmel
epidermis of WT and Ifngr1 KO mice at 4 months after vitiligo induction. j, QPCR T cells in the spleen, only WT host mice exhibit robust VB13+ Pmel T cell
analysis of GZMA, GZMB, PRF1, and CCL5 expression in skin CD8+ T cells isolated infiltration in skin, whereas the Ifngr1 KO host mice did not. u–v, Representative
from WT and Ifngr1 KO mice at Day 33 after vitiligo induction. k–p, Analysis of tail skin images (u, Day 60) and whole-mount immunofluorescent images
WT and Ifngr1 KO mice with Pmel transgenic spontaneous vitiligo mouse (v, Day 26) of WT and Ifngr1 KO mice post vitiligo model-induction. Scale bars,
model. k, Schematic diagram. l–m, Representative tail skin images (l, P70) and 500 µm. w, Representative FACS profiles and quantification of epidermal
whole-mount immunofluorescent staining images (m, P42) of Pmel;WT and CD117+ melanocytes in WT and Ifngr1 KO mice at Day 26 post vitiligo model-
Pmel; Ifngr1 KO mice. n, o, Representative FACS profiles (n) and quantification induction. x, Scatter plots and correlation analysis of melanocyte number
(o) of CD117+ melanocytes and CD45+CD3+CD8+ T cells in tail skin epidermis of versus CD8+ T cell number in each scale in WT and Ifngr1 KO mice at Day 26 post
WT, Ifngr1 KO, Pmel;WT and Pmel;Ifngr1 KO mice at P42. p, Scatter plots and vitiligo model-induction. Scale bars, 500 µm (b, c, g, i, l, m, s, u, v). For exact p
correlation analysis of melanocyte number versus CD8+ T cell number in each values, see Source Data. For statistics, p summary and sample sizes,
scale in Pmel;WT and Pmel;Ifngr1 KO tail skin at P42. q–x, Analysis of WT and see Methods.
Extended Data Fig. 5 | IFNγ-responsive skin stromal cells are required for mice rather than the grafted skin. These data indicate the graft infiltrated
vitiligo progression. a–c, Analysis of six main immune cell types in tail skin of CD8+ T cells were derived from the host mice rather than the grafted skin.
WT and Ifngr1 KO mice with different vitiligo induction methods. a, Day 33 after g, Quantification of melanocytes and CD8+ T cells number in grafted skin
melanoma/Treg-induced vitiligo model induction. b, P42 of Pmel;WT and without vitiligo model induction in host mice. Donor skin pairs grafted onto
Pmel;Ifngr1 KO mice. c, Day 26 after adoptive transfer-based vitiligo the same host mouse are linked by lines. h, Representative immunofluorescent
model-induction. d, Timeline of the graft and vitiligo induction assay. staining of the junction region between grafted C57 tail skin and host dorsal
Representative whole-mount immunofluorescent staining images of grafted skin of a membrane-Tomato (mT) transgenic mouse. After full-thickness C57 tail
WT tail skin epidermis on WT host at Day 0 and Day 21 with or without vitiligo skin was grafted onto membrane-tdTomato (mT)-expressing host mice
induction were showed. These data indicate tail skin did not spontaneously (Rosa-mT), in which all host cells are genetically labelled as mT+, host derived
develop vitiligo after the graft procedure alone; only after vitiligo induction in dermal cells indeed invaded into the grafted skin dermis. But neither K14+
the host mice did the grafted skin develop vitiligo as indicated by melanocyte keratinocytes nor DCT+ melanocytes migrated from the host to grafted skin.
loss and large amounts of CD8+ T cells infiltration. e, f, Lineage tracing i, Representative immunofluorescent staining image of pSTAT1 signal in
experiment of CD8+ T cells: e, Schematic diagram WT tail skin graft on immune cell (CD45+), smooth muscle cell (a-SMA+), Langerhans cells
CD4 Cre;mTmG host and representative immunofluorescent staining images of (Langerin+) of grafted Ifngr1 KO tail skin on WT host after vitiligo induction on a
CD8+ T cells in the grafted skin epidermis after vitiligo induction; f, FACS host mouse. Enlarged image on the right represents Langerin+ cells in the
analysis and quantification of spleen cells from CD4Cre;Ai6 mouse to determine epidermis. Scale bars, 50 µm (d, e, h, i). For exact p values, see Source Data. For
GFP labelling efficiency in CD8+ T cells. Cells were pre-gated on CD45+ live statistics, p summary and sample sizes, see Methods.
singlets, indicating that the infiltrated CD8+ T cells were derived from the host
Article
Extended Data Fig. 9 | Anatomically distinct dermal fibroblasts have in the dorsal and paw dorsal skin of mice at P60. g, Representative
intrinsic differences in the IFNγ response. a, Heat map of upregulated genes immunofluorescent images, and quantification of hair follicle located
in human dermal fibroblasts from different anatomical regions after IFNγ melanocytes in ventral and paw ventral skin of control mice and vitiligo mice
treatment (> 2-fold, t test p < 0.01). b, HOX expression pattern (based on (at Day 300). h, FACS analysis and quantification of skin CD3+CD8+ T cells and
RNA-seq results) in human dermal fibroblasts from eight different anatomic CD117+ melanocytes from paw dorsal, paw ventral, dorsal, and ventral regions
regions with or without IFNγ treatment during in vitro culture. c, QPCR analysis at Day 300 after vitiligo induction. i, Schematic diagram of fibroblast isolation
of HOXB8, HOXC8, HOXB13, and HOXD11 in human dermal fibroblasts from and qPCR analysis of CXCL9 and CXCL10 in fibroblasts from four distinct
eight anatomic regions. d, JAK1, JAK2, IFNGR1, and STAT1 expression pattern regions of long-term vitiligo mice. j, Quantification of migration ratio using
(based on RNA-seq results) in human dermal fibroblasts from eight different medium without fibroblast, corresponding to Fig. 4g. k, Relationship between
anatomic regions with or without IFNγ treatment. e, QPCR analysis of TNFSF10, melanocyte number and infiltrated CD8+ T cell number in each scale after
IL6, and CTSH in human dermal fibroblasts from eight anatomically distinct vitiligo induction based on quantification data in Extended Data Fig. 4h. Scale
regions after IFNγ treatment. f, Left, representative whole-mount images and bars, 100 µm (f), 50 μm (g). For exact p values, see Source Data. For statistics,
quantification of hair follicle density in the dorsal and paw dorsal skin of mice p summary and sample sizes, see Methods.
at P21. Right, representative section image and quantification of skin thickness
Extended Data Fig. 10 | See next page for caption.
Article
Extended Data Fig. 10 | Mathematical modelling reveals that fibroblasts or two directions (third row, fibroblasts have 1/2 normal chemotactic effect,
direct collective CD8+ T cell local activity. a, The mathematic model also shown in Supplementary Videos 5, 6). c, The mathematic model predicts
developed to predict local CD8+ T cell recruitment and clonal expansion large-scale T cell clone expansion over the long term. Four representative time
behaviour. The model is a 3D square lattice with two layers. The upper points from T0 to T450 show the initial state and the subsequent T cell cluster
epidermal layer contains T cells and melanocytes. The lower dermal layer expansion process over 450 time units. CD8+ T cells in the normal chemotaxis
contains fibroblasts with different chemotactic abilities. The CD8+ T cell model efficiently coordinate so as to achieve clonal expansion and melanocyte
population in skin is considered to be a decentralized system. Each CD8+ T cell clearance. CD8+ T cells in the no chemotaxis model fail to undergo clonal
is equipped with the means of sensing a change in density. Over time, the expansion and melanocyte death is detected. CD8+ T cells in the weak
behaviour of each cell changes according to its state and the states of its chemotaxis model (migration ability decreases to 1/2) generate small T cell
neighbouring cells and the surrounding signals. The collective pattern can be clones and expand slowly. d, The mathematic model predicts the T cell clone
globally modulated by changing the parameters governing local cell-cell distribution pattern under regional variant fibroblasts with different levels of
interactions. Once the T cell surrounding a melanocyte exceeds a threshold chemotactic effect to T cell. In this model, the fibroblasts marked in blue have a
number, melanocyte death and IFNγ secretion occur. The IFNγ signal induce normal chemotactic effect, and those marked in green have a weak
neighbouring fibroblast chemotactic effect to recruit nearby CD8+ T cells, chemotactic effect. In this region, T cell chemotaxis (migration ability)
reaching the local CD8+ T cell density threshold for adjacent melanocyte decrease to 1/2 of normal value. Six representative time points from T0 to T600
cytotoxicity and IFNγ secretion. This positive feedback loop between CD8+ show the initial state and subsequent T cell cluster expansion (also shown in
T cells and fibroblasts could ensure T cell clonal expansion and vitiligo Supplementary Video 7). The results show that the T cell clones are more likely
progression. b, The mathematic model predicts the expansion process of a to expand and generate white patches on the normal chemotactic region. Both
single T cell clone. Five representative time points from T0 to T80 show the white patches and T cell clone patterns are highly correlated with the regional
initial T cell cluster state and the subsequent T cell expansion process over 80 fibroblast variants. e, f, QPCR analysis validated high knockdown efficiency
time units. In model 1 with normal chemotactic fibroblasts, full spectrum of (e) and the effect blocking IFNγ downstream signal (f) in vitro for shRNAs
T cell cluster formation and expansion patterns observed in WT vitiligo mouse targeting IFNGR1, JAK1, or STAT1. g, Representative epidermis whole-mount
model were reproduced (first row, fibroblasts with normal chemotactic effect, immunofluorescent staining images and density plot of shRNA-mediated
also shown in Supplementary Videos 1, 2). In model 2, in which the fibroblasts knockdown assay, relative to the corresponding dermis in Fig. 4k. h, Box-
were incapable of chemotaxis, we obtained patterns observed in whisker plots and correlation analysis of T cell number versus percentage of
Pdgfra CreER ;IFNGR1 fl/fl cKO mice. In this model, although T cells could still infected fibroblasts (upper panels), and melanocyte number versus
randomly aggregate in the epidermis, they failed to propagate this effect and percentage of infected fibroblasts (lower panels) in each scale of in vivo mosaic
recruit more T cells to the initial site (second row, fibroblasts with no fibroblast knockdown experiment. i, Scatter plots of median of melanocyte
chemotactic effect, also shown in Supplementary Videos 3, 4). In model 3, in number versus percentage of infected fibroblasts in each scale of in vivo
which the chemotactic effect of fibroblasts was turned down to 1/2 of the mosaic fibroblast knockdown experiment. Scale bars, 500 µm(g). For exact
normal value, random CD8+ T cell aggregates only recruited a limited number p values, see Source Data. For statistics, p summary and sample sizes,
of T cells, resulting in slow T cell cluster expansion and melanocyte loss in one see Methods.
nature portfolio | reporting summary
Corresponding author(s): Ting Chen
Last updated by author(s): Sep 28, 2021
Reporting Summary
Nature Portfolio wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Portfolio policies, see our Editorial Policies and the Editorial Policy Checklist.
Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.
Data analysis Imaris (7.4.2), GraphPad Prism (6.0), R (3.5.1), RStudio (1.1.456), GO analysis (http://geneontology.org), FlowJo (v10.0.7), Seurat (V2.3.2),
TopHat (v2.0.13), Cufflinks (v2.2.1), STAR (v 2.6.1a), FeatureCount (v2.2.1). The code for the vitiligo progression model is available at: https://
github.com/hydrays/vitiligo.
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and
reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Portfolio guidelines for submitting code & software for further information.
Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
March 2021
The single-cell RNA-sequencing data, mouse fibroblast RNA-seq data, and human fibroblast RNA-seq data have been deposited in the GSA with accession number
PRJCA006797 (https://ngdc.cncb.ac.cn/bioproject/browse/PRJCA006797).
Source data for all figures are provided with the paper.
1
nature portfolio | reporting summary
Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf
Data exclusions No data were excluded, with the exception of vitiligo mouse model induction experiment. Mice that fail to developed primary tumors after
intradermally inoculated B16F10 cells were excluded for further analysis. Mice with recurrent primary tumors after surgery were excluded for
further analysis.
Replication For in vivo experiment, biological replicates and independent mice from at least 2 independent litters were used. For in vitro experiments,
biological replicates and technical triplicates were used.
Randomization Participants, mice, and samples were assigned into different experimental groups randomly. In the experiments with specific genotypes,
samples were sorted into groups based on their genotypes. Mice were littermate when available or age matched and co-housed so as to
reduce additional variables beyond their genotype difference.
Blinding Investigator were blinded to the genotype of mice during vitiligo induction and data collection.
Antibodies
Antibodies used anti-human Tyr (Ting Chen Lab made, 1:500)
anti-human KRT14 (Ting Chen Lab made, 1:1000)
anti-human DCT (Ting Chen Lab made, 1:500)
anti-human CD45 (BD, Cat: 11-9459-41, Clone: 2D1, Dilution:1:300)
Anti-human CD117 (BD, Cat 555714, Clone: YB5.B8, Dilution: 1:300)
anti-human pSTAT1 (CST, Cat: 9167, Clone: 58D6, Dilution: 1:500)
anti-human pdgfra (abcam, Cat: ab203491, Clone: EPR22059-270, Dilution: 1:500)
anti-human CD31 (abcam, Cat: ab24590, Clone: P2B1, Dilution: 1:500)
anti-human aSMA (Sigma, Cat: C6198, Clone: 1A4, Dilution: 1:500)
anti-human CD11c (abcam, Cat: ab 52632, Clone: EP1347Y, Dilution: 1:500)
March 2021
2
anti-mouse CXCR3, (Biolegend, Cat: 126506, Dilution: 1:300)
anti-mouse CXCR3, (Biolegend, Cat: 126511, Dilution: 1:300)
The signal of human pSTAT1, human CD8a, human pdgfra, human CD11c, human Langerin, human CD3e, human Foxp3, human
KLRB1, human mouse pdgfra, mouse pSTAT1 was amplified by ABC Kit and TSA Kit.
Validation All the commercial antibodies were validated by the suppliers as indicated in the quality assurance literature provided by the supplier
either through flow cytometry or immunofluorescence assays.
The home made antibodies were validated by immunofluorescence assays in this study (Tyr, 1:500) and previous papers (Krt14,
1:1000, Xu et al, eLife, 2015; DCT, 1:500, Lu et al., eLife, 2020).
Authentication All cell lines were kept at low passages in order to maintain their identity.
Mycoplasma contamination All cell lines were tested negative for mycoplasma contamination.
Commonly misidentified lines No commonly misidentified cell lines were used in this work.
(See ICLAC register)
Ethics oversight Mice were bred and maintained in National Institute of Biological Sciences specific pathogen-free facility in accordance with the
March 2021
Guide for the Care and Use of Laboratory Animals of the NIBS. Procedures were approved by the Laboratory Animal Management
Committee of NIBS and were in compliance with all relevant ethical regulations.
Note that full information on the approval of the study protocol must also be provided in the manuscript.
3
Human research participants
In this study of anatomically distinct human fibroblasts, we collected skin biopsies from human embryos (female, 22 to 24
weeks of gestation).
Recruitment Vitiligo patients who were scheduled to have surgery for pathological analysis in Beijing hospital were recruited if they
consented to donate tissue. All the patients were newly diagnosed with vitiligo and none of the patients had received therapy
or UVB treatment. All vitiligo skin tissue used are based on their availability, which would not affect our conclusion. Healthy
donors who were scheduled to have surgery for plastic operation in Beijing hospital were recruit if they consented to donate
tissue. All healthy donors’ tissue used are based on their availability, which would not affect our conclusion.
Research donors of human embryo were recruited from Peking University Third Hospital. Before giving consent, donors have
a suitable opportunity to receive proper counseling about the implication of the donation and potential risk. Embryos were
collect with written informed consent from the donors in this study. Both human embryo used are based on their availability,
which would not affect our conclusion.
Ethics oversight The study of single cell RNA-seq and immunofluorescent analysis of human vitiligo skin biopsy was reviewed and approved by
Ethics Committee of Beijing Hospital. For minors, informed consent was obtained from the LAR’s of the minors. The study of
anatomically distinct human fibroblast was reviewed and approved by the Peking University Third Hospital Medical Science
Research Ethics Committee.
Note that full information on the approval of the study protocol must also be provided in the manuscript.
Flow Cytometry
Plots
Confirm that:
The axis labels state the marker and fluorochrome used (e.g. CD4-FITC).
The axis scales are clearly visible. Include numbers along axes only for bottom left plot of group (a 'group' is an analysis of identical markers).
All plots are contour plots with outliers or pseudocolor plots.
A numerical value for number of cells or percentage (with statistics) is provided.
Methodology
Sample preparation For FACS quantification and collection of epidermal immune cells including CD8+ T cells, CD4+ T cells, Langerhans cells, and
γδ T cells, skin from middle section of the tail was harvested, flattened, cut into 1x1 cm pieces, and then floated dermis side
down in 2 mL of 2.4 U/mL dispase solution and incubated at 37˚C with shaking at 80 rpm for 50 min. To obtain most of the
epithelial cells, the epidermis was then carefully removed from dermis along the anterior-posterior direction with a fine-
tipped tweezer. The removed epidermis was floated dermis side down in 3 mL TrypLE™ Express Enzyme (Gibco) at 37˚C for
10 min, afterwards 5% FBS was added to neutralize TrypLE.
For FACS quantification of epidermal melanocyte, melanocytes located in the hair follicle must be removed. We found out
prior to skin collection, wax using VEET wax paper facilitates subsequent hair follicle separation from epidermis. Waxed tail
skin was flattened and floated dermis side down in 2 mL of 2.4 U/mL dispase solution at 37˚C with shaking at 80 rpm for 50
min. To efficiently obtain hair follicle-free epidermis, the epidermis was swiftly peeled from the dermis along the posterior-
anterior direction with a fine-tipped tweezer. The occasional remaining hair follicles attached to epidermis were removed
using tweezers. The hair follicle-free epidermis was then floated dermis side down in 3 mL Trypsin-EDTA (0.25%) at 37˚C for
10 min, afterwards 5% FBS was added to neutralize trypsin.
Single cell suspensions were obtained by repeatedly aspirating and dispensing with a pipette for 20 times followed by
March 2021
filtering through 40 μm strainers. The cells were then stained with anti-mouse CD117 (ebioscience, Cat: 13-1171-82, Clone:
2B8, Dilution: 1:200) for melanocyte detection, anti-mouse CD45 (BD, Cat: 559864, Clone: 10-F11, Dilution: 1:300) for
immune cell detection; anti-mouse CD3 (Biolegend, Cat: 100220, clone: 17A2, Dilution: 1:300) and anti-mouse CD8a
(ebioscience, Cat: 17-0081-83, Clone: 53-6.7, Dilution: 1:300) for CD8+ T cell detection; anti-mouse CD3 and anti-mouse CD4,
(Biolegend, Cat: 100451, Dilution: 1:300) for CD4+ T cell detection; anti-mouse MHCII (Biolegend, Cat: 107629, Dilution:
1:300) and anti-mouse CD207, (Biolegend, Cat: 144206, Dilution: 1:300) for Langerhans cell detection; anti-mouse CD3 and
anti-mouse TCRγδ (Biolegend, Cat: 118108, Dilution: 1:300) antibodies for γδ T cell detection.
For FACS quantification and collection of dermal fibroblasts, endothelial cells, immune cells including macrophage and
4
dendritic cells from tail skin, skin from middle section of the tail was harvested, flattened, cut into 1x1 cm pieces, and then
floated dermis side down in 2 mL of 2.4 U/mL dispase solution and incubated at 37˚C with shaking at 80 rpm for 50 min. To
For analysis and isolation of immune cells from spleens, the spleen was mechanically dissociated and the red blood cells were
lysed with RBC lysis buffer according to the manufacturer’s instructions (Beyotime). Single cell suspensions were obtained by
filtering with strainers (40 μm) and then cells were stained with anti-mouse CD3, anti-mouse CD45, and anti-mouse CD8
antibodies. anti-mouse VB13 (BIolegend, Cat: 140703, Clone: MR12-4, Dilution:1:300 ) antibody was used in Pmel T cell
adoptive transfer experiments.
For CXCR3 expression analysis in skin immune cell types of vitiligo mice, the entire tail skin was harvested, flattened, then
floated dermis side down in 5 mL Trypsin-EDTA (0.25%) at 37˚C for 30 min. The skin was inverted and scraped with a scalpel
to get total skin cell suspensions. Single cell suspensions were obtained by filtering with strainers (70 μm and 40 μm), then
cells were co-stained with anti-mouse CXCR3, (Biolegend, Cat: 126506, Dilution: 1:300) against immune cell markers
mentioned above for further analysis.
For FACS analysis of CD8+ T cells and melanocytes from mouse dorsal, ventral, paw dorsal, and paw ventral skin, skin biopsies
were harvested, flattened, and floated dermis side down in 3 mL of TrypLE and incubated at 37˚C with shaking at 80 rpm for
30 min. The skin epidermis was gently scraped with a scalpel to get the epithelial cell suspensions. Single cell suspensions
were obtained by filtering with strainers (40 μm), then cells were stained with anti-mouse CD45, CD3, CD8, and CD117
antibodies for further analysis.
For re-isolation of injected RFP-labeled dorsal or paw fibroblasts from IFNGR1 KO mouse tail, tail skin was separated into
epidermis and dermis post dispase treatment. After digestion of dermis with collagenase and neutralization, single cell
suspensions were obtained by filtering through 70 μm and 40 μm strainers. Then RFP+ fibroblasts were isolated for further
qPCR analysis.
Cell population abundance For sorting for human single cell RNA sequencing, more than 10,000 immune cells and 2,000 melanocytes were collected.
The purity was more than 95% and was verified by single cell RT-qPCR. For sorting for mouse RNA-seq, at least 200,000
fibroblast were collected. The purity was verified by expression of Pdgfra and Krt14 with RT-qPCR. For immune analysis in
mice, 1,000 CD45+ singlets were collected. For CD117+ melanocytes analysis, 20,000 singlets were collected.
Gating strategy In the FACS for single cell RNA sequencing, cells were gated on live (DAPI-) and FSC/SSC- area. CD45+ and c-Kit+ were used to
enrich immune cells and melanocytes; CD45-, c-Kit- cells are predominantly keratinocytes. In dermis, CD45 was used to
distinguish immune cells from dermal cells (Extended Data Figure 1b).
In the FACS for qPCR validation and mouse fibroblast RNA sequencing, cells were gated on live (DAPI-) and FSC/SSC- area.
CD45+ and c-Kit+ markers were used to enrich immune cells and melanocytes; CD45-, c-Kit- cells are predominantly
keratinocytes. In dermis, CD45, CD31 were used to distinguish immune cells and endothelial cells from fibroblast, PDGFRA
were used to enrich fibroblasts (Extended Data Figure 6a).
For immune analysis, cells were gated on live (DAPI-) and FSC/SSC- area. CD45+ were used to enrich immune cells; CD3
+(medium), CD8+ cells are predominantly CD8+ T cells; CD3+(medium), CD4+ cells are predominantly CD4+ T cells;
F4/80+, CD11b+ cells are predominantly Macrophage; CD207+, MHC-II+ cells are predominantly Langerhans cells; TCR γδ+,
CD3+ cells are predominantly γδ T cells; CD11c+, MHC-II+ cells are predominantly dendritic cells (Extended Data Figure 3f).
Positive and negative gates were set using fluorescence minus one (FMO) background intensity controls. Fluorophores were
chosen to minimize spectral overlap.
Tick this box to confirm that a figure exemplifying the gating strategy is provided in the Supplementary Information.
March 2021