Designing yeast cell factories to produce early-step Taxol® precursors: from CRISPR toolkit development to computer-aided design and optimisation of a CRISPR-based molecular diagnosis method

Abstract

CRISPR methods for yeast generally rely on pre-assembled DNAs and extra cloning steps to deliver gRNA, Cas protein, and donor DNA. These laborious steps might hinder its usefulness. In this thesis, I propose a convenient, rapid and standardisable CRISPR method, named Assembly and CRISPR-targeted in vivo Editing (ACtivE), in Chapter 3. ACtivE relies on in vivo assembly of linear DNA fragments for plasmid and donor DNA construction. In this way, these parts can be easily selected and combined accordingly from a repository. Eight autonomously replicating sequence (ARS) - proximal genomic loci were also characterised in terms of integration and gene expression efficiencies and the impacts of the disruptions of these regions on cell fitness. The multiplexing capacity of the ACtivE was evaluated using different strategies such as multi-locus or single-locus integrations to insert two and three genes simultaneously into the genome.

The following chapter (Chapter 4) covers the study in which a system biology approach was used to design yeast strains expressing heterologous genes from the Taxol® (paclitaxel), a blockbuster anticancer drug, biosynthetic pathway. The computer-aided design was employed with the help of the COnstraint-Based Reconstruction and Analysis (COBRA) Toolbox to predict gene or reaction candidates to be deleted or overexpressed to enhance the fluxes towards the Taxol® pathway. The genomic modifications were screened to determine the promising modifications to increase the titres of Taxol® precursors. The wet-lab results were compared with the in silico simulations and the best-performing strains were then used for higher-scale productions in shake flasks and mini-scale bioreactors. With this approach, a yeast strain (KM32) producing the maximum titers of early step Taxol® precursors with 215 mg/L of taxadiene, 43.65 mg/L of taxa-4(20),11-dien-5α-ol (T5α-ol), and 26.2 mg/L of taxa-4(20),11-dien-5-α-yl acetate (T5αAc) has been designed.

In Chapter 5, the use of CRISPR was expanded to develop an ultra-sensitive molecular diagnostics method where a statistical design of experiments (DoE) was performed to accelerate the development and optimisation of a CRISPR/Cas12a-recombinase polymerase amplification (RPA)-based one-pot COVID19 detection method for the first time. The factors with a significant effect on performance were elucidated and optimised, facilitating the detection of two copies/µL of the full-length SARS-CoV-2 (COVID-19) genome using simple instrumentation. The screening revealed that the addition of a reverse transcription buffer and an RNase inhibitor, components generally omitted in one-pot reactions, improved performance significantly, and optimisation of reverse transcription was critical on the sensitivity. The findings presented in this thesis demonstrate the capability of computer-aided microbial system designs and the potential of CRISPR technologies for both yeast strain development and molecular diagnostics that can greatly interest the synthetic biology research community.