Redefining the origins and pathogenesis of epidermal cysts using spatial transcriptomics

Summary: Abstract Body: Hematoxylin-eosin (H&E) staining has been a cornerstone of histopathology for over a century, enabling morphological studies of tissues. However, functional insights into tissue dynamics remain limited. The advent of spatial transcriptomics, allowing molecular and spatial analyses, opens new avenues for understanding complex pathologies. Epidermal cysts, benign tumors common in dermatology, were long believed to originate from hair follicle infundibula. However, their presence in hairless areas such as palms and soles challenges this view. To investigate epidermal cyst origins, we analyzed a specimen using 10x Genomics’ Visium Spatial Gene Expression platform. The cyst epithelium exhibited a unique expression profile distinct from both epidermis and hair follicle infundibula. While keratinization resembled the epidermis, lipid metabolism and immune response pathways were absent. Instead, the epithelium showed active inflammation and tissue remodeling. This discovery contradicts the theory of cyst formation due to clogged pores, providing a new explanation for their occurrence in hairless regions and resolving a long-standing dermatological question. Spatial analysis revealed ripple-like gene expression patterns surrounding the cyst, with an inflammatory core transitioning into tissue remodeling. Functional enrichment analyses highlighted tissue changes undetectable with conventional techniques. These findings underscore spatial transcriptomics’ ability to provide spatiotemporal insights into tissue pathology. Our results challenge traditional paradigms of epidermal cyst origins and highlight the necessity of integrating spatial transcriptomics into dermatopathology. This approach enables unprecedented insights into pathogenesis, advancing precision diagnostics and therapeutic strategies. Yukihiko Kato¹, Chiho Yanai¹, Yoshihiro Umebayashi¹, Tsuyoshi Hachiya² 1. Dermatology, Tokyo Ika Daigaku Hachioji Iryo Center, Hachioji, Tokyo, Japan. 2. Genome Analytics Japan Inc., Shinjuku, Tokyo, Japan. Bioinformatics, Computational Biology, and Imaging