O08 Using single-cell transcriptomics to characterize early mechanisms of psoriasis resolution

Abstract The use of biological drugs targeting the interleukin (IL)-23/IL-17 axis has greatly improved outcomes in people with psoriasis, with an increasing number achieving clear skin (‘remission’). However, the early mechanisms of biologic-induced psoriasis resolution are yet to be fully explored. Longitudinal sampling of resolving disease tissue provides the opportunity to uncover the dynamic cellular and molecular events triggered by these powerful treatments. Identifying the early sequence of events and key drivers underpinning psoriasis resolution may help to guide future treatment strategies. We performed single-cell RNA sequencing (scRNA-seq) on skin biopsies from five individuals with psoriasis who achieved remission upon treatment with the exemplar IL23p19 inhibitor biologic risankizumab. We profiled whole skin at baseline, and at days 3 and 14 of treatment. We filtered, integrated, scaled and normalized the scRNA-seq data set using R. Ingenuity pathway analysis and the CellChat R package were used to explore canonical pathways and receptor–ligand interactions, respectively. RNA in situ hybridization (RNA ISH) was used to identify the spatial context of cell populations of interest using three independently ascertained patients, sampled at days 0 and 14 of risankizumab treatment. The final scRNA-seq data set comprised 164 741 cells We identified 39 clusters in whole skin, including keratinocyte, fibroblast, myeloid cell and T-cell subsets. We found marked differences in transcript levels between day 0 and day 3 of biological treatment using differential gene expression analysis, with the largest number of modulated genes observed among fibroblast and myeloid cell types. We identified a novel fibroblast population which was reduced in relative abundance at day 14 of treatment in lesional psoriasis skin and absent in nonlesional skin. Pathway analysis of this fibroblast population showed an association with inflammatory signalling, and receptor–ligand interaction analysis revealed proinflammatory cytokine signalling to differentiated keratinocytes. The presence of this fibroblast population in psoriasis skin was validated by deconvolution of independent, publicly available bulk RNA sequencing data sets. Finally, we confirmed the localization of this fibroblast population in the upper dermis of psoriasis skin and the reduced abundance after 14 days of risankizumab treatment using RNA ISH of lesional skin from three independently ascertained patients. Our findings provide a high-resolution atlas of the early changes in the skin microenvironment of resolving psoriasis. We identify a response in a novel fibroblast population in the early stages of biological treatment. We are currently exploring the role of this fibroblast population as a driver cell type in psoriasis resolution.