DVP workflow and cell-type-specific proteome of keratinocytes in cutaneous drug reactions. From Nordmann et al., 2024. “Spatial proteomics identifies JAKi as treatment for a lethal skin disease“, Nature 635, 1001–1009 (2024). https://doi.org/10.1038/s41586-024-08061-0. Licensed under the terms of the Creative Commons CC-BY 4.0 license.

Toxic epidermal necrolysis (TEN) is a severe, life-threatening cutaneous adverse drug reaction characterised by extensive epidermal detachment and a 33% fatality rate. Despite various proposed mechanisms, the main drivers of keratinocyte cytotoxicity and epidermal detachment remain poorly understood, with no established treatment beyond supportive care. In a groundbreaking study, Nordmann et al. from the Mann lab at the Max Planck Institute of Biochemistry aimed to address this knowledge gap by investigating molecular mechanisms of cutaneous drug reactions and identifying potential therapeutic interventions for TEN.

To to identify inflammatory drivers and therapeutic targets, Nordmann et al. employed Deep Visual Proteomics (DVP), a spatial proteomics technique combining high-content imaging, AI-guided cell segmentation, and laser microdissection with mass spectrometry to analyse FFPE lesional skin biopsies from patients with varying severities of cutaneous adverse drug reactions to explore the spatial proteome of keratinocytes and immune cells.

Samples were analysed using the Evosep One LC system using the Whisper 40 SPD method, coupled to the Orbitrap Astral with an EASY-Spray source in data-independent acquisition (DIA) mode. Peptides were separated using an IonOpticks Aurora Elite XT 15×75 C18 UHPLC column, enabling sensitive detection from minute FFPE tissue regions. Approximately 5,000 proteins were quantified per cell type, while the phosphoproteomic analysis measured over 16,000 class I phosphosites.

A key finding was the marked upregulation of the JAK/STAT pathway in TEN, particularly in keratinocytes adjacent to the epidermal detachment sites compared to less severe drug reactions. Specifically, STAT1 was identified as one of the main drivers, with a prominent interferon signature in both keratinocytes and immune cells, particularly macrophages. Leveraging these findings, the study demonstrated that JAK inhibitors effectively treated TEN in in vitro models, mouse models, and seven human patients who all survived after treatment.

This research represents a successful translation of spatial proteomics to clinical benefit, establishing JAK inhibitors as a promising therapeutic approach for TEN and potentially other inflammatory conditions. By mapping spatially resolved molecular changes directly to treatment response, this work illustrates how deep tissue proteomics can close critical gaps between disease diagnosis and the development of targeted interventions, while the use of robust DIA workflows and automated phospho-enrichment reflects a broader shift towards high-throughput, clinically actionable proteomics. Looking ahead, Deep Visual Proteomics holds significant promise not only for advancing dermatological therapeutics but also for addressing a wide array of tissue-specific diseases where localised molecular insights are critical.

Publication
Nature

Authors

Thierry M. Nordmann, Holly Anderton, Akito Hasegawa, Lisa Schweizer, Peng Zhang, Pia-Charlotte Stadler, Ankit Sinha, Andreas Metousis, Florian A. Rosenberger, Maximilian Zwiebel, Takashi K. Satoh, Florian Anzengruber, Maximilian T. Strauss, Maria C. Tanzer, Yuki Saito, Ting Gong, Marvin Thielert, Haruna Kimura, Natasha Silke, Edwin H. Rodriguez, Gaetana Restivo, Hong Ha Nguyen, Annette Gross, Laurence Feldmeyer, Lukas Joerg, Mitchell P. Levesque, Peter J. Murray, Saskia Ingen-Housz-Oro, Andreas Mund, Riichiro Abe, John Silke, Chao Ji, Lars E. French, & Matthias Mann;

Title

Spatial proteomics identifies JAKi as treatment for a lethal skin disease