Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Dec 10;57(12):2827-2842.e5.
doi: 10.1016/j.immuni.2024.11.010.

Skin immune-mesenchymal interplay within tertiary lymphoid structures promotes autoimmune pathogenesis in hidradenitis suppurativa

Affiliations

Skin immune-mesenchymal interplay within tertiary lymphoid structures promotes autoimmune pathogenesis in hidradenitis suppurativa

Wei-Wen Yu et al. Immunity. .

Abstract

Hidradenitis suppurativa (HS) is a chronic, debilitating inflammatory skin disease characterized by keratinized epithelial tunnels that grow deeply into the dermis. Here, we examined the immune microenvironment within human HS lesions. Multi-omics profiling and multiplexed imaging identified tertiary lymphoid structures (TLSs) near HS tunnels. These TLSs were enriched with proliferative T cells, including follicular helper (Tfh), regulatory (Treg), and pathogenic T cells (IL17A+ and IFNG+), alongside extensive clonal expansion of plasma cells producing antibodies reactive to keratinocytes. HS fibroblasts express CXCL13 or CCL19 in response to immune cytokines. Using a microfluidic system to mimic TLS on a chip, we found that HS fibroblasts critically orchestrated lymphocyte aggregation via tumor necrosis factor alpha (TNF-α)-CXCL13 and TNF-α-CCL19 feedback loops with B and T cells, respectively; early TNF-α blockade suppressed aggregate initiation. Our findings provide insights into TLS formation in the skin, suggest therapeutic avenues for HS, and reveal mechanisms that may apply to other autoimmune settings, including Crohn's disease.

Keywords: CCL19; CXCL13; HS; TLS; TLS on a chip; TNF blockade; hidradenitis suppurativa; inflammatory fibroblast; lymphocyte clonal expansion; single-cell VDJ sequencing; single-cell transcriptomics; tertiary lymphoid structure.

PubMed Disclaimer

Conflict of interest statement

Declaration of interests A patent application derived from this work has been filed and is currently pending.

Figures

Figure 1.
Figure 1.. Immune cell proliferation occurs predominantly in tertiary lymphoid structure (TLS) in HS.
A. H&E image, showing HS lesional skin with dermal tunnel surrounded by dense immune aggregates. Scale bar, 2mm. B. Scoring maps of multiplexed IHC, showing representative TLSs in HS skin (Top, CD19+ and CD19+Ki67+; Bottom, CD3+ and CD3+Ki67+). C. Bar graphs, showing the density of CD3+Ki67+ and CD19+Ki67+ in various areas as indicated. Each dot represents one area counted. Scale bars, 100 μm. D. Graphic summary of the experimental workflow. Whole skin samples were collected from HS patients and healthy donors and processed for scRNA-seq and subsequent computational analyses. The immune cells were further re-clustered to show in E. E. UMAP plot of 17,772 cells showing the immune cell clusters, color-coded by cell types. F. Heatmap, showing representative marker genes for immune cell clusters. cDC, conventional dendritic cell; mDC, migratory dendritic cell. G. Bar graph, showing fractions of the major immune cell types from the normal (N=4), HS perilesional (N=7) and lesional (N=8) skin. H. Bar graph, showing fractions of the major immune cell types from individual samples, including four healthy donors. H, healthy donor; Pt, patient (lesional and perilesional skin). I. UMAP, showing expression of marker genes for T, B, Plasmablast (PB) and Myeloid cells in the MKI67+ cycling cell cluster.
Figure 2.
Figure 2.. In situ clonal expansion of Tfh, Treg, IL17A+ and IFNG+ T cells occurs within TLS.
A. UMAP, showing 5,218 T cells split in four unsupervised and color-coded clusters. B. Feature plot, showing T cells expressing Tfh signature genes and color-coded as indicated. Each dot represents a single cell. C. Bar graph, showing percentage of IL21+ T helper cells in normal, perilesional and lesional skin. Each dot represents one patient. Data are presented as mean. D. Spatial feature plot, showing Tfh expression markers (CD3+CXCL13+BCL6+) within HS lesional skin, either within the immune aggregate (yellow region) or outside the immune aggregate. Dots represent a 10 μm2 area. Scale bars, 300 μm. E. Bar graph, showing percent composition of areas with Tfh signature inside or outside of TLS in D. F. Scoring maps of multiplexed IHC, showing CD4+ and CD4+ FoxP3+ cells in representative areas from HS skin (Top: TLS area; Bottom: Perivascular infiltrate area). G. Bar graph, showing the density of CD4+FOXP3+ Treg cells by multiplexed IHC in areas as indicated. Each dot represents one area counted. H. Representative FACS plot, showing IL17 versus FOXP3 expression in cells from peripheral blood and HS lesional skin (with tunnel), gated on live, singlet T cells. I. Feature plot, showing the distribution of FOXP3+ and FOXP3+IL17A+ cells in cytotoxic (Tc), helper (Th), or cycling T cells. Each dot represents a single cell. Note a small portion of both FOXP3+ and FOXP3+IL17A+ cells are detected to be actively proliferating in skin lesions. J. Feature plot, showing the distribution of IFNG+, IL17A+, and IFNG+IL17A+ cells in cytotoxic (Tc), helper (Th), or cycling T cells. Each dot represents a single cell. K. Bar graph, showing quantification of J, split by normal, HS perilesional and lesional skin. L. UMAP, showing IL17A+ clones (in Pt5), IL17A+IFNG+ clones (Pt6), and IFNG+ clones (Pt6). M. Circos plots, showing paired TRBV and TRBJ for clonally expanded T-cells in HS Pt5, 6, and 7. Stars indicate clonotypes that have individual cells expressing IFNG, IL17A, FOXP3, and/or IL21, as color-coded.
Figure 3.
Figure 3.. In situ clonal expansion of B cell and plasma cells with preferential occurs predominantly in TLSs.
A. Scoring maps of multiplexed IF images, showing CD19+ and CD38+ cells in representative areas from HS lesional skin as indicated. Dashed line, basement membrane. Scale bars, 100 μm. B. Bar graph, showing density of plasmablasts (CD38+Ki67+) cells in representative areas as indicated. Each dot represents one area counted. C. Representative FACS plot, showing the percentage of CD138+ plasma cells in peripheral blood, HS lesional epidermis, and lesional dermis with tunnel, gated on live, singlet CD19+ cells. PC, plasma cell. D. Expression plots of memory (CD27), naïve (IGD), follicular (BCL6, AICDA), and anti-inflammatory (IL10) B-cell signatures in HS B cell cluster with percentage of positive cells. Each dot represents one cell. E. Bar graph showing cell counts of all clonotypes in blood, perilesional and lesional skin samples from four patients. Clonal size (cell number within each clonotype) is grouped by 1, 2, 3–9, and >10 cells and color-coded as shown. The percentage labeled on the top of each bar indicates the percentage of cells from expanded clonotypes (>10 cells). F. UMAPs showing top 5 clones in B cell, plasmablasts (PB), and plasma cells (PC) of patient 6, split by two distinct skin regions. G. Circos plot connecting the CDR3 amino acid sequence and the corresponding IGHV of the top 150 BCR-seq clonotypes from the skin lesion (left) and blood (right) of Pt7. Clonal size are color-coded. Red links: ≥ 10 cells; orange links: 3–9 cells; grey links: 1–2 cells. Star, indicating clonotypes with cells present in both the blood and skin. H. Circos plot, same as in G for Pt8. Note that there was less clonal expansion in the skin lesion of this patient and less starred clonotypes were found in the blood.
Figure 4.
Figure 4.. Plasma cells produce autoreactive antibodies that target keratinocytes in HS lesional skin.
A. Bar graph, showing the percent distribution of plasma cell IgG isotypes. B. Circos plots, showing the top expanded BCR clonotypes from Pt5 and Pt8, with recombined V and J heavy chain genes. Stars indicate clones from different patients that use the same V-J recombination in Ig heavy chains. C. Graphic workflow, showing the generation of HS lesional skin-derived antibodies and examination of their reactivity to skin tissues. D. IF images, showing staining of Pt5-derived antibody on normal, Pt5, Pt8 and Pt25 lesional skin. Scale bars, 20 μm. The skin regions are also displayed at a lower magnification in Figure S4C. E. IF images, showing co-staining of Pt5- and Pt8-derived antibodies on Pt5 lesional skin surface and LOR (Loricrin, labeling stratum granulosum and corneum). Boxed area, split channels are shown on the right. Note that Pt5- and Pt8-derived antibodies show reactivities to the same skin layer, but their staining patterns do not completely overlap. Scale bars, 20 μm. F. IF images, showing direct staining of anti-human IgG antibody to detect the deposition of endogenous IgG in the skin. KRT16 labels keratinocytes. Note IgG deposition in HS surface and tunnel epithelium. Boxed area, shown in higher magnification on the right. Star, showing IgG deposition in the keratinocytes. Scale bars as indicated. G. IF image, showing indirect staining using HS patient serum (1:200), co-stain with KRT16 for keratinocytes. Boxed area, shown in higher magnification at the bottom. Scale bars as indicated. H. IF images, showing staining of neutrophils (MPO, myeloperoxidase), plasma cells (CD38) and keratinocytes (panKRT) in HS lesional tunnel epithelium with a cocktail of DNA-conjugated antibodies by CODEX. Top, neutrophil infiltration into tunnel epithelium. Bottom, neutrophil accumulation at the terminally differentiated keratinocyte layer in tunnel epithelium with formation of neutrophil extracellular traps (NETs) in the lumen. Scale bars, 200 μm.
Figure 5.
Figure 5.. CXCL13 and CCL19 are highly expressed by HS lesional fibroblasts
A. UMAP of 7,849 fibroblasts, color-coded by the four unsupervised clusters. B. Heatmap of the top 15 differentially expressed genes (DEGs; log2 fold change >1) for the four fibroblast clusters. C. Expression plots of CXCL13 and CCL19 in fibroblast clusters. D. Fractions of the four fibroblast clusters in different sample types. E. Fractions of the four fibroblast clusters in each patient (including lesional and perilesional skin). F. Monocle 3 pseudotime analysis of the four fibroblast clusters. “1” represents an arbitrarily selected starting point of trajectory calculations. G. Monocle 3 psuedotime analysis of the (left) CXCL13+ and reticular fibroblast clusters and the CCL19+ and papillary clusters (right), with corresponding upregulation of genes through the pseudotime analysis. H. Spatial transcriptomic analysis of CXCL13 and CCL19 expression in PDGFRA+ populations within HS lesional skin, either within the immune aggregate (white dash) or outside the immune aggregate. Dots represent 10 μm spaces. Scale bars, 300 μm. I. Quantification percent composition of CXCL13+ PDGFRA+ and CCL19+ PDGFRA+ spatial dots from H.
Figure 6.
Figure 6.. Skin fibroblasts upregulate CXCL13 and CCL19 expression in response to immune cytokines and promote lymphocyte aggregation
A. Graphic summary of potential ligand-receptor interactions between immune cells and fibroblasts based on scRNAseq results, showing receptors expressed on HS fibroblasts and their respective ligands (cytokines) produced by the immune cells. The size of the dots presents the relative expression level as in Figure S6A-B. Created with BioRender.com. B. Workflow of HS lesional skin culture experiment. Surgically excised lesional skin tissues were placed in culture dishes as explants, which can be used for cytokine treatment as shown in Figure S6C-D and allow fibroblasts to migrate out from tissue and be used for experiments shown in Figure 6C-D and 6G-K. Scale bars, surgical excision (top) 4cm, fibroblasts (bottom) 50 μm. C. Bar graph, showing the expression of CXCL13 in fibroblasts from normal abdomen, normal groin, and HS lesional fibroblasts (N = 3) in response to various cytokines by RT-qPCR. D. Bar graph, showing the expression of CCL19 in fibroblasts from normal abdomen, normal groin, and HS lesional fibroblasts (N = 3) in response to various cytokines by RT-qPCR. E-F. Image and graphic diagram of a PDMS-etched microfluidic chip with two compartments: fibroblasts are stationary in the outer ring, and a mixture of B and T cells is cultured within the inner space. The two compartments are separated by pillars with spacing to allow media exchange. Scale bar, 5mm in E. G. IF Images, showing CellTrace-violet staining of B and T cell mixture in the inner space of the microfluidic device, with normal or HS fibroblasts in the outer ring. Scale bars, 100 μm. H-I. Quantification of G, showing area (H) for aggregate size and cell number (I) (N=17, triplicate each). J. IF images, showing the formation of lymphoid aggregates with T cell in red and B cells in blue in response to rhCXCL13 and rhCCL19 and their respective blocking antibodies, as indicated. Scale bars, 200 μm. K. Quantification of J, showing areas as the size of lymphoid aggregates (Experimental repeats, N=4).
Figure 7.
Figure 7.. Immune-mesenchymal positive feedback loop is crucial for the initiation of lymphocyte aggregation.
A. Bar graph, showing mRNA expression of TNF, CXCR5 and CCR7 by B and T cells in the absence and presence of CXCL13 and CCL19, respectively (N=3). B. Graphical diagram, showing HS fibroblasts express high level of CXCL13 and CCL19, which may upregulate the expression of their respective receptors (CXCR5 and CCR7) in B and T cells, as well as the expression of TNF from both B and T cells, together forming a signaling feedback loop. C. Scheme of the experiments: anti-TNFα antibody was added to the co-culture at the different time points as indicated by red arrows (Day1,2, and 3). IF Images were taken on Day4, showing the aggregate formation. Scale bars, 200 μm. D. Quantification of C, showing area as the size of lymphoid aggregates (N=4).

Comment in

References

    1. Gago da Graca C, van Baarsen LGM, and Mebius RE (2021). Tertiary Lymphoid Structures: Diversity in Their Development, Composition, and Role. J Immunol 206, 273–281. 10.4049/jimmunol.2000873. - DOI - PubMed
    1. Pipi E, Nayar S, Gardner DH, Colafrancesco S, Smith C, and Barone F (2018). Tertiary Lymphoid Structures: Autoimmunity Goes Local. Front Immunol 9, 1952. 10.3389/fimmu.2018.01952. - DOI - PMC - PubMed
    1. Sautes-Fridman C, Petitprez F, Calderaro J, and Fridman WH (2019). Tertiary lymphoid structures in the era of cancer immunotherapy. Nat Rev Cancer 19, 307–325. 10.1038/s41568-019-0144-6. - DOI - PubMed
    1. Fridman WH, Meylan M, Pupier G, Calvez A, Hernandez I, and Sautes-Fridman C (2023). Tertiary lymphoid structures and B cells: An intratumoral immunity cycle. Immunity 56, 2254–2269. 10.1016/j.immuni.2023.08.009. - DOI - PubMed
    1. Morbe UM, Jorgensen PB, Fenton TM, von Burg N, Riis LB, Spencer J, and Agace WW (2021). Human gut-associated lymphoid tissues (GALT); diversity, structure, and function. Mucosal Immunol 14, 793–802. 10.1038/s41385-021-00389-4. - DOI - PubMed

Publication types