Spatial transcriptomics unveils myeloid cell dynamics in autoimmune skin diseases

Summary: Abstract Body: Spatial Transcriptomics Unveils Myeloid Cell Dynamics in Autoimmune Skin Diseases Single-cell RNA sequencing (scRNA-seq) captures cellular heterogeneity but lacks spatial context. Sequential fluorescence in situ hybridization (seqFISH) outperformed current spatial methods with its subcellular resolution, high sensitivity and specificity in skin samples, but is limited to pre-selected genes. We addressed this limitation by integrating scRNA-seq with seqFISH in the study of dermatomyositis (DM), cutaneous lupus erythematosus (CLE), psoriasis, and vitiligo. Using scRNA-seq data from 44 patients, we designed a 511-gene panel targeting cell type markers and disease-related genes. SeqFISH analysis of 13 sections from 7 skin biopsies profiled ~138,000 cells, detecting 594 transcripts per cell. To examine cell interactions, we developed an asymmetric colocalization heatmap, uncovering typical skin features and disease-specific patterns. Gene expression showed interaction-dependent changes. We also traced skin surface to measure cell depth. This approach provided new insights into myeloid cell biology. Plasmacytoid dendritic cells (pDCs), a source of IFNα, were present in all diseases, but localized deeper in CLE lesions. CD163+ myeloid cells (MCs) transitioned from LYVE1+ among mesenchymal fibroblasts (FBs) to MMP9+ in immune infiltrates near proinflammatory FBs and the superficial dermis, indicating resident MCs gained inflammatory phenotypes in certain regions of the skin. This aligns with the need for stimulated-fibroblast co-culture to induce MMP9 in DCs. IL6, an inducer of MMP9, is highly expressed by proInf FB in top dermis. scRNA-seq showed a substantial expansion of moDC, the source of IFNβ, in DM and lupus. This moDC mapped to MMP9+ MCs based on markers. Our study highlights seqFISH’s value for spatially resolving immune cell functions in autoimmune skin diseases. Yuqing Wang¹, Khashayar Afshari¹, Nazgol Haddadi¹, Chee-Huat L. Eng², Kirsten Frieda², Mehdi Rashighi¹, Manuel Garber¹ 1. University of Massachusetts Chan Medical School, Worcester, MA, United States. 2. Spatial Genomics, Pasadena, CA, United States. Bioinformatics, Computational Biology, and Imaging