Deep Visual Proteomics Reveals New Drug Targets in Lethal Skin Disease
Deep Visual Proteomics uncovered a hidden molecular driver of toxic epidermal necrolysis, a lethal skin disease.
Toxic epidermal necrolysis (TEN) is a devastating condition that affects an estimated 1.9 million adults per year in the US alone. Yet, until recently, its molecular underpinnings remained unknown, making its treatment challenging.
In late 2024, researchers led by Professor Matthias Mann published a paper in Nature that marked a significant step forward for the TEN community. A Deep Visual Proteomics (DVP) workflow was used to analyze patient tissue samples and systematically map molecular changes associated with TEN. By integrating high-resolution imaging, AI-driven cell classification, laser microdissection and ultra-sensitive mass spectrometry (MS), DVP revealed hyperactivation of the JAK/STAT pathway as a driver of the disease. Subsequently, existing FDA-approved drugs targeting the JAK pathway – known as JAK inhibitors (JAKi) – were tested in seven critically ill TEN patients, all of whom recovered fully and rapidly.
Dr. Falk Schlaudraff, senior research scientist at Leica Microsystems, was part of the team that developed the DVP workflow used in this study. He joined Technology Networks for an interview to discuss the workflow from tissue biopsy to data integration and how DVP could be used for purposes beyond TEN.
TEN is a lethal skin disease. Can you explain what happens in the body when a patient has this diagnosis?
Falk Schlaudraff, PhD (FS):
TEN is one of the most severe skin reactions caused by common medications. It begins as a rash but can quickly progress to widespread skin peeling, resembling a burn affecting over 30% of the body. Underneath, skin cells – especially keratinocytes – suddenly die off. This leads to the outer layer of skin detaching, which is extremely painful and dangerous.
For a long time, doctors didn’t know exactly why this happened. The immune system seemed to be involved, but the molecular drivers were unclear. For this reason, treatment was mostly supportive, keeping patients stable without stopping the disease itself.
MC:
DVP was key to uncovering the role of the JAK/STAT and interferon signaling pathways in TEN. What makes this technique uniquely powerful compared to previous investigative methods?
FS:
DVP is a powerful method that combines spatial context with deep molecular data. Most established methods like spatial multiplexing rely on antibodies to detect known proteins. This requires you to determine what you’re looking for in advance using predefined antibody panels. DVP doesn't work that way. It’s unbiased and can detect thousands of proteins without needing predefined targets. Researchers are able to combine spatial resolution with unbiased proteomic depth.
DVP is antibody-independent and enables the untargeted quantification of thousands of proteins directly from morphologically defined regions, even at the single-cell level.
This makes it uniquely powerful for discovering novel disease mechanisms without making any assumptions beforehand.
In this case, the DVP workflow enabled researchers to analyze thousands of proteins in individual cells from patient tissue. This helped them uncover the molecular mechanisms behind TEN and identify potential therapeutic targets. Through this approach, they discovered that certain immune pathways, particularly the JAK/STAT and interferon signaling pathways, were overactive in TEN.
This insight led to the use of JAKi, which helped patients recover quickly. While I did not contribute to the clinical treatment, I supported the development and application of the DVP workflow. This work enabled the generation of molecular insights that ultimately informed therapeutic decisions, and it’s a discovery that may have remained hidden using traditional, antibody-based methods.
MC:
Can you walk us through the DVP workflow – from tissue biopsy to data integration – and highlight where laser microdissection plays a critical role?
FS:
DVP bridges morphology and molecular content. The workflow begins with a tissue biopsy, typically taken from a patient sample. The tissue is stained to highlight its structures and then imaged at a high resolution. The image is then analyzed using AI to segment the tissue into different cell types or regions of interest.
These regions are then transferred to a laser microdissection (LMD) system, which is where our instrument comes in. It uses a laser to precisely cut out the selected cells without touching the sample. This step is critical because it ensures that only the cells of interest are collected for analysis.
The isolated cells are sent to an ultra-sensitive mass spectrometer, which quantifies thousands of proteins in each of the collected regions. All the data are then integrated to create a map of protein activity throughout the tissue. This provides researchers with a detailed view of the molecular processes occurring in areas where the disease is active. Using DVP, researchers can map cellular pathways and heterogeneity with spatial and molecular precision, from tissue images to proteomic profiles.
MC:
JAKi – drugs already approved for other conditions – were tested on critically ill TEN patients with rapid recovery. How significant is this for the future of drug repurposing?
FS:
The treatment of TEN patients with JAKi was conducted off-label and based on careful risk assessment by the responsible clinicians. While the results were remarkable, with all treated patients recovering, formal approval for TEN would still require the standard regulatory process for drug approval.
By its design, the DVP workflow is especially powerful at identifying disease-specific biomarkers and actionable molecular targets. DVP clearly helps open new therapeutic avenues by providing spatially resolved molecular insights that were previously inaccessible. The strength of DVP lies in its ability to profile complex tissues at single-cell resolution in an unbiased, spatially resolved manner making it a highly versatile tool for investigating a wide range of severe and rare diseases. Whether these findings will lead to further successful drug repurposing depends on many factors.
MC:
Do you see DVP being applicable to uncovering disease mechanisms and therapeutic targets in other severe or rare conditions?
FS:
Yes, absolutely. DVP is already being used for purposes beyond TEN. For instance, in a study of α1-antitrypsin deficiency (AATD), a common liver disease, DVP revealed stress pathways and potential treatments. In the case of pancreatic cancer, DVP identified early biomarkers and new drug targets. Despite being very different diseases, DVP was effective in both cases.
The strength of DVP is that it doesn't rely on prior knowledge. It can analyze complex tissues at the level of individual cells and identify unexpected patterns. This makes DVP ideal for rare or poorly understood conditions where traditional methods might miss key details. I think we’re just beginning to see what DVP can do.
MC:
What are the biggest challenges in scaling DVP – whether technical, analytical or clinical – to make it a widely used tool for drug discovery?
FS:
DVP is rapidly gaining traction but scaling it for widespread use in drug discovery comes with a few key challenges. One challenge is the complexity of the process. DVP involves several advanced steps: imaging, AI analysis, laser microdissection and MS. Each step requires expertise, and the entire workflow must be carefully coordinated.
To scale up, the process must be more automated, robust and user-friendly. This requires dedicated software to handle all the information and standardized protocols. Additionally, we must make data analysis more accessible so that researchers without extensive technical backgrounds can still obtain meaningful results.
If we can achieve these goals, DVP could become a routine tool in translational research and drug discovery. However, achieving this will require collaboration across disciplines.
