Embargoed until 2026-05-23
Author
Date
2023Type
- Doctoral Thesis
ETH Bibliography
yes
Altmetrics
Abstract
The past few decades have seen a dramatic increase in the use of biocatalysts in commercial chemical processes, shifting the emphasis from energy-intensive traditional chemistry to sustainable chemistry. Unsurprisingly, significant effort has gone into modifying and improving the characteristics of naturally occurring enzymes for use in specific biotechnological applications. Current enzyme engineering techniques, such as directed evolution, require the production and testing of large libraries of mutations to identify commercially valuable variants. Unfortunately, traditional screening approaches are unable to screen such large mutagenesis libraries in a robust and timely manner. Droplet-based microfluidic systems are able to produce, process and sort picoliter droplets at kilohertz rates and have emerged as a potentially powerful and high-throughput tool for library screening. However, the reliance of these screening approaches on inline fluorescence detection either restricts their use to a limited number of natural substrates and enzyme classes or involves the use of surrogate substrates, which bias the enzyme optimization process.
This thesis describes both the development of novel enzyme screening assays and the implementation of existing assays in droplet-based microfluidic platforms, with the aim of eliminating the need to use surrogate substrates from certain enzyme classes. In addition, this work expands the existing droplet-based microfluidic toolbox by establishing a new and powerful droplet absorbance detection and sorting methodology, which has particular utility in the screening of natural substrates. To this end, three different assays and implementations of microfluidic technology are presented herein.
First, a universal NADH-based fluorescence assay focused on evolving a poorly-performing alcohol dehydrogenase from Sphingomonas species A1 (SpsADH) was established. A library of 50,000 variants was screened against the non-native substrate L-guluronate using fluorescence-activated droplet sorting (FADS). Significantly, we discovered an enzyme variant with a 2.6-fold increase in catalytic efficiency towards the non-native substrate after only one round of mutation.
Second, a miniaturized and ultra-sensitive halide detection assay was used to improve the performance of an ancestral haloalkane dehalogenase enzyme. A droplet-based microfluidic platform and miniaturized assay was used to screen a large library against three natural substrates (1,2-dibromoethane, 1-bromohexane and bromocyclohexane) to select the most active mutant in each case.
Finally, a novel absorbance-activated droplet sorting platform was developed for library screening. Direct absorbance measurements from picoliter-size droplets was achieved using a lithographic mask and refractive index matched fluid phases. Such an approach allowed the sensitive interrogation and sorting of droplets at kHz rates. The efficiency of the sorter was showcased through the rapid screening of a 105-member aldehyde dehydrogenase library with approximately 2x coverage, successfully obtaining six mutant variants with improved enzymatic properties, with the most successful variant exhibiting a 51% improvement in catalytic efficiency towards D-glyceraldehyde. Show more
Permanent link
https://doi.org/10.3929/ethz-b-000612985Publication status
publishedExternal links
Search print copy at ETH Library
Publisher
ETH ZurichSubject
microfluidics; directed evolution; droplet-based microfluidics; droplet sorting; enzyme engineeringOrganisational unit
03914 - deMello, Andrew / deMello, Andrew
Funding
176011 - Advanced Droplet-Based Technologies for Engineering Dynamic Elements in Proteins (SNF)
Related publications and datasets
More
Show all metadata
ETH Bibliography
yes
Altmetrics