Genetic tool development and metabolic engineering in model and non-model yeasts

Schultz, John C

Permalink

https://hdl.handle.net/2142/113804 Copy

Description

Title

Genetic tool development and metabolic engineering in model and non-model yeasts

Author(s)

Schultz, John C

Issue Date

2021-09-03

Director of Research (if dissertation) or Advisor (if thesis)

Zhao, Huimin

Doctoral Committee Chair(s)

Zhao, Huimin

Committee Member(s)

• Rao, Christopher V
• Kraft, Mary L
• Jin, Yong-Su

Department of Study

Chemical & Biomolecular Engr

Discipline

Chemical Engineering

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Saccharomyces cerevisiae
• metabolic engineering
• CRISPR/Cas9
• Rhodosporidium toruloides
• fatty alcohols

Abstract

The use of microbial cell factories such as yeast to convert plant-derived sugars and biomass into biofuels and other products offers a route towards a sustainable, carbon-neutral economy. Saccharomyces cerevisiae has been extensively engineered in efforts produce many classes of compounds. However, much of the genome of this best-studied eukaryote remains poorly characterized for fermentation-relevant phenotypes and therefore inaccessible to rational engineering. Genome-scale engineering involves the creation vast numbers of mutants, targeting all genes in the host genome and then screening the resultant library to identify novel genetic determinants of phenotypes of interest. Building upon previously developed genome-scale engineering strategies and CRISPR-based tools, we created a tri-functional CRISPR mutant library covering overexpression, downregulation, and deletion of each ORF in the BY4742 genome with 4-6 gRNAs per gene using CRISPR activation, interference, and deletion, respectively. The library was screened to identify genetic determinants of furfural tolerance, protein display, and increased production of S-adenosylmethionine (SAM) using next-generation sequencing, flow cytometry, and a fluorescent biosensor. Downregulation of SIZ1 combined with overexpression of NAT1 and downregulation of PDR1 was found to dramatically improve furfural tolerance and display of Trichoderma reesei endoglucanase II was improved by deletion of HOC1 and repression of NUP157. For SAM overproduction, FACS screening using a SAM-responsive biosensor identified the combination of upregulation of SNZ3, RFC4, and RPS18B to improve SAM productivity by 2.2-fold and 1.6-fold in laboratory and industrial yeast strains, respectively. Transcriptomic and metabolomic analyses of the engineered SAM-producing strains indicated that upregulation of vitamin B6 metabolism greatly increases the activity the transsulfuration pathway used to produce SAM. Using genome-scale engineering of laboratory yeast strains to inform and guide industrial yeast strain engineering presents an effective approach to design microbial cell factories for industrial applications. While S. cerevisiae has proven to be a highly effective platform for conversion of glucose to ethanol, its inability to grow natively on lignocellulosic sugars such as xylose and arabinose, and its preference for utilizing carbon flux for anaerobic fermentation over aerobic respiration limit its capability to process lignocellulosic biomass, and to produce acetyl-CoA-derived bioproducts such as terpenoids and fatty acids and their derivatives. Although S. cerevisiae has been extensively engineered to improve its capabilities in these roles, an alternative increasingly investigated by the scientific community is the exploration of other, “non-model” organisms with different metabolic properties. The oleaginous, red yeast Rhodosporidium toruloides, which can grow natively on lignocellulosic sugars and can naturally produce high levels of lipids and isoprenoids is one such organism. However, its large phylogenetic distance from other yeast species means that genetic tools and understanding of the metabolism of R. toruloides remain limited. We have sought to develop tools facilitating the genetic manipulation of R. toruloides and to apply them in the metabolic engineering of this non-model yeast. Firstly, a high-efficiency CRISPR knockout system was developed in the strain NP11. Cas9 expression was optimized using codon optimization and screening of various nuclear localization sequences and constitutive promoters of different strengths. Due to the lack of a suitable RNA polymerase III promoter for gRNA expression, a novel self-splicing fusion 5S rRNA-tRNA promoter was developed, allowing greater than 95% gene knockout for various genetic targets. Additionally, multiplexed double-gene knockout mutants were obtained using this method with an efficiency of 78% by expressing two tandem gRNA cassettes. This CRISPR knockout tool was applied in combination with overexpression of native and heterologous genes to optimize the titer of a fatty alcohol-producing strain expressing a fatty acyl-CoA reductase gene from Marinobacter aqueolei. Following co-expression of this gene and a Cas9 gene to facilitate creation of knockout mutants, a panel of metabolic engineering targets were explored. Two overexpression targets (ACL1 and ACC1, improving cytosolic acetyl-CoA and malonyl-CoA production, respectively) and two deletion targets (the acyltransferases DGA1 and LRO1) resulted in significant (1.8 to 4.4-fold) increases to the fatty alcohol titer in culture tubes. Combinatorial exploration of these modifications in bioreactor fermentation culminated in a 3.7 g/L fatty alcohol titer in the LRO1Δ mutant. As LRO1 deletion was not found to be beneficial for fatty alcohol production in other yeasts, a lipidomic comparison of the DGA1 and LRO1 knockout mutants was performed, finding that DGA1 is the primary acyltransferase responsible for triacylglyceride production in R. toruloides, while LRO1 disruption simultaneously improved fatty alcohol production, increased diacylglyceride and triacylglyceride production, and increased glucose consumption. The DNA-protein mapping technique CUT&RUN was used to identify the centromeric regions of R. toruloides IFO0880, an essential genetic element for the creation of a stable episomal plasmid (in combination with a replication origin). CUT&RUN was used to map genomic regions associated with the centromeric histone H3 protein Cse4, a marker of centromeric DNA. Fifteen putative centromeres ranging from 8 to 19 kb in length were identified and analyzed, and several were tested for, but did not show activity as replication origins in their own right. These centromeric sequences contained below average GC content, below average but variable gene density, and in most cases low to moderate sequence conservation. Future efforts to identify a replication origin in this yeast can utilize these centromeric DNA sequences to improve the stability of episomal plasmids derived from putative origin replication elements. Finally, two ongoing efforts in genetic tool development in R. toruloides are described. CRISPR activation and interference were explored through the fusion of dSpCas9 with various trans-activation and trans-repression elements. Efficient CRISPRi was obtained by fusing Cas9 with two tandem repression domains from S. cerevisiae transcription factors, while the commonly used VP64, VP64-p65, and VPR activation domains were not functional. Two Cas orthologs, SaCas9 and LbCpf1 were also tested for gene deletion in R. toruloides and were found to be functional following codon optimization. Efforts were also made to develop an inducible Cre/lox selection marker recovery system in the strain IFO0880. A Cre gene was codon optimized and found to be functional under constitutive expression. However, the inducible promoters tested so far based on reports in different R. toruloides strains, pLAD1, pCTR3, and pNAR1 were found to be either too leaky or nonfunctional in this strain. Characterization of additional promoters with tight but strongly inducible expression will enable the completion of this system in the future.

Graduation Semester

2021-12

Type of Resource

Thesis

Permalink

http://hdl.handle.net/2142/113804

Copyright and License Information

Copyright 2021 John Schultz

Owning Collections