O13 Unravelling the molecular pathogenesis of steatocystoma multiplex using spatial transcriptomics and lipidomics

Introduction and aims

Steatocystoma multiplex (SM) is a genodermatosis usually caused by KRT17 variants, presenting at puberty with multiple inflamed cysts that significantly affect quality of life. These cysts resemble sebaceous ducts and contain sebocytes in the wall. We previously observed sebaceous differentiation in the epidermis overlying early cysts, but the molecular pathways remain unknown. Historical thin-layer chromatography suggested that cyst contents resemble sebum, yet modern characterization has not been attempted. Our aim was to investigate SM pathogenesis using spatial transcriptomics and lipidomics of cysts.

Methods

We used Visium HD spatial transcriptomics on formalin-fixed paraffin-embedded samples from early cysts, established cyst wall and overlying epidermis in five patients (four men, 1 woman; aged 21–83 years). Analyses included receptor–ligand interactions (CellPhoneDB), trajectory (Monocle), and transcription factor profiling (Decoupler). Lipidomic profiling of cyst contents, sebum and stratum corneum was performed via mass spectrometry–coupled gas and liquid chromatography, with two normal skin controls.

Results

We identified a strong interaction between S100A8/9 and CD36, a regulator of cell lipid uptake, in the epidermis overlying an early cyst showing sebaceous differentiation. Sebocyte reclustering and trajectory analysis identified a sebocyte subcluster specific to cysts expressing keratinization genes such as CARD18, CDSN and ASPRV1. Transcription factor analysis demonstrated that early keratinocytes show activation of the Wnt pathway.

Lipidomic cyst content analysis revealed two phenotypes with three of five resembling sebum and two of five resembling stratum corneum. Sphingoid bases and N-acylethanoamines were also identified which are infrequently seen in normal sebum.

Conclusions

Early cyst formation may involve alarmin–CD36 interactions, Wnt–β-catenin activation, and sebaceous differentiation in the epidermis. We identified two lipid profiles and two lipid classes rarely seen in normal sebum. Future work will validate these findings with multiplex imaging and pathway targeting in cell models.