Unraveling cell signaling intricacies with affinity mass spectrometry for therapeutic development

Cell signaling governs fundamental biological processes and is frequently dysregulated in disease. This thesis investigates how signaling networks are perturbed in metabolic disorders, inflammatory bowel disease (IBD), and cancer, and how cells adapt to such perturbations through compensatory mechanisms. Using multi-dimensional proteomics, including phosphoproteomics, ubiquitylomics, and affinity proteomics, I explored context-specific signaling rewiring in cells, tissues, and patient-derived models. In metabolic disease (Chapter 2), phosphoproteomics revealed how obesogenic stress remodels bile acid synthesis, β-oxidation, and lipogenesis, illustrating adaptive signaling responses to preserve lipid homeostasis. In Chapter 3, we found that knocking out MUC13, a glycoprotein overexpressed in IBD and colorectal cancer (CRC), paradoxically led to increased expression and membrane localization of tight junction proteins. This upregulation strengthens the epithelial barrier, exemplifying the intestine’s compensatory mechanism to maintain tissue integrity. In CRC organoids (Chapter 4), we showed that FBXW7 hotspot mutations impair its ubiquitin ligase function, leading to the accumulation of EGFR, MYC, and STAT3. This enhances proliferative signaling, a key oncogenic process. Intriguingly, FBXW7-mutant organoids also upregulate EGF-like ligands (AREG, EREG) at the transcript level to sustain EGFR signaling even under low-ligand conditions. These findings highlight intricate feedback loops and adaptive mechanisms sustaining essential signaling under stress or mutation. To systematically characterize such signaling adaptations, we leveraged state-of-the-art mass spectrometry (MS) techniques. Enrichment strategies for phosphopeptides, ubiquitin remnants, and membrane proteins improved coverage of low-abundance proteins and PTMs, uncovering key molecular vulnerabilities. However, truly unraveling compensatory signaling requires resolving heterogeneity at single-cell resolution. Emerging single-cell proteomics (SCP) technologies, featuring ultra-sensitive MS (e.g., timsTOF SCP), advanced chromatography, and DIA-based quantification enables the detection of thousands of proteins across individual cells. While SCP faces challenges in detecting PTMs due to limited peptide abundance, advances in multiplexing and large-scale analyses offer promising workarounds. Moreover, spatial proteomics complements SCP by preserving subcellular context, mapping region-specific protein functions in intact tissues. Novel approaches such as single-cell deep visual proteomics (scDVP) further integrate spatial, temporal, and PTM information, providing nuanced insights into signaling dynamics in native environments. Looking ahead, integrating these proteomic data with genomics, transcriptomics, and metabolomics will be essential for a systems-level understanding. However, interpreting high-dimensional, multi-modal datasets demands robust computational frameworks. Here, artificial intelligence (AI), especially explainable AI (XAI), offers transformative potential. By identifying key regulatory features and predicting compensatory responses, XAI models could guide therapeutic strategies and overcome drug resistance. Yet, realizing this potential requires addressing challenges like data heterogeneity, small sample sizes, and batch effects. Altogether, this thesis underscores the centrality of compensatory mechanisms in disease and demonstrates how multi-omics and cutting-edge proteomics technologies can unravel the adaptive plasticity of signaling networks. These insights not only enhance our biological understanding but also inform the rational design of therapies that target network vulnerabilities in disease.