The typically low thermodynamic and kinetic stability
of enzymes
is a bottleneck for their application in industrial synthesis. Baeyer–Villiger
monooxygenases, which oxidize ketones to lactones using aerial oxygen,
among other activities, suffer particularly from these instabilities.
Previous efforts in protein engineering have increased thermodynamic
stability but at the price of decreased activity. Here, we solved
this trade-off by introducing mutations in a cyclohexanone monooxygenase
from
Acinetobacter
sp., guided by a combination of
rational and structure-guided consensus approaches. We developed variants
with improved activity (1.5- to 2.5-fold) and increased thermodynamic
(+5 °C
T
m
) and kinetic stability
(8-fold). Our analysis revealed a crucial position in the cofactor
binding domain, responsible for an 11-fold increase in affinity to
the flavin cofactor, and explained using MD simulations. This gain
in affinity was compatible with other mutations. While our study focused
on a particular model enzyme, previous studies indicate that these
findings are plausibly applicable to other BVMOs, and possibly to
other flavin-dependent monooxygenases. These new design principles
can inform the development of industrially robust, flavin-dependent
biocatalysts for various oxidations.