A general strategy for functionalization of de novo designed helical bundle proteins

Description

Tremendous progress has been made in protein design in part due to advances in computation power and growth of protein structure databases. However, the ability to reliably introduce function into genetically encodable de novo proteins is still an unsolved task. Among the plethora of protein functions, the introduction of catalytic functionality into de novo proteins has proven to be particularly hard. One of the bottlenecks in this endeavor is the limited variability of starting backbones, assembled from naturally occurring protein fragments. A remedy for this lack of diversity is parametric design, which uses the Crick coiled-coil generating equations to generate thousands of very similar, yet different backbone geometries for a given fold as starting points. It has been shown previously that proteins designed in this fashion feature remarkably high thermodynamic stabilities, which they most likely owe to their geometric ideality.
Here we show that it is possible to trade some of this thermodynamic stability to generate structures with local deformations that harbor binding sites.
We used this approach to establish a general method for the de novo design of metal-dependent protein catalysts with diverse metal cofactors of increasing complexity. Surprisingly, most of our designs retain their high stability. Our show-case examples include copper (as a cofactor for nitrite reductases and quercetinases) and ruthenium half-sandwich complexes.
Biochemical and biophysical characterization of these designs show that they are readily expressed in Escherichia coli, display α-helical circular dichroism signals as well as SAXS profiles that agree well with the designs and are thermostable up to 95°C. UV/Vis spectroscopic analysis and binding assays of these designs show that they indeed bind the desired metal ion. Assays for nitrite reductase and quercetinase activity reveal their, albeit slow, catalytic properties. copper-based The experimentally determined structure of one of these designs is well superimposable onto the design model with a Cα-rmsd values of roughly 2Å.

Period	25 Sept 2023
Event title	European Rosettacon: Protein design in the age of artificial intelligence
Event type	Conference
Location	Leipzig, GermanyShow on map
Degree of Recognition	International