MARKHAM-DISSERTATION-2018.Pdf

2018

The Dissertation Committee for Kelly Ann Markham Certifies that this is the approved version of the following dissertation:

Expanding the Product Portfolio of Yarrowia lipolytica through Metabolic Engineering and Synthetic Biology Tool Development

Committee:

Hal Alper, Supervisor

Lydia Contreras

George Georgiou

Adrian Keatinge-Clay Expanding the Product Portfolio of Yarrowia lipolytica through Metabolic Engineering and Synthetic Biology Tool Development

Kelly Ann Markham

Dissertation Presented to the Faculty of the Graduate School of The University of Texas at Austin in Partial Fulfillment

of the Requirements for the Degree of

Doctor of Philosophy

The University of Texas at Austin August 2018 Dedication

For those who have been made to feel small.

You are strong. You are important. You deserve to take up space.

Acknowledgements

First and foremost, I want to thank Dr. Hal Alper for being a supportive advisor and helping me grow in all aspects of being a scientist. From helpful tricks for personal development like working with Microsoft (that alignment function will always haunt me) to high expectations requiring proper experimental design, I have learned an amazing amount from Hal in the last five years. Hal has proven willing to take time to make sure that students are taken care of and meeting their full potential. Thanks for taking that time to make us all better researchers and giving us room to explore different projects and ideas.

I am forever grateful for the opportunity to work with Hal and the Alper lab.

Next, I would like to thank Dr. Lydia Contreras, Dr. George Georgiou, and Dr.

Adrian Keatinge-Clay for serving as my dissertation committee. Each of them took time out of their busy schedules to provide helpful advice throughout my thesis work. Their questions and recommendations from my preliminary proposal led to a strengthened set of aims and challenged me to become a better scientist. Special thanks to Dr. Contreras for always being willing to help with fellowship applications and the job process, especially for writing letters of recommendation on my behalf. I am thankful for having the opportunity to work on the same floor as her group. An extra thank you to Dr. Keatinge-

Clay for taking time to meet with Claire and me to help us learn more about the chemistry of polyketides and troubleshoot when we were making products we could not identify.

Thank you to my committee members for all your suggestions and feedback.

v Next, it would not have been the same graduate experience without the awesome members of the Alper Lab. Thanks to everyone in the lab for all the collaboration, suggestions, questions, and willingness to show up for presentations. I am especially grateful for Team Yarrowia including Claire Palmer, James Wagner, Lauren Cordova,

Haibo Li, Leqian Liu, and Andrew Hill. Working with this nonconventional system is hard, and you all made it better by struggling side by side with me, sharing in successes and failures, and working to develop and share best practices for the lab. Everyone else who I had the opportunity to work with in the lab have also been instrumental to my success-

Kate Curran, Nathan Crook, Sun-mi Lee, Jie Sun, John Leavitt, Joseph Cheng, Joe

Abatemarco, Aaron Lin, Nick Morse, Matt Deaner, Joseph Yuan, Kevin Reed, Kristin

Presnell, and Xiunan Yi. Thank you for all your collaborations, life chats, TV suggestions, food adventures, and genuinely making the Alper lab a great place to work.

In addition to all the graduate students I have had the opportunity to work with in the lab, undergraduates and high school students deserve a major shout out. Thank you to

Clare Murray, Arvind Swaminathan, Sofia Vazquez, Ishani Chakravarty, Sarah Coleman,

Valerie Vines, Mallika Maheshwary, Cedric Ginestra, Cecilia Barnhill, and Grace Lawler.

Your energy kept me on my toes and made lab a fun place to work. Each of you made me a better teacher, and this work would have been impossible without your help.

Collaborations in the lab have been abundant, but I also had the opportunity to work with Dr. Nate Lynd and Gosia Chwatko. Thank you for taking the chance on working with bio-sourced products and making the TAL paper a truly great story. It was fun to see a process all the way from development to application, and that would have been impossible vi without your help along with the collaboration with Claire Palmer and the hard work of

Clare, Arvind, Sofia, and Ishani who will likely hold the lab record for number of simultaneous genomic DNA preps for life.

Labs do not work in isolation, and the support of staff members has not gone unnoticed. Thank you to Shallaco McDonald, Tony Le, and Jim Smitherman for keeping

CPE and our lab equipment safe, functioning, running an amazing shop, and always having the best candy. Thank you to Kevin Haynes, Ben Hester, Tammy McDade, and Eddie

Ibarra for ensuring that resources were promptly procured. Additional support from

Courtney Hazlett, Randy Rife, Jason Barborka, Kate West Baird, and Melinda Heidenreich ensured that I had all the tools to succeed. Each of these staff members have gone out of their way to help me to be a successful graduate student and always did so with kindness.

Outside of the department, I would like to especially thank my family for their support. Thank you to my Mom, Kate, Kyle, Kurt, Joel, Margeaux, Esme, and my grandparents. You all helped to make sure that I was a functioning human, especially with

Kate and Joel keeping me caffeinated. Though not related by blood, my girl squad has supported me like a family. Thank you Haley Streff, Rachel Casciato, Erin Christensen,

Sarah Waldschmidt, Callie Larson, Laura Chemler, AK Rockwell, Melissa Donahue,

Andrea DiVenere, and Claire Palmer. You challenge me to be a better person and I am grateful for your unconditional support. Thank you as well to my graduate school cohort and other Austin friends for making this city a fun place to live and explore.

Finally, thanks to 7-Eleven and podcasts for fueling this graduate degree.

vii Expanding the Product Portfolio of Yarrowia lipolytica through Metabolic Engineering and Synthetic Biology Tool Development Kelly Ann Markham, Ph.D. The University of Texas at Austin, 2018

Supervisor: Hal S. Alper

Yarrowia lipolytica is a potent microbial cell factory for metabolic engineering that will enable the sustainable production of chemicals beyond first generation biofuels and polymer precursors. As an alternative host, there are many limitations yet to be resolved that require much more intensive research. Yet, this oleaginous yeast has already proven superior to its conventional yeast counterpart Saccharomyces cerevisiae for the production of a variety of different products most notably including lipids. The work presented here offers two metabolic engineering stories demonstrating the potential of Y. lipolytica to produce value-added chemicals at high titers. Additionally, seeking to alleviate bottlenecks in the engineering process, this work presents details of an optimized and streamlined protocol for transformation as well as a strategy for tuning gene expression via synthetic terminators.

First, the focus was set on engineering a high-lipid strain for cyclopropane fatty acid production. This work leveraged a strain of Y. lipolytica that had been previously metabolically engineered to produce high levels of fatty acids. Cyclopropane fatty acids (a non-native molecule to Y. lipolytica) serve as a simple way to improve biodiesel properties and are additionally useful in lubricants. Through engineering, 3 g/L production was enabled. In a second study, the inherent flux towards lipids was hijacked and redirected towards polyketides. Specifically, extensive metabolic engineering led to production of 36 viii g/L of triacetic acid lactone, an important biorenewable precursor molecule. Third, the development and optimization of an electroporation transformation protocol alleviated a critical bottleneck in the engineering process, allowing for highly efficient incorporation of exogenous DNA into Y. lipolytica. This protocol streamlines transformations and does not require the addition of carrier DNA. Fourth, synthetic terminators originally developed for S. cerevisiae were tested in Y. lipolytica enabling improved tuning of expression. Incorporation of these synthetic terminators allowed for a 70% increase in gene expression over a commonly used terminator. Collectively, these chapters demonstrate the potential of Y. lipolytica as a host for metabolic engineering through study of tools and investigation of native pathways that can be redirected towards non-native products.

ix Table of Contents

List of Tables ...... xv

List of Figures ...... xvi

Chapter 1: Introduction ...... 1

1.1 Chapter Summary ...... 1

1.2 Introduction ...... 1

1.3 Recent advances in Y. lipolytica metabolic engineering ...... 3

1.3.1 Oleochemicals ...... 5

1.3.2 Organic acids ...... 6

1.3.3 Nutraceuticals ...... 6

1.3.4 Specialty chemicals ...... 7

1.4 The expanding synthetic toolbox for Y. lipolytica engineering ...... 8

1.4.1 New synthetic parts enable fine-tuned gene expression ...... 10

1.4.2 CRISPR-Cas9 systems increase targeted integration efficiencies ...... 12

1.4.3 DNA assembly toolboxes enable rapid testing of pathways ...... 13

1.4.4 Model based approaches guide the future of metabolic engineering ...14

1.5 Future outlook for Y. lipolytica engineering ...... 17

Chapter 2: Engineering Yarrowia lipolytica for the production of cyclopropane fatty acids ...... 19

2.1 Chapter Summary ...... 19

2.2 Introduction ...... 19

2.3 Results and discussion ...... 24

2.3.1 CFA production in a previously engineered-strain of Y. lipolytica ENGR ...... 24

x 2.3.2 CFA production in an alternative strain of Y. lipolytica, L36-DGA1 ..26

2.3.3 Bioreactor production and characterization of CFA production ...... 29

2.4 Conclusions ...... 32

Chapter 3: Rewiring Yarrowia lipolytica toward triacetic acid lactone for novel materials generation ...... 33

3.1 Chapter Summary ...... 33

3.2 Introduction ...... 34

3.3 Results and Discussion ...... 36

3.3.1 Y. lipolytica can support TAL production...... 36

3.3.2 A previously uncharacterized Y. lipolytica pyruvate bypass pathway improves TAL production ...... 38

3.3.4 Modification of β-oxidation can likewise improve TAL production...43

3.3.5 Strain engineering effectively diverted flux from lipids to TAL ...... 45

3.3.6 Culture conditions differentially alter TAL production ...... 46

3.3.7 Bioreactor cultivation boosts overall TAL titer ...... 52

3.3.8 Biosourced TAL can be incorporated into a polymer through O- functionalization ...... 55

3.4 Supplemental Data ...... 59

3.4.1 Toxicity of TAL in Y. lipolytica ...... 59

3.4.2 Testing of previously identified g2ps1 mutants ...... 60

3.4.3 Identification of uncharacterized overexpression targets ...... 60

3.4.4 Copy number assay ...... 62

3.4.5 Acetylaldehyde dehydrogenase activity assay ...... 62

3.4.6 Further media optimization ...... 63

3.5 Conclusions ...... 64 xi Chapter 4: High efficiency transformation of Yarrowia lipolytica using electroporation ...... 66

4.1 Chapter Summary ...... 66

4.2 Introduction ...... 67

4.3 Results and discussion ...... 70

4.3.1 Alterations to pre-culture conditions and competent cell preparation can improve overall transformation efficiency ...... 71

4.3.2 Transformation conditions can be modified to further improve efficiency of linear integrations ...... 76

4.3.3 Longer post-transformation recovery improves net transformation efficiency...... 79

4.3.4 Transformation efficiency of plasmid DNA is higher than linear transformation and integration ...... 81

4.4 Concluding remarks ...... 82

Chapter 5: Testing synthetic terminators in Y. lipolytica ...... 84

5.1 Chapter Summary ...... 84

5.2 Introduction ...... 85

5.3 Results and discussion ...... 86

Chapter 6: Conclusions and major findings ...... 90

Chapter 7: Proposals for future work ...... 95

Chapter 8: Materials and methods ...... 99

8.1 Methods for Chapter 2 ...... 99

8.1.1 Plasmid and strain construction ...... 99

8.1.2 Media conditions ...... 100

8.1.3 Fermentation conditions ...... 100

8.1.4 Bioreactor fermentation ...... 101 xii 8.1.5 Cell density measurement ...... 101

8.1.6 Lipid isolation and quantification ...... 101

8.1.7 Glucose quantification ...... 103

8.1.8 Copy number assay ...... 103

8.2 Methods for Chapter 3 ...... 104

8.2.1 Plasmid and strain construction ...... 104

8.2.2 Media conditions ...... 105

8.2.3 Tube fermentation ...... 105

8.2.4 HPLC quantification of TAL ...... 106

8.2.5 HPLC quantification of fermentation byproducts...... 106

8.2.6 Lipid isolation and quantification ...... 107

8.2.8 Bioreactor fermentation ...... 108

8.2.9 Copy number assay ...... 109

8.2.10 Acetylaldehyde dehydrogenase activity assay ...... 109

8.2.11 Crispr targeting to TRP1 ...... 110

8.2.12 Chemicals used for materials generation ...... 110

8.2.13 Equipment used for materials generation ...... 110

8.2.14 Purification of TAL from fermentation media...... 112

8.2.15 Synthesis of poly[(epichlorohydrin)-co-(epoxy triacetic acid lactone)] ...... 112

8.3 Methods for Chapter 4 ...... 114

8.3.1 Media conditions ...... 114

8.3.2 Plasmid preparations ...... 114

8.3.3 Strains and yeast cultivation ...... 115 xiii 8.3.4 Transformation and quantification ...... 116

8.4 Methods for Chapter 5 ...... 117

8.4.1 Strains and cultivation...... 117

8.4.2 Plasmid construction ...... 117

8.4.3 Flow cytometry ...... 118

References ...... 119

xiv List of Tables

Table 1-1: Models that have been developed for Y. lipolytica metabolism...... 16 Table 2-1: Comparative study of native cyclopropane fatty acid production levels in bacterial and plant hosts demonstrates a range of CFA species and

compositions ...... 22

Table 2-2: Genomic copy number of ycoCFA in all strains used in this study...... 28

Table 3-1: Summary of titers, yields, and productivities achieved in this work...... 55

Table 3-2: Characteristics of copolymers in comparison to PECH...... 58 Table 3-3: Overexpression gene targets. The gene abbreviations used in this paper are mapped to the systematic gene identifier and KEGG enzyme

number...... 61 Table 3-4: Copy number of overexpressed genes in the top TAL producing strains

built in this study...... 62 Table 4-1: Comparative analysis of transformation methods developed and optimized

for Y. lipolytica...... 69

Table 5-1: Terminator sequences used in this study...... 87

xv List of Figures

Figure 1-1: Developments in synthetic biology have elevated Y. lipolytica as a host

of industrial interest...... 4

Figure 1-2: Schematic of components in the synthetic biology toolbox...... 9

Figure 2-1: Mechanism of action for enzymatic CFA formation...... 22

Figure 2-2: Lipid distribution of wildtype PO1f...... 25

Figure 2-3: Overexpression of ycoCFA enables production of C19CP in ENGR...... 26

Figure 2-4: Overexpression of ycoCFA enables production of C19CP in L36-DGA1. ...28 Figure 2-5: Pulse-fed bioreactor fermentation of ENGR HPH NAT ycoCFA a) Replicate 1 glucose level, C19CP titer, and viable cell count. b) Lipid

composition of final reactor time point...... 30 Figure 2-6: Pulse-fed bioreactor fermentation of ENGR HPH NAT ycoCFA a) Replicate 2 glucose level, C19CP titer, and viable cell count. b) Lipid

composition of final reactor time point...... 31

Figure 3-1: TAL inhibits growth rate of Y. lipolytica but is not lethal...... 37 Figure 3-2: Previously identified g2ps1 mutations do not promote TAL production in

Y. lipolytica...... 38 Figure 3-3: Strain engineering scheme to evaluate native Yarrowia lipolytica

pathway potential for the production of TAL...... 39 Figure 3-4: Difference of means plot demonstrating the effect of overexpressing the

citrate pathway genes...... 40 Figure 3-5: Difference of means plot demonstrating the effect of overexpressing the

pyruvate dehydrogenase complex genes...... 41

xvi Figure 3-7: Difference of means plot demonstrating the effect of overexpressing β-

oxidation targets...... 44

Figure 3-8: Lipid production as a function of TAL production...... 45

Figure 3-9: The effect of fatty acid synthesis inhibitor cerulenin on TAL production. ...46

Figure 3-10: Impact of nitrogen starvation on TAL production...... 47 Figure 3-11: The effect of vitamin supplementation and alternative carbon sources on

TAL production...... 49 Figure 3-12: Impact of pyruvate bypass pathway intermediates spike on TAL

production...... 50

Figure 3-13: Acetate spike consumption...... 50 Figure 3-14: Relative acetylaldehyde dehydrogenase activity in NAD+ versus

NADP+...... 51

Figure 3-15: The effect of nitrogen limitation on TAL production...... 51

Figure 3-16: Bioreactor cultivation of pyruvate bypass overexpression strain...... 53

Figure 3-17: Bioreactor cultivation of pyruvate bypass overexpression strain...... 54 Figure 3-18: Production of poly[(epichlorohydrin)-co-(epoxy triacetic acid lactone)]

or PETAL...... 57 Figure 3-19: Further NMR characterization of poly(epichlorohydrin-co-

epoxytriacteic lactone)...... 58 Figure 4-1: Transformation efficiency of Y. lipolytica integrations with variable wash

strategies and pre-culture scale...... 72 Figure 4-2: Transformation efficiency of Y. lipolytica integrations from variable

length pre-cultures with one and two washes...... 74 Figure 4-3: Transformation efficiency of Y. lipolytica integrations at various

competent cell concentrations...... 75 xvii Figure 4-4: Transformation efficiency of Y. lipolytica integrations at varying

electroporation voltages...... 77 Figure 4-5: Transformation efficiency of Y. lipolytica integrations with varying

competent cell volumes...... 78 Figure 4-6: Transformation efficiency of Y. lipolytica integrations with increased

recovery time...... 80 Figure 4-7: Transformation efficiency of Y. lipolytica integration compared to

transformation efficiency of an episomal plasmid...... 82

xviii Chapter 1: Introduction1

1.1 CHAPTER SUMMARY

The oleaginous yeast Yarrowia lipolytica is quickly emerging as the most popular non-conventional (i.e. non-model organism) yeast in the field. With a high propensity for flux through TCA-cycle intermediates and biological precursors like acetyl-CoA and malonyl-CoA, this host is especially suited to meet our industrial chemical production needs. Recent progress in synthetic biology tool development has greatly enhanced our ability to rewire this organism with advances in genetic part design, CRISPR technologies, and modular cloning strategies. In this introduction, recent developments in metabolic engineering are investigated and new tools being developed are described. Together, these studies help realize the full industrial potential of this host. Finally, necessary developments to enhance future engineering efforts are proposed.

1.2 INTRODUCTION

Metabolic engineering enables inexpensive and renewable routes for the production of essential chemicals. Benefiting from concurrent advances in synthetic biology, the speed of strain engineering and industrialization of microbes has been enhanced [1]. In this regard, fungal systems (both conventional and non-conventional) are emerging as unique

1 Content in this chapter adapted from a previously authored publication written by KAM. Markham, K. A., & Alper, H. S. (2018). Synthetic Biology Expands the Industrial Potential of Yarrowia lipolytica. Trends in Biotech. Copyright © 2018 Elsevier Ltd.

1 hosts for industrial bioconversion [2]. Among these hosts, the oleaginous yeast Yarrowia lipolytica is beginning to emerge as a host with unique potential due to rapidly developing synthetic biology tools as well as metabolic plasticity. Y. lipolytica is an oleaginous yeast with generally regarded as safe (GRAS) [3] status commonly studied for its growth on (and hence tolerance to [4]) a range of chemicals, hydrocarbons, and waste carbon [5, 6]as well as for its extracellular protein production [7]. Quickly, Y. lipolytica has achieved a track- record of industrial success for the production of lipase [8] and lipids, including fish oil substitutes [9-11], with substantial engineering efforts initially led by researchers at

Codexis, DuPont, Microbia, and Pfizer. While initial applications were stymied by limited tools, synthetic biology and molecular biology tool developments including transformation

[12], overexpressions [13], marker recycle [14], and knockouts [14] have enabled the production of a variety of new chemicals.

In recent years, Y. lipolytica has gained in popularity as a host organism across both academic and industrial research groups due to its aforementioned strengths and metabolic potential. Yet, further challenges remain with respect to the propensity for non-homologous end joining [15] and rather unstable episomal plasmid systems [16-19]. These limitations notwithstanding, Y. lipolytica engineering is benefiting from an expanded synthetic biology toolkit. In this introduction chapter, the industrial relevance of Y. lipolytica is benchmarked by focusing on recent engineering progress toward oleochemicals, organic acids, nutraceuticals, and specialty chemicals. Next, advances in synthetic biology are highlighted including genetic parts like promoters, CRISPR-Cas9 systems, DNA assembly toolboxes, and model based engineering approaches that are rapidly expanding the 2 potential of this host. Finally, prospects for rapid synthetic engineering in Y. lipolytica for an expanded array of industrial applications are explored.

1.3 RECENT ADVANCES IN Y. LIPOLYTICA METABOLIC ENGINEERING

Y. lipolytica has a unique propensity for high flux through acetyl-CoA and malonyl-

CoA that can be diverted into many heterologous products beyond its native metabolic pools. Likewise, the ability to use a variety of different carbon substrates (a topic reviewed recently [20]) enhances the competitive advantages inherent in using this host. Recently, this competitive advantage was improved by engineering Y. lipolytica to assimilate phosphite to outcompete contaminating strains in low cost feedstocks [21]. As a result of these advantages, the field has seen a rapid expansion in the number of endogenous and synthetic pathways explored in Y. lipolytica (Figure 1-1). Here, recent advances for the areas of oleochemicals, organic acids, nutraceuticals, and specialty chemicals are highlighted.

3 Figure 1-1: Developments in synthetic biology have elevated Y. lipolytica as a host of industrial interest.

Y. lipolytica has seen industrial production of natively overproduced products like lipids, citric acid, and lipase. With recent improvements in the synthetic biology toolbox, metabolic engineers have expanded this industrial potential to include chemicals in four main categories: oleochemicals, specialty chemicals, nutraceuticals, and organic acids. Oleochemicals include a variety of functionalized lipid-based molecules that serve purpose in a variety of industries, most notably including biodiesel. Organic acids are derived from the TCA-cycle for primary use in the food industry, where nutraceuticals also serve as food additives. Finally, specialty chemicals find placement in a range of industries, including pharmaceuticals. The highest scale of production of specific metabolites is denoted by a symbol as per the legend in the figure. This figure is meant to serve as a schematic showcasing select areas of improvement for this host.

4 1.3.1 Oleochemicals

As an oleaginous yeast, Y. lipolytica is a logical host to explore for the production of lipids and lipid derived products (e.g. functionalized fatty acids, fatty alcohols, and single cell oil). As such, the topic of oleochemical production in Y. lipolytica has been well reviewed in the field [4, 22, 23]. Recent advances have improved lipid synthesis and related products through both shifting cofactor utilization for improved yield [24] and gaining an understanding of lipogenesis and metabolic regulation [25]. In doing so, titers of nearly

100 g/L TAGs (process yield 0.229 g/g), 10 g/L free fatty acids, and 2 g/L fatty alcohols were achieved (enabling upwards of 100-fold improvements over initial starting strains).

Although previous studies have reported production of g/L levels of TAGS and free fatty acids in Y. lipolytica, titers of fatty alcohols have only reached tens of milligrams per liter

[25]. For context, production values of fatty acids in un-engineered model organisms such as E. coli and S. cerevisiae are 27 mg/L [26] and 42.7 mg/L [27] respectively. Additional recent studies aimed to fundamentally understand lipogenesis. One such study has identified a C2H2-type zinc finger protein, Mhy1p, that when knocked out, results in transcriptional changes in a quarter of annotated genes to rewire the lipogenic phenotype

[28]. Taken together with the study of unique nitrogen sources [29], these advances continue to demonstrate the industrial potential of Y. lipolytica to produce both native and exogenous oleochemicals.

5 1.3.2 Organic acids

The oleaginous nature of Y. lipolytica is driven by a highly active TCA-cycle, a trait that has also been exploited for the production of citric acid, an important chemical used in a variety of industries including food, pharmaceutical, and specialty chemicals [30].

In many cases, fermentation conditions (e.g. nitrogen starvation) can enhance the productivity of citric acid from a variety of inexpensive carbon sources [4, 31]. Recent efforts have continued to focus on citric acid metabolism in an effort to elucidate a mechanism underlying overproduction from glucose [32]. Beyond citric acid, metabolic engineering efforts have shifted metabolism toward high titers (all in the g/L range with values between 5 and 110 g/L depending on the acid) of intermediates including α- ketoglutaric acid [33, 34], pyruvic acid [35], succinic acid [36], and itaconic acid [37].

These efforts demonstrate both the versatility and industrial potential of Y. lipolytica as an organic acid production platform.

1.3.3 Nutraceuticals

The GRAS status of Y. lipolytica has spurred interest in using this host for the production of nutraceuticals, including carotenoids and sugar alcohols. Early and extensive work from DuPont and Microbia demonstrated the potential for producing four carotenoids at appreciable concentrations (374 mg/g DCW)—lycopene, β-carotene, zeazanthin, and canthaxanthin [38, 39]. Subsequent efforts of rational engineering for the case of lycopene have increased yields by increasing expression of known rate limiting enzymes [40] and using flux balance analysis [41]. β-carotene is a promising nutraceutical product produced 6 from Y. lipolytica with titers reaching 6.5 g/L in a fed-batch bioreactor [42] and has led to studies evaluating safety and bioavailability of Y. lipolytica derived products [43]. Further engineering of β-carotene-producing Y. lipolytica led to the production of astaxanthin as well as intermediates canthaxanthin and echinenone [44]. Similar to carotenoids, the essential vitamin riboflavin has also been produced in Y. lipolytica [45].

Outside of carotenoids, Y. lipolytica has been explored for the production of erythritol, a common additive food sweetener. Early reports demonstrated a high yield production (0.56 g/g) from raw glycerol in an acetate-negative mutant strain [46].

Subsequent strain engineering increased erythritol titer [47, 48] and productivity [49].

Additionally, fundamental studies have characterized an erythrose reductase [50] and erythritol utilization factor [51], both targets for future metabolic engineering efforts.

Collectively, these efforts increased production and concomitant yield to 0.67 g/g. Taken together, these efforts demonstrate industrial applications of nutraceutical products in Y. lipolytica.

1.3.4 Specialty chemicals

As a final class of products, specialty chemicals have been explored as new development areas for Y. lipolytica-based production. One example is erythrulose, produced from the abundant erythritol discussed above (through erythritol dehydrogenase overexpression). Erythrulose is easily converted into glyceraldehyde acetonide, a precursor to biologically active pharmaceuticals compounds [50]. In this regard, erythrulose has been produced at high rates and yields through strain engineering efforts (0.116 g/g DCW/h,

7 0.64 g/g) [52]. Another pharmaceutical precursor, campesterol, has been produced in Y. lipolytica from mevalonate with appreciable titers after screening a variety of 7- dehydrocholesterol reductases in engineered strain backgrounds [53]. Also stemming from mevalonate, researchers have produced the monoterpene alcohol linalool by successfully importing a synthetic pathway [54]. Finally, researchers have recently redirected the large acetyl-CoA and malonyl-CoA pools natively used to generate lipids to polyketides [55].

Specifically, through the investigation of three native pathways, production of the polyketide triacetic acid lactone reached unprecedented levels of 36 g/L. This effort demonstrates that Y. lipolytica is an excellent host for the potential industrial production of this important class of molecules. Collectively, the extensive engineering efforts described here demonstrate an ever expanding range of products with high titers produced using Y. lipolytica.

1.4 THE EXPANDING SYNTHETIC TOOLBOX FOR Y. LIPOLYTICA ENGINEERING

The conversion of a non-conventional yeast into a mainstream yeast requires both industrial potential (as seen above) and an improved set of genetic and synthetic tools.

Initial approaches for transformation remain limited when compared with mainstream host organisms like Saccharomyces cerevisiae which regularly achieves transformation efficiencies exceeding 107 transformants per μg DNA [56]. After an early patent for Y. lipolytica transformation was granted to Pfizer Inc. in 1992 [57], further progress has been made to improve rates including high throughput methods in microtiter plates. However, these approaches typically reach only 103 transformants per μg DNA with limited

8 efficiency for locus-specific integration (17.3%-82.5% depending on the genotype) [58].

Despite this lower efficiency, new approaches for genetic parts, CRISPR-Cas9 systems,

DNA assembly toolboxes, and model-based design greatly expand the potential of this host

(Figure 1-2).

Figure 1-2: Schematic of components in the synthetic biology toolbox.

The synthetic biology toolbox for engineering Y. lipolytica contains elements including those that enable tunable gene expression (genetic parts), targeted integrations (CRISPR- Cas9), modular cloning techniques (DNA assembly), and computer aided strain design (models).

9 1.4.1 New synthetic parts enable fine-tuned gene expression

An expanded set of synthetic control elements are required for facile engineering, and the field has seen recent developments of parts including promoters and terminators for Y. lipolytica. As starting points, native promoters have been studied and exhibit a variety of expression levels [59]. Early efforts into a hybrid promoter approach characterized two upstream activating sequences (UAS) native to the XPR2 gene- UAS1 and UAS2. In doing so, researchers demonstrate the efficacy of hybridizing several tandem repeats of a truncated UAS1 (named UAS1B) to increase promoter activity and enable enhanced constitutive expression [60] under different expression contexts [61]. This work was expanded when it was demonstrated that promoter activity was enhancer limited when paired with TEF and leucine minimal cores [62] and thus hybrid promoter engineering can enable a nearly 300-fold range of mRNA output. Subsequent efforts focused on truncating native promoter elements to identify functional UAS fragments. In doing so, a strong

UASTEF sequence was identified that could interface with the UAS1B element to establish a strong constitutive promoter regardless of culture conditions [63]. These initial studies established strong, constitutive expression to develop promoters that have been widely used by researchers for metabolic engineering of this host.

As a second level of control, inducible promoters add a dimension of temporal control to gene expression that can help decouple growth and production phases in industrial fermentation. To this end, researchers have utilized a similar hybrid promoter approach using a UAS from the native oleic acid inducible POX2 promoter to create strong promoters that were induced by exogenous oleic acid [64]. While promising, these first 10 elements suffered from leaky expression in glucose. To improve the performance of this element, the POX2 oleic acid inducible element was further characterized to achieve further repression in glucose resulting in a 48-fold induction by oleic acid. This inducible element was found to be relatively promiscuous and can also be induced by other lipids

[65].

Similar strategies of identifying and hybridizing metabolite responsive elements have expanded the suite of inducibility to include alkanes and erythritol. Alkane responsive elements from the ALK1 promoter, inducible by n-decane, were identified and hybridized to enable 5-fold higher expression [66]. Likewise, an upstream inducible element from

EYK1 has recently been explored to enable dose-dependent induction by erythritol and erythrulose, a feature that may enable a more fine-tuned induction over the previously described lipid and alkane inducible promoters [67]. Ongoing efforts in developing new synthetic promoters for this host will further enhance the capability of gene expression control for Y. lipolytica.

Synthetic fine-tuned gene expression control extends beyond promoters and into terminators. For fungal systems, terminators can improve protein yields from transcripts and can increase enzyme abundances [68] often through altered mRNA stability and half- life [69]. While terminators have not been fully characterized in Y. lipolytica, synthetic designs for S. cerevisiae remain functional in Y. lipolytica and present a 70% improvement in gene expression compared with common, endogenous terminators [70]. Taken together, the field of synthetic biology has had a great impact in expanding the set of strong and

11 inducible gene expression control elements in Y. lipolytica. In turn, these elements have led to improved success in strain engineering efforts.

1.4.2 CRISPR-Cas9 systems increase targeted integration efficiencies

Traditional genome engineering in Y. lipolytica is dominated by NHEJ favoring random integration of exogenous DNA over homologous repair. While this molecular mode can be circumvented by using a ku70 deficient strain, long-term stability disfavors this genotype. In recent years, the use of CRISPR technologies has begun to emerge as a facile way to improve engineering efforts in non-conventional yeasts [71]. These applications were first implemented via the use of synthetic RNA polymerase III (Pol III) promoters as well as an RNA polymerase II TEF promoter using hammerhead and HDV ribozymes to drive sgRNA expression [72] and were first patented by DSM [73]. In the initial study by Schwartz et al., three genes were targeted for disruption (PEX10, MFE1, and KU70), with varying degrees of success, reporting single gene disruption efficiencies above 90% using a Pol III promoter to drive sgRNA expression [72]. In parallel work, increased disruption frequencies were seen to enable multiplex gene disruptions (two to three genes) in a single step, albeit with low efficiencies [74]. Recently, T7-transcribed sgRNAs have also been used in Y. lipolytica to enable CRISPR-Cas9 mediated knockouts with 60% efficiency [75]. These initial reports demonstrate promise, but several variables

(especially conditions and outgrowth steps) still require optimization.

Advances in the field continue to push the boundary of CRISPR-Cas9 applications in yeast. For example, researchers have developed two separate systems, one capable of

12 targeted markerless gene integration [76] and another that enables CRISPR-mediated gene repression [77]. In the first case, five loci were ultimately identified that could serve as rapidly accessible hot spots for expression. The utility of this approach was demonstrated by integrating four genes comprising a semisynthetic lycopene biosynthesis pathway into four of the five identified loci [76]. In the second case, repression via a dCas9-Mxi fusion was used to target KU70 and KU80 for repression, thus enabling a higher frequency of homologous recombination in selected cells. This repression was temporally controlled and could be reversed by clearing the plasmid encoding the dCas9 system [77]. Finally, this growing toolkit of CRISPR-Cas9 technologies have been incorporated into several cloning tool sets (including EasyCloneYALI [78] and YaliBricks [79]) described below.

1.4.3 DNA assembly toolboxes enable rapid testing of pathways

Cloning standardization efforts have created a set of toolboxes for cloning vectors and integrating genes into target loci for Y. lipolytica. The first use of a multi-gene DNA assembly method for this organism was reported in 2014. In this experiment, a single in vivo homologous recombination step was used to concatenate the three genes comprising the β-carotene production pathway with a positive selection marker. This four-gene construct (~11 kb designed with distinct promoter/terminator pairs for each gene) was successfully integrated into the genome with a highest reported efficiency of 21% [80].

Recently, pathway assembly and construction has been improved through the Golden Gate

DNA assembly platform. In this regard, this approach is more cost and time efficient and results in higher success rates compared to methods relying on homologous recombination

13 in Y. lipolytica. The efficacy of this Golden Gate platform was once again demonstrated by assembling the β-carotene pathway with a positive selection marker. Using this approach, researchers reported efficiencies ranging from 67%-90% when their fully assembled construct was transformed into yeast [81].

As alternatives to Golden Gate cloning methods, modular multi-part assembly has been developed and tested using BioBrick standards called YaliBricks. The efficacy of this approach was successfully tested through assembly of a 12kb, five gene violacein pathway in a parallel cloning reaction in under a week to produce the pigment in Y. lipolytica, though no efficiency was reported [79]. A final modular cloning technique termed

EasyCloneYALI utilizes CRISPR-Cas9 to enable markerless integration of up to five vectors into targeted loci with efficiencies above 80% [78]. This vector set comprises eleven dominant selection marker-containing vectors and a protocol to knockout two genes simultaneously using their vector sets. Finally, recent developments have established a vector set expanding dominant selection schemes to enable footprint-free genome editing using a piggyBac transposase [82]. In this study, researchers demonstrated the efficacy of genome-wide mutagenesis to identify improvements in pigment and lipid formation. These modular cloning and assembly approaches enable facile construction of rapid and complex cargo for Y. lipolytica.

1.4.4 Model based approaches guide the future of metabolic engineering

The aforementioned advances in metabolic engineering and synthetic biology for

Y. lipolytica are complemented by enhanced ‘omics technologies and model-guided

14 engineering in this host (Table 1-1). The field is beginning to see the use of model based approaches to better understand and predict metabolism (esp. lipid accumulation). The first integrated data study of Y. lipolytica involved developing a genome-scale metabolic model

(GEM) coupled with metabolomics and lipidomics to elucidate the linkage between different growth conditions and lipid formation. Specifically, this study highlighted the regulatory role amino acid biosynthesis (rather than regulation of lipid metabolism itself) exerts on lipogenesis, thus demonstrating that the mode of lipid metabolism is similar to ethanol overflow metabolism in S. cerevisiae [83]. A further study on lipid metabolism transcriptomics identified the role of the MHY1 gene as described above [28], a finding that further highlights the complex interaction between amino-acid and lipid biosynthesis.

In this study, they used RNA-seq to elucidate differential gene expression then classified those genes and used pathway enrichment to determine the effect of a MHY1 gene disruption. Further global models of gene regulation have identified and then subsequently experimentally validated six identified master regulators of lipid metabolism leading to a wide variation in lipid accumulation through construction of a genome-wide coregulatory network [84]. Assuming quasi-steady state allowed for decomposition of metabolic networks into minimal pathways that were used to build dynamic metabolic models to describe lipid accumulation and predict potential targets [85]. While most of the above efforts have focused on lipogenic phenotypes, recent studies have used metabolome data for principal components analysis to extract pathway information and construct a basic visualization network to improve the production of heterologous lycopene [86]. Genome scale-models and global metabolic understanding are continuing to enhance the systems 15 metabolic engineering approach in Y. lipolytica and provide better starting points for industrial strain engineering efforts.

Table 1-1: Models that have been developed for Y. lipolytica metabolism.

Approach Outcome Reference Researchers used metabolite profiling and lipidomics as well as Identified transcriptional transcriptional quantification using regulation of lipid metabolism is 83 RNA-seq from chemostat cultures associated with regulation of under carbon and nitrogen limited amino-acid biosynthesis. conditions to build this GEM. Researchers used RNA-seq to Identified complex regulatory compare regulation in a Mhy1p role of MHY1 including deficient strain to wildtype yeast by increased lipid biosynthesis, 28 analyzing differentially expressed increased cell cycle rate, and genes through pathway enrichment. inability to form hyphae. Researchers constructed a genome- Identified context-specific wide coregulatory network from transcrition factors and timecourse transcriptomic data where experimentally validated 84 Y. lipolytica was fermented under transcrition factors involved in nitrogen-limited conditions. lipid accumulation. Researchers developed three dynamic Three models were built, metabolic models using decomposed statistically validated, and metabolic networks assumed to be at compared to one another, but no 85 quasi-steady state from data collected experimental validation was at a 20L growth scale. sought. Researchers used metabolomics to extract pathway information and Identified regulation effects of construct a basic visualization eleven pathways effected by 86 network comparing a control strain production of lycopene. and lycopene-producing strain of Y. lipolytica.

16 1.5 FUTURE OUTLOOK FOR Y. LIPOLYTICA ENGINEERING

Y. lipolytica is quickly becoming a model non-conventional organism with strong applications in industrial metabolic engineering. Distinct metabolic advantages including high flux through the TCA cycle makes Y. lipolytica a highly attractive host for chemical production in a variety of areas including oleochemicals, organic acids, nutraceuticals, and specialty chemicals. Likewise, industrially advantageous traits including the lack of tight metabolic regulation, tolerance to chemicals, and high protein production and secretion abilities make Y. lipolytica an attractive fungal host. Recent advances in synthetic biology are enabling a rapid industrialization of products derived from Y. lipolytica metabolism.

Although considerable progress has occurred, several challenges remain in improving transformation efficiency, building new genetic tools, full implementation and utilization of CRISPR-Cas9, and developing more descriptive models (see Outstanding

Questions). We believe that progress is stymied by the inability to screen large libraries due to limited targeted integration efficiencies. At least for the short term, it seems unlikely that a stable episomal plasmid system is realizable, so integration efficiency needs improved to grasp the full potential of library screening in this host. To remedy the inability to screen large libraries, new models providing a clearer picture of Y. lipolytica metabolism will increase the speed at which new strains can be engineered by guiding selection of potential targets for gene manipulation. An additional disadvantage arises from the dimorphic nature of Y. lipolytica that can lead to difficulty in bioprocessing specifically when cells are filamentous, though studies have demonstrated ways to limit growth to the

17 yeast form. The discovery space for improving tools in Y. lipolytica is large and provides great potential for high-impact work in this host.

Though there is abundant need for researchers to develop new tools, current limitations are not precluding recent progress in the field of non-conventional yeast engineering. Lessons learned in developing these new and necessary tools for Y. lipolytica will lead to methodologies that can be applied to other promising oleaginous yeast hosts like Lipomyces starkeyi, Rhodosporidium toruloides, and Trichosporon oleaginosus. The rapid adoption of Y. lipolytica as an academic and industrial host of inquiry gives promise that this host will find new avenues for industrial-scale production in the coming years.

18 Chapter 2: Engineering Yarrowia lipolytica for the production of cyclopropane fatty acids2

2.1 CHAPTER SUMMARY

Traditional synthesis of biodiesel competes with food sources and has limitations with storage, particularly due to limited oxidative stability. Microbial synthesis of lipids provides a platform to produce renewable fuel with improved properties from various renewable carbon sources. Specifically, biodiesel properties can be improved through the introduction of a cyclopropane ring in place of a double bond. In this study, we demonstrate the production of C19 cyclopropane fatty acids in the oleaginous yeast Yarrowia lipolytica through the heterologous expression of the Escherichia coli cyclopropane fatty acid synthase. Ultimately, we establish a strain capable of 3.03 ± 0.26 g/L C19 cyclopropane fatty acid production in bioreactor fermentation where this functionalized lipid comprises over 32% of the total lipid pool. This study provides a demonstration of the flexibility of lipid metabolism in Y. lipolytica to produce specialized fatty acids.

2.2 INTRODUCTION

Until recently, the energy and fuel sector has relied almost exclusively on nonrenewable, petroleum-based sources. For the case of liquid transportation fuels, two major alternatives are available in the United States—corn ethanol and soybean-based biodiesel. However, these two products are not without their own issues as their production

2 Thank you to Andrew Hill for his initial work on CFA synthase enzymes in Yarrowia. 19 competes with land and food resources [87]. The production of second generation ethanol seeks to ameliorate this issue [88]. For the case of biodiesel, microbial oil production is emerging as an alternative to more traditional soybean oil-based processing [89, 90].

Microbial-derived biodiesel is most commonly made through transesterification of biogenic lipids into fatty acid methyl esters (FAMEs) [91]. The direct microbial conversion of sugars into FAME has been explored in the field, but suffers from limited titers [92]. Oleaginous yeasts such as Yarrowia lipolytica are particularly promising hosts for biodiesel production as they accumulate and store lipids almost exclusively as triacylglycerides (TAGs) in lipid bodies [93]. In this regard, TAGs are the preferred lipid substrate for transesterification as they lack the phosphorous, sulfur, and nitrogen present in other lipid species and they contain a high percentage of constituent fatty acids [91]. As a result, conversion rates of over 99% can be achieved in converting TAGs into biodiesel

[91].

The chemical composition of FAMEs can influence overall biodiesel storage capacity and performance. Specifically, FAMEs that have unsaturation are essential for fuel fluidity, but provide a chemical moiety that is subject to oxidation, thus presenting a major challenge for the long-term storage of biodiesel [94]. In part, this limitation can be overcome through cyclopropanation of unsaturated fatty esters as a means to maintain biodiesel fluidity with increased energy density [94]. Moreover, further chemical modification by a metal catalyst can lead to the synthesis of branched-chain fatty acids—a molecule with utility in lubricants [95]. Thus, cyclopropane fatty acids (CFAs) are inherently useful in the lubrication and oleochemical industries [96]. 20 Natively, CFAs are produced in a variety of different organisms ranging from bacteria to plants and comprise varying percentages of the total fatty acid pool, as high as

46% (Table 2-1). For plants, this lipid pool is stored in seeds, thus requiring large volumes of processing to extract these compounds [97]. In bacteria, CFAs serve as membrane protectant molecules that promote membrane fluidity in face of adverse conditions [98].

For example, in E. coli, CFAs are preferentially produced in late exponential and early stationary phases [99] and allow for increased resistance to acids [100-102] and phenols

[103]. Synthesis of these modified fatty acids are catalogued in many bacteria beyond E. coli [104] including in Pseudomonas putida where these molecules help provide freeze- drying protection [105]. Mechanistically, CFAs are formed through a common path in bacteria whereby the synthase transfers a methyl group from S-adenosylmethionine across the cis double bond to form a cyclopropane ring (Figure 2-1) [106]. Within these native hosts, CFA synthesis is typically isolated to membrane-bound phospholipids, a feature that leads to a less desired pool for biodiesel production compared with TAGs. Therefore, there is promise in using an alternative host that can create CFAs in the form of TAGs.

21 Table 2-1: Comparative study of native cyclopropane fatty acid production levels in bacterial and plant hosts demonstrates a range of CFA species and compositions

Species CFA Species % CFA Reference Escherichia coli C17CP, C19CP 14, 10 [101] Pseudomonas putida C17CP, C19CP 33, 4 [105] Chlorobaculum tepidum C17CP 46 [107] Streptococcus faecalis C17CP, C19CP 23, <1 [108] Serratia marcescens C17CP, C19CP 31, 5 [109] Pseudomonas fluorescens C17CP, C19CP 3, 8 [109] Cotton C19CP <1 [110] Litchi chinensis C11CP-C19CP 42 [111]

Figure 2-1: Mechanism of action for enzymatic CFA formation.

The CFA synthase transfers a methyl group from S-adenosylmethionine across the cis double bond of a lipid to form a cyclopropane fatty acid.

In this work, we sought to upgrade the fatty acid profile in Y. lipolytica by producing CFAs from the ample precursor pool of unsaturated fatty acids that get stored as TAGs. Y. lipolytica is a promising host for the production of a variety of acetyl-CoA and malonyl-CoA derived molecules including TAGs [112] and polyketides [55].

Improved lipogenesis in this host has been reported [113] and reviewed in the literature including efforts to convert these precursors into value-added chemicals like polyunsaturated fatty acids, fatty alcohols, fatty esters, and fatty alkanes [4, 23, 114].

22 Likewise, this host has been engineered to create designer fatty acids like ricinoleic acid

[115, 116].

In prior work from the Alper research group, two high-lipid overproducing strains of Y. lipolytica based on the wildtype strain PO1f were established. First, a rational metabolic engineering approach was used to achieve a 60-fold improvement in lipid titer

(exceeding 25 g/L in bioreactor fermentation) through the disruption of β-oxidation by knocking out pex10 and mfe1 paired with overexpression of DGA1 [113]. For the case of this study, we will refer to this strain as ENGR for simplicity. Second, a reverse engineering approach was used to identify and characterize a mutant regulatory protein encoded by

MGA2 to create a strain named L36 which, when complemented with DGA1 overexpression, likewise produces 25 g/L of lipids in bioreactor fermentation [117]. For the case of this study, we will refer to this strain as L36-DGA1 for simplicity. Both of these strains accumulate significant quantities of oleic acid, thus providing a promising starting point for resulting C19 CFA (C19CP) production. Here, we demonstrate that complementation of the E. coli CFA synthase into these high-lipid producing strains of Y. lipolytica can enable C19CP production that reaches over 32% of the total lipid pool with titers of over 3 g/L in bioreactor fermentation.

23 2.3 RESULTS AND DISCUSSION

2.3.1 CFA production in a previously engineered-strain of Y. lipolytica ENGR

Natively, the E. coli CFA synthase converts unsaturated bonds in phospholipids into cyclopropane groups through the simultaneous conversion of s-adenosylmethionine into s-adenosylhomocysteine (Figure 2-1). Unlike E. coli, Y. lipolytica primarily stores lipids as TAGs. However, previous work in microalgae demonstrated that E. coli CFA synthase overexpression can lead to cyclopropanated TAG species (although detrimental growth effects were observed) [118]. Using this preliminary result, we sought to enable production in Y. lipolytica through the expression of a codon optimized version of this synthase (referred to as ycoCFA).

As a production host for CFAs, we sought to utilize strains ENGR and L36-DGA1

(described above) based on their increased lipid content and C18:1 skewed fatty acid pools compared to wildtype PO1f. Specifically, this wild-type cell produces only 31.4% of fatty acids as C18:1 with a resulting titer of only 52.3 mg/L—a level that is insufficient as a starting point for high-level C19CP production (Figure 2-2). In contrast, the first strain in this study, ENGR, produces 45.0% of its total lipids as C18:1, a value that results in 616 mg/L titer in tube fermentations (Figure 2-3). Initial overexpression of ycoCFA in ENGR resulted in 118 mg/L production of C19CP in tube fermentations (Figure 2-3), a value that is equivalent to 10.0% of total lipids. Indeed, the production of C19CP resulted in a conversion of C18:1 and reduced the size of this pool to 34.8% of total fatty acids without having any impact on saturated fatty acid production. To further improve production, we

24 sought to include a second copy of ycoCFA which resulted in a near doubling of production to 190 mg/L of C19CP in tube fermentation (Figure 2-3). In this resulting strain, total

C19CP content was 20.5% of the total lipid pool leaving only 21.8% C18:1 with a relatively untouched saturated fatty acid distribution, albeit with a slightly reduced total lipid production. Notably, for both ycoCFA overexpression strains, the C16:1 pool is decreased suggesting additional, marginal production of C17CP. Similarly, both overexpression strains demonstrate an increase in C18:2 production compared to ENGR. Collectively, this initial engineering in ENGR enabled C19CP production of almost 200 mg/L demonstrating that Y. lipolytica is a viable host for TAG-based production of C19CP.

Figure 2-2: Lipid distribution of wildtype PO1f.

The lipid titer of each lipid species produced in wildtype Y. lipolytica after a three day tube fermentation in CSM complete identified via GC. Error bars represent the standard deviation of n=3.

25 Figure 2-3: Overexpression of ycoCFA enables production of C19CP in ENGR.

The lipid titer of each species of lipid produced by strain ENGR and cyclopropane- producing derivatives of this lineage after three day tube fermentation in CSM complete as measured by GC. Error bars represent the standard deviation of n=3. Significance was tested using Dunnett’s test.

2.3.2 CFA production in an alternative strain of Y. lipolytica, L36-DGA1

To compare the results of CFA production in ENGR, we sought to evaluate production in an alternative engineered strain, L36-DGA1. Like ENGR, this strain has a higher production of C18:1 lipids than wildtype (in L36-DGA1 C18:1 production is 506 mg/L), which is 20% less in titer than ENGR, but comprises a higher percentage of the total lipids (Figure 2-4). Accordingly, the lipid distribution in the L36-DGA1 strain is unique as it has a much lower percentage of saturated fatty acids (19.9% total compared with 41.2% in ENGR). As a result, this shifted profile toward unsaturated fatty acids can be more advantageous for biodiesel production, thus we conducted a similar engineering strategy for CFA production in his host. The initial pool of C18:1 in L36-DGA1 is 26 equivalent to 55.3% of the total lipids and initial overexpression of ycoCFA generated 191 mg/L of C19CP (Figure 2-4). This C19CP titer represents 19.2% of the total lipid pool in this strain. Likewise, a concomitant reduction in the C18:1 lipid pool was seen to comprise

39.5% of the total lipids. Further overexpression of the ycoCFA synthase increased C19CP only slightly to 22.3% of the total lipids, but resulted in a substantially decreased overall lipid production and thus only 171 mg/L of desired product was observed. A copy number analysis of these strains (Table 2-2) suggests that this L36-based strain is still limited in copy number, thus the ENGR strain was used for further study. Nonetheless, in similar fashion to the ENGR strain above, overexpression of ycoCFA in L36-DGA1 did not impact the saturated fatty acid content. Yet, increases in C18:2 were not observed as they were in

ENGR. Thus, engineering in these two different backgrounds demonstrates the impact that the initial lipid distribution profile can have on product distribution.

27 Figure 2-4: Overexpression of ycoCFA enables production of C19CP in L36-DGA1.

The lipid titer of each species of lipid produced by strain L36-DGA1 and cyclopropane- producing derivatives of this lineage after three day tube fermentation in CSM complete as measured by GC. Error bars represent the standard deviation of n=3. Significance was tested using Dunnett’s test

Table 2-2: Genomic copy number of ycoCFA in all strains used in this study

Strain ycoCFA Copy Number PO1f 0 ENGR 0 ENGR HPH ycoCFA 1 ENGR HPH NAT ycoCFA 2 L36-DGA1 0 L36-DGA1 NAT ycoCFA 1 L36-DGA1 NAT URA3 ycoCFA 1

28 2.3.3 Bioreactor production and characterization of CFA production

Based on promising initial tube fermentations of ENGR NAT HPH ycoCFA for the production of C19CP in Y. lipolytica, we performed a scaled-up production test at the 3 L bioreactor scale using a pulse-fed bioreactor scheme similar to our previously published methodology [113]. In this experiment, the ENGR NAT HPH ycoCFA strain fermented in minimal media with a pulse feeding of glucose was able to produce a final C19CP titer of

3.13 ± 0.34 g/L in a representative bioreactor run (Figure 2-5a). This titer corresponds to

32.7% of the total lipid titer (the complete distribution of which is shown in Figure 2-5b).

Notably, this value of 32.7% C19CP greatly surpasses the 20.5% observed at the tube fermentation scale, though the C18:1 composition is similar under both conditions. These seemingly contradictory results are explained by the accumulation of C16:0 and C18:2 in tube fermentations, a phenomenon absent in these bioreactors. To this end, the composition of fatty acids that are C18:1 or derived from C18:1 exceeds 63.9% in the bioreactor compared with only 56.9% in tube fermentation. These results demonstrate a higher overall flux toward C18:1 production. Collectively, these bioreactor results demonstrate benefits of both (i) improved conversion toward C19CP products and (ii) a reduction in the oxidation-prone C18:2 pool. An independent replicate 3 L fermentation resulted in nearly similar characteristics with a C19CP titer of 2.94 ± 0.16 g/L (Figure 2-

6a) and CFAs comprising 30.9% of the total lipid pool (Figure 2-6b).

29 Figure 2-5: Pulse-fed bioreactor fermentation of ENGR HPH NAT ycoCFA a) Replicate 1 glucose level, C19CP titer, and viable cell count. b) Lipid composition of final reactor time point.

a) Three samples were taken from pulse-fed bioreactor fermentation in minimal media every 24 hours and analyzed via GC for C19CP lipid titer, HPLC for glucose level, and plating for viable cell count. Error bars represent the standard deviation of n=3 for each of these measurements. b) Lipid composition was measured via GC. Each bar represents the percent of a specific lipid species out of the total lipid content for the final time point of the representative bioreactor fermentation. Error bars represent the standard deviation of n=3

30 Figure 2-6: Pulse-fed bioreactor fermentation of ENGR HPH NAT ycoCFA a) Replicate 2 glucose level, C19CP titer, and viable cell count. b) Lipid composition of final reactor time point.

31 2.4 CONCLUSIONS

Through this work, bioreactor fermentation enabled titers exceeding 3 g/L for the specialized, cyclopropane fatty acid species. These results demonstrate the promise of Y. lipolytica as a host for producing functionalized fatty acids. Furthermore, these high titers represent over 30% of the total fatty acid pool, a value that is comparable to the highest percentage of a single CFA species observed in bacteria. Likewise, this value also approaches the 42% observed in Litchi chinesis, a value that is reported as the sum of a variety of different CFA species [111]. The overall reduction in unsaturated bonds enabled by expression of ycoCFA in the high-lipid ENGR parental stain provides a TAG pool that has advantageous, upgraded properties for biodiesel production.

In summary, this work demonstrates the first reported production of a cyclopropanated fatty acid species in the oleaginous host Y. lipolytica. Through importing the CFA synthase from E. coli into two high-lipid producing engineered strains of Y. lipolytica, production approaching 200 mg/L of C19CP was achieved in tube fermentations. Moreover, titers exceeding 3 g/L of C19CP were achieved through bioreactor fermentation in minimal media. This bioreactor production of C19CP represented up to 32.7% of the total lipids produced and rivals the values observed in native

CFA producers. Engineering in two different parental lineages demonstrates the importance of lipid distribution on downstream product engineering. Collectively, this work expands the industrial potential of Y. lipolytica for the production of fatty acid- derived products.

32 Chapter 3: Rewiring Yarrowia lipolytica toward triacetic acid lactone for novel materials generation3

3.1 CHAPTER SUMMARY

Polyketides represent an extremely diverse class of secondary metabolites that are often explored for their bioactive traits. These molecules are also attractive building blocks for chemical catalysis and polymerization. However, the use of polyketides in larger-scale chemistry applications is stymied by limited titers and yields from both microbial and chemical production. Here we demonstrate that an oleaginous organism (specifically,

Yarrowia lipolytica) can overcome such production limitations owing to a natural propensity for high flux through acetyl-CoA. By exploring three distinct metabolic engineering strategies for acetyl-CoA precursor formation, we demonstrate that a previously uncharacterized pyruvate bypass pathway supports increased production of the polyketide triacetic acid lactone (TAL). Ultimately, we establish a strain capable of producing over 35% of the theoretical conversion yield to TAL and an averaged maximum observed titer of 35.9 ± 3.9 g/L with an achieved productivity of 0.21 ± 0.03 g/L/h in bioreactor fermentation. Additionally, we illustrate that a β-oxidation related overexpression can support high TAL production and is capable of achieving over 43% of the theoretical conversion yield under nitrogen starvation. Next, through use of this

3 The content in this chapter is adapted from a previously authored publication. KAM and CMP equally contributed to experiments and analyses, and collectively wrote the manuscript comprising this chapter. MC conducted all materials experiments and wrote the materials sections with NAL. Reprinted with permission from Markham KA, Palmer CM, Chwatko M, Wagner JM, Murray C, Vazquez S, Swaminathan A, Chakravarty I, Lynd NA & Alper HS (2018) Rewiring Yarrowia lipolytica toward triacetic acid lactone for materials generation. Proceedings of the National Academy of Sciences of the United States of America 115: 2096-2101. 33 bioproduct, in collaboration with the Lynd lab at UT-Austin, we demonstrate the utility of polyketides like TAL to modify commodity materials such as poly(epichlorohydrin) resulting in an increased molecular weight and shift in glass transition temperature.

Collectively, these findings both establish an engineering strategy enabling unprecedented production from a type III polyketide synthase as well as establish a route through O- functionalization for converting polyketides into new materials.

3.2 INTRODUCTION

The growing demand for renewable chemicals and fuels has spurred great interest in using cells as biochemical factories [119]. Metabolic engineering enables this goal by rewiring cells’ metabolism toward desirable chemical compounds [2, 4, 112]. Among possible molecules, polyketides are an interesting class of secondary metabolites produced by microbes and plants with native roles in processes such as cellular defense and communication [120-122]. While many polyketides can serve as potent antibiotics, this class of molecules also encompasses chemicals with other useful properties such as pigments, antioxidants, antifungals, and other bioactive traits [120, 123]. However, the use of polyketides in more unique and non-medical applications has been partially limited due to low natural abundance and difficult cultivation of native hosts. Specifically, polyketide producing organisms are typically unusual plants and bacteria that are not well suited for high-level industrial production [124]. Synthetic production of these molecules in model host organisms has also proven quite difficult with titers and yield insufficient for industrial

34 production [125-128]. Likewise, traditional chemical synthesis of polyketides is limited by low concentrations and challenging chiral centers [129]. While the scale and price-point for pharmaceuticals can tolerate plant-based sourcing of polyketides or challenging syntheses, this is not an option for any larger-scale chemistry application.

Here, we focus on the interesting, yet simple polyketide, triacetic acid lactone

(TAL) as it is derived from acetyl-CoA and malonyl-CoA—two common precursors for many other polyketides. TAL has been demonstrated as a platform chemical that can be converted into a variety of valuable products traditionally derived from fossil fuels including sorbic acid, a common food preservative with a global demand of 100,000 metric tons [119, 130-133]. However, meeting this annual demand using the low concentrations of TAL derived from native plants like gerbera daisies [124] would require four-times the quantity of global arable land. As a result, utilization of polyketides for unique industrial applications including polymers, coatings, and even commodity chemical production has not been considered despite the promising chemical nature of these molecules. To address these limitations, previous efforts have explored microbial production of polyketides like

TAL. However, these efforts have been restricted to conventional organisms (like E. coli

[125, 134] and S. cerevisiae [126-128]) and are limited with respect to titer only reaching

5.2 g/L with low yields [128].

In this work, we explore the unique application of an oleaginous, non-conventional yeast (Yarrowia lipolytica) based on its potential for high flux through the key polyketide precursors, acetyl-CoA and malonyl-CoA. By investigating three distinct pathways toward

CoA precursor formation along with targets hypothesized to enhance β-oxidation, we 35 demonstrate the utility of a previously uncharacterized pyruvate bypass pathway for significantly increasing TAL production. After subsequent optimization, our final strain achieved over 35% of the theoretical conversion yield to TAL and a maximum observed titer of 35.9 ± 3.9 g/L in bioreactor operation. We demonstrate that a higher yield strain

(43% of theoretical conversion yield) is possible by overexpressing a β-oxidation related target. Finally, we demonstrate the chemical opportunities gained by high polyketide titers for novel materials modification by O-functionalization of biosourced TAL with commodity poly(epichlorohydrin) to tune and upgrade thermal properties of the parent material. This work both establishes a novel host organism for polyketide overproduction and demonstrates the potential utility of polyketides for materials synthesis and modification.

3.3 RESULTS AND DISCUSSION

3.3.1 Y. lipolytica can support TAL production

High lipid flux in oleaginous organisms like Y. lipolytica suggests a high potential for these organisms to produce alternative acetyl-CoA derived products like polyketides.

Moreover, Y. lipolytica exhibits an extraordinary tolerance to many chemicals [135] including TAL to concentrations approaching the soluble limit (Figure 3-1). With these two features in place, we first established heterologous TAL production in Y. lipolytica through expression of the codon-optimized Gerbera hybrida 2-pyrone synthase gene, g2ps1. While this initial strain produced TAL, further amplifying the gene copy number to four enabled 2.1 g/L production in tube fermentations. This strain, named YT, was selected 36 as the starting point for further metabolic engineering work. Additional studies of previously characterized mutants of g2ps1 established in E. coli [125] were also tested, but produced lower titers than the wild-type allele (Figure 3-2).

Figure 3-1: TAL inhibits growth rate of Y. lipolytica but is not lethal.

0.3

0.25

) 0.2

(

a 0.15

G 0.1

0.05

0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 TAL Concentration (g/L)

The wildtype Y. lipolytica strain, PO1f, was fermented in defined media, in the presence of varying concentrations of TAL. Growth was assessed via OD600nm using a Bioscreen C. Error bars represent the standard error of n=3.

37 Figure 3-2: Previously identified g2ps1 mutations do not promote TAL production in Y. lipolytica.

0.6

0.5

0.4

)

( *

T 0.3

e 0.2

0.1

*** 0 g2ps1 g2ps1 L261N g2ps1 L261F g2ps1 L202G L261N TAL production was assessed following 48-hour tube fermentations in defined media. Error bars represent standard deviation of n=10-14. Significance was tested using Dunnett’s test to compare mutants to wildtype g2ps1, * p< 0.05, ** p< 0.01, *** p<0.001.

3.3.2 A previously uncharacterized Y. lipolytica pyruvate bypass pathway improves

TAL production

To increase metabolic flux through acetyl-CoA and malonyl-CoA, we investigated three independent metabolic engineering strategies (Figure 3-3). First, we explored the citrate route—a pathway that has been extensively studied for its capacity to increase lipid production in Y. lipolytica [136-138]. Overexpression of ACC1, the enzyme that converts acetyl-CoA to malonyl-CoA, has also been shown to promote lipid production [139, 140].

38 When the pathway enzymes (ACL1, ACL2, and AMPD) were concurrently overexpressed,

TAL production was significantly reduced and only marginally improved with the addition of ACC1 (Figure 3-4).

Figure 3-3: Strain engineering scheme to evaluate native Yarrowia lipolytica pathway potential for the production of TAL.

Four overall schemes were tested in this work. Illustrated here are the three anabolic pathways targeted for overexpression in this work: the citrate route shown in blue, the pyruvate dehydrogenase complex in green, and the pyruvate bypass pathway in purple. Additionally, shown in red are two potential β-oxidation upregulation targets. These color schemes are maintained in future Figures to enable continuity and rapid identification.

39 Figure 3-4: Difference of means plot demonstrating the effect of overexpressing the citrate pathway genes.

TAL titers were measured following four-day tube fermentations in defined media and presented as the increase in titer over the YT parental strain (the color scheme used in Figure 3-3 has been retained). Error bars represent the standard error of n≥2. Significance was tested using Dunnett’s test, * p< 0.05, ** p< 0.01, *** p<0.001. The effect of sequential overexpression of genes involved in the citrate pathway.

Second, we explored the pyruvate dehydrogenase (PDH) complex pathway. In contrast to the citrate route, this pathway has not been extensively studied in Y. lipolytica, but would theoretically enable a direct path to convert pyruvate to acetyl-CoA. The related alpha-ketoglutarate dehydrogenase complex, which shares one subunit with the PDH complex, has been previously overexpressed to promote alpha-ketoglutaric acid production

[141], suggesting a similar strategy may be successful here. To test this approach, we

40 established a coordinated overexpression of the different subunits for this complex

(encoded by PDA1, PDE2, PDE3, and PDB1). By combining this pathway with ACC1 overexpression, overall TAL production was significantly improved by 23%, achieving 2.5 g/L (Figure 3-5).

Figure 3-5: Difference of means plot demonstrating the effect of overexpressing the pyruvate dehydrogenase complex genes.

41 Third, we investigated the pyruvate bypass pathway, which converts pyruvate to acetaldehyde through pyruvate decarboxylase (PDC), then to acetate through acetylaldehyde dehydrogenase (ALD), and finally to acetyl-CoA via acetyl-CoA synthetase (ACS) [142, 143]. Previous work has targeted this pathway using heterologous enzymes [25], however the function and potential of the native Y. lipolytica pyruvate bypass pathway has not been previously explored. While a single ACS gene had been previously characterized in Y. lipolytica [142] two PDC homologs (arbitrarily named

PDC1 and PDC2) and five potential ALD homologs were identified based on previous yeast homology studies [144]. Next, we established a full combinatorial assembly of this pathway in the YT strain background. Unlike the previous two approaches, ACC1 overexpression did not consistently increase production for all constructs. Nevertheless, four genetic combinations emerged as the top TAL producing strains including ACS1,

ALD5, PDC2, ACC1 (64.7% improvement over YT), ACS1, ALD3, PDC1, ACC1 (61.1% improvement), ACS1, ALD2, PDC2, ACC1 (31.7% improvement), and ACS1, ALD3,

PDC2, ACC1 (17.9% improvement) (Figure 3-6). The top strain from this effort (YT-

ACS1, ALD5, PDC2, ACC1) produced 2.8 g/L of TAL in tube fermentations, equivalent to

30.4% of the theoretical yield.

42 Figure 3-6: Difference of means plot demonstrating the effect of overexpressing the pyruvate bypass pathway genes.

3.3.4 Modification of β-oxidation can likewise improve TAL production

As an alternative (and potentially complementary) approach to increase acetyl-CoA pools, we targeted participants in the β-oxidation pathway for overexpression. Specifically, we evaluated the transcription factor POR1 (reported in other hosts to increase polyketide 43 formation [145]) and the peroxisomal matrix protein Pex10. When overexpressed in the

YT background, POR1 had no effect on TAL production whereas PEX10 overexpression increased TAL titer by 22% (2.4 g/L) (Figure 3-7), suggesting β-oxidation upregulation as strategy for acetyl-CoA recycling if it cannot be shuttled away from lipid synthesis effectively. This result is intriguing as Pex10p is not directly involved in the catalytic conversion of fatty acids to acetyl-CoA and thus provides an area for further biochemical study.

Figure 3-7: Difference of means plot demonstrating the effect of overexpressing β- oxidation targets.

TAL titers were measured following four-day tube fermentations in defined media and presented as the increase in titer over the YT parental strain. Error bars represent the standard error of n≥2. Significance was tested using Dunnett’s test, * p< 0.05, ** p< 0.01, *** p<0.001.

44 3.3.5 Strain engineering effectively diverted flux from lipids to TAL

To evaluate the true efficacy of these four rewiring strategies, both lipid and TAL production were evaluated from tube fermentations. We observed a clear, inverse correlation between TAL titer and lipid titer with an R2 of 0.89 (Figure 3-8). Collectively, these results demonstrate that the PDH complex, pyruvate bypass, and PEX10 overexpressions can divert acetyl-CoA from lipids into TAL whereas the citrate pathway is strongly coupled to lipid formation.

Figure 3-8: Lipid production as a function of TAL production.

Following defined media tube fermentation of YT and strains containing the overexpressions outlined in Figure 3-3, average TAL titer and average total lipids were assessed. An inverse correlation between TAL titer and lipid titer was observed, R2 = 0.89. Error bars represent the standard deviation of n=3.

45 3.3.6 Culture conditions differentially alter TAL production

Y. lipolytica is highly responsive to environmental factors and thus we evaluated a

series of conditions for increased TAL production. First, we evaluated the impact of

nitrogen starvation, a strategy commonly used in Y. lipolytica to induce lipid formation

[137] . Under these conditions, the impact was varied across the various rewiring schemes

(Figure 3-9) with the most significant improvement observed in the PEX10 overexpression

strain (Figure 3-10). When grown in C20N2 media, this strain achieved a greater than 2-

fold increase over the YT strain under normal conditions (reaching 4.1 g/L in a test tube,

43.4% theoretical yield). Further nitrogen limitation did not improve TAL production as a

result of the expense to growth.

Figure 3-9: The effect of fatty acid synthesis inhibitor cerulenin on TAL production.

A 4.5 4

3.5

) C20N5 Cerulenin

( 3 C20N5 No Spike

i C20N2.5 Cerulenin

Overall Avg fold 1.135676 T 2.5

L C20N2.5 No Spike

2 e C20N2 Cerulenin

a r 1.5 e C20N2 No Spike

A 1 C20N1 Cerulenin C20N1 No Spike 0.5

0 YT YT- ACS1, ALD3, YT- PDA1, PDE2, YT- ACL1, ACL2, YT- ACS1, ALD5, PDC1, ACC1 PDE3, PDB1, ACC1 AMPD, ACC1 PDC2, ACC1 B 6

TAL 5production was assessed following 4-day tube fermentation in defined media. Error

)

barsg represent the standard deviation of n=3. The effect of cerulenin supplementation and

(

r 4

nitrogeni limitation on TAL production. Fermentations with cerulenin produced

L * C20N5

3 *

significantly more TAL than those without for an average increase of 12.8% (p=0.003).

* *

e *

* C20N2.5

* *

* * * *

* *

e * * C20N2

A C20N1 1 46 0 YT YT- ACS1, YT- ACS1, YT YT- ACS1, YT- ACS1, YT YT- ACS1, YT- ACS1, ALD3, PDC1, ALD5, PDC2, ALD3, PDC1, ALD5, PDC2, ALD3, PDC1, ALD5, PDC2, ACC1 ACC1 ACC1 ACC1 ACC1 ACC1 No Spike Acetate Glucose Figure 3-10: Impact of nitrogen starvation on TAL production.

TAL production under different conditions was assessed following tube fermentation in defined media. Error bars represent the standard deviation of n=3. Statistical significance was determined by a Dunnett’s test, each new condition was compared to the relevant control (C20N5 in the case of nitrogen limitation), * p< 0.05, ** p< 0.01, *** p<0.001. Nitrogen limitation enhances the effect of gene overexpressions related to β-oxidation.

Second, a series of supplements (see Supplementary Data 3.4.6) and different carbon sources were tested as spikes or sole carbon sources along with glucose as a control

(Figure 3-11). The largest gains in TAL titer were realized through providing an acetate spike when the pyruvate bypass pathway was overexpressed. Under these conditions, the top strain from the pyruvate bypass pathway produced 4.9 g/L TAL in a test tube

(representing over 35% of the theoretical conversion yield) (Figure 3-12). Analysis of acetate consumption indicates this result is not simply due to acetate acting as a carbon source for TAL production, because a substantial portion of the acetate fed remains at the end of the fermentation (Figure 3-13). This result suggests a more regulatory or redox related impact on metabolism. Further to this point, in Saccharomyces cerevisiae, acetate feeding has been shown to induce changes to metabolism mediated through protein acetylation [146, 147], a possible mechanism to be explored here. Additionally, 47 acetylaldehyde dehydrogenase activity assays suggest pathway engineering (especially with ALD5) resulted in altered redox cofactor usage favoring NAD+ over NADP+ (Figure

3-14). Thus, an overall redox and regulatory mechanism may explain the almost 2-fold increase in TAL production observed in this study. Intriguingly, this improvement was not seen under nitrogen-limited conditions (Figure 3-15), suggesting that these two strategies

(acetate feed and nitrogen starvation) are not compatible.

48 Figure 3-11: The effect of vitamin supplementation and alternative carbon sources on TAL production.

A 3.5

)

g 2.5

(

t *

T 2 Control

T *** * Biotin e 1.5

g **

a B5

A 1 TPP

0.5

0 YT YT- ACS1, ALD3, YT- ACS1, ALD5, YT- PDA1, PDE2, YT- ACL1, ACL2, PDC1, ACC1 PDC2, ACC1 PDE3, PDB1, ACC1 AMPD, ACC1

B 3 C 2.5

2.5 **

) )

L L

/ /

g g

( ( 2 *

r r

e ** e

t t

i i 1.5

T T

L 1.5 L

A A

T T

Control e *** Control e 1

g g

a 1 a Citrate r Pyruvate r

e e

v v A A 0.5 0.5

0 0 YT YT- ACS1, YT- ACS1, YT- PDA1, YT YT- ACL1, ALD3, ALD5, PDE2, ACL2, AMPD, PDC1, PDC2, PDE3, ACC1 ACC1 ACC1 PDB1, ACC1

TAL production under different media conditions was assessed following 4-day tube fermentations in defined media. Error bars represent the standard deviation of n=3, * p< 0.05, ** p< 0.01, *** p<0.001. A) The effect of vitamin supplementation. Significance was tested using Dunnett’s test to compare each condition to the corresponding control. B) Pyruvate was tested as an alternative carbon source using the pyruvate bypass pathway and PDH complex overexpression strains. Significance was tested using a two- sample T test. C) Citrate was tested as an alternative carbon source for the citrate pathway overexpression strain. Significance was tested using a two-sample T test.

49 Figure 3-12: Impact of pyruvate bypass pathway intermediates spike on TAL production.

TAL production under different conditions was assessed following tube fermentation in defined media. Error bars represent the standard deviation of n=3. Statistical significance was determined by a Dunnett’s test, each new condition was compared to the relevant control (glucose in the case of feeding spikes), * p< 0.05, ** p< 0.01, *** p<0.001. Feeding assay demonstrating the effect of adding 10 g/L carbon molar equivalent of glucose as a feeding spike 24 hours into standard tube fermentation.

Figure 3-13: Acetate spike consumption.

)

(

t 15

A 10

*** 5

0 YT YT-ACS1, YT YT-ACS1, YT YT-ACS1, ALD5, ALD5, ALD5, PDC2, PDC2, PDC2, ACC1 ACC1 ACC1 no spike acetate 1X (13.7g/L) acetate 2X (27.4 g/L) The acetate remaining following 4-day tube fermentations in defined media was assessed. Error bars represent the standard deviation of n≥2, *** p<0.001. Significance was tested using a two-sample T test to compare the engineered strain to YT in each condition. 50 Figure 3-14: Relative acetylaldehyde dehydrogenase activity in NAD+ versus NADP+.

100

D 10

A 4.5 0.14 unmodified ALD1 ALD2 ALD3 ALD5 ALD6 3.5

) C20N5 Cerulenin

The/ relative acetylaldehyde dehydrogenase activity of lysates of pyruvate bypass

( 3 C20N5 No Spike

pathwaye overexpression strains was assessed in the presence of both NAD+ and NADP+

i C20N2.5 Cerulenin

Overall Avg fold 1.135676 T 2.5

usingL a diaphorase-resazurin system. Activity ratios are organized by the ALDC20 Nhomolog2.5 No Spike

2 e C20N2 Cerulenin

overexpressed,g protein lysates were prepared in biological triplicate. The population of

a r 1.5 e C20N2 No Spike strainsv containing ALD5 overexpression was found to be significantly different than the

A unmodified1 group using Dunnett’s test (n=5, p=7.26x10-5). C20N1 Cerulenin C20N1 No Spike 0.5

0 Figure 3-15:YT The effectYT- ACofS1 nitrogen, ALD3, YT limitation- PDA1, PDE2, onYT TAL- ACL1, ACproduction.L2, YT- ACS1 , ALD5, PDC1, ACC1 PDE3, PDB1, ACC1 AMPD, ACC1 PDC2, ACC1 B 6

)

(

r 4

L * C20N5

3 *

* *

e *

* C20N2.5

* *

* * * *

* *

e * * C20N2

A C20N1 1

0 YT YT- ACS1, YT- ACS1, YT YT- ACS1, YT- ACS1, YT YT- ACS1, YT- ACS1, ALD3, PDC1, ALD5, PDC2, ALD3, PDC1, ALD5, PDC2, ALD3, PDC1, ALD5, PDC2, ACC1 ACC1 ACC1 ACC1 ACC1 ACC1 No Spike Acetate Glucose TAL production was assessed following 4-day tube fermentation in defined media. Error bars represent the standard deviation of n=3. The effect of nitrogen limitation in combination with feeding spikes. Significance was tested using Dunnett’s test to compare each condition to the corresponding control, * p< 0.05, ** p< 0.01, *** p<0.001. 51 3.3.7 Bioreactor cultivation boosts overall TAL titer

Fermentation of the pyruvate bypass overexpression strain (YT- ACS1, ALD5,

PDC2, ACC1) with acetate spiking was scaled up to the 3 L bioreactor scale. After optimization, fermentation resulted in production of 35.9 ± 3.9 g/L of TAL (Figure 3-

16A). Although a longer period was necessary to achieve this high titer, this fermentation time is comparable to that of previously published work for TAL [127]. Moreover, we achieve a substantially improved productivity over the batch cultures (upwards of 4-fold increase to 0.21 ± 0.03 g/L/h (Table 3-1)) and achieve values that exceed these previous reports in S. cerevisiae by up to 10-fold. As this production level is well outside the soluble range of TAL, substantial in situ precipitation occurred, an attractive feature for industrial production, but a unique source of sampling error. Likewise, we observed a diauxic shift from glucose to (produced) citrate utilization (Figure 3-16A) that leads to an increased production as cell viability stagnates and even decreases throughout the process (Figure 3-

16B).

52 Figure 3-16: Bioreactor cultivation of pyruvate bypass overexpression strain.

YT-ACS1, ALD5, PDC2, ACC1 was fermented in a 3 L bioreactor with YP media, 180 g/L glucose and a 13.7 g/L sodium acetate spike at 36 hours. This figure demonstrates a representative run with a duplicate presented in Figure 3-17. Concentrations of TAL, citrate and glucose were determined from three independent samples taken at each time point, error bars represent standard deviation of n=3. Viable cell count was determined by plating different dilutions of sample, error bars represent standard deviation of n≥2.

53 Figure 3-17: Bioreactor cultivation of pyruvate bypass overexpression strain.

YT-ACS1, ALD5, PDC2, ACC1 was fermented in a 3 L bioreactor with YP media, 180 g/L glucose and a 13.7 g/L sodium acetate spike at 36 hours. Concentrations of TAL, citrate and glucose were determined from three independent samples taken at each time point, error bars represent standard deviation of n=3. Viable cell count was determined by plating different dilutions of sample, error bars represent standard deviation of n≥2.

54 Table 3-1: Summary of titers, yields, and productivities achieved in this work.

Media Titer Yield Productivity Strain Figure Conditions (g/L) (g/g) (g/L/h) YT C20N5 2.1 0.105 0.022 Figure 3-2 YT-PDA1, PDE2, PDE3, PDB1, C20N5 2.5 0.125 0.026 Figure 3-5 ACC1 YT-ACS1, ALD5, C20N5 2.8 0.14 0.029 Figure 3-6 PDC2, ACC1 YT-PEX10 C20N5 2.4 0.12 0.025 Figure 3-7 YT-PEX10 C20N2 4.1 0.205 0.043 Figure 3-10 YT-ACS1, ALD5, C20N5, C10 4.9 0.245 0.051 Figure 3-12 PDC2, ACC1 acetate spike Overallb: C180YP, 0.12 YT-ACS1, ALD5, Figures 3-16 C10 acetate 35.9 N/Aa Glucose PDC2, ACC1 and 3-17 spike Phasec: 0.21 a Yield is not calculated in the bioreactor due to the presence of complex media components b Overall productivity is calculated as the total TAL produced divided by total fermentation time c Glucose phase productivity is calculated as the TAL produced during the peak glucose fermentation phase after biomass accumulation (typically 50 – 150 hours of the fermentation)

3.3.8 Biosourced TAL can be incorporated into a polymer through O- functionalization4

Finally, we leveraged the newfound bulk-availability of polyketides such as TAL to demonstrate their utility in the modification of polymer properties in collaboration with the Lynd lab at UT-Austin. TAL serves a dual role as both a polymer modifier but also a

4 Materials experimental work and text was contributed by Malgorzata Chwatko from Dr. Nathaniel A. Lynd’s lab at UT-Austin. 55 functional adduct for later chemical derivatization owing to its lactone and unsaturated functionalities [148]. TAL was rapidly extracted and purified from fermentation broth and ultimately used for the modification of structure and properties of commodity poly(epichlorohydrin). This was achieved through heat and an activating organic base 1,8-

Diazabicyclo(5.4.0)undec-7-ene (DBU) (Figure 3-18A) [148, 149]. In this process, the displacement of chloride through the O-functionalization of TAL was evident spectroscopically, chromatographically, and thermally through the use of NMR spectroscopy (Figure 3-18B and 3-19), size exclusion chromatography (SEC), and differential scanning calorimetry (DSC) (Table 3-2), respectively. The amount of TAL incorporated along the poly(epichlorohydrin) backbone was tuned stoichiometrically from

50% to 83% by mole resulting in poly[(epichlorohydrin)-co-(epoxy triacetic acid lactone)] or PETAL. We measured the compositionally-dependent changes in the glassy amorphous solid based on molecular weight and glass transition temperature of PETAL. Molecular weight was seen to increase proportionately to the amount of TAL incorporated as measured by size exclusion chromatography with multi-angle light scattering detection to yield absolute number-average molecular weights (Mn) (Table 3-2). While poly(epichlorohydrin) exhibits a native glass transition temperature (Tg) of –30ºC, this value was markedly increased and strongly dependent on TAL incorporation with the highest TAL composition (83% by mole) exhibiting a Tg of 70ºC. The final product can be formed into a film and is seen to exhibit a unique hue and relative transparency (Figure 3-

18C). In particular, the orange color of the polymer is due to the native color of covalently bound TAL, as no other characteristic spectroscopic differences were observed via UV- 56 Vis spectroscopy (Figure 3-20). It should be noted that the reaction rate obtained with this purified, biosourced TAL was comparable to a reaction conducted with commercially sourced TAL. Finally, the residual lactone and unsaturated functionality from TAL repeat units offer a unique future strategy toward further modification of material properties.

Figure 3-18: Production of poly[(epichlorohydrin)-co-(epoxy triacetic acid lactone)] or PETAL.

A) Reaction scheme to create PETAL. B) H NMR characterization of PETAL which shows distinct new peaks and shifts from the start molecules. C) Photo of PETAL pressed into a film.

57 Figure 3-19: Further NMR characterization of poly(epichlorohydrin-co-epoxytriacteic lactone).

A B

A) COSY (HH) NMR characterization of PECH-c-eTAL shows correlations of the TAL protons and backbone of polymer. B) HSQCAD (CH) NMR characterization of PECH-c- eTAL shows correlations protons to carbon on the molecule.

Table 3-2: Characteristics of copolymers in comparison to PECH.

a Monomer Feed Polymer c d Monomer b Mn Tg Pairs ECH: TAL : Composition DBU ECH:TAL (g/mol) (°C) PECH 1: 0 17,300 -30 P(ECH-TAL) 1: 1.5: 0.75 1: 1 27, 200 30 PECH 1:0 7,000 -30 P(ECH-TAL) 1: 1.5: 1 1: 5 9,400 70

Polymer characteristics were measured for PETAL. aDetermined by gravimetry. bDetermined by 1H NMR spectroscopy. cNumber-average molecular weight determined by size exclusion chromatography in chloroform using light scattering and differential refractometer detectors. dThermal properties determined by differential scanning calorimetry.

58 Figure 3-20: Ultraviolet–visible spectroscopy of PETAL and TAL.

Characterization of PETAL and TAL via UV-Vis spectrophotometer reveals no significant changes between the two compounds.

3.4 SUPPLEMENTAL DATA

3.4.1 Toxicity of TAL in Y. lipolytica

Previous work has indicated that Y. lipolytica is highly resistant to many different conditions including bioactive molecules similar to TAL [137, 138]. In order to determine the resistance of Y. lipolytica to TAL specifically, the wildtype strain, PO1f [61] was grown in media containing TAL concentrations varying from 2 g/L to 4.5 g/L. TAL slowed the growth rate of PO1f, as rates varied from 0.26 to 0.07 hours-1 (Figure 3-1). Although growth rate was inhibited, no TAL concentration was lethal, and cells were able to reach a similar final cell density under all conditions. This small decrease in growth is not surprising based on past work in fungal hosts [126] and is vastly superior to the high toxicity observed in E. coli [125].

59 3.4.2 Testing of previously identified g2ps1 mutants

Previous work in E. coli identified g2ps1 active site mutations that improved TAL production [125]. These mutants were constructed and tested in Y. lipolytica where a significant increase in TAL titer was not observed (Figure 3-2). In fact, the L261F mutant and L202G L261N double mutant produced significantly less TAL when compared to the wild type. Following this test, all further production strains were constructed using wildtype g2ps1.

3.4.3 Identification of uncharacterized overexpression targets

Overexpression gene targets were identified by KEGG enzyme number through searching the KEGG Yarrowia lipolytica genome T01033 (Table 3-3). The five potential

ALD genes identified were annotated as aldehyde dehydrogenases and were all tested in combination with the other pathway genes in order to determine which genes could act as acetylaldehyde dehydrogenases specifically. Updates to the iGenoLevures database identified ALD1 as a HFD3 fatty aldehyde dehydrogenase. ALD2 and ALD3 are putative mitochondrial aldehyde dehydrogenases. Finally, ALD5 and ALD6 are putative cytoplasmic aldehyde dehydrogenases. Note the arbitrary assignment of aldehyde dehydrogenase gene numbers (ALD1-6) used here do not match those used in the iGenoLevures annotation.

60 Table 3-3: Overexpression gene targets. The gene abbreviations used in this paper are mapped to the systematic gene identifier and KEGG enzyme number.

Abbreviation Gene Identifier KEGG EC Selective Agent g2ps1 N/A N/A CSM- LEU, -URA ACC1 YALI0C11407 6.4.1.2 nourseothricin sulfate PDC1 YALI0D10131 4.1.1.1 mycophenolic acid PDC2 YALI0D06930 4.1.1.1 mycophenolic acid ACS1 YALI0F05962 6.2.1.1 CSM-TRP AMPD YALI0E11495 3.5.4.6 CSM-TRP ALD1 YALI0B01298 1.2.1.3 hygromycin B ALD2 YALI0C03025 1.2.1.3 hygromycin B ALD3 YALI0E00264 1.2.1.3 hygromycin B ALD4 YALI0F23793 1.2.1.3 hygromycin B ALD5 YALI0D07942 1.2.1.5 hygromycin B ALD6 YALI0F04444 1.2.1.5 hygromycin B PDA1 YALI0F20702 1.2.4.1 CSM-TRP PDB1 YALI0E27005 1.2.4.1 chlorimuron ethyl PDE2 YALI0D23683 2.3.1.12 hygromycin B PDE3 YALI0D20768 1.8.1.4 mycophenolic acid ACL1 YALI0E34793 2.3.3.8 hygromycin B ACL2 YALI0D24431 2.3.3.8 CSM-TRP PEX10 YALI0C01023 N/A CSM-TRP POR1 YALI0D12628 N/A CSM-TRP

61 3.4.4 Copy number assay

Genomic copy number was determined in the engineered strains with the highest TAL yields- YT-PEX10 and YT-ACS1, ALD5, PDC2, ACC1. Copy number of these overexpressed genes can be found in found in Table 3-4.

Table 3-4: Copy number of overexpressed genes in the top TAL producing strains built in this study.

Strain YT-PEX10 Estimated Copy Standard Deviation Gene Number from Assay g2ps1 4 1.0 PEX10 2 0.4

Strain YT- ACS1, ALD5, PDC2, ACC1 Estimated Copy Standard Deviation Gene Number from Assay g2ps1 5 0.6 ACS1 2 0.2 ALD5 2 0.2 PDC2 5 0.5 ACC1 2 0.2

3.4.5 Acetylaldehyde dehydrogenase activity assay

The overall acetylaldehyde dehydrogenase activity of pyruvate bypass overexpression strains was assessed in the presence of NAD+ and NADP+. Strains containing overexpressed ALD5 were found to have a significantly increased ratio of lysate activity in the presence of NAD+ versus NADP+, suggesting that the overall acetylaldehyde dehydrogenase activity in these cells favors NAD+ over NADP+ (Figure 3-14).

62 3.4.6 Further media optimization

Cofactor limitation was tested by supplementing fermentations of top strains from each of the three pathway overexpression strategies with cofactors necessary for the upregulated enzymes in those strains. Biotin is necessary for the carboxylation activity performed by ACC1 [137, 139], overexpressed in conjugation with each acetyl-CoA pathway tested. Thiamine pyrophosphate, a decarboxylation cofactor, is essential for the activity of both pyruvate decarboxylases as well as the PDH complex [150]. In addition to these cofactors, we tested the limitation of CoA availability through supplementation of pantothenic acid (vitamin B5) necessary for CoA synthesis [7]. In each genetic background tested, the supplementations did not significantly increase TAL titers, with some supplements leading to significant decreases in TAL titer (Figure 3-11A). These results suggest that the availability of these nutrients is not limiting TAL production.

In addition to testing cofactors, inhibition of fatty acid synthesis through dosing with a small molecule was tested. The addition of cerulenin [145] produced a significant increase with an average of 12.8% increase in titer across all conditions tested (Figure 3-

9). This feeding (while too expensive and toxic for large scale) suggests further gains in production capacity by decreasing lipid synthesis.

In addition to testing supplementations, different sole carbon sources were tested and compared: glucose, pyruvate, and citrate. Pyruvate was fed to strains with the PDH complex or pyruvate bypass pathways overexpressed, as pyruvate is where acetyl-CoA production diverges from central metabolism for these pathways. When pyruvate was supplied as a sole carbon source, TAL production was hindered in all strains as a result of 63 decreased cell growth (Figure 3-11B). Citrate was fed as the sole carbon source to YT and

YT-ACL1, ACL2, AMPD, ACC1. As expected, citrate had no effect on the production of

TAL in YT, but improved TAL production 8% when the citrate pathway was overexpressed (Figure 3-11C).

Following the discovery that an acetate spike 24 hours into fermentation stimulates

TAL production, acetate consumption was assessed (Figure 3-13). Following fermentation

YT-ACS1, ALD5, PDC2, ACC1 had significantly less acetate remaining than YT, suggesting about half of the acetate fed had been consumed. However when an additional

48-hour acetate spike was added the difference between the two strains was no longer significant suggesting additional acetate is not helpful. We therefore hypothesize that the ability of acetate to promote TAL production is not simply a carbon catabolism effect.

3.5 CONCLUSIONS

In summary, this work demonstrates the first use of an oleaginous organism for high-level production of an acetyl-CoA and malonyl-CoA derived polyketide. Moreover, we establish a previously uncharacterized pyruvate bypass pathway as superior for rewiring

CoA flux from lipid biosynthesis and into high-level TAL production reaching a titer of

35.9 ± 3.85 g/L and a yield of 0.164 g/g. Additional engineering efforts related to the β- oxidation pathway increased yields to 0.203 g/g. These values far exceed previous efforts in the field with conventional organisms (a summary of the achieved titers and yields in this work is provided in Table 3-1). This high-level production enabled a rapid purification and conversion into a unique polymer with favorable molecular weight and glass transition

64 temperature. This work and resulting strain provides a path forward for microbial production of other acetyl-CoA and malonyl-CoA derived polyketides for novel applications such as polymers and chemical conversion.

65 Chapter 4: High efficiency transformation of Yarrowia lipolytica using electroporation5

4.1 CHAPTER SUMMARY

Yarrowia lipolytica is an industrial host organism with incredible potential for metabolic engineering. However, the genetic tools and capacities in this host lag behind those of conventional counterparts. In this study, we sought to increase the transformation efficiency of Y. lipolytica by creating a simple protocol using electroporation. Efficiency was increase by optimizing wash buffers, pre-culture growth time, and OD600 of competent cells, voltage, competent cell volume, and recovery time. The outcome of these optimizations led to a simple protocol with maximum linear fragment transformation efficiency of 7.8 x 103 transformants per μg DNA and 2.8 x 104 transformants per μg DNA for episomal plasmid transformation. The protocol presented here is superior to other Y. lipolytica transformation protocols as it requires no lengthy pretreatment and no required carrier DNA to achieve efficiencies on par with, or exceeding, previously reported methods.

5 Sofia Vazquez and Cecilia Barnhill helped with conducting initial transformation experiments. Special thanks to Clare Murray and Arvind Swaminathan for help counting plates. 66 4.2 INTRODUCTION

In recent years, Yarrowia lipolytica has become the most well studied oleaginous yeast and has incredible potential as an industrial production host [112]. To this end, metabolic engineering strategies in this host have successfully increased the production of native metabolites such as citric acid [151, 152] and lipids [25, 28, 113] and nonnative chemicals such as vitamins [42, 45] and polyketides [55]. These successes notwithstanding, the genetic tools available in this host lag behind that of traditional, conventional hosts like

Escherichia coli and Saccharomyces cerevisiae. One particular genetic challenge in this host is its highly active non-homologous end joining (NHEJ) DNA repair mechanism [45].

This NHEJ activity means that exogenous DNA present in a transformation (including carrier DNA) is often spontaneously and randomly integrated into the genome. Just as importantly, Y. lipolytica suffers from a lack of efficient and facile transformation techniques compared to traditional host organisms. In this regard, transformation efficiency improvements will enable an increased capacity for library based screening and protein engineering in this host, an area underdeveloped in Y. lipolytica. Moreover, improved transformation removes an important bottleneck in Y. lipolytica and can potentially inform genetic engineering strategies for other nonconventional and oleaginous yeasts.

Many approaches to improve genetic transformation efficiency in Y. lipolytica have been explored over the past 30 years (Table 4-1), with a heavy reliance on chemical transformation. One of the more recent, optimized protocols achieves up to 2.2 x 103 transformants per μg DNA, but requires large quantities of carrier DNA [58]. This dependence on carrier DNA is similar to previously reported chemical transformation 67 methods in this host [12, 153, 154]. Specifically, in the absence of carrier DNA, chemical transformation of Y. lipolytica is barely effective with reported efficiencies of 16 transformants per μg DNA [12] and 20 to 200 transformants per μg DNA [155]. While carrier DNA is a relatively benign participant in S. cerevisiae transformations, its presence is not desirable in a host where the NHEJ DNA repair is prominent and highly active. In particular, carrier DNA can be incorporated into the host DNA alongside cassettes of interest [156]. For previously reported chemical transformation methods in Y. lipolytica, this carrier DNA is present at amounts up to 5,000-fold over the DNA of interest [58]. As a result, this additional exogenous DNA can inadvertently integrate into and impact the genotype and phenotype of transformed strains. An alternative to this chemical transformation is electroporation. Transformation of Y. lipolytica using electroporation has been previously reported with efficiencies up to 2.1 x 104 transformants per μg DNA, but requires the use of cells pretreated in lithium acetate (a potential mutagen) and many temperature-controlled wash steps [157] (Table 4-1).

68 Table 4-1: Comparative analysis of transformation methods developed and optimized for Y. lipolytica.

Max. Competent Carrier DNA Competent Type of Efficiency Cell Reagents (per Cell Ref. Transformation (transformants Incubation Required transformation) Washes per µg DNA) Time

250 µg single Chemical 2.2E+03 strand salmon 1 hr 4 1 [58] sperm DNA

50 µg sonicated Chemical 1.0E+04 2 hr 45 min 5 1 [12] E. coli DNA

25 µg single 1 hr, plus Chemical 6.0E+05 strand salmon >1 hr to dry 4 None [153] sperm DNA plate

25 µg single Chemical* 2.0E+02 strand salmon 2 hr 25 min 4 3 [155] sperm DNA

Electroporation 2.1E+04 None 1 hr 3 4 [157]

This Electroporation 2.8E+04 None None 2 1 study

*Reported use of transformation method from Barth and Gaillardin [154]

69 To bypass the limitations in transformation efficiency in this host and drawbacks of previous approaches, we sought to develop a simplified method that provides high transformation efficiencies for Y. lipolytica genetic engineering endeavors. In this study, we establish an improved protocol by investigating parameters in each of the three major stages of transformation: (i) pre-culture conditions and competent cell preparation, (ii) transformation conditions, and (iii) post-transformation conditions and recovery. In particular, for pre-culture conditions, we investigated the impact of pre-culture fermentation time, cultivation vessel, cell concentration, and washing schemes. For transformation conditions, we investigated the impact of electroporation waveform and voltage as well as competent cell volume. Finally, for post-transformation, we evaluated the impact of recovery time. Collectively, these optimizations led to an optimized and simplified protocol suitable for linear and plasmid DNA that successfully surpasses previously published maximum transformation efficiencies without the use of carrier

DNA.

4.3 RESULTS AND DISCUSSION

To develop a streamlined and efficient electroporation protocol for Y. lipolytica that bypasses the limitations of existing transformation approaches, we investigated parameters in each of the three major stages of transformation: (i) pre-culture conditions and competent cell preparation, (ii) transformation conditions, and (iii) post-transformation conditions and recovery. Each of these experiments was conducted using the

70 transformation of linear DNA fragments into the wildtype lab strain PO1f (ATCC MYA-

2613).

4.3.1 Alterations to pre-culture conditions and competent cell preparation can improve overall transformation efficiency

The typical approach to competent cell preparation involves culturing yeast and then washing in a solution to obtain a specified cell density. To optimize this procedure for

Y. lipolytica transformation, a variety of different rich media based pre-culture conditions and wash conditions were evaluated. Efficiencies were obtained by counting the number of resulting transformation events using a linearized DNA fragment containing an auxotrophic marker enabling leucine biosynthesis paired with a strong hybrid promoter driving the expression of hrGFP as described in the Materials section above.

Potential wash strategies were guided by literature review. Previously developed electroporation methods for fungal species use 1 M sorbitol as the wash reagent in competent cell preparation [158, 159]. Likewise, it is commonplace to use 10% glycerol when preparing competent cells—especially for bacterial hosts [160, 161]. For this study, we evaluated these two wash conditions in conjunction with evaluating the pre-culture scale (i.e. test tube or shaker flask) (Figure 4-1). It is immediately clear that 10 % glycerol serves as an inadequate wash solution for preparation of Y. lipolytica competent cells compared with 1 M sorbitol (102 vs. 3.9 x 103 transformants per μg DNA). Though the transformation efficiency of cells grown in a tube slightly surpasses that of cells grown in a flask (3.9 x 103 vs. 3.2 x 103 transformants per μg DNA with 1 M sorbitol wash), the

71 advantage of flask culture and larger scale transformations warrants its use for most applications. As a result, all further experiments were conducted with pre-cultures grown at the flask scale.

Figure 4-1: Transformation efficiency of Y. lipolytica integrations with variable wash strategies and pre-culture scale.

Pre-cultures were grown overnight in a tube or flask and washed with either 1 M sorbitol or 10% glycerol. Competent cells were prepared at an OD600 of 30, and a 400 µL volume was electroporated with 5 µg DNA at 2700 V with a 1 hr recovery. Transformation efficiency was measured by colony count. Error bars show standard deviation of n = 2 transformations.

72 Multiple washes are typically employed during electrocompetent cell preparation, especially for E. coli [160]. To evaluate the impact of multiple washes, we evaluated this strategy using pre-cultures that were grown in flasks for varying amounts of time (Figure

4-2). Three sets of pre-cultures were grown consisting of: (i) 21 hours, (ii) 18 hours, or (iii)

21 hours with a 3 hour outgrowth after back diluting to 20% of the starting cell concentration. For each of these cases, a second wash with sorbitol did not improve transformation efficiency. From a practical standpoint, a single wash is much more desirable as it simplifies the transformation protocol and saves time. Moreover, these results demonstrated the efficacy of room temperature reagents for the wash, an improvement over a previously reported electroporation method for Y. lipolytica [157]. The only condition of pre-culture growth that seemed to increase efficiency was the back dilution after long overnight pre-culture trial (improvement of 1.5 x 103 transformants per

μg DNA from 1.1 x 103 transformants per μg DNA). While slightly beneficial, this step does involve more time, media, and materials, thus can be bypassed except when maximum transformation efficiencies are necessary.

73 Figure 4-2: Transformation efficiency of Y. lipolytica integrations from variable length pre-cultures with one and two washes.

Cells were pre-cultured for 21 hrs, 18 hrs, or 21 hrs with an extra 3 hr back dilution and washed either once or twice with 1 M sorbitol. Competent cells were prepared at an OD600 of 30, and a 400 µL volume was electroporated with 5 µg DNA at 2700 V with a 1 hr recovery. Transformation efficiency was measured by colony count. Error bars show standard deviation of n = 2 transformations.

As a final parameter tested in competent cell preparation, we investigated the impact of diluting the cells prior to transformation. To do so, different competent cell densities were prepared by testing OD600 and diluting in 1 M sorbitol to OD600 levels ranging from 5 to 40, in step changes of 5 OD600 units (Figure 4-3). As a general trend, transformation efficiency increased as a function of cell concentration (with the exception of OD600 of 40). However, it should be noted that electroporation at higher cell densities have an increased chance of arcing. Based on these experiments, an OD600 of 30 (which corresponds to cell counts of 1.2 x 1011 ± 7.0 x 109 cells mL-1 as measured by dilution

74 plating on rich media) was observed to have the highest transformation efficiency while still maintaining the ability to complete a pulse without arcing. It should be noted that the data in this particular experiment is slightly lower than the values in the prior experiments owing to additional incubation time in the 1 M sorbitol prior to electroporation (as a result of setting up all the conditions simultaneously).

Figure 4-3: Transformation efficiency of Y. lipolytica integrations at various competent cell concentrations.

Competent cells were prepared from overnight cultures and diluted to various cell densities. A 400 µL volume of cells was electroporated with 5 µg DNA at 2700 V with a 1 hr recovery. Transformation efficiency was measured by colony count. Error bars show standard deviation of n = 2 transformations.

Collectively, these results demonstrate that the best Y. lipolytica competent cell preparation steps include cells cultured overnight in rich medium washed one time with 1

75 M sorbitol and diluted to an OD600 of 30. Small volumes of highly competent cells can be prepared from tube cultures and larger volumes of competent cells can be prepared from flasks with an extra boost in efficiency occurring after back-dilution and an additional three hour growth phase.

4.3.2 Transformation conditions can be modified to further improve efficiency of linear integrations

Following optimization of the competent cell preparation protocol, we next evaluated various parameters associated with the electroporation condition itself including the waveform, voltage, and competent cell volume. First, the use of both exponential and square waveforms were tested. Square wave electroporation did not successfully transform

Y. lipolytica, thus we proceeded to optimize the exponential waveform voltage. To do so, we evaluated voltages spanning 1500 and 2900 V (Figure 4-4). Transformation efficiency increased as a function of voltage, up to a maximum of 1.4 x 103 transformants per μg DNA occurring at 2700 V. At tested voltages above this level, the electroporation arced and transformation efficiency subsequently dropped. Therefore, 2700 V was selected as the best parameter in this protocol.

76 Figure 4-4: Transformation efficiency of Y. lipolytica integrations at varying electroporation voltages.

A variety of different electroporation voltages were tested. Competent cells were prepared from overnight cultures at an OD600 of 30, and a 400 µL volume was electroporated with 5 µg DNA with a 1 hr recovery. Transformation efficiency was measured by colony count. Error bars show standard deviation of n = 2 transformations.

Next, we evaluated the impact of competent cell volumes loaded into a 2 mm gap electrocuvette between 50 μL to 400 μL (Figure 4-5). In general, efficiencies were all within error of one another and had a maximum with 200 μL of competent cells which resulted in 7.8 x 103 transformants per μg DNA. Taken together, maximum transformation efficiency was observed when 200 μL of competent cells was used for electroporation at

2700 V.

77 Figure 4-5: Transformation efficiency of Y. lipolytica integrations with varying competent cell volumes.

Competent cells were prepared from overnight cultures at an OD600 of 30, and different volumes were aliquoted into electrocuvettes. Each competent cell volume was electroporated with 5 µg DNA with a 1 hr recovery. Transformation efficiency was measured by colony count. Error bars show standard deviation of n = 2 transformations.

78 4.3.3 Longer post-transformation recovery improves net transformation efficiency

As a final stage in the transformation optimization, we evaluated the impact of recovery time in rich media following the electroporation step. Specifically, recovery times ranging from zero to eight hours were tested (Figure 4-6). In addition to measuring transformation efficiency (through plating on selective media), an assessment of average total viable cells was conducted to see whether outgrowth was resulting in cell growth and thus division of previously transformed cells. In this experiment, we observed a linear increase in transformation efficiency as a function of recovery time. In all prior experiments above, a one hour recovery time was used, however, it was found that among the times tested, eight hours was superior. Specifically, the extended recovery time yielded

3.8 x 103 transformants per μg DNA compared with the 1.9 x 103 transformants per μg

DNA obtained in a 1 hour recovery. Throughout the recovery stage, average viable cells did not increase significantly suggesting that this increase in transformants was due to new integration events in the genome of transformed DNA rather than cell outgrowth and division of initially transformed cells. This result demonstrates the importance of optimizing recovery time during transformation.

79 Figure 4-6: Transformation efficiency of Y. lipolytica integrations with increased recovery time.

Competent cells were prepared from overnight cultures at an OD600 of 30, and a 400 µL volume was electroporated with 5 µg DNA. Cells were recovered in rich media in a rotary drum for varying amounts of time. Transformation efficiency was measured by colony count (black dot) with viable cells also measured (gray triangle). Error bars show standard deviation of n = 2 transformations.

80 4.3.4 Transformation efficiency of plasmid DNA is higher than linear transformation and integration

All experiments described above were tested with linear fragments of DNA to obtain stable genome integrations. Nevertheless, episomal plasmids are useful for testing synthetic biology tools and expressing genes that are undesirable for integration like

CRISPR-Cas9 [72] and Cre recombinase [14]. To initially evaluate the efficiency of electroporation-based transformation of plasmids, we used an episomal plasmid with similar structure to the linear DNA (Figure 4-7) and obtained 2.8 x 104 transformants per

μg plasmid DNA. This efficiency was obtained when using 400 μL of competent cells at an OD600 of 30 from a 21 hour flask pre-culture electroporated at 2700 V and recovered for

1 hour. This efficiency value is the highest reported for an electroporation method of Y. lipolytica (Table 4-1). It is possible that further improvements can be made by combining additional factors including competent cell volume and longer recovery times.

Nevertheless, this electroporation condition is suitable for both plasmid and linear fragments.

81 Figure 4-7: Transformation efficiency of Y. lipolytica integration compared to transformation efficiency of an episomal plasmid.

Constructs carrying similar cargo either on a linear fragment for integration or episomal plasmid were tested for transformation efficiency under standard conditions. Competent cells were prepared from overnight cultures at an OD600 of 30, and each competent cell volume was electroporated with 5 µg DNA with a 1 hr recovery. Transformation efficiency was measured by colony count. Error bars show standard deviation of n = 2 transformations.

4.4 CONCLUDING REMARKS

Most current transformation methods for Y. lipolytica involve the complicated preparation of competent cells or the use of large quantities of carrier DNA. In this work, we optimized a simple protocol for efficient electroporation transformation of Y. lipolytica that involves only a single wash with a common reagent at room temperature. This protocol simplifies day to day metabolic engineering efforts in this nonconventional yeast while

82 providing transformation efficiencies up to 7.8 x 103 transformants per μg DNA for linear integration and 2.8 x 104 transformants per μg DNA for episomal plasmid transformation.

These efficiencies were achieved by optimizing pre-culture conditions, competent cell preparation, competent cell volume, voltage, and recovery time. The resultant protocol providing these efficiencies uses cells pre-cultured in rich media overnight, washed once, resuspended to an OD600 of 30, electroporated at 2700 V with 200 μL cells used in linear integration and 400 μL used for plasmid transformation, and recovered for 1 hr. Finally, this transformation method has the advantage of not requiring the use of any carrier DNA.

These results improve the set of genetic tools available for Y. lipolytica.

83 Chapter 5: Testing synthetic terminators in Y. lipolytica6

5.1 CHAPTER SUMMARY

In general, tool development in Y. lipolytica lags behind conventional hosts, as engineering is more cumbersome without highly stable episomal plasmids or easy targeted integration methods. Though the development of tools is difficult, some elements have been built and tested to control expression. Modulation of expression in metabolic engineering is often achieved at the most basic level through varying the promoter driving expression. On the other hand, terminators are an overlooked yet essential element directing cellular machinery to create a functional transcript. Although not yet common practice, an enhanced level of control is achievable through exchanging the terminator completing the transcription process. Due to the presumed conservation of function of terminators in yeast hosts, we tested the applicability of transferring S. cerevisiae terminators to Y. lipolytica. In this chapter, a subset of short, synthetic terminators designed for S. cerevisiae were imported into Y. lipolytica. Here, the impact of those synthetic terminators on production of a green fluorescent protein reporter in Y. lipolytica is described.

6 This chapter contains some content adapted from a previously authored manuscript. Reprinted with permission from Curran KA, Morse NJ, Markham KA, Wagman AM, Gupta A & Alper HS (2015) Short Synthetic Terminators for Improved Heterologous Gene Expression in Yeast. ACS Synthetic Biology 4: 824-832. ©2015 American Chemical Society

84 5.2 INTRODUCTION

Metabolic engineering in Y. lipolytica has been enabled by the development of basic genetic tools, and modulation of expression is achievable through control elements like promoters and terminators. At the most basic level, native promoters have different strengths leading to a variety of different expression levels. A common strategy for increasing promoter strength, particularly for Y. lipolytica, is to identify and concatenate upstream activating sequences in front of a core promoter [60-63]. Adding an additional level of control is achievable by hybridizing inducible elements such as for induction with oleic acid [64, 65], n-decane [66], and erythritol [67]. Further control is achievable by using synthetic terminators [70].

Strong terminators improve yields by increasing target enzymes abundance [68].

This improvement in yield is attributed to the terminators effects on mRNA stability and half-life [69]. In this chapter, synthetic terminators that were initially characterized in S. cerevisiae were imported into Y. lipolytica. These synthetic terminators were developed by varying a minimal set of elements described by Guo et al. [162] including the efficiency element, positioning element, poly(A) element, and spacers [70]. A subset of the synthetic terminator elements were tested in Y. lipolytica resulting in a range of expression levels of green fluorescent protein.

85 5.3 RESULTS AND DISCUSSION

Unlike promoters that heavily rely on specific transcription factor binding, terminators are potentially a more universal synthetic part. Y. lipolytica is phylogenetically distinct from S. cerevisiae [163], thus provides an interesting platform for testing the universality of synthetic terminators developed in the latter host. A subset of seven synthetic terminators (Tsynth2, Tsynth7, Tsynth8, Tsynth10, Tsynth22, Tsynth27, and

Tsynth30) along with TGuo1, CYC1, and TEF1 terminators (Table 5-1) were cloned into a heterologous expression cassette expressing the green fluorescent protein gene hrGFP along with a strong hybrid promoter developed previously [63]. The CYC1 terminator from S. cerevisiae and a 250 bp region immediately following the stop codon, assumed to be TEF1 and CYC1 terminators, from Y. lipolytica were also included for comparison. With the exception of Tsynth30, the seven tested synthetic terminators exhibited a similar performance in Y. lipolytica as that seen in S. cerevisiae, indicating the portability of these synthetic parts across yeast strains (Figure 5-1). Included among this set is one synthetic terminator, Tsynth10, which performed poorly in both hosts (due to high GC content in a linker region), demonstrating that Y. lipolytica terminators are likewise sensitive to the same constraints of sequence space seen in S. cerevisiae. The majority of synthetic terminators performed as well as or better than the native terminators, including a nearly

70% improvement over the currently used CYC1 terminator. These results demonstrate that short synthetic terminators can be highly functional across species and can potentially simplify the design and testing of constructs across multiple yeast hosts. Additionally, they provide an easily synthesizable set of control elements that can be used to tune gene expression for metabolic engineering efforts in Y. lipolytica.

86 Table 5-1: Terminator sequences used in this study.

Terminator Sequence Name Tguo1 TATATAACTGTCTAGAAATAAAGAGTATCATCTTTCAAA TATATATATATATATATATATATAACTGTCTAGAAATAA Tsynth2 AGAGTATCATCTTTCAAA Tsynth7 TATATAACTGTCTAGAAATAAATTTTTTCAAA TATATAAACTCATTTACTTATGTAGGAATAAAGAGTATC Tsynth8 ATCTTTCAAA TATATACACCCGTCGAGCCTGTCCGAAATAAAGAGTATC Tsynth10 ATCTTTCAAA TGGGTGGTATATATAACTGTCTAGAAATAAAGAGTATCA Tsynth22 TCTTTCAAA TGGGTGGTATATATATATATATATATATATATAACTGTCT Tsynth27 AGAAATAAAGAGTATCATCTTTCAAA TTTTTTTTTATATATATATATATATATATATAAACTCATTT Tsynth30 ACTTATGTAGGAATAAATTTTTTCAAA TCATGTAATTAGTTATGTCACGCTTACATTCACGCCCTCC CCCCACATCCGCTCTAACCGAAAAGGAAGGAGTTAGAC S. cerevisiae AACCTGAAGTCTAGGTCCCTATTTATTTTTTTATAGTTAT CYC1 GTTAGTATTAAGAACGTTATTTATATTTCAAATTTTTCTT TTTTTTCTGTACAGACGCGTGTACGCATGTAACATTATAC TGAAAACCTTGCTTGAGAAGGTTTTGGGACGCTCG GCGTCTACAACTGGACCCTTAGCCTGTATATATCAATTG ATTATTTAAAGATTTGGTCGGTAGGCGGTTCGTATTGTA CAATGGGATCTGTTACTGAGGTGGATCTACCCAACTTGC Y. lipolytica GAGATTCAATTGCGAGATTCAATCGCGAGATTCAATTGC CYC1 GAGAATCAGTTGCGAGTTGTTCTAACACTCAGCTTCTAC GAGCGCTTGTATTAGGACGAGTGATACTCCGTGGGGCGA CGGCTTCTCTTGCGTC GCTGCTTGTACCTAGTGCAACCCCAGTTTGTTAAAAATT AGTAGTCAAAAACTTCTGAGTTAGAAATTTGTGAGTGTA GTGAGATTGTAGAGTATCATGTGTGTCCGTAAGTGAAGT Y. lipolytica GTTATTGACTCTTAGTTAGTTTATCTAGTACTCGTTTAGT TEF1 TGACACTGATCTAGTATTTTACGAGGCGTATGACTTTAG CCAAGTGTTGTACTTAGTCTTCTCTCCAAACATGAGAGG GCTCTGTCACTCAGT

87 Figure 5-1: Fluorescence of hrGFP in Y. lipolytica using synthetic terminators.

Fluorescence resulting from different terminators on a vector expressing hrGFP was measured. Values are relative to the construct containing the S. cerevisiae CYC1 terminator expressed in Y. lipolytica. Error bars represent standard deviation of three biological replicates.

88 Through this initial study, the portability of terminators was demonstrated. In importing a variety of synthetic terminators from the model host S. cerevisiae to the phylogenetically distinct Y. lipolytica resulting in equivalent fluorescence outputs to a native Y. lipolytica terminator, the versatility of this synthetic part was demonstrated. This versatility is presumably due to how terminators contain conserved sequence elements leading to generic structure and function ultimately guiding mRNA stability and corresponding mRNA half-life. Ultimately, this set of synthetic terminators expands the synthetic biology toolkit for Y. lipolytica and facilitates maximizing expression from constructs used in metabolic engineering.

89 Chapter 6: Conclusions and major findings

Until relatively recently, the field of metabolic engineering has focused on the development of model host systems like Escherichia coli and Saccharomyces cerevisiae for the renewable production of chemicals. Although decades of research have led to a deep understanding of these hosts and creation of powerful tools for their genetic manipulation, they are not always the most advantageous choice. The nonconventional yeast Yarrowia lipolytica holds many superior features over these model systems, namely high flux through the TCA cycle, high flux through acyl-CoA intermediates, diverse substrate utilization, and a natively high chemical tolerance. This native flux through crucial metabolites linked to the oleaginous nature of Y. lipolytica enables the production of a diverse set of compounds beyond lipids and TCA cycle intermediates.

This dissertation seeks to move beyond the improvement of native metabolite production in Y. lipolytica to showcase the full potential of this host as a powerhouse for metabolic engineering. This was achieved by upgrading the lipid pool in previously engineered high-lipid strains to form the non-native product of cyclopropane fatty acids.

Additionally, the high acyl-CoA precursor flux was diverted into the polyketide triacetic acid lactone demonstrating the first production of this class of chemicals in Y. lipolytica.

These metabolic engineering successes were enabled by improvements in the basic genetic engineering technique of transformation, removing the bottleneck of low transformation efficiency for overexpressing gene targets. Finally, genetic tools in this organism were upgraded through a demonstration of the modulation of gene expression by synthetic

90 terminators in Y. lipolytica. Collectively, these efforts begin to demonstrate the wide range of possibilities for future engineering endeavors in this nonconventional host.

As an initial demonstration of the versatility of the high production lipid platform provided by Y. lipolytica, we sought to enzymatically upgrade the fatty acid pool to improve biodiesel properties in Chapter 2. The distributions of fatty acids in Y. lipolytica strains ENGR and L36 both contain a large percentage of unsaturated fatty acids which are prone to oxidation. Due to the desired chemical characteristics of biodiesel, e.g. fluidity, it was necessary to convert this pool to something other than saturated fatty acids.

Cyclopropane fatty acids allow for maintaining fluidity while improving oxidative stability. To achieve improved biodiesel properties, the E. coli cyclopropane fatty acid synthase was codon optimized and imported into ENGR and L36. With increased copy number and bioreactor scale up, over 32% of the total lipid pool was converted to C19 cyclopropane fatty acids. Ultimately, these efforts produced over 3 g/L of the desired C19 cyclopropane fatty acid product and demonstrated the promise of further engineering to create functionalized fatty acids in Y. lipolytica.

Building on previous successes in engineering Y. lipolytica to overproduce fatty acids, in Chapter 3, we hypothesized that the same precursor elements (acetyl-CoA and malonyl-CoA) could be diverted to biosynthesis of polyketides. Investigation of the simple polyketide triacetic acid lactone (TAL), built from the iterative condensation of two malonyl-CoA extender units on an acetyl-CoA starter followed by autocyclization, served as a demonstration of the potential of Y. lipolytica as a polyketide production platform.

TAL formation is catalyzed by the 2-pyrone synthase g2ps1 from Gerbera hybrida. This 91 gene was codon optimized and imported into wildtype Y. lipolytica. After amplifying the copy number to four through iterative transformation (resulting in strain YT), 2.1 g/L of

TAL was produced in a test tube, outperforming bioreactor production of E. coli. After demonstrating successful production, we sought to further divert flux towards TAL formation by investigating overexpression strategies related to four different native pathways- citrate, pyruvate bypass, pyruvate dehydrogenase complex, and β-oxidation.

Each of the first three overexpression strategies was paired with overexpression of ACC1, which catalyzes the conversion of acetyl-CoA to malonyl-CoA. Additionally, various media supplementations and nitrogen starvation strategies were tested in these engineered strains. Of these overexpression strategies, overexpression of the peroxisomal matrix protein Pex10p paired with nitrogen starvation resulted in the highest yield fermentation,

43% of the theoretical yield. The highest titer fermentation was achieved through dosing acetate mid-fermentation of strain YT- PDC2, ALD5, ACS1, ACC1, the best pyruvate bypass pathway overexpression strain. In a test tube, this strain produced 35% of theoretical yield (4.9 g/L TAL). After bioreactor optimization, this same strain produced 36 g/L of

TAL, approximately an order of magnitude improvement over production in S. cerevisiae.

This chapter effectively demonstrated a case where engineering in a nonconventional host was superior to its conventional counterparts while characterizing some native pathways that had previously been unexplored.

The aforementioned metabolic engineering efforts were enabled by the development and optimization of an electroporation transformation protocol, detailed in

Chapter 4. Chemical transformation protocols previously developed for Y. lipolytica 92 involve the need for large amounts of carrier DNA. The use of carrier DNA in a system so proficient in non-homologous end joining is undesired as it will invariably lead to additional exogenous DNA being integrated into the host’s genome. In addition, chemical transformation methods proved largely ineffective in the high lipid strains that have been developed in-house. Optimization of an electroporation transformation method was done by investigating three major factors in the transformation process: (i) pre-culture conditions and competent cell preparation, (ii) transformation conditions, and (iii) post-transformation conditions and recovery. Ultimately, the optimized electroporation transformation method provided many improvements over the chemical transformation previously used in our lab- simple competent cell preparation leading to a significant decrease in time required, removal of the need to use carrier DNA, and most importantly, increased transformation efficiency. This improvement in a basic genetic tool speeds up the rate of metabolic engineering in Y. lipolytica as it alleviates the bottleneck of testing constructs in this host.

Finally, the modulation of gene expression in Y. lipolytica was explored in Chapter

5. Previously, most constructs used in Y. lipolytica have used promoters of various strengths in order to control gene expression. In this chapter, we explored the use of synthetic terminator elements that had been previously developed for use in S. cerevisiae.

The synthetic terminators developed provided a way to modulate gene expression through short pieces of DNA that are easily synthesizable. This is effective, as terminators can increase enzyme abundance by increasing mRNA stability and half-life. A subset of these terminators was tested in Y. lipolytica. Ultimately, these synthetic terminators performed well in this alternative yeast host without any optimization. The best terminator was 93 equivalently strong to the native TEF1 terminator and resulted in a 70% increase in fluorescence over the currently used S. cerevisiae CYC1 terminator. This improvement in green fluorescent protein output was achieved using a strong Y. lipolytica hybrid promoter demonstrating that improvements in expression can be achieved even when genes are already highly expressed. This discovery provides a platform to enhance future engineering endeavors when high gene expression is required. With improvements in output simply through altering the terminator on a construct, the need for multiple iterative integrations of the same gene with different markers can be mitigated saving time and increasing the number of metabolic engineering steps before markers have to be recycled.

Collectively, the chapters contained in this dissertation provide demonstration that there is unmet potential in Y. lipolytica. Improvements in basic genetic tools will lead to the ability to develop more complicated synthetic biology resources, a field that is constantly evolving. Together, advancements in genetic and synthetic biology tools will lead to the ability to tackle complex metabolic engineering schemes increasing the chemical repertoire and industrial impact of Y. lipolytica.

94 Chapter 7: Proposals for future work

This work has furthered the development of Yarrowia lipolytica as an industrial powerhouse for the sustainable production of chemicals. In this chapter, further questions sparked by this research are discussed with potential areas for future study presented.

In Chapter 2, overexpression and increased copy number enabled production of over 3 g/L of cyclopropane fatty acids. While this is a promising demonstration of the capacity for Y. lipolytica to produce high titers of functionalized fatty acids, further optimization could increase the impact of this engineering endeavor. First, approximately

2 g/L of oleic acid, the immediate precursor to our molecule of interest, was produced during bioreactor fermentation. Improvements in conversion could potentially be made through enzyme engineering of the cyclopropane fatty acid synthase. Additionally, knockout of the gene encoding the delta 12 desaturase would remove flux from the desired pathway towards undesirable linoleic acid production. Finally, optimized bioreactor fermentation conditions would likely lead to improved output without any additional strain engineering. Specifically, it is well documented that at lower temperatures, more unsaturated fatty acids are formed [164, 165] providing more flux towards precursors for cyclopropanation. In a broader sense, this research could serve as a starting point for developing more complex designer fatty acids with desirable functional groups to be used in a variety of applications.

After demonstrating the flexibility of the lipid platform provided by Y. lipolytica to produce high titers of a specific lipid species, in Chapter 3 this platform was rewired to instead produce triacetic acid lactone. By increasing flux towards acetyl-CoA while 95 simultaneously funneling that flux away from lipids, product diversification was enabled.

The extensive pathway engineering and characterization of acetyl-CoA accumulation strategies in this chapter provide a broad starting point that can be applied to any number of products including other polyketides. While our engineering strategies proved successful, as the first study overexpressing many of these native genes, important questions were raised. Notably, the effects of PEX10 overexpression are largely unknown, and the complex regulatory shifts that occur under nitrogen starvation are not fully understood, though nitrogen starvation appears to induce both lipid formation [166] and

TAL production. Additionally, overexpression of the pyruvate bypass pathway genes appears to shift overall redox cofactor utilization. With further study, these effects could be better understood informing future engineering for production of molecules constructed from critical acyl-CoA building blocks.

From a less broad sense, TAL production could likely be improved through further fermentation optimization and strain modifications. During bioreactor fermentation, the

TAL-producing strain accumulates significant citrate and undergoes a diauxic shift after glucose depletion. The productivity of TAL production from citrate late in fermentation is high, and that effect could likely be exploited by employing a fed-batch fermentation strategy. Though this would likely improve titers, additional costs would be incurred in industrial fermentation by requiring a citrate feed and pH adjustment. An alternative approach could use strain engineering to reduce flux towards citrate, potentially using a

CRISPRi system [77]. To gain a better understanding of carbon loss, a full carbon balance using off-gas analysis and metabolomics would be beneficial in guiding additional strain 96 modifications. Critically, having a comprehensive view of carbon sinks in the engineered

Y. lipolytica strains would guide selection of gene deletions or knockdowns, which are cumbersome to test thus require rationally guided design. All the engineering for TAL production was done solely through gene overexpressions, and further improvements to flux are likely achievable through complementation of gene overexpressions with gene knockdown or knockout strategies. Overall, there are many interesting scientific pursuits arising from this study that would allow for a deeper understanding of the cellular and molecular biology of Y. lipolytica.

Although many metabolic engineering successes have been reported, general tool development for Y. lipolytica lags behind that of conventional microbial counterparts. The bottleneck of the import of exogenous DNA was addressed by developing and optimizing an electroporation transformation protocol in Chapter 4. This tool is essential for day to day engineering efforts, but still has inherent weaknesses. Due to highly active non- homologous end joining being the dominant form of DNA repair in this host, it is quite cumbersome to integrate at specific loci. In addition to difficulty targeting loci, episomal plasmids are largely ineffective. These two factors together make it nearly impossible to screen libraries of enzyme variants as the impact of genomic integration location often outweighs any change made by altering the enzyme, e.g. there is a low signal to noise ratio, as is also seen in mammalian cells [167]. To overcome this obstacle, targeted integrations could be improved by pairing this optimized transformation protocol with tools that will inhibit the random integration of DNA like CRISPRi [77] or induced protein degradation

[168] of the Ku heterodimer required for NHEJ without having to use a strain deficient of 97 ku70 or ku80. Improvements in methods to target integrations would not only lead to the ability to screen libraries of enzyme variants, but also would improve the protocol for creating full gene deletions. Hopefully, the transformation protocol presented here serves as a starting point for improving basic genetic manipulations in Y. lipolytica.

Lastly, a very basic testing of synthetic terminators in Chapter 5 demonstrated effective tuning of gene expression in Y. lipolytica. These terminators were tested only in a single context, and it is possible that the context surrounding the synthetic terminators will have a large impact on performance. Thus, these terminators should be tested with various promoters and reporters to demonstrate that they are robust tools. Additionally, further characterization of native terminators is warranted. Overall, the synthetic parts available for use in Y. lipolytica are lacking. These terminators serve as a useful starting point, but development of short, synthetic promoter elements could have a large impact on the field as the currently used promoters are often bulky and repetitive [62].

In summary, the work presented here begins to demonstrate the full potential of Y. lipolytica through showcasing production of two new products and developing an improved protocol for basic genetic manipulations. We have yet to achieve the full potential of metabolic engineering in this host, and future studies will help to advance the field. In investigating some of the questions that arose in this work, a deeper understanding of metabolism in oleaginous hosts will be gained. Finally, this work demonstrates how utilizing the strengths of nonconventional hosts like Y. lipolytica can help move the chemical industry towards a more renewable future.

98 Chapter 8: Materials and methods7

8.1 METHODS FOR CHAPTER 2

8.1.1 Plasmid and strain construction

The E. coli CFA synthase was codon optimized and synthesized by Blue Hero Biotech,

LLC. Synthesized DNA was amplified via PCR and cloned into overexpression constructs containing the strong constitutive promoter 16dTEF [62, 63] using InFusion (Clontech).

Plasmids were transformed into Stellar Competent Cells, derived from E. coli HST08

(Clontech), that were included with the InFusion cloning kit following the enclosed protocol. Vectors were linearized with NotI-HF (NEB) and transformed into Y. lipolytica strains L36-DGA1 [117] and pO1f ∆pex10 ∆mfe1 DGA1 (ENGR) [113] via electroporation. These strains were built from wildtype PO1f (ATCC MYA-2613), which was also used in this study. Auxotrophic selection for S1 (URA3) was achieved on

Complete Supplement Mixture lacking uracil (MP Biomedicals). Dominant selection occurred through use of the appropriate antifungal selection agents for HPH and NAT, hygromycin B (Invitrogen) and nourseothricin sulfate (Gold Bio), respectively [82].

7 Some content in this chapter is adapted from previously authored publications. KAM and CMP collectively wrote the materials and methods for Chapter 3. Reprinted with permission from Markham KA, Palmer CM, Chwatko M, Wagner JM, Murray C, Vazquez S, Swaminathan A, Chakravarty I, Lynd NA & Alper HS (2018) Rewiring Yarrowia lipolytica toward triacetic acid lactone for materials generation. Proceedings of the National Academy of Sciences of the United States of America 115: 2096-2101. Chapter 5 methods adapted from Curran KA, Morse NJ, Markham KA, Wagman AM, Gupta A & Alper HS (2015) Short Synthetic Terminators for Improved Heterologous Gene Expression in Yeast. ACS Synthetic Biology 4: 824-832. ©2015 American Chemical Society

99 Following transformation, individual colonies were isolated, stocked in 25% glycerol, and genomic DNA was extracted (Wizard gDNA kit, Promega). PCR of genomic DNA confirmed integration of ycoCFA.

8.1.2 Media conditions

Complete Synthetic Mixture (MP Biomedicals) prepared at the recommended concentration with Yeast Nitrogen Base containing 5 g/L of ammonium sulfate (BD Difco) and 20 g/L glucose (MP Biomedicals) was used for tube fermentations. Rich media containing 20 g/L yeast extract (VWR), 40 g/L peptone (VWR), 40 g/L glucose, 100 mg/L tryptophan (Acros), and 20 mg/L adenine (Sigma-Aldrich) was used to ferment cultures overnight for glycerol stocking [55]. YPD plates contained 10 g/L yeast extract, 20 g/L peptone, 20 g/L glucose, and 25 g/L agar (Teknova). Auxotrophic selection was achieved by plating on CSM-URA and dominant selection used YPD plates supplemented with

Hygromycin B (Thermo Fisher Scientific) at 300 mg/L or nourseothricin sulfate (Gold

Biotechnology) at 1 mg/mL [82].

8.1.3 Fermentation conditions

Tube cultures were fermented in 14 mL round-bottom tubes (Falcon). Starter cultures for tube fermentations were inoculated from glycerol stock into 2 mL media and grown for

24 hours at 28°C in a rotary drum. Precultures were diluted back to an OD600nm of 0.1 in 2 mL fresh media and grown for 72 hours at 28°C in a rotary drum.

100 8.1.4 Bioreactor fermentation

Bioreactor fermentation used a BioFlo 115 3 L fermenter (New Brunswick) with a final operating volume of 1.7 L under conditions similar to previous work in the ENGR strain [113]. Specifically, fermentations were inoculated into media containing 80 g/L glucose and 13.4 g/L Yeast Nitrogen Base without amino acids with ammonium sulfate.

An additional 80 g/L glucose was dosed in after 49 hours of fermentation. pH was controlled at 3.5 with 2.5 M NaOH. Temperature was controlled at 28 °C. Dissolved oxygen was controlled at 50% via agitation cascade from 250-800 RPM with inlet sparge set at 3.5 SPLM. Bioreactor seed cultures were inoculated by thawing an entire 1 mL glycerol stock and adding it to 50 mL of the initial media conditions in a 250 mL baffled flask. Seed cultures were fermented for 24 hours at 28 °C prior to bioreactor inoculation at

OD600nm of 0.1. Three samples of approximately 3 mL were taken from each reactor every

24 hours. Antifoam 204 (Sigma) was added as needed.

8.1.5 Cell density measurement

Samples from the bioreactor were diluted with PBS (Corning) and plated on YPD to measure viable cell density throughout the fermentation. Dilutions ranged from 105 –

7 10 as cell density increased throughout the time course.

8.1.6 Lipid isolation and quantification

For bioreactor samples, 500 µL of culture was analyzed for the 24 hour time point and 100 µL of culture was used for all following samples. For tube fermentations, 1 mL of 101 culture was analyzed. Lipids were extracted using a modified Folch extraction [169].

Specifically, cells were pelleted via centrifugation at 5000 xg for 3 minutes then washed with 1 mL PBS. An internal standard of nonadecanoic acid 98% + (Acros) was added to the pellet at 100 mg/L for tube samples, 500 mg/L for 24 hour bioreactor samples, and 1 g/L for all remaining bioreactor samples. Pellet was resuspended in 700 µL 2:1 v/v choloroform:methanol (Fisher), transferred to a screw cap vial containing 0.5 mm zirconia beads (BioSpec Products), and vortexed for 20 minutes. After vortexing, vials were centrifuged at 21,000 xg for 2 minutes. Supernatant was transferred to a fresh vial containing 150 µL 0.3% NaC1 in water and briefly vortexed then centrifuged at 1000 xg for 10 minutes. The lower phase was recovered and dried under nitrogen. Lipids were transesterified via 1 hr reaction at 85 °C in a mixture containing 86.75% methanol, 5% dimethoxypropane (Sigma Aldrich), and 8.25% concentrated HCl (Fisher) by volume.

FAMEs were extracted into 300 µL hexanes. Samples were analyzed on a Trace 1310 GC

(Thermo Fisher) equipped with a 30 m, 0.32 mm, 0.25 µm film DBwaxETR column

(Agilent), helium carrier gas (Airgas), and flame-ionization detector. Oven conditions were held at 40 °C for 1 min, ramped to 250 °C at a rate of 5 °C/min, and finally held at 250 °C for 16 min. Standard curves used to quantify FAMEs were prepared from methyl nonadecanoate (Fluka analytical), RM-6 (Supelco), and methyl cis-9,10- methyleneoctadecanoate (Matreya LLC).

102 8.1.7 Glucose quantification

To quantify glucose, 1 mL of sample was centrifuged at 10,000 xg for 10 minutes and syringe filtered through a nylon filter. Samples outside of the linear range were diluted with water and rerun. Separation of metabolites was achieved on HPLC (Thermo Fisher) via Aminex 87H (300 x 7.8 mm) with isocratic flow of 5 mM H2SO4 at 0.6 mL/min and

50 °C for 25 min. Glucose was detected with RI. Standard curves were prepared with D- glucose (MP Biomedicals).

8.1.8 Copy number assay

Genomic copy number was determined from gDNA using the Applied Biosystems

Power SYBR Green PCR Mastermix following the enclosed protocol. A plasmid built in a pUC19 vector containing both actin and ycoCFA was used to create a standard reference curve with primers KM1 (AGTCCAACCGAGAGAAGATG) and KM2

(GGGGAGAGAGAAACCAGAG) for actin and KM31

(CTTCCTTCGGTTCGACAAAT) and KM32 (CGTTCATACCAGGCCATAAG) for ycoCFA. Primers were designed using PrimerQuest (IDT) targeting a melting temperature of 60 °C. Copy number was calculated using the standard reference curve compared to actin.

103 8.2 METHODS FOR CHAPTER 3

8.2.1 Plasmid and strain construction

The 2-pyrone synthase gene¸ g2ps1, was codon optimized using Blue Heron

Biotech’s algorithm with the publically available Y. lipolytica codon table and synthesized as a gblock by IDT. Point mutations were introduced into g2ps1 through hybrid PCR [170].

Native overexpression targets were amplified from gDNA; the genes explored in this study have the same sequence in both CLIB122 and PO1f genomes. Genes were cloned into Y. lipolytica integration constructs containing the synthetic promoter 16dTEF [62, 171] and assembled using the In-Fusion HD Cloning Kit (Clontech).

Assembled plasmids were linearized using NotI-HF (NEB) and were sequentially transformed into the wild type Y. lipolytica strain PO1f [61] via electroporation. Genes were stably integrated using either dominant or auxotrophic markers. Auxotrophic selection was achieved with Complete Supplement Mixture lacking leucine, uracil, or tryptophan (MP Biomedicals). Dominant selection was achieved using hygromycin B

(Invitrogen), nourseothricin sulfate (Gold Bio), mycophenolic acid (Sigma-Aldrich), or chlorimuron ethyl (Supelco). After each transformation, individual colonies were isolated and glycerol stocked. Genomic DNA was extracted (Wizard gDNA kit, Promega) and gene integration was confirmed via PCR. Positive transformants were screened for TAL production via tube fermentation followed by HPLC analysis. As each gene overexpression is randomly integrated into the genome, screening of multiple colonies was essential. At each stage, the strain with the highest TAL titer was selected for further engineering.

104 8.2.2 Media conditions

Rich 2X YPADW was prepared with 40 g/L D-glucose (MP Biomedicals), 40 g/L peptone (VWR), 20 g/L yeast extract (VWR), 100 mg/L tryptophan (Acros) and 20 mg/L adenine (Sigma-Aldrich). Cultures grown in 2X YPADW were used to create 25% glycerol stocks. Growth media for tube fermentations consisted of Complete Supplement Mixture

(MP Biomedicals) at the recommended concentration, Yeast Nitrogen Base (BD Difco) at

5 g/L of ammonium sulfate (unless otherwise specified) and 20 g/L D-glucose (MP

Biomedicals). YPD plates used for viable cell counts consisted of 20 g/L peptone (VWR),

20 g/L D-glucose (MP Biomedicals), 10 g/L yeast extract (VWR), and 20 g/L agar

(Teknova).

8.2.3 Tube fermentation

Starter cultures for tube fermentations were inoculated from single colony glycerol stocks and grown for 24 hours at 28°C. Fermentations were inoculated from starter cultures at an OD600nm of 0.1 and grown for an additional 96 hours prior to sample preparation unless otherwise noted. For vitamin supplementation, filter sterilized vitamin solutions were prepared and added to media at previously published [172] concentrations: thiamine pyrophosphate 20.5 mg/L, biotin 0.075 mg/L and D-pantothenic acid hemicalcium salt 1.5 mg/L (Sigma-Aldrich). A 1000X stock of Cerulenin (EMD Millipore Biosciences) was prepared in DMSO and spiked into cultures grown in regular media at 24 hours for a final concentration of 1 μg/ml. Alternative carbon source feeding experiments were performed through supplementation of sodium pyruvate, sodium citrate monobasic or sodium acetate 105 solutions (Sigma-Aldrich). Sodium pyruvate and sodium citrate were used in place of glucose as sole carbon source at a carbon molar equivalent to 20 g/L of glucose. Sodium acetate was spiked into cultures grown in regular media after 24 hours of fermentation at a carbon molar equivalent to 10 g/L of glucose.

8.2.4 HPLC quantification of TAL

Samples were prepared for HPLC by diluting to the linear range, vortex mixing on high, and centrifuging at 500 xg for 5 minutes to remove cells. Supernatant was then filtered with a nylon syringe filter. TAL was quantified from filtered supernatant on a

Dionex UltiMate 3000 (Thermo) fitted with an Agilent LS Eclipse Plus C18 column (3.0 x 150mm, 3.5 µm) with detection at 280 nm. Column oven was held at 25 °C with 1% acetic acid in water or acetonitrile as the mobile phase being ramped from 5% organic to

100% organic over 40 minutes. A standard curve was prepared using >98.0% purity TAL from TCI America.

8.2.5 HPLC quantification of fermentation byproducts

Samples were prepared for HPLC as previously described. Fermentation byproducts were identified and quantified on a Dionex UltiMate 3000 (Thermo) fitted with an Aminex HPX-87H column (300 x 7.8 mm) with UV detection at 210 nm and RI. Column oven was held at 50°C with 5 mM sulfuric acid as the mobile phase. A constant flow rate of 0.60 ml/min was used over the course of the 25-minute sequence. Standard curves were

106 prepared using D-glucose (MP Biomedicals), sodium citrate (Sigma-Aldrich) and sodium acetate (Sigma-Aldrich).

8.2.6 Lipid isolation and quantification

Lipids were extracted from 1 mL of culture following the procedure described by

Folch et al. and modified for yeast [173]. After extraction, solvent was evaporated under nitrogen. Lipids were transesterified in 0.5 mL of a mixture of 8.25% concentrated hydrochloric acid, 5% dimethoxypropane, and 86.75% methanol by volume incubated at

85 °C for 1 hour. Fatty acid methyl esters (FAMEs) were extracted with the addition of 0.5 mL 0.9% sodium chloride and 0.3 mL of hexanes. FAMEs were separated with a 30 m,

0.32 mm, 0.25 μm film DBwaxETR column (Agilent) and detected via FID on a Trace

1310 GC (Thermo Fisher). FAMEs were quantified using RM-6 (Supleco) and normalized via addition of nonadecanoic acid (Acros Organics) as an internal standard prior to extraction.

8.2.7 TAL toxicity assay

TAL toxicity assay was performed using a Bioscreen C (Growth Curves USA). The wild type strain PO1f was inoculated at an OD600nm of 0.1 into media containing Complete

Supplement Mixture, Yeast Nitrogen Base and glucose at previously described concentrations, as well as varying concentrations of TAL (TCI America). An 8 g/L stock solution of TAL was prepared in water and filter sterilized. Cells were grown at 28°C with

107 continuous shaking for 50 hours. OD600nm was measured every 15 minutes and the experiment was performed in triplicate.

8.2.8 Bioreactor fermentation

Bioreactor fermentations were performed using a 3 L New Brunswick Bioflo115 system. The pyruvate bypass strain YT-ACS1, ALD5, PDC2, ACC1 was fermented in YP media prepared with 180 g/L D-glucose (MP Biomedicals), 40 g/L peptone (VWR), and

20 g/L yeast extract (VWR). A 13.7 g/L sodium acetate anhydrous (Sigma-Aldrich) spike was added after 36 hours of fermentation. Air was sparged into the fermenter at 3.5 splm.

Dissolved oxygen was controlled at 50% with an agitation cascade of 250-800 RPM. The pH was base controlled to 6.5 using a 2 M NaOH solution. Three separate samples were taken every 24 hours; this helped to compensate for the heterogeneous nature of the fermentation as TAL began to precipitate. Samples for HPLC quantification of TAL were diluted 20-100X in a 10% triethylene glycol (TCI America) solution and solubilized by 30 minutes of vortex mixing on high. Supernatant was then prepared for HPLC as previously described. TAL standards were also prepared in 10% triethylene glycol for quantification.

Samples were prepared for HPLC analysis of fermentation productions as previously described. Samples were also diluted and plated on YPD; the resulting number of colonies was used to determine viable cell count.

108 8.2.9 Copy number assay

Genomic copy number was determined from gDNA using the Applied Biosystems

Power SYBR Green PCR Mastermix following the enclosed protocol. A standard reference curve was built in a pUC19 vector and contained both actin and g2ps1 for determination of copy number in YT. Two other plasmids were built containing actin, g2ps1, and PEX10 as well as actin, g2ps1, ACS1, ALD5, PDC2, and ACC1 to determine copy number in the further engineered strains. Copy number was calculated using the standard reference curve compared to actin (Figure S10 and S11). The sequences of qPCR primers used are provided in Table S2.

8.2.10 Acetylaldehyde dehydrogenase activity assay

Acetylaldehyde dehydrogenase activity was assessed as the rate of NAD(P)H production, detected using the diaphorase-resazurin system adapted from Zhu, et al. [174].

A crude lysate was generated from saturated overnight cell cultures via bead milling in 50 mM HEPES buffer (pH 8.0). Protein concentration was determined using a Bradford assay and 0.01 mg/ml of protein was added to an assay buffer containing 50mM HEPES buffer

+ + (pH 8.0), 1mM MgCl2, 100 μM NADP or NAD , 10 μM resazurin, and 0.2 U/ml diaphorase. The acetylaldehyde dehydrogenase reaction was initiated by adding acetylaldehyde to a final concentration of 50 μM. Fluorescence at 590 nm was measured every minute for 30 minutes using excitation at 530 nm in a Cytation 3 imaging reader

(BioTek).

109 8.2.11 Crispr targeting to TRP1

To increase available selection markers, TRP1 was mutated for loss of function through creation of an indel. TRP1 phenotypic deletion was accomplished using CRISPR-

Cas9 in the YT strain background following the protocol used in Schwartz et al. [76]. Prior to use in YT, the pCRISPRyl plasmid marker was swapped to allow hygromycin resistance using standard cloning techniques, as the strain was already prototrophic for leucine.

Targeted gRNA were designed using CRISPRdirect [175].

8.2.12 Chemicals used for materials generation8

Sodium hydroxide pellets (Fisher Chemical), 1,8-Diazabicyclo[5.4.0]undec-7-ene

(Aldrich), and methanol (Fisher Chemical) were used as received. Anhydrous N,N- dimethylformamide was obtained from a JC Meyer solvent system. Polyepichlorohydrin was prepared by using a previously described procedure [176]. It was cleaned via 0.1M sodium hydroxide wash, followed by precipitation in methanol and drying at reduced pressure. All air and moisture sensitive reactions were prepared under a dry nitrogen atmosphere.

8.2.13 Equipment used for materials generation

1H NMR and 13C NMR spectroscopy were performed on a 400 MHz Agilent MR spectrometer at room temperature and referenced to the residual solvent signal of CDCl3.

8 Materials experimental work and text was contributed by Malgorzata Chwatko from Dr. Nathaniel A. Lynd’s lab at UT-Austin. 110 Size exclusion chromatography (SEC) was carried out on one of two systems: (1) an

Agilent system with a 1260 Infinity isocratic pump, degasser, and thermostated column chamber held at 30 °C containing Agilent PLgel 10 μm MIXED-B and 5 μm MIXED-C columns with a combined operating range of 200–10,000,000 g mol−1 relative to polystyrene standards, or (2) an Agilent system with a 1260 Infinity II isocratic pump, degasser, and thermostated column chamber held at 30 °C containing Agilent PLgel 10 μm

MIXED-B with a combined operating range of 500–10,000,000 g mol−1 relative to polystyrene standards. Chloroform with 50 ppm amylene was used as the mobile phase on both systems. System (1) was equipped with an Agilent 1260 Infinity refractometer, dual angle dynamic and static light scattering. System (2) was equipped with a suite of detectors from Wyatt Technologies, which provided measurement of polymer concentration, molecular weight, and viscosity. Static light scattering was measured using a DAWN

HELEOS II Peltier system with differential refractive index measured with an Optilab

TrEX, and differential viscosity measured using a Viscostar II. Differential scanning calorimetric (DSC) tests were conducted on a TA250 instrument with an initial heating and

–1 –1 cooling rate of 10°C min and second rate of 2°C min under a N2 atmosphere and the data from the second heating curve were collected. UV-Vis spectroscopy was done on

PerkinElmer Lambda 35 UV.VIS spectrometer. The polymers were analyzed in chloroform, while TAL was analyzed in methanol.

111 8.2.14 Purification of TAL from fermentation media

Bioreactor product was centrifuged at 3000 xg for 5 min to separate cells from the supernatant. Cell-free broth (500 mL) was extracted with three 500 mL portions of ethyl acetate/acetic acid (99:1, v/v). The organic phases were combined and the solution was concentrated to dryness in a rotary evaporator under reduced pressure. The resulting material was recrystallized from ethyl acetate.

8.2.15 Synthesis of poly[(epichlorohydrin)-co-(epoxy triacetic acid lactone)]

Poly(epichlorohydrin )(PECH) (2 g) was added to a reaction vial after which 1.5 molar excess of triacetic acid lactone was added to create a ratio of 1 to 1.5. To this monomer mixture, 6 mL of dry N,N-dimethylformamide (DMF) was added after which the vial was sealed and kept in an inert nitrogen atmosphere. 1,8-Diazabicyclo[5.4.0]undec-7- ene (DBU) was added in either 0.75 or 1.0 molar equivalence. The reaction was then stirred and heated to 110°C for 24 hours. The reaction was terminated and precipitated in water.

The polymer was then dried in vacuo. The sample at this time was characterized by NMR

1 spectroscopy and SEC. H NMR Spectroscopy (400 MHz, CDCl3), integrals reported relative to repeat unit (RU): δ 5.78 (s, 1H, -O-C-CH-C(=O)-O-C-(CH3)-CH-), δ 5.38 (s,1H,

O-(C-CH-C(=O)-O-C-(CH3)-CH-)), δ 4.05 (dd, -CH2-CH-(CH2-O-)-O-), δ 3.68 (m, -CH2-

CH-(CH2-Cl)-O-) and (m, -CH2-CH-(CH2-O-)-O-), δ 2.19 (s, 1H, -O-(C-CH-C(=O)-O-C-

13 (CH3)-CH-)), C NMR (100 MHz, CDCl3): δ 170.30 ( -O-(C-CH-C(=O)-O-C-(CH3)-CH-

)), δ 164.64 ( -O-(C-CH-C(=O)-O-C-(CH3)-CH-)), δ 162.6 ( -O-(C-CH-C(=O)-O-C-

(CH3)-CH-)),δ 100.27 ( -O-(C-CH-C(=O)-O-C-(CH3)-CH-)), δ 87.87 ( O-(C-CH-C(=O)- 112 O-C-(CH3)-CH-)), δ 77.89 (-CH2-CH-(CH2-Cl)-O-), δ 77.25 ( -CH2-CH-(CH2-O-)-O-), δ

68.94 (-CH2-CH-(CH2-O-)-O-), (-CH2-CH-(CH2-Cl)-O-) and ( -CH2-CH-(CH2-O-)-O-), δ

43.33 (-CH2-CH-(CH2-Cl)-O-), δ 19.17 (-O-(C-CH-C(=O)-O-C-(CH3)-CH-)).

113 8.3 METHODS FOR CHAPTER 4

8.3.1 Media conditions

LB broth (Teknova) was prepared with 100 μg mL-1 ampicillin (GoldBio). Selective plates of Complete Synthetic Mixture lacking Leucine (CSM-Leu, MP Biomedicals) were prepared at the recommended concentration with 20 g L-1 glucose (MP Biomedicals), 25 g

L-1 agar (Teknova), and Yeast Nitrogen Base containing 5 g L-1 ammonium sulfate (BD

Difco). Rich media plates containing 10 g L-1 yeast extract (VWR), 20 g L-1 peptone

(VWR), 20 g L-1 glucose, and 25 g L-1 agar were prepared for total cell counts. Rich media containing 20 g L-1 yeast extract, 40 g L-1 peptone, 40 g L-1 glucose, 100 mg L-1 tryptophan

(Acros), and 20 mg L-1 adenine (Sigma-Aldrich) was prepared for competent cell cultivation. Sorbitol (Acros) was prepared at 1 M concentration. Glycerol (Fisher) was prepared at 10% concentration by dilution with deionized water.

8.3.2 Plasmid preparations

The linear DNA fragment used for testing transformation efficiency was built from an integration vector containing a leucine marker surrounded by LoxP sites and the strong

16dTEF promoter driving gene expression, all flanked by 26s rDNA integrative sequences, as previously described for engineering in Y. lipolytica [62, 113]. The gene of interest from this previous study was removed via digestion with AscI (New England Biolabs Inc) and

PacI (New England Biolabs Inc) to create an empty backbone. Backbone was gel extracted and purified with GeneJET Gel Extraction Kit (Thermo Fisher Scientific). The fluorescent

114 reporter, green fluorescent protein (hrGFP), was amplified from pMCS-16dTEF-hrGFP

[62] using PCR and assembled with the backbone using InFusion (Clontech). Cells for plasmid isolation of the linear vector were inoculated from glycerol stock into 200 mL

LB/AMP media grown overnight in a 500 mL flask at 37 °C with 225 RPM orbital shaking.

DNA for transformation was prepared via maxiprep with the ZymoPURE Plasmid

Maxiprep Kit (Zymo Research) following the enclosed protocol. The plasmid tested for linear integration was linearized with NotI-HF (New England Biolabs) in an overnight digestion. Overnight digestions were purified with the GeneJET PCR Purification Kit

(Thermo Fisher Scientific). Cells harboring DNA for testing plasmid transformation efficiency (pMCS-16dTEF-hrGFP [62]) were inoculated into several 5 mL volumes of

LB/AMP in 14 mL round-bottom tubes and grown overnight at 37 °C with orbital shaking at 225 RPM. DNA from these cultures was isolated using the GeneJET Plasmid Miniprep

Kit (Thermo Fisher Scientific) following the enclosed protocol.

8.3.3 Strains and yeast cultivation

Transformation methods were tested in a laboratory wildtype strain PO1f (ATCC

MYA-2613). Pre-cultures were inoculated from a glycerol stock stored at -80 °C into 50 mL of rich media in a 250 mL baffled flask. Yeast was grown at 28 °C with orbital shaking at 225 RPM for 21 hours if time is not specified otherwise. Back dilutions were prepared by inoculating 10 mL of overnight culture into 40 mL fresh rich media in a flask. For tube cultures, cells were inoculated into 2 mL of rich media in 14 mL round-bottom tubes

(Falcon) and grown at 28 °C in a rotary drum.

115 8.3.4 Transformation and quantification

Competent cells were prepared from overnight cultures of PO1f. Culture was transferred to a 50 mL conical tube (Falcon) and centrifuged at 500 x g for 5 min.

Supernatant was discarded and pellets were washed once with 50 mL of 1 M sorbitol, unless otherwise specified. After washing, cells were centrifuged 500 x g for 5 min. Again,

supernatant was discarded. Cells were then resuspended to the proper OD600 in 1 M sorbitol.

An OD600 of 30 was used unless otherwise specified. A volume of these competent cells

(400 μL unless otherwise specified) was mixed with 5 μg of linearized DNA in an electroporation cuvette with 2 mm gap (VWR). Cells were electroporated with a Gene

Pulser Xcell Electroporation System (Bio-Rad) with the following settings- exponential decay pulse, 25 µF, 200 ohm, 0.2 cm cuvette, and 2700 V (unless voltage is otherwise specified). After electroporation, cells were quenched with 600 μL rich media two times, and as much liquid as could be recovered with a 1 mL pipette was transferred to a 14 mL round-bottom tube. Cells were recovered in a rotary drum at 28 °C for one hour unless otherwise specified. After recovery, cells were centrifuged at 500 x g for 5 min and washed with 1 mL of 1 M sorbitol. After an additional centrifugation, cells were resuspended in 1 mL of 1 M sorbitol. An appropriate dilution was plated on CSM-Leu selective plates.

Dilution plating for viable cells and competent cell counts was done with rich media plates.

Plates were incubated for two days at 28 °C then counted to obtain colony counts.

Efficiencies were calculated by dividing the average colony count by the amount of DNA normalized to the cassette of interest (lacking the additional linearized puc19 fragment).

116 8.4 METHODS FOR CHAPTER 5

8.4.1 Strains and cultivation

Yarrowia lipolytica strain PO1f (MatA, leu2–270, ura3–302, xpr2–322, axp-2)

(ATCC no. MYA-2613) was used in this work and propagated at 30 °C in synthetic complete medium lacking leucine (YSC-Leu) containing 20 g/L glucose and 6.7 g/L yeast nitrogen base with ammonium sulfate. Escherichia coli strain DH10β was used for all cloning and plasmid propagation. DH10 β was grown at 37 °C in 5 mL Luria–Bertani (LB) broth supplemented with 50 μg/mL of ampicillin. E. coli strains were cultivated with 225 rpm orbital shaking. Y. lipolytica strains were cultivated in a rotary drum (CT-7, New

Brunswick Scientific) at speed seven to facilitate oxygenation of the cultures. Yeast and bacterial strains were stored at −80 °C in 25% glycerol.

8.4.2 Plasmid construction

Y. lipolytica plasmids were modified centromeric, replicative plasmids as described in earlier work [63]. The vector pMCS-UAS1B8-hrGFP was used to create all plasmids.

Synthetic terminators were prepared with annealing and extending in HF Phusion buffer as described by Curran et al. [70]. The native terminator regions were obtained via PCR from genomic DNA purified with the Wizard genomic DNA purification kit (Promega). PCR reactions were run with recommended conditions using HF Phusion polymerase (New

England BioLabs, Inc.). Cloned terminators were inserted into the plasmid vector via PacI and PmeI restriction sites using T4 DNA ligase (Thermo Scientific). All plasmids were

117 transformed using the Zymo EZ freeze yeast transformation kit II (Zymo Research, Irvine,

CA) according to manufacturer’s instructions.

8.4.3 Flow cytometry

Y. lipolytica strains were initially propagated from individual colonies on YSC-Leu plates into 2 mL of fresh YSC-Leu media. After 48 h of incubation in a rotary drum, cultures were normalized to an OD600 of 0.03 in 2 mL of fresh YSC-Leu media. Cultures were grown 48 h before being harvested. To harvest, cultures were spun at 1000g for 5 min, washed with 5 mL of ice-cold water, and then 100 μL of this wash was added to 1 mL of ice-cold water. Fluorescence from Y. lipolytica expressing the hrGFP gene was measured using the GFP fluorochrome, a voltage of 319, and 10 000 events. Day-to-day voltage variability was mitigated by measuring all comparable strains on the same day.

FlowJo (Tree Star Inc., Ashland, OR) was used to analyze data and to compute mean fluorescence values.

118 References

1. Sun, J. and H. Alper, Synthetic Biology: An Emerging Approach for Strain Engineering, in Industrial Biotechnology. 2017, Wiley-VCH Verlag GmbH & Co. KGaA. p. 85-110. 2. Liu, L., H. Redden, and H.S. Alper, Frontiers of yeast metabolic engineering: diversifying beyond ethanol and Saccharomyces. Current Opinion in Biotechnology, 2013. 24(6): p. 1023-1030. 3. Groenewald, M., et al., Yarrowia lipolytica: safety assessment of an oleaginous yeast with a great industrial potential. Crit Rev Microbiol, 2014. 40(3): p. 187-206. 4. Markham, K.A., et al., Yarrowia lipolytica as a Cell Factory for Oleochemical Biotechnology, in Consequences of Microbial Interactions with Hydrocarbons, Oils, and Lipids: Production of Fuels and Chemicals, S.Y. Lee, Editor. 2017, Springer International Publishing: Cham. p. 1-18. 5. Fickers, P., et al., Hydrophobic substrate utilisation by the yeast Yarrowia lipolytica, and its potential applications. FEMS Yeast Res, 2005. 5(6-7): p. 527- 43. 6. Papanikolaou, S. and G. Aggelis, Biotechnological valorization of biodiesel derived glycerol waste through production of single cell oil and citric acid by Yarrowia lipolytica. Lipid Technology, 2009. 21(4): p. 83-87. 7. Barth, G. and C. Gaillardin, Physiology and genetics of the dimorphic fungus Yarrowia lipolytica. FEMS Microbiology Reviews, 1997. 19(4): p. 219-237. 8. Brígida, A.I.S., et al., Lipase from Yarrowia lipolytica: Production, characterization and application as an industrial biocatalyst. Journal of Molecular Catalysis B: Enzymatic, 2014. 101: p. 148-158. 9. Howard Damude, P.G., Daniel Macool, Stephen Picataggio, Dana Pollak, James Ragghianti, Zhixiong Xue, Narendra Yadav, Hongxiang Zhang, Quinn Zhu Current High arachidonic acid producing strains of Yarrowia lipolytica. 2005, E I du Pont de Nemours and Co 10. Damude Howard, G., et al., High Eicosapentaenoic Acid Producing Strains Of Yarrowia lipolytica. 2011, DU PONT: US. 11. Xie, D., E.N. Jackson, and Q. Zhu, Sustainable source of omega-3 eicosapentaenoic acid from metabolically engineered Yarrowia lipolytica: from fundamental research to commercial production. Applied Microbiology and Biotechnology, 2015. 99: p. 1599-1610. 12. Davidow, L.S., et al., Integrative transformation of the yeast Yarrowia lipolytica. Current Genetics, 1985. 10(1): p. 39-48. 13. Juretzek, T., et al., Vectors for gene expression and amplification in the yeast Yarrowia lipolytica. Yeast, 2001. 18(2): p. 97-113. 14. Fickers, P., et al., New disruption cassettes for rapid gene disruption and marker rescue in the yeast Yarrowia lipolytica. Journal of Microbiological Methods, 2003. 55(3): p. 727-737.

119 15. Wagner, J.M. and H.S. Alper, Synthetic biology and molecular genetics in nonconventional yeasts: Current tools and future advances. Fungal Genetics and Biology, 2016. 89: p. 126-136. 16. Yamane, T., et al., Dissection of centromeric DNA from yeast Yarrowia lipolytica and identification of protein-binding site required for plasmid transmission. J Biosci Bioeng, 2008. 105(6): p. 571-8. 17. Liu, L., et al., Increasing expression level and copy number of a Yarrowia lipolytica plasmid through regulated centromere function. FEMS Yeast Res, 2014. 14(7): p. 1124-7. 18. Fournier, P., et al., Colocalization of centromeric and replicative functions on autonomously replicating sequences isolated from the yeast Yarrowia lipolytica. Proc Natl Acad Sci U S A, 1993. 90(11): p. 4912-6. 19. Vernis, L., et al., An origin of replication and a centromere are both needed to establish a replicative plasmid in the yeast Yarrowia lipolytica. Molecular and Cellular Biology, 1997. 17(4): p. 1995-2004. 20. Ledesma-Amaro, R. and J.-M. Nicaud, Metabolic Engineering for Expanding the Substrate Range of Yarrowia lipolytica. Trends in Biotechnology, 2016. 34(10): p. 798-809. 21. Shaw, A.J., et al., Metabolic engineering of microbial competitive advantage for industrial fermentation processes. Science, 2016. 353(6299): p. 583-6. 22. Beopoulos, A., et al., Yarrowia lipolytica as a Cell Factory for Oleochemical Biotechnology, in Handbook of Hydrocarbon and Lipid Microbiology, K.N. Timmis, Editor. 2010, Springer Berlin Heidelberg: Berlin, Heidelberg. p. 3003- 3010. 23. Ledesma-Amaro, R. and J.-M. Nicaud, Yarrowia lipolytica as a biotechnological chassis to produce usual and unusual fatty acids. Progress in Lipid Research, 2016. 61: p. 40-50. 24. Qiao, K., et al., Lipid production in Yarrowia lipolytica is maximized by engineering cytosolic redox metabolism. Vol. 35. 2017. 25. Xu, P., et al., Engineering Yarrowia lipolytica as a platform for synthesis of drop- in transportation fuels and oleochemicals. Proc Natl Acad Sci U S A, 2016. 113(39): p. 10848-53. 26. Liu, T., H. Vora, and C. Khosla, Quantitative analysis and engineering of fatty acid biosynthesis in E. coli. Metabolic Engineering, 2010. 12(4): p. 378-386. 27. Runguphan, W. and J.D. Keasling, Metabolic engineering of Saccharomyces cerevisiae for production of fatty acid-derived biofuels and chemicals. Metabolic Engineering, 2014. 21: p. 103-113. 28. Wang, G., et al., Comparative transcriptome analysis reveals multiple functions for Mhy1p in lipid biosynthesis in the oleaginous yeast Yarrowia lipolytica. Biochim Biophys Acta, 2018. 1863(1): p. 81-90. 29. Brabender, M., et al., Urea and urine are a viable and cost-effective nitrogen source for Yarrowia lipolytica biomass and lipid accumulation. Appl Microbiol Biotechnol, 2018. 102(5): p. 2313-2322. 120 30. Connor, J.M., The Citric Acid Industry, in Global Price Fixing. 2008, Springer: Berlin, Heidelberg. 31. Levinson, W.E., C.P. Kurtzman, and T.M. Kuo, Characterization of Yarrowia lipolytica and related species for citric acid production from glycerol. Enzyme and Microbial Technology, 2007. 41(3): p. 292-295. 32. Kamzolova, S.V. and I.G. Morgunov, Metabolic peculiarities of the citric acid overproduction from glucose in yeasts Yarrowia lipolytica. Bioresour Technol, 2017. 243: p. 433-440. 33. Zeng, W., et al., Biosynthesis of keto acids by fed-batch culture of Yarrowia lipolytica WSH-Z06. Bioresour Technol, 2017. 243: p. 1037-1043. 34. Zhou, J., et al., Screening of a thiamine-auxotrophic yeast for alpha-ketoglutaric acid overproduction. Lett Appl Microbiol, 2010. 51(3): p. 264-71. 35. Morgunov, I.G., et al., Pyruvic acid production by a thiamine auxotroph of Yarrowia lipolytica. Process Biochemistry, 2004. 39(11): p. 1469-1474. 36. Cui, Z., et al., Engineering of unconventional yeast Yarrowia lipolytica for efficient succinic acid production from glycerol at low pH. Metab Eng, 2017. 42: p. 126- 133. 37. Blazeck, J., et al., Metabolic engineering of Yarrowia lipolytica for itaconic acid production. Metab Eng, 2015. 32: p. 66-73. 38. Pamela L. Sharpe, R.W.Y., Quinn Qun Zhu Carotenoid production in a recombinant oleaginous yeast. 2006, E I du Pont de Nemours and Co 39. Trueheart, J., et al., Production Of Carotenoids In Oleaginous Yeast And Fungi. 2010: WO. 40. Matthäus, F., et al., Production of Lycopene in the Non-Carotenoid-Producing Yeast Yarrowia lipolytica. Applied and Environmental Microbiology, 2014. 80(5): p. 1660-1669. 41. Nambou, K., et al., Flux Balance Analysis Inspired Bioprocess Upgrading for Lycopene Production by a Metabolically Engineered Strain of Yarrowia lipolytica. Metabolites, 2015. 5(4): p. 794-813. 42. Larroude, M., et al., A synthetic biology approach to transform Yarrowia lipolytica into a competitive biotechnological producer of β-carotene. Biotechnology and Bioengineering, 2018. 115(2): p. 464-472. 43. Grenfell-Lee, D., et al., The safety of beta-carotene from Yarrowia lipolytica. Food Chem Toxicol, 2014. 65: p. 1-11. 44. Kildegaard, K.R., et al., Engineering of Yarrowia lipolytica for production of astaxanthin. Synthetic and Systems Biotechnology, 2017. 2(4): p. 287-294. 45. Wagner, J.M., et al., A comparative analysis of single cell and droplet-based FACS for improving production phenotypes: Riboflavin overproduction in Yarrowia lipolytica. Metabolic Engineering, 2018. 47: p. 346-356. 46. Rymowicz, W., A. Rywinska, and M. Marcinkiewicz, High-yield production of erythritol from raw glycerol in fed-batch cultures of Yarrowia lipolytica. Biotechnol Lett, 2009. 31(3): p. 377-80.

121 47. Mirończuk, A.M., A. Biegalska, and A. Dobrowolski, Functional overexpression of genes involved in erythritol synthesis in the yeast Yarrowia lipolytica. Biotechnology for Biofuels, 2017. 10(1): p. 77. 48. Rakicka, M., et al., Polyol production from waste materials by genetically modified Yarrowia lipolytica. Bioresour Technol, 2017. 243: p. 393-399. 49. Carly, F., et al., Enhancing erythritol productivity in Yarrowia lipolytica using metabolic engineering. Metab Eng, 2017. 42: p. 19-24. 50. Janek, T., et al., Characterization of erythrose reductase from Yarrowia lipolytica and its influence on erythritol synthesis. Microb Cell Fact, 2017. 16(1): p. 118. 51. Rzechonek, D.A. and C. Neuveglise, EUF1 - a newly identified gene involved in erythritol utilization in Yarrowia lipolytica. 2017. 7(1): p. 12507. 52. Carly, F., et al., Identification and characterization of EYD1, encoding an erythritol dehydrogenase in Yarrowia lipolytica and its application to bioconvert erythritol into erythrulose. Bioresource Technology, 2018. 247: p. 963-969. 53. Zhang, Y., et al., Improved campesterol production in engineered Yarrowia lipolytica strains. Biotechnol Lett, 2017. 39(7): p. 1033-1039. 54. Cao, X., et al., Enhancing linalool production by engineering oleaginous yeast Yarrowia lipolytica. Bioresour Technol, 2017. 245(Pt B): p. 1641-1644. 55. Markham, K.A., et al., Rewiring Yarrowia lipolytica toward triacetic acid lactone for materials generation. 2018. 56. Kawai, S., W. Hashimoto, and K. Murata, Transformation of Saccharomyces cerevisiae and other fungi: Methods and possible underlying mechanism. Bioengineered Bugs, 2010. 1(6): p. 395-403. 57. Davidow, L.S. and J.R. Dezeeuw, Process for transformation of Yarrowia lipolytica. 1992, Google Patents. 58. Leplat, C., J.-M. Nicaud, and T. Rossignol, High-throughput transformation method for Yarrowia lipolytica mutant library screening. FEMS Yeast Research, 2015. 15(6): p. fov052-fov052. 59. Howard Glenn Damude, P.J.G., Daniel Joseph Macool, Stephen K. Picataggio, Dana M. Walters Pollak, James John Ragghianti, Zhixiong Xue, Narendra S. Yadav, Hongxiang Zhang, Quinn Qun Zhu High eicosapentaenoic acid producing strains of Yarrowia lipolytica. 2006, E I du Pont de Nemours and Co US. 60. Madzak, C., et al., Functional analysis of upstream regulating regions from the Yarrowia lipolytica XPR2 promoter. Microbiology, 1999. 145(1): p. 75-87. 61. Madzak, C., B. Treton, and S. Blanchin-Roland, Strong hybrid promoters and integrative expression/secretion vectors for quasi-constitutive expression of heterologous proteins in the yeast Yarrowia lipolytica. J Mol Microbiol Biotechnol, 2000. 2(2): p. 207-16. 62. Blazeck, J., et al., Tuning gene expression in Yarrowia lipolytica by a hybrid promoter approach. Appl Environ Microbiol, 2011. 77(22): p. 7905-14. 63. Blazeck, J., et al., Generalizing a hybrid synthetic promoter approach in Yarrowia lipolytica. Appl Microbiol Biotechnol, 2013. 97(7): p. 3037-52.

122 64. Shabbir Hussain, M., et al., Engineering Promoter Architecture in Oleaginous Yeast Yarrowia lipolytica. ACS Synthetic Biology, 2016. 5: p. 213-223. 65. Shabbir Hussain, M., I. Wheeldon, and M.A. Blenner, A Strong Hybrid Fatty Acid Inducible Transcriptional Sensor Built From Yarrowia lipolytica Upstream Activating and Regulatory Sequences. Biotechnology Journal, 2017. 12: p. 1-11. 66. Sumita, T., et al., YlALK1 encoding the cytochrome P450ALK1 in Yarrowia lipolytica is transcriptionally induced by n-alkane through two distinct cis-elements on its promoter. Biochemical and Biophysical Research Communications, 2002. 294: p. 1071-1078. 67. Trassaert, M., et al., New inducible promoter for gene expression and synthetic biology in Yarrowia lipolytica. Microbial Cell Factories, 2017. 16(1): p. 141. 68. Ito, Y., et al., Characterization of five terminator regions that increase the protein yield of a transgene in Saccharomyces cerevisiae. Journal of Biotechnology, 2013. 168(4): p. 486-492. 69. Geisberg, J.V., et al., Global analysis of mRNA isoform half-lives reveals stabilizing and destabilizing elements in yeast. Cell, 2014. 156(4): p. 812-24. 70. Curran, K.A., et al., Short Synthetic Terminators for Improved Heterologous Gene Expression in Yeast. ACS Synth Biol, 2015. 4(7): p. 824-32. 71. Raschmanová, H., et al., Implementing CRISPR-Cas technologies in conventional and non-conventional yeasts: Current state and future prospects. Biotechnology Advances, 2018. 72. Schwartz, C.M., et al., Synthetic RNA Polymerase III Promoters Facilitate High- Efficiency CRISPR-Cas9-Mediated Genome Editing in Yarrowia lipolytica. ACS Synth Biol, 2016. 5(4): p. 356-9. 73. Meijrink, B., et al., A crispr-cas system for a lipolytic yeast host cell. 2017, Google Patents. 74. Gao, S., et al., Multiplex gene editing of the Yarrowia lipolytica genome using the CRISPR-Cas9 system. Journal of Industrial Microbiology & Biotechnology, 2016. 43(8): p. 1085-1093. 75. Morse, N.J., et al., T7 Polymerase Expression of Guide RNAs in vivo Allows Exportable CRISPR-Cas9 Editing in Multiple Yeast Hosts. ACS Synthetic Biology, 2018. 7(4): p. 1075-1084. 76. Schwartz, C., et al., Standardized Markerless Gene Integration for Pathway Engineering in Yarrowia lipolytica. ACS Synthetic Biology, 2017. 6(3): p. 402- 409. 77. Schwartz, C., et al., CRISPRi repression of nonhomologous end-joining for enhanced genome engineering via homologous recombination in Yarrowia lipolytica. Biotechnol Bioeng, 2017. 114(12): p. 2896-2906. 78. Holkenbrink, C., et al., EasyCloneYALI: CRISPR/Cas9-Based Synthetic Toolbox for Engineering of the Yeast Yarrowia lipolytica. Biotechnol J, 2018. 79. Wong, L., et al., YaliBricks, a versatile genetic toolkit for streamlined and rapid pathway engineering in Yarrowia lipolytica. Metabolic Engineering Communications, 2017. 5: p. 68-77. 123 80. Gao, S., et al., One-step integration of multiple genes into the oleaginous yeast Yarrowia lipolytica. Biotechnol Lett, 2014. 36(12): p. 2523-8. 81. Celinska, E., et al., Golden Gate Assembly system dedicated to complex pathway manipulation in Yarrowia lipolytica. Microb Biotechnol, 2017. 10(2): p. 450-455. 82. James M Wagner, E.V.W., Hal S Alper, Developing a piggyBac transposon system and compatible selection markers for insertional mutagenesis and genome engineering in Yarrowia lipolytica. Biotechnology Journal, 2018. 83. Kerkhoven, E.J., et al., Regulation of amino-acid metabolism controls flux to lipid accumulation in Yarrowia lipolytica. 2016. 2: p. 16005. 84. Trebulle, P. and J.M. Nicaud, Inference and interrogation of a coregulatory network in the context of lipid accumulation in Yarrowia lipolytica. 2017. 3: p. 21. 85. Robles-Rodriguez, C.E., et al., Dynamic metabolic modeling of lipid accumulation and citric acid production by Yarrowia lipolytica. Computers & Chemical Engineering, 2017. 100: p. 139-152. 86. Zhao, C., et al., Metabolomic changes and metabolic responses to expression of heterologous biosynthetic genes for lycopene production in Yarrowia lipolytica. J Biotechnol, 2017. 251: p. 174-185. 87. Hill, J., et al., Environmental, economic, and energetic costs and benefits of biodiesel and ethanol biofuels. Proceedings of the National Academy of Sciences, 2006. 103(30): p. 11206. 88. Dias, M.O.S., et al., Integrated versus stand-alone second generation ethanol production from sugarcane bagasse and trash. Bioresource Technology, 2012. 103(1): p. 152-161. 89. Li, Q., W. Du, and D. Liu, Perspectives of microbial oils for biodiesel production. Applied Microbiology and Biotechnology, 2008. 80(5): p. 749-756. 90. Peralta-Yahya, P.P., et al., Microbial engineering for the production of advanced biofuels. Nature, 2012. 488(7411): p. 320-8. 91. Milla, S., et al., Macroalgae in biofuel production. Phycological Research, 2015. 63(1): p. 1-18. 92. Sherkhanov, S., et al., Production of FAME biodiesel in E. coli by direct methylation with an insect enzyme. Sci Rep, 2016. 6: p. 24239. 93. Beopoulos, A., et al., Control of lipid accumulation in the yeast Yarrowia lipolytica. Appl Environ Microbiol, 2008. 74(24): p. 7779-89. 94. Langlois, A. and O. Lebel, To Cyclopropanate or Not To Cyclopropanate? A Look at the Effect of Cyclopropanation on the Performance of Biofuels. Vol. 24. 2010. 95. Kai, Y. and E.H. Pryde, Production of branched‐chain fatty acids from sterculia oil. Journal of the American Oil Chemists' Society, 1982. 59(7): p. 300-305. 96. Schmid Katherine, M., Cyclopropane Fatty Acid Expression In Plants. 1999, SCHMID; KATHERINE M.: US. 97. Murphy, D.J., Structure, function and biogenesis of storage lipid bodies and oleosins in plants. Progress in Lipid Research, 1993. 32(3): p. 247-280.

124 98. Poger, D. and A.E. Mark, A Ring to Rule Them All: The Effect of Cyclopropane Fatty Acids on the Fluidity of Lipid Bilayers. The Journal of Physical Chemistry B, 2015. 119(17): p. 5487-5495. 99. Grogan, D.W. and J.E. Cronan, Jr., Cloning and manipulation of the Escherichia coli cyclopropane fatty acid synthase gene: physiological aspects of enzyme overproduction. J Bacteriol, 1984. 158(1): p. 286-95. 100. Chang, Y.Y. and J.E. Cronan, Jr., Membrane cyclopropane fatty acid content is a major factor in acid resistance of Escherichia coli. Mol Microbiol, 1999. 33(2): p. 249-59. 101. Brown, J.L., et al., Acid habituation of Escherichia coli and the potential role of cyclopropane fatty acids in low pH tolerance. Int J Food Microbiol, 1997. 37(2-3): p. 163-73. 102. Shabala, L. and T. Ross, Cyclopropane fatty acids improve Escherichia coli survival in acidified minimal media by reducing membrane permeability to H+ and enhanced ability to extrude H+. Research in Microbiology, 2008. 159(6): p. 458- 461. 103. Keweloh, H., R. Diefenbach, and H.-J. Rehm, Increase of phenol tolerance of Escherichia coli by alterations of the fatty acid composition of the membrane lipids. Archives of Microbiology, 1991. 157(1): p. 49-53. 104. Grogan, D.W. and J.E. Cronan, Cyclopropane ring formation in membrane lipids of bacteria. Microbiology and Molecular Biology Reviews, 1997. 61(4): p. 429- 441. 105. Munoz-Rojas, J., et al., Involvement of cyclopropane fatty acids in the response of Pseudomonas putida KT2440 to freeze-drying. Appl Environ Microbiol, 2006. 72(1): p. 472-7. 106. Wang, A.Y., D.W. Grogan, and J.E. Cronan, Jr., Cyclopropane fatty acid synthase of Escherichia coli: deduced amino acid sequence, purification, and studies of the enzyme active site. Biochemistry, 1992. 31(45): p. 11020-8. 107. Mizoguchi, T., et al., Cyclopropane-ring formation in the acyl groups of chlorosome glycolipids is crucial for acid resistance of green bacterial antenna systems. Bioorg Med Chem, 2013. 21(13): p. 3689-94. 108. Jungkind, D.L. and R.C. Wood, Physiological differences between cyclopropane fatty acid-deficient mutants and the parent strain of Streptococcus faecalis. Biochim Biophys Acta, 1974. 337(2): p. 298-310. 109. Brian, B.L. and E.W. Gardner, A simple procedure for detecting the presence of cyclopropane fatty acids in bacterial lipids. Appl Microbiol, 1968. 16(4): p. 549- 52. 110. Yu, X.H., R. Rawat, and J. Shanklin, Characterization and analysis of the cotton cyclopropane fatty acid synthase family and their contribution to cyclopropane fatty acid synthesis. BMC Plant Biol, 2011. 11: p. 97. 111. Gaydou, E.M., A. Ralaimanarivo, and J.P. Bianchini, Cyclopropanoic fatty acids of litchi (Litchi chinensis) seed oil. A reinvestigation. Journal of Agricultural and Food Chemistry, 1993. 41(6): p. 886-890. 125 112. Markham, K.A., Alper, H.S., Synthetic biology expands the industrial potential of Yarrowia lipolytica Trends in Biotechnology, 2018. 113. Blazeck, J., et al., Harnessing Yarrowia lipolytica lipogenesis to create a platform for lipid and biofuel production. Nat Commun, 2014. 5. 114. Adrio, J.L., Oleaginous yeasts: Promising platforms for the production of oleochemicals and biofuels. Biotechnol Bioeng, 2017. 114(9): p. 1915-1920. 115. Lima Da Silva, N.d., et al., Optimization of biodiesel production from castor oil. Applied Biochemistry and Biotechnology, 2006. 130(1): p. 405-414. 116. Beopoulos, A., et al., Metabolic engineering for ricinoleic acid production in the oleaginous yeast Yarrowia lipolytica. Appl Microbiol Biotechnol, 2014. 98(1): p. 251-62. 117. Liu, L., et al., Surveying the lipogenesis landscape in Yarrowia lipolytica through understanding the function of a Mga2p regulatory protein mutant. Metab Eng, 2015. 31: p. 102-11. 118. Machida, S., Y. Shiraiwa, and I. Suzuki, Construction of a cyanobacterium synthesizing cyclopropane fatty acids. Biochim Biophys Acta, 2016. 1861(9 Pt A): p. 980-987. 119. Shanks, B.H. and P.L. Keeling, Bioprivileged molecules: creating value from biomass. Green Chem., 2017. 19: p. 3177-3185. 120. Austin, M.B. and J.P. Noel, The chalcone synthase superfamily of type III polyketide synthases. Natural product reports, 2003. 20: p. 79-110. 121. Yu, D., et al., Type III polyketide synthases in natural product biosynthesis. IUBMB life, 2012. 64: p. 285-295. 122. Lim, Y.P., M.K. Go, and W.S. Yew, Exploiting the Biosynthetic Potential of Type III Polyketide Synthases. Molecules, 2016. 21: p. 1-37. 123. Eckermann, S., et al., New pathway to polyketides in plants. Nature, 1998. 396: p. 387-390. 124. Yrjönen, T., et al., Application of centrifugal force to the extraction and separation of parasorboside and gerberin from Gerbera hybrida. Phytochemical Analysis, 2002. 13: p. 349-353. 125. Tang, S.-Y., et al., Screening for enhanced triacetic acid lactone (TAL) production by recombinant Escherichia coli expressing a designed TAL reporter. Journal of the American Chemical Society, 2013. 135: p. 10099-10103. 126. Cardenas, J. and N.A. Da Silva, Metabolic engineering of Saccharomyces cerevisiae for the production of triacetic acid lactone. Metabolic engineering, 2014. 25: p. 194-203. 127. Saunders, L.P., et al., Triacetic acid lactone production in industrial Saccharomyces yeast strains. Journal of industrial microbiology & biotechnology, 2015. 42: p. 711-721. 128. Cardenas, J. and N.A. Da Silva, Engineering cofactor and transport mechanisms in Saccharomyces cerevisiae for enhanced acetyl-CoA and polyketide biosynthesis. Metabolic engineering, 2016. 36: p. 80-89.

126 129. Keatinge-Clay, A.T., Stereocontrol within polyketide assembly lines. Natural product reports, 2016. 33: p. 141-149. 130. Plc, S.I. Product FAQS, sorbic acid. [Web Page] 2011; Available from: http://www.sorbicinternational.com/media/product-faqs.php. 131. Schwartz, T.J., et al., Engineering catalyst microenvironments for metal-catalyzed hydrogenation of biologically derived platform chemicals. Angewandte Chemie - International Edition, 2014. 53: p. 12718-12722. 132. Kraus, G.A., K. Basemann, and T. Guney, Selective pyrone functionalization: Reductive alkylation of triacetic acid lactone. Tetrahedron Letters, 2015. 56: p. 3494-3496. 133. Kraus, G.A., U.K. Wanninayake, and J. Bottoms, Triacetic acid lactone as a common intermediate for the synthesis of 4-hydroxy-2-pyridones and 4-amino-2- pyrones. Tetrahedron Letters, 2016. 57: p. 1293-1295. 134. Xie, D., et al., Microbial synthesis of triacetic acid lactone. Biotechnology and Bioengineering, 2006. 93: p. 727-736. 135. Madzak, C., Yarrowia lipolytica: recent achievements in heterologous protein expression and pathway engineering. Appl Microbiol Biotechnol, 2015. 99(11): p. 4559-77. 136. Holdsworth, J.E., M. Veenhuis, and C. Ratledge, Enzyme activities in oleaginous yeasts accumulating and utilizing exogenous or endogenous lipids. Journal of general microbiology, 1988. 134: p. 2907-2915. 137. Beopoulos, A., et al., Yarrowia lipolytica as a model for bio-oil production. Progress in lipid research, 2009. 48: p. 375-387. 138. Nicaud, J.-M., Yarrowia lipolytica. Yeast, 2012. 29: p. 409-418. 139. Tai, M. and G. Stephanopoulos, Engineering the push and pull of lipid biosynthesis in oleaginous yeast Yarrowia lipolytica for biofuel production. Metabolic Engineering, 2013. 15: p. 1-9. 140. Yuzbasheva, E.Y., et al., A metabolic engineering strategy for producing free fatty acids by the Yarrowia lipolytica yeast based on impairment of glycerol metabolism. Biotechnology and Bioengineering, 2017: p. 1-11. 141. Holz, M., et al., Overexpression of alpha-ketoglutarate dehydrogenase in Yarrowia lipolytica and its effect on production of organic acids. Applied Microbiology and Biotechnology, 2011. 89: p. 1519-1526. 142. Kujau, M., Weber, Herbert, Barth, Gerold, Characterization of Mutants of the Yeast Yarrowia lipolytica Defective in Acetyl-Coenzyme A Synthetase. Yeast, 1992. 8: p. 193-203. 143. Flores, C.L., et al., Carbohydrate and energy- yielding metabolism in nonconventional yeasts. FEMS Microbiol Rev, 2000. 24: p. 517-529. 144. Sherman, D., et al., Génolevures: comparative genomics and molecular evolution of hemiascomycetous yeasts. Nucleic acids research, 2004. 32: p. D315--D318. 145. Luo, X., K.J. Affeldt, and N.P. Keller, Characterization of the Far Transcription Factor Family in Aspergillus flavus. G3 (Bethesda), 2016. 6(10): p. 3269-3281.

127 146. Cai, L., et al., Acetyl-CoA Induces Cell Growth and Proliferation by Promoting the Acetylation of Histones at Growth Genes. Molecular Cell, 2011. 42: p. 426-437. 147. Shi, L. and B.P. Tu, Acetyl-CoA induces transcription of the key G1 cyclin CLN3 to promote entry into the cell division cycle in Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences, 2013. 110: p. 7318-7323. 148. Burns, M.J., et al., 4-Hydroxy-6-alkyl-2-pyrones as nucleophilic coupling partners in Mitsunobu reactions and oxa-Michael additions. Beilstein J Org Chem, 2014. 10: p. 1159-65. 149. Hansen, C.A. and J.W. Frost, Deoxygenation of polyhydroxybenzenes: an alternative strategy for the benzene-free synthesis of aromatic chemicals. J Am Chem Soc, 2002. 124(21): p. 5926-7. 150. Guo, H., et al., Effects of pyruvate dehydrogenase subunits overexpression on the alpha-ketoglutarate production in Yarrowia lipolytica WSH-Z06. Appl Microbiol Biotechnol, 2014. 98(16): p. 7003-12. 151. Rywinska, A. and W. Rymowicz, High-yield production of citric acid by Yarrowia lipolytica on glycerol in repeated-batch bioreactors. J Ind Microbiol Biotechnol, 2010. 37(5): p. 431-5. 152. Tan, M.J., et al., Enhanced citric acid production by a yeast Yarrowia lipolytica over-expressing a pyruvate carboxylase gene. Bioprocess Biosyst Eng, 2016. 39(8): p. 1289-96. 153. Chen, D.C., J.M. Beckerich, and C. Gaillardin, One-step transformation of the dimorphic yeast Yarrowia lipolytica. Appl Microbiol Biotechnol, 1997. 48(2): p. 232-5. 154. Barth, G. and C. Gaillardin, Yarrowia lipolytica, in Nonconventional Yeasts in Biotechnology: A Handbook. 1996, Springer Berlin Heidelberg: Berlin, Heidelberg. p. 313-388. 155. Nicaud, J.-M., C. Gaillardin, and G. Pignede, Method For Non-homologous Transformation Of Yarrowia lipolytica. 2003, AGRONOMIQUE INST NAT RECH CENTRE NAT RECH SCIENT: US. 156. Irie, T., et al., Efficient transformation of filamentous fungus Pleurotus ostreatus using single-strand carrier DNA. Applied Microbiology and Biotechnology, 2001. 55(5): p. 563-565. 157. Wang, J.H., W. Hung, and S.H. Tsai, High efficiency transformation by electroporation of Yarrowia lipolytica. J Microbiol, 2011. 49(3): p. 469-72. 158. Benatuil, L., et al., An improved yeast transformation method for the generation of very large human antibody libraries. Protein Eng Des Sel, 2010. 23(4): p. 155-9. 159. Liu, H., et al., Fast and efficient genetic transformation of oleaginous yeast Rhodosporidium toruloides by using electroporation. FEMS Yeast Res, 2017. 17(2). 160. Dower, W.J., J.F. Miller, and C.W. Ragsdale, High efficiency transformation of E.coli by high voltage electroporation. Nucleic Acids Research, 1988. 16(13): p. 6127-6145.

128 161. Schenk, S. and R.A. Laddaga, Improved method for electroporation of Staphylococcus aureus. FEMS Microbiol Lett, 1992. 73(1-2): p. 133-8. 162. Guo, Z. and F. Sherman, Signals sufficient for 3'-end formation of yeast mRNA. Molecular and Cellular Biology, 1996. 16(6): p. 2772-2776. 163. Fitzpatrick, D.A., et al., A fungal phylogeny based on 42 complete genomes derived from supertree and combined gene analysis. BMC Evolutionary Biology, 2006. 6(1): p. 99. 164. Neidleman, S.L., Effects of temperature on lipid unsaturation. Biotechnol Genet Eng Rev, 1987. 5: p. 245-68. 165. Renaud, S.M., et al., Effect of temperature on growth, chemical composition and fatty acid composition of tropical Australian microalgae grown in batch cultures. Aquaculture, 2002. 211(1): p. 195-214. 166. Beopoulos, A., T. Chardot, and J.-M. Nicaud, Yarrowia lipolytica: A model and a tool to understand the mechanisms implicated in lipid accumulation. Biochimie, 2009. 91(6): p. 692-696. 167. K., C.J., et al., Identifying and retargeting transcriptional hot spots in the human genome. Biotechnology Journal, 2016. 11(8): p. 1100-1109. 168. Claudine, H., et al., NPI1, an essential yeast gene involved in induced degradation of Gap1 and Fur4 permeases, encodes the Rsp5 ubiquitin—protein ligase. Molecular Microbiology, 1995. 18(1): p. 77-87. 169. Folch, J., M. Lees, and G.H. Sloane Stanley, A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem, 1957. 226(1): p. 497- 509. 170. Horton, R.M., et al., Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension. Gene, 1989. 77(1): p. 61-8. 171. Blazeck, J., et al., Heterologous production of pentane in the oleaginous yeast Yarrowia lipolytica. J Biotechnol, 2013. 165(3-4): p. 184-94. 172. Ghosalkar, A., V. Sahai, and A. Srivastava, Optimization of chemically defined medium for recombinant Pichia pastoris for biomass production. Bioresour Technol, 2008. 99(16): p. 7906-10. 173. Schneiter, R. and G. Daum, Extraction of yeast lipids. Methods Mol Biol, 2006. 313: p. 41-5. 174. Zhu, A., R. Romero, and H.R. Petty, An enzymatic fluorimetric assay for glucose- 6-phosphate: application in an in vitro Warburg-like effect. Anal Biochem, 2009. 388(1): p. 97-101. 175. Naito, Y., et al., CRISPRdirect: software for designing CRISPR/Cas guide RNA with reduced off-target sites. Bioinformatics, 2015. 31(7): p. 1120-3. 176. Rodriguez, C.G., et al., Ring-Opening Polymerization of Epoxides: Facile Pathway to Functional Polyethers via a Versatile Organoaluminum Initiator. Macromolecules, 2017. 50(8): p. 3121-3130.

129