MSACL 2025 Abstract

Spatial Multi-Omics Using DESI Imaging-Guided Laser Microdissection for LC-MS/MS Proteomics

Brittannie Willis (1), Virag Sagi-Kiss (1), Yuchen Xiang (1), Lauren Ford (1), Duncan Roberts (1), Stefania Maneta Stavrakaki (1), Zoltan Takats (1), Nelofer Syed (3), Elizabeth Want (1), Harry Whitwell (1,2)
(1) Section of Bioanalytical Chemistry, Division of Systems Medicine, Department of Metabolism, Digestion, and Reproduction, Imperial College London, London, United Kingdom (2) National Phenome Centre and Imperial Clinical Phenotyping Centre, Department of Metabolism, Digestion and Reproduction, IRDB Building, Imperial College London, London, United Kingdom (3) Department of Brain Sciences, Imperial College London, London, United Kingdom

Brittannie Willis, BSc Medical Sciences, MRes Medicine (Presenter) Imperial College London >> POSTER (PDF)

Presenter Bio: I am a Stratified Medicine PhD student at Imperial College London, based in the Department of Metabolism, Digestion, and Reproduction. My project integrates Analytical Chemistry with Neuro-Oncology under the supervision of Dr Harry Whitwell, Dr Nelofer Syed, and Dr Elizabeth Want.

The aim of my PhD is to develop a spatially resolved multi-omics pipeline to investigate the metabolic and proteomic alterations occurring in glioblastoma following arginine deprivation and radiation therapy. By employing DESI imaging and LC-MS/MS, this research could provide crucial insights into GBM metabolism.

Relevant Financial Disclosures (within past 24 months, reported on Sep 03, 2025) No relevant financial relationship(s) to disclose. Conflict mitigation NOT REQUIRED.

INTRODUCTION:
Glioblastoma (GBM) is a highly aggressive and treatment-resistant brain cancer with a poor prognosis, presenting significant challenges to effective treatment despite interventions including surgery, radiation, and chemotherapy. The tumour's genetic mutations and immune microenvironment contribute to treatment resistance, further complicating therapeutic strategies. To improve clinical outcomes, it is critical to understand GBM's molecular heterogeneity in response to treatment and explore innovative metabolic therapeutic strategies, such as arginine deprivation.

To understand drug responses in GBM, it is essential to consider the spatial context of proteins and metabolites within the tumour and its microenvironment. A powerful approach is the ability to capture metabolomic and proteomic data from the same tissue section. Mass spectrometry imaging (MSI) techniques, such as matrix-assisted laser desorption/ionisation (MALDI-MSI) and desorption electrospray ionisation mass spectrometry imaging (DESI-MSI), are valuable tools for metabolomic and lipidomic analyses. However, they do not provide the same depth of proteomic coverage as liquid chromatography-tandem mass spectrometry (LC-MS/MS).

Furthermore, spatial multi-omics approaches often face challenges, such as poor protein recovery due to laser ablation or interference from embedding matrices, leading to biases in spatial representation. Additionally, many workflows often require consecutive tissue sections, which introduces variability and may cause loss of spatial context. To address these limitations, we have developed a spatially resolved multi-omics pipeline that integrates DESI-MSI for metabolomic imaging with LC-MS/MS-based quantitative proteomics from a single tissue section.

OBJECTIVE:
To develop a spatially resolved multi-omics pipeline that integrates DESI imaging-guided laser microdissection with LC-MS/MS proteomics, enabling enhanced molecular profiling from a single tissue section while preserving spatial context.

METHODS:
Fresh-frozen mouse brain tissue was sectioned at 8 μm thickness and analysed using DESI-MSI on a XEVO-G2-XS-QTOF mass spectrometer (Waters) to acquire spatial metabolomic profiles. Data were acquired in both positive and negative ion modes, with a spatial resolution of 50–100 μm. Data processing was performed using an in-house MATLAB pipeline for spectral preprocessing and peak picking. Regions of interest (ROIs) were identified based on high metabolomic intensities and resected using laser microdissection (Leica LMD7).

Proteins from these resected ROIs were extracted, digested using trypsin, and analysed via LC-MS/MS on a ZenoTOF 7600 mass spectrometer (SCIEX). Peptides were separated using a 2.6 μm Kinetex XB-C18 100 Column, 150 × 0.3 mm (Phenomenex). Both DESI-MSI and LC-MS/MS conditions were optimised to maximise protein recovery while preserving spatial integrity. Database searching was performed using DIA-NN (version 1.8.1) and FragPipe (version 21.1) for protein identification and quantification.

RESULTS:
Our pipeline successfully combines spatial metabolomics and proteomics, preserving sample integrity and enhancing reproducibility. Over 1,100 proteins were identified with high sensitivity and spatial resolution within 10 × 10 μm ROIs, enabling detailed and comprehensive molecular profiling. By eliminating the need for consecutive tissue sections, variability was minimised, allowing for direct comparisons between metabolic and proteomic data. Additionally, optimised DESI-MSI conditions ensured maximal protein recovery, supporting accurate downstream analyses and a deeper understanding of molecular processes.

CONCLUSION:
In conclusion, we demonstrate the basis for a spatially resolved multi-omics workflow that integrates DESI-MSI and LC-MS/MS, enabling high-resolution mapping of both metabolites and proteins from a single tissue section. This approach overcomes the limitations of traditional methods by maximising data extraction while maintaining the spatial resolution, making it particularly advantageous for studies involving rare or limited clinical samples.

Using mouse brain tissue as a model, we highlight the potential of our pipeline to uncover detailed differences in tissue composition through region-specific molecular profiling. This method offers a robust, quantitative approach for tissue profiling, with strong potential for advancing our understanding of therapy response and biomarker discovery. Our approach enables relative quantification by scaling spatial signals against a bulk reference channel and has demonstrated sensitivity in detecting 10 × 10 µm ROIs, ensuring robust and reproducible results. We believe that this workflow has the potential for widespread application in both research and clinical settings, facilitating the broader adoption of spatial multi-omics across diverse tissue types and disease models.