T cells promote distinct transcriptional programs of cutaneous inflammatory disease in keratinocytes and dermal fibroblasts

Abstract

T cells and structural cells coordinate appropriate inflammatory responses and restoration of barrier integrity following insult. Dysfunctional T cells precipitate skin pathology occurring alongside altered structural cell frequencies and transcriptional states, but to what extent different T cells promote disease-associated changes remains unclear. We show that functionally diverse circulating and skin-resident CD4⁺CLA⁺ T cell populations promote distinct transcriptional outcomes in human keratinocytes and fibroblasts associated with inflamed or healthy tissue. We identify T_h17 cell-induced genes in keratinocytes that are enriched in psoriasis patient skin and normalized by anti-IL-17 therapy. We also describe a CD103⁺ skin-resident T cell-induced transcriptional module enriched in healthy controls that is diminished during psoriasis and scleroderma and show that CD103⁺ T cell frequencies are altered during disease. Interrogating clinical data using immune-dependent transcriptional signatures defines the T cell subsets and genes distinguishing inflamed from healthy skin and allows investigation of heterogeneous patient responses to biologic therapy.

Figure 1.. Isolation and functional characterization of circulating and skin-resident human CD4⁺CLA⁺ T cell populations.

(a) Experimental design and study schematic: Blood and skin CD4⁺CLA⁺ T_h cell isolation, activation, and subsequent stimulation of keratinocytes (KC) and fibroblasts (Fib), analysis by bulk RNA-seq, and comparison to public gene expression data from human skin. (b) CD4⁺CLA⁺ T_h cell sorting strategy for indicated blood and skin populations. T Conv, conventional T cells; T Reg, regulatory T cells. (c) Quantitative cytokine bead array measurement of stimulated T cell supernatants for the 8 indicated analytes (n=5–6 donors per population). Full statistical analysis of cytokine production is provided in Table S1.

Figure 2.. CD4⁺CLA⁺ T cells induce distinct transcriptional states in epithelial keratinocytes.

(a) Principal component (PC) analysis of healthy donor, primary human KCs cultured for 24 hours with the indicated activated blood- or skin-derived T cell supernatants, compared with matched unstimulated controls. (b) PC analysis showing each blood (n=7) and skin (n=5) healthy donor-treated KC sample. (c,d) Top: Heat map showing z-score expression changes of KC genes (rows) occurring in response to culture with the indicated donor T cell population supernatant (columns). Bottom: Heat map showing cytokine production for each donor T cell population, based on quantification in Figure 1. All measures are scaled by quantile. IL-5, IL-17a, and IL-17f are truncated at the 95% quantile due to extreme outliers (c) The top 20 differentially expressed KC genes are shown in response to each indicated donor T cell population. (d) Expression of 18 well-characterized inflammatory and proliferative response elements in KC averaged across all donor samples for each T cell subset. (e) Dot plot showing functional enrichment analysis of differentially expressed, T cell-induced KC genes within the indicated modules (n=5–6 donors per population). The -log10 adjusted P value is plotted as a statistical measure for enrichment within each module.

Figure 3.. CD4⁺CLA⁺ T cells induce distinct transcriptional states in dermal fibroblasts.

(a) Principal component (PC) analysis of healthy donor, primary human KCs cultured for 24 hours with the indicated activated blood- or skin-derived T cell supernatants, compared with matched unstimulated controls. (b) PC analysis showing each blood (n=7) and skin (n=5) healthy donor-treated Fib sample. (c,d) Top: Heat map showing z-score expression changes of Fib genes (rows) occurring in response to culture with the indicated donor T cell population supernatant (columns). Bottom: Heat map showing cytokine production for each donor T cell population, based on quantification in Figure 1. All measures are scaled by quantile. IL-5, IL-17a, and IL-17f are truncated at the 95% quantile due to extreme outliers. (c) The top 20 differentially expressed Fib genes are shown in response to each donor T cell population. (d) Expression of 18 well-characterized inflammatory and proliferative response elements in Fibs averaged across all donor samples for each T cell subset. (e) Dot plot showing functional enrichment analysis of differentially expressed, T cell-induced Fib genes within the indicated modules (n=5–6 donors per population). The -log10 adjusted P value is plotted as a statistical measure for enrichment within each module.

Figure 4.. T cell-dependent gene signatures are enriched in the epidermis during psoriasis and normalized by anti-IL-17A.

(a) PC analysis of gene expression for paired donor Ps lesional or non-lesional skin compared with healthy controls (HC) from the indicated public data sets. Subjects were treated with an anti-IL-17A therapy – secukinumab (Sec), ixekizumab (Ixe) – or placebo. Merged data sets were batch corrected to adjust for study-specific effects. (b) Heat map showing z-score expression changes of KC genes determined to be T cell-dependent (Figure 2) within public data, as shown in ‘a’ for the indicated groups. (c) Plots showing blood T cell-induced gene set scores, an averaged measure for the effect of each T cell subset on healthy donor KCs per subject group. (d) Radar plot showing T cell-induced KC signature gene enrichment values for the indicated pairwise comparisons from public data with -log10 p-values plotted along the radial axis. (e) Bivariate plot of the log2 fold change between subject groups for T cell-dependent KC genes (Ps vs HC, ‘Disease associated’; Ps vs anti-IL17A, ‘Drug responsive’). Individual genes are colored according to their T cell signature membership. (f) Plots showing batch corrected gene expression for each subject group for genes that have at least one expression-modulating Ps-associated SNP (i.e. eQTL). All statistical measures shown are compared to the Ps Lesional D0 group. Error bars indicate mean ± SD; ns = not significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001 (Student’s t-test).

Figure 5.. T cell-dependent gene signatures are enriched in the epidermis during atopic dermatitis and normalized by anti-IL-4Rα therapy.

(a) PC analysis of gene expression for AD paired donor lesional and non-lesional skin compared with healthy controls (HC) from the indicated publicly available data sets. AD patients were treated with dupilumab (Dupi), an anti-IL-4Rα drug. Merged data sets were batch corrected to adjust for study-specific effects. (b) Heat map showing z-score expression changes of KC genes determined to be T cell-dependent (Figure 2) within public data, as shown in ‘a’ for the indicated groups. (c) Plots showing blood T cell-induced gene set scores, an averaged measure for the effect of each T cell subset on healthy donor KCs per subject group. (d) Radar plot showing T cell-induced KC signature gene enrichment values across the indicated pairwise comparisons from public clinical trial data (AD and HC) with -log10 p-values plotted along the radial axis. (e) Bivariate plot of the log2 fold change for T cell-dependent KC genes within public clinical trial data from AD or HC subjects (AD vs HC, ‘Disease associated’; AD vs anti-IL17A, ‘Drug responsive’). Individual genes are colored according to their T cell signature membership. Error bars indicate mean ± SD; ns = not significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001 (Student’s t-test).

Figure 6.. Skin T cell-induced gene signatures and CD4⁺CLA⁺ CD103⁺ T cells are elevated in healthy subjects and reduced during inflammatory skin disease.

(a-j) T cell-induced Fib gene sets (Figure 3) were used to analyze (a-d) public scleroderma and (e-j) psoriasis skin scRNA-seq data, filtered on cells annotated as ‘fibroblasts’. Ps lesional and non-lesional skin is from paired donors. (a-d) Uniform Manifold Approximation and Projection (UMAP) dimensionality reduction plots showing assignment by (a) subject group and (b) defined Fib subset [GSE195452]. (c) Heat map showing the enrichment of T cell-dependent gene signatures for each Fib population pooled across all subject groups. (d) Plot showing T cell-dependent gene signature enrichment in Fib subsets, pseudobulked at the subject level. (e-g) UMAP plots showing assignment (e) by subject group and (f) skin T cell-dependent gene set enrichment, (g) quantified per group [GSE173706]. (h-j) UMAP plots of patient data [GSE228421] before (Day 0) and after (Days 3, 14) IL-23 blockade (Risankizumab, Rk) showing assignment (h) by subject group and (i) T cell-dependent gene set enrichment, (j) quantified per group. (k-m) Cytometric analysis of blood CD4⁺CLA⁺ CD103⁺ T cells. (k) Representative fluorescence cytometry plots and (l) quantification of T cells in HC (n=15) or SSc (n=19) patient blood, gated on CD103, integrin β7, and pre-gated on CLA^mem, as in Figure 1. SSc samples: circle = limited disease, square = diffuse, X = unclassified. (m) Representative mass cytometry plots and (n) quantification of T cells in HC (n=11) or Ps (n=12) patient blood, gated on CLA^mem, as in Figure S8. Error bars indicate mean ± SD; ns = not significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001 (d, Student’s t-test; l,n, Mann-Whitney U).