Collaborative Research: Directed Enzyme Evolution Accelerated by Machine Learning for Enhancing the Biodegradation of Emerging Contaminants

点击翻译按钮获取中文摘要

The Environmental Chemical Science Program in the Division of Chemistry supports Professors Mengyan Li and Edgardo Farinas at the New Jersey Institute of Technology and Professor Wenwei Zheng at Arizona State University for this project. An increasing number of human-made chemicals are being released into the environment where they can deteriorate the water quality. Bacteria living in nature can secrete enzymes to degrade some of these chemicals and exploit them as carbon sources. However, the naturally occurring biodegradation process can be slow or inadequate to remove these contaminants. In this project, bacterial enzymes will be modified and optimized in the laboratory to speed up their degradation rates. The experimental results will be input data to train a computational model that simulates the interaction between the enzyme and the contaminant. In return, this computational model will promote the design of new enzymes with greater degradation rates. This project will initiate vigorous activities to engage graduate and undergraduate students, especially those who are members of underrepresented groups. Summer exchange workshops will be organized to promote communications between students and researchers from the New Jersey Institute of Technology and Arizona State University. Outreach efforts will be made to promote the green treatment of water contaminants and technology innovation by combining laboratory-based enzyme evaluation and computation-based machine learning.Some emerging contaminants pose imminent threats to public health and natural biota due to their frequent detection and enduring persistence in the environment. The project selects 1,4-dioxane (dioxane) as a model contaminant and employs state-of-the-art directed enzyme evaluation to gain fundamental biochemical insights into essential bacterial enzymes and ultimately optimize their biocatalytic performance for dioxane removal. Directed enzyme evolution will be used to mimic and accelerate the natural evolution in a laboratory setup, creating enzyme mutants with increased degradation efficiency towards dioxane. The rich empirical data set provided by directed enzyme evolution can be used to guide a machine learning process to predict key molecular determinants that link the protein sequence with its function and suggest new mutations for further improvement of their catalytic performance. This integrative framework will advance our fundamental knowledge regarding the biochemistry of bacterial enzymes and promote the technological transformation to combating the global challenges caused by emerging contaminants in water.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.