Native Mass Spectrometry-Based Proteoform Explicit Metrics System (PrEMS) Informs Structural Improvement of Artificial Photocyclase Enzymes

Abstract

Abstract Directed evolution is a powerful strategy for enzyme engineering and has successfully enabled unique stereoselective biosynthesis of high-value organic molecules. Enantiomeric excess (ee) remains the gold standard to assess the enzyme performance, but only reflects the final reaction outcome and provides limited mechanistic insights, leaving design efforts effectively a ″black box″, especially for multimeric proteins and heterogeneous systems. Here, we introduce a Proteoform Explicit Metrics System (PrEMS) that links protein point mutations and catalytic enantioselectivity via structural features measured by native mass spectrometry (native MS). Using 11 artificial photocyclase mutants, we demonstrate that single-point mutations can have a profound impact on protein structure, resulting in a highly convoluted effect on protein function not easily explainable by a single variable. We derived a multiple linear regression (MLR) model from the proteoform explicit native MS data based on the enzyme’s oligomeric state, photosensitizer labeling efficiency, susceptibility to oxidation, and photodegradation half-life. PrEMS showed significantly higher correlation (adjusted R2 increases from 0.053 to 0.869) to the enzyme’s enantioselectivity than the conventional physicochemical descriptors of the mutated amino acids. PrEMS serves as a quantitative framework to fill the knowledge gap in structural descriptors between sequence mutation and the enzyme’s function, rationalizing enzyme improvement strategies. For example, our analysis indicated that photodamage can be suppressed by incorporating aromatic residues near the active site. Our study outlines a tangible path with quantifiable metrics for integrating native MS data into enzyme design workflows.