Back

Spatial Transcriptomics Identifies Immune-Stromal Niches Associated with Cancer in Adult Dermatomyositis

Anufrieva, K. S.; Shahriari, N.; Gao, C.; Castillo, R. L.; Liu, J.; Prell, S. A.; Kazerounian, S.; Afshari, K.; Kazakova, A. N.; Theisen, E.; Bowman, T.; LaChance, A.; Hashemi, K.; Korsunsky, I.; Rashighi, M.; Vleugels, R. A.; Wei, K.

2025-03-19 immunology
10.1101/2025.03.19.644147 bioRxiv
Show abstract

Adult-onset dermatomyositis (DM) is an autoimmune inflammatory myopathy with distinct cutaneous manifestations and a strong malignancy association. Through comparative analysis with cutaneous lupus erythematosus (CLE), our integrated spatial and single-cell transcriptomics analysis revealed unique immune and stromal niches associated with DM subtypes. Unexpectedly, we found an association between cancer-associated DM skin lesions and the presence of dispersed immune infiltrates enriched with macrophages, CD8+ T cells, plasma cells, and B cells with preserved vascular architecture. In contrast, non-cancer associated DM skin exhibited dense myeloid cell infiltrates, including neutrophils, monocytes, and macrophages, with elevated expression of IL1B and CXCL10 localized near injured vascular endothelia. Cytokines produced by these myeloid infiltrates together with local tissue hypoxia triggered dramatic stromal remodeling, leading to loss of vascular-associated fibroblasts. In addition to the CXCL10+ myeloid signature, non-cancer-associated DM skin with pDC presence showed the emergence of specific cellular pairs: PD-L1-expressing mregDCs and activated Tregs expressing NFKB2 and TNF receptors. While both DM and CLE showed strong interferon signatures, DM uniquely displayed IFN-{beta} expression. Together, our study provides the first comprehensive spatial mapping of immune and stromal cells in adult-onset DM.

Published in Nature Communications (predicted rank #2) · training set

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.