Ancestral Nitrilase Mining and Semi-Rational Engineering for Enhanced Thermal Stability in Rapeseed Meals-Derived Nitriles Degradation

Rapeseed meal (RSM), a protein-rich byproduct, holds potential as a high-quality animal feed, but nitrile compounds derived from glucosinolates (GSLs) in RSM pose a toxicity risk. Nitrilases, enzymes that hydrolyze toxic nitriles to carboxylic acids, offer a potential solution for detoxification. However, the low thermal stability of nitrilases restricts their industrial applicability. We herein identified eight ancestral nitrilases through sequence-based mining using 6803NIT as a probe enzyme. Among these, ancestral enzyme A1 exhibited the highest specific activity (58.3 U/mg) and half-life (t_1/2 = 3.5 h at 40 °C). To enhance thermal stability, we engineered a quadruple mutant A1M_4C, which exhibited a 4.7-fold increase in half-life (t_1/2 = 16.3 h) and a 2-fold increase in specific activity (118.5 U/mg). Kinetic analysis revealed a reduction in K_m from 14.9 to 10.5 mM and an increase in k_cat/K_m from 1.9 to 4.37 s^-1·mM^-1. Mechanistic studies indicated that enhanced stability in A1M_4C was due to increased hydrogen bonding and stronger amino acid interactions. Simulated feed pelletization at 90 °C for 2 min showed that A1M_4C acquired a 22.2-fold improvement toward nitriles degradation over wild-type A1. These findings demonstrate the potential of ancestral enzyme mining to develop thermostable nitrilases for industrial feed applications.