Rosalie Lipsh

Conference 2023 Presentation

 

Project title

Modularly designed protein fragments combine into thousands of active and structurally diverse enzymes

 

Authors and Affiliations

Rosalie Lipsh-Sokolik1, Sarel J. Fleishmam1

1. Department of Biomolecular Sciences, Weizmann Institute of Science, 7610001, Rehovot, Israe

 

Abstract

Background

Functional diversity in enzymes often requires significant backbone changes, but even subtle mutations in an enzyme may negatively impact its activity.

Methods

We introduce CADENZ, an atomistic and machine-learning strategy to design modular enzyme fragments that combine all-against-all to generate diverse, low-energy structures with stable catalytic constellations.

Results

Applied to endoxylanases, CADENZ designed nearly a million unique proteins, and activity-based protein profiling with an endoxylanase-specific probe rapidly recovered 3,114 active enzymes that adopt 756 distinct backbones. Functional designs exhibit higher active-site preorganization than nonfunctional ones and more stable and compact packing outside the active site. Implementing these lessons into CADENZ led to a tenfold improved hit rate.

Conclusions

This design-test-learn loop can be applied, in principle, to any modular enzyme or binding protein family, yielding huge diversity and general lessons on protein design principles.