Postdoc University of California, Los Angeles Los Angeles, California, United States
Introduction/Rationale: Keloids are characterized by excessive fibroblast activation and aberrant extracellular matrix (ECM) deposition. These features are potentially driven by profibrotic gene programs and pathogenic crosstalk between inflammatory cells and fibroblasts. However, the spatial organization of these cellular interactions and gene programs within keloid tissue has not been fully resolved.
Methods: We applied Xenium spatial transcriptomics (ST) to normal skin and keloid tissues to profile cellular composition and spatial architecture. Secondary clustering resolved fibroblast and keratinocyte subtypes. To define conserved keloid-associated fibroblast programs, we integrated scRNA-seq datasets from keloid and healthy skin across four Fitzpatrick skin types.
Results: ST analysis identified 11 major cell types and five fibroblast subtypes, including mesenchymal, papillary, reticular, pro-inflammatory, and cycling, with mesenchymal fibroblasts markedly enriched in keloids. Cross-dataset integration revealed 98 genes consistently upregulated in keloids; 26 were fibroblast-enriched, including 12 highly expressed in mesenchymal fibroblasts. These genes encode proteins involved in ECM synthesis, degradation, and remodeling. Their coordinated expression pattern suggests that inflammatory signals from cytokine-producing macrophages, T cells, and keratinocytes promote a pathological mesenchymal shift in fibroblasts.
Conclusion: Our integrative spatial and single-cell analysis identifies a distinct mesenchymal fibroblast program enriched in keloids and implicates inflammatory–fibroblast communication as a key driver of keloid formation. Xenium ST enables high-resolution mapping of skin architecture and reveals pathogenic fibroblast states that contribute to keloid development.
