Repository logo
 

Ultrahigh-Throughput Microfluidic Droplet Screening of Metagenomic Libraries for Esterases and Kemp Eliminases


Type

Thesis

Change log

Authors

Abstract

In the search for new enzymes, functional metagenomic libraries are particularly interesting since they give access to the genomes of the 99% of microorganisms which cannot be cultured in the laboratory. However, using classical biochemical assays to screen such libraries is impeded by the fact that enzymes for a given reaction are incredibly rare (1 in 105 on average). Ultrahigh-throughput droplet microfluidics has emerged as an effective technology to overcome this limitation, with the ability to screen 107 reactions per day. However, fewer than ten enzyme assays have been implemented in droplets to date and only once has droplet microfluidics been applied to functional metagenomics. Therefore, for the wider adoption of this technology, it is critical to develop more enzyme assays in droplets. To address this need, first I built an ultrahigh-throughput droplet sorting instrument and then set out to establish two new enzyme assays in droplets: one for the industrially important esterase reaction and another for the Kemp elimination, an artificial reaction. In Chapter 2, I describe the state-of-the-art fluorescence-activated droplet sorter (FADS) I developed for use with fluorogenic enzyme substrates. I improved this instrument incrementally based on the needs of collaborative projects, which ensured the success of these projects in screening enzyme libraries. In Chapter 3, I established an esterase assay and performed the first reported functional metagenomic screen for esterases in droplets. Over 30 million droplets were sorted, amounting to 10× coverage of a metagenomic library consisting of over one million members; making this the largest esterase screen performed to date. Twelve clones encoding thirteen novel esterases were isolated. The majority were members of the α/β-hydrolase super-family of proteins. Four came from small families taking up less than 1% of the sequence space within the super-family. Therefore, functional screening at ultrahigh-throughput provided access to thinly populated sequence-space. It is unlikely that these sequences would have been explored using prediction-based methods. Eight out of eleven enzymes had detectable thioesterase activity, three had β-lactamase activity, one had β-galactosidase activity, and one had Kemp eliminase activity. This finding corroborates the idea that the ability of enzymes to catalyse more than one reaction, a property called enzyme promiscuity, is widespread. In Chapter 4, I established a Kemp eliminase assay in droplets with the aim of exploring how widespread promiscuous enzymes catalysing this non-natural reaction are. Using the substrate 5-nitro-1,2-benzisoxazole in combination with absorbance-activated droplet sorting (AADS). I describe the enrichment of a previously reported Kemp eliminase, HG3.17, over a negative control using droplet microfluidics and use this method to screen substitution, insertion and deletion libraries of HG3.17. Active library variants were enriched and the variants with improved soluble expression isolated. Five locations were identified that tolerate insertions and deletions and, in one investigated case, have improved soluble expression. These variants may serve as starting points to explore previously inaccessible mutational trajectories to improve the catalytic parameters of HG3.17. Due to the limited sensitivity of the assay, functional metagenomic screening was not possible using this substrate. In Chapter 5, I report the newly-discovered fluorogenic Kemp substrate 5-azido-1,2-benzisoxazole, reported and characterised here for the first time. I established this substrate in droplets and, using the FADS instrument, enriched the Kemp eliminase HG3.17 over a negative control. This assay was sufficiently sensitive to detect Kemp eliminase activity in a metagenomic library. The apparent hit rate was comparable to that of the esterases, suggesting that promiscuous enzymes capable of catalysing this reaction are commonplace. A large number of false positives impeded the straightforward isolation of the responsible library members. This constraint is likely to be overcome in future by using a bias-free metagenomic library. Here, a combination of functional re-screening and sequencing allowed the identification of one lead: a predicted class IV adenylyl cyclase. In conclusion, I have built a highly sensitive droplet sorting instrument using which I isolated thirteen new esterases, established the first Kemp eliminase assay in droplets, used both for mutant library and metagenomic screening, and additionally enabled the improvement of numerous enzymes through library screenings in collaboration with others. This work contributes to the success of ultrahigh-throughput droplet microfluidics in furthering our understanding of enzymes in nature and our ability to tailor them for green chemistry applications in industry.

Description

Date

2019-10-28

Advisors

Hollfelder, Florian

Keywords

Microfluidics, Metagenomics, Enzymes, Ultrahigh-Throughput, Screening, Esterases, Kemp Elimination

Qualification

Doctor of Philosophy (PhD)

Awarding Institution

University of Cambridge
Sponsorship
EPSRC Centre for Doctoral Training in Sensor Technologies and Application Grant Number: EP/L015889/1