Tofacitinib blocks IFN-regulated biomarker genes in skin fibroblasts and keratinocytes in a systemic sclerosis trial

Abstract

BACKGROUNDSystemic sclerosis (SSc) is an autoimmune, connective tissue disease characterized by vasculopathy and fibrosis of the skin and internal organs.METHODSWe randomized 15 participants with early diffuse cutaneous SSc to tofacitinib 5 mg twice a day or matching placebo in a phase I/II double-blind, placebo-controlled trial. The primary outcome measure was safety and tolerability at or before week 24. To understand the changes in gene expression associated with tofacitinib treatment in each skin cell population, we compared single-cell gene expression in punch skin biopsies obtained at baseline and 6 weeks following the initiation of treatment.RESULTSTofacitinib was well tolerated; no participants experienced grade 3 or higher adverse events before or at week 24. Trends in efficacy outcome measures favored tofacitnib. Baseline gene expression in fibroblast and keratinocyte subpopulations indicated IFN-activated gene expression. Tofacitinib inhibited IFN-regulated gene expression in SFRP2/DPP4 fibroblasts (progenitors of myofibroblasts) and in MYOC/CCL19, representing adventitial fibroblasts (P < 0.05), as well as in the basal and keratinized layers of the epidermis. Gene expression in macrophages and DCs indicated inhibition of STAT3 by tofacitinib (P < 0.05). No clinically meaningful inhibition of T cells and endothelial cells in the skin tissue was observed.CONCLUSIONThese results indicate that mesenchymal and epithelial cells of a target organ in SSc, not the infiltrating lymphocytes, may be the primary focus for therapeutic effects of a Janus kinase inhibitor.TRIAL REGISTRATIONClinicalTrials.gov NCT03274076.FUNDINGPfizer, NIH/National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) R01 AR070470, NIH/NIAMS K24 AR063120, Taubman Medical Research Institute and NIH P30 AR075043, and NIH/NIAMS K01 AR072129.

Keywords: Clinical Trials; Rheumatology; Skin.

Figures

Figure 1. Transcriptomes and proportions of cell populations of study participants.

Combined analysis of scRNA-Seq transcriptomes of skin from 15 participants with dcSSc at baseline and 6 weeks after treatment with tofacitinib (n = 10) or placebo (n = 5; A). Cell clusters (n = 49) are numbered with cell types based on known marker genes indicated to the right. The proportion of cells in each cluster by subgroups of participants: placebo-treated baseline (PBO-Bsl) and 6-week (PBO-W6) and tofacitinib-treated baseline (TOFA-Bsl) and 6-week (TOFA-W6) biopsies (B). Stars indicate pericyte and endothelial cell clusters showing increased proportions of cells after tofacitinib (P < 0.05 by paired 2-tailed t test corrected for multiple comparisons by Bonferroni’s method).

Figure 2. IPA of scRNA-Seq from fibroblast populations.

Pathway analysis of fibroblast scRNA-Seq data from baseline study biopsies (n = 15), analyzed together with scRNA-Seq data from previously described dcSSc (n = 12) and healthy skin (n = 10). Selected pathways from clusters 1 and 9, clustered with analogous cells in previous studies, representing SFRP2/DPP4 fibroblasts (A) and clusters 6 and 7 clustered with analogous cells in previous studies representing MYOC/CCL19 fibroblasts (B). Genes correlating with baseline mRSS (uncorrected P < 0.05) were included in the pathway analysis. Only selected significant pathways (–log P < 1.4) are indicated. SFRP2/DPP4. Yellow bars indicate positive associations with IPA-expected direction of regulation; blue bars show negative associations with expected direction of regulation; and gray bars indicate no expected direction of regulation.

Figure 3. Genes and pathways changing in tofacitinib-treated patient SFRP2/DPP4 fibroblasts.

Pathway analysis of scRNA-Seq data from tofacitinib-treated baseline compared with week 6 gene expression (n = 10) by SFRP2 fibroblasts (clusters 1 and 9; A). Average gene expression in these clusters (pseudo-bulk gene expression) showing decreased expression at week 6 compared with baseline were included in the IPA (uncorrected P < 0.05). Only selected significant pathways (–log P < 1.4) are indicated. IPA used right-sided Fisher’s exact test to calculate the significance scores (shown on the y axis). Blue bars indicate pathways downregulated (z score less than –2), orange bar upregulated (z score greater than 2), white bars without direction of regulation, and gray bars with no expected direction of regulation after tofacitinib compared to baseline. The intensity of the shading indicates the level of the z score. Heatmap of gene expression of the genes associated with the IFN pathway seen in A (B). Changes in inflammatory gene signatures at week 6 compared with baseline in the placebo and tofacitinib groups for SFRP2/DPP4 fibroblasts (C). Clustering of changes in pseudo-bulk gene expression in SFRP2/DPP4 fibroblasts at week 6 compared with baseline in tofacitinib-treated participants of all (filtered) genes (D), showing IFN-regulated genes clustering with STAT1 (indicated by a red star).

Figure 4. Genes and pathways changing in tofacitinib-treated patient CCL19/MYOC fibroblasts.

Pathway analysis of scRNA-Seq data from tofacitinib-treated patients at week 6 compared with baseline gene expression (n = 10) by CCL19/MYOC fibroblasts (clusters 6 and 7; A). Average gene expression in these clusters (pseudo-bulk gene expression) showing decreased expression at week 6 compared with baseline were included in the IPA (uncorrected P < 0.05). Only selected significant pathways (–log P < 1.4) are indicated. IPA used right-sided Fisher’s exact test to calculate the significance scores (shown on the y axis). Blue bars indicate pathways downregulated (z score less than –2), orange bar upregulated (z score greater than 2), white bars without direction of regulation, and gray bars with no expected direction of regulation after tofacitinib compared to baseline. The intensity of the shading indicates the level of the z score. Heatmap of gene expression of the genes associated with the IFN pathway seen in A (B). Changes in inflammatory gene signatures at week 6 in the placebo and tofacitinib groups for CCL19/MYOC fibroblasts (C). Clustering of changes in pseudo-bulk gene expression in CCL19/MYOC fibroblasts at 6 weeks compared with baseline in tofacitinib-treated participants of all (filtered) genes (D), showing IFN-regulated genes clustering with JAK2 and STAT1 (indicated by red stars).

Figure 5. Reduction of type I and type I IFN and other inflammatory signatures by tofacitinib in epidermal keratinocytes.

Changes in inflammatory signature by week 6 in the placebo and tofacitinib groups for basal (KRT14), differentiated (KRT10), and keratinized (FLG) epidermal keratinocytes (A–C). Fold change was computed using median value in baseline versus week 6 groups with P value calculated using Wilcoxon rank sum test. (D) Dot plot showing the most significantly enriched functions among genes only downregulated in the tofacitinib (but not placebo) group. Only significant results (i.e., FDR ≤ 1%) are shown.

Figure 6. Downregulated expression of IFN-regulated genes after tofacitinib in fibroblast populations.

Dot plots showing markers for the SFRP2 (cluster 1 and 9), MYOC (cluster 6), CCL19 (cluster 7), as well as CRABP1 (dermal papilla, cluster 36) and ANGPTL7 (cluster 41) fibroblasts. IFN-regulated genes (IFI35, IFITM1, IFITM3, OAS1, and MX1) decreased after tofacitinib in clusters 1/9 (indicated by aqua bar) and in clusters 6/7 (IFNAR2, ISG15, IFI44L, and OAS3), indicated by maroon bar), but not in myofibroblasts (A). The lack of effect of tofacitinib treatment on expression of genes associated with myofibroblast differentiation in individual participants (B).