Expanding the universe of functional metalloproteins through AI-assisted computational design

点击翻译按钮获取中文摘要

Metalloproteins are an abundant class of proteins. They perform some of the most difficult and life-sustaining reactions on earth with high efficiency. They have potential as green biocatalysts for applications in energy production and bioremediation. Adding new functions to these proteins or generating completely novel metalloproteins is challenging. Computational methods will be used to design new functional metalloproteins. The project will also introduce high school students and undergraduates to protein engineering.The goal of this project is to expand the universe of functional metalloproteins. To address the limitations of native proteins and the current challenges facing traditional design approaches, latest advances in machine learning in combination with protein design software will be applied to generate: (a) a library of functional electron transfer proteins with diverse shapes that contain a well-known 2His-1Cys copper binding motif (T1Cu site), (b) a library of functional catalytic proteins with 3His zinc binding motifs (hydrolase active site), and (c) a new platform for metal coordination using the non-canonical bpy-ala amino acid. It is hoped that the results obtained will provide new insight into the role of protein shape on tuning properties of the metal center and the rules for designing new metalloproteins. Methods developed in this proposal can be generalized to the design of other metalloproteins (dinuclear, heme binding, etc.) and small molecule-binding proteins, largely expanding the current scope of functional protein design.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.