Abstract
Background Cutaneous lupus erythematosus (CLE) is an often disfiguring autoimmune disease affecting systemic lupus erythematosus (SLE) patients and a roughly similar number of CLE-only patients without systemic disease. Effective therapies for CLE are limited, and additional studies are needed to better understand the cellular composition and mediators of CLE pathology in order to improve treatment modalities.
Methods Lesional and non-lesional skin biopsies and peripheral blood mononuclear cells (PBMCs) from 7 patients with active CLE lesions and 14 healthy control skin biopsies were subjected to scRNA-seq on the 10X platform. Cells were clustered using Seurat and annotated according to putative cell type-specific markers. Keratinocytes (KCs), fibroblasts (FBs), T cells, and myeloid cells were sub-clustered and cell subsets annotated. Cytokine signatures were generated by treating KCs and FBs with panels of cytokines and identifying induced genes. For KC and FB sub-clustering analyses, each cell was scored for each cytokine signature. Cell-cell communication was examined using CellPhoneDB to perform ligand-receptor analysis for cell types. To further investigate cellular interactions, imaging mass cytometry and spatial transcriptomics were performed, and cell signatures derived from scRNA-seq were used to plot the proximity of different cell types in discoid lupus lesions. Finally, myeloid cells were identified from skin and PBMC samples and subclustered. Pseudotime analysis was performed on cells bridging the connection between circulating and skin-resident myeloid cells.
Results Both lesional and non-lesional SLE/CLE KCs exhibited strong type I interferon scores. Lesional and nonlesional SLE/CLE skin exhibited dramatic shifts in cell-cell crosstalk; CD16+ DCs were highly enriched in CLE lesions relative to control and were among the most active communicators as expressers of both ligands and receptors. Spatial transcriptomics demonstrated CD16+ DCs localizing most prominently in the superficial dermis, enabling interaction with KCs. This was confirmed using imaging mass cytometry. Pseudotime analysis of paired circulating and skin myeloid cells revealed CD16+ DCs may arise from non-classical monocytes, with discrete shifts in myeloid cell transcriptional states, including a robust IFN education in the skin, detectable across this transition.
Conclusions Non-lesional skin of patients with SLE and CLE exists in a type I IFN-rich, ‘prelesional’ state. This affects gene transcription in all major skin cell types and dramatically alters cell-cell communication. Non-classical monocytes may infiltrate this environment to become CD16+ DCs that engage in crosstalk with diverse cell types as one of the earliest steps in the evolution of CLE.
