Descripción del proyecto
Health and disease states result from dynamic cellular interactions within spatially defined regions in tissues and organs. In diseases such as cancer, these interactions are often disturbed, but their systematic analysis with respect to their impact on the proteome, a close proxy for cellular function, has so far remained elusive. To overcome this major bottleneck in molecular biosciences, I propose to develop and apply Deep Spatial Proteomics (DSP), a multimodal strategy, which for the first time will link distinct cellular neighbourhoods within biological samples to functional proteome states. DSP will combine multiplex immunofluorescence imaging and machine-learning driven cellular neighbourhood profiling with single-cell sensitivity mass spectrometry (MS) based proteomics. Our preliminary data support the feasibility and strong potential of DSP to uncover novel disease mechanisms, drug targets and predictive biomarkers. After development and rigorous benchmarking, we will apply DSP to an already available retrospective cohort of advanced head and neck squamous cell carcinoma, where response rates for anti-cancer immunotherapy are only below twenty percent. The correlation of cell states and spatial neighbourhoods with clinical outcomes will allow us to identify cell communities of highest likelihood to be critical for treatment response and hence patient survival. Through their functional characterisation by deep MS based proteomics, we will not only gain unique biological insights into immunotherapy resistance and potential therapeutic targets, but also identify predictive candidate markers to improve patient stratification. This new concept will have strong implications for basic and translational research, far beyond the study of cancer immunotherapy. DSP could pave the way for a plethora of spatial proteomics applications with countless opportunities for discovery-driven biomedical research.