The increasing production of Polyethylene terephthalate (PET) combined with inadequate waste management and a lack of effective recycling technologies has led to the accumulation of PET material and significant environmental pollution. Bio-catalysis is a promising solution to address the recycling or upcycling of PET in an environmentally friendly and cost-effective manner. However, currently the low efficiency of known PET hydrolysing enzyme (PHE)s and the limited scope to scale up their production get in the way of using them at an industrial scale. Efforts have been made to identify novel PHEs and to improve the catalytic efficiency of known PHEs by using rational design approaches, applying homology searches, and/or machine learning. However, the capability of assessing the efficiency of PHEs is limited to low-throughput functional assays. Since PHEs stem from different enzyme families general findings on one PHE cannot easily be transferred to other known PHEs. Thus functional assays that are compatible with high throughput screens would enable far larger explorations of protein sequence space and thus increase the probability to identify highly active and industrially relevant PHEs.

The present PhD thesis is aiming to develop a sensitive and reliable screening platform compatible with both low and ultra-high-throughput assay formats to support the directed evolution of known PHEs. Such a screening platform would facilitate the acquisition of large and high quality data sets of structure-function relationships of single residues within a PHE. These data-sets can thus be used to support computationally aided rational design approaches for enzyme improvement as well as to uncover a deeper understanding sequence landscape and function of PHEs which can in turn be applied to identify novel PHEs from metagenome data-sets.

The core of screening platform is thereby formed by a platform organism, which contains both the gene encoding a PHE variant, as well as a transcription factor based synthetic gene circuit that senses and reports on the accumulation of the PET hydrolysis product terephthalic acid (TPA). Thus, the platform organism allows the use of PET as a substrate and enables a direct link of PET hydrolysis activity to a detectable signal, enabling genotype-phenotype coupling while facilitating the downstream screening process (Fig. 1).

To screen for PHE activity, single platform organisms that carry the PHE library are compartmentalised (in e.g. microtiter plates or droplet microfluidics) to allow the extracellular PET hydrolysis reaction to feedback to the host organism and trigger the synthetic circuit while avoiding cross activity between two reactions and PHE variants. Upon completion of the reaction, the platform organism which now contains information on both genotype and phenotype of the PHE variant, can be pooled with the other host organisms and subjected to FACS for high-throughput screening of the library.

In the course of this PhD project the host organisms E. coli, P. putida and S. cerevisiae have been identified as suitable platform organisms that have been validated to secrete a range of PHEs. Furthermore, functional TPA sensors have been developed for each of the determined host organisms. The functionality of both the PHE expression as well as the TPA sensor unit have successfully been tested individually and in concert but are yet to be tested when hosted within one and the same organism.

This work offers a promising approach to improve the efficiency and stability of PET hydrolases by enabling the screen of large variant libraries, and thus to contribute to a sustainable PET waste management and a circular PET economy.