저작자표시-비영리-변경금지 2.0 대한민국

이용자는 아래의 조건을 따르는 경우에 한하여 자유롭게

l 이 저작물을 복제, 배포, 전송, 전시, 공연 및 방송할 수 있습니다.

다음과 같은 조건을 따라야 합니다:

l 귀하는, 이 저작물의 재이용이나 배포의 경우, 이 저작물에 적용된 이용허락조건을 명확하게 나타내어야 합니다.

l 저작권자로부터 별도의 허가를 받으면 이러한 조건들은 적용되지 않습니다.

저작권법에 따른 이용자의 권리는 위의 내용에 의하여 영향을 받지 않습니다.

이것은 이용허락규약(Legal Code)을 이해하기 쉽게 요약한 것입니다.

Disclaimer

저작자표시. 귀하는 원저작자를 표시하여야 합니다.

비영리. 귀하는 이 저작물을 영리 목적으로 이용할 수 없습니다.

변경금지. 귀하는 이 저작물을 개작, 변형 또는 가공할 수 없습니다.

http://creativecommons.org/licenses/by-nc-nd/2.0/kr/legalcodehttp://creativecommons.org/licenses/by-nc-nd/2.0/kr/

1

Doctoral Thesis

Engineering Escherichia coli

to Increase the Production of Free Fatty Acids

Kwang Soo Shin

Department of Biomedical Engineering

Graduate School of UNIST

2018

2

Engineering Escherichia coli

to Increase the Production of Free Fatty Acids

Kwang Soo Shin

Department of Biomedical Engineering

Graduate School of UNIST

3

Engineering Escherichia coli

to Increase the Production of Free Fatty Acids

A thesis/dissertation

submitted to the Graduate School of UNIST

in partial fulfillment of the

requirements for the degree of

Doctor of Philosophy

Kwang Soo Shin

01. 03. 2018

4

Engineering Escherichia coli

to Increase the Production of Free Fatty Acids

Kwang Soo Shin

This certifies that the thesis/dissertation of Kwang Soo Shin is

approved.

01. 03. 2018

5

Engineering Escherichia coli

to increase the production of free fatty acids

Abstract

Microbial production of free fatty acids (FFAs) and their derivatives from renewable plant-

derived biomass is considered a promising approach for replacement of petroleum-based production

of fuels and chemicals. Microbial production has several advantages, such as sustainable and cost-

effective production, and direct synthesis of chemicals from renewable. For these reasons, many

studies have attempted to redesign microbes through metabolic engineering and synthetic biology for

increased production of FFA.

One of efficient strategies is to push/pull and block (push: activating first or rate-limiting reaction

in a metabolic pathway, pull: activating final reaction, block: inactivating product-degrading).

Previously, the genes encoding acetyl-CoA carboxylase and thioesterase were overexpressed to push

and pull the carbon flux toward FFA synthesis. Furthermore, genes in the β-oxidation pathway were

also deleted to prevent the degradation of FFA. The combined strategy that employed push/pull and

block increased FFA production, but the strain exhibited around 0.048 g of FFA/g of glucose, which is

equivalent to 14% of the theoretical yield, indicating that further manipulations are required to

enhance FFA production.

In this thesis, several metabolic strategies were used to construct engineered E. coli based on

push/pull and block strategy. First, acetyl-CoA carboxylation bypass was introduced to avoid

regulation in the carboxylation. The bypass also provided an alternative route for malonyl-CoA

synthesis, altered precursor specificity, and redirected carbon flux from the TCA cycle to FFA

synthesis, resulting in 2.5-fold increase in FFA production. Second, mutant thioesterase that is more

catalytically active was constructed. The change of amino acid residue in the non-conserved region

increased catalytic activity up to two-fold. Employing the mutant thioesterase overcame limited FFA

production driven by protein-protein interaction in the fatty acid synthetic pathway, resulting in two-

fold higher FFA production than employing wild-type thioesterase. Finally, identifying and

engineering membrane transport system, indirectly involved in increased FFA production, improved

both extracellular and total FFA production. Further engineering including overexpression of

transcriptional regulator FadR and phosphoenolpyruvate carboxylase (PPC) increased FFA production.

Combinatorial manipulation of the strategies increased FFA production yield (0.24 g of FFA/g of

glucose, corresponding to 69% of theoretical yield) which is 5-fold higher than that of the benchmark

strain.

6

The works presented in this thesis provide viable strategies for improving production of the FFA,

precursor for value-added chemicals such as biodiesel, consumer product (cosmetics, shampoo), and

industrial product (lubricants, surface coating). While success was attained in an increase in FFA

production titer and yield, the optimized FFA-producing strain could be constructed through further

engineering such as increasing NADPH pool, developing active expression system in stationary phase,

and reprogramming metabolic pathway.

7

8

Table of Contents

Chapter 1. Introduction-overview of fatty acid metabolism

1.1 Motivation ………………………………………………………………………………………..1

1.2 Fatty acid biosynthesis pathway in Escherichia coli ……………………………………...……..1

1.3 Successful engineering to increase fatty acid production ………………...…………….………..3

1.4 Objectives ……………………………………………………………………….………………..4

Chapter 2. Introduction of an acetyl-CoA carboxylation bypass

2.1 Abstract …………………………………………………………………………………………..5

2.2 Introduction ……………………………………………………………………………..………..6

2.3 Materials and methods

2.3.1 Bacterial strains and plasmids …………………………………………………….………..7

2.3.2 Media and cultivation conditions ……………………………………………...……….....10

2.3.3 Metabolite quantification …………………………………………….………...…………10

2.4 Results and Discussion

2.4.1 MMC-overexpressing E. coli showed improved FFA production ……………….....…….11

2.4.2 Overexpression of PPC together with MMC further enhanced FFA production …………14

2.4.3 Redirecting TCA cycle intermediates into fatty acid synthesis increased FFA

production ………………………………………………………………………………...17

2.5 Conclusions ……………………………………………………………...………...……………19

Chapter 3. Construction of a mutant Acyl-CoA thioesterase I with high specific activity

3.1 Abstract …………………………………………………………………………………………20

9

3.2 Introduction …………………………………………………………………………………..…21

3.3 Methods

3.3.1 Bacterial strains and plasmids ………………………………………………...…………..22

3.3.2 Media and cultivation conditions ………………………………………...……………….25

3.3.3 Mutant library construction ………………………………………………………...……..25

3.3.4 Enrichment and isolation of ‘TesA mutants with increased activity …………...………....26

3.3.5 Analysis of FFA and glucose …………...…………………………..………………...…..26

3.3.6 Enzyme expression and kinetic analysis …………...…………………………...…….…..27

3.4 Results and discussion

3.4.1 Construction and characterization of a high-throughput screening system …………..…..28

3.4.2 Genotypic and phenotypic analysis of isolated ‘TesA mutants ………………..…...……..31

3.4.3 The effect of mutation at Arg64 of ‘TesA on enzymatic activity …………………..……..33

3.4.4 Increased FFA production driven by high specific activity of ‘TesAR64C ...........................35

3.5 Conclusions …………...………………..…………………………………………………...…..37

3.6. Abbreviations used …………...……………………………..………………………....…...…..37

Chapter 4. Disruption of membrane transport systems

4.1 Abstract …………...……………………………..…………………………...…………......…..38

4.2 Introduction …………...……………………………..…………………………….…...…...…..39

4.3 Materials and methods

4.3.1 Bacterial strains and plasmids …………...…………………………..….………..…...…..42

4.3.2 Mutant library construction …………...………………………….……….……...…...…..42

4.3.3 Screening of the mutant library and identification of insertion sites ………….....…...…..43

4.3.4 Media and cultivation conditions …………...........................................................…...…..43

10

4.3.5 Free fatty acid measurement and statistical analysis …………....................................…..44

4.4 Results

4.4.1 Screening and identification of high FFA-producing mutants …………...............…...…..44

4.4.2 Improvement of FFA production by envR deletion …………...………………........….....46

4.4.3 Enhanced FFA production by multiple gene disruption …………...……………….....…..49

4.5 Discussion …………...………………………………...…..…………………………...…...…..51

4.6 Conclusion …………...……………………………..……………………..…………...…...…..54

4.7 Abbreviations used …………...…………………………………………………………......…..54

Chapter 5. Construction of high FFA-producing strain with combined manipulation

5.1 Abstract …………...……………………………..…………...………………………...…...…..55

5.2 Introduction …………...……………………………..……………….………………...…...…..56

5.3 Methods

5.3.1 Bacterial strains and plasmids …………...…………………………………….....…...…..56

5.3.2 Media, cultivation conditions, and metabolite quantification ………….......................…..58

5.3.3 Fed-batch fermentation …………...……………………………..…...……...…...…...…..58

5.4 Results and discussion

5.4.1 Combined manipulation of the strategies to increase FFA production ………….........…..59

5.5 Conclusions …………...……………………………..………….……………………...…...…..63

Chapter 6. Summary and future perspectives

6.1 Summary of the findings …………...…………………………………………………..…...…..64

6.2 Recommendations for future work

11

6.2.1 Increasing NADPH pool by controlling carbon flux between EMP and PPP ……......…..65

6.2.2 Developing active expression system in stationary phase …………............................…..66

6.2.3 Reprogramming metabolic pathway to develop microbial cell factories for

efficient FFA production ………..…………………. …………..............................……...67

7. References ………………………………………...…………. …………...............................…..69

8. Acknowledgements …………………………. …………........................................................…..83

12

List of figures

1.1 Metabolic pathway of fatty acid metabolism in E. coli …………………………………………2

2.1 Metabolic pathway of free fatty acid synthesis in E. coli via acetyl-CoA carboxylase (ACC) and

methylmalonyl-CoA carboxyltransferase (MMC) ……………………………………………......7

2.2 FFA production, cell growth, and acetate accumulation in E. coli strains overexpressing ACC or

MMC ……………………..…………………………………………………………………….12

2.3 Effect of the expression level of ACC or MMC on FFA production ……………………….…..13

2.4 Free fatty acid production with overexpression of PPC …………………………………..……15

2.5 Effect of expression of phosphoenolpyruvate carboxylase (PPC) or addition of aspartic acid on

cell growth and glucose consumption of engineered strains ……………………………...…....16

2.6 Free fatty acid production levels of engineered strains overexpressing MaeB or supplemented

with aspartic acid …………………………………………………………………….………...18

2.7 Fold change in FFA production of engineered strains …………………………………..……...19

3.1 A schematic diagram of fatty acid synthesis in E. coli ………………………………...……….22

3.2 Construction of the FFA-sensing plasmid and its performance …………………………..…….29

3.3 Response of the FAB biosensor to endogenous FFA ………………………..………………….30

3.4 Free fatty acid production of selected strains from the ‘TesA mutant library ………………......32

3.5 Free fatty acid production from ‘TesA mutants at position 64 generated by site-directed

mutagenesis …………………………………………..………………………………………...33

3.6 Batch culture of the SBF06 and SBF08 in mini-bioreactor ………………………...…………..34

3.7 Kinetic analysis of the ‘TesA and ‘TesAR64C ……………………………………………..……..35

3.8 Free fatty acid production and protein expression levels at various concentrations of IPTG ….36

4.1 Free fatty acid production by transposon mutants …………………………………………..….45

4.2 Cell growth of engineered mutants ……………………………………...……………………...46

4.3 Effect of gene deletion on response to exogenous fatty acids in strains with oleic acid ……….47

13

4.4 Comparison of FFA production in strains SBF06 and SBF22 ………………………..………...48

4.5 Effect of deletion of additional genes on FFA production ………………………..…………….49

4.6 Time course of FFA production by strains SBF06, SBF25, and SBF37 …………….………….50

5.1 Batch culture of engineered strains at 72 h post-induction …………………………..…………60

5.2 Fed-batch fermentation of SBF43 strain ………………………………………………………..61

5.3 Schematic diagram of developed strain in this study …………………………………..……….61

14

List of tables

2.1 Strains and plasmids used in this study ………………………………………………………….8

2.2 Primers used in this study ………………………………………………………………………..9

2.3 Results of batch fermentations with engineered strains ………………………………………...14

3.1 Strains and plasmids used in this study …………………………………………………………23

3.2 Primers used in this study ………………………………………………………………………24

4.1 Strains and plasmids used in this study …………………………………………………………40

4.2 Primers used in this study ………………………………………………………………………41

4.3 Results of batch fermentations with engineered strains ………………………………………...53

5.1 Strain and plasmids used in this study ………………………………………………………….57

5.2 Literature summary of FFA production ………………………………………………………...62

1

Chapter 1

Introduction-overview of fatty acid production

1.1. Motivation

Current world economy has been supported by oil refinery that refines petroleum into useful

products. The easy availability of the petroleum and its derivatives have made humanity highly

dependent on the petroleum. However, limited reserves, increased demands, and environmental

problems such as greenhouse effect have raised development of sustainable and eco-friendly

production process for manufacturing the fuels and value-added chemicals.

As substitutes of petrochemicals, oleochemicals are widely used in lots of applications

including biodiesel, lubricants, and bioplastics. Generally, the oleochemicals have been synthesized

from plant oils and animal fats, alternative resources instead of petroleum. However, concern about

the sustainability and environmental impact of production of the lipid sources has raised the bio-

refinery process that converts biomass to the oleochemicals [1, 2]. Microbial production of free fatty

acid (FFA) could provide a precious precursor for the production of oleochmicals. The microbial

production of the FFA has several advantages: sustainable and cost-effective production from

renewable biomass, low land-use, and feasible extraction process [3]. These benefits of microbial

process have raised a lot of studies aimed at producing and increasing FFA through metabolic

engineering and synthetic biology.

1.2. Fatty acid biosynthesis pathway in Escherichia coli

The mechanism of fatty acid biosynthesis (FAB) is well understood in E. coli since

numerous studies were reported and well summarized in several reviews [1, 3-6]. Prokaryotic

microorganism including E. coli harbors type II fatty acid synthase to synthesize cellular membrane.

The genes encoding type II fatty acid synthase are composed of discrete and monofunctional enzymes

to produce fatty acid. Thus, the intermediates of type II system could be accessible and converted into

wide range of fatty acid products. The FAB pathway employs iterative condensation reaction to

extend acyl chain from building blocks including acetyl-CoA. To synthesize one mole of fatty acid,

lots of enzymatic reactions and intracellular molecules are involved (Fig. 1.1).

2

Figure. 1.1. Metabolic pathway of fatty acid metabolism in E. coli. The genes highlighted by blue are involved in fatty

acid synthesis. The genes highlighted by green are involved in fatty acid degradation.

Initiation

The precious molecule for fatty acid synthesis is acetyl-CoA which could act as primer and

extender precursor in initiation and elongation step, respectively. Fatty acid synthesis could be

initiated by converting the acetyl-CoA into malonyl-CoA by action of acetyl-CoA carboxylase (ACC)

[7]. The malonyl-CoA should be subsequently transferred to acyl carrier protein (ACP) to participate

fatty acid initiation and elongation. The reaction is catalyzed by action of malonyl-CoA-ACP

transacylase (FabD). Next step for fatty acid synthesis is condensation reaction of the acetyl-CoA and

malonyl-ACP to synthesize acetoacetyl-ACP. Although three enzymes are currently existed for the

condensation, β-ketoacyl-ACP synthase III (FabH) is previously known as an unique enzyme that

catalyzes initial condensation [8]. However, further study reported that its deletion results in growth

reduction but not lethality, implying that other enzymes must be involved in the initial condensation

reaction [9].

Elongation

The β-ketoacyl-ACP, resulting molecules of condensation, is reduced to β-hydroxyacyl-ACP

by NADPH-dependent β-hydroxyacyl-ACP reductase (FabG). Kinetic analysis of FabG behavior

revealed that it has a wide range of β-ketoacyl-ACP species [10, 11]. Deletion of fabG in E. coli is

lethal to growth [12], indicating that it participates every elongation cycle during fatty acid synthesis.

There are two β-hydroxyacyl-ACP dehydratases that catalyze dehydration of their substrate to

3

synthesize enoyl-ACP. Two enzymes, FabZ and FabA, have broad substrate specificity with different

activity on chain length and saturation. The FabZ prefers short and long chain substrates, whereas the

FabA has high specificity toward medium chain substrates [13]. The enoyl-ACP reductase encoded by

fabI catalyzes reduction of double bond of enoyl-ACP to produce acyl-ACP [14]. The acyl-ACP could

be used as primer to extend its chain length. This reaction is catalyzed by β-ketoacyl-ACP synthase I

and II (FabB and FabF, respectively). The FabB has broad substrate specificity toward C2-ACP to

C14-ACP, but insufficient activity to longer substrates [15, 16]. Unique role of FabB is elongation of

cis-3-decenoyl-ACP, a product of FabA-mediated isomerization, to synthesize cis-5-dodecenoyl-ACP

and cis-7-tetradecenoyl-ACP [17].

Thioesterase

Thioesterase catalyze hydrolysis of thioester bond in the acyl-ACP or acyl-CoA to produce

free fatty acid and ACP or CoA. In E. coli, there are three thioesterase that convert the substrates into

free fatty acid. Among them, acyl-CoA thioesterase I (TesA, encoded by tesA) has been frequently

employed to increase the FFA production as its characteristics are well-studied. Catalytic mechanism,

substrate specificity, and crystal structure of the TesA were well-identified by several studies [18-22].

The TesA prefers long chain acyl-CoAs or acyl-ACPs (more than 12 carbon length) [23, 24] and is

located in periplasmic space [25]. Cytosolic form of TesA (‘TesA) by removing leader sequence

significantly increased fatty acid production as it relieves feedback inhibition resulted from

accumulation of the acyl-ACP [26]. A number of thioesterase have been identified from various

organisms and expressed in E. coli to produce a variety of products such as short, medium, or

unsaturated fatty acid [27-30]. Thus, utilization of thioesterase not only accelerates the FFA

production but also alters type of fatty acid produced.

1.3. Successful engineering to increase fatty acid production

It is well-known approach that increasing precursor availability by redirecting carbon flux

generally leads to increase in production of desired products. One of efficient methods to redirect

carbon flux is blocking or overexpressing a certain pathway. Several competing pathways at branch

point in pyruvate to acetyl-CoA were inactivated to provide high level of acetyl-CoA, leading to

improved titer of desired products [31, 32]. To investigate the effect of acetyl-CoA consuming

pathway in the FFA production, acetate formation pathway was also inactivated [33]. To be used as

elongation unit, the acetyl-CoA should be converted into malonyl-CoA through the ACC activity. As a

gatekeeping enzyme, expression and activity of ACC is tightly regulated, resulting in low

4

concentration of malonyl-CoA. This limitation has raised several studies aimed at increasing

intracellular concentration of malonyl-CoA through metabolic engineering. The overexpression of

ACC was shown to significantly increase fatty acid production by 6-fold over short period of time,

however, fold-increase in fatty acid production decreased as incubation time increase [34]. Another

important finding is detrimental effect of the overproduction of E. coli ACC on cell viability [34].

Thus, high activity of ACC is not an efficient method to improve the FFA production as further studies

also revealed that the overexpression of ACC lead to small increase in fatty acid production and the

production sharply reduced at excess activity of the ACC [35-37].

Preventing degradation is one of conventional approaches to enhance the production of

desired products. The inactivation of β-oxidation significantly improved fatty acid production in

several studies [35, 38]. However, contrary results were also reported that inactivation of β-oxidation

did not have a positive effect in the FFA production [33, 39]. The different achievements might be due

to that β-oxidation is subjected to carbon catabolite repression, resulted from different culture

condition such as glucose concentration [3].

1.4. Objectives

The main objective of this thesis is to engineer E. coli to improve the FFA production though

‘pull/push and block’ strategy. This study would also help in providing several engineering strategies

that are redesigning metabolic pathway, developing biosensor, constructing mutant library, and

identifying mutant candidates for improving the FFA production. A combination of the main strategies

also was carried out to enhance the FFA production.

5

Chapter 2

Introduction of an acetyl-CoA carboxylation bypass

2.1. Abstract

This study investigated the effect of the methylmalonyl-CoA carboxyltransferase (MMC) of

Propionibacterium freudenreichii on production of free fatty acid (FFA) in Escherichia coli.

Overexpression of the MMC exhibited a 44% increase in FFA titer. Co-overexpression of MMC and

phosphoenolpyruvate carboxylase (PPC), which supplies the MMC precursor, further improved the

titer by 40%. Expression of malic enzyme (MaeB) led to a 23% increase in FFA titer in the acetyl CoA

carboxylase (ACC)-overexpressing cells, but no increase in the MMC-overexpressing cells. The

highest FFA production in the MMC-overexpressing strain was achieved through the addition of

aspartic acid, which can be converted into oxaloacetate (OAA), resulting in a 120% increased titer

compared with that in the ACC-overexpressing strain. These findings demonstrate that MMC provides

an alternative pathway for malonyl-CoA synthesis and increases fatty acid production.

6

2.2. Introduction

In Escherichia coli, fatty acid synthesis involves a set of enzyme reactions beginning with

the carboxylation of acetyl-CoA to malonyl-CoA. This is an irreversible reaction catalyzed by the

acetyl-CoA carboxylase (ACC) complex (Fig. 2.1a). The expression of ACC genes is tightly regulated

at the transcriptional and translational levels [40-42], as well as at the level of enzyme activity [43,

44]. In order to enhance the production of FFAs and flavonoids, homologous or heterologous

overexpression of ACC genes has been extensively used for increasing intracellular malonyl-CoA

levels [32, 45-47]. However, this strategy has shown some limitations: ACC is feedback inhibited by

the presence of the acyl-acyl carrier protein (acyl-ACP), the product of the fatty acid synthetic

pathway [44]; the increased flux from central metabolism to malonyl-CoA through acetyl-CoA also

upregulates other metabolic pathways such as the tricarboxylic acid (TCA) cycle or acetate formation

[47]; and the overproduction of E. coli ACC has been shown to affect cell viability [32, 34, 35].

In order to avoid these rate-limiting steps, in some cases, a bypass strategy has been applied

for enhanced production of desired products. For example, the mevalonate-dependent pathway from

Saccharomyces cerevisiae was expressed in E. coli to bypass the host’s native regulatory system,

resulting in a 20-fold increase in terpenoid production [48]. Another bypass in the mevalonate

pathway led to high isopentenol production by reducing energy requirements, intrinsic regulation, and

intermediate-related toxicity [49]. Recently, the introduction of a pyruvate dehydrogenase bypass,

composed of pyruvate oxidase (encoded by poxB) and acetyl-CoA synthetase (encoded by acs),

avoided a complicated regulatory system involved in the expression and activity of pyruvate

dehydrogenase and resulted in a 1.3-fold increase in isopropanol production [50]. The expression of

malonyl-CoA synthetase that converts malonate into malonyl-CoA enhanced the production of

polyketide and streptavidin in a malonate-containing medium [51-53].

It has been reported that methylmalonyl-CoA carboxyltransferases (MMCs) from

Propionibacterium spp. catalyze the conversion of methylmalonyl-CoA and pyruvate into propionyl-

CoA and oxaloacetate (OAA) [54] as well as catalyzing the conversion of acetyl-CoA and

oxaloacetate into malonyl-CoA and pyruvate [55]. In this study, the MMC genes from

Propionibacterium freudenreichii were expressed in E. coli to bypass the complex system regulating

acetyl-CoA carboxylation (Fig. 2.1b). The introduction of the MMC bypass also redirected carbon

flux to FFA synthesis from the TCA cycle, reduced cell growth inhibition, and reduced acetate

formation. Finally, the engineered strain was capable of producing 2.5-fold more FFA than the control

strain upon addition of aspartic acid.

7

Figure 2.1. Metabolic pathway of free fatty acid synthesis in E. coli via (a) acetyl-CoA carboxylase (ACC) and (b)

methylmalonyl-CoA carboxyltransferase (MMC). ACC carboxylates acetyl-CoA to malonyl-CoA via the transfer of the

carboxyl group from carbon dioxide. MMC carboxylates acetyl-CoA to malonyl-CoA via the transfer of the carboxyl group

of OAA, which is converted into pyruvate. Dashed line indicates multiple enzymatic steps. Tables indicate the reactions

involved in each enzymatic pathway and the net reaction from phosphoenolpyruvate.

2.3. Materials and methods

2.3.1. Bacterial strains and plasmids

The strains and plasmids constructed in this study are listed in Table 2.1. E. coli MG1655

was used as a parental strain to construct all strains. Strain SBF06 overexpressing a plasmid-encoded

leaderless version of E. coli acyl-CoA thioesterase I (ʹTesA) was used as a control strain. P.

freudenreichii shermanii KCTC5753 was purchased from the Korean Collection for Type Cultures

(KCTC, Daejeon, Korea).

8

Table 2.1. Strains and plasmids used in this study

All plasmids used in this study were derived from BioBrick plasmids [57]. Briefly, the gene

encoding red fluorescent protein (RFP) in the pBbA2k-rfp vector was removed with BglII and XhoI.

The gene encoding E. coli accD was PCR-amplified from E. coli MG1655 genomic DNA with a

forward primer containing a BglII site and a reverse primer containing BamHI and XhoI sites and was

then cloned into the BglII–XhoI sites of pBbA2k. Full-length DNAs of accA, accB, and accC were

also PCR-amplified with a forward primer containing a BglII site and a reverse primer with an XhoI

site and were then cloned into the above construct after digestion by BamHI and XhoI. The

phosphoenolpyruvate carboxylase (ppc) and malic enzyme (maeB) genes of wild-type E. coli

MG1655 were cloned into the EcoRI–XhoI and BglII–XhoI sites of pBbS2a and pBbB2a, respectively,

to create pBbS2a-ppc and pBbB2a-maeB. To create pBbA2k-mmc, the P. freudenreichii MMC genes

were amplified individually with the primers listed in Table 2.2 and cloned into pBbA2k using the

Gibson assembly method [58]. Ribosome binding site (RBS) sequences were designed for each gene

using the 5ʹ-UTR designer [59] and were included in the forward primers.

Strains and plasmids Genotype and description Reference

Strains

MG1655 E. coli K-12 F–λ–ilvG–rfb-50rph-1 [56]

SBF06 MG1655 with pBbB6c-‘tesA This study

SBF14 SBF06 with pBbA2k-accABCD This study

SBF15 SBF06 with pBbA2k-mmc This study

SBF16 SBF06 with pBbS2a-ppc This study

SBF17 SBF14 with pBbS2a-ppc This study

SBF18 SBF15 with pBbS2a-ppc This study

SBF19 SBF06 with pBbB2a-maeB This study

SBF20 SBF14 with pBbB2a-maeB This study

SBF21 SBF15 with pBbB2a-maeB This study

Plasmids

pBbB6c-‘tesA pBbB6c-gfp with Δgfp::‘tesA, CmR This study

pBbA2k-rfp p15A ori, carrying Ptet promoter and rfp, KmR [57]

pBbB2a-gfp BBR1 ori, carrying Ptet promoter and gfp,

AmpR

[57]

pBbS2a-rfp pSC101 ori, carrying Ptet promoter and rfp,

AmpR

[57]

pBbA2k-accDABC pBbA2k-rfp with Δrfp::accD, accA, accB, and

accC, KmR

This study

pBbA2k-mmc pBbA2k-rfp with Δrfp::mmc, KmR This study

pBbB2a-maeB pBbB2a-gfp with Δgfp::maeB, AmpR This study

pBbS2a-ppc pBbS2a-rfp with Δrfp::ppc, AmpR This study

9

Table 2.2. Primers used in this study

Primers Sequence (5’-3’)

AccA_FP TCAAAAGATCTTTTAAGAAGGAGATATACATATGAGTCTGAATTTCCTTGA

TTTTG

AccA_RP TCCTTACTCGAGTTTGGATCC TTACGCGTAACCGTAGCTCATCAGG

AccB_FP TCAAAAGATCTTTTAAGAAGGAGATATACATATGGATATTCGTAAGATTAA

AAAAC

AccB_RP TCCTTACTCGAGTTTGGATCC TTACTCGATGACGACCAGCGGCTC

AccC_FP TCAAAAGATCTTTTAAGAAGGAGATATACATATGCTGGATAAAATTGTTATT

GCCA

AccC_RP TCCTTACTCGAGTTTGGATCC TTATTTTTCCTGAAGACCGAGTTTT

AccD_FP TCAAAAGATCTTTTAAGAAGGAGATATACATATGAGCTGGATTGAACGAAT

TAAAA

AccD_RP TCCTTACTCGAGTTTGGATCC TCAGGCCTCAGGTTCCTGATCCGGT

M18870_FP TCAAAAGATCTTCTGCTGAGGAGGCCACAATTTAAAATGAGTCCGCGAGA

AATTGAGGTTT

M18870_RP TTAGTGTACCCCCGCCTATAGGTAGTCACGCCTGCTGAACGGTGACTTCG

MmdA_FP CTACCTATAGGCGGGGGTACACTAAATGGCTGAAAACAACAATTTGAAGC

MmdA_RP CTTAAATATTTCCCCCCGCATTCAATCAGCAGGGGAAGTTTCCATGCTTC

HY_FP TTGAATGCGGGGGGAAATATTTAAGATGGCTGATGAGGAAGAGAAGGAC

C

HY_RP AGCGACCCTCCTACGACTTGTGTGTTCAACGAATGGAATGGTTCTGCAGA

BCCP_FP ACACACAAGTCGTAGGAGGGTCGCTATGAAACTGAAGGTAACAGTCAAC

G

BCCP_RP AGATCCTTACTCGAGTTTGGATCCTCAGCCGATCTTGATGAGACCCTGAC

pBbA2k_FP GGATCCAAACTCGAGTAAGGATCT

PBBA2K_R

P TTTAAATTGTGGCCTCCTCAGCAGAAGATCTTTTGAATTCTTTTCTCTAT

MaeB_FP GAAAAGAATTCAAAAGATCTTTTAAGAAGGAGATATACATATGGATGACC

AGTTAAAACAAAGTG

MaeB_RP CCTTACTCGAGTTTGGATCCTTACAGCGGTTGGGTTTGCGCTTCT

Ppc_FP TCAAAAGATCTTTTAAGAAGGAGATATACATATGAACGAACAATATTCCGC

ATTGC

Ppc_RP CCTTACTCGAGTTTGGATCC TTAGCCGGTATTACGCATACCTGCC

Note: Underlined sequences indicate restriction enzyme sites.

10

2.3.2. Media and cultivation conditions

Luria-Bertani (LB) medium and M9 minimal medium were used for routine bacterial culture

and fatty acid production, respectively. For fatty acid production, a single colony was picked from an

agar plate and incubated in 3 mL of LB medium at 37 °C at 200 rpm. When the medium was turbid

(around 10 h cultivation), the culture was 1:100 transferred to 5 mL of the M9 medium supplemented

with additional 100 mM sodium phosphate buffer (pH 7.0), 2% glucose (wt/vol), and trace elements

[60] and cultured overnight with appropriate antibiotics: ampicillin (100 µg/ml), kanamycin (50 µg/ml)

or chloramphenicol (30 µg/ml). The overnight culture was 1:50 inoculated into 25 mL of the M9

medium in a 250-mL shaking flask at 37 °C at 200 rpm. Expression of each enzyme (ʹTesA, ACC,

MMC, PPC, or MaeB) was induced by the addition of 0.2 mM IPTG for ʹTesA or 25 nM tetracycline

for the others at an OD600 of 1–1.5. Cells were incubated at 37 °C for 48 h post-induction with

vigorous shaking (200 rpm). Aspartic acid (665 mg/L) was added to the culture medium to supply a

high level of OAA. The OD600 was measured on a spectrophotometer (Libra S22).

2.3.3. Metabolite quantification

FFA concentration was measured as previously described [61]. Briefly, 500 µL of culture

broth was supplemented with 50 µL of 6 N HCl and 50 µL of 1 g/L methyl nonadecanoate (Sigma-

Aldrich) as an internal standard. FFA was extracted twice with 500 µL of ethyl acetate by vigorous

vortexing. After centrifugation, the organic layer was separated and supplemented with 100 µL of a

mixture of MeOH:6 N HCl (9:1, v/v) and 100 µL of trimethylsilyl-diazomethane (Sigma-Aldrich) in

order to methylate FFA into fatty acid methyl ester (FAME). Then, the FAME was analyzed using a

gas chromatograph (Agilent 7890A) equipped with a flame ionization detector (FID) and a DB-5

column (30 m 0.25 mm, Agilent Technologies). The FAME concentration was identified using

external standards composed of C10–22 FAMEs (Sigma-Aldrich).

The quantification of glucose and acetate was assessed as previously described [62]. Briefly,

the supernatant was separated and heated for 1 h prior to a second centrifugation at 16,100 g for 30

min at 25 °C. After appropriate dilution, the supernatant was injected into a Shimadzu high-

performance liquid chromatography (HPLC) station equipped with an SIL-20A auto-sampler

(Shimadzu) and a refractive index detector (Shimadzu) or a UV detector (Shimadzu) for

quantification of glucose or acetate, respectively. HPLC separation was performed using an HPX-87P

column (Bio-Rad) at 0.6 mL/min with HPLC-grade water for glucose quantification and an HPX-87X

column (Bio-Rad) at 0.5 mL/min with 5 mM H2SO4 for acetate quantification.

11

2.4. Results and Discussion

2.4.1. MMC-overexpressing E. coli showed improved FFA production

When either the ACC or MMC complex was overexpressed in SBF06, E. coli overexpressing

thioesterase (ʹTesA) that catalyzes the conversion of acyl-ACP or acyl coenzyme A (acyl-CoA) into

FFA, the cells overexpressing the MMC (strain SBF15) produced 1043 mg/L, which is a 44% increase

in FFA titer over that of SBF06 (Fig. 2.2a). The cells overexpressing the ACC (strain SBF14)

produced 816 mg/L (Fig. 2.2a), which represents a 12% increase in FFA titer over that of SBF06, and

this result is consistent with previous reports [35, 36]. SBF14 showed a reduction in cell growth (Fig.

2.2b) and low glucose consumption (Fig. 2.2c), compared with those of SBF06 and SBF15.

Furthermore, the amount of acetate formed in SBF14 was 0.171 g per g of glucose consumed, which

is 45% higher than that obtained from SBF15 (Fig. 2.2d, Table 2.3). The acetate formation level in

SBF14 is similar to that observed in the previous study that overexpressed the ACC in E. coli [63].

This may indicate that the MMC bypass redirects metabolic flux from PEP to malonyl-CoA via OAA,

resulting in decrease in the acetate synthesis.

12

Figure 2.2. FFA production, cell growth, and acetate accumulation in E. coli strains overexpressing ACC or MMC. (a)

Free fatty acid production of engineered strains: SBF06, a control strain overexpressing only thioesterase; SBF14, a strain

overexpressing thioesterase and ACC; and SBF15, a strain overexpressing thioesterase and MMC. The table indicates the

enzymes overexpressed from plasmids. All strains were harvested for analysis of FFA production at 48 h post-induction. (b)

Cell growth, (c) glucose consumption, and (d) acetate accumulation in the three strains: SBF06 (●), SBF14 (▲), and SFB15

(■). The expression of thioesterase and carboxylase was induced with 0.2 mM IPTG and 25 nM tetracycline, respectively, at

an OD600 of around 1.0. Error bars indicate standard deviations of three independent cultivations.

13

Next, in order to observe the relationship between the expression levels of the enzymes and

FFA production, SBF14 and SBF15 were cultivated with different levels of inducers. The FFA

production in SBF14 increased as the inducer was added up to a concentration of 25 nM, but declined

when the inducer concentration exceeded this level (Fig. 2.3). This drop in FFA production was

consistent with a previous report [37]. In contrast, increasing the expression of MMC by increasing

the inducer concentration did not reduce FFA production (Fig. 2.3).

Figure 2.3. Effect of the expression level of ACC or MMC on FFA production. Cells (SBF14 and SBF15) were induced

with different concentrations of tetracycline ranging from 0 to 100 nM, incubated for 48 h post-induction, and harvested for

analysis of FFA production. Error bars indicate standard deviations of three independent cultivations.

The net reactions of MMC and ACC have the same stoichiometry from phosphoenolpyruvate

(PEP) (Fig. 2.1). Acetate formation is one of main competitive reactions for the consumption of

acetyl-CoA and inhibits the growth rate of E. coli [64]. In a previous study, the co-expression of ACC

and acetyl-CoA synthetase that converts acetate to acetyl-CoA resulted in increased flavanone

production and reduced acetate accumulation [32]. These results suggest that the accumulation of the

product should be avoided to prevent loss of cell viability and production titer. The MMC bypass

succeeded in improving FFA production (Fig. 2.3) and reducing acetate formation (Fig. 2.2d, Table

2.3), presumably indicating that the MMC expression redirects metabolic flux from PEP to malonyl-

CoA with reduced flux to acetyl-CoA. The MMC bypass improved the FFA production titer and yield

as compared with those of ACC-mediated acetyl-CoA carboxylation (Table 2.3). This bypass might be

also beneficial for the production of other malonyl-CoA-derived chemicals such as flavanone,

polyketide, or 3-hydroxypropionic acid [52, 65, 66].

14

Table 2.3. Results of batch fermentations with engineered strains

Strain

Optical

density

(OD600)

FFA

concentration

(g/L)

FFA yield

(g of FFA

/g of glucose)

Acetate

concentration

(g/L)

Acetate yield

(g of acetate

/g of

glucose)

SBF06 7.09 0.725 0.044 4.55 0.230

SBF14 5.57 0.816 0.059 2.99 0.171

SBF15 7.53 1.043 0.065 2.63 0.118

SBF16 7.17 0.488 0.034 ND ND

SBF17 5.56 0.454 0.033 ND ND

SBF18 7.23 1.462 0.077 ND ND

SBF19 7.35 0.839 NDa ND ND

SBF20 5.43 1.006 ND ND ND

SBF21 7.57 1.086 ND ND ND

SBF06 with Aspb 6.65 0.612 0.031 ND ND

SBF14 with Aspb 6.85 0.780 0.042 ND ND

SBF15 with Aspb 7.41 1.775 0.090 ND ND

Note: All data were obtained or calculated after 48h of post-induction. a ND, Not determined. b 665 mg/L aspartic acid was supplemented.

2.4.2. Overexpression of PPC together with MMC further enhanced FFA production

MMC produces malonyl-CoA and pyruvate from acetyl-CoA and OAA, and this pyruvate

will be converted into acetyl-CoA by the action of the pyruvate dehydrogenase complex (Fig. 2.1b).

Thus, the net reaction of MMC-mediated carboxylation is the conversion of OAA to malonyl-CoA,

indicating that the MMC bypass is an alternative route for synthesizing the precursor of FFA synthesis.

OAA can be formed from PEP by phosphoenolpyruvate carboxylase (encoded by ppc) or

intermediates in the TCA cycle. It is hypothesized that enhancing the MMC bypass by supplying

OAA would further improve FFA production.

15

Figure 2.4. Free fatty acid production with overexpression of PPC. The free fatty acid production levels of three strains

(SBF16, SBF17, and SBF18) overexpressing phosphoenolpyruvate carboxylase were measured. Expression of carboxylase

was induced with 25 nM tetracycline at an OD600 of around 1, and cells were cultivated for 48 h post-induction. The table

indicates the enzymes overexpressed from plasmids. Error bars indicate standard deviations of three independent cultivations.

To investigate this, a plasmid harboring the ppc gene under the control of a tetracycline-

inducible promoter was transformed into strain SBF06, SBF14, and SBF15, generating SBF16,

SBF17, and SBF18, respectively. Expression of PPC in the ACC-overexpressing strain negatively

affected the production of FFA, whereas SBF18 overexpressing both MMC and PPC produced about

1.5 g/L, which represents a 40% increase over that of SBF15 (Fig. 2.4). This result indicates that

increasing the intracellular concentration of OAA is important for enhancing FFA production through

the MMC reaction. The overexpression of PPC in the strains, SBF06, SBF14, and SBF15, exhibited

no effect on their cell growth (Fig. 2.5a), but did reduce glucose consumption in exponential growth

phase compared with their parental strains (Fig. 2.5b). This reduced glucose consumption may reflect

a reduced glucose uptake rate caused by competition for PEP between FFA synthesis and the PEP-

dependent glucose uptake system [67]. These findings are consistent with a previous report that the

PPC expression reduced glucose consumption but did not affect cell growth [67].

Employing the MMC bypass changes the precursor specificity from acetyl-CoA to OAA in

acetyl-CoA carboxylation. Cells overexpressing MMC can produce malonyl-CoA from two precursors:

acetyl-CoA (by chromosomal ACC) and OAA (by episomal MMC). The expression of PPC may

further reduce the metabolic flux from PEP to acetyl-CoA, resulting in low levels of fermentative

16

byproducts such as acetate [67]. Thus, the change in the precursor and the redirected carbon flux

caused by the MMC bypass serve to increase FFA production titer and yield (Table S2). This result

suggests that the activated MMC bypass provides an alternative route for malonyl-CoA synthesis from

OAA formed from PEP.

Figure 2.5. Effect of expression of phosphoenolpyruvate carboxylase (PPC) or addition of aspartic acid on cell growth

and glucose consumption of engineered strains. (a) Cell growth and (b) residual glucose concentration of four strains

[SBF06 (overexpressing the ꞌTesA), SBF16 (overexpressing the ꞌTesA and PPC), SBF17 (overexpressing the ꞌTesA, ACC,

and PPC), and SBF18 (overexpressing the ꞌTesA, MMC, and PPC)] were measured. (c) Cell growth and (d) residual glucose

concentration of three strains (SBF06, SBF14, and SBF15) were measured with the addition of 665 mg/L aspartic acid.

Expression of PPC was induced by the addition of 25 nM tetracycline at an OD600 of 1. Error bars indicate standard

deviations of three independent cultivations.

17

2.4.3. Redirecting TCA cycle intermediates into fatty acid synthesis increased FFA

production

The fatty acid biosynthetic pathway, acetate formation, and TCA cycle compete for their

precursor, acetyl-CoA, which is derived from the glycolysis pathway. For example, expression of

thioesterase increases carbon flux from PEP to acetyl-CoA in the central metabolism [68]. As a

consequence, this increases carbon flux toward fatty acid synthesis and the TCA cycle [68]. Down-

regulation of the TCA cycle can, therefore, be regarded as a strategy to increase the metabolic flux

towards FFA synthesis. Thus, it is hypothesized that the utilization of MMC might not only provide

acetyl-CoA carboxylation but might also redirect carbon flux from the TCA cycle to FFA synthesis.

Previously, overexpression of the malic enzyme (encoded by maeB) in E. coli redirected carbon flux

from malate (intermediate component of the TCA cycle) to pyruvate (precursor of acetyl-CoA) and

resulted in a 1.5-fold increase in fatty acid production in non-thioesterase-expressing E. coli [69].

Therefore, it was explored whether MMC is capable of directing the OAA in the TCA cycle to FFA

synthesis by comparing the FFA production levels in cells overexpressing or not overexpressing

MaeB.

The expression of MaeB in SBF06 and SBF14 led to 15% and 23% increases, respectively,

in FFA production titer as compared with production in the parental strains, indicating that the

redirection of malate to pyruvate by MaeB further enhanced FFA production in both strains (Fig. 2.6).

These increases suggest that directing the TCA cycle intermediates into the fatty acid synthetic

pathway has a positive effect on FFA production, as previously reported [70]. NADPH, synthesized

by MaeB activity, may not influence FFA production, as little perturbation in the NADPH

concentration was observed in a previous study [71]. In contrast, little increase in FFA production was

observed when MaeB was expressed in MMC-overexpressing cells (Fig. 2.6). Thus, MMC efficiently

directs TCA cycle intermediates towards fatty acid synthesis during the acetyl-CoA carboxylation

bypass.

18

Figure 2.6. Free fatty acid production levels of engineered strains overexpressing MaeB or supplemented with

aspartic acid. Expression of carboxylases or malic enzyme was induced with 25 nM tetracycline at an OD600 of around 1,

and cells were cultivated for 48 h post-induction. Aspartic acid (665 mg/L) was supplemented in the medium. The table

indicates the enzymes overexpressed from plasmids or the addition of aspartic acid. Error bars indicate standard deviations

of three independent cultivations.

To confirm that an increase in FFA production can be produced by supplying TCA

intermediates, aspartate (which is converted into OAA) was used as a representative amino acid for

evaluating FFA production. In the presence of aspartate, only cells overexpressing MMC exhibited an

increase in FFA production titer of up to 70% to around 1.8 g/L as compared with that of SBF15

without aspartate (Fig. 2.6). In contrast, no increase in FFA production was observed in cells

overexpressing ACC, as expected (Fig. 2.6). Cells overexpressing MMC exhibited elevated glucose

uptake and cell growth (Fig. 2.5c, d), whereas cells overexpressing ACC grew more slowly than the

other two strains, SBF06 and SBF15, in exponential phase but reached a final cell mass similar to

those of the other strains (Fig. 2.5c). The added aspartic acid (665 mg/L) is corresponding around 95

mg/L (C14 fatty acid) that accounts for around 15% of increase in FFA production (compared with

FFA in SBF15 without aspartic acid), indicating that addition of aspartic acid increases FFA

production through not only providing OAA and also regulating other metabolism such as glucose

consumption (Fig. 2.5d).

In this study, an alternative metabolic pathway, called MMC bypass, has been introduced

into E. coli for malonyl-CoA synthesis. The MMC bypass had a strong impact on FFA production by

19

avoiding the regulatory systems involved in E. coli’s native acetyl-CoA carboxylation, altering

precursor specificity, and redirecting carbon flux toward FFA synthesis. Further improvements were

obtained by increasing the flux of PEP carboxylation or supplementing with aspartate. The highest

yield of the MMC bypass (0.09 g of FFA/g of glucose) was obtained when the aspartic acid was

supplemented, which is equivalent to 26.5% of the theoretical yield. Given that the added aspartic

acid could be additional carbon source, the yield of FFA production from only glucose was 0.085 g of

FFA/ g of glucose. However, the yield is almost two-fold higher than that obtained in the previous

study that co-overexpressed ACC and thioesterases [35]. Further improvement might be achieved

through reducing the carbon flux toward byproduct formation [72] or preventing the fatty acid

degradation [38]. Finally, it is believed that the bypass of acetyl-CoA carboxylation may potentially

open up new avenues for the production of FFA, as evidenced by the increased FFA production

through the MMC bypass (Fig. 2.7).

Figure 2.7. Fold change in FFA production of engineered strains. The FFA production titer of each strain was compared

with that of SBF06 overexpressing only thioesterase. Symbols (+ or -) indicate overexpression of enzymes or addition of

chemicals. Error bars indicate standard deviations of three independent cultivations.

2.5. Conclusions

Fatty acid production begins with the carboxylation of acetyl-CoA to malonyl-CoA by the

ACC reaction in E. coli. However, the complex regulation of ACC expression/activity leads to

limitations in FFA production. The use of the acetyl-CoA carboxylation bypass showed a potential

way to avoid multiple regulatory systems and optimize carbon flux. Combining MMC expression

with the addition of aspartate, which is converted into OAA, resulted in an approximately 2.5-fold

increase in FFA production. Based on the findings, it is hopefully expected that this engineering

strategy will increase the production of FFA.

20

Chapter 3

Construction of a mutant Acyl-CoA thioesterase I

with high specific activity

3.1. Abstract

Microbial production of oleochemicals has been actively studied in the last decade. Free fatty

acids (FFAs) could be converted into a variety of molecules such as industrial products, consumer

products, and fuels. FFAs have been produced in metabolically engineered Escherichia coli cells

expressing a signal sequence-deficient acyl-CoA thioesterase I (‘TesA). Nonetheless, increasing the

expression level of ‘TesA seems not to be an appropriate approach to scale up FFA production because

a certain ratio of each component including fatty acid synthase and ‘TesA is required for optimal

production of FFAs. Thus, the catalytic activity of ‘TesA should be rationally engineered instead of

merely increasing the enzyme expression level to enhance the production of FFAs. Here, we

constructed a sensing system with a fusion protein of TetA (tetracycline resistance protein) and RFP

(red fluorescent protein) under control of a FadR-responsive promoter to select the desired mutants.

Fatty acid-dependent growth and RFP expression allowed for selection of FFA-overproducing cells. A

‘TesA mutant that produces a twofold greater amount of FFAs was isolated from an error-prone PCR

mutant library of E. coli ‘TesA. Its kinetic analysis revealed that substitution of Arg64 with Cys64 in the

enzyme causes an approximately twofold increase in catalytic activity. Because the expression of

‘TesA in E. coli for the production of oleochemicals is an almost indispensable process, the proposed

engineering approach has a potential to enhance the production of oleochemicals. The use of the

catalytically active mutant ‘TesAR64C should accelerate the manufacture of FFA-derived chemicals and

fuels.

21

3.2. Introduction

FFAs have been synthesized from acyl-ACPs—intermediates of iterative cycles of fatty acid

synthetic pathway—by means of thioesterases. Among them, E. coli acyl-CoA thioesterase I (TesA)

has been frequently employed because its characteristics are well- understood. TesA has Ser10-Asp154-

His157 as a catalytic triad and belongs to the family of SGNH-hydrolases [19]. The catalytic

mechanism of the enzyme is as follows: the hydroxyl group of a serine residue nucleophilically

attacks a thioester bond of the acyl-ACP or acyl-CoA with the assistance of the histidine residue [21].

TesA in a leaderless form (‘TesA) has been used to produce FFAs in several studies [35, 38, 72, 73].

The optimal ratio of enzymes in a metabolic pathway is believed to be important for

improvement of product titers [74, 75]. In the fatty acid synthetic pathway, nine distinct enzymes

(FabA, B, D, F, G, H, I, Z, and ACP) interact with extended acyl-ACPs to synthesize long-chain acyl-

ACPs. An optimal molar ratio of the fatty acid synthase (FAS) components is the key for

maximization of FAS activity. It was reported that the relative molar ratio of ‘TesA to each FAS

component is also required for maximal production of FFAs [73]. In addition, expression of ‘TesA

enhances FAS activity by hydrolyzing the long-chain acyl-ACPs that inhibit several enzymes

involved in the fatty acid synthesis [76]. According to these findings, overexpression of the wild type

‘TesA may not be sufficient to maximize the production of FFAs in E. coli (Fig. 3.1a). We

hypothesized that engineering ‘TesA with a high specific activity is necessary for further improvement

of the FFA production (Fig. 3.1b).

In this study, we constructed a mutant ‘TesA library through error-prone PCR and developed

a high-throughput screening (HTS) method to obtain FFA-overproducing strains. The catalytic

properties of the desired mutant enzyme were characterized by biochemical assays. Finally, we

demonstrated that the mutant ‘TesA enzyme yielded a twofold greater amount of FFAs than the wild

type enzyme did.

22

Figure. 3.1 A schematic diagram of fatty acid synthesis in E. coli. (a) FFA production is improved up to a certain protein

level of a wild type thioesterase; however, it does not improve further even at its highest level because of stoichiometric

protein-protein interactions involved in fatty acid synthesis. (b) FFA production is improved when a catalytic active

thioesterase is devised. Abbreviations: AccABCD, acetyl-CoA carboxylase; MAT, malonyl-CoA:ACP transacylase; DH,

hydroxyacyl-ACP dehydratase; ER, enoyl-ACP reductase; KR, ketoacyl-ACP reductase; KS III, ketoacyl-ACP synthase III;

KS, ketoacyl-ACP synthase I and II; TE, thioesterase; ACP, acyl carrier protein.

3.3. Methods

3.3.1 Bacterial strains and plasmids

Strains, plasmids, and primers used in this study are listed in Tables 3.1 and 3.2. E. coli

MG1655 served as a parental strain for construction of all the strains used in this study. Deletion of

chromosomal fadE (acyl-CoA dehydrogenase gene) was performed by means of λ-recombination

system as described elsewhere [77].

23

Table 3.1. Strains and plasmids used in this study

Strains and plasmids Genotype and description Source

Strains

MG1655 E. coli K-12 F–λ–ilvG–rfb-50rph-1 [56]

SBF01 MG1655 with pFAB This study

SBF02 MG1655 containing pFAB and pBbB6c-‘tesA This study

SBF03 MG1655 with ΔfadE::FRT This study

SBF04 MG1655 with ΔfadE::FRT containing pFAB This study

SBF05 MG1655 with ΔfadE::FRT containing pFAB and

pBbB6c-‘tesA

This study

SBF06 MG1655 with pBbB6c-‘tesA This study

SBF07 MG1655 with pBbB6c-‘tesAA120D, A171V This study

SBF08 MG1655 with pBbB6c-‘tesAR64C This study

SBF09 MG1655 with pBbB6c-‘tesAD74G This study

SBF10 MG1655 with pBbB6c-‘tesAR64T This study

SBF11

SBF12

SBF13

MG1655 with pBbB6c-‘tesAR64Q

MG1655 with pBbB6c-‘tesAFT

MG1655 with pBbB6c-‘tesAR64CFT

This study

This study

This study

SBL01 MG1655 library containing pFAB and mutated ‘TesA This study

BL21(DE3) E. coli B F–ompT gal dcm lon hsdSB(rB-mB

-) λ(DE3 acI

lacUV5-T7 gene 1 ind1 sam7 nin5])

[78]

BL21(DE3)-‘TesA BL21(DE3) with pET28a-‘tesA This study

BL21(DE3)-‘TesAR64C BL21(DE3) with pET28a-‘tesAR64C This study

Plasmids

pBbE8a-rfp ColE1 ori, carrying PBAD promoter and rfp, AmpR [57]

pBbE8a-fadR pBbE8a with Δrfp::fadR (from MG1655), AmpR This study

pBBR1 MCS-3 Cloning vactor carrying MCS and lacZ alpha, TcR [79]

pFAB pBbE8a-fadR carrying PLR-tetA-rfp on AatII site This study

pBbB6c-gfp BBR1 ori, carrying PL-lacO1 promoter and gfp, CmR [57]

pBbB6c-‘tesA pBbB6c-gfp with Δgfp::‘tesA, CmR This study

pBbB6c-‘tesAA120D, A171V pBbB6c-gfp with Δgfp::‘tesAA120D, A171V, CmR This study

pBbB6c-‘tesAR64C pBbB6c-gfp with Δgfp::‘tesAR64C, CmR This study

pBbB6c-‘tesAD74G pBbB6c-gfp with Δgfp::‘tesAD74G, CmR This study

pBbB6c-‘tesAR64T pBbB6c-gfp with Δgfp::‘tesAR64T, CmR This study

pBbB6c-‘tesAR64Q

pBbB6c-‘tesAFT

pBbB6c-‘tesAR64CFT

pBbB6c-gfp with Δgfp::‘tesAR64Q, CmR

pBbB6c- gfp with Δgfp::flag tagged‘tesA, CmR

pBbB6c- gfp with Δgfp::flag tagged‘tesAR64C, CmR

This study

This study

This study

pET28a (+) Expression vector with (His)6-tag, KmR Novagen

pET28a-‘tesA pET28a with C-terminally (His)6-tagged ‘tesA This study

pET28a-‘tesAR64C pET28a with C-terminally (His)6-tagged ‘tesAR64C This study

24

To generate all the plasmids harboring ‘tesA variants, the BioBrick plasmid pBbB6c-gfp

served as a starting vector, and we replaced the gfp insert with PCR-amplified ‘tesA alleles after

digestion of the plasmid and PCR products with EcoRI and XhoI. The ‘tesA and tesAR64C alleles were

also cloned into the NcoI- and XhoI-digested pET28a expression vector for purification of His-tagged

proteins. The biosensor plasmid pFAB was constructed by cloning E. coli fadR and a tetA-rfp fusion

gene with a FadR-responsive promoter into pBbE8a-rfp. In brief, fadR of E. coli MG1655 was

amplified by means of primers, FadR FP and FadR RP, and ligated into pBbE8a-rfp after digestion

with BglII and XhoI, resulting in the pBbE8a-fadR vector. The translational fusion of tetA and rfp

linked by a flexible peptide linker with GGGS4 [80] was ligated with PL promoter and FadR-binding

sites [81] using splice overlap extension (SOE)-PCR [82], and the product was cloned into the AatII-

digested pBbE8a-fadR vector.

Table 3.2. Primers used in this study

Primers Sequence (5’-3’)

fadE_del_FP CCATATCATCACAAGTGGTCAGACCTCCTACAAGTAAGGGGCTTTTCGTTGTGTAGGCTGGAGC

TGCTTC

fadE_del_RP TTACGCGGCTTCAACTTTCCGCACTTTCTCCGGCAACTTTACCGGCTTCGATTCCGGGGATCCGT

CGACC

fadE_seq_FP AAAAGTTAGCCAGCGTTTCCGCCGC

fadE_seq_RP ACGTTGGGAGATGAGACGTATCAGG

FadR FP AAAGATCTTTTAAGAAGGAGATATACATATGGTCATTAAGGCGCAAAG

FadR RP TACTCGAGTTATCGCCCCTGAATGGCTA

TesA FP AAAGAATTCAAAAGATCTTTTAAGAAGGAGATATACATATGGCGGACACGTTATTGAT

TesA RP TTACTCGAGTTATGAGTCATGATTTACTA

TesA seq FP AATTGTGAGCGGATAACAATTGAC

PLR FP ACCTGACGTCGCTAGCATCTGGTACGACCAGATTTGACAATCTGGTACGACCAGATGATACTGA

GCACATCAGCAGGACGCACTGA

TetA FP ACATCAGCAGGACGCACTGACCGAATTCAATTTAAGAAGGAGATATACATATGAAATCTAACAA

TGCGCTCATCG

TetA RP AGAACCGCCTCCAGAACCACCACCGGAGCCGCCGCCGCTTCCACCGCCGGTCGAGGTGGCCC

GGCTCCATGCA

RFP FP GGCGGTGGAAGCGGCGGCGGCTCCGGTGGTGGTTCTGGAGGCGGTTCTATGGCGAGTAGCGAA

GACGTTATCA

RFP RP ATAAGACGTCTACCGCCTTTGAGTGAGCTG

TEShis FP ATATACCATGGCGGACACGTTATTGATTCTGGGTG

TEShis RP TGGTGCTCGAGTGAGTCATGATTTACTAAAGGCTGC

TEShis seq RP CTAGTTATTGCTCAGCGG

TESflag RP CCTTACTCGAGTTACTTGTCATCGTCATCCTTGTAATCTGAGTCATGATTTACTAAAGGCTGC

TesA SD FP TGCTGAAACAGCATCAGCCGNNSTGGGTGCTGGTTGAACTGGGCGGCAAT

TesA SD RP CGGCTGATGCTGTTTCAGCA

Underlined sequences indicate restriction enzyme sites.

25

3.3.2. Media and cultivation conditions

The response of the fatty acid sensing system to fatty acids was measured as previously

described [83]. An overnight culture grown in rich broth (RB) (10 g tryptone, 5 g NaCl and 1 g yeast

extract per liter) was inoculated at 1:100 into 5 mL of the fresh RB medium supplemented with the

various concentrations of oleic acid (0, 0.1, 0.5, 1, or 2 mM) and 40 µg/mL tetracycline. Tween 20

was added to final concentration of 0.5% to solubilize the oleic acid. After 24 h of growth at 37 °C,

RFP fluorescence and optical density at 600 nm (OD600) were measured on a 96-well microplate

reader (Infinite M200 TECAN, Austria) and a spectrophotometer (Libra S22, England). The

fluorescence was then normalized to OD600 (RFP fluorescence per OD600 unit). The response of

sensing system to internally produced FFAs was measured as cultivating various strains in the M9

medium supplemented with 2 mM MgSO4, 0.1 mM CaCl2, additional 100 mM sodium phosphate

buffer (pH 7.0), and 0.6% glycerol. Addition of 0.3 mM IPTG and 40 µg/mL tetracycline was

conducted at OD600 of 0.5. Each parameter such as OD600, RFP intensity, and FFA concentration was

measured as mentioned above and ‘Analysis of FFA and glucose’ section, respectively.

To analyze the FFA production, a single colony was picked from an agar plate and incubated

with 5 mL of a LB medium at 37 °C. An overnight culture was washed with distilled water twice and

inoculated at the inoculum size of 3% into 5 mL of the M9 medium supplemented with 2 mM MgSO4,

0.1 mM CaCl2, additional 100 mM sodium phosphate buffer (pH 7.0), and 2% glucose. The cells were

cultivated at 30 °C for 48 h. The expression of thioesterases was induced by addition of 0.3 mM IPTG

at OD600 of 0.5. To measure several parameters such as cell density, glucose concentration, and FFA

concentration in the SBF06 and SBF08, cells were cultivated in mini-bioreactors as previously

described [62]. Briefly, cells were grown in LB medium, which was used to inoculated 1:100 into 5

mL of the M9 medium supplemented with 2 mM MgSO4, 0.1 mM CaCl2, 3% glucose, 1 g/L yeast

extract, and trace elements. The trace elements consist of 2.4 g FeCl3∙6H2O, 0.3 g CoCl2∙6H2O, 0.15 g

CuCl2∙2H2O, 0.3 g ZnCl2, 0.3 g Na2MO4∙2H2O, 0.075 g H3BO3, and 0.495 g MnCl2∙4H2O per liter.

Overnight grown cells in the minimal medium were harvested and resuspended in 70 mL of the fresh

M9 medium. The pH was maintained at 7.0 with 2 N NaOH. IPTG (0.3 mM) was added to induce the

expression of thioesterase at OD600 of 1.

3.3.3. Mutant library construction

Error-prone PCR mutagenesis of ‘tesA was conducted, and the PCR products were cloned

into EcoRI and XhoI sites of pBbB6c-gfp by the Mutagenex Inc (www.mutagenex.com). The mutant

library containing 106 independent colonies was constructed without hot spots of mutations.

http://www.mutagenex.com/

26

Site-directed mutagenesis was performed by PCR amplification using the overlapping

oligonucleotide primers containing degenerate codon NNS [84] to randomly replace the arginine at

position 64 of ‘TesA with another amino acid. Briefly, two overlapping PCR products were generated

from tesA by means of the primer sets (TesA FP, TesA SD RP, TesA SD FP, and TesA RP).

Overlapping PCR products were then spliced using the appropriate primer set (TesA FP and TesA RP).

The final constructs were obtained by the same procedure that was used for plasmid construction. In

total, 1016 colonies were generated.

3.3.4. Enrichment and isolation of ‘TesA mutants with increased activity

To confirm the enrichment efficiency, two strains in a 19:1 ratio (SBF02 to SBF05) were

mixed and incubated in the LB medium supplemented with 0.5% glucose, 0.1 mM IPTG, and 40

µg/mL tetracycline. The cells were subcultured at 1:50 into the fresh medium every 24 h. After each

round of enrichment, cells from each mixed culture were plated on an agar plate and incubated at

37 °C overnight. Fifteen of the resulting colonies were randomly picked and subjected to colony PCR

and DNA sequencing of the amplified fadE locus using primers (fadE seq FP and fadE seq RP) to

evaluate the proportion of SBF05 in each mixed population. For enrichment of FFA-overproducing

cells, the ‘TesA mutant library was transformed into strain SBF01, and 7.2 109 cells harboring

mutant ‘TesA were obtained. Three successive subcultures were performed by transferring 2% (v/v) of

the culture to the new subculture at 24-h intervals. The concentration of tetracycline was increased at

every transfer (40, 50, and 60 µg/mL).

Cells enriched with tetracycline were sorted based on the RFP signal using a FACS Calibur

instrument (BD Bioscience, USA). The sorted cells were recovered in 50 mL of the SOC medium at

37 °C for 1 h prior to plating on an agar plate. For single-cell culture analysis, colonies grown on the

agar plate were cultivated in a deep 96-well plate with 200 µL of the LB medium at 30 °C for 16 h.

Overnight cultures were transferred at 1:100 into a new deep 96-well plate filled with 400 µL of the

LB medium supplemented with 0.5% glucose and 0.1 mM IPTG. After 24 h of cultivation, the RFP

signal was measured as described above.

3.3.5. Analysis of FFA and glucose

The concentration of FFAs was measured as previously described [38]. In a 2.0-mL tube, 500

µL of cultures was taken and acidified by addition of 50 µL of 6 N HCl. The FFAs were extracted by

vigorous vortexing for 30 sec with 500 µL of ethyl acetate and 50 µL of 1 g/L methyl nonadecanoate

(Sigma-Aldrich) as an internal standard. The organic layer was separated after centrifugation for 2

27

min, and the same procedure was repeated with additional 500 µL of ethyl acetate. Extracted FFAs

were subjected to methylation with 100 µL of the mixture MeOH: 6 N HCl (9:1, v/v) and 100 µL of

TMS-diazomethane (Sigma-Aldrich). After incubation for 10–15 min, fatty acid methyl esters

(FAMEs) were analyzed using a gas chromatograph (Agilent 7890A) equipped with a flame

ionization detector (FID) and a DB-5 column (30 m 0.25 mm, Agilent Technologies). The

identification and concentration of FAMEs were analyzed by comparing peaks with the external

standards composed of C10–22 FAMEs (Sigma-Aldrich). Oven temperature of 60 °C was held for 2.5

min and ramped to 250 °C at 20 °C/min, and then held for 4 min. Finally, temperature reached 325 °C

at 10 °C/min and was held for 5 min.

Residual glucose was analyzed using a Shimadzu HPLC station equipped with a refractive

index detector (Shimadzu) and a SIL-20A auto-sampler (Shimadzu) as previously described [62].

Supernatant of each culture was collected and heated at 80 °C for 1 h prior to second centrifugation

for 30 min. The final samples were injected into a HPX-87P column (Bio-Rad) at 0.6 mL/min with

HPLC-grade water.

3.3.6. Enzyme expression and kinetic analysis

BL21(DE3) cells harboring pET28a-‘tesA and pET28a-‘tesAR64C were cultured in the LB

medium at 37 °C until OD600 reached 0.6. The expression of ‘TesA and ‘TesAR64C was induced by

addition of 0.1 mM IPTG, and the cells were allowed to grow at 18 °C for 20 h. The cells were

harvested by centrifugation at 4000 g for 20 min, resuspended in lysis buffer (40 mM Tris-HCl, pH

8.0), and disrupted by ultrasonication (Sonic Dismembrator Model 500, Fisher Scientific). The cell

debris was removed by centrifugation at 13000 g for 40 min, and the supernatant was loaded onto

Ni-NTA agarose (QIAGEN). The bound proteins were eluted with 300 mM imidazole in lysis buffer.

The purified protein was concentrated to 3.1 × 104 mM in 40 mM Tris-HCl pH 8.0.

Thioesterase activity was measured on a UV-1800 UV-Vis Spectrophotometer (Shimadzu,

Japan). The reaction mixture contained 50 mM potassium phosphate buffer (pH 7.0), 80 µg/mL BSA,

0.1 mM DTNB [5,5-dithiobis-(2-nitrobenzoic acid)], 3.1 × 105 mM enzyme, and a substrate (0, 2, 4,

6, 8, 12, 16, 20, or 24 µM palmitoyl-CoA) in the volume of 1 mL [85]. The reduction of DTNB by

released CoA was monitored at 412 nm for 1–3 min. The kinetic parameters were calculated by

nonlinear repression plots of the Michaelis-Menten equation.

28

3.4. Results and discussion

3.4.1. Construction and characterization of a high-throughput screening system

Various sensing systems were previously employed to detect specific cellular signals and to

improve production of valuable chemicals [86-88]. In this study, a fatty acid sensing system was

developed based on an existing system [81] to screen FFA-overproducing mutants in the ‘TesA mutant

library. We placed the TetA-RFP fusion protein rather than reporter RFP under the control of a FadR-

regulated promoter, comprised of the PL promoter and FadR-binding sites. This screening system was

expected to allow two selections: (1) FFA-dependent tetracycline resistance for enrichment of the

expected mutants, and (2) FFA-dependent RFP expression to identify the desired mutants and to

eliminate false positive recombinants. FadR, an acyl-CoA-responsive transcriptional regulator,

represses the expression of several genes prior to binding to long-chain acyl-CoAs [83]. In E. coli,

acyl-CoAs can be synthesized from FFAs by acyl-CoA ligase (FadD). Thus, the intracellular

concentration of FFAs and acyl-CoAs well correlate in E. coli [81]. At the low concentrations of FFAs

or acyl-CoAs, FadR may consistently inhibit the expression of TetA-RFP in our sensing system.

However, the high level of FFAs or acyl-CoAs should antagonize FadR and activate the expression of

TetA-RFP. This sensing system was designated as the Fatty Acid Biosensor (FAB, Fig. 3.2a).

29

Figure. 3.2 Construction of the FFA-sensing plasmid and its performance. (a) The plasmid map of the fatty acid

biosensor (pFAB); tetA : tetracycline resistance gene; rfp: red fluorescent gene; Ampr: ampicillin resistance gene. (b) Effect

of extracellular FFA concentrations on cell growth. At 40 µg/mL tetracycline, SBF01 was cultivated with different

concentrations of oleic acid (mM) [2 (●), 1 (■), 0.5 (▲), 0.1 (◆), and 0 (○)]. (c) Effect of extracellular FFA concentrations

on RFP expression. At 40 µg/mL tetracycline, SBF01 was cultivated with various concentrations of oleic acid (d) Screening

efficiency. The strain SBF02 and SBF05 were cultivated in the 19:1 ratio to confirm that the FFA-overproducing cells were

dominant in enrichment culture condition containing 40 µg/mL tetracycline. The population of SBF05 was determined by

colony PCR and DNA sequencing of the fadE locus of 15 randomly picked colonies after every enrichment cycle. Error bars

mean standard deviations of three independent experiments.

First, we confirmed a dose response of the FAB toward exogenous FFAs. When E. coli

MG1655 harboring pFAB, designated strain SBF01, was cultivated with various concentrations of

oleic acid and 40 µg/mL tetracycline, the fastest growth was observed in cells cultivated with the

highest level of oleic acid tested (Fig. 3.2b). In addition, the RFP intensity of cells was well correlated

with the growth of cells (Fig. 3.2c), indicating that the sensing system could confer the corresponding

tetracycline resistance and RFP intensity depending on the levels of FFAs. Furthermore, we measured

the functionality of sensing system to internally produced FFAs. Cell growth was measured in E. coli

30

strains MG1655 harboring pFAB (SBF01), MG1655 harboring pFAB and pBbB6c-‘tesA (SBF02),

and MG1655 with fadE deletion harboring pFAB and pBbB6c-‘tesA (SBF05). When tetracycline (40

µg/mL) was added to exponentially growing cells, the SBF05 grew faster than other two strains (Fig.

3.3a). As expected, the SBF05 produced the highest level of FFAs, around 365 mg/L, 1.4 and 6.1-fold

higher than others (Fig. 3.3b). In addition, RFP intensity was also correlated with the cell growth and

FFA production of each strain (Fig. 3.3c). The fastest growth and highest RFP intensity of the SBF05

should be resulted from its high FFA production. These results imply that the constructed sensing

system shows a functional response to FFAs or acyl-CoAs.

Figure 3.3. Response of the FAB biosensor to endogenous FFA. Three strains (SBF01, SBF02, and SBF05) containing the

biosensor were cultivated in minimal medium to confirm its response to internally produced FFAs. Tetracycline and IPTG

were added in exponentially growing cells (around OD600 of 0.5, time point 0 h). (a) Cell growth of three strains. (b)

Concentration of internally produced FFAs. (c) Normalized RFP intensity in three strains. (d) Effect of internally produced

FFAs on cell growth in enrichment medium. The cell growth of two strains (SBF02 and SBF05) was measured to confirm

enrichment. Each symbol represents SBF01 (◆), SBF02 (■), and SBF05 (●). Error bars mean standard deviations of three

independent experiments.

31

Finally, we tested whether the sensing system could act as a screening tool for the

enrichment of FFA-overproducing cells. The fadE-undeleted strain (SBF02) and the fadE-deleted

strain (SBF05) were mixed in the 19:1 ratio and incubated with 40 µg/mL tetracycline. After four

cycles of the enrichment, the population of the SBF05 considerably increased to 30% from 5% after

four cycles of the enrichment (Fig. 3.2d). This increased population of the FFA-overproducing strain

(SBF05) clearly showed that the constructed sensing system detected the level of FFAs and

selectively supported the growth of strains synthesizing a large amount of FFAs.

The FFA-overproducing cells showed short doubling time even at a high concentration of

tetracycline (Fig. 3.3d), resulting from the increase in the ratio of the FFA-overproducing strain after

four cycles of the enrichment culture (Fig. 3.2d). In addition, the different levels of intracellular and

extracellular fatty acids resulted in changes of tetracycline resistance and RFP expression. These

results indicate that the constructed sensing system effectively confers the corresponding tetracycline

resistance and red fluorescence in response to FFAs produced by the cells. This genetic sensing

system was used to screen ‘TesA mutants for those with high enzymatic activity that enables cells to

produce more FFAs.

3.4.2. Genotypic and phenotypic analysis of isolated ‘TesA mutants

Three recombinant cells [(MG1655 expressing a mutant ‘TesAA120D/A171V (SBF07), ‘TesAR64C

(SBF08), or ‘TesAD74G (SBF09)] produced more FFAs than MG1655 expressing a wild type ‘TesA

(SBF06) (Fig. 3.4). The SBF07 had a mutant ‘TesA with two point mutations: A120D and A171V. It

was difficult to explain how two alterations improve the FFA production because there are no studies

on the Ala171 residue. On the other hand, the Ala120 residue is a part of the loop (residues 111–120)

located around a substrate-binding region. The loop slightly moves when ‘TesA interacts with its

substrate [22]. Thus, we can hypothesize that the mutation positively influences substrate-binding

affinity and improves the FFA production. Strain SBF08, which expresses a mutant ‘TesA with a

substitution of longer and basic arginine with sulfur-containing cysteine at position 64th, showed the

highest production of FFAs (~1.1 g/L) among the three recombinants. This concentration was almost

twofold higher than the amount produced by the SBF06. The SBF09 strain harboring ‘TesAD74G

produced ~1.5-fold higher FFAs than the SBF06. As an N-terminus of the flexible loop (residues 75–

80) in ‘TesA, the Asp74 controls movements of the loop during substrate binding [22]. Thus, the

alteration of Asp74 may increase the FFA production in a manner similar to that of the SBF07. The

three mutants showed quite similar fatty acid distribution as compared with the SBF06 (Additional

file 1: Table S1), indicating that the substitutions are not likely to significantly influence substrate

specificity.

32

Figure. 3.4. Free fatty acid production of selected strains from the ‘TesA mutant library. The expression of ‘TesA was

induced with 0.3 mM IPTG, and the cells were cultivated for 48 h post-induction. Parenthesises indicate mutation of the

‘TesA. Filled circles (●) indicate the optical density of each culture after 48 h. Error bars mean standard deviations of three

independent experiments.

Alteration of enzyme characteristics such as substrate specificity [89], activity [90], and

deregulated allosteric inhibition [91] has been used to increase production of desired products. The

three mutants (strain SBF07, SBF08, and SBF09) isolated from the ‘TesA mutant library accounted

for ~22%, 33%, and 22% of 18 selected mutants, respectively. This result indicates that the

constructed sensing system selectively supported the growth of FFA-overproducing cells in the

mutant library. The mutations in ‘TesAA120D/A171V and ‘TesAD74G might cause a conformational change

in the loop regions mentioned above and subsequently improve the substrate-binding affinity as

previously reported [22], resulting in FFA overproduction. Given that no studies have shown the

improvement of FFA production by altering amino acids in ‘TesA to date, our results may offer target

amino acid residues for the engineering of ‘TesA.

33

Fig. 3.5. Free fatty acid production from ‘TesA mutants at position 64 generated by site-directed mutagenesis. The

expression of ‘TesA was induced with 0.3 mM IPTG, and the cells were cultivated for 48 h post-induction. Parenthesises

indicate mutation of the ‘TesA. Filled circles (●) indicate the optical density of each culture after 48 h. Error bars mean

standard deviations of three independent experiments.

3.4.3. The effect of mutation at Arg64 of ‘TesA on enzymatic activity

In order to identify the substitution leading to the highest thioesterase activity, we performed

site-directed mutagenesis on the Arg64 residue. Replacement of arginine with threonine (‘TesAR64T) or

glutamine (‘TesAR64Q) showed a 1.6- and 1.2-fold increase in the FFA production, respectively (Fig.

3.5). Among the cells isolated from the targeted random mutant library, a mutant with ‘TesAR64C like

strain SBF08 produced the largest amount of FFAs (Fig. 3.5). This result indicates that this

substitution might provide high catalytic activity although its location is far from an active site in the

protein’s structure [18]. In addition, the SBF08 showed comparable growth rate and glucose

consumption rate with the SBF06 (Fig. 3.6a and 3.6b), but significantly higher FFA production (Fig.

3.6c), indicating that the expression of ‘TesAR64C could improve final titer, productivity, and yield.

Kinetic parameters of the ‘TesA and ‘TesAR64C were analyzed in assays with palmitoyl-CoA

to confirm the correlation between the increased FFA production level and increased enzymatic

activity. As expected, Km value of the ‘TesAR64C toward palmitoyl-CoA was ~60% of that of the ‘TesA,

indicating that the substitution increases the substrate affinity (Fig. 3.7). The kcat value of the

engineered enzyme was 30% higher than that of the wild type enzyme, suggesting that the ‘TesAR64C

more rapidly converts palmitoyl-CoA to palmitic acid. Therefore, these kinetic parameters can be

considered supporting evidence of the highest amount of FFA synthesis in the SBF08.

34

Figure 3.6. Batch culture of the SBF06 and SBF08 in mini-bioreactor. Two strains were cultivated in the defined M9

minimal medium (see Methods) to measure (a) cell density, (b) glucose consumption, and (c) FFA production. Each symbol

represents SBF06 (■) and SBF08 (●). Error bars mean standard deviations of three independent experiments.

Several residues of ‘TesA have been studied to understand its kinetic behavior. The

movement of the switch loop (residues 75–80) influences substrate specificity by stabilizing the

Michaelis complex (MC) when ‘TesA interacts with its substrate [22]. Moreover, mutation of Trp23

(not in the active site) significantly reduces the catalytic activity [20]. The promise of ‘TesA

engineering was previously seen in the ‘TesAL109P, which shows altered substrate specificity [22]. Its

use in engineered E. coli increased a proportion of short-chain FFAs, but not a total amount of FFAs

[92]. It was previously reported that ‘TesA and its substrate rapidly form MC; however, conversion

into the tetrahedral complex is slow, suggesting that ‘TesA is not likely to efficiently synthesize FFAs

from acyl-ACPs [85]. Nevertheless, none of mutations in ‘TesA were found to increase the production

of FFAs to date. The ‘TesAR64C has higher substrate affinity and reaction rate than the wild type ‘TesA,

indicating that the increased specific activity of ‘TesA improves the FFA production in E. coli as we

hypothesized.

35

Figure. 3.7. Kinetic analysis of the ‘TesA (●) and ‘TesAR64C (■). The values are average of three independent experiments.

The kinetic parameters were calculated by nonlinear repression plots of the Michaelis-Menten equation. The enzyme

concentration was 3.1 105 mM.

3.4.4. Increased FFA production driven by high specific activity of ‘TesAR64C

Multiple enzymatic activities are required to synthesize FFAs from acetyl-CoAs. It has been

reported that the relative ratios of protein abundance of the FAS components are crucial for generating

maximum synergy for the fatty acid synthesis [73]. Based on this fact, we