Our Research
Decoding protein function through the lens of protein-metabolite co-regulation
The human genome encodes over 20,000 protein-coding genes, yet over the past two centuries, life science research has primarily focused on only ~6,000–7,000 well-studied proteins. A large portion of the proteome therefore remains understudied, with the functions of many proteins still unknown.
Our work centers on developing mass spectrometry (MS)-based technologies to uncover the roles of orphan and understudied proteins in regulating cellular and systemic metabolism. We apply MS and machine learning approaches to investigate how proteins regulate metabolites on a proteome-wide scale and in vivo, and how metabolic cues in turn modulate protein abundance and function to maintain metabolic homeostasis. Using this approach, we deorphanized Leucine-Rich Repeats Containing Protein 58 (LRRC58) and defined its role in controlling cysteine metabolism. We found that LRRC58 functions as the substrate adaptor of an E3 ubiquitin ligase that mediates proteasomal degradation of CDO1, the rate-limiting enzyme that channels cysteine into the catabolic pathway toward hypotaurine and taurine. Moreover, we discovered that cysteine itself serves as the molecular signal that turns on and off LRRC58-mediated CDO1 degradation. In this way, the LRRC58-CDO1 partnership is able to sense cellular cysteine levels and maintain cysteine homeostasis.
Currently, we are focused on understanding how metabolism is rewired during aging and cancer, with the goal of identifying therapeutically actionable molecular targets to extend healthspan and combat cancer.
Representative publications:
We recognize that endogenous metabolite regulation of protein function is conceptually analogous to chemical perturbation by small molecules such as chemical probes and drugs. In both cases, small molecules bind proteins and modulate their activity. The key distinction is that metabolites have evolved to function as physiological signals, conveying intracellular and environmental metabolic states to specific pathways to elicit appropriate cellular responses, whereas synthetic compounds are designed to modulate protein function for mechanistic investigation or therapeutic benefit. Building on this conceptual framework, we develop mass spectrometry- and data science-driven technologies to systematically map the ligandability and druggability of the human proteome. Our chemoproteomic platforms provide systems-level views of protein-small molecule interactions, while computational approaches prioritize molecular targets for mechanistic investigation. For high-value targets, we combine our MS technologies with phenotypic screening to identify chemical leads and establish a foundation for probe and drug development.
Current efforts focus on developing chemoproteomic and machine learning approaches to design or repurpose small molecules that selectively target proteins of therapeutic interest. In parallel, we are generating large-scale, high-resolution maps of protein-ligand interactions to enable AI-guided prediction of protein ligandability. Together, these efforts aim to establish chemical strategies for perturbing protein function to investigate biological mechanisms and advance therapeutic discovery
Representative publications:
Defining protein-small molecule interactions for chemical perturbation of protein function
We are currently pursuing the following directions:
1. Developing mass spectrometry and machine learning approaches to elucidate molecular mechanisms of metabolic rewiring in cancer and aging.
2. Designing activity-based proteome profiling (ABPP) strategies to guide the development of chemical leads that target undrugged proteins or proteins prone to drug resistance.
3. Combining mass spectrometry with genetic perturbations to mechanistically define protein functions at scale.
4. Creating automation strategies for large-scale proteomics and metabolomics.
5. Defining the role of protein LRRC58 in physiology and disease, and developing chemical leads to target LRRC58 for potential therapeutic applications.