Li»

 

 

 

 

 

 

 

3 1:03

(2,.“

n
u.

at;
Ir

5.5... ‘
In: 1

5mg?” ,.
, “Eva;

:2.
fi.‘

mam.“

3.

Han:

.

, memafisd<

2;

.3395}.
1:21.. ..

fill... was“

xix... 35:. 11.5

I...
1.: «A. .
=11. h. w.
.3. yo...“

3.]..SIYI7L-Ax Its. ‘

h.
L ,‘K-«Iai
5'): 5. .il§..
99:57:12..» 3...» k
. 1.1.... x { ... 31
.. (1:351.
i.

:9. 33.3}.
c L};

5‘22... ..

4.9} 1‘ 12: .1
:...........§...$

i. \94, 9x...

:13?!

I

 

,2

(o
Ci J l
/

This is to certify that the
dissertation entitled

lNCREASING THE CARBON FLOW INTO THE AROMATIC
COMMON PATHWAY: BIOSYNTHESIS OF 3-DEHYDROSHIKIMIC
ACID FROM D-GLUCOSE

presented by
Jian Yi

has been accepted towards fulfillment
of the requirements for the

P—h. D. degree in Organic Chemistry

mm

Major Professor’ 5 Signature

2/2.} /09

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

 

.UBRARY
M'Chigan State
University

 

 

 

to remove this checkout from your record.
on or before date due.
rlier due date if requested.

PLACE IN RETURN BOX
TO AVOID FINES return

MAY BE RECALLED with ea
DATE DUE DATE DUE

 

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p.15

 

INCREASING THE CARBON FLOW INTO THE AROMATIC
COMMON PATHWAY: BIOSYNTHESIS OF 3-DEHYDROSHIKIMIC ACID
FROM D-GLUCOSE
By

Jian Yi

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Chemistry

2004

ABSTRACT

INCREASING THE CARBON FLOW INTO THE AROMATIC
COMMON PATHWAY: BIOSYNTHESIS OF 3-DEHYDROSHIKIMIC ACID FROM
D-GLUCOSE
By

Jian Yi

Metabolic engineering, DNA microarray and recombinant DNA technology were
used to develop strategies to increase the carbon flow directed into the Shikimate
pathway. 3-Dehydroshikimic acid (DHS) is the most advanced intermediate shared by
both biosynthesis of aromatic amino acids and biocatalytic syntheses of a variety of
value-added chemicals. Strategies elaborated to increase the yield and concentration of
DHS synthesized from glucose are thus applicable to the microbial synthesis of a wide
range of chemicals.

Shikimate pathway product yields in microbial synthesis are ultimately limited by
the glucose transport mechanism. To increase the availability of phosphoenolpyruvate
(PEP), overexpression of PEP synthase or alteration of glucose transport by Glf-mediated
facilitated diffusion or the GalP galactose permease in E. coli constructs led to increased
yields of DHS and Shikimate pathway byproducts. Overexpression of PEP synthase was
currently leading to the synthesis of higher yields of DHS and Shikimate pathway
byproducts relative to alteration of glucose transport. In addition, the production of DHS
can be enhanced by changing fermentor-controlled pH from 7 to 6.

DNA microarray technology was used to study gene expression profiles under

fed-batch fermentor conditions. The results showed that a simple overexpression of PEP

 

synthase led to dramatic expression changes of many genes. Based on the microarray
study, a more stable DAHP synthase (AroGFBR) was isolated by PCR mutagenesis and has

FFBR) because of its stability over

its advantage over previously used DAHP synthase (Aro
the course of a fermentation run.

The export Of shikimic acid or DHS was also studied. Two approaches to identify
the Shikimate export protein were attempted without success. The first approach
attempted to identify a mutant deficient in Shikimate export while the second approach
was based on the overexpression of a Shikimate export protein. The failure of both
approaches is consistent with the possibility that multiple proteins exist for the export of
shikimic acid. Furthermore, the determination that intracellular shikimic acid

concentrations are much lower than extracellular concentrations suggests that export of

this hydroaromatic is not a limiting factor during microbial synthesis.

Copyright by
.fian‘Yi

2004

 

 

To my parents and my wife

For their love and support

ACKNOWLEDGMENTS

First, I would like to thank my research adviser, Professor John W. Frost, for his
patience, guidance and understanding during my time in the group. His excellent spoken
and written skill, his aggressiveness, his pertaining to detail and hard working attitude
will influence me for my entire life. In addition, I want to thank the members of my
graduate committee, Professor Baker Gregory, Professor John McCracken, and Professor

Geiger James for their discussion of the thesis.

I am grateful for Dr. Karen Draths for her friendship, patient guidance in the
experimental techniques and intelligent input to my research. Her word “ one complete
experiment was better than 10 incomplete experiments” will be remembered forever.
Meanwhile I am appreciative of current and former group members: Dr. Kai Li, Dr. Feng
Tian, Dr. Spiros Kambourakis, Dr. Sunil Chandran, Dr. Jessica Barker, Dr. Dave Knop,
Dr. Chad Hansen, Dr. Padmesh Venkitasubramanian, Dr. Dongming Xie, Dr. Jihane
Achkar, Ningqing Ran, Wei N in, Jiantao Guo, Heather Steuben, Mapitso Molefe, Xiaofei
Jia, Wensheng Li, Jinsong Yang, Justas Jancauskas, Mankit Lau, Adam Bush and Lee

Stephen Sing Kin, for their assistance and friendship. We are a big family!

This thesis is dedicated to my parents, Guangyou Yi and Liangying Yao, and my

wife, Chunhui Li. Without their love and support, this thesis would never have

happened.

vi

TABLE OF CONTENTS

LISTOFFIGURES.................... ................................................................................ x

LIST OF ABBREVIATIONS x111

CHAPTER 1

Introduction 1
Microbial syntheses of value-added chemicals utilizing the Shikimate pathway 5
Use of microarray technique to investigate the consequences of metabolic

engineering .................................................................................. 25
Export study ................................................................................ 30

CHAPTER 2

Availability of phosphoenolpyruvate and biosynthesis of DHS from glucose ........... 42
Introduction ................................................................................. 42
Modulation of PEP synthase expression increases Shikimate pathway product
yields ..................................................................................... 48
Altered glucose transport and Shikimate pathway product yields ................... 59
Comparison of KL3ar0Kar0L and SP1.1 ............................................... 80
Effect of pH on the production of DHS ................................................ 84

CHAPTER 3

Microarray analysis of gene expression changes in response to overexpression of PEP

synthase ............................................................................................... 95
Introduction ................................................................................. 95
How many genes are differentially expressed with phosphenolpyruvate synthase
overexpression during the fermentation? ............................................. 104
Pathway diagram .......................................................................... 106
Feedback insensitive isozyme of DAHP synthase ..................................... 126
E-4-P limitation and superoxide protection ......................................... 133

CHAPTER 4

Studies of Shikimate export ..................................................................... 140
Background .............................................................................. 140
Shikimate export is a carrier-mediated process and not a bottleneck for Shikimate
production ................................................................................ 142
Approaches to identify the loci involved in the export of Shikimate ......... 145

CHAPTER 5

Experimental ....................................................................................... 161

vii

LIST OF TABLES

Table 1. Concentrations and yields of DHS and byproducts synthesized by E. coli
KL3/pJYl .216A cultured under glucose-rich conditions ................................................. 54

Table 2. PEP synthase and DAHP synthase specific activities for E. coli KL3/pJYl.216A
cultured under glucose-rich conditions .............................................................................. 54

Table 3. Concentrations and yields of DHS and byproducts synthesized by E. coli
KL3/pJY1.216A cultured under glucose-limited conditions ........................................ 58

Table 4. PEP Synthase and DAHP synthase specific activities for E. coli KL3/pJYl .216A
cultured under glucose-limited conditions ...................................................... 58

Table 5. Yields and concentrations of synthesized DHS and Shikimate pathway

byproducts 65
Table 6. DAHP synthase specific activities 70
Table 7. Glucokinase specific activities 70

Table 8. Comparison of the yields and concentrations of synthesized Shikimate pathway
products and byproducts as a function of the strain and strategy employed to increase
PEP availability 76

Table 9. Comparison of the yields and concentrations of synthesized shikimic acid and
Shikimate pathway byproducts as a function of the progenitor host under glucose-rich

conditions 83

Table 10. Yields and concentrations of synthesized DHS and Shikimate pathway
byproducts as a function of pH under glucose-rich conditions 86

Table 11. Comparison of data reproducibility (coefficient of variation) for all the samples
obtained from E. coli KL3/pJY1.216A fermentation ............................................... 101

Table 12. Number of differentially expressed ORFs in response to PEP synthase (PpsA)
overexpression ..................................................................................... 104

Table 13. Functional group of the upregulated expressed ORFS (ratio >20) in response to
PEP synthase overexpression 105

Table 14. Functional group of the downregulated expressed ORFs (ratio >20) in
response to PEP synthase overexpression106

viii

Table 15. Expression data for gluconeogenic genes and all the genes involved in

Embden-Meyerhof pathway (EMP) ............................................................ 113
Table 16. Expression data for genes involved in pentose pathway ................................ 116
Table 17. Expression data for all genes involved in TCA cycle ........................... 117
Table 18. Expression data for global regulators 118
Table 19. Expression data for genes involved in the Shikimate pathway 119
Table 20. Expression data for selected genes related to stress conditions 121
Table 21. Expression data for selected transport genes 125

Table 22. Comparison of the amino acid changes and sensitivity to feedback inhibition

of AroG and AroF Mutants ...................................................................... 128
Table 23A. Concentrations and yields of 3-dehydroshikimic acid and byproducts
synthesized by E. coli KL3/pJY1.216A and KL3/pJY6.23l 132
Table 23B. DAHP synthase specific activities for E. coli KL3/pJY1.216A and
KL3/pJY6.231 cultured under glucose-rich conditions 132
Table 24. Shikimate intracellular and extracellular concentration during the fermentation
of SP1.1/pKD12.138 under glucose-rich conditions 143
Table 25. Shikimate intracellular and extracellular concentration during the fermentation
of SP1.1/pKD15.071 under glucose-rich conditions 144
Table 26. SP1.4 mutants ........................................................................... 149

Table 27. Intracellular and extracellular shikimic acid accumulation of SP1.4 mutants
under shake flask conditions 151

Table 28. Production of DHS by KL3 derivatives/pJY1.216A under glucose-rich
conditions ........................................................................................... 1 52

Table 29. PEP synthase specific activities for fed-batch fermentation of
KL3ile/pJY1.216A and KL3/pJY1.2l6A with 0.5 mL 100 mM IPTG addition ....153

Table 30. Identified genes to resist (6S)-6-fluoroshikimic acid toxicity 153

Table 31. Bacterial strains and plasmids ........................................................... 162

LIST OF FIGURES

Figure 1. The common pathway of aromatic amino acid biosynthesis and pathways
beyond chorismate .................... 3

 

Figure 2. Chemicals synthesized from tryptophan, tyrosine, and phenylalanine .............. 6
Figure 3. Shikimic acid’s use in synthetic chemistry 6

Figure 4. Comparison of p-hydroxybenzoic acid (PHB) synthesis from glucose versus
potassuim phenoxide .......................... R

 

Figure 5. Comparison of benzoquinone and hydroquinone synthesis from glucose and

 

 

 

benzene ....................................... 10
Figure 6. Value added-chemicals synthesized from glucose via DHS intermediacy ......12
Figure 7. Synthesis of gallic acid from glucose 13
Figure 8. Comparison of microbial and chemical synthesis of catechol ......................... 15
Figure 9. Comparison of microbial and chemical synthesis of vanillin 15
Figure 10. Comparison of microbial and chemical synthesis of adipic acid ................... 17
Figure 11. Reactions catalyzed by (a) transketolase and (b) transaldolase 21
Figure 12. Different strategies to remove PEP limitation 22
Figure 13. Overview of Microarray process 27
Figure 14. Synthesis of DHS from D-glucose 43
Figure 15. Theoretical flux distribution for directing carbon into common pathway with
glucose as the carbon source ................................ 46
Figure 16A. Construction of plasmid pJY1.131A and pJY1.143A 49
Figure 16B. Construction of plasmid pJY1.207A and pJY1.211A 50
Figure 16C. Construction of plasmid pJY1.216A 51

Figure 17. E. coli KL3/pJY1. 216A cultured under glucose- -rich conditions with 12 mg of

IPTG added at 18, 24, 30, and 36 h. 56
Figure 18. Restriction enzyme maps of plasmids .............................................. 60
Figure 19. Construction of plasmid pJY1.282 and pJY1.283A ........................................ 62
Figure 20. Construction of plasmid pJY3.2lB and pJY3.29B ......................................... 63

Figure 21. Growth and synthesis of DHS during cultivation of E. coli JY1/pJY2.183A
under glucose-limited conditions (gray bars) and E. coli JY1.3/pKL5. 17A under glucose-
rich conditions (open bars) ............................................................................. 68

Figure 22. Synthesis of Shikimate pathway byproducts during cultivation of E. coli
JY1/pJY2.183A under glucose-limited conditions and E. coli JY1.3/pKL5.17A under
glucose-rich conditions .............................................................................................. 68

Figure 23. DHS production by E. coli KL3/pJY1.216A with and without IPTG
addition ............................................................................................... 92

Figure 24. DHQ, DAH and GA production by E. coli KL3/pJYl. 216A with and without

Figure 25. Cluster analysis of expression data by each condition 103
Figure 26. Pathway diagram of expression ratio at 18 h (+IPTG/-IPTG) .................. 107
Figure 27. Pathway diagram of expression ratio at 24 h (+IPTG/-IPTG) . ............108
Figure 28. Pathway diagram of expression ratio at 30 h (+IPTG/-IPTG) . . . . . . . ......109
Figure 29. Pathway diagram of expression ratio at 36 h (+IPTG/-IPTG) . . . . . . . ....1 10

Figure 30. Futile cycle involving Adk (adenylate kinase), PpsA (PEP synthase) and

PykAF (pyruvatekinase) ............................................................................ 112

Figure 31. The specific activity of AroFFBR (open bars) and aroFFBR-encoded mRNA

transcript (filled circles) over the course of an E. coli KL3/pJY1.216A fermentor run with

IPTG addition 120

xi

Figure 32A. Construction of plasmid pJY6.186 ............................................ 130

Figure 32B. Construction of plasmid pJY6.231 ............................................ 131
Figure 33. (A) Y-linker. (B). PCR of transposon-flanking sequences ...................... 148
Figure 34. Summary of the aromatic biosynthetic pathway 156

xii

 

 

Ac
ADP
ATP

CIAP
Cm

C mR
COMT
DAHP
DCU
DEAE
DHQ
DHS
DO
DTT
E4P
EMP
EPSP
FBR
GA

LIST OF ABBREVIATIONS

acetyl

adenosine diphosphate

adenosine triphosphate

ampicillin

ampicillin resistance gene

base pair

chorismic acid

calf intestinal alkaline phosphatase
chloramphenicol

chloramphenicol resistance gene
catechol-O-methyltransferase
3-deoxy-D-arabin0-heptulosonic acid 7-phosphate
digital control unit
diethylaminoethyl
3—dehydroquinic acid
3-dehydroshikimic acid

dissolved oxygen

dithiothreitol

D-erythrose 4-phosphate
Embden-Meyerhof pathway
5-enolpyruvoylshikimate 3-phosphate
feedback resistant

gallic acid

xiii

 

HPLC

Kan
KanR

kb

Km
LB

M9
min
mL

11L
mM
11M
mRN A
NAD
NADH
NADP

NADPH
NMR
OD
ORF
PCA
PEG

hour

high pressure liquid chromatography

isopropyl B-D-thiogalactopyranoside

kanamycin

kanamycin resistance gene

kilobase

kilogram

Michaelis constant

luria broth

molar

minimal salts

minute

milliliter

microliter

millimolar

micromolar

messenger RNA

nicotinamide adenine dinucleotide, oxidized form
nicotinamide adenine dinucleotide, reduced form
nicotinamide adenine dinucleotide phosphate, oxidized
form

nicotinamide adenine dinucleotide phosphate, reduced form
nuclear magnetic resonance spectroscopy

optical density

open reading frame

protocatechuic acid

polyethylene glycol

xiv

PEP
pfu
PHB
PID
PCR
Phe
psi
PTS

QA

SA
SAM
SDS
S3P
Tc
TCA
Trp
TSP
Tyr
UV

 

phosphoenolpyruvic acid
plaque forming units
p-hydroxybenzoic acid
proportional-integral-derivative
polymerase chain reaction
L-phenylalanine

pounds per square inch
phosphotransferase system
quinic acid

rotations per minute

shikimic acid
S-adenosylmethionine

sodium dodecyl sulfate
Shikimate 3-phosphate
tetracycline

tricarboxylic acid

L-tryptophan

sodium 3-(trimethylsily1)propionate-2,2,3,3-d4
L-tyrosine

ultraviolet

XV

CHAPTER 1

Introduction

Chemistry is moving into an era in which renewable resources and starting materials
such as D-glucose, D-xylose, and L-arabinose will play an important role in industrial
chemical manufacture.l Current chemical manufacture typically relies on abiotic, chemical
catalysts and on starting materials derived from petroleum. As a nonrenewable natural
resource, petroleum has several negative environmental and geopolitical problems associated
with its use. For example, aromatics are currently derived predominantly from the benzene,
toluene, xylene (BTX) fraction of petroleum refining. With annual production levels at 1.9
billion gallons in the United States, benzene is the most important component of the BTX
fraction.2 Benzene is a potent carcinogen,3 and it is also included by the Environmental
Protection Agency on the list of chemicals covered by the Chemical Manufacturing Rule
that requires drastic reductions in emissions of hazardous organic air pollutants.4

Beyond the health costs associated with benzene, the costs of deriving this starting
material from nonrenewable fossil fuel feedstocks are important to consider.5 Oil spills and
land reclamation along with geopolitical complications substantially add to the true cost of
aromatic manufacture from fossil fuel-derived BTX. By contrast, plant-derived starch,
cellulose, and hemicellulose are abundant, renewable sources of glucose, xylose, and
arabinose, which can be used as carbon source for microbial synthesis of a variety of
chemicals.6 The avoidance of toxic intermediates, reagents, and byproducts, and the ability
to synthesize a chemical by microbial biocatalysis from renewable feedstocks and nontoxic
starting materials presents itself as an appealing alternative to traditional chemical
manufacture.6

In chapter 2 of this thesis, two basic strategies were examined to surmount the
fundamental limitation on the yields of 3-dehydroshikimic acid and shikimic acid imposed
by PTS—mediated glucose transport. In chapter 3 of this thesis, DNA microarray

technology was applied to detect differential transcription profiles of relative “high-yield”

cells and relative “low-yield” cells. The gene targets for further yield improvement for
biocatalyzed synthesis of 3-dehydroshikimic acid were identified. In Chapter 4 of this
thesis, efforts towards identification of the shikimic acid efflux protein is reported and
whether efflux of shikimic acid is a limiting factor during microbial synthesis of shikimic
acid is established. All the chapters focus on developing strategies to increase the carbon
flow directed into the Shikimate pathway to improve titers and yields of biosynthesized
chemicals. 3—Dehydroshikimic acid (DHS) is the most advanced intermediate shared by
both biosynthesis of aromatic amino acids (Figure 1) and biocatalytic syntheses of a variety
of value-added chemicals. Strategies elaborated to increase the yield and concentration of 3-
dehydroshikimic acid synthesized from glucose are thus applicable to the microbial

synthesis of a wide range of chemicals.

The Shikimate pathway

The Shikimate pathway, also referred to as the common pathway of aromatic amino
acid biosynthesis, has been the subject of considerable study due to its role in the
transformation of simple carbohydrate precursors into aromatics in plants, bacteria, fungi,
and molds.7 It consists of seven enzyme-catalyzed reactions converting
phosphoenolpyruvic acid (PEP) and erythrose 4-phosphate (E4P) into chorismic acid
(Figure 1). Three terminal pathways then lead from chorismic acid to L-tryptophan, L-
tyrosine, and L-phenylalanine. In addition, biosynthetic pathways leading to ubiquinone,

folic acid, enterochelin and huge number of secondary metabolites also branch away from

the common pathway at chorismic acid (Figure 1).5 Folic acid-derived coenzymes are
frequently involved in the biosynthetic transfer of one carbon fragment, ubiquinones are
involved in electron transport, and enterochelin is an iron chelator responsible for iron
uptake in numerous microorganisms.

Individual pathway enzymes have received attention due to the novel mechanisms

H203PO
C02H H 3Po4

phosphoenolpyruvate t

 

OH O aroF aroB aroD
M H aroH H203PO OH OH
H203PO OH DAHP 3-dehydroquinic acid
erythrose 4-phosphate
COH CO H CO H
2 NADPH NADP 2 ATP ADP 2 PEP H3PO4
O’JigH‘OH 3’05 Ho“ OH 2:3}; HzosPo‘“ 5 OH 3’“
OH OH
3- -dehydrosl.-lhikimic acid shikimic acid shikimate 3-phosphate
COQH H3PO4 COgH COZH
OJL C $©JL —>’ Ubiquinones
H203P Po“ 002H 3’0 o’lL COZH
OH OH
EPSP choriZmic acid hydroxybenzoic acid

COZH
COzH COQHH HOZC“ COZH
KEN” 1L
OH

0 C02 H NH 2
2-aminobenzoic acid isochorismic acid 4-aminobenzoic acid prephenic acid
_ menaquinones folic acid L- rosine
L tryptophan enterobactins L-ttilhenylalanine

Figure 1. The common pathway of aromatic amino acid biosynthesis and pathways
beyond chorismate. Intermediates (abbreviations): 3-deoxy-D-arabino-heptulosonic acid
7-phosphate (DAHP), 3-dehydroquinic acid (DHQ), 5-enolpyruvyl-shikimate—3-phosphate
(EPSP). Enzymes (genes): aroF, DAHP synthase (Tyrosine); aroG, DAHP synthase
(phenylalanine); aroH, DAHP synthase (Tryptophan); aroB, DHQ synthase; aroD, DHQ
dehydratase; aroE, shikimate dehydrogenase; aroL, shikimate kinase; aroK, shikimate
kinase; aroA, EPSP synthase; aroC, chorismate synthase.

employed during turnover of substrate into product and as targets for inhibition. The
existence of the shikimate pathway in plants and bacteria but not in humans provides an
appealing mode of action for enzyme-targeted herbicides and antibiotics.
The substrates and products of the seven enzymatic reactions that convert PEP and
E4P into chorismic acid (Figure 1) were identified by the early 19605 from studies of
bacterial auxotrophs of Escherichia coli and Klebsiella aerogenes. The first committed
step of aromatic amino acid biosynthesis involves the condensation of PEP and E4P to form
3-deoxy-D-arabin0-heptulosonic acid 7-phosphate (DAHP) and is catalyzed by DAHP
synthase.5 Three isozymes of DAHP synthase exist in E. coli, each of which is sensitive to
feedback inhibition by one of the three aromatic amino acids. The genes aroF, aroG and
aroH encode for tyrosine-sensitive, phenylalanine-sensitive, and tryptophan-sensitive
isozymes of DAHP synthase, respectively. DAHP is converted into 3-dehydroquinic acid
(DHQ) by DHQ synthase which is encoded by aroB8 in a complex reaction where NAD’r
is a catalytic requirement.9 A syn elimination of water from DHQ affords 3-
dehydroshikimic acid (DHS).IO This reaction is catalyzed by DHQ dehydratase, which is
encoded by aroD. Reduction of DHS to shikimic acid in the presence of NADPH is
catalyzed by aroE-encoded shikimate dehydrogenase." Shikimic acid is further converted
‘0 Shikimate 3-phosphate by phosphoryl group transfer from ATP. This reaction is
catalyzed by two isozymes of shikimate kinase encoded by the loci aroL12 and aroK.l3 5-
EnOlpyruvoylshikimate 3-phosphate (EPSP) synthase, encoded by aroA,l4 catalyzes the
reversible condensation of shikimate 3-phosphate and PEP. Product EPSP is formed along
With inorganic phosphate. The last enzyme of the common pathway is chorismate
Syllthase.15 Encoded by aroC, it catalyzes the elimination of inorganic phosphate from

EPSP tO afford chorismic acid.

 

 

Microbial syntheses of value-added chemicals utilizing the shikimate pathway
L-Phenylalanine, L-tryptophan and L-tyrosine
Aromatic amino acids such as L-phenylalanine, L-tryptophan and L—tyrosine are

prominently among the chemicals being microbially manufactured from glucose.6' '6 L-

Tryptophan (market volume 1.1 — 1.2 x 107 kg/year”) is produced predominantly for use
as a feed additive. Introduction of naphthalene dioxygenase into a tiyptophan-synthesizing
microbe that also expresses tryptophanase results in biocatalytic synthesis of indigo (Figure

2),”3 the vat dye that gives blue jeans their faded-blue coloration. L-Phenylalanine (market

volume 5 -— 6 x 105 kg/year”) is produced predominantly for the production of the low-
calorie sweetener aspartame (Figure 2).16 Other uses for L-phenylalanine are its use in
infusion fluids, in food additives, as intermediate for synthesis of pharmaceuticals (HIV

protease inhibitor, anti-inflammatory drugs, rennin inhibitors, etc.) and as a flavor

enhancer.6b L-tyrosine (market volume 1.5 x 105 kg/yearl 7) is produced at a small scale and
its use for the production of the anti-Parkinson’s drug 3,4—dihydroxyphenyl L-alanine (L-
DOPA)l9 (Figure 2), for the treatment of Basedow’s disease and as a dietary supplement.6b
Microbial production of L—phenylalanine has been focused mainly on E. coli, C.
glutamicum and Brevibacterium strains.20 Classic methods have been applied to screen for
aUXOtrophs and mutants with feedback deregulated key enzymes. Sugimoto et al. cloned
the genes encoding feedback resistant forms of DAHP synthase, aroFFBR, and chorismate
mutase/prephenate dehydratase, pheAFBR, into a temperature-controllable expression vector.
An L-tyrosine auxotrophic E. coli strain carrying this plasmid can reach a titer of 46 g/L
and a productivity of 0.85 g/L/h.2| To obtain L-tyrosine overproducers, most attention has
been focused on screening for regulatory and auxotrophic mutants. These were mostly
strains Of E. coli, Bacillus subtilis or various coryneform bacteria.22 By application of
recombinant DNA technology additional improvements of these L-tyrosine producing

strains have been

H o
Indigo
CHsozC o
N NH2
L-phenylalanine —_->, H
COZH
aspartame

HO COZH
- . —>
L tyrosme ____, HOmHZ

3,4-dihydroxyphenyl L-alanine

Figure 2. Chemicals synthesized from tryptophan, tyrosine, and phenylalanine.

 

 

 

 

O O
COZH \/
Ho“ . on o“ ”NH.
OH /:r HN\"/
shikimic acid 0
l tamiflu
H O
o N R. \\-NH
3 \C _, HAN-Oi; b
. O O i . i
HO '/6 H ’0

2.18 * 105 molecules
Figure 3. Shikimic acid’s use in synthetic chemistry.

made.23 Metabolic engineering for production of L-tryptophan, has primarily been carried

out in E. coli (reviewed by Berry“), Corynebacterium25 and Bacillus.26 Alterations in the

central carbon metabolism and the common aromatic amino acid pathway, including its

regulation, were carried out. In addition, overexpression of the genes encoding the

tryptophan branch of the aromatic amino acid pathway was performed.24

Shikimic acid

Shikimic acid has been exploited by Schreiber and coworkers in their use of this
hydroaromatic as the core scaffold for the synthesis of large combinatorial libraries of
molecules (Figure 3).27 In addition, shikimic acid is a valuable chiral synthon used in the
synthesis of tamiflu (soeltaminvir, Figure 3),28 which is a potent inhibitor of both influenza
A and influenza B neuraminidases. Tamiflu was approved by the FDA and is marketed by
Roche as an orally administered antiinfluenza agent. Tamiflu’s six-membered carbocyclic
ring, carboxylic acid, and backbone stereocenters have led to the use of shikimic acid as the
starting material. Nonetheless, shikimic acid’s use in synthetic chemistry has been
restricted by its limited availability and prohibitive cost. Shikimic acid is isolated from plant
sources. However, isolation of shikimic acid from the pericarps of Illicium is a tedious
procedure.29 The supply of [llicium does not benefit from monoculture.30 When Tamiflu
came out of clinical trials, Roche was confronted by a price for shikimic acid that precluded
its viable use as a pharma starting material.

To alleviate supply issues, E. coli SP1.1/pKDl2.138 was constructed for microbial
conversion of glucose into shikimic acid.31 Shikimic acid production using E. coli
SP1.1/pKDl2.138 has been scaled up to a 50,000 L process and is used by Roche as one
Of its sources of shikimic acid. This shikimate-synthesizing biocatalyst resulted from

disruption of the genomic aroL and aroK loci (Figure 1) in E. coli and overexpression of

. . . FBR .
feedback-msenSItive, aroF -encoded DAHP synthase to channel more carbon Into the

common pathway.32

Another advantage of a microbe-catalyzed conversion of glucose into
shikimic acid is the ability to make tons of shikimic acid from glucose within a 48 h time

frame. This abundant, reliable source of shikimic acid contrasts with the need to wait for an

entire growing season for Illicium plants to mature and circumvents the problems of

adverse weather destroying production for an entire Illicium growing season.

 

 

 

 

CO2H COzH
AOH see Fig 1 see Fig 1
—> ——> JL
—> V —>
s OH HO‘ i OH , o COZH
HO OH OH OH
D-glucose shikimic acid chorismic acid
COZK A COZH Am
20 atm 002 H+
-———> ——-->
180-250 °C
OK OH OH
potassium potassium PHB
phenoxide p-hydroxybenzoate

Figure 4. Comparison of p-hydroxybenzoic acid (PHB) synthesis from glucose
versus potassuim phenoxide. Enzyme: Chorismate-pyruvate 1yase (UbiC).

p-Hydroxybenzoic acid

p-Hydroxybenzoic acid is a component of liquid crystal polymers such as Xydar,33
which have attracted considerable attention because of their use in high-performance
applications. Esters of p-hydroxybenzoic acid are also widely used as food preservatives.34
p-Hydroxybenzoic acid is currently manufactured by Kolbe-Schmitt reaction of dried
potassium phenoxide with 20 atm dry carbon dioxide at 180-250 °C (Figure 4).34 Product
p-hydroxybenzoate potassium salt is converted to its free acid upon addition of mineral
acid. Besides the required temperatures and pressures, p-hydroxybenzoic acid manufacture
has to contend with handling of phenol which is listed as a highly toxic, corrosive
chemical.35

Chorismic acid can also be converted directly into p-hydroxybenzoic acid in a

reaction catalyzed by ubiC-encoded36 chorismate-pyruvate 1yase (Figure 4). An E. coli

biocatalyst has been constructed consisting of a genome-localized aroAaroLaroBaroCkanR

cassette and plasmid-localized aroFFBR, rktA, and ubiC. The genes encoding o
chorismate-utilizing enzymes are disrupted to prevent biocatalytic conversion of choris
acid into anthranilic acid and prephenic acid (Figure l). The biocatalyst can synthesize
g/L of p-hydroxybenzoic acid from glucose under fed-batch fermentor conditions.37
Hydroxybenzoic acid can also be synthesized by means of chemical dehydratior
shikimic acid. This reaction is catalyzed by l M sulfuric acid in refluxing acetic I

(Figure 4).37

Quinic acid and hydroquinone

Quinic acid is another useful molecule that can be microbially synthesized thro
the utilization of the shikimate pathway (Figure 5).38 Quinic acid’s six-membe
carbocyclic ring arrayed with stereogenic centers has been extensively used as a ct
synthon and starting material in natural product synthesis.39 It is currently obtained by

expensive isolation from plant sources. An E. coli biocatalyst has been constructed

mutational inactivation of aroD-encoded DHQ dehydratase, overexpression of aroFI
encoded feedback-insensitive DAHP synthase and overexpression of aroE—enco
shikimate dehydrogenase. The biocatalyst can synthesize 49 g/L of quinic acid f1
glucose in 20% (mol/mol) yield.32 Quinic acid is then chemically converted ‘
hydroquinone and benzoquione by simple chemical oxidation (Figure 6).“7’8 Other hi
yielding chemical methodology has also been elaborated for conversion of quinic acit
these fermentation broths into hydroquinone.40 Oxidative decarboxylation of quinic acit
clarified, decolorized, ammonium ion-free fermentation broth with NaOCl and subseqi
dehydration of the intermediate afforded purified hydroquinone in 87% yield. Halide-l
oxidative decarboxylation of quinic acid in fermentation broth with stoichiometric quant:
of (NH4)2Ce(SO,,)3 and V205 afforded hydroquinone in 91% and 85% yield, respectiv
Conditions suitable for oxidative decarboxylation of quinic acid with catalytic amount

metal oxidants were also identified. Ag3PO4 at 2 mol % relative to quinic acit

fermentation broth catalyzed the formation of hydroquinone in 74% yield with K,s,<
serving as the cooxidant. Selective reduction of photoactivated silver ion with hydroquino
is the basis for this organic's widespread use in photography. Benzoquinone is

important chemical precursor in the manufacture of various chemicals.4|

NH2
M002
benzene aniline stk C O
O

OH see Fi 12H/ benzoquinone
\ __>g ‘ ArOE H‘ —
H‘O“ OH a, b

OI'C :

HO OH0 \ OH
D-glucose 3-dehydroquHinic acid quinic Oacid MnOZ
O HO

2
/ hydroquinone
II A New

benzene p-diisopropyl
benzene

 

 

 

 

 

 

 

Figure 5. Comparison of benzoquinone and hydroquinone synthesis from gluco
and benzene. Enzyme: Shikimate dehydrogenase (AroE). Key a) i. NaOCl,
isopropanol, iii. reflux; b) (NH 4).,Ce(SO,,)3 (2.4equiv), 30 min; reflux, 10 h. c) Ag3PO4 (
mol %), K23208, 50 °C, ii.reflux.

Hydroquinone is produced globally at volumes of 4.5 — 5.0 x 107 kg/year.42 T
dominant route to manufacture hydroquinone is through hydroperoxidative synthesis
The p-diisopropylbenzene is synthesized by zeolite-catalyzed Friedel-Crafts reaction
benzene or cumene with propylene or isopropanol. Air oxidation of the .
diisopropylbenzene proceeds at 90 - 100 °C in an aqueous N aOH solution also containi
organic bases along with cobalt or copper salts. Hydroperoxycarbinol and dicarbinol :
produced along with the dihydroperoxide during air oxidation. Treatment with acid 3

H202 converts the hydroperoxycarbinol and dicarbonyl to the dihydroperoxide which

10

cleaved to form acetone and hydroquinone. During the acidic cleavage, explosive orgz
peroxide can form which presents a safety hazard.42 Benzoquinone is prima
manufactured from oxidation of aniline (Figure 5). The oxidant employed. is MnOg
aqueous H28 04 (Figure 5). Benzoquinone can also be reduced by Fe or hydrogenatec
afford hydroquinone. Although accounting for approximately 10% of hydroquinc

production,42 this manufacturing route generates large quantities of MnSO4, (NH4)2S

and iron oxide salts.42

3-Dehydroshikimic acid

3-Dehydroshikimic acid (DHS), a powerful antioxidant,43 is an intermediate in
biosynthesis of aromatic natural products derived from chorismic acid and is also a bra;
point from which aromatic natural products can be biosynthesized without intermediacy
chorismic acid (Figure 6). The aromatization of 3-dehydroshikimic acid to protocateck
acid is crucial to the value of 3-dehydroshikimic acid as an intermediate towards
synthesis of several commodity and fine chemicals. Protocatechuic acid is a branch poin
the biosynthesis of catechol, cis, cis-muconic acid, vanillin, gallic acid and pyrogal
Hydrogenation of cis-cis-muconic acid affords adipic acid. The Frost group deveIOpec
coli KL3/pKL5.17A as a construct capable of converting glucose into 3-dehydroshikii
acid.44 This 3-dehydroshikimate-synthesizing biocatalyst resulted from disruption of

genomic aroE loci (Figure 1) in E. coli and overexpression of feedback-insensit}

aroFFBR-encoded DAHP synthase and transketolase to channel more carbon into
common pathway.44 Progress made in improving the synthesis of 3-dehydroshikimic 2
can be viewed as progress made towards the synthesis of increased yields of shikirr

pathway products biosynthesized either with or without the intermediacy of chorismic aci

11

 

 

COZH

 

 

 

 

  

NH2
COZH
HO OH
HOZC OH
HO adipic acid pyrogallol
OH HO
L-DOPA T OH i
\ catechol COZH
C02H /
HO OH
HO OH
OH \ / gallicacid\
protocatechuic acid 0 O
COZH \/\
o H /
0 5H OH HO OH
CH3O OH
OH DHS propyl gallate
vanillin H
OH
.\OH
5 OH
HO OH
D-glucose

Figure 6. Value added-chemicals synthesized from glucose via DHS intermediacy.

Gallic acid and pyrogallol

Gallic acid can be isolated from gall nuts or from tara powder made from ground
seed pods of the Peruvian tree species Coulteria tinctoria. Although gallic acid is available
from natural sources, an inherent danger is its undependable supply. Neither the gall nuts
acquired from insect carapaces nor the tara powder obtained from ground seed pods are
cultivated in a controlled fashion. Both sources are highly susceptible to drought or flood
conditions. Chemical oxidation of DHS can afford gallic acid.45

converted to gallic acid biocatalytically via protocatechuic acid (PCA) intermediary (Figure

12

DHS can also be

 

7). DHS dehydratase, which is encoded by the aroZ locus isolated from Klebsiella
pneumoniae, converts DHS to PCA. Hydroxylation of PCA catalyzed by a mutant isozyme
of p-hydroxybenzoate hydroxylase (p0bA*) isolated from Pseudomonas fluorescens then
affords gallic acid. An E. coli biocatalyst has been constructed which synthesizes

approximately 20 g/L of gallic acid from glucose under fed-batch fermentation conditions.45

OH

 
 

 

COZH
\OH see _>Fig 1
0 AroZ
s. OH
HO OH
D-glucose 3-dehydroquHinic acid protocatechuic acid
Cu2+/Zn2\‘ flobA"
COZH
HO OH
OH
gallic acid

 

 

 

Figure 7. Synthesis of gallic acid from glucose. Enzyme: DHS dehydratase (AroZ), p-
hydroxybenzoate hydroxylase mutant (p0bA*).

Derivatives of gallic acid include: propyl gallate (Figure 6), an important food-grade
antioxidant; trimethoprim, an antibacterial agent; and pyrogallol, a product of chemical or
enzymatic decarboxylation of gallic acid. Pyrogallol is used in photographic developing
solutions and is a key building block in the manufacture of the insecticide bendiocarb.46
Decarboxylation of gallic acid to pyrogallol was conducted by the construct E. coli
RB791serAzzaroB/pSK6234 expressing a copy of aroY-encoded protocatechuic acid
decarboxylase. Addition of gallic acid to a batch culture of E. coli
RB791serA::aroB/pSK6.234 at stationary phase and at 33 °C and neutral pH afforded

pyrogallol in 97% yield.45 The toxicity of pyrogallol towards growing E. coli cells

13

precluded the direct synthesis of pyrogallol from D-glucose using a single microbiz

construct.45

Catechol

Global, non-captive production of catechol is approximately 2.5 x 104 tons p6
year.47 Estimating total world production is difficult since catechol is often part of captiv
markets where it is produced and then converted into higher, value-added products befor
ever seeing the open market as catechol. Chemical products derived from catechol includ
pharmaceuticals (L-DOPA, adrenaline, papaverine), flavors (vanillin, eugenol, isoeugenol
agrochemicals (carbofuran, propoxur), and polymerization inhibitors and antioxidants (4
tert-butylcatechol, vera'trol).48 Most catechol production begins with Friedel-Crafl
alkylation of benzene to afford cumene (Figure 8). Subsequent Hock-type, air oxidation c
the cumene leads to formation of acetone and phenol. The phenol is then oxidized to
mixture of catechol and hydroquinone using 70% hydrogen peroxide either in the presenc
of transition metal catalysts or in formic acid solution where performic acid is the actuz
oxidant. Catechol and hydroquinone are separated by distillation (Figure 8).48

Catechol is also microbially synthesized from DHS (Figure 8).49 A geneticall
modified E. coli strain mediates the dehydration of DHS to form protocatechuic aci
(Figure 8), which then undergoes an enzyme-catalyzed, non-oxidative decarboxylation t-
catechol. As in biocatalytic synthesis of gallic acid, dehydration of DHS to protocatechui
acid is catalyzed by aroZ—encoded DHS dehydratase. Conversion of protocatechuic aci:
into catechol is catalyzed by aroY-encoded PCA decarboxylase, which was isolated from K

pneumoniae. Although DHS and protocatechuic acid are intermediates in the conversion 0
glucose into catechol, only catechol accumulates in the culture supernatant. 49 However, th
toxicity of catechol towards growing E. coli cells precluded the direct synthesis of catechc

from D-glucose using a single microbial construct.49

l4

acetone hydroquinone

%Y.OCOZ b Ho’ .C \CL‘ no”c
benzene cumene

phenol OH
catechol

 

   

 

OH

 
 

COQH/
(OH see _>Fig 1 Amy
HAroZ
s OH HO
HO OH
D-glucose 3-dehydroquHinic acid protocatechuic acid

Figure 8. Comparison of microbial and chemical synthesis of catechol. Enzyme
DHS dehydratase (AroZ), PCA decarboxylase (AroY). Key: a) propylene solid H3PO

catalyst, 200-260 °C, 400-600 pSi.; b) 02, 80-130 °C then 302, 60-100 °C;, C) 70% H202
EDTA, Fe2+ or C021”, 70-80 °C.

  

 

 

 

 

OH HO COQH COZH
_\OH see Fig 1 COZH
, . AroZ COMT
5 OH O OH HO H3 co
HO OH OH
D-glucose 3-dehydroquinic acid protocatechuic acid vanillic acid
aryl-aldehyde
dehydrogenas
HO COZH COZH O H
HO EH8 H3CO!Hb H3CO C)H3OCOiH—H: H3CO I
OH
cateochol guaicacol mandelic acid 4- -hydroxy-3-methoxy vanillin
-phenyl glyoxylic acid

Figure 9. Comparison of microbial and chemical synthesis of vanillin Enzymes

DHS dehydratase (AroZ), catechol-0-methyltransferase (COMT). Key: a) (CH3O)ZSO2
b) glyoxylic acid.

Vanillin

Vanillin (4-hydroxy-3-methoxybenzaldehyde; Figure 9) is the major component c

natural vanilla, which is one of the most widely used and important flavouring material

15

worldwide. The source of vanilla is the bean, or pod, of the tropical Vanilla orchiI
(principally Vanilla planifolia Andrews, syn. V. fragrans (Salisb. Ames)).50 Vanillin iI
fact occurs in trace amounts in other plants, including commercial products such a
tobaccoS‘;however, the pods of the Vanilla orchid still remain the only commercial some
of natural vanillin. Although more than 12,000 tons of vanillin are produced each year, les
than 1% of this is natural vanillin form Vanilla; the remainder is synthesized much morI
cheaply via chemical processes.5 0

One synthetic route to produce vanillin begins with the condensation of guaiaco
with glyoxylic acid (Figure 9). Air oxidation of the resulting mandelic acid afford
phenylglyoxylic acid. Crude vanillin is obtained by acidification and decarboxylation of 4
hydroxyl-3-methoxypheny1 glyoxylic acid solution. Commercial grades are obtained b}
vacuum distillation and subsequent recrystallization.” Although this process is currenth
the major source for synthetic vanillin, it has several inherent problems. Guaiacol ha:
historically been obtained from condensation of dimethyl sulfate with catechol. Dimethy
sulfate is classified as a highly toxic, cancer-suspect agent. Catechol is listed as being toin
and corrosive while guaiacol is a toxic irritant. Catechol is currently manufactured fron
benzene (Figure 9).

Synthesis of vanillin from glucose has been achieved using a microbe-catalyzec
conversion of glucose into vanillic acid followed by an enzyme-catalyzed reduction of tin
vanillic acid to afford vanillin (Figure 9).53 A genetically engineered E. coli synthesizec
5.0 g/L of vanillic acid from glucose under fed-batch fermentor conditions.S 3 Aryl-aldehydI
dehydrogenase purified from Neurospara crassa was used to reduce vanillic acid H

vanillin in 66% isolated yield.5 3

16

0“ HOZC

OH see F' 8

" lg 0 CatA I
—> .__>
—’ HO l

HO OH OH COQH

   

 

D-glucose catechol cis, cis-muconic acid

COZH
adipic acid

 

 

y“
OH o
. b
—> —> +
benzene cyclohexane cyclohexanol cyclohexanone

Figure 10. Comparison of microbial and chemical synthesis of adipic acid. Enz
CatA, catechol 1,2-dioxygenase. Key: a) Ni-A1203, H2, 370-800 psi, 150-250 °C; b) Co
120-140 psi., 150-160 °C; c) Cu, NH4VO3, 60% HNO3, 60-80 °C.
Adipic acid

Adipic acid is primarily used for production of nylon-6,6. The annual g
demand for nylon exceeds 4.4 x 10 9 kg.54 Benzene is the principal starting material
which adipic acid is currently synthesized. Hydrogenation of benzene to pro
cyclohexane is followed by air oxidation to yield a mixture of cyclohexanol
cyclohexanone (Figure 10). Nitric acid oxidation that yields adipic acid and nitrous ox
might make a measurable contribution to global warming and ozone depletion. An i
industry group was formed to facilitate development of new technologies for nitrous (
abatement. 56 So far all technologies have demonstrated the capability to very effici
abate nitrous oxide emissions.5 6

Inclusion of a catA gene encoding catechol l,2~dioxygenase in the cateI
producing microbe converts catechol into cis,cis-muconic acid (Figure 10).“ 58 The
gene was isolated from Acinetobacter calcoaceticus.59 E. coli WNl/pWN2.248

developed that synthesized 36.8 g/L of cis,cis-muconic acid in 22% (mol/mol) yield i

glucose after 48 h of culturing under fed—batch fermentor conditions.58 After fed-t

17

fermentation was complete, the cells were removed from the broth, which was treated with
activated charcoal and subsequently filtered to remove soluble protein. Hydrogenation of
the resulting solution with 10% Pt on carbon (5% mol/mol) at 3400 kPa of H2 pressure for
2.5 h at ambient temperature afforded a 97% (mol/mol) conversion of cis,cis-muconic acid

into adipic acid. 57' 58

Metabolic engineering of E. coli to increase product titers and yields

To compete with chemical synthesis, microbial synthesis must produce value-added
chemicals in a high yield (mol/mol percent conversion) and high titer (product
concentration) while taking advantage of abundant, inexpensive feedstocks. Metabolic
engineering of biocatalytic syntheses utilizing the shikimate pathway has focused on the
elimination of feedback inhibition, overexpression of specific enzymes leading to the
desired product, and elimination of enzymes and pathways resulting in depletion of the
desired product. For example, in commercial tryptophan-producing strains, the tna gene
encoding the tryptophan-degrading enzyme tryptophanase is deleted, and the genes in the
terminal pathway are all overexpressed."0

Alteration of central metabolism in E. coli to channel increased carbon flow into the
common pathway has been the focus of considerable research activity.“ In addition to
overexpressing feedback-insensitive DAHP synthase, strategies have been evaluated to
increase E4P and PEP availability. The rate of aromatic amino acid biosynthesis is
controlled by modulation of the catalytic activity of DAHP synthase. There are three
isozymes of DAHP synthase expressed in E. call, each of which is sensitive to feedback
inhibition by one of the three aromatic amino acids. The genes aroF, araG and aroH
encode tyrosine-sensitive, phenylalanine-sensitive, and tryptophan-sensitive DAHP synthase
isozymes, respectively. The in vivo activities of these enzymes are dictated by their
expression levels, feedback inhibition by aromatic amino acids, and the availability of their

E4P and PEP substrates.

18

Overexpression of DAHP synthase does not necessarily mean that the in viv
activity of DAHP synthase has been increased because of the prominent regulatory rol
played by feedback inhibition."2 Several alleles that encode feedback insensitive DAH.
synthase have been obtained by mutation of the aroF,63 aroG,"4 and araH65 loci. Extensiv
mutation is not required to make the DAHP synthase feedback-insensitive. For instance,
single amino acid change can render AroF catalytic activity insensitive to the concentratio
of Tyr.63a Insensitivity to feedback inhibition by aromatic amino acids increases the in viv
catalytic activity of each molecule of DAHP synthase.

In 1999, several different strategies were explored in this lab for circumventin
transcriptional repression of aroF.61f One strategy placed plasmid-localized aroFFBR unde
strong promoters such as Pm. Another strategy utilized increased copies of aroFFBR an
(or) its umnodified native promoter to titrate away the cellular supply of TyrR protein the
binds the promoter region of aroFFBR and represses transcription of aroFFBR. This was th
reason to design strategies to localize one araFFBR locus per plasmid, two aroFFBR loci pe
plasmid, and one aroFFBR locus accompanied by one PmF promoter per plasmid.6
However, higher DAHP synthase activity did not necessarily translate into higher shikimat
pathway yields or titers as evidenced by the relationship between DAHP synthase specifi
activity and DHS synthesized when plasmid-localized aroFFBR was under Pm control. 6”

Ultimately, the catalytic activity of DAHP synthase increases to a point wher
further amplification of even feedback-resistant DAHP synthase does not lead to improve
synthesis of either aromatic amino acids or precursors to these end products. Historicallj
attention was focused on the in vivo availability of phosphoenolpyruvic acid (PEP) at tit

6' A number of different cellular processes and enzymes compete with DAH

juncuue.
synthase for PEP including pyruvate kinase, PEP carboxylase, and the PEP-dependeI
carbohydrate:phosphotransferase system (PTS) for uptake of glucose and structural]

related sugars. Efforts to improve intracellular PEP availability began with mutation:

61c, 66

inactivation of PEP carboxylase and pyruvate kinase.67 More recent efforts ha\

19

focused on circumventing PTS-mediated uptake of glucose in microbes such as E. coli th
will be discussed in detail next. Discovery that E4P availability is a critical limiting factor '
aromatic amino acid biosynthesis separated these two periods of research focusing c
intracellular PEP availability. 6 ‘ "

Mutational inactivation of PEP carboxylase and pyruvate kinase did not lead I
improvements in aromatic amino acid biosynthesis that were significant or practical. FI

example, mutational inactivation of ppc encoded PEP carboxylase results in a slow-growir

E. coli ppc“ strain that requires succinic acid supplementation.61C Although E. coli ppI
produces 10-fold higher phenylalanine titers relative to E. coli with native expression (

ppc, these phenylalanine titers are 10-fold lower than the amount of acetic acid which

produced. Also, the 1.5 g/L titers of phenylalanine produced by E. coli ppc' are rath<
small relative to the 46 g/L of phenylalanine which can be produced by E. coli.68

In 1990, Frost and coworkers published the first work indicating that E4
availability was also an important factor limiting in vivo DAHP synthase activity.69 As 2
aldose phosphate which can not exist in solution in a cyclic form, E4P is prone I
dimerization, trimerization and polymerization.7O Dissociation of these E4P forms back I
monomeric E4P is quite slow. This is probably the underlying reason why nature closel
matches the rate of E4P synthesis with the rate of E4P utilization, thereby maintaining 10x
steady-state concentrations of E4P. Low steady-state concentrations of E4P likely limit th
substrate's availability thereby limiting in vivo activity of DAHP synthase. Inspection of tl
enzymes which catalyze reactions where E4P is either a substrate or a product led to tl
pentose phosphate pathway and the enzyme transketolase (Figure 11). Overexpression (
transketolase increases the levels of E4P available to the cell for channeling into t1
common pathway.69 Two of the three intracellular E4P-generating reactions are catalyze
by transketolase while the third reaction is catalyzed by transaldolase (Figure 11

Transketolase also serves to generate the substrate D-sedoheptulose-7-phosphate for tl

20

E4P—generating reaction catalyzed by transaldolase. Thus, transketolase plays a major ro

 

 

 

 

H203PO OH O H203PO O H203PO OH HZO3PO OH OH
- + - H é _ H + .. -
OH OH OH OH OH O OH O
o-fructose o-glyceraldehyde o-xylulose
6-phosphate 3-phosphate 5-phosphate
H203PO OH O HZO3PO OH O a H203PO OH HZO3PO OH OH C
2. + WH ‘—__—“ WH + Z. 2
OH OH OH OH OH OH O OH OH O
o-fructose o-ribose o-sedoheptulose
6-phosphate 5-phosphate 7-phosphate
HZO3PO OH OH OH H203PO O b HZOBPO OH HZOSPO OH O
z : - + 9H ‘—-—_ K‘AfH + .2.
OH OH O OH OH O OH OH OI
o—sedoheptulose o-glyceraldehyde erythrose o-fructose
7-h ht 3-h ht 0' 6— hos hat
p osp ae p osp ae 4-phosphate p p e
(E4P)

Figure 11. Reactions catalyzed by (a) transketolase and (b) transaldolase.

in increasing the levels of E4P available to the cell for aromatic production. DAH
synthase, which catalyzes the first and irreversible reaction of the common pathwa
commits elevated levels of carbon generated by transketolase to aromatic amino ac
biosynthesis only if sufficient PEP is around. Work combining expression of feedbacl
insensitive aroG with overexpression of tktA has revealed that increased DAHP syntha:
catalytic activity in tandem with increased transketolase catalytic activity leads to a twofo
increase in carbon flow directed into the common pathway above that achieved wi
overexpression of DAHP synthase alone.“ By combining fed—batch fermentor contr
with overexpression of a feedback-insensitive isozyme of DAHP synthase a1
overexpression of transketolase, shikimate pathway products were synthesized in 36‘
(mol/mol) yield from glucose,61f also a two-fold increase above the achieved yield wi

overexpression of DAHP synthase alone. Overexpression of transaldolase also reliev

E4P limitation in the presence of amplified PEP synthase, but no further improvements

21

aromatic amino acid biosynthesis are observed relative to when transketolase is also

overexpressed.7 1

altered
transport Glf ATP ADP
OH ——>Ga'P G'k OH
.\OH _ ‘ AOH

 

 

     

i OH PTS i
HO OH ' 7? ‘ HZO3PO OH

glucose OPO3H2 0 glucose

6-phosphate
C02H COZH
PEP pyruvic

Pps aCld
— \(4ATP -
AMP + Pi

Figure 12. Different strategies to remove PEP limitation. Glucose facilitator (Glf),
galactose permease (GalP), glucokinase (le), phosphoenolpyruvate synthase (Pps),
phosphoenolpyruvate:carbohydrate phosphotransferase system (PTS).

recycle

 

 

Liao and coworkers were the first to examine the impact of amplifying the
expression of PEP synthase on shikimate pathway product yields.61d PEP synthase (Figure
12) catalyzes the generation of PEP from pyruvate with the expenditure of two of the high-
energy, phosphodiester linkages of ATP to form AMP.72 Using an E. coli aroB construct
with overexpression of a feedback-insensitive isozyme of DAHP synthase, transketolase,
and PEP synthase, DAH was synthesized in 90% (mol/mol) yield from glucose.6ld' 6” This
result indicated a strategy for circumventing the limitation imposed on product yields by the
PTS employed by E. coli and many other microbes.

Some approaches to remove PEP limitation uses the principle of circumventing
PTS-mediated glucose transport. One approach is to employ a different source of carbon
for microbial synthesis of shikimate pathway products. For example, higher shikimate

pathway product yields have been achieved when E. coli is cultured on D—xylose or L-

22

arabinose due to the transport of these pentoses in E. coli by permeases driven by the
conversion of ATP to ADP.6'““ 73 Unfortunately, commercial D-xylose and L-arabinose
stream that are pure, abundant, and inexpensive, are not currently available.

Another approach is to transport glucose by non-PTS mechanism including
facilitated diffusion mediated by the Z. mobilis glf-encoded glucose facilitator and
galactose-proton symport mediated by the E. coli galP-encoded galactose permease
(Figure 12). Ingram and coworkers74 were the first to demonstrate that E. coli mutants
lacking PTS-mediated glucose transport and phosphorylation can grow on glucose with
heterologous expression of the Z. mobilis glf—encoded glucose facilitator protein and Z.
mobilis glk-encoded glucokinase (Figure 12). Heterologous expression of the Z. mobilis
glf-encoded glucose facilitator75 increased the concentrations of L-phenylalanine
synthesized by various E. coli constructs.

Various groups studied the impact of replacement of PTS with the GalP galactose-
proton symport system to transport glucose on shikimate pathway products yields in E. coli
(Figure 12).76 Valle and coworkers examined DAHP synthesis in E. coli NF9/pRW5tkt,
which relied on GalP-mediated glucose transport with inactivation of PTS-mediated glucose
transport.7621 E. coli NF9/pRW5tkt synthesized 2.4-fold higher concentrations of DAHP
under shake-flask culture conditions relative to E. coli PBlO3/pRW5tkt, which retained
PTS-mediated glucose transport?“ Plasmid pRWStkt contained tktA encoding
transketolase and araGFBR encoding feedback-insensitive DAHP synthase. A 71%
(mol/mol) yield of DAHP and a detailed analysis of carbon metabolism have recently been
reportedm' 76° for E. coli constructs where GalP was recruited for glucose transport. In
another study, Bailey and coworkers76c examined phenylalanine synthesis in E. coli
PPA3 16/pSY1 30-14, which also relied on GalP-mediated glucose transport with inactivation
of PTS-mediated glucose transport. In contrast to Valle’s result, E. coli PPA3l6/pSY130-
14 under fermentor—controlled culture conditions synthesized only 0-67% of the

phenylalanine synthesized by E. coli PPA305/pSY130-14, which relied on a PTS for

23

glucose transport?“ Plasmid pSYl30-l4 contained aroFFBR encoding feedback-insensitiv:
DAHP synthase and pheAFBR encoding chorismate mutase-prephenate dehydratase. Th:
different impact of GalP-mediated glucose transport on the synthesis of phenylalanine"
relative to the synthesis of DAHP76a can be attributed to several factors. The Bailey study76
did not overexpress either transketolase or transaldolase. Increased availability 0
phosphoenolpyruvate in E. coli is not reflected in increased shikimate pathway produc
yields until the availability of D-erythrose 4-phosphate is increased with overexpression o
transketolaseéle These studies“ 76° also did not examine the microbial synthesis of thI
same shikimate pathway product, and microbes were cultivated under different conditions.

Another unique strategy to remove PEP limitation employs adjuncts that can b
readily converted to PEP. For example, addition of succinic acid to glucose-limited culture
of E. coli constructs overexpressing phosphoenolpyruvate carboxylase leads to synthesis 0
increased concentrations of shikimate pathway products. 77

Most of the experiments mentioned above were performed in shake-flask and no
under fed-batch fermentor conditions. Under typical shake-flask conditions, cells were firs
grown to stationary phase in rich medium using antibiotics as the selection pressure tI
maintain plasmids. These cells were then harvested and resuspended in minimal medium
Glucose in the minimal medium is then converted to the desired molecule with limited or m
cell growth. Although this type of cell cultivation has been widely employed in evaluaan
pathway engineering, there are several inherent problems associated with such experimenta
procedures. For example, the use of rich medium and antibiotics can dramatically increaSI
the cost if such culturing methods were scaled up. Centrifugation and resuspension of cell
also complicate large-scale manufacture. Under shake-flask conditions, because cells an
grown in rich medium prior to resuspension in minimal medium, calculated yields do no

accurately reflect the percent conversion of carbohydrate into product. DAH yield

exceeding the theoretical maximum yield were repeatedly reported when aroGFBR, tktA anI

pps were overexpressed in an E. coli aroB cultured under shake-flask conditions.6“" 6'“: 6‘

24

In addition, under shake-flask conditions, cultures begin in glucose-rich environment and
end in a glucose—deficient environment, and both oxygenation levels and pH are difficult to
control.

In this thesis, fed-batch fermentor conditions as opposed to shake-flask culture
conditions were employed to analyze the efficiency of conversion of glucose to product.
Use of a fermentor allows glucose, dissolved oxygen concentrations, and pH to be easily
controlled. One of the advantages of using fed-batch fermentor conditions is that it can
increase cell density more than 10-fold relative to the shake-flask cultivation. In addition,
fed-batch fermentor conditions overcome most of disadvantages of shake-flask conditions.
For example, only minimal medium is used in the process, and the growth of the cells and
the production of DHS occur simultaneously under fed-batch fermentor conditions. A
strategy of using nutritional pressure to maintain the plasmid has also been developed. As a

result, antibiotics are no longer used in the fermentor.

Use of microarray technique to investigate the consequences
of metabolic engineering

Functional genomics for metabolic engineers

During the last decade of the 20th century, basic biological research underwent a
major revolution as life scientists switched their level of analyses from studying the
expression of single genes and proteins to studying large numbers of genes and gene
products simultaneously. During the first portion of the genomics era, researchers
concentrated on accumulating DNA sequence information from a variety of bacteria. As
genomic information has become available on a broad range of organisms, a new
postgenomics era has begun in which data is used as a resource base to characterize gene
expression under different conditions. Because of the emphasis on gene function, this
research is also referred to as functional genomics. The purpose of functional genomics is

to use the information made available upon sequencing a genome to quantitatively determine

25

the spatial and temporal accumulation patterns of specific mRNAs, proteins, and importan
metabolites using high-throughput technologies.78

For metabolic engineers who are largely interested in using living organisms tI
produce proteins and metabolites for commercial purposes, functional genomics provide
new tools and approaches for understanding, modeling, and ultimately manipulatin;
organisms. Functional genomics relies heavily on three levels of high-throughput analyses
transcriptomics (or RNA profiling) for measuring levels of mRNA, proteomics fo
determining concentrations of individual proteins, and metabolomics (metabolite profiling
for determining the amounts of important metabolites. Tradeoffs exist among the analytica
power of these systems in the amounts of data generated. and the usefulness of the result
obtained. These tradeoffs can readily be visualized in the context of the central dogma 0
molecular biology, in which DNA can be transcribed into mRNAs, mRN As translated intI
proteins, and proteins act to catalytically interconvert metabolites.

As you move through the central dogma from DNA to metabolites, the infonnatiOI
becomes increasingly useful in terms of function. While DNA sequence indicates wha
genes are present in a bacterium, mRNA measurements show which of theses genes an
expressed. Similarly, protein measurements identify those specific mRNAs that are being
translated and the amount of the specific enzymes that are present. This may or may not b
reflected in the mRNA level. Finally the amount of metabolite present (particularly if flu:
information can be deduced) may be more important than determining the potential fo
product formation as estimated by measuring enzyme levels. Unfortunately, while thI
information might become more valuable as we read through the central dogma, it become

more difficult to obtain, often less quantitative, and certainly more fragmentary.

26

Fermentation

l

Extract total RNA

l

Generate 33P-Iabeled cDNA by reverse transcription

l

Hybridize to arrays

l

Phosphorimaging
l
Data analysis
Figure 13. Overview of Microarray process.
Microarray (RNA profiling)

Analysis of the transcriptome in a given sample yields a measurement of the relative
level of each mRNA. This level reflects the balance between the transcription rate of the
gene and the rate at which that specific mRNA is being degraded. Several methods exist for
determining the levels of cellular mRNAs; most utilize nucleic acid “probes” covalently
bound to glass slides or nylon membranes. A major technology used is DNA
microarray.79 Microarrays are robotically printed sets of PCR products or conventionally
synthesized oligonucleotides. High quality, pre-made oligonucleotide and PCR product
arrays can be purchased from a number of sources, and increasing competitive pressure can
be expected to continue to drive prices down and quality up. There are two major types of
DNA microarrays. One is the oligonucleotide-based array and the other is the PCR
product-based array. For example, the Panorama microbial gene arrays from Sigma
Genosys make it possible for investigators to study expression at the transcriptional level of

all open reading frames (ORFs) simultaneously in one experiment. The Panorama E. coli

27

gene may contains 4,290 PCR-amplified ORFs, representing all of the protein-encoding
genes provided on a pair of duplicate nylon membranes. Affymetrix chips, on the other
hand, synthesized high-density arrays of oligonucleotide probes (25-mers) in situ using
light-directed chemistry.80 The probes are designed to maximize sensitivity, specificity, and
reproducibility, allowing consistent discrimination between specific and background signals,
and between closely related target sequences. The major advantages of the Panorama E.
coli gene array over Affymetrix chips are low capital investment and relatively easy
operating procedure. Therefore, Panorama E. coli gene array was chosen as a platform to
perform microarray analysis in our study. However, the Panorama E. coli gene array does
not contain intergenic regions that include promoters or regions encoding small RNA
molecules with possible regulatory functions.81

Thousands of gene probes can be represented on a single chip. mRNA is isolated
from tissue or bacteria samples and used as a template to prepare a “target”, for example,
cDNA (or cRNA for Affymetrix chips) that is labeled, usually with fluorescent by active
labeling compounds such as biotin, Cy3 and CyS or P33 by incorporation radioactive
nucleotides. The labeled target is hybridized to the probe on the microarray. Individual
cDNAs/cRNAs from the target hybridize (bind) with the corresponding probe
proportionally to their representation in a sample. A specialized scanner detects the
hybridized molecules by fluorescence or radiolabel. The whole procedure used. in this

study is summarized in Figure 13.

Application of microarray to metabolic engineering

Traditional approaches of biocatalysis optimization use random screening,
mutagenesis and engineering improvement. While these methods are still very effective, a
better understanding of the underlying physiology using genomic tools can accelerate these
efforts. Information obtained by the DNA microarrays can help pathway engineering and

process optimization in several ways. (i) Regulatory circuitry and coordination of gene

28

expression among different pathways under different growth conditions can be measured
by DNA microarray.82 (ii) The physiological state of the cells during fermentation can be
assessed by the genome-wide transcriptional pattems.“ (iii) DNA microarray can help
identify genes involved in a production process if they are coregulated.84 (iv) The
differences in genetic contents and expression profiles between wild-type and improved
strains can be compared.85 (v) The actual array data can be incorporated into the
mathematical models to describe a cellular process.86 Finally, general applications of DNA
microarray technology to understand microbial physiology will continue to generate very
large amounts of information that will ultimately benefit the pathway engineering and
fermentation optimization effort.

Previous studies have reported differences in expression during the aerobic growth
of E. coli on different carbon sources,87 in minimal and complex media,87 in high cell
density culture conditions,83 in response to heat shock,88 in response to stress,82b and in

84 Additional studies have identified changes that contribute to

response to pH change.
ethanol tolerance in ethanologenic Escherichia coli by comparison of KOll (parent) to
LY01 (resistant mutant).85 Expression levels for 205 of the ORFs were found to differ
significantly between K011 and LY01 under each of six different growth conditions.
Statistical evaluation of differentially expressed genes according to various classification
schemes identified physiological areas of importance. A large fraction of differentially
expressed ORFs were globally regulated, leading to the discovery of a nonfunctional fnr
gene in strain LY01. In agreement with a putative role for FNR in alcohol tolerance,
increasing the copy number of fnr in KOll decreased ethanol tolerance but had no effect
on growth in the absence of ethanol. Other differences in gene expression provided
additional clues that permitted experimentation. Tolerance appears to involve increased
metabolism of glycine and increased production of betaine. Changes in the expression of

several genes concerned with the synthesis of the cell envelope components were also noted,

which may contribute to increased ethanol tolerance.

29

Expression arrays offer a unique opportunity to investigate the consequences of
metabolic engineering and may provide clues to limitations in metabolic flux.89 The fed-
batch fermentation of E. coli under pH and oxygen controlled conditions is an attractive
system for such studies due to the simplicity of culture conditions and the lack of
complications from changes in oxygen availability or pH. In this thesis, the physiological
basis for the increase in yield of shikimate pathway products90 attendant with
overexpression of PEP synthase was examined by gene array technology to detect
differential transcription profiles of relative “high-yielding (51% mol shikimate pathway
product/ mol glucose)” cells when PEP synthase was optimally overexpressed and relative
“low-yielding (37% mol shikimate pathway product/ mol glucose)” cells when PEP
synthase was not overexpressed. This work detailed the investigation of PEP synthase
overexpression effect in E. coli under fed-batch fermentor conditions and the results
demonstrated that the transcript abundance of a range of genes was altered by the single
modification of phosphoenolpyruvate synthase gene expression. Several gene targets were

identified for future strain development.

Export Study

The question of how the desired product that is synthesized within the cell becomes
excreted has been a neglected field until recently. Among other hypotheses, it was formerly
assumed that unspecific leaks were present, or that diffusion enables the passage of the
desired product into the culture medium. However, these rationales cannot explain why L-
glutamate, for instance, is excreted “uphill” (i.e., at low intracellular concentrations toward
the higher extracellular concentration).91 It also cannot explain why a cellular
discrimination between the physically very similar amino acids L-glutamate and L-aspartate
occurs. Therefore, specific carriers must be present. It is now verified that at least the
excretion of L-glutamate,91 L-lysine,92 L—isoleucine,93 and L—threonine94 by C. glutamicum,

of cysteine95 and L-threonine96 by E. coli, are carrier-mediated processes.

30

A special situation prevails for the L-isoleucine export of C. glutamicam.93 Because
this amino acid is hydrophobic, both active export and diffusion actually contribute to the
transport. This set of properties has serious consequences for the removal of L-isoleucine
from the cell. Since diffusion is not a vectorial process, a high diffusion flux via the
membrane always takes place at high L-isoleucine concentrations, irrespective of whether
from the inside or outside of the cell. Therefore, in addition to exporting the L-isoleucine
synthesized, the carrier activity must also be sufficient to export the L-isoleucine that
reentered the cell by diffusion. Otherwise the exposure of the interior of the cell to high
concentrations of the hydrophobic amino acid might influence vital functions. It is
remarkable in this context that for the more hydrophobic aromatic and branched-chain
amino acids, far lower arrrino acid concentrations are obtained in productions than with the
charged arrrino acids, where extremely high concentrations (>150 g/L) are reached.

Shikimic acid synthesis affords an opportunity to elaborate an efflux system that is
likely uniquely different from previously elaborated efflux systems given the structurally
distinct class of molecule exported. Several bacterial proteins responsible for efflux of
specific amino acids have recently been identified.97 The LysE protein responsible for L-
lysine export in Corynebacterium glutamicum is perhaps the best characterized amino acid
efflux system.92 LysE is a 25 kDa protein, which appears to span the cellular membrane six
times. C. glutamicum lysE mutants fail to accumulate L-lysine extracellularly and have
high intracellular lysine concentrations. Conversly, C. glutamicum strains possessing
multiple copies of lysE export L-lysine at a rate five-fold higher than wild-type C.
glutamicum and have relatively low intracellular lysine concentrations. Efflux of cystein
and several related metabolites is encoded by 017299 in E. coli,95 and the rgtB locus in this
same microbe has been found to participate in efflux of homoserine.96

Shikimic acid is a hydrophilic molecule and is likely transported out of the cell by a

carrier-mediated process instead of diffusion. The identification of the carrier gene and

31

study of its regulation and mechanism is central to our goal of achieving high titer, high-

yield syntheses of shikimic acid.

32

REFERENCE

1 Industry of the Future: New Biocatalysis: Essential tools for a Sustainable 215t
Centu ry C he mic a1 Ind u 5 try. 19 9 9, www.oit.doe.gov/chemicals/pdfs/
biocatalysis_roadmap.pdf.

2 PRODUCTION: DOWN BUT NOT OUT. Chem. Eng. News 2001, 80, 60-65.

3 Lewis, R. J. Carcinogenically Active Chemicals, Van Nostrand Reinhold, New
York, 1991, p. 68.

4 Ember, L. Toxic Air Emissions - Tough EPA Rule Will Hit 112 Chemicals. Chem.
Eng. News 1994, 72, 4.

5 (a) Yoshida, J .; Inomata, M. Trends in developments of aromatics production
technologies. Aramatikkusu 2002, 54, 123-135. (b) Tullo, A. H. A New Source. Chem.
Eng. News 2003, 81, 16-17.

6 (a) Frost, J. W.; Lievense, J. Prospects for Biocatalytic Synthesis of Aromatics in
the 2lst— Century. New J. Chem. 1994, 18, 341. (b) Bongaerts, J .; Kramer, M.; Muller, U.;
Raeven, L.; Wubbolts, M. Metabolic Engineering for Microbial Production of Aromatic
Amino Acids and Derived Compounds. Metabal. Eng. 2001, 3, 289-300.

7 (a) Haslam, E. The Shikimic Pathway; Wiley: New York, 1974. (b) Bentley, R. The
Shikimate Pathway- a Metabolic Tree with Many Branches. Crit. Rev. Biochem. Mal.
Biol. 1990, 25, 307-384. (c) Herrmann, K. M. In Amino Acids: Biasynthesis and Genetic
Regulation; Herrmann, K. M., Somerville, R. L., Ed.; Addison-Wesley: Reading, 1983: p.
301.

8 Frost, J. W.; Bender, J. L.; Kadonaga, J. T.; Knowles, J. R. Dehydroquinate
synthase from Escherichia cali: purification, cloning, and construction of overproducers of
the enzyme. Biochemistry 1984, 23, 4470-4475.

9 Carpenter, E. P.; Hawkins, A. R.; Frost, J. W.; Brown, K. A. Structure of
dehydroquinate synthase reveals an active site capable of multistep catalysis. Nature 1998,
394, 299-302.

10 (a) Smith, B. W.; Turner, M. J .; Haslem, E. Shikimate Pathway .4. Stereochemistry
of 3-Dehydroquinate Dehydratase Reaction and Observations on 3-Dehydroquinate
Synthetase. J. Chem. 506., Perkins Trans. 1 1975, I, 52-55. (b) Duncan, K.; Chaudhuri,
S.; Campbell, M. S.; Coggins, J. R. The Overexpression and Complete Amino-Acid-
Sequence of Escherichia coli 3-Dehydroquinase. Biochem. J. 1986, 238, 475-483.

11 (a) Anton, I. A.; Cogins, J. R. Sequencing and Overexpression of the Escherichia
coli AroE Gene Encoding Shikimate Dehydrogenase. Biochem. J. 1988, 249, 319-326. (b)
Chaudhuri, S.; Coggins, J. R. The Purification of Shikimate Dehydrogenase from
Escherichia coli. Biochem. J. 1985, 226, 217-223.

12 (a) DeFeyter, R. C.; Pittard, J. Genetic and Molecular Analysis of AroL, the Gene
for Shikimate Kinase-II in Escherichia coli K-12. J. Bacterial. 1986, 165, 226-232. (b)
DeFeyter, R. C.; Pittard, J. Purification and Properties of Shikimate Kinase-II from
Escherichia caliK-12. J. Bacterial. 1986, 165, 331-333.

33

l3 Laner-Olesen, A.; Marinus, M. G. Identification of the Gene (Arok) Encoding
Shikimic Acid Kinase-I of Escherichia coli. J. Bacterial. 1992, 174, 525-529.

14 Duncan, K.; Coggins, J. R. The Serc-AroA Operon of Escherichia coli: a Mixed
Function Operon Encoding Enzymes from 2 Different Amino-Acid Biosynthetic Pathways.
Biochem. J. 1986, 234, 49-57. (b) Duncan, K.; Lewendon, A.; Coggins. J. R. The
Purification of 5-Enolpyruvylshikimate 3-Phosphate Synthase from an Overproducing
Strain of Escherichia coli. FEBS Lett. 1984, 165, 121-127.

15 White, P. J .; Millar, G.; Coggins, J. The Overexpression, Purification and Complete
Amino-Acid Sequence of Chorismate Synthase from Escherichia coli-K12 and Its
Comparison with the Enzyme from Neuraspara crassa. Biochem. J. 1988, 251, 313-322.

16 Hodgson J. BULK Bulk Amino-Acid Fermentation - Technology and Commodity
Trading. Biotechnology 1994, 12, 152-155.

17 Chemical Econorrrics Handbook, SRI International, 1999.

18 (a) Murdock, D.; Ensley, B. D.; Serdar, C.; Thalen, M. Construction of Metabolic
Operons Catalyzing the Denovo Biosynthesis of Indigo in Escherichia coli. Biotechnology
1993, 11, 381—386. (b)Berry, A.; Battist, S.; Chotani, G.; Dodge, T. C.; Peck, 8.; Power, S.;
Weyler, W. Biosynthesis of indigo using recombinant E. coli: development of a biological
system for the cost-effective production of a large volume chemical. Proceedings-Biomass
Conference Americas: Energy, Environment, Agriculture and Industry, 2”, Portland, Oreg.
Aut. 21-24, 1995, p 1121-1129. (c) Berry, A.; Dodge, T. C.; Pepsin, M.; Weyler, W.
Application of metabolic engineering to improve both the production and use of biotech
indigo. Indust. Microbial. Biotech. 2002, 28, 127-133.

19 Haq, I. U.; Ali, S. Microbiological Transformation of L-Tyrosine to 3,4-
Dihydroxyphenyl L-Alanine (L-Dopa) by a Mutant Strain of Aspergillus. 2002, 45, 88-93.

20 De Boer, L.; Dijkhuizen, L. Microbial and enzymatic processes for L-phenylalanine
production. Adv. Biochem. Eng./Biatechnal. 1990, 41, 1-27.

21 Takagi, M.; Nishio, Y.; Oh, G.; Yoshida, T. Control of L-phenylalanine production
by dual feeding of glucose and L-tyrosine. Biotechnol. Bioeng. 1996, 52, 653-660.

22 Maiti, T. K.; Roy, A.; Mukherjee, S. K. Chatterjee, S. P. Microbial production of L-
Tyrosine: A review. Hindustan Antibiot. Bull. 1995, 37, 51-65.

23 (a) Ito, H.; Sato, K.; Enei, H.; Hirose, Y. Improvement of microbial production of L-
Tyrosine by gene dosage effect of araL gene encoding shikimate kinase. Agric. Biol.
Chem. 1990,54, 823-824. (b) Ikeda, M.; Katsumate, R. Metabolic engineering to produce
tyrosine or phenylalanine in a tryptophan-producing Carynebacterium glutamicum strain.
Appl. Environ. Microbial. 1992, 58, 781-785. (c) Ikeda, M.; Okamoto, K.; Katsumata, R.
Cloning of the transketolase gene and the effect of its dosage on aromatic amino acid
production in Carynebacterium glutamicum. Appl. Microbial. Biotechnol. 1999, 51, 201-
206.

24 Berry, A. Improving production of aromatic compounds in Escherichia coli by
metabolic engineering. Trends Biotechnol. 1996, 14, 250-256.

34

25 Katsumata, R.; Ikeda, M. Hyperproduction of tryptophan in Carynebacterium
glutamicum by pathway engineering. Biotechnology 1993, 11, 921-925.

26 Kurahashi, O.; Tsuchida, T.; Kawashima, N.; Ei, H.; Yamane, K. L-Tryptophan
production by transformed Bacillus subtilis. 1984, JP 61096990.

27 (a) Tan, D. S.; Foley, M. A.; Shair, M. D.; Schreiber, S. L. Stereoselective Synthesis
of over Two Million Compounds Having Structural Features Both Reminiscent of Natural
Products and Compatible with Miniaturized Cell-Based. Assays. J. Am. Chem. Soc. 1998,
120, 8565-8566. (b) Tan, D. S.; Foley, M. A.; Stockwell, B. R.; Shair, M. D.; Schreiber, S.
L. Synthesis and Preliminary Evaluation of a Library of Polycyclic Small Molecules for
Use in Chemical Genetic Assays. J. Am. Chem. Soc. 1999, 121, 90073-9087.

28 Kim, C. U.; Lew, W.; Williams, M. A.; Liu, H.; Zhang, L.; Swarrrinathan, S.;
Bischofberger, N.; Chen, M. S.; Mendel, D. B.; Tai, C. Y.; Laver, W. G.; Stevens, R. C.
Influenza neuraminidase inhibitors possessing a novel hydrophobic interaction in the
enzyme active site: Design, synthesis, and structural analysis of carbocyclic sialic acid
analogues with potent anti-influenza activity. J. Am. Chem. Soc. 1997, 119, 681-690.

29 Weber, W. F. Hoffmann-La Roche, Ltd., personal communication.
30 Leuenberger, H. Hoffmann-La Roche, Ltd., personal communication.

31 Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von Daeniken, R.;
Weber, W.; Frost, J. W. Hydroaromatic Equilibration During Biosynthesis of Shikimic
Acid. J. Am. Chem. Soc. 2001, 123, 10173-10182.

32 Draths, K. M.; Knop, D. R.; Frost, J. W. Shikirrric Acid and Quinic Acid: Replacing
Isolation from Plant Sources with Recombinant Microbial Biocatalysis. J. Am. Chem.
Soc. 1999, 121, 1603-1604.

33 Kirsch, M. A.; Williams, D. J. Understanding the Thermoplastic Polyester
Business. CHEMTECH 1994, 24, 40-49.

34 Szmant, H. H. In Organic Building Blocks of the Chemical Industry; New York:
Wiley, 1989, p. 467.

35 Lenga, R. E.; Votoupal, K. L. The Sigma—Aldrich Libray of Regulatory and Safety
Data; Sigma-Aldrich: Milwaukee, WI, 1993.

36 Siebert, M.; Berchthold, A.; Melzer, M.; May, U. Berger, U. FEBS Lett. Ubiquinone
Biosynthesis - Cloning of the Genes-Coding for Chorismate Pyruvate-Lyase and 4-
Hydroxybenzoate Octaprenyl Transferase from Escherichia coli. 1992, 307, 347-350.

37 Barker, J .; Draths, K. D.; Frost, J. W. Microbial synthesis of p-hydroxybenzoic
acid from glucose. Biotechnol. Bioeng. 2001, 76, 376-390.

38 Draths, K. M.; Ward, T. L.; Frost, J. W. Biocatalysis and 19th-Century Organic-
Chemistry - Conversion of D-Glucose into Quinoid Organics. J. Am. Chem. Soc. 1992,
114, 9725-9726.

39 (a) Hanessian, S.; Beaulieu, P.; Dube, D. A Novel Synthetic Route to the
Hexahydrobenzofuran Subunit of the Averrnectins and Milbemycins. Tetrahedron lett.

35

1986, 27, 5071-5074. (b) Flack, J. R.; Yadagiri, P. Enantiospecific Synthesis of D-Myo-
Inositol 1,4,5-Trisphosphate from (-)-Quinic Acid. J. Org. Chem. 1989, 54, 5851-5852.
(c) Molin, H. ; Pring, B. G. Regioselective and Stereoselective Synthesis of the Carbocyclic
Analog of 3-Deoxy-Beta-D-manna-2—Octulopyranosonic Acid (Beta— Kdo) from (-)-Quinic
Acid. Tetrahedron lett. 1985,26, 677-680. (d) Montchamp, J .-L.; Piehler, L. T.; Frost, J.
W. Diastereoselection and In vivo Inhibition of 3-Dehydroquinate Synthase. J. Am. Chem.
Soc. 1992, 114, 4453-4459. (e) Widlanski, T.; Bender, S. L.; Knowles, J. R.
Dehydroquinate Synthase: a Sheep in Wolfs Clothing. J. Am. Chem. Soc. 1989, 111,
2299-2300.

40 Ran, N.; Knop, D. R.; Draths, K. M.; Frost, J. W. Benzene-Free synthesis of
Hydroquinone. J. Am. Chem. Soc. 2001, 123, 10927-10934.

41 Szmant, H. H. In Organic Building Blocks of the Chemical Industry; New York:
Wiley, 1989, p. 512.

42 Krumenacker, L.; Constantini, M.; Pontal, P.; Sentrnac, J. In Kirt-Othmer
encyclopedia of chemical technology, 4th ed. Kroschwitz, J. 1., Howe-Grant, M., Eds.;
Wiley, New York, 1995, vol 13, p 996.

43 Chang, Y.; Almy, E. A.; Blamer, G. A.; Gray, J. 1.; Frost, J. W.; Strasburg, G. M.
Antioxidant Activity of 3-Dehydroshikimic Acid in Liposomes, Emulsions, and Bulk Oil. J.
Agric. Food Chem. 2003, 51, 2753—2757.

44 Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von Daeniken, R.;
Weber, W.; Frost, J. W. Hydroaromatic Equilibration During Biosynthesis of Shikimic
Acid. J. Am. Chem. Sac.2001,123, 10173-10182.

45 Kambourakis, S.; Draths, K. M.; Frost, J. W. Synthesis of Gallic Acid and
Pyrogallol from Glucose: Replacing Natural Product Isolation with Microbial Catalysis. J.
Am. Chem. Soc. 2000,122, 9042-9043.

46 Szmant, H. H. In Organic Building Blocks of the Chemical Industry; New York:
Wiley, 1989, p. 519.

47 Krumenacker, L.; Costantini, M.; Pontal, P.; Sentenac, J. In Kirt-Othmer
encyclopedia of chemical technology, Kroschwitz, J. I., Howe-Grant, M., Eds; Wiley, New
York, 1995, vol 13, p 996.

48 (a) Franck, H.-G.; Stadelhofer, J. W. Industrial Aromatic Chemistry; Springer-
Verlag: New York, 1988; p. 183-190. (b) Varagnat, J. In Kirt-Othmer encyclopedia of
chemical technology, 3rd ed.; Grayson, M., Ed.; Wiley: New York, 1981; vol. 13, p. 39-69.
(c) Szmant, H. H. In Organic Building Blocks of the Chemical Industry; New York:
Wiley, 1989, p. 512-519.

49 Draths, K. M.; Frost, J. W. Environmentally Compatible Synthesis of Catechol
from D-Glucose. J. Am. Chem. Soc. 1995,117,2395-2400.

50 Walton, N. J.; Mayer, M. J.; Narbad, A. Molecular of Interest: Vanillin.
Phytochemistry 2003, 63, 505-515.

51 Makkar, H. P. S.; Becker, K. Isolation of tannins from leaves of some trees and
shrubs and their properties. J. Agric. F aad Chem. 1994, 42, 731—734.

36

52 (a) Esposito, L.; Formanek, K.; Kientz, G.; Mauger, F.; Maureaux, V.; Robert, G.;
Truchet, F. In Kirk-Othmer Encyclopedia of Chemical Technology; Fourth Ed,
Kroschwitz, J. I.; Howe-Grant, M., Ed.; Wiley: New York, 1997; Vol. 24, p 812. (b) Van
Ness, J. H. In Kirk-Othmer Encyclopedia of Chemical Technology, 3rd Edition; Wiley:
New York, 1983, Volum 23, p. 704.

53 Li, K.; Frost, J.W. Synthesis of Vanillin from Glucose. J. Am. Chem. Soc. 1998,
120, 10545-10546.

54 World Nylon 6 & 66 Supply/Demand Report. http://www.thepcigroup.com/fibres/

55 Thiemens, M. H.; Trogler, W. C. Nylon Production - an Unknown Source of
Atmospheric Nitrous- Oxide. Science 1991, 251 , 932-934.

56 Reimer, R. A.; Slaten, C. S.; Seapan, M.; Koch, T. A.; Triner, V. G. Adipic acid
industry- NZO abatement implementation of technologies for abatement of N20 emission
associated with adipic acid manufacture. Non-CO2 Greenhouse Gases: Scientific
Understanding, Control and Implementation, Proceedings of the International Symposium,

2'”, Noordwijkerhout, Netherlands, sept. 8-10, 1999, 347-358.

57 Draths, K. M.; Frost, J. W. Environmentally Compatible Synthesis of Adipic Acid
from D-glucose. J. Am. Chem. Soc. 1994, 116, 399-400.

58 Niu, W.; Draths, K. M.; Frost, J. W. Benzene-Free Synthesis of Adipic Acid.
Biatechnal. Prag. 2002, 18, 201-211.

59 Neidle, E. L.; Ornston, L. N. Cloning and Expression of Acinetabacter
calcaaceticus Catechol 1,2-Dioxygenase Structural Gene Cata in Escherichia coli. J.
Bacterial. 1986, 168, 815-820.

60 (a) Aiba, S.; Tsunekawa, H.; Imanaka, T. New Approach to Tryptophan Production
by Escherichia coli: Genetic Manipulation of Composite Plasmids Invitro. Appl. Env.
Microbial. 1982, 43, 289-297. (b) Tsunekawa, H.; Imanaka, T.; Aiba, S. Phenotypic
Stability of Trp Operon Recombinant Plasmids in Escherichia coli. J. Gen. Microbial.
1980, 118, 253-261.

61 (a) Draths, K. M.; Pompliano, D. L.; Conley, D. L.; Frost, J. W.; Berry, A.;
Disbrow, G. L.; Staversky, R. J .; Lievense, J. C. Biocatalytic Synthesis of Aromatics from
D-Glucose: The Role of Transketolase. J. Am. Chem. Soc. 1992, 114, 3956-3962. (b)
Gubler, M., Jetten, M., Lee, S. H., Sinskey, A. J. Cloning of the Pyruvate-Kinase Gene
(Pyk) of Carynebacterium glutamicum and Site-Specific Inactivation of Pyk in a Lysine-
Producing Carynebacterium lactafermentum Strain. Appl. Env. Microbial. 1994, 60,
2494-2500. (c) Miller, J. E.; Backrnan, K.C.; O'Conner, M. J .; Hatch, R. T. Production of
Phenylalanine and Organic-Acids by Phosphoenolpyruvate Carboxylase-Deficient Mutants
of Escherichia coli. J. Ind. Microbial. 1987, 2, 143-149. ((1) Patnaik, R.; Liao, J. C.
Engineering of Escherichia coli Central Metabolism for Aromatic Metabolite Production
with Near Theoretical Yield. Appl. Environ. Microbial. 1994, 60, 3903-3908. (e) Patnaik,
R.; Spitzer, R. G.; Liao, J. C. Pathway Engineering for Production of Aromatics in
Escherichia coli: Confirmation of Stoichiometric Analysis by Independent Modulation of
AroG, TktA, and Pps Activities. Biotechnol. Bioeng. 1995, 46, 361-370. (f) Li, K.; Mikola,
M. R.; Draths, K. M.; Worden, R. M.; Frost, J. W. Fed-Batch Fermentor Synthesis of 3-
Dehydroshikimic Acid Using Recombinant Escherichia coli. Biotechnol. Bioeng. 1999, 64,

37

61—73.

62 Ogino, T.; Garner, C.; Markley, J. L. ; Herrmann, K. M. Biosynthesis of Aromatic-
Compounds - Cl 3 NMR-Spectroscopy of Whole Escherichia coli Cells. Prac. Natl. Acad.
Sci. USA 1982, 79, 5828-5832.

63 (a) Weaver, L. M.; Herrmann, K. M. Cloning of an AroF Allele Encoding a
Tyrosine-Insensitive 3-Deoxy-D-arabina-heptulosonate 7-phosphate Synthase. J.
Bacterial. 1990,172, 6581. (b) Mikola, M. R.; Widman, M. T.; Worden, R. M. In situ
mutagenesis and chemotactic selection of microorganisms in a diffusion gradient chamber.
Appl. Biochem. Biotechnol. 1998, 70-72, 905-918.

64 (a) Kikuchi, T.; Sotochi, N .; Fukase, K.; Kojima, H.; Kurahashi, O.; matsui, Y.
Aromatic amino acid manufacture enhancement with recombinant microorganism. JP Patent
04248983, 1993. (b) Tonouchi, N.; Kojima, H.; Matsui, H. The use of feedback-insensitive
enzymes in the production of aromatic amino acids by fermentation. Eur. Pat. Appl.
488424, 1992.

65 Ray, J. M.; Yanofsky, C.; Bauerle, R. Mutational Analysis of the Catalytic and
Feedback Sites of the Tryptophan-Sensitive 3-Deoxy-D-Arabina-Heptulosonate-7-
Phosphate Synthase of Escherichia coli. J. Bacterial. 1988, 170, 5500-5506.

66 Backman, K.C. US. Patent 5,169,768, 1992.
67 Mori, M.; Yokota, A. ; Sugitomo, S.; Kawamura, K. Patent JP 62,205,782, 1987.

68 (a) Konstantinov, K. B.; Nishio, N.; Yoshida, T. Glucose Feeding Strategy
Accounting for the Decreasing Oxidative Capacity of Recombinant Escherichia coli in Fed-
Batch Cultivation for Phenylalanine Production. Ferment. Bioeng. 1990, 70, 253-260. (b)
Konstantinov, K. B.; Nishio, N.; Seki, T.; Yoshida, T. Physiologically Motivated Strategies
for Control of the Fed-Batch Cultivation of Recombinant Escherichia cal i for
Phenylalanine Production. Ferment. Bioeng. 1991, 71, 350-355.

69 (a) Draths, K. M.; Frost, J. W. Synthesis Using Plasmid-Based Biocatalysis:
Plasmid Assembly and 3~Deoxy-D-Arabina-Heptulosonate Production. J. Am. Chem.
Soc. 1990, 112, 1657-1659. (b) Frost, J. W. US. Patent 5,168,056, 1992.

70 Williams, J. F.; Blackmore, P. F.; Duke, C. C. MacLeod, J. K. fact, uncertainty and
speculation concerning the biochemistry of D-erythrose-4-phosphate and its metabolic
roles. Int. J. Biachem.1980, 12, 339-344.

71 (a) Farabaugh, M. MS. Thesis, Michigan State University, April 1996. (b) Lu, J .—
L.; Liao, J. C. Metabolic engineering and control analysis for production of aromatics: Role
of transaldolase. Biotechnol. Bioeng. 1997,53, 132-138.

72 Niersbach, M.; Kreuzaler, F.; Geerse, R. H.; Postma, P. W.; Hirsch, H. J. Cloning
and nucleotide sequence of the Escherichia coli K-12 ppsA gene, encoding PEP synthase.
Mol. Gen. Genetics 1992, 231, 332-336.

73 Li, K.; Frost, J. W. Microbial Synthesis of 3-Dehydroshikimic Acid: A Comparative

Analysis of D-Xylose, L-Arabinose, and D-Glucose Carbon Sources. Biatechnal. Prag.
1999, 15, 876-883.

38

74 (a) Parker, C.; Barnell, W. O.; Snoep, J. L.; Ingram, L. 0.; Conway, T.
Characterization of the Zymamanas mabilis glucose facilitator gene product (glf) in
recombinant Escherichia coli: Examination of transport mechanism, kinetics and the role of
glucokinase in glucose transport. Mal. Microbial. 1995, 15, 795-802. (b) Snoep, J. L.;
Arfman, N.; Yomano, L. P.; Fliege, R. K.; Conway, T.; Ingram, L. 0. Reconstruction of
Glucose Uptake and Phosphorylation in a Glucose-Negative Mutant of Escherichia coli by
Using Zymamanas mabilis Genes Encoding the Glucose Facilitator Protein and
Glucokinase. J. Bacterial. 1994, 176, 2133-2135. (c) Weisser, P.; Kramer, R.; Sahm, H.;
Sprenger, G. A. Functional Expression of the Glucose Transporter of Zymamanas mabilis
Leads to Restoration of Glucose and Fructose Uptake in Escherichia coli Mutants and
Provides Evidence for Its Facilitator Action. J. Bacterial. 1995, 177, 3551-3554.

75 (a) Sprenger, G.; Sahm, H.; Karutz, M.; Sonke, T. Microbial Preparation of
Substances from Aromatic Metabolism/H. W0 98/ 18937, 1998. (b) Kraemer, M.; Karutz,
M.; Sprenger, G.; Sahm, H. Microbial Preparation of Substances from Aromatic
Metabolism/III. WO 99/55877, 1999. (c) Sprenger, G.; Siewe, R.; Sahm, H.; Karutz, M.;
Sonke, T. Microbial Preparation of Substances from Aromatic Metabolism/I. US. Patent
6,316,232 B1, 2001.

76 (a) Flores, N.; Xiao, J .; Berry, A.; Bolivar, F.; Valle, F. Pathway Engineering for the
Production of Aromatic Compounds in Escherichia coli. Nature Biotechnol. 1996, 14,
620-623. (b) Chen, R.; Yap, W. M. G. J.; Postma, P. W.; Bailey, J. E. Comparative Studies
of Escherichia coli Strains using Different Glucose Uptake Systems: Metabolism and
Energetics. Biotechnol. Bioeng. 1997, 56, 583-590. (c) Chen, R.; Hatzimanikatis, V.; Yap,
W. M. G. J.; Postma, P. W.; Bailey, J. E. Metabolic Consequences of Phosphotransferase
(PT S) Mutation in a Phenylalanine-Producing Recombinant Escherichia coli. Biatechnal.
Prag. 1997, 13, 768-775. (d) Baez, J. L.; Bolivar, F.; Gosset, G. Determination of 3-Deoxy-
D-Arabina-Heptulosonate 7-Phosphate Productivity and Yield from Glucose in Escherichia
coli Devoid of the Glucose Phosphotransferase Transport System. Biotechnol. Bioeng.
2001, 73, 530-535. (6) Flores, S.; Gosset, G.; Flores, N.; de Graaf, A. A.; Bolivar, F.
Analysis of Carbon Metabolism in Escherichia coli Strains with an Inactive
Phosphotransferase System by 13C Labeling and NMR Spectroscopy. Metabal. Eng.
2002, 4, 124-137.

77 Li, K.; Frost, J. W. Utilizing Succinic Acid as a Glucose Adjunct in Fed-Batch
Fermentation: Is Butane a Feedstock Option in Microbe-Catalyzed Synthesis? J. Am.
Chem. Soc. 1999, 121 , 9461-9462.

78 Oliver, D. J .; Nikolau, B.; Wurtele, E. S. Functional Genorrrics: High-Throughput
mRNA, Protein, and Metabolite Analyses. Met. Engin. 2002, 4, 98-106.

79 Brown, P. O.; Botstein, D. Exploring the new world of the genome with DNA
microarrays. Nat. Genet. 1999, 21, 33-37.

80 (a) Eisen, M. B.; Brown, P. 0. DNA arrays for analysis of gene expression.
Methods Enzymal. 1999, 303, 179-205. (b) Lipshutz, R. J .; Fodor, S. P.; Gingeras, T. R.;
Lockhart, D. J. High density synthetic oligonucleotide arrays. Nat. Genet. 1999, 21, 20-24.

81 Lease, R. A.; Belfort, M. A trans-acting RNA as a control switch in Escherichia

coli: DsrA modulates function by forming alternative structures. Prac. Natl. Acad. Sci. U.
S. A. 2000, 97, 9919-9924.

39

82 (a) Zheng, M.; Wang, X.; Templeton, L. J .; Smulski, D. R.; LaRossa, R. A.; Storz,
G. DNA rrricroarray-mediated transcriptional profiling of the Escherichia coli response to
hydrogen peroxide. J. Bacterial. 2001, 183, 4562-4570. (b) Tucker, D. L.; Tucker, N.;
Conway, T. Gene expression profiling of the pH response in Escherichia coli. J. Bacterial.
2002, 184, 6551-6558.

83 Yoon, S. H.; Han, M.-J.; Lee, S. Y.; Jeong, K. J.; Yoo, J .-S. Combined transcriptome
and proteome analysis of Escherichia coli during high cell density culture. Biotechnol.
Bioeng. 2003, 81, 753-767.

84 Gonzalez, R.; Tao, H.; Shanmugam, K. T.; York, S. W.; Ingram, L. 0. Global Gene
Expression Differences Associated with Changes in Glycolytic Flux and Growth Rate in
Escherichia coli During the Fermentation of Glucose and Xylose. Biotechnology Progress
2002, 18, 6-20.

85 Gonzalez, R.; Tao, H.; Purvis, J. E.; York, S. W.; Shanmugam, K. T.; Ingram, L. 0.
Gene Array-Based Identification of Changes That Contribute to Ethanol Tolerance in
Ethanologenic Escherichia coli: Comparison of Koll (Parent) to Ly01 (Resistant Mutant).
Biatechnal. Prag. 2003, 19, 612-623.

86 Sabatti, C.; Rohlin, L.; Oh, M.-K.; Liao, J. C. Co-expression pattern from DNA
microarray experiments as a tool for operon prediction. Nucl. Acid Res. 2002, 30, 2886-
2893.

87 (a) Tao, H.; Bausch, G; Richmond, C.; Blattner, F. R.; Conway, T. Functional
genorrrics: expression analysis of Escherichia coli growing on minimal and rich media. J.
Bacterial. 1999, 181, 6425-6440.

88 Richmond, C. S.; Glasner, J. D.; Mau, R.; Jin, H.; Blattner, F. R. Genome-wide
expression profiling in Escherichia coli K-12. Nucl. Acid Res. 1999, 2 7, 3821-3839.

89 Oh, M.-K.; Liao, J. C. In Book of Abstracts, 219th ACS National Meeting, San
Francisco, CA, March 26-30, 2000, 2000, pp BIOT-161.

90 Yi, J .; Li, K.; Draths, K. M.; Frost, J. W. Modulation of Phosphoenolpyruvate
Synthase Expression Increases Shikimate Pathway Product Yields in E. coli. Biatechnal.
Prag. 2002, 18, 1141-1148.

91 (a) Hoischen, C.; Kramer, R. Evidence for an Efflux Carrier System Involved in the
Secretion of Glutamate by Carynebacterium glutamicum. Arch. Microbial. 1989, 151,
342-347. (b) Kramer, R. Secretion of Amino-Acids by Bacteria - Physiology and
Mechanism. FEMS Microbial. Rev. 1994,13,75-93.

92 (a) Bellmann, A.; Vrljic, M.; Patek, M.; Sahm, H.; Kramer, R.; Eggeling, L.
Expression Control and Specificity of the Basic Amino Acid Exporter LysE of
Carynebacterium glutamicum. Microbialagy-ng 2001, 147, 1765-1774. (b) Vrljic, M.;
Sahm, H.; Eggeling, L. A New Type of Transporter with a New Type of Cellular Function:
L-Lysine Export from Carynebacterium glutamicum. Mal. Microbial. 1996, 22, 815-826.
(c) Vrljic, M.; Kronemeyer, W.; Sahm, H.; Eggeling, L. Unbalance of L-Lysine Flux in
Corynebacterium glutamicum and Its Use for the Isolation of Excretion-Defective
Mutants. J. Bacterial. 1995, 177, 4021-4027.

40

 

93 (a) Kennerknecht, N .; Sahm, H.; Yen, M. R.; Patek, M.; Saier, M. H.; Eggeling, L.
Export of L-Isoleucine from Carynebacterium glutamicum: A Two Gene-Encoded
Member of a New Translocator Family. J. Bacterial. 2002, 184, 3947-3956. (b)
Ebbighausen, H.; Weil, B.; Kramer, R. Isoleucine Excretion in Carynebacterium-
Glutamicum - Evidence for a Specific Efflux Carrier System. Appl. Microbial. Biotech.
1989, 31, 184-190. (c) Hermann, T .; Kramer, R. Mechanism and Regulation of Isoleucine
Excretion in Corynebacterium glutamicum. Appl. Environ. Microbial. 1996, 62, 3238-
3244. (d) Morbach, S.; Sahm, H.; Eggeling, L. L-Isoleucine Production with
Carynebacterium glutamicum: Further Flux Increase and Limitation of Export. Appl.
Environ. Microbiol.1996, 62, 4345-4351.

94 (a) Simic, P.; Willuhn, J.; Sahm, H.; Eggeling, L. Identification of Glya (Encoding
Serine Hydroxymethyltransferase) and Its Use Together with the Exporter ThrE to Increase
L-Threonine Accumulation by Carynebacterium Glutamicum. Appl. Environ. Microbial.
2002, 68, 3321-3327. (b) Sirrric, P.; Sahm, H.; Eggeling, L. L-Threonine Export: Use of
Peptides to Identify a New Translocator from Corynebacterium glutamicum. J. Bacterial.
2001, 183, 5317-5324. (c) Palrrrieri, L.; Bems, D.; Kramer, R.; Eikmanns, M. Threonine
Diffusion and Threonine Transport in Corynebacterium glutamicum and Their Role in
Threonine Production. Archives of Microbiology 1996, 165, 48-54.

95 Dassler, T.; Maier, T.; Winterhalter, C.; Bock, A. Identification of a Major Facilitator
Protein from Escherichia Cali Involved in Efflux of Metabolites of the Cysteine Pathway.
Mal. Microbiol. 2000, 36, 1101-1112.

96 Zakataeva, N. P.; Aleshin, V. V.; Tokmakova, I. L.; Troshin, P. V.; Livshits, V. A
The Novel Transmembrane Escherichia coli Proteins Involved in the Amino Acid Efflux.
FEBS Letters 1999, 452, 228-232.

97 (a) Liu, J. Y.; Miller, P. F.; Gosink, M.; Olson, E. R. The Identification of a New
Family of Sugar Efflux Pumps in Escherichia coli. Mal. Microbial. 1999, 31, 1845-1851.
(b) Carole, S.; Pichoff, S.; Bouche, J. P. Escherichia coli Gene YdeA Encodes a Major
Facilitator Pump Which Exports L-Arabinose and Isopropyl-Beta—D-
Thiogalactopyranoside J. Bacterial. 1999, 181, 5123-5125.

41

CHAPTER 2

Availability of phosphoenolpyruvate and biosynthesis of DHS from glucose

Introduction

3-Dehydroshikimic acid (DHS) is an important intermediate in benzene-free
syntheses of a variety of industrial chemicals: commodity Chemicals (adipic acid,I
phenolz), pseudocommodity chemicals (catechol,3 hydroquinone,4 p-hydroxybenzoic
acids), fine chemicals (vanillin,6 indigo7), and ultrafine chemicals (gallic acid,8
pyrogallole). Strategies developed to improve the synthesis of DHS can thus be applied
to the microbial synthesis of a wide range of molecules. In addition, DHS can be used as
a potent antioxidantf’

The key step for the biosynthesis of DHS from glucose is the first step in the
common pathway of aromatic amino acid biosynthesis: condensation of
phosphoenolpyruvate (PEP) with D-erythrose 4-phosphate (E4P) catalyzed by 3-deoxy-D-
arabina-heptulosonic acid 7-phosphate (DAHP) synthase (Figure 14). PEP is drived
from the Embden-Meyerhof pathway (glycolysis) and E4P is derived from pentose
phosphate pathway. Overexpression of DAHP synthase isozymes insensitive to feedback
inhibition formed the initial basis for increasing the carbon flow directed into the
shikimate pathway.10 The in vivo availability of PEP and E4P limit the catalytic activity
of overexpressed, feedback-insensitive DAHP synthase in Escherichia coli K-l2.ll
Access to E4P is increased upon overexpression of tktA-encoding transketolase.12 With
improved E4P availability, the availability of PEP becomes a key limiting factor in DHS

biosynthesis as discussed in Chapter 1.[3

42

ACOZH

pyruvate
ATP
glucose aHzosPO AMP + P,
%COZH

PEP?» Ho,, COZH
OH"\O/'/'
HZO3PO OH \ R0 0”
esp H R: H DAH
H203P0 9“ R: P03H2 DAHP
E4P
COZH COzH H01 COQH
DHS DHQ
3y
COZH
Ho“' 5 OH
OH
SA

Figure 14. Synthesis of DHS from D- glucose. Key: Enzymes (genes): (a)
PEP. carbohydrate phosphotransferase system, (b) PEP synthase (ppsA), (c) DAHP
synthase (araFFBR), (d) DHQ synthase (araB), (e) DHQ dehydratase (araD), (f) shikimate
dehydrogenase (araE), (g) uncharacterized. Intermediates (abbreviations); D-glucose 6-
phosphate (G6P), phosphoenolpyruvate (PEP), D-erythrose 4-phosphate (E4P), 3-deoxy—
D-arabina-heptulosonic acid (DAH) 7-phosphate (DAHP), 3-dehydroquinic acid (DHQ),
3-dehydroshikirnic acid (DHS), gallic acid (GA), shikimic acid (SA).

PEP is a key intermediate involved in several cellular processes such as
phosphoenolpyruvate:carbohydrate phosphotransferase system (PTS), aromatic amino
acid biosynthesis, glycolysis, 3-deoxy-D-manna-octulosonate biosynthesis and
peptidoglycan biosynthesis. When wild-type E. coli grows on glucose, the major
consumer of PEP is the PTSl4 system responsible for the simultaneous transport and
phosphorylation of glucose (Figure 12, Chapter 1). One molecule of PEP is converted
into pyruvic acid for each molecule of glucose transported into the cytoplasm. PTS-

generated pyruvic acid is oxidized through the TCA cycle to carbon dioxide and

apparently not recycled to PEP under aerobic culture conditions where glucose is the sole

43

carbon source. As a consequence, stoichiometric analysis (Figure 15A, discussed in
detail in the next section) indicated that the maximum theoretical molar yield for DAHP
synthesis from glucose without recycling of pyruvic acid to PEP is equal to 0.43
mol/mol.” In theory, this 43% maximum yield can be doubled (Figures 15B and 15C,
discussed in detail in the next section) if a strain can recycle PTS-generated pyruvic acid
back to PEP (the first strategy) or transport glucose without coupling this process to PEP
utilization (the second strategy). To evaluate the impact of increased PEP availability on
DHS biosynthesis, this study examined these two basic strategies to alleviate PEP
limitation in constructs cultured under fed-batch fermentor conditions.

The first strategy employed overexpressed ppsA-encoded PEP synthase (Figure
12, Chapter 1) to recycle the pyruvic acid generated by PTS-mediated glucose transport
to PEP. PEP synthase converts pyruvic acid to PEP along with conversion of ATP to
AMP and one molecule of inorganic phosphate (Figure 12, Chapter 1).16 In a second
strategy, expenditure of PEP during glucose transport was partially or completely avoided
by a non-PTS glucose transport mechanism (Figure 12, Chapter 1). Non-PTS glucose
transport includes facilitated diffusion mediated by the Z. mabilis glf-encoded glucose
facilitator17 and galactose-proton symport mediated by the E. coli galP-encoded galactose
permease (Figure 12, Chapter 1).18 Glf is specific for glucose and has an apparent Km for
glucose of approximately 1.1-2.9 mM, consistent with the function of Glf as a low-
affinity glucose facilitator.l7 GalP is a member of the major facilitator family (MFS) that
can function by solute/H” symport.‘9 Constructs were assembled where PTS-mediated
glucose transport was replaced or augmented with heterologous expression of the

Zymamanas mabilis glf-encoded glucose facilitator protein. Alternatively, PTS-mediated

44

glucose transport is replaced with the E. coli galP-encoded galactose permease. Genomic
E. coli glk-encoded glucokinase alone or in combination with plasmid-localized Z.
mabilis glk-encoded glucokinase phosphorylates Glf- or GalP—transported glucose to
glucose-6-phosphate using ATP as the phosphor donor. Based on the yield of the same
product and byproduct mixture synthesized by E. coli constructs cultivated under a
uniform set of fermentor—controlled conditions, the impact of strategies employed to

alleviate PEP limitation was evaluated.

Theoretic maximum yield calculation

The theoretic maximum yield is calculated based on the assumption that the
branching pathways are blocked and that the carbon flow is directed by the most efficient
pathway with minimum loss to carbon dioxide and other metabolites. Under these
conditions, the relative flux through each step at the steady state can be calculated by
balancing the input (production) and output (consumption) fluxes from each metabolite
pool. For instance, the input flux for glyceraldehydes 3-phosphate is 12 mol (6 mol + 6
mol) while the output flux for glyceraldehyde 3-phosphate is also 12 mol (10 mol + 2
mol) in Figure 15A. As shown in Figure 15A, if a strain transports glucose by PTS
without overexpression of PEP synthase, the input of 7 mol of glucose can produce 3 mol
of DAHP (43% molar yield) and 7 mol of pyruvate, which is further metabolized. To
avoid the waste of pyruvate when glucose is used as the carbon source, two approaches
are possible: recycling pyruvate to PEP (Figure 15B) or isolation of PTS mutants that use
ATP as the phosphate donor (Figure 15C). These approaches all allow 100% theoretical

carbon yield, or 86% theoretical molar yield of DAHP (6 mol) from glucose (7 mol).

45

Glucose

 

 

 

 

GAP S7P

1
E4P F6P-//

DAHP synthase

DAHP

1

common pathway

(B) Glucose

 

 

 

 

 

 

\

Ribulose-SP
2 / \2

X5P RSP
2 Tkl

GAP

S7P
2 Ta
1
E4P F6P

 

DAHP synthase

DAHP

1

common pathway

46

 

(C)

 

 

 

Glucose
3: \
Ribulose-SP
7i 2/ 2
F6P X5P R5P
l 2 Tkt
5
1,6 FDP GAP S7P
f/ \i 2 Ta] 2
DHAP -—5-> GAP E4P F6P j

6

DAHP synthase

DAHP

l

common pathway

Figure 15. Theoretical flux distribution for directing carbon into common pathway
with glucose as the carbon source. (A) Without recycling pyruvate to PEP. (B)
Recycling pyruvate to PEP using PEP synthase (Pps). (C) Using non-PTS glucose
transport. The numbers are the relative fluxes needed to convert 7 mol of glucose into
DAHP. Enzymes: Pps, PEP synthase; Tkt, transketolase; Tal, transaldolase.
Intermediates (abbreviations): G6P, glucose 6-phosphate; F6P, fructose 6-phosphate;
1,6FDP, 1,6-fructose diphosphate; DHAP, dihydroxyacetone phosphate; GAP,
glyceraldehyde 3-phosphate; R5P: ribose 5-phosphate; X5P, xylulose 5- phosphate; S7P,
sedoheptulose 7-phosphate; PYR, pyruvate.

The relative flux through each intermediate step is also shown in Figures 15A,
15B and 15C. For instance, 2 (or 4) mol out of 3 (or 6) mol input E4P is coming from the
transketolase reaction by coupling Embden-Meyerhof pathway intermediate fructose-6-
phosphate and glyceraldehydes 3-phosphate (Figures 15A, 15B and 15C). In reality, this

is not likely to happen without overexpression of transketolase.

47

Modulation of PEP synthase expression increases shikimate pathway product yields
Biocatalyst Design

E. coli KL3/pJY1.216A was constructed to study the impact of
phosphoenolpyruvate synthase expression on the synthesis of DHS. E. coli KL3ll carried
an araE mutation resulting in the absence of shikimate dehydrogenase (Figure 14). In
addition, a second araB locus encoding 3-dehydroquinate synthase was inserted into its
serA locus encoding 3-phosphoglycerate dehydrogenase. The insertion had two
functions: increased 3-dehydroquinate synthase resulting from the two genomic araB loci
in E. coli KL3 had been previously observed to eliminate formation of DAH as a
byproduct when carbon flow directed into the shikimate pathway was increased;20
disruption of the genomic serA locus destroyed the ability of the E. coli KL3 to
synthesize L-serine and served as a nutritional pressure for maintenance of serA localized
plasmid in minimal salts medium.

In addition to the serA locus, plasmid pJY1.216A (Figure 16C) contained ppsAl6

encoding PEP synthase directly behind a tac promoter (P ) and ribosomal binding site.

As the lac]Q gene encoding lac repressor protein was also localized in the plasmid
pJY1.2l6A, the expression of PEP synthase was controlled by the concentration of

isopropyl B-D-thioglucopyranoside (IPTG) added to the culture medium. Plasmid

pJY1.216A also contained the araFFBR encoding feedback-insensitive

48

 

pKL4.668
$1) BamHI and Sa/l digest
BarnHl Safl

L47 ‘L3kb l
k

 

aroFFBR

1) BamHI and Sa/l digest
2) CIAP treatment

T4 Ligase

PCR ppsA ORF from R8791 genomic DNA
*1) EcoRl digest

 

EcoRl EcoRl
I 2.4 kb I
ppsA
1) EcoRl digest
2) CIAP treatment

T4 Ligase

 

Figure 16A. Construction of plasmid pJY1.131A and pJY1.143A

49

    
   
      
 

BamHl
Smal
EcoRl
pr1.143A PCR from pMF63A
ppsA 9.0 kb 1) BamHl digest
BamHI BamHl
0.15 kb
EcoRl Pia ParoF
1) BamHldigest
2) CIAP treatment
T4 Ligase

p02625

pJY1.207A _
‘1) Oral and EcoRV dIgest

9.2 kb

ppsA
Dr I EcoRV
1.9 kb
serA
Ptac
ECOR' 1) Smal digest
2) CIAP treatment
T4 Ligase

pJY1.211A

11.1 kb

 

Figure 16 B. Construction of plasmid pJY1.207A and pJY1.211A

3-deoxy-D-arabina-heptulosonic acid 7-phosphate (DAHP) synthase, an additional copy

of promoter region of araF (P ), and the tktA encoding transketolase. In the absence of

araF

50

   

pJY1.211A

1 1 .1 kb
/ 2.2 kb
(BamHl) (BamHl)
,3, | WA L

1) Hindlll digest
2) Klenow treatment
3) CIAP treatment

  

pMF51A

1) BamHl digest
2) Klenow treatment

  

 
 
  

 

T4 Ligase

(Hindlll)

 

 

S |
( ma)ECOR| PPSA EcoRl

Figure 16C. Construction of plasmid‘pJY 1.216A
shikimate dehydrogenase activity, aromatic amino acids and aromatic vitamins had to be

added to the culture medium in order for 3-dehydroshikimate-synthesizing constructs to
grow. Plasmid-localization of a feedback-insensitive isozyme of DAHP synthase
encoded by araFFBR in all 3-dehydroshikimate-synthesizing constructs was thus essential
to circumvent feedback inhibition caused by the aromatic amino acid supplements.

Plasmid-localized P has three binding sequences for the TyrR repressor protein that

aroF

represses transcription of araFFBR,2' and its inclusion on a plasmid had been shown to

titrate away the cellular supply of TyrR and increased expression of DAHP synthase.ll

51

E4P (Figure 14) availability was increased by the expression of plasmid-localized tktA

encoding transketolase.12

Fermentation conditions

The impact of expression level of PEP synthase in E. coli KL3/pJYl .216A on the
yields and concentrations of DHS and shikimate pathway byproducts synthesized under
fed-batch fermentor conditions was examined under both glucose-rich conditions and
under glucose-limited conditions in a 2.0-L working volume fermentor. A concentration
range of 55-170 mM glucose was maintained under glucose—rich conditions while a
steady state concentration of approximately 0.2 mM glucose was maintained under
glucose-limited conditions. According to the literature, reductions in product yields
associated with excessive CO2 generation are minimized under glucose-rich conditions.22
However, glucose-rich conditions can lead to excessive generation of acetic acid, which
is toxic to E. coli and many other microbes.22 Glucose-limited conditions minimize
generation of acetic acid but can lead to excessive CO2 generation resulting in lower
product yields.22 Thus both glucose—rich and glucose—limited conditions were employed
to find the glucose concentration leading to the synthesis of the highest possible yield of
DHS and shikimate pathway byproducts. Temperature was maintained at 36 °C, pH was
maintained at 7.0 and dissolved oxygen was maintained at 20% air saturation under both
glucose-rich and glucose—limited conditions. All the fermentations were run in duplicate

and reported results represent an average of two runs.

52

DHS and shikimate pathway byproducts synthesized by E. coli KL3/pJY 1.216A
under glucose-rich conditions

E. coli KL3/pJY1.216A was first tested under glucose-rich conditions. The
dissolved oxygen was maintained at 20% air saturation by varying the impeller speed. A
glucose concentration of 55-170 mM was maintained by manually adjusting the glucose
feeding rate. The fermentation runs stopped at 42 h after inoculation of the culture
medium because DHS synthesis did not continue beyond 42 h. To induce the expression
of PEP synthase, the indicated quantities of IPTG were added at 18, 24, 30, and 36 h after
inoculation of the culture medium. Aliquots of the culture medium at certain time points
were removed and concentrations of DHS, 3-dehydroquinic acid, gallic acid, 3—deoxy-D-
arabino-heptulosonic acid (DAH), and acetic acid were determined by lH NMR using
response factors determined with authentic samples of all molecules (Table l). The
specific activities of PEP synthase and DAHP synthase were also measured from aliquots
removed from the fermentor at 24, 36, and 42 h after inoculation of the culture medium
(Table 2).

At 24 h, PEP synthase specific activities varied from 0.0093 U/mg (Entry 1, Table
2) when no IPTG was added to 0.15 U/mg (Entry 6, Table 2) when 48 mg IPTG was
added at 6 h intervals. As expected, PEP synthase activities were higher with addition of
more IPTG. For comparison, DAHP synthase specific activities did not vary over more
than 2-fold range as a function of IPTG addition, indicating that increased PEP synthase
activities did not negatively impact DAHP synthase activities. For all IPTG
concentrations examined, both PEP synthase specific activity and DAHP synthase

specific activity declined over the course of the fermentation runs after 24 h.

53

With no IPTG addition and wild-type PEP synthase activity (0.0093 U/mg at 24
h), 52 g/L of DHS was synthesized in 28% (mol/mol) yield from glucose by E. coli

KL3/pJYl.2l6A (Entry 1, Table l) and provided the basis for evaluating the impact on

Table 1. Concentrations and yields of DHS and byproducts synthesized by E. coli
KL3/pJY1.216A cultured under glucose-rich conditions.

 

Dry cell

 

Entry IPTG [DHS]" DHS [DAH] [DHQ] [GA] Total weight [acetate]

addn."" (g/L) Yieldc (g/L) (g/L) (g/L) Yield“ (9,” (g/L)
1 0 52 28% 3.1 7.3 7.1 37% 21 0.30
2 6.0 58 31% 11 8.9 9.1 45% 21 0.30
3 12 69 35% 10 12 13 51% 21 0.30
4 18 62 34% 1 1 9.8 8.0 48% 21 0.30
5 24 66 34% 9.0 9.8 10 47% 22 0.50
6 48 58 29% 12 5.6 7.0 40% 21 0.80

 

“Amount (mg) of isopropyl B-D-thioglucopyranoside (IPTG) added at 18, 24, 30, and 36
h. b Abbreviations: 3-dehydroshikimic acid (DHS), 3-deoxy-D-arabina-heptulosonic acid
(DAH), 3-dehydroquinic acid (DHQ), gallic acid (GA). C (mol DHS)/(mol glucose
consumed). d (mol DHS + mol DAH + mol DHQ + mol GA)/ (mol glucose consumed).

Table 2. PEP synthase and DAHP synthase specific activities for E. coli
KL3/pJY1.216A cultured under glucose-rich conditions.

 

 

 

Ent IPTG PEP synthase (U/mg)" DAHP synthase (U/mg)‘
ry addn-a 24 h 36 h 42 h 24 h 36 h 42 h
1 O .0093 .0061 .0042 .40 .32 .17
2 6.0 .031 .018 .015 .76 .27 .13
3 12 .069 .058 .049 .53 .27 .19
4 18 .081 .058 .046 .91 .41 .28

5 24 .1 O .069 .031 .58 .44 .28

6 48 .15 .13 .10 .87 .62 .20

 

" Amount of IPTG (mg) added at 18, 24, 30, and 36 h. b One unit (U) of PEP (PEP)
synthase corresponds to consumption of 1 umole of pyruvate per nrin at 30 °C. C One unit
(U) of DAHP synthase corresponds to formation of 1 umole of DAHP per nrin at 37 °C.

the concentration and yield of synthesized DHS attendent with overexpression of PEP

synthase. As PEP synthase specific activities increased from 0.0093 U/mg to 0.15 U/mg,

higher concentration and yield of DHS was synthesized. Both the concentration and

54

yield of DHS reached an apparent maximum and then decreased (Entries 1-6, Table 1).
With 12 mg of IPTG addition and 7-fold increase in PEP synthase specific activity (0.07
U/mg) at 24 h, 69 g/L of DHS was synthesized in 35% (mol/mol) yield from glucose
(Entry 3, Table 1). However, with 48 mg of IPTG addition and lS-fold increase in PEP
synthase specific activity (0.15 U/mg) at 24 h, a significantly lower concentration of DHS
(58 g/L) was synthesized in reduced yield (29%, mol/mol) from glucose (Entry 6, Table
1) relative to the concentrations and yields of DHS synthesized with 12 mg of IPTG
addition (Entry 3, Table 1).

In addition to DHS, shikimate pathway byproducts DAH, 3-dehydroquinic acid,
and gallic acid also accumulated in the culture medium (Table 1). During a typical
fermentation run of E. coli KL3/pJY1.216A (Figure 17), the concentration of DAH
reached the maximum at 30 h, the concentration of 3-dehydroquinic acid reached the
maximum at 36 h and the concentration of gallic acid reached the maximum at the end of
the fermentation run 42 h. The extracellular accumulation of byproducts DAH and 3-
dehydroquinic acid (Figure 14) generated from the accumulated DAHP and 3-
dehydroquinic acid in vivo indicated that the turnover rate of 3-dehydroquinate synthase
and 3—dehydroquinate dehydratase were unable to keep up with the rate of increased
carbon flow into the shikimate pathway.ll Gallic acid (Figure 14) is derived from DHS,
although the route for this transformation remains to be discovered. Possibilities include
either direct oxidation of DHS or dehydration of this hydroaromatic to protocatechuic
acid followed by hydroxylation. As a function of PEP synthase specific activity, the
biosynthesized concentrations of DAH, 3-dehydroquinic acid, and gallic acid showed

similar trends to that of biosynthesized DHS (Table 1, Figure 17).

55

The total yield of shikimate pathway products was based on the combined yields
of DHS, DAH, 3-dehydroquinic acid, and gallic acid. At the optimum level of PEP
synthase expression, 69 g/L DHS, 10 g/L DAH, 12 g/L 3-dehydroquinic acid, and 13 g/L
gallic acid were synthesized in 51% total yield (Entry 3, Table 1) by E. coli

KL3/pJYl.216A under glucose-rich conditions. The maximum concentration of gallic

 

 

 

 

 

 

 

 

 

 

 

 

A. B.
A80 A20
3370 —— g
EGO- _ #516,.
a) G
7:50 ‘5
5 312..
540- <‘
o (5
1€30 -- d 8--
8:204 H ° ' ° ‘8
z - 4..
010-- 0 I
<
0 e : e 1".“ 1 fl: - 1- 1 1 D 0 ._
0 6 12 18 24 30 35 42 0 6 12 18 24 30 36 42

Time (11) Time (11)

Figure 17. E. coli KL3/pJY1.216A cultured under glucose-rich conditions with 12
mg of IPTG added at 18, 24, 30, and 36 h. Legend: (A) DHS (open bars), dry cell
weight (black circles). (B) 3-deoxy-D-arabina-heptulosonic acid (black bars), 3-
dehydroquinic acid (open bars), gallic acid (hashed bars), acetic acid (open circles).
acid produced (Entry 3, Table 1) by E. coli KL3/pJY1.216A approximates that reported
for E. coli KL7/pSK6.16l that was specifically designed to synthesize gallic acid with
overexpression of both 3-dehydroshikimate dehydratase and protocatechuate
hydroxylase.8

E. coli KL3/p1 Y1.216A also synthesized relatively low concentrations of acetate
(0.3 g/L — 0.8 g/L, Table 1) under glucose-rich conditions. When PEP synthase

expression was increased beyond its optimal level for biosynthesis of DHS, acetate

accumulation was observed to increase slightly (Entries 5 and 6, Table 1). However, the

56

increase of acetate concentration did not affect the cell growth as measured by dry cell
weight (Entries 1 — 6, Table 1).

The cell mass was approximately 21 g/L at the end of all the fermentation runs
(Entries 1 — 6, Table 1). The control profiles of all the fermentation runs were very
similar. The same amount of glucose (approximately 1.4 mol) was consumed for all the
fermentation runs of E. coli KL3/pJYl.2l6A under glucose-rich conditions, indicating
that increasing the expression level of PEP synthase didn’t increase the glucose

consumption rate for E. coli KL3/pJY1.2l6A.

DHS and shikimate pathway byproducts synthesized by E. coli KL3/pJY1.216A
under glucose-limited conditions

E. coli KL3/pJYl.216A was also tested under glucose-limited conditions.
Automatic oxygen—sensor-controlled glucose feeding was used to maintain a steady-state
low glucose concentration. The dissolved oxygen in the medium was also maintained at
20% air saturation. Under glucose-limited conditions, the fermentation runs stopped at
48 h after inoculation of the culture medium because DHS synthesis did not continue
beyond 48 h. The indicated quantities of IPTG were added 18, 24, 30, 36 and 42 h after
inoculation of the culture medium. Aliquots of the culture medium at certain time points
were removed and the concentrations of DHS and shikimate pathway byproducts were
determined by 'H NMR (Table 3). The specific activities of PEP synthase and DAHP
synthase in these aliquots were also measured from the fermentor at 24, 36, and 42 h after

inoculation of the culture medium (Table 4).

57

 

 

Table 3. Concentrations and yields of DHS and byproducts synthesized by E. coli
KL3/pJY1.216A cultured under glucose-limited conditions.

 

 

Em IPTG [DHS]” DHS [DAH] [DHQ] [GA] Total 3310:? acetate]
9’ addnf' (g/L) Yieldc (g/L) (g/L) (g/L) Yield" (g/QL) (g/L)

1 0 36 20% 0 4.7 4.5 25% 22 0.10

2 12 52 29% 4.6 6.8 11 41% 22 0.10

3 24 48 27% 3.3 6.9 8.4 36% 22 0.10

 

" Amount of IPTG (mg) added at 18, 24, 30, and 36 h. b Abbreviations: 3-
dehydroshikimic acid (DHS), 3-deoxy-D-arabina-heptulosonic acid (DAH), 3-
dehydroquinic acid (DHQ), gallic acid (GA). C (mol DHS)/(mol glucose consumed).
d(mol DHS + mol DAH + mol DHQ + mol GA)/ (mol glucose consumed).

Table 4. PEP Synthase and DAHP synthase specific activities for E. coli
KL3/pJY1.216A cultured under glucose-limited conditions.

 

 

 

Ent IPTG PEP synthase (U/mg)” DAHP synthase (U/mg)c
ry alddn-a 24 h 36 h 42 h 24 h 36 h 42 h
1 0 .0050 .0060 .0047 .25 .16 .07
2 12 .067 .045 .036 .65 .42 .15
3 24 .12 .069 .12 .29 .33 .13

 

" Amount of IPTG (mg) added at 18, 24, 30, and 36 h. b One unit (U) of PEP (PEP)
synthase corresponds to consumption of l umole of pyruvate per rrrin at 30 °C. C One unit
(U) of DAHP synthase corresponds to formation of 1 umole of DAHP per min at 37 °C.
As a function of IPTG addition, the specific activities (Table 4) of both PEP
synthase and DAHP synthase measured under glucose-limited conditions were
comparable to those measured under glucose-rich conditions (Table 2). However, E. coli
KL3/pJY1.216A consumed significant less glucose under glucose-limited conditions
(approximately 1.2 mol) than under glucose-rich conditions (approximately 1.4 mol).
Both the concentrations and the yields of DHS synthesized by E. coli KL3/pJY1.21 6A
under glucose-limited conditions were lower than the concentrations and yields
synthcsized under glucose-rich conditions at comparable levels of PEP synthase

express ion. With no IPTG addition and wild-type PEP synthase activity, E. coli

Kimmy 1.216A synthesized 36 g/L of DHS in 20% (mol/mol) yield from glucose (Entry

58

1, Table 3). DAH was not detected in the medium, although both 3-dehydroquinic acid
(4.7 g/L) and gallic acid (4.5 g/L) accumulated, and the total yield was 25%. When PEP
synthase expression was increased approximately 10-fold with 12 mg IPTG addition, 52
g/L of DHS was synthesized in 29% yield (mol/mol) from glucose (Entry 2, Table 3).
Accumulation of DAH was once again observed and the total yield was 41% (mol/mol).
Consistent with the observation under glucose-rich conditions, increasing PEP synthase
expression beyond this level (approximately 0.07 U/mg at 24 h) had a negative impact on
shikimate pathway product synthesis. When PEP synthase expression was increased
approximately 20-fold with 24 mg IPTG addition (Entry 3, Table 3), 48 g/L of DHS was
synthesized in 27% (mol/mol) yield from glucose, and the total yield decreased to 36%
(mol/mol). As expected, E. coli KL3/p1 Y1.216A synthesized significantly lower acetate
concentrations (0.1 g/L, Entries 1 - 3, Table 3) under glucose-limited conditions than the
acetate concentrations synthesized under glucose-rich conditions (0.3 - 0.8 g/L, Table 1).
The cell mass at the end of the fermentation runs was approximately the same both under

glucose-rich conditions and under glucose-limited conditions.

Altered glucose transport and shikimate pathway product yields in E. coli

Biocatalyst design

The host strain E. coli KL3 used for synthesis of DHS was described in the

previous section.” E. coli KL3 utilizes the wild-type phosphoenolpyruvate:carbohydrate

phosphotransferase system (PTS) to transport and phosphorylation of glucose. E. coli

59

 

 

pKor1.291A
H N E ESa X XSH
I l l J l l 111
— —> <— ‘—4—
CmR Plac aroFFBR serA Pam;

pKL5.17A
H N E ESa X XS(H) (H)
I 1 l J l 1 Jll 1

 

_ _. ‘— ‘_‘_
CmR Plac aroFFBR serA Pamp tktA

 

pKL6.239A
H N E ESa x X(S) (S)H
l 1 I I l 141 l J

CmR Plac aroFFBR serA ParoF Ptacglt

pKL7J15A
H N E ESa X X(S) (S) (H) (H)
l I I l l I II I I a

—— —><— ._.____,

—’
CmR Plac aroFFBR serA Pam; Ptacg" tktA

pKL7.85A
H N E ESa x X(S) H (S)H
l l l l I I ll 1 I l

— —"—' <—-d——-'——’

CmR Plac aroFFBR serA Pam; Ptacg" glk

pJY2.183A
H N E ESa Sa x X(S) H (S)H
I g l J I 1 I ll 1 J |

 

CmR Plac aroFFBR tktA serA ParoF Ptacglt glk

pJY3.21B
H N E ESa X XSH H
I I I IL I III I

 

— —> <——- «n—C— ‘——
CmR Plac aroFFBR serA ParoF glk

pJY3.298

H N E ESa Sa XXSH H
l I I ll 1 I Ill 1

CmFl Plac aroFFBR tktA serA Pam; glk

 

Figure 18“. Restriction enzyme maps of plasmids.

3 Restriction enzyme sites are abbreviated as follows: B = BamHI, E = EcaRI, H =
Hindlll, N = NcaI, S = Sall, Sa = SacI, X = Xbal. Parentheses indicate that the designed

enzyme site has been eliminated.

 

JYl (KL3AptsHIcrr, devoid of the ptsH, ptsI, and crr genes) was made from E. coli KL3
by P1 phage transduction using E. coli UE7923 as a donor strain to inactivate PTS-
mediated transport and phosphorylation of glucose. The ptsH—encoded protein and ptsI-
encoded protein are proteins involved in all PTS systems. The crr-encoded protein is

specific for glucose PTS transport. E. coli JYl was characterized by its growth pattern

60

 

on glucose, maltose containing fosfomycin, xylose, and xylose containing CAMP.24 The
isolation of spontaneous mutants that can grow on glucose from E. coli strains lacking the
pts genes has been reported previously.25 Using a similar approach, cultivation of E.
caliJYl/pRC55B under fermentor-controlled conditions in M9 glucose minimal medium
with aromatic supplementation led to the isolation of a faster-growing variant that was
designated E. coli JY1.2. Plasmid pRC55B containing serA gene enabled the growth of
E. coli KL3-derived strains in minimal salts medium without serine supplementation.
Repetitive subculturing of E. coli JY1.2/pRC55B in M9 glucose minimal medium
resulted in the isolation of E. coli JY1.3. E. coli JYl.2 and E. coli JY1.3 differed in their
growth rates in M9 glucose minimal medium. The inactive PTS in E. coli JYl.2 and E.
coli J Y1.3 was confirmed by growth on maltose containing fosfomycin and no growth on
mannitol. The GalP dependency for glucose transport in E. coli JY1.2 and E. coli JY1.3
was confirmed by transduction of a galP::Tn10 mutation to generate strains that were
unable to grow on glucose.

The plasmids used to evaluate the impact of different mechanisms for glucose
transport on the concentration and yield of synthesized DHS are summarized in Figure
18. Plasmid pKL5.17A contained tktA, an additional copy of the POW promoter region of
aroF, and araFFBR. All the other plasmid contained Z. mobilis glf encoding the glucose
facilitator protein and (or) Z. mobilis glk encoding glucokinase in addition to plasmid
pKL5.17A. E. coli KL3/pKL5.17A transports glucose by wild—type PTS system. E. coli
J Y1.2/pKL5.l7A and E. coli J Y1.3/pKL5.17A transports glucose by galactose permease

(GalP) followed by phosphorylation of glucose with native glucokinase. Plasmid

61

pJY2.183A (Figure 19), which carried both a Z. mobilis glfinsert and a Z. mobilis glk

pSK4.203A

$1) BamHl digest
BamHl BamHl

‘ 2.2 kb I

tktA

1) BamHl digest
2) CIAP treatment

 

 

 

1 cl
1) Sad digeSt ' EcoHl
l 2) CIAP treatment 1) 5301 dlgeStl

T4 Ligase

 

Sacl EcoRl
Figure 19. Construction of plasmid pJY 2.182 and pJY2.183A

62

 

pT0325
l1) Hindlll digest
2

Hi I” Hindlll
i 2. kb I

glk
1) Hindlll digest
2) CIAP treatment

 

T4 Ligase

 

p] Y 2. l 8 2
$ 1) Sad digest
Sad 22 kb Sacl
lktA
ECORI ¢ 1) Sad digest ¢
2) CIAP treatment

(Smal)
.I'er

 

Figure 20. Construction of plasmid pJY3.21B and pJY3.29B

63

 

insert, was used in E. coli JY1/pJY2.l83A to transport glucose by Glf followed by
phosphorylation of glucose with overexpressed glucokinase (le). Plasmid pKL7.1 15A
carried only a Z. mobilis glf insert and was used in E. coli KL3/pKL7.115A to transport
glucose by both PTS and Glf. Plasmid pJY3.29B (Figure 20) carried only a Z. mobilis
glk insert and was used in E. coli JY1.2/pJY3.29B and E. coli JY1.3/pJY3.29B to
transport glucose by GalP followed by phosphorylation of glucose with overexpressed
glucokinase (le).
Concentrations and yields of synthesized DHS
All constructs were cultured at l L scale in a 2.0-L working volume fermentor at
pH 7.0 and 36 °C. Dissolved oxygen was maintained at 20% air saturation under both
glucose-rich and glucose-limited conditions. Glucose-rich conditions were important for
evaluation of both the Glf system and the GalP system since the driving force to transport
glucose by facilitated diffusion (Glf) or galactose permease (GalP) included the
concentration gradient of glucose between the culture medium and cytoplasm. Aliquots
of the culture medium at certain time points were removed and concentrations of DHS, 3-
dehydroquinic acid, gallic acid, 3-deoxy-D-arabina-heptulosonic acid (DAH), and acetic
acid were determined by 1H NMR using response factors determined with authentic
samples of all molecules (Table 5). The specific activities of DAHP synthase and
glucokinase were also measured from aliquots removed from the fermentor at certain
time points (Table 6 and Table 7). All the fermentations were run in duplicate and
reported results represent an average of two runs.
With native PTS-mediated glucose transport, E. coli KL3/pKL5.17A synthesized

49 g/L DHS in 26% yield under both glucose-limited (Entry 1, Table 2) and glucose-rich

64

conditions (Entry 2, Table 5) and provided the basis for evaluating the impact of different

glucose transport mechanisms on concentration and yield of synthesized DHS.

Table 5. Yields and concentrations of synthesized DHS and shikimate pathway
byproducts.

 

 

Entry Strain HostC/Plasmidd Time“ DHSr DAH, DHQ, Total AcOHf
(h) (g/L) GA Yield“ (g/L)
Yieldg (g/L)
1" KL3/pKL5.17A pls+ 48 (49) 26% (0.0) (6.0) (4.8) 33% 0.1
2b KL3/pKL5.17A pts+ 42 (49) 26% (0.0) (7.9) (6.1) 33% 0.1
3" KL3/pKL7.115A ptsl/glf 48 (45) 25% (0.0) (5.1) (6.0) 32% 0.0
4b KL3/pKL7.115A ptsl/glf 48 (63) 27% (4.1) (7.7) (7.8) 36% 0.6
5a JY1/pJY2.183A ptslglfglk 48 (60) 34% (0.0) (5.5) (8.6) 41% 0.2
6b JY1/pJY2.183A pts‘/glfglk 42 (31) 19% (0.0) (5.0) (4.4) 25% 9.5
7b JY1.2/pKL5.17A plat-gallD+ 96 (40) 20% (0.0) (5.4) (5.1) 25% 0.0
8b JY1.2/pJY3.29B pts’galPl/glk 84 (35) 19% (0.0) (4.2) (5.7) 24% 0.8
9b JY1.3/pKL5.17A pts‘galrr 60 (60) 36% (0.0) (5.7) (7.1) 43% 0.0
10b JY1.3/pJY3.29B pts'galPVglk 6O (54) 29% (0.0) (5.2) (13) 38% 0.0
11“ JY1.3/pKL5.17A pts‘galV 6O (37) 21% (0.0) (4.1) (3.4) 25% 0.0

 

“ glucose-limited conditions. b glucose-rich conditions. ‘ in addition to araE353
serAzzaraB (see Table 1). d in addition to araFFBR, ParoF’ tktA, and serA plasmid inserts
(see Figure 2). e fed-batch fermentor culture times. fAbbreviations: DHS (DHS), 3-
deoxy-D-arabina-heptulosonic acid (DAH), 3-dehydroquinic acid (DHQ), gallic acid
(GA), acetic acid (AcOH). g (mol DHS)/(mol glucose consumed). h (mol DHS + mol
DAH + mol DHQ + mol GA)/(mol glucose consumed).
Accumulation of DAH was not observed and the total yields of DHS and byproducts 3-
dehydroquinic and gallic acids synthesized by E. coli KL3/pKL5.17A (33%) were also
the same under both glucose-limited (Entry 1, Table 5) and glucose-rich conditions
(Entry 2, Table 5). Unlike E. coli KL3/pJY1.216A, E. coli KL3/pKL5.17A consumed the
same amount of glucose (approximately 1.4 mol) under both glucose-rich and glucose-
limited conditions.

E. coli KL3/pKL7.115A was constructed to transport glucose by both PTS and
Glf-mediated facilitated diffusion. Under glucose-rich conditions, E. coli

KL3/pKL7.115A consumed significantly more glucose (approximately 1.8 mol) relative

to E. coli KL3/pKL5.17A (approximately 1.4 mol). Under glucose-limited conditions, E.

65

coli KL3/pKL7.115A consumed the same amount of glucose (approximately 1.8 mol) as
E. coli KL3/pKL5.17A, indicating that the outside glucose concentration had a significant
impact on the consumption of glucose by Glf-mediated glucose transport. Consistent
with more glucose consumption, E. coli KL3/pKL7.115A synthesized significantly more
DHS (63 g/L) in almost the same yield (27%) from glucose under glucose-rich conditions
(Entry 4, Table 5) relative to E. coli KL3/pKL5.17A (Entry 2, Table 5). The increase in
the total yield of DHS and shikimate pathway byproducts by E. coli KL3/pKL7.115A
(36%) was primarily due to formation of DAH. By contrast, E. coli KL3/pKL7.115A
under glucose-limited conditions synthesized the same amount of DHS and shikimate
pathway byproducts (Entry 3, Table 5) as E. coli KL3/pKL5.17A (Entry 1, Table 5),
suggesting that Glf may not transport glucose under glucose-limited conditions when
PTS is available.

E. coli JY1/pKL7.115A was constructed to transport glucose by only Glf-
mediated facilitated diffusion. However, E. coli JY1/pKL7.1 15A grew too slow to be
used in the high-density culture fermentations where glucose was the carbon source. In E.
coli JY1/pKL7.l 15A, the glucose transported inside the cytoplasm via Glf-mediated
facilitated diffusion was phosphorylated by E. coli chromosomal glk-encoded
glucokinase. Heterologous overexpression of Z. mobilis glucokinase in addition to E.
coli native glucokinase led to normal growth rates of E. coli J Y1/pJY2.183A in minimal
salts medium where glucose was the carbon source. E. coli JYl/pJY2.183A under
glucose-limited conditions synthesized significantly more DHS (60 g/L) in higher yield
(34%) from glucose (Entry 5, Table 5) relative to E. coli KL3/pKL5.17A (Entry 1, Table

5). However, E. coli JYl/pJY2.183A under glucose-rich conditions (Entry 6, Table 5)

66

led to a very steep decline in the synthesis of DHS. Acetic acid, which increased in
concentration throughout the course of the fermentation run of E. coli JY1/pJY2.l83A,
reached a final concentration of 9.5 g/L (Entry 6, Table 5). The accumulation of acetic
acid appeared to inhibit the cell growth as indicated by the decreased dry cell weight (15
g/L compared to 22 g/L for all the other fermentation runs) at the end of the fermentation
run.

E. coli JYl.2 and the more rapidly growing E. coli JY1.3 was constructed to
transport glucose by only GalP-mediated Hl symport. The glucose transported inside the
cytoplasm via GalP-mediated H+ symport was phosphorylated by E. coli chromosomal
glk-encoded glucokinase. E. coli JY1.2/pKL5.17A synthesized less DHS (40 g/L) in
lower yield (20%) from glucose (Entry 7, Table 5) relative to E. coli KL3/pKL5.17A
under glucose-rich conditions (Entry 2, Table 5). By contrast, the more rapidly growing
E. coli JY1.3/pKL5.17A (Entry 9, Table 5) synthesized more DHS (60 g/L) in higher
yield (36%) relative to E. coli KL3/pKL5.17A under glucose-rich conditions (Entry 2,
Table 5). However, E. coli JY1.3/pKL5.17A under glucose-limited conditions (Entry 11,
Table 5) led to a very steep decline in the synthesis of DHS. Although the Z. mobilis glk
plasmid insert had a major impact on the synthesis of DHS in E. coli JY1/pJY2.183A, a
Z. mobilis glk plasmid insert in E. coli JY1.2/pJY3.29B (Entry 8, Table 5) and E. coli
JY1.3/pJY3.29B (Entry 10, Table 5), respectively, did not lead to a higher concentration
and yield of biosynthesized DHS relative to JY1.2/pKL5.l7A (Entry 7, Table 5) and

JY1.3/pJY5.17A (Entry 9, Table 5) under glucose-rich conditions.

67

 

 

 

 

 

 

 

 

 

 

 

 

time (11)

Figure 21. Growth and synthesis of DHS during cultivation of E. coli
JY1/pJY2.183A under glucose-limited conditions (gray bars) and E. coli
JY1.3/pKL5.17A under glucose-rich conditions (open bars). Dry cell weight formed
during growth of E. coli JY1/pJY2.183A (black circles) and E. coli JY1.3/pKL5.17A
(open circles).

 

DHQ, GA (g/L)
U'l

 

 

 

 

 

 

 

 

 

 

 

 

0 18 54 60

time (h)
Figure 22. Synthesis of shikimate pathway byproducts during cultivation of E. coli
JY1/pJY2.183A under glucose-limited conditions and E. coli J Y1.3/pKL5.17A under
glucose-rich conditions. 3-Dehydroquinic acid (black bars) and gallic acid (gray bars)
synthesized by E. coli JY1/pJY2.183A. 3-Dehydroquinic acid (open bars) and gallic acid
(stipled bars) synthesized by E. coli JY1.3/pKL5.17A.

Heterologous expression of Z. mobilis glf and glk in E. coli JY1/pJY2.183A under
glucose-limited conditions and. utilizing GalP in E. coli JY1.3/pKL5.17A under glucose-
rich conditions led to the highest titers and yields of biosynthesized DHS in this study

(Figure 21). One major difference between those two fermentation runs was the

68

cultivation time. E. coli JY1/pJY2.183A reached the stationary phase of its growth at
approximately 24 h while E. coli JY1.3/pKL5.17A reached the stationary phase of its
growth at approximately 30 h (Figure 21). E. coli JYl/pJY2. 183A was cultivated for 48
h until the synthesis of DHS stopped while E. coli JY1.3/pKL5.17A was cultivated for 60
h until the synthesis of DHS stopped. The concentrations of both 3-dehydroquinic acid
and gallic acid reached the maximum at the end of the fermentation runs of both E. coli

JYl/pJY2.183A and E. coli JY1.3/pKL5.17A (Figure 22).

DAHP synthase and glucokinase expression

The specific activity of DAHP synthase varied widely from 0.01 U/mg to 0.98
U/mg as a function of the employed. glucose transport system, culture conditions and
cultivation times. Under glucose-rich conditions, higher DAHP synthase specific
activities (10-26-fold) were observed for PTS—utilizing E. coli KL3/pKL5.17A (Entry 2,
Table 6) relative to glucose-limited conditions (Entry 1, Table 6). In contrast, under
glucose-rich conditions, lower DAHP synthase specific activities were observed for PTS—
and. Glf- utilizing E. coli KL3/pKL7.115A (Entry 3 versus Entry 4, Table 6) and Glf—
utilizing E. coli JY1/pJY2.l83A (Entry 5 versus Entry 6, Table 6) relative to glucose-
limited conditions. DAHP synthase specific activities declined after 12 h over the course
of the fermentation runs of both PTS-utilizing E. coli KL3/pKL5.17A (Entry 2, Table 6)
and PTS- and Glf-utilizing E. coli KL3/pKL7.1 15A (Entry 3, Table 6). However, DAHP
synthase specific activities were stable for Glf-utilizing E. coli JY1/pJY2.183A under
glucose-limited conditions (Entry 5, Table 6) and even increasing over the course of the

fermentation runs of GalP-utilizing constructs under glucose-rich conditions (Entries 7-9,

69

 

Table 6). No correlation was observed between increased synthesized DHS and

increased DAHP synthase specific activities.

Table 6. DAHP synthase specific activities.

 

 

 

Entry Construct HostC/Plasmidd DAHP Synthase Specific Activity (U/mg)

12h 24h 36h 48h 60h 72h

1a KL3/pKL5.17A pts+ 0.037 0.031 0.033 0.025

2b KL3/pKL5.17A pts” 0.98 0.63 0.30 0.23

3“ KL3/pKL7.115A ptsl/glf 0.52 0.53 0.26 0.1 1

4b KL3/pKL7.115A ptsl/glf 0.12 0.64 0.15 0.01

5“ JY1/pJY2. 1 83A pts'/glfglk 0.29 0.21 0.20 0.25

6b JY1/pJY2.183A pts‘lglfglk 0.1 l 0.11 0.089 0.022

7" JY1.2/pKL5. 17A pts'galP+ 0.02 0.08 0.09

8b JY1.2/pJY3.29B pts‘galP’Vglk 0.09 0.08 0.15 0.20

9b JY1.3/pKL5.17A pts‘galP+ 0.10 0.10 0.12 0.20

 

a glucose-limited conditions. b glucose-rich conditions. C in addition to araE353
serAzzaraB (see Table 1). d in addition to araFFBR, PamF, tktA, and serA plasmid inserts
(see Figure 18).

Table 7. Glucokinase specific activities.

 

 

 

Entry Construct HostC/Plasmidd Glucokinase Specific Activity (U/mg)

12h 24h 36h 48h 60h 72h

5a JY1/pJY2. 183A pts'/glfglk 0.30 0.30 0.30 0.30

6b JY1/pJY2.183A pts‘/glfglk 1.0 1.0 1.2 1.0

7b J Y1.2/pKL5.17A pts'galP” 0.04 0.06 0.07 0.05

8" JY1.2/pJY3.2QB pts‘galP‘L/glk 0.08 0.19 0.16 0.16

9b JY1.3/pKL5.17A pts'galP+ 0.03 0.07 0.05

10b JY1.3/pJY3.29B pts'galPVglk 0.11 0.15 0.10

 

2‘ glucose-limited conditions. b glucose-rich conditions. C in addition to araE353
serAzzaraB (see Table 1). d in addition to araFFBR, PamF, tktA, and serA plasmid inserts
(see Figure 18).

Glucokinase specific activities were stable over the course of all the fermentation
runs (Entries 5-10, Table 7). Overexpression of glucokinase in Glf-utilizing E. coli
JY1/pJY2.183A was a critical factor to enable fast growth on glucose minimal salts
medium. The glucokinase specific activity for E. coli JYl/pJY2.183A was six-fold
higher under glucose-limited conditions (Entry 5, Table 7) and approximately twenty-
fold higher under glucose-rich conditions (Entry 6, Table 7) than the average native E.

coli glucokinase specific activity of 0.05 U/mg (Entry 9, Table 7). For GalP-utilizing

70

constructs, glucokinase specific activity was observed to be the same for both slow-
growing E. coli JY1.2/pKL5.17A (Entry 7, Table 7) and faster growing E. coli
JY1.3/pKL5.17A (Entry 9, Table 7) under glucose-rich conditions, indicating that the
reason for the different growth rate between E. coli JY1.2 and E. coli JY1.3 was probably
due to different expression levels of GalP between E. coli JY1.2 and E. coli JY1.3. The
Z. mobilis glk insert in E. coli JY1.2/p1 Y3.29B led to a two- to three-fold increase in the
specific activity of glucokinase relative to E. coli JY1.2/pKL5.17A (Entry 8 versus Entry
7, Table 7). The similar increase in glucokinase specific activity was observed for E. coli
JY1.3/pJY3.29B relative to E. coli JY1.3/pKL5.17A (Entry 10 versus Entry 9, Table 7).
However, increased glucokinase activity had little impact on the biosynthesized DHS for

GalP-utilizing constructs (Entry 7-10, Table 5).

Discussion

Metabolic engineering to design and construct microorganisms suitable for the
production of shikimate pathway products requires control of a complicated metabolic
network. 3-Deoxy-D-arabina-heptulosonic acid 7-phosphate (DAHP) synthase, the first
enzyme in the shikimate pathway, is always the key target for manipulation.
Overexpression of DAHP synthase isozymes insensitive to feedback inhibition formed
the initial basis for increasing the carbon flow directed into the shikimate pathway and
increasing the yield of natural products synthesized from glucose via this pathway. With
overexpression of a feedback insensitive isozyme of DAHP synthase,10 Frost and
coworkers suggested that the availability of E4P was a key factor limiting DAHP

synthase activity.” ‘2 Overexpression of transketolase under shake flask11 and fermentor-

71

 

controlled12 culture conditions led to significant increase in the concentrations and yields
of shikimate pathway products. With the increased availability of E4P, the availability of
DAHP synthase’s other substrate, PEP, remains as a target for manipulation.

As discussed in the introduction chapter, previous examination of altered glucose
transport and overexpressed PEP synthase under shake flask conditions have many
disadvantages. Fermentor-controlled cultivation was employed throughout this
examination of microbial synthesis of DHS from glucose. An accurate evaluation of
yields was thus possible when microbial growth and synthesis of DHS and shikimate
pathway byproducts occurred during a single fermentation run.

Glucose-limited or glucose-rich concentrations were maintained in the
fermentation. The glucose-rich conditions were previously avoided because of reports of
acetate accumulation and low-yield conversion.22 However, Frost and coworkers
previously investigated the production of shikimic acid under both glucose-rich
conditions and glucose-limited conditions and observed that concentrations of
biosynthesized shikimic acid were significantly higher under glucose-rich conditions.26
Thus the glucose—rich conditions were employed in this study and were the major
difference between this study and earlier study of the impact of overexpression of

transketolase and DAHP synthase on the biosynthesis of DHS.ll

Increased flux to shikimate pathway product by PEP synthase overexpression
Overexpression of PEP synthase in E. coli KL3/pJYl.216A led to substantial
increases in the total yields of DHS and other shikimate pathway byproducts (Table 1 and

Table 3) under both glucose-limited and glucose-rich conditions. The increase in yield of

72

DHS and shikimate pathway byproducts (Table 1) synthesized by E. coli
KL3/pJY1.216A overexpressing PEP synthase under glucose-rich conditions was
particularly significant. As PEP synthase specific activity was almost the same under
both glucose-rich and glucose-limited conditions with IPTG addition (Table 2 and Table
4), the yield difference between glucose-rich and glucose-limited conditions with PEP
synthase overexpression may due to the different pyruvic acid availability for PEP
synthase under different conditions. Under glucose-rich conditions, E. coli
KL3/pJY1.216A was observed to consume glucose more rapidly (1.4 mol in 42 h)
relative to glucose-limited conditions (1.2 mol in 48 h). Due to PTS-modulated glucose
transport in E. coli KL3/pJY1.216A, the increased glycolysis flux measured as the
glucose consumption rate over the course of the fermentation runs under glucose-rich
conditions should lead to increased in vivo concentrations of pyruvic acid under glucose-
rich conditions. At 51% (mol/mol), the highest total yield (Entry 3, Table l) of DHS,
DAH, 3-dehydroquinic acid, and gallic acid synthesized by E. coli KL3/pJY1.2l6A was
well above the maximum theoretical yield of 43% (mol/mol) expected in the absence of
recycling of pyruvate to PEP (Figure 15A). The 51% yield, although still well below the
maximum theoretical yield of 86% (mol/mol) expected in the presence of recycling of all
the PEP synthase-generated pyruvate back to PEP, indicated that the limits imposed on
synthesis of DHS by PTS-mediated glucose transport have been successfully
circumvented. In the theoretical calculations, neither 86% nor 43% yield takes into
consideration the amount of glucose that must be converted into cell biomass or

converted into energy to assemble cell mass.

73

An optimal level of PEP synthase expression (approximately 0.07 U/mg at 24 h)
was observed to result in the highest titer and yield of DHS and shikimate pathway
byproducts synthesized from glucose by E. coli KL3/pJY1.216A under both glucose-rich
and glucose-limited conditions. Exceeding this optimal level of PEP synthase expression
led to decreases in both the yield and concentration of DHS and shikimate pathway
byproducts synthesized from glucose (Table 1 and Table 3). With high PEP synthase
expression, it is very likely that the cell’s ability to generate energy for cellular processes
becomes limited by reduced pyruvic acid availability when the glucose consumption
stayed the same for all the fermentation runs under glucose-rich conditions or under
glucose-limited conditions. However, the 0.15 U/mg of the highest PEP synthase
specific activity observed in this study was much lower than the previous report27 of 0.29
U/mg of PEP synthase specific activity that only stimulated oxygen consumption and
0.60 U/mg of PEP synthase specific activity that stimulated both oxygen and glucose
consumption by approximately 2-fold.

An increase in acetate accumulation (Table 1 versus Table 3) was observed under
glucose—rich conditions for E. coli KL3/pJY1.216A. However, these concentrations of
acetate did not negatively impact microbial growth based on the dry cell weight of E. coli
KL3/pJY1.216A (Table 1 versus Table 3). Aerobic acetate production is a manifestation
of the imbalance between glucose uptake and the demands for both biosynthesis and
energy production.28 The most common arguments for the cause of such imbalance are
that the glucose uptake rate is improperly controlled and that the activity of the
tricarboxylic acid (TCA) cycle is limiting.29 So the absence of pyruvate and the modest

concentrations of acetate generated by E. coli KL3/p1 Yl .216A can likely be explained by

74

the properly controlled glucose uptake rate and the activated tricarboxylic acid cycle

throughout the fermentation.

Increased flux to shikimate pathway product by altered glucose transport

All of the previous examinations of different transport systems have been
reviewed in the introduction chapter and there was no direct comparison between GalP-
mediated glucose transport and Glf-mediated glucose transport under a common set of
culture conditions. In this study, the uniform set of fermentor-controlled as opposed to
shake-flask culture conditions were used to maintain glucose, dissolved oxygen
concentrations, and pH. Altered glucose transport led to significant increases in the total
yields of DHS and shikimate pathway byproducts (Entries 4 and 5, Table 8) relative to
the 33% total yield of DHS, 3-dehydroquinic, and gallic acids synthesized by E. coli
KL3/pKL5.17A utilizing PTS (Entry 1, Table 8). Replacement of the native PTS with
Glf-mediated transport and le-catalyzed phosphorylation of glucose in E. coli
JY1/pJY2.183A (Entry 4, Table 8) led to a 41% total yield of DHS, 3-dehydroquinic, and
gallic acids under glucose-limited conditions. Replacement of the native PTS with GalP-
mediated glucose transport and le-catalyzed phosphorylation of glucose in E. coli
JY1.3/pKL5.17A (Entry 5, Table 8) led to a 43% total yield of DHS, 3-dehydroquinic,
and gallic acids under glucose-rich conditions. The results demonstrated that in vivo PEP
availability for biosynthesis of shikimate pathway products increased when the PTS was

replaced with systems that do not expend PEP during the glucose transport process.

75

 

Table 8. Comparison of the yields and concentrations of synthesized
shikimate pathway products and byproducts as a function of the strain
and strategy employed to increase PEP availability.

 

 

 

Glucose Pyruvate product/byproduct (g/L)c Total

Entry Strain transportc recyclingd DAHr DHQf G ,4,f DI—ISr QAf S Ar Yieldg
1b KL3/pKL5.17A PTS — 0 7.9 6.1 49 0 0 33
2" KL3/pJY1.216A PTS + 10 12 13 69 0 0 51
3" KL3/pKL7.115A PTS/Glf - 4.1 7.7 7.8 63 0 0 36
4“ JY1/pJY2.183A Glf — 0 5.5 8.6 60 0 0 41
5" JY1.3/pKL5.17A GalP — 0 5.7 7.1 60 0 0 43
6"h SP1.1/pKD12.138A PTS — 0 0 0 1 1 4 52 24
7"-h SP1.1/pKD15.07lB PTS + 0 0 0 l6 4 66 29
8"“ SP1.1/pSC5.1128 PTS/Glf - 0 0 0 19 6 70 32
9"h SP1.1pts/pSC6.09OB Glf — 0 0 0 15 5 71 34

 

a b

glucose-rich conditions. glucose-limited conditions. ° PEcharbohydrate
phosphotransferase (PTS), augmentation of PTS with facilitated diffusion (PTS/Glf),
facilitated diffusion (Glf), galactose permease (GalP). d native expression (-), amplified
expression (+) of phosphoenolpruvate synthase. '3 product targeted for synthesis is in
boldface. fAbbreviations: 3-deoxy-D-arabina-heptulosonic acid (DAH), 3-dehydroquinic
acid (DHQ), gallic acid (GA), DHS (DHS), quinic acid (QA), shikimic acid (SA). g (mol
DAH + mol DHQ + mol GA + mol DHS + mol QA + mol SA)/(mol glucose consumed);
Unit: percent. hSee Ref. 36.

E. coli JY1.3/pKL5.17A utilizing GalP (Entry 5, Table 8) without overexpression
of Z. mobilis glk-encoded glucokinase synthesized approximately the same
concentrations of DHS and byproducts 3-dehydroquinic and gallic acids in the same
yields as those achieved with E. coli JY1/pJY2.183A (Entry 4, Table 8) utilizing Glf with
overexpression of Z. mobilis glk-encoded glucokinase. However, the slow rates for
growth and synthesis of DHS (Figure 21) and shikimate pathway byproducts (Figure 22)
is a major disadvantage for GalP-utilizing E. coli JY1.3/pKL5.17A relative to Glf-
utilizing E. coli JYl/pJY2.l83A. In addition, the basis for the improvement in growth
rate and yields of shikimate pathway products for E. coli JY1.3 relative to E. coli JY1.2 is

unknown and complicates formulation of future metabolic engineering strategies to

further improve growth rates and shikimate pathway product yields by GalP system. A

76

disadvantage of Glf-utilizing E. coli JYl/pJY2.183A is that only glucose-limited
conditions can be used for the production of DHS. Under glucose-rich conditions, the
acetate produced by E. coli J Yl/pJY2.183A (Entry 6, Table 5) significantly impaired cell
growth and metabolic activities and suggested an imbalance between glucose uptake rate

and TCA cycle activities.

Metabolic consequences of PTS inactivation

The presence of a rapidly metabolizable carbon source such as glucose in the
growth medium of a microorganism can inhibit the synthesis of enzymes involved in the
metabolism of other carbon-containing compounds. This phenomenon is called catabolic
repression. PTS not only functions in glucose transport but also plays an important and
complex regulatory role in catabolic repression.24"‘ 30 It is expected that a strain without
PTS should have impaired catabolic repression and may consume sugar mixtures faster
than a wild-type strain. For example, one PTS mutant can assimilate glucose and

' Therefore, it would be interesting to determine the growth

arabinose simultaneously.3
characteristics of non-PTS strains E. coli JYl and E. coli JY1.3 on different carbon

sources or sugar mixtures.

Comparison of altered glucose transport and PEP synthase overexpression

The strategy of altering glucose transport to increase PEP availability focuses on
the competition for PEP molecule by different cellular processes, the same principle used
by mutational inactivation of PEP carboxylase32 and pyruvate kinase.33 Alternatively,

overexpression of PEP synthase recycles PTS-generated and pyruvate kinase-generated

77

pyruvate back to PEP.34 The strategy of PEP synthase overexpression has the advantage
of recycling not only the PTS generated pyruvate back to PEP but also the pyruvate
kinase-generated pyruvate back to PEP. The supply of PEP thus depends on the
competition for pyruvate by PEP synthase and the other cellular processes that consume
pyruvate. E. coli KL3/pJY1.216A, when expressing the optimal specific activity of PEP
synthase,34 synthesized 69 g/L of DHS in a combined yield with DAH, 3-dehydroquinic,
and gallic acids of 51% (Entry 2, Table 8) under glucose-rich conditions. This yield is
significantly more than the yields of DHS and shikimate pathway byproducts realized
with alteration of glucose transport (Entries 3, 4, and 5, Table 8) and may imply that
overexpression of PEP synthase is a better strategy for increasing PEP availability.
However, the enzyme activities of PEP synthase were optimized in this study while the
activities of enzymes involved in altered glucose transport were not optimized. In
addition, a number of other factors need to be considered. The acetic acid accumulation
with cultivation of Glf-utilizing E. coli J Y1/pJY2.183A (Entry 6, Table 6) under glucose-
rich conditions prevented use of the glucose-rich conditions observed (Entry 2, Table 8)
to lead to the highest yields of DHS and shikimate pathway byproducts when PEP
synthase expression was amplified. The underlying reason for the faster growth rate and
the higher combined yield of DHS, 3-dehydroquinic, and gallic acids synthesized by
constructs based on GalP-utilizing E. coli JY1.3 relative to E. coli JY1.2 (Table 5)

remains to be elaborated.

78

 

Comparison of shikimic acid biosynthesis and DHS biosynthesis

Overexpression of PEP synthase and Glf-mediated facilitated diffusion were also
examined in E. coli for synthesis of shikimic acid (Entries 6-9, Table 8).35 GalP-mediated
transport was not examined for shikimic acid synthesis because of the uncertain genotype
of cells relying on GalP for glucose transport. All shikimate-synthesizing constructs
(Entries 6-9, Table 8) contained plasmid-localized araFFBR encoding feedback-insensitive
DAHP synthase, tktA encoding transketolase and araE encoding shikimate
dehydrogenase. E. coli SP1.1/pSC5.1 12B transported glucose using both Glf and PTS
(Entry 8, Table 8). E. coli SP1.1pts/pSC6.090B transported glucose using only Glf
(Entry 9, Table 8). Recycling of PTS-generated pyruvic acid to PEP with overexpression
of PEP synthase was tested in E. coli SP1.1/pKD15.071B while the plasmid-localized
ppsA gene was under its native promoter (Entry 7, Table 8). Shikimic acid, DHS and
quinic acid were accumulated in the culture medium and the total yield of hydroaromatic
products was based on the combined yields of shikimic acid, DHS, and quinic acid
(Entries 6-9, Table 8).

Contrary to synthesis of DHS (Entries 1-5, Table 8), replacing (Entry 9, Table 8)
or augmenting PTS with Glf (Entry 8, Table 8) were led to higher yields relative to
overexpression of PEP synthase (Entry 7, Table 8) during synthesis of shikimic acid.
However, this may not be an apprOpriate comparison because the expression level of PEP
synthase was not optimized during the shikimate study. Synthesized yields (29-34%) of
shikimic acid and pathway byproducts (Table 8) were consistently lower than the
synthesized yields (36-51%) of DHS and pathway byproducts (Table 8) no matter what

strategies were employed to increase PEP availability. The variability of the host strain

79

 

for synthesis of DHS (E. coli KL3“) and synthesis of shikimic acid (E. coli SP1.135) may
:ontribute to these yield differences. The consumption of NADPH by shikimate
lehydrogenase (AroE) may also contribute to the yield differences due to the redox
lalances inside the cells. Feedback inhibition of shikimate dehydrogenase by shikimic

Icid may also be limiting yields for synthesis of this hydroaromatic.

Ionclusion

Overexpression of PEP synthase or alteration of glucose transport by Glf-
rediated facilitated diffusion or the GalP galactose permease in E. coli constructs led to
rcreased yields of DHS and the combined yields of DHS, 3-dehydroquinic, and gallic
cids relative to the native PTS without PEP synthase overexpression. Both strategies
lere compared under a common set of fermentor-controlled conditions. Overexpression
f PEP synthase is currently leading to the synthesis of higher yields of DHS and higher
ambined yields of DHS and shikimate pathway byproducts relative to alteration of
lucose transport. Beyond the optimizations remaining to be made in Glf-utilizing and
alP-utilizing E. coli strains for DHS biosynthesis, the general applicability of strategies
>r increasing PEP availability remained to be tested. The consequences of a given
.rategy for increasing PEP availability differed as a function of whether DHS or
likimic acid is targeted for synthesis. The respective contribution to this variability

ising from the progenitor host E. coli strain was investigated in the next section.

Comparison of KL3araKaraL and SP1.1 for the production of shikimic acid

locatalyst design

80

Both shikimate kinase isozymes (Figure 1) encoded by araK and araL were
inactivated by successive P1 phage-mediated transductions of araL47822Tn10 and
araKzszR into the appropriate E. coli host (KL3 and RB791) to generate E. coli
constructs (KL3araKaraL and SP1.1).35 The araE gene localized on plasmid
complements genomic mutation of araE in KL3. While ensuring that carbon flow
directed into the common pathway did not proceed beyond synthesis of shikimic acid,
inactivation of the shikimate kinase also precluded de novo biosynthesis of aromatic
amino acids and aromatic vitamins. Growth of all constructs therefore required
supplememntation with L-phenylalanine, L-tyrosine, L-trptophan, p-hydroxybenzoic acid,
p-aminobenzoic acid, and 2,3-dihydroxybenzoic acid. These supplements could
potentially create a problem in that feedback inhibition of DAHP synthase by aromatic
amino acids play a prominent role in controlling carbon flow directed into the common
pathway. Accordingly, an isozyme of DAHP synthase encoded by plasmid-localized
araFFBR that was insensitive to feedback inhibition by aromatic amino acids was carried
as a plasmid-localized insert. An additional copy of the araE encoding shikimate
dehydrogenase, ppsA encodign PEP synthase, and tktA encoding transketolase, were also
localized in the plasmid pKD15.07lB35 to enhance the carbon flow into the shikimate

pathway.

Shikimic acid titers and yields as a function of the progenitor host
E. coli SP1.1/pKD15.07lB and KL3araKaraL/pKD15.07lB were initially

cultured at 1 L scale in a 2.0-L working volume fermentor for 60 h at pH 7.0 and 33 °C.

Glucose-rich conditions were employed where the concentration of glucose in the

81

medium was maintained in the range of 55-170 mM throughout the run. These culture
conditions have been identified to suppress equilibration of shikimic acid and quinic
acid.36 Formation of quinic acid results from the reduction of 3-dehydroquinic acid
catalyzed by araE encoding shikimate dehydrogenase. The formation of quinic acid
during biosynthesis of shikimic acid results from a microbe-catalyzed equilibration
involving transport of initially synthesized shikimic acid back into the cytoplasm and
operation of the common pathway of aromatic amino acid biosynthesis in the reverse of
its normal biosynthetic direction. Shikimic acid transport has been shown to subject to
catabolite repression.36a Hence increasing glucose availability in the culture (glucose-rich
conditions) can repress shikimic acid transport and minimize formation of quinic acid.
Under these culture conditions, all constructs (Entries 1 and 2, Table 9) entered the
stationary phase of their growth 18 h after inoculation of the fermentor.

In addition to shikimic acid, the concentrations of DHS and quinic acid were
determined. The total yield of hydroaromatic products was based on the combined yields
of shikimic acid, DHS, and quinic acid. Shikimic acid, DHS, and quinic acid were
determined by 1H NMR using response factors determined with authentic samples of all
products. E. coli SP1.1/pKD15.07lB and KL3araKaraL/pKD15.07lB provided the
comparative strains for evaluating the impact on the concentrations and yields of
synthesized shikimic pathway products attendant with altering the progenitor strain.

Both E. coli SP1.1/pKD15.071B and KL3araKaraL/pKD15.07lB synthesized
shikimate pathway products in 37% total yield (Table 9). Varying the progenitor strain
had no impact on the yield of the synthesized shikimate pathway products. Based on this

result, the difference between the achieved total yield of 51% for DHS and shikimate

82

pathway byproducts synthesized by E. coli KL3/pJYl.216A and yield of 37% for
shikimic acid and shikimate pathway byproducts synthesized by E. coli
KL3araKaraL/pKD15.071B will focus on the effect of shikimate dehydrogenase.

Table 9. Comparison of the yields and concentrations of synthesized shikimic acid

and shikimate pathway byproducts as a function of the progenitor host under
glucose-rich conditions.

 

 

. 1 , .1 Total

Entry Straln SA DHS QA Yield”
1 SP1.1/pKD15.071B 67 18 4.4 37%

2 KL3araKaraL/pKD1507 l B 60 18 5.4 37%

 

aAbbreviations: DHS (DHS), quinic acid (QA), shikimic acid (SA); Unit: g/L.
b (mol DHS + mol QA + mol SA)/(mol glucose consumed).

Shikimate dehydrogenase is inhibited by shikimic acid exhibiting linear mixed-
type inhibition with an inhibition constant of 0.16 mM and may limit the yield of
synthesized shikimate pathway products.37 Alternatively, because shikimate
dehydrogenase needs NADPH to act as a cofactor, the demand of ample NADPH inside
the cells with the overexpression of shikimate dehydrogenase may also limit the yield of
synthesized shikimate pathway products. To examine the first hypothesis, isolation of a
shikimate dehydrogenase isozyme insensitive to shikimic acid was essential. To examine
the second hypothesis, approaches to improve the NADPH supply inside the cells are
needed.

In the literature, it is a proven concept that in Carynebacterium glutamicum the
production of L-lysine correlates with increased intracellular NADPH generation.38
NADPH can be generated by glucose 6-phosphate dehydrogenase in the pentose
pathway, isocitrate dehydrogenase in the TCA cycle and malic enzyme. It has been
proved that pentose pathway was the most important source for intracellular NADPH

generation in C. glutamicum because the contribution of both malic enzyme39 and

83

isocitrate dehydrogenase38b' 39“ to the generation of NADPH is of minor importance. One
metabolic engineering strategy to increase NADPH availability entails redirection of
pathway by inactivation of pgi-encoded phosphoglucose isomerase, which is the first
enzyme specific for glycolysis. When an E. coli pgi mutant was derived from an E. coli
tryptophan production strain, the efficiency of tryptophan formation was doubled.40
Applying the same strategy for C. glutamicum strain DSM5715 the L-lysine formation
was increased by 1.7-fold.4| It was also shown that the reducing power imbalance caused
by N ADPH overproduction in an E. coli pgi mutant could be recovered to some extent by
introducing an NADPH-consuming pathway such as the poly-3-hydroxybutyrate
biosynthetic pathway of E. coli pgi.42 Thus it would be interesting to introduce a pgi
mutant into the shikimate-synthesizing E. coli constructs to evaluate the impact of

NADPH availability.

Effect of pH on the production of DHS

Introduction
In nature, a wide range of proton concentrations is encountered: from pH 1 in
acidic sulfur springs to pH 11 in soda lakes. Bacteria exist in all of these environments,
although the number of species that can grow in extremely acidic habitats (acidophilic
bacteria) and those (alkalophilic bacteria) that can grow in extremely alkaline ones are
more restricted than those (neutrophilic bacteria) that can grow over the mid range from

approximately pH 5.0 to 9.0. E. coli, which is representative of the large neutrophilic

84

class, grows at a maximum rate between pH 6.0 and pH 8.0 and more slowly at half a pH
unit or so beyond these limits.43

The pH is an important parameter of the rate of many types of reactions, not only
ionization of acids and bases but also solvolysis and oxidoreduction. As a result,
enzymes and other macromolecules function optimally only over a narrow range of pH,
usually close to neutrality, a range that is much more restricted than the pH range of the
external environment over which growth of the bacterium is possible.

E. coli, as well as many other neutrophiles, has evolved remarkably effective
mechanisms for homeostasis of intracellular pH.44 E. coli regulates its pH at 7.4 to 7.8
during growth over the external pH range of 5.0 to 9.0.45 The transmembrane pH
difference (ApH) is a component of the proton potential, which drives processes of
transport, motility, and coupling of respiration.44 Most of these processes are
“secondary”, in the sense that they entail no covalent bond exchanges, but a few are
“primary” because they are used to drive a chemical reaction. Conceptually speaking, the
simplest are the various porters, membrane transport proteins that couple the downhill
flux of protons into the cell to the uphill flux of some metabolite inward or outward.

There are three kinds of porters: uniporter, symport and antiporter. Uniport
proteins transport a substrate down its electrochemical gradient and are unlinked to any
ion. The glycerol uniporter of E. coli is one example, but there are not many other
examples. Symport protein is exemplified by the galactose permease (GalP).18 This
porter protein mediates the movement of two substrates in the same direction, a molecule
of sugar plus one proton. The obligatory coupling between the two fluxes ensures that

the electrochemical driving force upon the proton powers the concurrent accumulation of

85

the sugar. Another case is the accumulation of K+ ions; this is believed to occur by
symport with H“. Antiport protein is so articulated that the flux of protons into the cell
powers the efflux of some substrate from the cytoplasm.

In this study, the pH was lowered from 7.0 down to 6.5 and to 6.0. The increased
transmembrane pH difference (ApH) will increase the proton potential across the
membrane. Its effect on the production of DHS was examined by E. coli strains using
either galactose permease (E. coli JY1.3/pKL5.17A) or wild-type PTS (E. coli
KL3/pJY1.216A) to transport glucose inside the cells. Under our fed-batch fermentation
conditions, culture pH is easily controlled.

Table 10. Yields and concentrations of synthesized DHS and shikimate pathway
byproducts as a function of pH under glucose-rich conditions.

 

 

d .1 d d Glucose
Entry Strain Host“/Plasmidb pHC (g 01.))HYSiel d° DAH ’ (lg/P1132 ’ GA 5:123, Uptakeg
(mmol)

l JY1.3/pKL5.17A prs'galP+ 7.0 (60) 36% (0.0) (5.7) (7.1) 43% 1222

2 JY1.3/pKL5.17A pts'galP+ 6.5 (70) 37% (0.0) (6.8) (3.1) 43% 1369

3 JY1.3/pKL5.17A pts'galP* 6.0 (77) 42% (0.0) (6.8) (2.1) 48% 1242

4 KL3/pJYl.2l6A pts+/ ppsA 7.0 (69) 35% (10.0) (12.0) (13.0) 51% 1372

5 KL3/pJY1.216A pts+/ ppsA 6.0 (83) 39% (12.0) (14.1) (1.6) 51% 1623

 

21in addition to araE353 serAzzaraB. b in addition to araFFBR, PamF, tktA, and serA plasmid
inserts. ° pH of culture medium. d Abbreviations: DHS (DHS), 3-deoxy-D-arabina-
heptulosonic acid (DAH), 3-dehydroquinic acid (DHQ), gallic acid (GA), acetic acid
(AcOH). "' (mol DHS)/(mol glucose consumed). f (mol DHS + mol DAH + mol DHQ +
mol GA)/(mol glucose consumed). g glucose consumed throughout the fermentation.
DHS titers and yields as a function of pH

The DHS-synthesizing strains were initially cultured at 1 L scale in a 2.0-L
working volume fermentor for 42 h (KL3/pJY1.216A) or 60-72 h (JY1.3/pKL5.17A) at

36 °C. Glucose-rich culture conditions were employed where the concentration of

glucose in the medium was maintained in the range of 55-170 mM throughout the run.

86

Dissolved oxygen was maintained at 20%. All the fermentations were run in duplicate
and reported results represent an average of two runs.
Substantial increases in the titer of DHS were observed (Table 10) with
decreasing pH for both E. coli JY1.3/pKL5.17A and KL3/pJYl.2l6A cultured under
glucose-rich conditions. With the change of pH from 7 to 6, approximately 15 g/L more
DHS was synthesized for both strains (Entry 1 versus 3, Table 10; Entry 4 versus 5, Table
10). In addition, with the change of pH from 7 to 6, the total yield of the shikimate
pathway products increased 5% for E. coli JY1.3/pKL5.17A (Entry 1 versus 3, Table 10)
while remaining the same for E. coli KL3/pJY1.216A (Entry 4 versus 5, Table 10). The
glucose consumption for E. coli JY1.3/pKL5.17A didn’t change much at low pH while
the glucose consumption for E. coli KL3/pJY1.216A increased about 20% at low pH
(data not shown). The relationship between pH and glucose consumption by wild-type
PTS or galactose permease is unexpected. At low pH, the increased transmembrane pH
difference (ApH) will increase the proton potential across the membrane. Glucose
tran3port mediated by galactose permease required proton potential and glucose
consumption throughout the fermentation was expected to increase with the increased
proton potential at low pH. Glucose transport mediated by wild-type PTS didn’t require
Proton potential to transport glucose and glucose consumption throughout the
fermentation and was expected to stay the same at low pH.
Another interesting observation is the decrease of gallic acid accumulation at low
PH- Gallic acid was accumulated normally during the stationary phase of the
fermentation. At pH 6.0, both E. coli JY1.3/pKL5.17A and KL3/pJY1.216A

accumulated only 2 g/L gallic acid compared to more than 7 g/L gallic acid at pH 7.0.

87

Gallic acid formation was due to either enzymatic or chemical catalysis.46 In vitro DHS
can react with O2 in a phosphate buffer to form gallic acid, protocatechuic acid,
tricarballyic acid and pyrogallol. Because the reaction of DHS was a general base-
catalyzed process first order in both DHS and inorganic phosphate with an overall second
order rate constant of 1.4 x 10'5 M’I $4.4" the acid environment at pH 6.0 may prevent
DHS from being converted to gallic acid and other byproducts. This is probably the
major reason for increased accumulation of DHS at low pH.

In conclusion, E. coli KL3/pJY1.2l6A and JY1.3/pKL5.17A synthesized higher
concentrations of DHS at pH 6.0 relative to at pH 7.0. The 83 g/L of DHS produced by
E. coli KL3/pJY1.216A at pH 6 is by far the highest titer achieved for DHS production
because of both 20% increase in glucose consumption and decrease in gallic acid
accumulation. However, the reason why glucose transport mediated by wild-type PTS is

more efficient at low pH and whether lowering the pH can become a general strategy to

increase the titer of shikimate pathway product remains to be investigated.

88

REFERENCE

1 Niu, W.; Draths, K. M.; Frost, J. W. Benzene-Free Synthesis of Adipic Acid.
Biotechnal. Prag. 2002, 18, 201-211.

2 Gibson, J. M.; Thomas, P. S.; Thomas, J. D.; Barker, J. L.; Chandran, S. S.;
Harrup, M. K.; Draths, K. M.; Frost, J. W. Benzene-Free Synthesis of Phenol. Angew.
Chem., Int. Ed. 2001, 40, 1945-1948.

3 Draths, K. M.; Frost, J. W. Environmentally Compatible Synthesis of Catechol
from D-Glucose. J. Am. Chem. Soc. 1995, 117, 2395-2400.

4 Ran, N.; Knop, D. R.; Draths, K. M.; Frost, J. W. Benzene-Free Synthesis of
Hydroquinone. J. Am. Chem. Soc. 2001, 123, 10927-10934.

5 (a) Barker, J. L.; Frost, J. W. Microbial Synthesis of p-Hydroxybenzoic Acid
from Glucose. Biotechnol. Bioeng. 2001, 76, 376-390. (b) Amaratunga, M.; Lobos, J. H.;
Johnson, B. F.; Williams, D. Genetically Engineered Microorganisms and Method for
Producing 4-Hydroxybenzoic Acid. 2000 US Patent 6030819.

6 Li, K.; Frost, J .W. Synthesis of Vanillin from Glucose. J. Am. Chem. Soc. 1998,
120, 10545-10546.

7 Ensley, B. D.; Ratzkin, B. J.; Osslund, T. D.; Simon, M. J.; Wackett, L. P.;
Gibson, D. T. Expression of Naphthalene Oxidation Genes in Escherichia coli Results in
the Biosynthesis of Indigo. Science, 1983, 222, 167-169.

8 (a) Kambourakis, S.; Draths, K. M.; Frost, J. W. Synthesis of Gallic Acid and
Pyrogallol from Glucose: Replacing Natural Product Isolation with Microbial Catalysis.
J. Am. Chem. Soc. 2000, 122, 9042-9043. (b) Kambourakis, 8.; Frost, J. W. Synthesis of
Gallic Acid: Cu2+-Mediated Oxidation of DHS. J. Org. Chem. 2000, 65, 6904-6909.

9 Chang, Y.; Almy, E. A.; Blamer, G. A.; Gray, J. 1.; Frost, J. W.; Strasburg, G. M.
Antioxidant Activity of DHS in Liposomes, Emulsions, and Bulk Oil. J. Agric. Food
Chem. 2003, 51 , 2753-2757.

10 (a) Ogino, T.; Garner, C.; Markley, J. L.; Herrmann, K. M. Biosynthesis of
aromatic compoundsz‘3C NMR spectroscopy of whole Escherichia coli cells. Prac. Natl.
Acad. Sci. USA 1982, 79, 5828-5832. (b) Weaver, L. M.; Herrmann, K. M. Cloning of an
AroF Allele Encoding a Tyrosine-Insensitive 3-Deoxy-D-arabina-heptulosonate 7-
phosphate synthase. J. Bacterial.1990, 172, 6581-6584.

11 Li, K.; Mikola, M. R.; Draths, K. M.; Worden, R. M.; Frost, J. W. Fed-Batch

Fermentor Synthesis of DHS Using Recombinant Escherichia coli. Biotechnol. Bioeng.
1999, 64, 61-73.

89

12 (a) Draths, K. M.; Frost, J. W. Synthesis Using Plasmid-Based Biocatalysis:
Plasmid Assembly and 3-deoxy-D-arabina-heptulosonate Production. J. Am. Chem. Soc.
1990, 112, 1657-1659. (b) Draths, K. M.; Pompliano, D. L.; Conley, D. L.; Frost, J. W.;
Berry, A.; Disbrow, G. L.; Staversky, R. J.; Lievense, J. Biocatalytic Synthesis of
Aromatics from D-Glucose: The Role of Transketolase. J. Am. Chem. Soc. 1992, 114,

3956-3962.

13 (a) Li, K.; Frost, J. W. Microbial Synthesis of DHS: A Comparative Analysis of
D-Xylose, L-Arabinose, and D-Glucose carbon sources. Biatechnal. Prag. 1999, 15, 876-
883. (b) Li, K.; Frost, J. W. Utilizing Succinic Acid as a Glucose Adjunct in Fed-Batch
Fermentation: Is Butane a Feedstock Option in Microbe-Catalyzed Synthesis? J. Am.
Chem. Soc. 1999, 121, 9461-9462.

14 Postma, P. W.; Lengeler, J. W.; Jacobson, G. R.
Phosphoenolpyruvate:Carbohydrate Phosphotranferase Systems. In Escherichia coli and
Salmonella: Cellular and Molecular Biology, 2”" Ed.; Neidhardt, F. C., Ed.; ASM Press:
Washington, DC, 1996; pp 1149-1174.

15 (a) Patnaik, R.; Liao, J. C. Engineering of Escherichia coli Central Metabolism
for Aromatic Metabolite Production with Near Theoretical Yield. Appl. Environ.
Microbial. 1994, 60, 3903-3908. (b) Liao, J. C.; Hou, S. Y.; Chao, Y. P. Pathway
Analysis, Engineering, and Physiological Considerations for Redirecting Central
Metabolism. Biotechnol. Bioeng. 1996, 52, 129-140.

16 Niersbach, M.; Kreuzaler, F.; Geerse, R. H.; Postma, P. W.; Hirsch, H. J. Cloning
and Nucleotide Sequence of the Escherichia coli K-12 PpsA Gene Encoding PEP
Synthase. Mol. Gen. Genet. 1992, 231, 332-336.

17 Parker, C.; Barnell, W. O.; Snoep, J. L.; Ingram, L. O.; Conway, T.
Characterization of the Zymamanas mobilis glucose facilitator gene product (glf) in
recombinant Escherichia coli: Examination of transport mechanism, kinetics and the role
of glucokinase in glucose transport. Mal. Microbial. 1995, 15, 795-802.

18 (a) Flores, N.; Xiao, 1.; Berry, A.; Bolivar, F.; Valle, F. Pathway Engineering for
the Production of Aromatic Compounds in Escherichia coli. Nature Biotechnol. 1996, 14,
620-623. (b) Saier, M. H. Jr.; Bromberg, F. G.; Roseman, S. Characterization of
constitutive galactose permease mutants in Salmonella typhimurium. J. Bacterial. 1973,
113, 512-514.

19 Pao, S. S.; Paulsen, I. T.; Saier, M. H. Jr. Major facilitator superfamily. Microbial.
Mol. Biol. Rev. 1998, 62, 1-34.

90

20 Snell, K. D.; Draths, K. M.; Frost, J. W. Synthetic Modifica
Escherichia coli Chromosome: Enhancing the Biocatalytic Conversion of (
Aromatic Chemicals. J. Am. Chem. Soc. 1996, 118, 5605-5614.

21 Cobbett, C. S.; Delbridge, M. L. Regulatory Mutants of the AroF-Tyr.
Escherichia coli K-12; J. Bacterial. 1987, 169, 2500-2506.

22 (a) Konstantinov, K. B.; Nishio, N.; Yoshida, T. Glucose Feedi
Accounting for the Decreasing Oxidative Capacity of Recombinant Escher
Fed-Batch Cultivation for Phenylalanine Production. J. Ferment. Bioeng. 19
260. (b) Konstantinov, K. B.; Nishio, N.; Seki, T.; Yoshida, T. Physiological]
Strategies for Control of the Fed-Batch Cultivation of Recombinant Escheri.
Phenylalanine Production. J. Ferment. Bioeng. 1991, 71, 350-355. (c) Kle
Strohl, W. R. Acetate Metabolism by Escherichia coli in High-C
Fermentation. Appl. Environ. Microbial. 1994, 60, 3952-3958.

23 Meyer, D.; Schneider-Fresenius, C.; Horlacher, R.; Peist, R.; Boos, VI
Characterization of Glucokinase from Escherichia coli K-l2. J. Bacterial.
1298-1306.

24 (a) Postma, P. W.; Lengeler, J. W.; Jacobson, G. R. Phosphoen
Carbohydrate Phosphotransferase Systems. In Escherichia coli and Salmone.
and Molecular Biology, 2nd ed.; Neidhardt, F. C., Ed.; ASM Press: Wash
1996; pp 1149-1174. (b) Cordaro, J. C.; Melton, T.; Stratis, J. P.; Atagt'in, M
C.; Hartman, P. E.; Roseman, S. Fosfomycin Resistance: Selection Method
and Extended Deletions of the Phosphoenolpyruvate:Sugar Phosphotransfera
Salmonella typhimurium. J. Bacterial. 1976, 128, 785-793.

25 Flores, N .; Xiao, J .; Berry, A.; Bolivar, F.; Valle, F. Pathway Enginet
Production of Aromatic Compounds in Escherichia coli. Nature Biatechna
620-623.

26 (a) Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von D
Weber, W.; Frost, J. W. Hydroaromatic Equilibration During Biosynthesis
Acid. J. Am. Chem. Soc. 2001, 123, 10173-10182. (b) Draths, K. M.; Knop, l
J. W. Shikimic Acid and Quinic Acid: Replacing Isolation from Plant S1
Recombinant Microbial Biocatalysis. J. Am. Chem. Soc. 1999, 121, 1603-16C

27 Patnaik, R.; Roof, W. D.; Young, R. F.; Liao, J. C. Stimulation
Catabolism in Escherichia coli by a Potential Futile Cycle. J. Bacteriol. 1992
7532.

28 El-Mansi, E. M. T.; Holms, W. H. Control of carbon flux to acetal

during growth of E. coli in batch and continuous culture. J. Gen. Microbiol
2875-2883.

91

29 Farmer, W. R.; Liao, J. C. Reduction of Aerobic Acetate Production by
Escherichia coli. Appl. Environ. Microbial. 1997 , 63, 3205-3210.

30 (a) Hogema, B. M.; Arents, J. C.; Bader, R.; Eijkemans, K.; Inada, T.; Aiba, H.;
Postma, P. W. Inducer exclusion by glucose 6-phosphate in Escherichia coli. Mal.
Microbial. 1998, 28, 755-765.

31 Hernandez-Montalvo, V.; Valle, F.; Bolivar, F.; Gosset, G. Characterization of
sugar mixtures by an Escherichia coli mutant devoid of the phosphotransferase system.
Appl. Microbial. Biotechnol. 2001, 57, 186-191.

32 (a) Backman, K.C. US. Patent 5,169,768, 1992. (b) Miller, J. E.; Backman, K.C.;
O'Conner, M. J.; Hatch, R. T. Production of Phenylalanine and Organic-Acids by
Phosphoenolpyruvate Carboxylase-Deficient Mutants of Escherichia coli. J. Ind.
Microbial. 1987, 2, 143-149.

33 Mori, M.; Yokota, A. ; Sugitomo, S.; Kawamura, K. Patent JP 62,205,782, 1987.

34 Yi, J .; Li, K.; Draths, K. M.; Frost, J. W. Modulation of Phosphoenolpyruvate
Synthase Expression Increases Shikimate Pathway Product Yields in E. coli. Biatechnal.
Prag. 2002, 18, 1141-1148.

35 Chandran, S. S.; Yi, J.; Draths, K. M.; von Daeniken, R.; Weber, W.; Frost, J. W.
Phosphoenolpyruvic Acid Availability and the Biosynthesis of Shikimic Acid.
Biatechnal. Prag. 2003, 19, 808-814.

36 (a) Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von Daeniken, R.;
Weber, W.; Frost, J. W. Hydroaromatic Equilibration During Biosynthesis of Shikimic
Acid. J. Am. Chem. Soc. 2001, 123, 10173-10182. (b) Draths, K. M.; Knop, D. R.; Frost,
J. W. Shikimic Acid and Quinic Acid: Replacing Isolation from Plant Sources with
Recombinant Microbial Biocatalysis. J. Am. Chem. Soc. 1999, 121, 1603-1604.

37 Dell, K. A.; Frost, J. W. Identification and removal of impediments to biocatalytic
synthesis of aromatics from D-glucose: Rate-limiting enzymes in the common pathway of
aromatic amino acid biosynthesis. J. Am. Chem. Soc. 1993, 115, 11581-11589.

38 (a) Genealogy profiling through strain improvement by using metabolic network
analysis: metabolic flux genealogy of several generations of lysine-producing
corynebacteria. Appl. Environ. Microbial. 2002, 68, 5843-5849. (b) Marx, A.; de Graaf,
A. A.; Wiechert, W.; Eggeling, L.; Sahm, H. Metabolic fluxes in Carynebacterium
glutamicum-identification of metabolic patterns by BC isotope analysis. In: Preprints of
the Seventh International Conference on Computer Applications in Biotechnology. May
31 to June 4, 1998, Osaka, Japan, pp. 387-392. (c) Sonntag, K.; Schwinde, J .; deGraaf, A.
A.; Marx, A.; Eikmanns, B. J .; Wiechert, W.; Sahm, H. 13C nuclear magnetic resonance

92

studies of the fluxes in the central metabolism of Corynebacterium glutamicum during
growth and overproduction of amino acids in batch cultures. Appl. Microbial. Biotechnol.
1995, 44, 489-495.

39 (a) Marx, A.; Striegel, K.; de Graaf, A. A.; Sahm, H.; Eggeling, L. Response of
the central metabolism of Carynebacterium glutamicum to different flux burdens.
Biotechnol. Bioeng. 1997,56, 168-180. (b) Marx, A.; Eikmanns, B. J.; Sahm, H.; de
Graaf, A. A.; Eggeling, L. Response of the central metabolism in Carynebacterium

glutamicum to the use of an NADH-dependent glutamate dehydrogenase. Metabal. Eng.
1999, I, 35-48.

40 Deletion of pgi alters tryptophan biosynthesis in a genetically engineered strain of
Escherichia coli. Appl. Environ. Microbial. 1991, 5 7, 2995-2999.

41 (a) Dunican, L. K.; Mccormback, A.; Stapelton, C.; Burke, K.; O’Donohue, M.;
Marx, A.; Mockel, B. Cloning and uses of a novel nucleotide sequence coding for
glucose-6-phosphate isomerase (pgi) from bacteria. 2001, Patent EP 1087015. (b) Marx,
A.; Hans, S.; Mockel, B.; Bathe, B.; deGraaf, A. A. Metabolic phenotype of
phosphoglucose isomerase mutants of C arynebacterium glutamicum. J. Biotechnol. 2003,
104, 185-197.

42 (a) Shi, H.; Nikawa, J .; Shimizu, K. Effect of modifying metabolic network on
p1y-3-hydroxybutyrate biosynthesis in recombinant Escherichia coli. J. Biasci. Bioeng.
1999, 87, 666-677. (b) Kabir, M. M.; Shimizu, K. Gene expression patterns for metabolic
pathway in pgi knockout Eshcerichia coli with and without phb genes based on RT-PCR.
J. Biotechnol. 2003, 105, 11-31.

43 Ingraham, J. L.; Marr, A. G. Effect of Temperature, Pressure, pH, and Osmotic
Stress on Growth. In Escherichia coli and Salmonella: Cellular and Molecular Biology,
2“d ed.; Neidhardt, F. C., Ed.; ASM Press: Washington, DC, 1996; pp 1570-1578.

44 (a) Booth, I. R. Regulation of cytoplasmic pH in bacteria. Microbial. Rev. 1985,
49, 359-378. (b) Padan, E.; Schuldiner, S. Intracelluar pH and membrane potential as
regulators in the prokaryotic cell. J. Membr. Biol. 1987 , 95, 189-198.

45 (a) Padan, E.; Zilberstein, D.; Rottenberg, H. The proton electrochemical gradient
in Escherichia coli cells. Eur. J. Biochem. 1976, 63, 533-541. (b) Slonczewski, J. L.;
Rosen, B. P.; Alger, J. R.; macnab, R. M. pH homeostasis in Escherichia coli:
measurement by 3 IP nuclear magnetic resonacnce of methylphosphonate and phosphate.
Prac. Natl. Acad. Sci. USA 1981, 78, 6271-6275. (c) Zilberstein, D.; Agmon, V.;
Schuldiner, S.; Padan, E. Escherichia coli intracellular pH, membrane potential, and cell
growth. J. Bacterial. 1984, 158, 246-252.

93

 

46 Kambourakis, S. Synthesis of Value-Added Chemicals from Glucose Using
Chemical and Microbial Catalysis: Gallic acid, Protocatechuic acid, Pyrogallol and
Catechol. Ph. D. thesis. Michigan State University, East Lansing, MI. 2000.

94

 

CHAPTER 3
Microarray analysis of gene expression changes in response to

overexpression of PEP synthase

Introduction

Researchers have employed a variety of strategies for improving the yield of
shikimate pathway product in E. coli. The availability of PEP acid emerges as a key
factor limiting carbon flow directed into the shikimate pathway after overexpression of
araFFBR-encoded DAHP synthase and tktA-encoded transketolase to remove feedback
inhibition and ameliorate restricted availability of D-erythrose 4-phosphate.l One
strategy to surmount the limitation of PEP employed overexpression of E. coli ppsA-
encoded PEP synthase to recycle PTS—generated pyruvic acid back to
phosphoenolpyruvic acid, as was discussed in the preceding chapter.la The resulting
strain E. coli KL3/pJYl.216A is currently synthesizing the highest titer and yield of
biosynthesis of 3-dehydroshikimic acid when PEP synthase is expressed at optimal level.
However, the achieved total yield 51% (mol/mol) of 3-dehydroshikimic acid and
shikimate pathway byproducts synthesized from glucose was still far removed from the
theoretical maximum 86% (mol/mol).

In order to make a better biocatalyst to produce more 3-dehydroshikimic acid,
expression arrays was used as a tool to investigate both the consequences of PEP
synthase overexpression during the time course of the fermentation and the expression
profile changes between different stages of the fermentation. The comparison between

“relative high yield” cells when IPTG was added to induce the overexpression of PEP

95

synthase and “relative low yield” cells when no IPTG was added may provide clues to
limitations in metabolic flux. In addition, the comparison of expression profiles between
different stages of the fermentations will help us understand the regulation of the genes
during the time course of the fermentation and may provide clues to limitations in
metabolic flux particularly during the end of the fermentation. An advantage for this
analysis is that the same strain E. coli KL3/pJYl.216A was used to generate all the
expression array data thus prevent the complications from different strain background.
This method is limited, however, by an inability to detect mutations that alter primary
sequence (and function) without affecting mRNA levels. For example, araF and araFFBR
cannot be differentiated by this method. Where examined, previous studies have
indicated that expression changes provide a good predictor for metabolic flux2 and
changes in protein levels.3 Recent improvements in methodology and simple statistical
tools now allow the reproducible measurement of changes that differ by less than 2-

fold. 2"" 4

F ed-batch fermentation of E. coli KL3/pJY1.216A for RNA isolation

E. coli KL3/p1 Y1 .216A, which was discussed in detail in Chapter 2, was cultured
under glucose-rich conditions at 36°C and pH 7.0 in a 2.0-L working volume fermentor.
The concentration of glucose in the medium was maintained in the range of 55-170 mM
throughout the run. Dissolved oxygen was maintained at 20% air saturation by allowing
the impeller speed to vary. To induce PEP synthase expression, 12 mg IPTG was added

when the preset maxima of 750 rpm (stir rate) and 1.0 L/L/min (air flow rate) were

96

 

 

70
60-
50-

E40-

3330-
20-
10-

 

.....

 

 

 

 

 

 

 

18 24 3O 36
time (h)
Figure 23. DHS production by E. coli KL3/pJY1.216A with and without IPTG

addition. Abbreviation: 3-dehydroshikimic acid (DHS). No IPTG added (black bars), 12
mg of IPTG added at 16 h, 22 h, 28 h and 30 h (open bars).

 

 

 

 

time (h)

Figure 24. DHQ, DAH and GA production by E. coli KL3/pJY1.216A with and
without IPTG. Abbreviations: 3-deoxy-D-arabina-heptulosonic acid (DAH), 3-
dehydroquinic acid (DHQ), gallic acid (GA). No IPTG added, DHQ (black bars); IPT G
added, DHQ (open bars); No IPTG added, DAH (black bars with dots); IPTG added,
D AH (open bars with dots); No IPT G added, GA (hashed bars); IPTG added, GA (open

bars with curves).

97

 

 

reached (usually 16 h), and every 6 h after the first addition (usually 22 h, 28 h, 34 h). In
36 h, a total yield of 39% (mol/mol) of DHS (Figure 23) and shikimate pathway
byproduct (Figure 24) was realized by E. coli KL3/p1 Yl.2l6A in the absence of IPTG
addition. This contrasts with the total yield of 52% (mol/mol) when E. coli
KL3/pJY1.2l6A was cultured with IPTG addition (Figure 23 and 24). Glucose-rich fed-
batch fermentations were run in triplicate and reported results represent an average of

three runs.

Microarray platform

Under the discovery-based approach, DNA microarrays are used as screening
tools to identify genes associated with the biological process of interest. Once target
genes are identified, additional experiments may be designed to validate this list. The
efficiency of knowledge discovery using this high-throughput experimental process
depends upon the reliability of the microarray technology used in the initial screening
experiments. Research planning to utilize microarray experiments for discovery-based
research must evaluate available commercial technologies when allocating laboratory
resources for prospective experiments. As discussed in the introduction chapter,
Panorama E. coli gene array was chosen to be the platform to perform microarray
analysis.

Each Panorama array contains PCR—amplified open reading frames (ORFs) from
the E. coli K-12 (MG1655) genome. The majority of ORFs have been amplified from
Start to stop codon. All of the 4,290 ORFs have been printed in duplicate at equal mass

per spot onto positively charged nylon hybridization membranes. Following printing,

98

each array was cross-linked with UV light. The arrays are 12 cm x 24 cm in size. The
genomic DNA is found in the corner of the arrays. During any expression profiling
experiment where labeled cDNA is used as probe, the genomic DNA spots will act as a
positive control and should always show significant signals. However, this signal

strength was not utilized in the analysis in this study.

Synthesis of Radiolabeled cDN A

RNA was purified from the cells at 18h, 24 h, 30 h and 36 h after initiation of the
run from the triplicate fermentation in which IPTG was added and the triplicate
fermentation in which IPTG was not added. Total RNA was isolated using the hot
phenol method and purified as described by Tao et al.4‘1 Radioactive cDNA was
synthesized as described by Gonzalez et a1.211 using [ct-33P]dCTP (New England Nuclear)
and a set of primers homologous to the 3’ ends of the predicted 4,290 ORFs in E. coli
(Sigma-Genosys). Approximately 70% of the added radioactivity was incorporated into

cDNA.

Hybridization
Samples of cDNA were hybridized to DNA probes on nylon membranes (Sigma-
Genosys). Hybridizations were performed in roller bottles at 65 °C. A total of eight
filters used in these experiments were from the same manufacturing lot. Each filter was
stripped and used 3 times. Each filter corresponds to each condition investigated (with
and without IPTG addition, each at 4 time points during the fermentation for a total of 8

different conditions).

99

 

Quantitation of cDNA as a relative measure of gene expression (mRNA)

Wet membranes were wrapped in Mylar film and exposed to a phosphorimager
screen (Molecular Dynamics) for 20 h. Exposed screens were scanned using a
phosphoimager at a pixel size of 50 um. Array Vision software (version 6.1, Imaging
Research Inc.) was used to quantify the intensity of each spot after correcting
automatically for background using the “surrounding entire template” method. The sum
of values for all 4,290 ORFs in each array was normalized to 100 million to allow
comparison of gene expression between experiments. Each ORF is expressed as a
percentage of the total (i.e., 100,000 corrsponds to 0.1% of total cDNA hybridized to the
array). Results for every time point represent an average of six determinations (three
hybridizations each, two arrays per filter). Statistical analyses and data transformations

were performed in Excel using the log of expression values.

Statistical analysis

Among the 4290 ORFs, 347 genes with an average expression level below
0.002% were considered unreliable. The coefficient of variation (standard deviation
expressed as a percentage of the mean) was used to estimate the error associated with the
measurement of the remaining ORFs. Table 11 lists the average coefficient of variation

for each data set. Considerable error exists in these measurements, with an average
coefficient of variation of 26%.

As pointed out by Arfin et al.,5 the significance of gene expression comparisons

cannot be interpreted based solely on the ratio of measured values without statistical

evaluation. The same parametric statistical methods that are routinely used with small

100

 

sets of biological and biochemical data can be employed provided gene expression data
can be shown to follow a normal distribution. Log expression has previously been shown
to approximate a normal distribution?" In our studies, log values of expression data were
generated and statistical significance of gene comparisons at the same time point between
“high yield” cells and “low yield” cells was evaluated by computing probability values
for the null hypothesis using two-tailed student t tests.6 This test produces a p value that
represents the probability that the difference is observed because of random chance. A
very small p value (p < 0.05) will indicate that the tested gene is likely to be differentially
expressed. Using student t test, many smaller but significant changes were identified.

Table 11. Comparison of data reproducibility (coefficient of variation) for all the
samples obtained from E. coli KL3/pJY1.216A fermentation“.

 

 

 

 

18 h 24 h 30 h 36 h
No IPTG 23% 26% 30% 30%
With IPTG 25% 23% 30% 22%

 

 

 

 

 

a Results for each condition are summarized as a coefficient of variation (CV), the
standard deviation expressed as a percentage of the mean. Values in each column
represent the average CV for all ORFs under condition. Genes with expression value
below 0.002% of the total message (threshold for reliable measurement”) were
eliminated from consideration.

Cluster analysis to define corresponding physiological states

Clu sting analysis7 is the standard means by which large sets of experiments and
transcripts are analyzed. Generally, some measure of similarity is used to place each
gene (and/or experiment) into a single group or a position within a hierarchy. The output

can be displayed as lists, but are more typically visualized by some variation of the

101

 

red/green matrix display originally introduced by Eisen et al .8 An interesting application
of this approach is the clustering of tumors to find new possible tumor subclasses.9
Cluster analysis was used to examine if there are significant global expression
differences between samples of all conditions and identification of the most appropriate
pairing of samples for comparison. Hierarchical clustering by average linkage8 was
carried out to intuitively examine gene expression variation among samples of different
conditions using Cluster software downloaded from Standford University. Basically, this
is an agglomerative process in which single-member clusters are fused to bigger and
bigger clusters. In somewhat more detail, the procedure starts by computing a pairwise
distance matrix between all the genes or conditions, the distance matrix is explored for
the nearest genes or conditions, and they are defined as a cluster. There are several
hierarchical clustering algorithms that differ in the way the distances are calculated.
After a new cluster is formed by agglomeration of two clusters, the distance matrix is
updated to reflect its distance from all other clusters. Then, the procedure searches for
the nearest pair of clusters to agglomerate, and so on. This procedure leads to a
hierarchical dendogram in which multiple clusters are fused in nodes according to their
similarity, finally resulting in a single hierarchical tree (Figure 25). Relationships among
objects (conditions) are represented by a tree whose branch lengths reflect the degree of
similarity between the objects (conditions), as assessed by a pairwise similarity function.
In agreement with growth state, the samples could be clustered into 2 big groups
(early fermentation at 18 h and 24 h; later stages of fermentation at 30 h and 36 h) and 4
small groups (Figure 25). For more convenient analysis of transcriptome profile

variations, the entire cultivation was divided into four growth stages as follows: 1.

102

 

exponential growth phase in which cells grow at a constant specific growth rate (18 h); 2.
early stationary phase (24 h); 3. middle stationary phase (30 h); 4. late stationary phase
(36 h). It is interesting to notice that, at 30 h and 36 h, expression profiles for E. coli
KL3/pJY1.2l6A with and without IPTG addition were not clustered together as expected.
0n the contrary, expression profiles for E. coli KL3/pJYl .216A with IPTG addition at 30
h and 36 h were clustered together. Expression profiles for E. coli KL3/pJYl.2l6A
without IPTG addition at 30 h and 36 h were also clustered together. This result indicates
that there are significant changes of physiological states at 30 h and 36 h with PEP
synthase overexpression relative to the control. Based on cluster analysis, pairwise
comparisons of the “high yield” cells and “low yield” cells are appropriate for the 18 h
and 24 h samples. Comparisons at 30 h and 36 h samples appear valid but may be

anomalous.

 

 

 

 

 

 

 

 

 

 

 

 

1 T1

18h 18h 24h 24h 30h 36h 30h 36h
- IPTG + IPTG - IPTG + IPTG - IPTG - IPTG + IPTG + IPTG

Figure 25. Cluster analysis of expression data by each condition.

103

 

Table 12. Number of differentially expressed 0RFs in response to PEP synthase

(PpsA) overexpression.

 

 

 

 

 

 

 

 

 

Time points 18 h 24 h 30 h 36 h

The total genes with p-values < 0.05 716 1150 747 2186

Increased by 2-fold or more with PpsA overexpression 27 227 53 214

Decreased by 2-fold or more with PpsA 17 52 l 69 1056
overexpressron

 

How many genes are differentially expressed with phosphenolpyruvate synthase
overexpression during the fermentation?

A simple method for finding sets of interesting genes that may provide clues to
limitations in metabolic flux is comparing expression profiles of two or more samples for
differentially expressed genes. Tables 12 through 14 summarize comparisons for gene
array data. Expression results were compared for E. coli KL3/pJYl.216A during the
fermentation of glucose either with or without IPTG induction of phosphenolpyruvate
synthase overexpression under the glucose-rich conditions at individual time points.
Comparisons were made on a statistical basis and are presented as a ratio. Table 12
indicated total ORFs that differed significantly (defined as p < 0.05) and total ORFs that
differed by 2-fold or greater for each comparison (also p < 0.05). As a result of the
relatively low standard deviations for ORFs in these data sets, the statistical approach
identified many more differentially expressed ORFs than the ratio approach. Table 13
and Table 14 break down those ORFs with 2-fold or greater difference by functional
groups based on the information provided by Sigma Genosys. The names of ORFs that

differed by 2-fold or greater simultaneously at two individual time points (i.e. at both 24

104

 

h and 30 h) were also shown in Table 13 and Table 14. Identification of these ORFs by
functional groups served as a guide for further investigation.

Table 13. Functional group of the upregulated expressed ORFs (ratio >2.0) in
response to PEP synthase overexpression.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Number of genes (ratio > 2.0) Gene name° (ratio > 2.0)
Functional group“ Tot
al" 18h 24h 30h 36h 18h & 24h 24h & 30h 30h & 36h
Carbon compound _ _
metabolism 1 29 0 4 2 l 2 p018 paAB
Regulatory protein 178 l 7 2 28 cyaA b0705
Phage, transposon 87 0 0 l 13
or plasmid
Structure protein 236 l 5 2 22
Transport 427 2 l4 5 51 crr msYB bfr patH bfr
Cell process 197 0 22 4 20 (II) S dnaK SOdA dnaK sadA
mopB
Energy 243 0 19 8 25 adhE
metabolism
DNA replicatron, 1 1 5 2 2 0 7
repalr
Transcription 55 0 l 3 2 rpaB
Translation 1 82 1 l 3 8 rpiL
Amino acid _ trpA trpB trpC
biosynthesis 131 0 8 5 6 an trpD trpE
Nucleotide 58 0 1 0 3
metabolism
Fattrc acid 48 0 2 0 1
metabolism
Central
intermediary 188 3 l4 2 3 ppsA gldA aceB gldA aceB
metabolism
Cofactors and 103 2 5 0 3
prosthetlgtgroups
Putative 277 6 6 l l pgiA
yde ydiA b1005 b1112
. 163 171836 b1 783 yhbH b1005 b1112
Hypo‘het‘ca] 5 9 l ‘6 15 9 hdeB hdeD 63238 b3242 b3238 yde
hdeAjifi‘ yde

 

 

 

 

 

 

 

 

 

 

a Functional Group was classified based on the information provided by Sigma Genosys.
b Total number of identified ORFs for this functional group.
° The names of gene that differed by 2-fold or greater simultaneously at two individual

time points.

105

Table 14. Functional group of the downregulated expressed ORFs (ratio >2.0) in

response to PEP synthase overexpression.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Number of genes Gene name
Functional groupa
Totalb 18h 24h 30h 36h 18h & 24h 24h & 30h 30h & 36h
Carbon compound 129 l 2 1 8 ascG
metabolism
Regulatory protein ”178 0 5 l 12 0105
Phage, transposon 87 0 7 1 14
or plasrmd
Structure protein 236 1 7 l 25
Transport 427 3 '10 6 34 be” CydD g atC
ugpA tnaB
Cell process 197 l 3 3 l4 asmY rhi V
Energy
metabolism 243 l 5 1 17 hch
DNA repllcat‘on’ 115 0 2 4 21 modF priA
repair
Transcription 55 0 0 3 l6 sch sbcC hst
Translation 182 1 0 2 3 1 yihK
Amino acid
biosynthesis 131 3 2 8 35 b1748 yagF
Nucleotide
metabolism 58 4 l 7 l6 adk apt adk purT
FM" 39" 48 0 0 4 12 fadA acs 64249
metabolism
Central pckA bisZ
intermediary 188 0 1 7 4l b0221 edd
metabolism b2463 hisQ
Cofacto rs and 103 l 0 3 l6 menF hemY
prosthetrc groups
Putative 277 0 1 18 85 16 genes
. 83 genes
Hypothetlc 1635 l 4 99 659 yebL including ye b L

 

 

 

“ Functional Group was classified based on the information provided by Sigma Genosys.
b Total number of identified ORFs for this functional group.

° The names of gene that differed by 2-fold or greater simultaneously at two individual
time points.

Pathway diagrams

There is substantial literature about molecular mechanisms and biochemical

pathways,‘0 and some information has been organized into metabolic pathway databases,

106

most notably EcoCyc,“ but also KEGG,12 and WIT.l3 Now it is possible to overlay all

the expression data into all published pathway of E. coli metabolism in the EcoCyc

 

 

 

 

 

990‘ g 0 3 000 $ 0900 DDOO Q DDOOOO fl HUD

' 8128:8882

== .‘Y § i} § 8888:8888

‘*' aLek-3,. : . a 3, H8548

:1 “a? ‘3’}; n.1,; 2 .§ 110 .§° 882 888.88:
E' as...oi A. I”. -E°.-?p° 83:223.;518 g
: .. ° 1' 135 5312-: n.3,» 883883.93 5
55;. g; 1.1132,? :_ °"" 2g 8823:“: 8: g
7 . o a "£1 a c D - in ‘0'. ha 3:8:X&:&8 o

: .87} £51,; TC? via; .19” 82212118”.
I 11.3 - /§ 06%,“: , , " v . 888188883°1g
15:"- ‘T 5;“. 88 :/.° . file assigning;
2E . 1.2 § LAN; . 8882:8888" .

; 8“}, ‘§§ oggtgo . 21' §:. 8881888818
L 1 Ag} - 41g: :39 :88 :zuxs§°

. ‘3 1‘5”?“ 1 . 2“?“ i «Eitéiffiii

° 3: elf/{‘1’}: 8 , ,1? {29:3 118888.888
N 88 3, §. . ,8; m 8188888318,“

. ‘2: xii \ '°' ° 2 3122:4488-

is": ck: - -\ m. «8.2 g 31822881888

8 :3 :1‘ 8: 3° 3 33°: XISXLETES

123 - g 1: ¢8§§ «(g 8831:8881:

1° ‘ 5.. 18 .E 2. gmzaam

I g: ’I" .§ g 3;: 2:3: 8 8888:...

fig; 5 12°; :5 88;; 888883368

, "- Egy” 1 “é .. :: 381234.933

A A ° '“fii‘tgfi
3% 3c» 0Y3 22;" WW :0? :3 g“: 93$

9-2 588C28=8°=811 8—8 288:81'8-833 {:5‘”
a between ooobboé hoe 12an o? I

A 3516666163 7
a Carbuhydratzs
0 Proteins

0 (Filled) Phospharylated
Expruslon Ratios
7

5

3 5
2.4
1 5
1

0 7
0 5
D 34
0.23

- No data available
nknuwn enzyme

Histogram of Expression Data

143

left sIde = genes in overview

right 511.12 = gems not in overview
I Mulu ple highlighu

0 440

Figure 26. Pathway diagram of expression ratio at 18 h (+IPTG/-IPTG). Images in
this thesis/dissertation are presented in color.

107

A Amino Acids

OODIW 0000 g D ’0 ca

990 time «pagan
., . ...

 

‘9. i 7
m w W
a m ,
m 3
m m Wm
M. e!
n. m e mmsi
E .l h
m u mm 3mm.
5 m nm in new.
.n ae De 8
P n V“ 3:..m,
) m Slldmw m nkb
1A .1 0 D.
m mmimuimumuim m «mm,
,5 t
< a mu a mum
m I It

1° 129
18
7.5
lo

328888.888

000 o 00 0 010$ O 000 OD.
lllli kiwi mi. a. a was: Mn. men...
$3:88zazxmsfizsssgosismoahfifia
?.uo.f9o..m?u.u.m.832888233888888
sumo—omgaaomamaaaaaaa a .ma?aaaamo?
”cease?aa..?¥m.e&29?&..&&&&&&&
z..oo.lnoIIggxgflfl888888:0046..00“Lg..oo
flog0°IQ:.-:I.....-.O.‘I.A028008908....Dflg:00
glint.2.0.0880088000°.a00800ngx.00008g2
VOOOOOOOOOOOgOA883882002283218832068
008289000800008288:890‘690000080000008Vfi00
o .0000 AAvAAA
900.2... oooQMMMua M on“ 0006.. «cm can “H” n

8";
q
f
6'
f
8
.‘A.
h
a
“an?
8 3
X
E":
‘20 06:0

0

o
I m I n a . o \o c Anson
\
t o I 0|. .0 I \ 4 9200
“M900. .13 _ J A+oo°¢P
I: .o 090
a \ _
i Annéobn ALL
A.‘ a: II... /\ I 0.0
e u i-ooae: a 0\
4660001340
an o nu“; ufinV 0°13. 0.. Fa .A Mac 9006 AA
QIIIA O
o l5'fil0l OI... Gin .0 To 1050001.. :I o
I O
‘0 . no... a notch: A_ooooA noomou
90th..- .En 0.0008006 90. coon A000: u“ uflgu
as ca
. on 00000 a at. 0190 on 33.00.00
9.. aJa .0662: _ a 000 0 A6 .UA
I f A
o .94: Izoouea A A ozfln 9608A 6?.0850 0
9a . (a .a ur:¢0vo 2.960640
u I .OA on 006 .a 3 coin In. Cocoon
._ 0900‘ .69ch J I on.
00. I mic mud 30.0.90! .00
a, 30 no... 4 90 a ll
9..., oil

mfg.

2:3

,3;. ) . 8 Ezki
5%?Jmfi.mw“&§

"r "arrvuuwunuuuurrur oAaoooom/u WM“

p .4. . a...» b» a; » fl.»».»b b

...... ..:. 0.. 0.09.0qu

4E

O
\

0.0

3;}:

3

18

$4.
108

or .
c‘auboé

1 O
6060 "(h

"‘33st

siwwwnmfi

0‘

0..
O

0

0"

3:3 2:8

"0
0

°—

0

 

Figure 27. Pathway diagram of expression ratio at 24 h (+IPTGl-IPTG). Images in

this thesis/dissertation are presented in color.

1

”0):.
”Y“
.. 1‘"

v
4" 'I
U
.m "
..
>0
sf
oo-o-I
o
D.
n".
”on

rib nu
”In {I
”I
I
. .
.00“

"O“.-
0..
~42,
D

“’3 §° O 8%.“? :3
IfiJk‘NiJfr/‘o H ‘8’!“
1‘11 «€228. 8% ‘1.
: "8 8 88; . i871
”4&3 : w

: 8s 8‘8; 8:33;.8‘7‘; -88 885...?
334/833 2 822‘ \ 10“". 813-0 Ai‘fi‘k

°AE’EV zigai '\—/°°='Q°‘;°°§
8’8... °1is: 828:
18.83: i EM: 3 88.3”.“ 283
8888. .8888 $8888

ii

 

    

9% Q 996 25". ‘0 “*3— °.°3¥“° Egg :3) ggX:?&?
9:3 3:3 8:3 2:2 {—13 g 3:: 3:2 2:3 0"; 3:. 233 .‘3‘. — —
3 03—33% $000 zoo in a? a

  

2828 .

8388 PPyrimldLnes

88:8 1:, .,

:3: a 1. (Filled) Phos'phorylamd
88:8: . 8.9 a smegma“
88:38‘E'8a'8 2 g 88:1
888888.88. 8° gj
8833813857 3 1.8
882388.:M‘8 g1 37
88838?&'&’ 3%. 0:45
888:3?88? 8 :1 0.29
zlez‘msr .1 31:
:88:8*&:8a iNodataavaflable
8823888£88° .1”

888882;: 8#° . HistogramofExpreslunDau
fizzxzs’azs 8&1

33833?&8$
:3:2:&&=8
333331.883
81383881'88 u

88833.1..i8‘3’ :.
33::3363g4’g1

888383A88 g m
23883358” glleflside=genalnuvernew
28221338: Hmmfgggggmmmw :

888888888185“ *

2651
.

  

 

Figure 28. Pathway diagram of expression ratio at 30 h (+IPTG/-IPTG). Images in

this thesis/dissertation are presented in color.

 

 

 

 

00” 9? 9 9 one ? 1109!: 55900 g guqoog n can

r .8 . ' ' 88:88:88”:

«2 ‘- 3 ,5, ' 83882~$98

a/I/f’il:: %; $ g}; I DU §:&l a 23:3:gg&8

: - °-:° o E E; -2 -§ 88.28&&?z
if ‘ g'. °J“ ..3 u E” 'J'u gggg:&&a‘g o

‘ In. I I' _ o '_ / o a.

":2 -:-‘3° ”.3 88..:&9$.g
, B E}; -.-' "'8' i 8 88:: 88.888: ;

2f: . ”O. a J. 882: z-,- {Mai 2282' new
5“ T8,, r'mil ,-- 88833&&?3 ..
o ‘ a,» ‘ ' 4‘ n n H 382:8&&?8Hg
1 g. 3 1? :/° . 35‘. if “ 28888?&?3§°
3 3 E E0 °4¥g i 3:! ,9 083:3?38? flu:
£§§fi $20!: é ‘3 §:o 333: 3?°8334°
A,“ “a 5:8 - Mg: 22:”; 888888832283;
8 8-33.5.8 it; -. . :3}: 3 §§§§§2§ §Z°
oi m; E? 3%. 8 a °~\\: 00°33 2°: Eiggz?u88 85
I 3 :0 ° 0 a :0 ‘. \Q»% o '388?&8$°¢3
.9 _ :3... $88 ° 8. if}? 33883&&88°”°
o ‘5 ‘ &°3§ a .7 a i‘:\ ,. $8.: § 33:32:88,331?

a?” E $2? . 2A; 0 £:‘ ‘2‘,“ ‘3‘. :1 a 33°. 8:83532733‘“
' 8i): :1§.fi1 828 88888 28(3‘ 8835883.. 7‘8
com 0 6° 2°2 82 A .. §°.. 0.00 a “a:
- a: ‘ g g: 83 53 38338888888318.
88088542, 883.83 ”flinging zzzzeaswig
c 3 °-. 28¢” ‘ ° ° 3: 333:8:‘3’33 .3
A L A szzzassgflg
3w. .8.» n»; .5“ 838;: ”w ‘5: f” :8 888.8882 '3
8‘” 2:8 2:: ° 2—8 81 8:8 ’3? .; —. 8 8 8:8 3:3 .3. h ' g

6 066658: coins-ion boo dog? 09

A Amino Amos

0 Py rimidines
LRNAs
O Other
0 lFillrd) Fhosphurylmpd

Expreslon Ratios

I Nu dam available
Unknown enzyme

Histogram of Expression Data

114
left Sid! : genes in overView
5h side : gens not in overView
I Muluple highlighu

365

Figure 29. Pathway diagram of expression ratio at 36 h (+IPTGl-IPTG). Images in

this thesis/dissertation are presented in color.

database.

In this study at each time point, the expression ratio of two samples (PEP

synthase overexpression and wild type PEP synthase expression) for each gene was

overlaid onto all the pathways in the EcoCyc database (Figures 26, 27, 28 and 29).

Reaction lines are color—coded according to the relative expression level of the gene that

codes for the enzyme that catalyzes that reaction step. In the pathway diagram the blue,

green and yellow color showed upregulation genes with PEP synthase overexpression.

The red color showed downregulated genes with PEP synthase overexpression.

Downloading expression ratios for each gene into the EcoCyc database provides a

convenient visual analysis of expression data. For example, it can be seen instantly in the

pathway diagram (Figure 27) that genes involved in the TCA cycle and glycolysis

110

pathway at 24 h after the initiation of the fermentation run were upregulated with the
overexpression of PEP synthase. No other method of visualization can provide as much
cellular context and quick knowledge acquisition as a pathway diagram combined with
expression data. In the following sections, the transcriptional changes of genes grouped

by their functions were examined.

Gluconeogenic genes and adenylate kinase

The ppsA gene encoding for PEP synthase was upregulated approximately 7-10
fold during early fermentation (18 h and 24 h) but was unchanged during later stages of
the fermentation even with the continuous induction of the tac promoter by IPTG
addition (Table 15). The increase in expression levels of ppsA gene at 24 h correlate well
with the 7 fold specific activity increase of PEP synthase (0.069 U/mg with IPTG
addition at 24 h, 0.0093 U/mg without IPTG addition at 24 h) as measured in Chapter 2
of this thesis. The increase in expression levels of ppsA gene at 30 h and 36 h without
IPTG addition (from 0.04% of total cDNA at 24 h to approximately 0.15% of total cDNA
at 30 h and 36 b) may reflect the cell’s need for metabolic flux from pyruvate to PEP
during the later stages of the fermentation. The underlying mechanism for this.
upregulation of ppsA gene at later stages of the fermentation without IPTG addition,
however, remains to be determined. The pckA gene encoding PEP carboxykinase that
converts the TCA cycle intermediate oxaloacetate to PEP was downregulated at all time
points with the overexpression of PEP synthase. It is expected because there is ample

supply of PEP with the overexpression of PEP synthase.

111

The adk gene (Table 15) encoding adenylate kinase was downregulated with the
overexpression of PEP synthase at all time points. Combining pyruvate kinase, PEP
synthase and adenylate kinase together catalyzes reactions in which ATP is converted to
ADP and Pi (Figure 30). When PEP synthase was overexpressed, the downregulation of

adk can prevent the formation of such futile cycle that only consumes ATP.

Adk
\ ATP + AMP ——> 2 ADP
. P A
Futile > Pyruvate + ATP L PEP + AMP + R
cycle
P kAF
/ PEP +ADP y——> Pyruvate + ATP

 

 

ATP ——> ADP + F1

Figure 30. Futile cycle involving Adk (adenylate kinase), PpsA (PEP synthase) and
PykAF (pyruvate kinase).
Embden-Meyerhof pathway

Table 15 summarizes the expression data for all 15 genes involved in the
metabolism of glucose to pyruvate (Embden-Meyerhof pathway). Although results at
each time point were normalized for total bound cDNA, higher values of expression level
were observed for most of these genes during early fermentation (18 h and 24 h) than
during the later stages of fermentation (30 h and 36 h). For many genes such as pgi
encoding phosphoglucose isomerase, fbaA encoding fructose bisphoshate aldolase, gapA

encoding glyceraldehydes 3-phosphate dehydrogenase, pgk encoding phosphoglycerate

112

Table 15. Expression data for gluconeogenic genes and all the genes involved in
Embden-Meyerhof pathway (EMP).

Ratiou with IPTG

Gene 24 he 30 ht 24 h 30 h

G

10 nc 34 17

0.7 0.4 . . 0.5 0.5

adk . 0.4 0.4 . . 2.6 0.6
Embden-Meyerhof

6.7 2.2
2.7 1.8
4.2 1.9
2.9 4.3
6.8 3.9
12 5.7
1.4 9.9
4.2 2.2
2.1 2.3
3.0 3.2
15 12
1.6 1.6
30 11
2.1 1.5
6.0 3.1

 

a Expression ratio was calculated as (+IPTG/-IPTG). b percent cDN A x102.

° no, no change based on student t-test (p > 0.05).

kinase, tpiA encoding triose phosphate isomerase, eno encoding enolase, pku encoding
pyruvate kinase, expression levels declined to 10-30% of initial values during the later
stages of the fermentation both with IPTG addition and without IPTG addition (data not
shown).

Most glycolytic genes were thought to be unregulated, with expression levels
varying less than 2-fold.M However, many genes involved in glycolysis were judged to
express at higher levels in E. coli KL3/pJY1.216A with PEP synthase overexpression
during early fermentation (18 h and 24 h). Increases in expression levels of 12 genes in

Embden-Meyerhof pathway were significant (p < 0.05) at 24 h as shown in Table 15.

113

Two genes (eno encoding enolase and pfl<B encoding 6-phosphofructokinase) were
significantly higher with PEP synthase overexpression at all time points.

However, even most of the glycolytic enzymes have been overexpressed at
individual time points, the average glycolytic flux during the fermentation stays the same
as measured by the total amount of glucose consumed during the fermentation. As
suggested by KoebmannIS and Ingram,l6 the control of glycolytic flux resides not inside
but outside the pathway (demand for ATP), i.e., with the enzymes that hydrolyze ATP.
Because PEP synthase uses ATP to synthesize PEP, the ATP demand should be higher
for PEP synthase overexpression strain. It appeared that the ATP demand controls the
expression level of glycolytic enzymes but has no control over the glycolytic flux in the
system examined here. A possible explanation for this lack of stimulation of the
glycolytic flux could be that glycolysis is already running close to its maximum capacity
under the high density fed-batch fermentation conditions. A second possibility is that
there are still rate-limiting enzymes in the pathway that have not been overexpressed, for
instance in the transport of sugar into the cell or in the reactions that consume the

products of glycolysis.

Pentose pathway

Table 16 summarizes the expression data for 10 genes involved in both oxidative
and non-oxidative pentose pathway. Although results at each time point were normalized
for total bound cDNA, expression levels of 7 genes stay almost the same at all the stages
Of the fermentation. Higher values were observed for three genes (tktA encoding

transketolase, gnd encoding 6-phosphogluconate dehydrogenase and pgl encoding 6-

114

phosphogluconolactonase) during early fermentation (18 h and 24 h) than during the later
stages of fermentation (30 h and 36 h). The gnd gene encoding 6-phosphogluconate
dehydrogenase is known to be under the control of the growth rate.[7 The change of
growth rate corresponds to the change of expression level of gnd gene. One particular
interesting observation is that transketolase expression levels dropped down
approximately 10 fold during the later stages of fermentation (30 h and 36 h). One
possible explanation could be that the transketolase promoter is downregulated during the
later stages of fermentation. However, the transketolase specific activities during the
fermentation were pretty stable during the fermentation (approximately 0.02 U/mg, data
not shown). I

Most pentose pathway genes did not vary significantly with the overexpression of
the PEP synthase. Two exceptions are MM and talB encoding two isozymes of
transaldolase, which are known to enhance E4P supply.l8 Two transaldolase genes were
significantly higher with PEP synthase overexpression. The increase in expression levels
of transaldolase genes with overexpression of PEP synthase would be expected to result
in higher specific activities for transaldolase and may provide a physiological basis for
the higher rates of E4P supply relative to the rates of E4P supply with wild type
expression of PEP synthase. As higher rates of E4P supply coupled with higher rates of
PEP supply in “high yield” cells, more carbon flow will be expected to direct into the
shikimate pathway. However, the possibility that E4P supply is still limiting the carbon
flow to the shikimate pathway with amply supply of PEP cannot be ruled out.

Another interesting observation is that although the expression level of pgi

encoding phosphoglucose isomerase, the first enzyme in the Embden-Meyerhof pathway,

115

 

 

declined to 10% of initial values during the later stages of the fermentation (Table 15),
the expression level of zwf encoding g1ucose—6—phosphate dehydrogenase, the first

Table 16. Expression data for genes involved in pentose pathway.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Gene Ratio“ Expression with IPTG additionb

18 h° 24 hc 30 hc 36 hc 18h 24h 30h 36h

ppsA 7.4 10 nc nc 10 34 17 13

Pentosepathway

talA nc 2.3 1.4 nc 5.8 4 4.3 3.4
talB 1.2 1.3 3.1 1.6 2.1 2.1 3.5 1.8

L zwf nc no me 1.7 2.0 1.5 2.1 1.6
L gnd me me 1.4 no 4.7 5.0 2.1 1.5
rpe 1.4 nc no no 1.0 0.6 1.9 2.2
rpiA 0.9 1.2 nc nc 2.1 1.4 1.0 0.9
tktA no no 0.63 0.44 320 198 47 28
tktB 1.3 no 1.5 2.6 7.9 3.4 7.3 7.6
rpiB nc no no no 0.1 0.2 0.3 0.4
pg] nc 1.9 no 0.5 6.7 2.2 0.7 0.6

 

 

 

 

 

 

 

 

 

 

3 Expression ratio was calculated as (+IPTG/-IPTG). b percent cDNA x102.
C nc, no change based on student t-test (p > 0.05).

enzyme in the oxidative pentose pathway, remained the same at all stages of the
fermentation (Table 16). Since the ratio of wa and Pgi determine the carbon distribution
between Embden-Meyerhof pathway and pentose pathway, the transcription result
suggested that the oxidative pentose pathway played a more important role for the
metabolism of glucose during the later stages of the fermentation than the Embden-

Meyerhof pathway.

TCA Cycle

Table 17 summarizes the expression data for all 24 genes involved in the TCA
CE’Cle. Although results at each time point were normalized for total bound cDNA, higher
values were observed for most of these genes during early fermentation (18 h and 24 h)
than during the later stages of fermentation (30 h and 36 h). For many genes, expression

levels declined to 10-30% of initial values (18 h) during the later stages of the

116

fermentation. Only three genes (aceK encoding isocitrate dehydrogenase phosphatase,
fumB encoding fumarase, sucC encoding succinyl-CoA synthetase) appeared to be stable
during the fermentation.

Table 17 . Expression data for all genes involved in TCA cycle.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Gene Ratioa Ex ression with IPTG additionb
18 Inc 24 h“ 30 hc 36 h“ 18 h 24 h 30 h 36 h
ppsA 7.4 10 nc no 10 34 l7 l3
TCA cycle
gltA 0.8 1.3 1.4 1.5 21 9.6 4.5 3.1
aan nc 1.7 2.2 nc 3.4 2 1.3 0.7
aan nc 5.9 nc nc 8.7 6.7 2 2.5
1’ch 0.7 2 2.2 1.6 14 13 4.7 3.2
aceA 0.9 4.7 nc nc 17 10 1.4 1.8
aceB nc 4 2.6 2.3 23 11 3.5 2.9
aceK 1.3 nc nc nc 3.5 1.8 2.3 4.1
mdh 1.5 nc 1.8 no 15 8 6.4 2.9
ppc nc 1.7 1.8 no 10 2.4 1.6 l
umA nc 1.4 1.3 0.4 1.6 1.1 1 0.4
fumB 1.7 no no 2.3 0.8 0.3 1.3 1.9
fumC nc 5.7 no 06 1.2 3.3 l 0.9
.9th 0.8 2.1 nc nc 1.7 0.5 0.2 0.4
sdhC 0.7 1.9 no no 3.1 1.2 0.7 0.4
.9th 0.8 2.1 nc nc 5.4 2 0.9 0.6
sdhA 0.8 1.3 no nc 6.6 3.5 1.3 l
sucC 0.9 1.8 nc 7.8 9.3 2.3 6.3 9.7
sucD nc 3.6 no no 9.9 3 0.9 0.7
sucB 0.8 2.3 nc 0.6 6.1 1.6 0.6 0.5
sucA nc 3.7 no 0.3 l l 2.7 0.4 0.4
[_mqo nc nc 0.3 0.1 1.1 1.6 0.5 0.3
gch no no no 0.1 0.5 0.1 0.1 0.1
M nc nc nc 0.4 1.5 1.4 0.6 0.6

 

 

 

a Expression ratio was calculated as (+IPTG/-IPTG). b percent cDNA x 102.
C no, no change based on student t-test (p > 0.05).

At 18 h (exponential growth phase), some genes involved in the TCA cycle were
downregulated (0.7-0.9) while at 24 h (early stationary phase) most of the genes involved
in TCA cycle were upregulated by 2- to 6— fold in E. coli KL3/pJY1.216A with PEP
synthase overexpression. The glyoxylate shunt genes (aceB encoding malate synthase,
aceA encoding isocitrate 1yase) were significantly higher with PEP synthase

overexpression at most of the time points. Because carbon loss to carbon dioxide can be

117

avoided using glyoxylate shunt, more usage of this pathway can lead to higher product
yield.

Table 18. Expression data for global regulators.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Gene Ratio“ Expression with IPTG additionh
18 hc 24 hC 30 h“ 36 h“ 18 h 24 h 30 h 36 h
ppsA 7.4 10 nc nc 1.0 34 l7 l3
Regulators
arcA 0.9 nc 0.7 0.6 4.9 6.5 2.7 1.6
(1ch nc 1.3 no 0.3 3.6 1.6 0.7 0.4
cg) 1.2 no nc nc 0.7 1.0 0.9 0.7
cyaA nc 2.1 4.0 1.8 1.8 1.0 2.8 1.6
fadR 0.7 0.8 0.7 0.7 5.5 5.3 3.1 2.0
iclR nc 0.9 nc 1.5 1.8 2.4 2.3 4.1
fruR nc nc nc 1.4 1.0 0.9 1.9 2.1
csrA nc 1.3 0.5 0.5 2.2 1.9 1.8 0.9
filr nc nc 0.8 no 2.6 3.1 2.4 2.4

 

 

 

 

 

 

 

 

“ Expression ratio was calculated as (+IPTG/-IPTG). b percent cDNA x102.
C nc, no change based on student t-test (p > 0.05).

The up-regulation of most TCA cycle genes during the stationary phase was
expected since overexpression of PEP synthase consumed a lot of ATP that must be
compensated by the TCA cycle. These TCA cycle genes are known to be regulated by
several regulators, such as cAMP-CRP, ArcA, and Fnr.19 From the expression data for
those control genes (Table 18), it is likely that the observed up-regulation of TCA cycle
genes with PEP synthase overexpression was mediated by cAMP-CRP, as CAMP level
might be increased with PEP synthase overexpression because the cyaA gene encoding
adenylate cyclase that is responsible for the synthesis of cAMP was upregulated 2- to 4-
fold with PEP synthase overexpression during the stationary phase (24 h, 30 h and 36 h).
For the glyoxylate shunt genes (aceA and aceB), the downregulation of FadR repressor

protein may lead to their upregulation.

118

 

Table 19. Expression data for genes involved in the shikimate pathway.

 

Expression with IPTG additionb

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Eerie Ratio8
18 hc 24 hc 30 h“ 36 he 18 h 24 h 30 h 36 h
ppsA 7.4 10 nc nc 10 34 l7 l3
Shikimate pathway

trpA 0.9 nc 5.7 2.9 1.6 l 12 6.2
trpB nc 2.8 10 4 0.2 0.3 1.5 0.6
trpC no no 7.6 3.8 0.9 0.6 22 12
trpD nc 1.5 7.2 3.2 1.1 0.8 32 14
trpE 1.4 no 3.8 2.5 0.7 0.3 43 24
aroF nc 0.5 0.7 nc 262 549 780 1070
aroG 0.9 0.4 0.5 1.8 10 6.0 23 21
aroH nc nc nc 0.5 0.4 0.2 0.8 0.4
aroB nc 0.9 nc 0.8 5.6 6.5 6.3 5.7
aroD nc nc nc 0.3 1.0 0.5 0.3 0.2
aroK nc 1.3 nc 0.4 2.6 1.8 0.9 0.5
aroL 0.7 nc nc 0.7 0.9 1.6 1.6 1.]
aroA nc 1.3 no nc 3.2 2.4 2.4 2.8
aroC 0.9 1.7 nc 0.4 1.7 1.0 0.4 0.4
aroE nc me me 0.5 2.2 2.1 2.1 1.2
pheA 0.5 no 1.4 0.6 2.0 1.5 1.1 0.7
tyrA nc 0.8 nc nc 9.7 27 35 32 #

 

 

 

 

 

 

 

 

 

3 Expression ratio was calculated as (+IPTG/-IPTG). b percent cDNA x102.
C no, no change based on student t-test (p > 0.05).

Shikimate pathway

Table 19 summarizes the expression data for 17 genes involved in the shikimate
pathway. The total amount of aroF cDNA was a large portion of the total cDNA. At 36
h, it represents about 10% of the total cDNA, indicating that the promoter PmF is very
strong during the later stages of the fermentation. With PEP synthase overexpression, the
continuous increase of expression level of aroFFBR during the course of the fermentations
did not lead to the continuous increase of DAHP synthase specific acitivites during the

FFBR may be

course of the fermentations (Figure 31). This result suggests that Aro
sensitive to proteolysis during the stationary phase. The expression level of aroB
encoding 3-dehydroquinate synthase was very stable during the course of the

fermentations both with IPTG addition and without IPTG addition. Thus the promoter of

aroB may be used as a weak promoter to express certain genes in the future.

119

.O
C)
_.L
N

 

 

 

 

 

 

 

 

 

 

 

 

 

O
a 0.5 " _______________________________ 10 <
E z
D _____________________________ 02
E; 0.4 t 8 (3
IE 9
2 O 3 __________ . —————————————————————— 6 E
0 c
t: o 2 4 ———————— J ————————————————————— 4 a:
'8 8
(%- 0.1 --——g ______________________ _ _ 2 D.
O I I I T O
18 24 30 36 42
Time (h)

Figure 31. The specific activity of AroFFBR (Open bars) and aroFFBR-encoded le
transcript (filled circles) over the course of an E. coli KL3/pJY1.216A fermentor 1
with IPTG addition.

For the genes involved in the biosynthesis of tryptophan, tyrosine z
phenyalanine (trpABCDE, tyrA, pheA), expression levels increase to approximately 4-
20- fold of initial values (18 h) during the later stages of the fermentation (30 h and 36
indicating the cell’s need for those three aromatic amino acids after the initial additior
those three aromatic amino acids has been consumed. Phenylalanine and tyrosine w
consumed completely before tryptophan was consumed at 24 h. The 13er gene encod
chorismate mutase that leads to the biosynthesis of tyrosine and phenylalanine has be
induced 3 fold while the trpA gene encoding tryptophan synthase that leads to
biosynthesis of tryptophan had not been induced at 24 h.

It is surprising to see that in high yield cells, aroFFBR encoding the first enzyme
the shikimate pathway was downregulated at 24 h and 30 ’h compared to low yield ce

However, this downregulation can be an effective way to reduce the metabolic burden

the cells with overexpressing both DAHP synthase and PEP synthase. Genes in

120

tryptophan pathway (trpABCDE) were unexpectedly upregulated 1.4- to 10- fold in high

yield cells during the later stages of the fermentation (30 h and 36 h). The underlying

mechanism for this upregulation was unknown.

Table 20. Expression data for selected genes related to stress conditions.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Gene Ratio“ ‘ Ex ression with IPTG additionb
18 11c 24 hc 30 h‘ 36 hc 18 h 24 h 30 h 36 h
ppsA 7.4 10 nc nc 10 34 17 13
Molecular chaperone
dnaK nc 2.6 3.2 3.1 13 5.9 5.6 5.9
dnaJ nc nc 1.8 1.4 3.3 1.8 5.4 4.5
mopB 0.9 7.5 2.6 1.6 4.7 2.2 0.8 0.6
grpE nc 1.2 1.3 nc 2.8 2.2 2.0 2.0
clpB nc 1.5 1.9 2 4.4 2.6 2.6 3.5
Protease
[on no nc nc nc 1.5 1.2 1.2 1.0
Sigma Factor
rpoE 1.1 2.5 nc 2.1 1.3 0.9 0.9 2.5
rpoH 0 7 nc 1.8 1.3 2.9 3.8 11 6.5
Oxidative stress
katE 1.8 3.1 nc nc 8.9 4.0 1.7 1.7
katG nc nc nc nc 2.5 1.8 1.0 1.7
sod/1 nc 4.1 2.3 2.3 4.8 2.5 1.6 2.2
50018 nc nc 0.8 1.3 6.1 4.1 4.1 4.1
gldA 1.4 2.1 3.5 2.5 1.2 1.9 4.2 2.4
poxB 1.5 7.0 3.6 2.4 4.6 2.3 1.8 1.9
[91005 1.6 5.7 3.3 2.2 1.8 2.8 1.9 2.3
Osmotic stress
kdpA nc nc 4.3 18 1.0 0.4 28 32
Acid stress
adA 1.8 1.7 no 0.5 1.3 0.7 0.4 0.4
adB 2.3 no 0.6 0.6 4.5 1.3 1.3 1.2
xasA 3.4 me no 0.3 0.9 0.2 0.2 0.2
hdeA 4.7 3.3 0.5 0.3 8.3 1.6 0.4 0.2
hdeB 4.5 4.4 0.4 0.1 4.7 0.6 0.1 0.1
marA nc 2.8 1.3 1.3 7.2 15.5 3.3 3.3
marB nc 2.0 no no 1.4 2.0 1.0 1.0
DNA protection
(1 s 1.4 5.3 2 1.8 7.0 19 24 29

 

 

 

 

a Expression ratio was calculated as (+IPTG/—IPTG). hpercent cDNA x102.

C me, no change based on student t-test (p > 0.05).
Genes related to stress conditions

A general function of the heat shock protein is to monitor and respond to the state
Of protein folding in the cell. Many of the heat shock proteins are molecular chaperones,

Whose function is to bind to newly synthesized, partially folded or unfolded proteins and

121

promote their folding and refolding by limiting the nonproductive interactions that lead
aggregation or misfolding. Other heat shock proteins are proteases that function

degrade misfolded or abnormal proteins. From this point of view, it is qui
understandable that the heat shock proteins are required for cell viability when the cel
are under stress in the high cell density culture.

Table 20 summarizes the expression data for selected genes related to stre
conditions. In general, major heat shock genes (dnaKJ, grpE, nwa) were upregulatr
in high yield cells during the stationary phase (24 h, 30 h and 36 h) than the low yie
cells. It may reflect the fact that more proteins in high yield cells were produced than
low yield cells. However, the expression level of major protease gene [on was unchangc
in both high yield and low yield cells. The increased expression levels of rpoE and rap
(encoding 0E and (7”, respectively) may be responsible for the increased expression leve
of major heat shock genes.

The sod/1 gene encoding superoxide dismutase and the karE gene encodir
hydroperoxidase II that are related to peroxide damage were expressed at significant
higher value (2-4 fold) in high yield cells at individual time point. This may indicate tl
existence of higher concentration of superoxide inside the high yield cells. In literatu
the sodAsodB strain of E. coli exhibited an auxotrophy for three aromatic amino acir
(phenylalanine, tyrosine and tryptophan) and superoxide was show20 to interfere with tl
production of erythrose-4-phosphate by oxidizing an intermediate in the transketola:
reaction with a rate constant of ~106 M'IS".21 The upregulation of the superoxi(
removing genes (sodA and katE) in high yield cells indicated a very important role 1

suPeroxide protection for the biosynthesis of aromatic amino acids. Pyruva

122

 

dehydrogenase (poxB), glycerol dehydrogenase (gldA) and a hypothetic NADPI
dependent dehydrogenase (1)1005) were also expressed at significantly higher value it
high yield cells. These genes are both effective to resist the oxidative stress in E. col
(36:113.

A very important gene related to osmotic stress, kdpA encoding potassium-
transporting ATPase was expressed at higher levels (4-fold to 18-fold) in high yield cell:
during the later stages of the fermentation (30 h, 36 h). The most rapid response t(
osmotic upshock is a stimulation of uptake of K“. deA is an inducible system tha
serves to scavenge K+ when the ion is present at low concentrations.22 The highe:
expression level of kdpA gene may indicate the higher osmolarity inside the high yielc
cells than the low yield cells.

Glutamate decarboxylase (gadA and gadB) and the glutamate transporter (xas) are
particularly important for acid tolerance and the maintenance of internal pH. Expressior
of these genes is related to both pH and the abundance of organic acids.23 Expressior
levels for gadA and the gadB-xasA operon were 2-4 fold higher with PEP synthase
overexpression during early fermentation at 18 h and 24 h, consistent with the productior
of higher levels of organic acids with PEP synthase overexpression. Some other genes
related to acid stress such as acid-resistant unknown genes (hdeA, hdeB), multidrug
resistance genes (marA, marB) were also expressed higher in high yield cells with PE]?
synthase overexpression during early fermentation at 18 h and 24 h.

Almiron et al.24 identified a DNA-binding protein, designated Dps (DNA-binding
protein from starved cells), which is produced primarily in stationary-phase cells of E

coli. Dps forms spherical dodecamers, homologous to ferritins, which sequester and

123

protect DNA from oxidative stress, nuclease, and UV light. Dukan and Touati25 showe-
that mutations of recA, recB, and dps in E. coli rendered cells more sensitive to damag
from hydroxyl radicals generated by HOCl. This suggests that not only DNA repair, bt
also DNA protection by dps, is pivotal for survival in extreme conditions. In high yiel'
cells with PEP synthase overexpression, dps gene was expressed at significant highe
levels during the course of fermentation relative to the low yield cells. This resul
confirmed with the above expression data that the high yield cells were under highe

stress (acid stress and oxidative stress) than the low yield cells.

Transport Genes

Table 21 summarizes the expression data for some selected transport genes. Th1
glucose transport genes ptsHI, crr were all upregulated in high yield cells with PE]
synthase overexpression compared to in low yield cells. However, the total amount 0
glucose consumed during the fermentation did not change. At later stages of th
fermentation (30 h and 36 h), many PTS transport genes (ptsN encoding unknown PT:
enzyme 11 that regulates N metabolism, cmtB encoding PTS enzyme 11 for mannitol, srli
encoding PTS enzyme 11 for sorbitol, manZ encoding PTS enzyme 11 for mannose) an.
other transport genes (192968 encoding unknown transport protein, bcp encodin
bacterioferritin comigratory protein, bfr encoding bacterioferritin storage protein, pot]
encoding putrescine ABC transporter, kdpA encoding potassium-transporting ATPase
were up-regulated 2- to 70- fold in high yield cells compared to in low yield cells. Th
uIJI‘Cgulation of PTS transport genes may reflect the cell’s willingness to find mor

nutrients when PEP was abundant inside the cells.

124

Table 21. Expression data for selected transport genes.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

{—Gene Ratio“ Expression with IPTG additionh
18 hc 24h° 30 hC 36 hc 18h 24h 30h 36h #

72pm 7.4 10 nc no 10 34 17 13

PTS transport
ptsH nc 1.8 2.3 4.6 4.1 1 2.7 2.3
ptsl nc 1.7 2.7 1.4 10 3.4 4.9 2.3
erA nc nc nc 10 0.4 0.1 2.2 6
crr nc 2.2 2.6 no 2.6 1.8 2.0 0.7
ptsN 1.4 nc 4.1 73 4 2 63 79 %
cmtB nc 0.4 1.9 15 0.3 0.5 4 7.1
rnanZ nc nc 1.7 4.4 2.2 0.6 3.2 5.7
ptsG nc nc nc nc 0.3 0.1 0.2 0.1
gva 0.9 nc nc 3.4 3.0 3.0 4.3 7.5

Other transport
kdpA no no 4.3 18 1.0 0.4 28 32
[)2 968 nc nc 2.1 11 1.9 1.2 13 30
bcp nc nc 2.6 17 1.4 1.3 6.5 11 p
bfr nc 2.6 3.2 2.6 1.8 1.6 3.4 1.6
potH 1.4 0.6 2.6 5.6 1.0 0.9 5.2 5.9

 

a Expression ratio was calculated as (+IPTG/-IPTG). b percent cDN A x102.
c nc, no change based on student t-test (p > 0.05).

Conclusion

This work demonstrates the utility of the DNA microarray technology in studying
gene expression profiles under fed-batch fermentor conditions. The results showed that a
simple overexpression of PEP synthase led to dramatic expression changes of many
genes. Some interesting expression changes happened in the TCA cycle, pentose
pathway, Embden-Meyerhof pathway, shikimate pathway, cell processes (resistance to
oxidative stress, acidic stress and osmotic stress) and transport proteins.

This comparison between high yield cells and low yields cells shows that
transcript levels may not be used to infer fluxes in a quantitative manner. For example,
the increase of expression levels of many glycolysis genes at 24 h and 30 h does not
correspond to the increase of glycolytic flux as measured by the amount of glucose

consumed during the course of the fermentation (data not shown). However, changes in

transcript levels reveal the existence of significant regulation and suggest possible

125

strategies of genetic modification. This expression profiles, combined with previous
knowledge of metabolic pathway and fermentation results, identified many interesting
gene targets for further strain development. For example, the large amount of aroFFBR
cDNA suggests the degradation of AroFFBR during the stationary phase of the
fermentation. The downregulation of tktA gene during the later stages of the
fermentation suggests that the availability of E4P may be the limiting factor for driving
carbon flow to shikimate pathway. The upregulation of sod/l gene in high yield cells
suggests that superoxide accumulated during the fermentation may be a limiting factor
for driving the carbon flow to shikimate pathway because superoxide can blocking an
intermediate in the transketolase reaction.20 The upregulation of some unknown transport
genes such as ptsN and 192968 at 36 h in high yield cells generates interests about their
functions. In the following section, two proposed experiments toward enhancing

biocatalyst development will be discussed and the first proposed experiment was tested.

Experiments toward enhancing biocatalyst development

Feedback insensitive isozyme of DAHP synthase

The activity of DAHP synthase has attracted attention as a critical determinant of
carbon flow directed into the shikimate pathway.26 Given the importance of feedback
inhibition in determining the activity of DAHP synthase in the microbial cytoplasm, use
Of mutant isozymes of DAHP synthase that are insensitive to feedback inhibition by
aromatic amino acids has been a centerpiece of virtually all efforts to synthesize

Shikimate pathway intermediates and products in high yield. A density gradient chamber

126

was used to isolate an isozyme of tyrosine-sensitive DAHP synthase that had reduced
sensitivity to feedback inhibition by aromatic amino acids.27 The gene encoding this
feedback-insensitive DAHP synthase, which was designated aroFFBR, corresponded to a

feedback-insensitive isozyme of DAHP synthase previously reported by Herrmann.26b

Although aroFFBR-encoded DAHP synthase has been our workhorse in studies of

hydroaromatics synthesis, we routinely observed that the specific activity of DAHP
synthase declined over the course of fermentor runs (Figure 31).28' 13' ‘C What became
even more alarming were the results from microarray analysis described above. These
experiments revealed (Figure 31) a decline in DAHP synthase activity but an increase in
the aroFFBR transcript over the course of fermentor runs. By the end of the fermentation,
aroFFBR transcript accounted for over 10% of the total transcriptome. This observation
raises the possibility that AroFFBR declined in specific activity even though transcription
and translation of this gene may have taken place over the entire course of the
fermentation runs. AroFFBR might be even more sensitive to proteolysis than indicated by
the declined specific activities of this enzyme over the course of a fermentation run.

The literature suggests that phenylalanine-sensitive AroG is more resistant
towards proteolysis than tyrosine-sensitive AroF although the degree of increased
Stability is open to question.29 Herrmann has reported that AroF activity declines with a
half-life of 3 h“ while under the same conditions, the specific activity of AroG remains
unchanged.29a Pittard has reported a 5.5 h'1 half-life for the specific activity of AroF and a

7 h'I half-life for the specific activity of AroG.29b Given our observations relative to the

Specific activity and transcription of aroFFBR and the literature concerning the proteolytic

127

 

lability of AroF relative to AroG, feedback insensitive isozyme of aroG-encoded DAHP
synthase was isolated to compare with AroFFBR.

Table 22. Comparison of the amino acid changes and sensitivity to feedback
inhibition of AroG and AroF Mutants.

 

 

 

 

 

 

Mutant Amino acid change Feedback inhibition‘1
AroGl Ser—211 —> Tyr 33%
AroG2 Met-147 —> Lys 42%
AroG3 Pro-150 —> Thr 53%
AroG4 Val-221 —> Met 54%
AroFFBR Pro-148 —> Leu 100%

 

 

 

 

 

3 Feedback inhibition = 100%- (the specific activity of DAHP synthase isozyme in a
solution containing 4 mM Phe, 4 mM Tyr, and 4 mM Trp / the specific activity of DAHP
synthase isozyme in the absence of any added amino acids).

PCR mutagenesis30 was used to generate variants of aroG. The top proiority of
mutagenic PCR is to introduce the various types of mutations in an unbiased fashion
rather than to achieve a high overall level of amplification. The protocol for mutagenic
PCR is derived from standard PCR conditions. The following changes were made to
enhance the mutation rate: the MgCl2 concentration was increased from 1.5 mM to 7 mM
to stabilize noncomplementary pairs; 0.5 mM MnCl2 was added to diminish the template
Specificity of the polymerase; the concentration of dCTP and TTP was increased to 1 mM
to promote misincorporation; the amount of Taq polymerase was increased to 5 units to
promote chain extension beyond positions of base mismatches. The E. coli aroG ORF
from plasmid pJY6.119A (wild-type aroG ORF amplified from E. coli RB791 genomic
DNA) was subject to PCR mutagenesis and the resulting DNA was ligated to pJF118EH.

The resulting ligation mixture was introduced by transformation into E. coli CB734,

Which is deficient in DAHP synthase activity due to a AaroFAaroGAaroH phenotype.“

128

 

Growth of E. coli CB734 in the presence of 20 mM phenylalanine supplementation,
which is adequate to prevent the growth of E. coli CB734/pJY6.119A, indicates an
aroGFBR isozyme plasmid insert.32 From the selection plates 13 colonies were picked up
and then assayed for the DAHP specific activity in the presence of 4 mM phenylalanine,
tyrosine and tryptophan. The concentrations of the three aromatic amino acids in the
enzyme assay were chosen to mimic the concentrations of the three aromatic amino acids
in the fermentation medium. Four feedback-insensitive isozymes (Table 22) have the
highest activity in the presence of all three aromatic amino acids were sequenced. A
Met-147-Ile and a Pro-150-Leu mutation have previously been reported to give rise to
aroGFBR isozymes.33 The AroGl mutant (Ser-211-Tyr mutation), which has the highest
resistance towards feedback inhibition, does not correspond to a position in AroG
previously reported for an isozyme insensitive to feedback inhibition by phenylalanine.
The AroGl mutant also has almost 100% resistance towards feedback inhibition of each
4 mM aromatic amino acid (data not shown). One of the surprises in obtaining the
aroGFBR isozymes is that our current aroFFBR displayed poor resistance to feedback
inhibition at the higher concentration of aromatic amino acids used in these
determinations (4 mM Phe, Tyr and Trp). In literature, the maximum concentration of

the aromatic amino acids used in the examination of aroFFBR feedback inhibition was

only 0.25 mM.34

Biocatalyst design and fermentation
To examine whether AroGl is more stable over the course of fermentation runs
than AroFFBR, plasmid pJY6.231 (Figure 32) that carried aroGFBR, PmF, tktA and ppsA

inserts was assembled to mimic plasmid pJYl .216A (the best DHS producer). The aroF

129

 

promoter was used for expression of aroGFBR in order to minimize changes in

( H indIII )

(RBS)
tktA ParoF

 

 

 

 

 

BglII

 

pJY6.186

12.2 kb

 

EcoRI
Figure 32A. Construction of plasmid pJY6.186

130

pJY6.186

122 kb PCR from AroGl encoding plasmid
l 1) BamHl digest

 

 

 

 

 

BamHI BamHl
1.1 kb
aroGI
1) Bglllldigest
2) CIAP treatment
T4 Ligase
(Hindlll)

   
  
     

Sal]
(Bglll)

(Bglll)
ParoF
(SmaI)
serA
P
(Smal) tac
Eco RI ppSA 5603/

Figure 32B. Construction of plasmid pJY6.231

transcription levels. Both plasmids were transformed into E. coli strain KL3 and
evaluated under fermentor-controlled glucose-rich conditions as described in Chapter 2.
The specific activities of DAHP synthase during the fermentation of E. coli
KL3/pJY6.231 (Table 23B) were not only stable but also higher than the DAHP synthase
activities during the fermentation of E. coli KL3/pJY1.216A. However, no increase in

titer and yield of DHS biosynthesis was observed (Table 23A). A reasonable explanation

131

for the observed ceiling in synthesized DHS concentrations with increasing DAHP
synthase activity is that the availability of either intracellular PEP or intracellular E4P is
still limiting even with the overexpression of both transketolase and PEP synthase.

Table 23A. Concentrations and yields of 3-dehydroshikimic acid and byproducts

synthesized by E. coli KL3/pJY1.216A and KL3/pJY 6.231 cultured under glucose-
rich conditions.

 

DHS

 

Construct IPTG [DHS]" Yieldc [DAH] [DHQ] [GA] (22:31,
addnf’ (g/L) (mol/mol (g/L) (g/L) (g/L) (mol/mol)

KL3/pJY6.231 12 64 35% 13 9.2 9.5 50%

KL3/pJY1.216A 12 69 35% 1o 12 13 51%

 

“Amount of IPTG (mg) added at 18, 24, 30, and 36 h. bAbbreviations: 3-dehydroshikimic
acid (DHS), 3-deoxy-D-arabin0-heptulosonic acid (DAH), 3-dehydroquinic acid (DHQ),
gallic acid (GA). C(mol DHS)/(mol glucose consumed). d(mol DHS + mol DAH + mol
DHQ + mol GA)/ (mol glucose consumed).

Table 23B. DAHP synthase specific activities for E. coli KL3/pJY1.216A and
KL3/pJY 6.231 cultured under glucose-rich conditions.

 

 

 

construct IPTG DAHP synthase (U/mg)”
addn-b 24 h 36 h 42 h
KL3/pJY6.231 12 .87 0.96 1.0
KL3/pJY1.216A 12 .53 .27 .19

 

“Amount of IPTG (mg) added at 18, 24, 30, and 36 h. bOne unit (U) of DAHP synthase
corresponds to formation of 1 umole of DAHP per min at 37 °C.

In conclusion, a stable AroGFBR was obtained by PCR mutagenesis and has its
advantage over AroFFBR because of its stability over the course of a fermentation run.
Further study will include testing the stability of the aroGFBR-encoded transcript and the
aroFFBR-encoded transcript over the course of a fermentation run using real-time reverse
transcriptase-PCR (RT-PCR).35 Real-time reverse transcriptase-PCR has become the
method of choice for the rapid and quantitative examination of the expression of specific

genes. Even small changes in the aroGFBR-encoded transcript or aroFFBR-encoded

132

 

transcript can be rapidly, reproducibly, and statistically determined using this method.
Another advantage of real-time RT-PCR is that the integrity of the aroGFBR-encoded
transcript or aroFFBR-encoded transcript can be estimated by using different pairs of
primers in the real-time RT-PCR to quantitate RNA of different lengths. Also future
study will include an assay to test the stability of the AroGFBR protein and AroFFBR protein
over the course of the fermentation run. Before aroGFBR was incorporated into our
constructs, it would be critical to determine whether there is an aroGFBR isozyme superior

to our currently employed aroFFBR.

E-4-P limitation and superoxide protection

The key determinant of carbon flow directed into a biosynthetic pathway is often
the in vivo activity of the first enzyme in the pathway. For the common pathway of
aromatic amino acid biosynthesis, the in vivo activity of DAHP synthase is dictated by
feedback inhibition, transcriptional repression, and the availability of substrates E4P and
PEP. In the system examined above (Tables 23A and 23B), the observed ceiling in
synthesized shikimate pathway product concentrations with increasing DAHP synthase
activity indicates that either intracellular PEP availability or intracellular E4P availability
is still limiting the carbon flow to the shikimate pathway even with the overexpression of
both transketolase and PEP synthase. The availability of PEP is not likely to be the
limiting factor based on two observations. First, higher PEP synthase activity didn’t
correspond to high yields of shikimate pathway product as examined in Chapter 2.la
Second, the upregulation of many PTS transport proteins in high yield cells at late

stationary phase (36 h) from the microarray analysis suggested there are still ample

133

supply of PEP at late stationary phase with the overexpression of PEP synthase. Thus the
availability of E4P appeared to be the limiting factor under the fed-batch fermentor
conditions employed. Then the question becomes why the supply of E4P is still limiting
even with overexpression of transketolase. One answer is that transketolase gene tktA is
downregulated approximately lO-fold during the later stages of the fermentation as
supported by the microarray data (Table 16). However, the specific activities of
transketolase remained the same over the course of the fermentation run of E. coli
KL3/pJYl.216A (approximately 0.02 U/mg, data not shown). Another answer is that
accumulation of superoxide (oxidative stress) during the fermentation process inhibits the
synthesis of E4P by blocking an intermediate in the transketolase reaction.20 The second
answer is supported by two evidences. First, the upregulation of superoxide removing
genes (sodA and katE, Table 20) and E4P producing genes (talA and talB, Table 16) in
high yield cells suggests that E4P production correlates with superoxide removing.
Second, since 3-dehydroshikimic acid as an antioxidant can react with O2 and H202 to
form gallic acid both in vitro and in vivo (mechanism is still unknown), gallic acid
accumulation in the culture medium by the DHS producer can be viewed as an indication
of high oxidative stress inside the cells. Once gallic acid was accumulating significantly
in the culture medium during the later stages of the fermentation, the DHS biosynthesis
ceased or stopped.

To further increase the availability of E4P, two options are available: increase the
expression level of tktA gene by replacing the native tktA promoter with an enhanced
expression promoter such as aroF promoter in a new plasmid similar to pJYl.216A;

insert the superoxide removing gene sod/1 into pJYl.216A to construct another new

134

plasmid. These two plasmids will be transformed to E. coli KL3 and tested under fed-
batch fermentation conditions. The attendant increases in the concentrations of
synthesized DHS would be interpreted as indicating that carbon flow directed into the

shikimate pathway is limited by the availability of E4P.

135

REFERENCE

1 (a) Yi, J.; Li, K.; Draths, K. M.; Frost, J. W. Modulation of PEP synthase
Expression Increases Shikimate Pathway Product Yields in E. coli. Biotechnol. Prog.
2002, 18, 1141-1148. (b) Chandran, S. S.; Yi, J.; Draths, K. M.; von Daeniken, R.;
Weber, W.; Frost, J. W. Phosphoenolpyruvic Acid Availability and the Biosynthesis of
Shikimic Acid. Biotechnol. Prog. 2003, 19, 808-814. (c) Yi, J.; Draths, K. M.; Li, K.;
Frost, J. W. Altered Glucose Transport and Shikimate Pathway Product Yields in E. coli.
Biotechnol. Prog. 2003, 19, 1450-1459.

2 (a) Gonzalez, R.; Tao, H.; Shanmugam, K. T.; York, S. W.; Ingram, I. O. Global
gene expression differences associated with changes in glycolytic flux and growth rate in
Escherichia coli during the fermentation of glucose and xylose. Biotechnol. Prog. 2002,
18, 6-20. (b) Oh, M. K.; Rohlin, L.; Kao, K. C.; Liao, J. C. Global expression profiling of
acetate-grown Escherichia coli. J. Biol. Chem. 2002, 277, 13175-13183. (0) Oh, M. K.;
Liao, J. C. Gene expression profiling by DNA microarrays and metabolic fluxes in
Escherichia coli. Biotechnol. Prog. 2000, 16, 278-286.

3 (a) Yoon, S. H.; Han, M. —J.; Lee, S. Y.; Jeong, K. J.; Yoo, J. —S. Combined
Transcriptome and Proteome Analysis of Escherichia coli During High Cell Density
Culture. Biotechnol. Bioeng. 2003,81, 753-766. (b) Khodursky, A. B.; Peter, B. J.;
Cozzarelli, N. R.; Botstein, D.; Botstein, P. O.; Yanofsky, C. DNA microarray analysis of
gene expression in response to physiological and genetic changes that affect tryptophan
metabolism in Escherichia coli. Proc. Natl. Acad. Sci. USA. 2000, 97, 12170-12175.

4 (a) Tao, H.; Gonzalez, R.; Martinez, A.; Rodriguez, M.; Ingram, L. 0.; Preston, J.
F.; Shanmugam, K. T. Engineering a homo-ethanol pathway in Escherichia coli:
Increased glycolytic flux and levels of expression of glycolytic genes during xylose
fermentation. J. Bacteriol. 2001, 183, 2979-2988. (b) Tseng, G. C.; Oh, M. K.; Rohlin,
L.; Liao, J. C.; Wong, W. H. Issues in cDNA microarray analysis: quality filtering,
channel normalization, models of variation and assessment of gene effects. Nucleic Acids
Res. 2001, 29, 2549-2557.

5 Arfin, S. M.; Long, A. D.; Ito, E. T.; Tolleri, L.; Richle, M. M.; Paegle, E. S.;
Hatfield, G. W. Global gene expression profiling in Escherichia coli K12: the effect of
integration host factor. J. Biol. Chem. 2000, 275, 29672-29684.

6 Samules, M. L.; Witmer, J. A. Statistics for the Life Sciences. Prentice Hall, New
Jersey, 1999.

7 Hartigan, J. Clustering Algorithms. New York: John Wiley & Sons, 1975.
8 Eisen, M. B.; Spellman, P. T.; Brown, P. O.; Botstein, D. Cluster analysis and

display of genome-wide expression patterns. Proc. Natl. Acad. Sci. USA 1998, 95 , 14863-
14868.

136

9 Webb, C.; Scollon, S.; Miller, J.; Teh, B. T. Gene expression profiling of
endocrine tumors by microarray analysis. Current Opinion in Endocrinology & Diabetes
2003, 10, 162-167.

10 Biochemical pathways: an atlas of biochemistry and molecular biology; Gerhard
M. , Eds.; Wiley; Heidelberg; Spektrum; New York, 1999.

ll Karp, P. D.; Riley, M.; Saier, M.; Paulsen, l. T.; Collado-Vides, J.; Paley, S. M.;
Pellegrini-Toole, A.; Bonavides, C.; Gama-Castro. The EcoCyc Database. Nucleic Acids
Res. 2002, 30, 56-58.

12 Minoru, K. The KEGG database. NOVARTIS FOUNDATION SYMPOSIUM
2002, 247, 91-101; discussion 101-3, 119-28, 244-52.

13 Overbeek, R.; Larsen, N.; Pusch, G. D.; Dsouza, M.; Selkov, E.; Kyrpides, N.;
Fonstein, M.; Maltsev, N.; SELKOV, E. WIT: integrated system for high-throughput

genome sequence analysis and metabolic reconstruction. Nucleic Acids Res. 2000, 28,
123—125.

14 Fraenkel, D. G. Glycolysis. In Escherichia coli and Salmonella; Neidhardt, F. C.,
Ed.; ASM press: Washington, D. C., 1996; pp 189-198.

15 (a) Koebmann, B. J .; Westerhoff, H. V.; Snoep, J. L.; Nilsson, D.; Jensen, P. R.
The Glycolytic Flux in Escherichia coli Is Controlled by the Demand for ATP. J.
Bacteriol. 2002, 184, 3909-3916. (b) Koebmann, B. J .; Westerhoff, H. V.; Snoep, J. L.;
Solem, C.; Pedersen, M. B.; Nilsson, D.; Michelsen, O.; Jensen, P. R. The extent to

which ATP demand controls the glycolytic flux depends strongly on the organism and
conditions for growth. Mol. Biol. Reports 2002, 29, 41-45.

16 Causey, T. B.; Zhou, S.; Shanmugam, K. T .; Ingram, L. O. Engineering the
metabolism of Escherichia coli W3110 for the conversion of sugar to redox-neutral and
oxidized products: Homoacetate production. Proc. Nat. Acad. Sci. 2003, 100, 825-832.

17 Pearse, A. J .; Wolf, R. E. Determination of the growth rate-regulated steps in
expression of the Escherichia coli K-12 gnd gene. J.Bacteriol. 1994, 176, 115-122.

18 Lu, J .-L.; Liao, J. C. Metabolic engineering and control analysis for production of
aromatics: Role of transaldolase. Biotechnol. Bioeng. 1997 , 53, 132-138.

19 Cronan, J. E.; Laporte, D. Tricarboxylic acid cycle and glyoxylate bypass. In

Escherichia coli and Salmonella; Neidhardt, F. C. Ed.; ASM Press: Washington, D. C.,
1999; PP206-216.

137

20 Benov, L.; Fridovich, 1. Why Superoxide Imposes an Aromatic Amino Acid
Auxotrophy on Escherichia coli. J. Biol. Chem. 1999, 274, 4202-4206.

21 (a) Asami, S.; Akazawa, T. Enzyme Formation of Glycolate to Chromatium: Role
of Superoxide Radical in Transketolase-type Mechanism. Biochemistry, 1977, I6, 2201-
2207. (b) Takabe, T.; Asami, S.; Akazawa, T. Glycolate Formation Catalyzed by Spinach
Leaf Transketolase Utilizing the Superoxide Radical. Biochemistry, 1980, 19, 3985-3989.

22 Epstein, W.; Wieczorek, L.; Siebers, A.; Altendorf, K. Potassium Transport in E.
coli: genetic and biochemical characterization of the potassium-transporting ATPase.
Biochem. Soc. Trans. 1984, 12, 235-236.

23 Foster, J. W. Microbial responses to acid stress. In Bacterial Stress Responses;
Storz, G., Hengge-Aronis, R., Eds; ASM Press: Washington, 2000; pp 99-115.

24 Almiron, M.; Link, A. J .; Furlong, D.; Kolter, R. A novel DNA-binding protein
with regulatory and protective roles in starved Escherichia coli. Genes Dev. 1992, 6,
2646-2654.

25 Dukan, S.; Touati, D. Hypochlorous acid stress in Escherichia coli: resistance,
DNA damage, and comparison with hydrogen peroxide stess. J. Bacterial. 1996, 178,
6145-6150.

26 (a) Ogino, T.; Garner, C.; markley, J. L.; Herrmann, K. M. “Biosynthesis of
aromatic compoundszl3C NMR spectroscopy of whole Escherichia coli cells. Prac. Natl.
Acad. Sci. USA 1982, 79, 5828-5832. (b) Weaver, L. M.; Herrmann, K. M. Cloning of an
AroF Allele Encoding a Tyrosine-Insensitive 3-Deoxy-D-araabina-heptulosonate 7-
phosphate synthase. J. Bacterial. 1990, 172, 6581-6584.

27 Mikola, M. R.; Widman, M. T.; Worden, R. M. In situ mutagenesis and
chemotactic selection of microorganisms in a diffusion gradient chamber. Appl. Biochem.
Biotechnol. 1998, 70-72, 905-918.

28 Li, K.; Mikola, M. R.; Draths, K. M.; Worden, R. M.; Frost, J. W. Fed-Batch
Fermentor Synthesis of DHS Using Recombinant Escherichia coli. Biotechnol. Bioeng.
1999, 64, 61-73.

29 (a) DeLucia, A.; Schoner, R.; Herrmann, K. M. Selective Isoenzyme Inactivation
in Biosynthesis of Aromatic Compounds. Abstracts of Papers, Xith International
Congress of Biochemistry, Toronto, Canada, 1979, 04—2-S4. (b) Tribe, D. E.; Pittard, J.
Hyperproduction of Tryptophan by Escherichia coli: Genetic Manipulation of the
Pathways Leading to Tryptophan Formation. Appl. Environ. Microbial. 1979, 38, 181-
190.

138

30 Cadwell, R. C.; Joyce, G. F. Mutagenic PCR. PCR Methods Applic. 1994, 4,
$136-$139.

31 Akowski, J. P.; Bauerle, R. Steady-State Kinetics and Inhibitor Binding of 3-
Deoxy-D-araabina-heptulosonate-7-Phosphate Synthase (Tryptophan sensitive) from
Escherichia coli. Biochemistry 1997, 36, 15817-15822.

32 Yoshimi, K.; Kumiko, T.; Osamu, K. Mutational Analysis of the Feedback Sites
of Phenylalanine-Sensitive 3-Deoxy-D-arabina-Heptulosonate- 7-Phosphate Synthase of
Escherichia coli. Appl. Environ. Microbial. 1997, 63, 761-762.

33 (a) Kikuchi, T.; Sotochi, N.; Fukase, K.; Kojima, H.; Kurahashi, O.; matsui, Y.
Aromatic amino acid manufacture enhancement with recombinant microorganism. JP
Patent 04248983, 1993. (b) Tonouchi, N.; Kojima, H.; Matsui, H. The use of feedback-
insensitive enzymes in the production of aromatic amino acids by fermentation. Eur. Pat.
Appl. 488424, 1992.

34 Mikola, M. R.; Widman, M. T.; Worden, R. M. In situ mutagenesis and
chemotactic selection of microorganisms in a diffusion gradient chamber. Appl. Biochem.

Biotechnol. 1998, 70-72, 905-918.

35 Walker, N. J. A Technique Whose Time Has Come. Science 2002, 296, 557-558.

139

CHAPTER 4

Studies of shikimate export

Background

As is evident from current genome analyses, a substantial number of bacterial genes
encode membrane transport proteins. These proteins enable the controlled exchange of
molecules between the cell and its environment. In Escherichia coli, more than 400 genes
encode transport proteins based on sequence homology to known transport genes.l
However, a lot of these putative transport proteins are functionally undefined. Many
characterized transport proteins are used to import nutrients such as carbohydrates, ions,
amino acids, or peptides. Other proteins are known to act as exporters. Most export
proteins catalyze the export of noxious substances. Examples are multidrug resistance
proteins,2 metal resistance proteins,3 and polysaccharide export proteins.4 Recent studies
have shown that there are also proteins that catalyze the export of sugars5 and amino acids.6
The first export protein to be identified was LysE6C of Carynebacterium glutamicum.
Exression of LysE is necessary during growth on complex medium or in the presence of
peptides rich in L-lysine or L-arginine.7 Under such special growth conditions, L-lysine or
L-arginine might accumulate to toxic levels, a situation which is prevented by their export.
Both L-lysine and L-arginine are exported. by LysE at a rate of 0.75 nmol min“
(mg dry wt) ".7 Proteins homologous to LysE are widespread and occur in bacteria and
archaea.8

Studies on the peptide uptake systems of Staphylococcus aureus,9 Streptococcus
faealisi, and E. coli‘0 indicated that amino acids are exported that was components of
peptides initially consumed by these bacteria. The export of amino acids also occurred with
Lactacaccus lactis grown on milk11 or in the presence of milk-derived peptides.12 Those
phenomena suggested the existence of more exporters of amino acids. Examination of the

export of L-threonine from E. coli and the export of L-glutamate, L-lysine, L-isoleucine, and

140

L-threonine from Corynebacterium glutamicum suggests the existence of active

transporters. ‘ 3

Based on these studies, it recently became possible to identify specific
export carriers at the molecular level. This has led to the identification of novel transport
families and to insights into how bacteria control the intracellular concentrations of
metabolites. Some examples of identified export proteins for amino acids and sugars
include: (a) the ThrEl4 protein of Corynebacterium glutamicum, which is involved in
threonine export; (b) the RhtB15 protein of E. coli, which exports the homoserine lactones,
threonine and other amino acids; (c) the Orf2996b protein of E. coli, which is involved in
cysteine export; (d) the YedAS‘1 protein of E. coli, which exports L-arabinose; and (e) the
SetA5b protein of E. coli for the export of glucose.

The export protein serves as a valve that exports a metabolite that increases to
excessive intracellular concentrations. If the intracellular concentration of a microbe-
synthesized product exceeds the extracellular concentration, the rate of export may limit
product concentrations and yields. In the literature, the export of L-lysine,l6 L-threoninel4
and L-isoleucinel7 in Carynebacterium glutamicum was an impediment to the development
of strains that synthesized these amino acids. One strategy to achieve high-yield production
in such fermentation process is to increase the rate of export of the product amino acid. For
example, threonine production by Escherichia coli was enhanced by overexpression of the
E. coli rhtB and rhtC genes or by heterologous overexpression of the thrE gene encoding
the Corynebacterium glutamicum threonine exporter.18 Shikimic acid and 3-
dehydroshikirrric acid are hydrophilic molecules and are likely to export out of the cells by a
canier-mediated process instead of diffusion. In this chapter active export of shikimic acid
was implicated, the hypothesis that the rate of export constitutes a bottleneck during the

fermentation was tested, and two approaches to identify the carrier were explored.

141

Shikimate export is a carrier-mediated process and not a bottleneck for shikimate
production.

The functional analysis of export processes is more complicated than studies of
nutrient uptake. The use of labeled transport substrates, which is the method of choice for
kinetic analysis of uptake processes is restricted since most substrates are metabolized in
the cell. In general, analysis of export systems requires quantification of intracellular and
extracellular concentrations of the product or biosynthetic intermediate. In order to
establish intracellular concentrations, the correct cytoplasmic volume must be determined.
In general, this is carried out by using silicone oil centrifugation to separate the cell from the
culture medium. The cytoplasmic volume has been measured by double labeling techniques
and a cytoplasmic volume of 1.7 irL/ mg dry cell weight was used in our calculation.”
Because of the highly unfavorable volume ratio (internal volume<< external volume), the
exact quantification of intracellular concentration is difficult. A recently reported
chromogenic assay facilitated rapid and easy quantification of shikimic acid concentration.20

E. coli shikimate producers SP1.1/pKDl2.l38 and SP1.1/pKD15.071 were used to
examine shikimate export system by determining the intracellular and extracellular
shikimate concentrations during the fermentation.21 The host strain SP1.1 used for
synthesis of shikimic acid was derived from E. coli RB791 by homologous recombination
of the araB gene into the serA locus followed by successive P1 phage-mediated
transductions to introduce araL478::Tn]0 and araK::CmR mutations. The resulting
RB791 serAzzaraB araLzzTnIO aroK::CmR lacked 3-phosphoglycerate dehydrogenase
(SerA) activity as well as the activities of both isozymes of shikimate kinase (AroK and
AroL). In the absence of shikimate kinase, carbon flow directed into the common pathway
was unable to proceed beyond biosynthesis of shikimic acid, which accumulates
extracellularly in the culture medium. Plasmid pKD12.138 contained araFFBR, tktA, serA,
PmcaraE while pKD15.071 contained a ppsA gene in addition to all of the aforementioned

inserts in pKD12.138.

142

At timed intervals during the fermentation, an aliquot of E. coli cells was centrifuged
and the resulting cell pellet was resuspended in ice-cold buffer. In order to measure
intracellular concentrations, the cellular reactions have to be rapidly terminated by mixing
with a quenching agent such as perchloric acid. In the silicone oil centrifugation method,22
the cells are on top, the silicon oil layer is in the middle and the acid layer is at the bottom
prior to centrifugation. During centrifugation, the cells pass through the silicone oil layer
and are stripped of their surrounding medium. The cell pellets in the perchloric layer were
then disrupted by sonication, and the mixture was neutralized. After centrifugation, the
shikimic acid levels in the supernatant were determined using a chromogenic assay.20 In
this assay, shikimic acid was oxidized by periodic acid to give trans-aconitic acid and a
dialdehyde and then basicified to pH 10 with addition of N aOH to give an intense yellow
characteristic of dialdehyde formation. Only shikimic acid, quinic acid and tryptophan have
been found to produce a chromophore having a yellow color after treatment with periodic

acid and subsequent basicification.

Table 24. Shikimate intracellular and extracellular concentration during the
fermentation of SP1.1/pKD12.138 under glucose-rich conditions.

 

 

 

 

 

 

 

 

 

 

 

Time (h) on600 SAW (mM) SAEm (mM)

12 7.5 4 3

18 57 10 19
24 54 10 95
30 50 6 145
36 49 10 185
42 47 8 225
48 49 10 231
54 47 8 278
60 45 7 305

 

 

 

 

143

 

Table 25. Shikimate intracellular and extracellular concentration during the
fermentation of SP1.1/pKD15.071 under glucose-rich conditions.

 

 

 

 

 

 

 

 

 

 

 

 

 

Time (h) or)600 SAW, (mM) SAEm (mM)
12 41. 10 6
18 24 9 19
24 52 28 82
30 49 11 171
36 49 10 215
42 49 8 240
48 42 19 300
54 44 9 329
60 40 6 387
66 42 6 389
72 42 7 351

 

 

 

 

The intracellular shikimate concentration of 7-30 mM measured over the course of
fermentation runs of E. coli SP1.1/pKD12.138 and SP1.1/pKD15.071 was low compared
to the extracellular concentration of shikimate, which reached a maximum concentration of
approximately 400 mM (Tables 24 and 25). In the literature, Morbach et a1. measured
intracellular and extracellular concentration of L-isoleucine during cultivation of
Carynebacterium glutamicum SM13. The intracellular L-isoleucine concentration was
always higher relative to the extracellular concentration of L-isoleucine. The intracellular
concentration rose to 110 mM, whereas extracellularly the extracellular concentration of L-
isoleucine measured to only a 60 mM concentration.l 7 Similar analysis of the E. coli L-
threonine producer strain BKIIM B-3396/pVIC40 also revealed that the intracellular L-
threonine concentration exceeded the extracellular concentration over the course of the entire

fermentor run.lga At times, the 250 mM intracellular L-threonine concentration was 10 times

144

 

higher than the extracellular concentration. The low intracellular shikimate concentration
during its microbial synthesis under fermentor conditions has two implications. First, the
export of shikimate is not likely to be the rate-limiting step for the production of shikimic
acid since no accumulation of intracellular shikimic acid was observed. Second, because the
export of shikimic acid is not driven by the difference between the intracellular and
extracellular concentration, the export of shikimic acid is most likely a carrier-mediated

process.

Approaches to identify the loci involved in the export of shikimate

First approach: isolation of export-defective mutants by mini-Tn10 mutagenesis
The same export protein may exist for both shikimate and DHS because of their
structure similarity. The first approach for identification of loci involved in the export of
shikimate and DHS used mini—Tn10 mutagenesis of the shikimate-producing strain E. coli
SP1.4 (serA22araFFBRaraB araLzzTnIO)23 to disrupt loci involved in shikimate export.
Carbon flow directed into the shikimate pathway is increased in E. coli SP1.4 with the
genomic insertion of araFFBR into the serA locus. By inactivating only one of the two loci
encoding shikimate kinase with the araLzzTnIO gene disruption, E. coli SP1.4 accumulated
0.2 mM DHS and 0.3 mM shikimic acid. in its culture medium under shake flask conditions
(see experimental) while allowing enough carbon flow through the shikimate pathway to
provide all of its aromatic amino acid and aromatic vitamin requirements. Phenylalanine
(0.6 mM) and tyrosine (0.9 mM) were also accumulated in the culture medium by E. coli
SP1.4 while no tryptophan was accumulated in the culture medium. E. coli SP1.4 was
submitted to transposon mutagenesis24 using ANK1324 obtained from ATCC, which is a
mini-TnlO construct marked with an insert encoding for resistance to chloramphenicol.
Infection of E. coli SP1.4 with ANK1324 was followed by selection for resistance to

chloramphenicol. These chloramphenicol-resistant colonies were then plated on solid

145

medium containing E. coli AB2847, which is an araB mutant that requires aromatic
supplementation. Halos of E. coli AB2847 grew around E. coli SP1.4 mutants that retained
the ability to export DHS and shikimic acid. In contrast, no halo of E. coli SP1.1
(araKaraL mutant that requires aromatic supplementation) was observed around E. coli
SP1.4. Since the difference between E. coli SP1.1 and E. coli AB2847 is that E. coli SP1.1
cannot utilize DHS and shikimic acid for aromatics biosynthesis, DHS and shikimic acid
exported by E. coli SP1.4 was metabolized by E. coli AB2847 to provide the tryptophan
and aromatic vitamins requirement for this auxotroph. E. coli SP1.4 mutants that lost
export capability were among those lacking halos or those displaying reduced halo sizes.
For isolation of transposition events into the bacterial chromosome, bacteriophages
are the most convenient type of delivery vehicle. Phage ANK1324 carrying the transponson
mini—Tn10 (approximately 1.4 kb) can be introduced into the host cell E. coli SP1.4 under
conditions where the phage genome didn’t replicate, kill or stably integrates into the host
cell.24 Mini-Tn10 was chosen so as to ensure random, stable insertions into the E. coli
SP1.4 chromosome. Transposons that do not contain a transposase gene within their
boundaries are generally refered to as minitransposons. By using the rrrinitransposons such
as mini-Tn10, stable transposon insertions that are unable to undergo additional rounds of
transposition can be obtained. Another advantage of ANK1324 is its usage of mutant
transposases that exhibit a much lower degree of insertion specificity than wild-type.24 The
ANK1324 phage lysate was prepared using standard procedures and then used to infect E.
coli SP1.4 under conditions that preclude lysogen formation.25 The cells that have acquired
the chloramphenicol-resistance marker from the mini-TnlO were selected on rich plates
containing chloramphenicol. Normally 100 to 200 colonies per plate were obtained by
optimizing the ratio of phage to cells. Serial dilutions (“pics”) on rich plates containing
chloramphenicol were used to isolate single colonies. The single chloramphenicol-resistant
colonies were replicate plated on M9 salts solid medium containing L-phenylalanine, L-

tyrosine and L-serine along with strain E. coli AB2847 at a density of 107 cells per mL.

146

After 24 h of growth in the solid medium, colonies were inspected for reduced shikimic acid
or DHS export as indicated by reduced halo formation of E. coli AB2847. With the supply
of L-phenylalanine and L-tyrosine in the selection plate, E. coli SP1.4 mutants with the
reduced export of L-phenylalanine and L-tyrosine showed normal halo size. From
approximately 9000 isolated colonies, 16 colonies were selected that showed significantly
decreased halo formation. Of the selected colonies 8 of 16 showed normal growth on M9
salts solid medium with L-serine indicating no essential gene in these colonies was
disrupted by mini-Tn10 insertion mutagenesis. All the other 8 colonies were auxotrophic,

suggesting a cross feeding between E. coli SP1.4 mutants and E. coli AB2847.

Y-linker PCR technique to identify the sequences that flank the transposon
insertions

The method of Kwon et al.26 that requires the sequence information of only
transposon-specific sequences (Figure 33) was used for specific amplification of DNA
sequences that flank a transposon insertion. The genomic DNA isolated from a mini-TnlO
mutant was digested to completion with NlaHI, a restriction enzyme with a 4-bp recognition
site. The digested DNA was ligated to Y linker (Figure 33) designed to have a 3’ overhang
complementary to the sticky end generated by the NlaIII (CATG). PCR was then
performed using the Y-linker primer and mini-Tn10 specific primer. The Y-linker has a
region of noncomplementary sequence on the 5’ end. The Y-linker primer is made from
this Y region and cannot anneal to the linker itself. Therefore, the fragments that do not
have the mini—TnlO sequence were not amplified in the PCR reaction. However, the mini-
Tn10 primer anneals to the fragments containing mini-TnlO sequences during the first PCR
cycle and extends DNA synthesis into the Y region of the ligated Y linker, where the Y
linker primer can anneal. Starting from the second cycle, the fragments containing mini-

Tn10 sequences and some genomic sequences were selectively amplified. The function of

147

/ Y linker primer

A 5'-CTGCTCGAATTCAAGCTTCT-3'
Y linker

/

5'-TTTCTG CTCGAATTCAAGCTI'CTAACGATGTACGGGGACACATG-3'
QAAGATTGCTACATGCCCCTGT-S' T

1 P90 PKG N/a/II overhang

B Genomic DNA extraction from a transposon-encoding mutant

1

Digestion with restriction enzyme with 4 bp recognition site (Nla Ill)

 

 

 

 

 

 

l transposon specific sequence
. . . \
l Ligation to Y linker /—
/ \ / \
PCR with transposon-specific primer
_ and Y linker-specific primer
First cycle
No sequence for binding of \_——\_‘————/
transponson-specific primer \
l Transposon-specific primer
Second cycle

/ Y linker primer
No sequence for binding of
Y linker primer E‘—

No amplification Amplification
Figure 33. (A) Y-linker. (B) PCR of transponson-flanking sequences.

the noncomplementary region on the 5’ end of the Y linker was to prevent nonspecific

background amplification. If the noncomplementary region on the 5’ end of the Y linker

148

was not present, the digested fragments that do not have a mini-TnlO sequence also would
provide an annealing site for Y linker primer and would be amplified with two Y linker

primers annealed at both ends.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 26. SP1.4 mutants.
Mutant Growth on Location of .
ID M9/glu/ser Tn10 insertion Gene functron
1 + i [v I Acetolactate synthase isozyme 111 large
subunit
2 + 1.le Acetolactate synthase isozyme 111 large
subunrt
3 + narU Nitrite extrusion protein
4 + 1.le Acetolactate synthase isozyme 1H large
subunrt
5 + ydaQ Unidentified protein
6 + tch tchBC Operon transcriptional activator
S-adenosylmethionine-6-N', N‘-adenosyl
7 + ksgA (rRNA) dimethyltransferase '
8 + ykgH Unidentified protein
9 - argE Acetylornithine deacetylase
10 — argG Argininosuccinate synthase
11 - argH Argininosuccinate 1yase
12 - cysN Sulfate adenylyltransferase
13 _ ur H Aminoirnidazole carboxamide ribonucleotide
p transformylase / IMP cyclohydrolase
Cobalamin-independent homocysteine
14 - metE
Transmethylase
15 — cysN Sulfate adenylyltransferase
Phospho-hydroxy-threonine
16 ' p dXA dehydrogenase/decarboxylase

 

 

 

 

 

 

As expected, a PCR product with length of 0.2-1.0 kb was amplified for each mini-
Tn10 E. coli SP1.4 mutant. The PCR product was purified and was sequenced in the
Genomics Technology Support Facility at Michigan State University. The resulting mini-
Tn10 flanking sequences were compared in the National Center for Biotechnology

Information (NCBI) database to determine the genomic location of the mini-Tn10 insertion.

149

The results of 16 mutants are summarized in Table 26. Mutants 9-16 included in Table 26
were auxotrophic for different nutrients: amino acids, nucleotides and cofactors. The
mutated argE, argG and argH loci are involved in the biosynthesis of arginine. The
employed selection protocol thus seemed to be a surprisingly good way of selecting
mutations in genes encoding arginine biosynthetic proteins. One possible explanation is
that an arginine residue plays an important role in the protein-mediated export of DHS or
shikimic acid. If the supply of arginine is limited, the export machinery for DHS or
shikimic acid slows down. Of the 16 mutants, only NarU is annoted as a transport protein
(for nitrite export).27

To further characterize the 16 mutants, two methods were used. First, the
intracellular and extracellular shikimate concentrations of the 16 mutants were measured
under shake flask conditions. Secondly, the impact of the mutation under fed-batch
fermentation conditions was evaluated. The mini-TnlO insertion in the genome of SP1.4
mutants (gene knockout) was transferred to the DHS producer E. coli KL3 by P1 phage
transduction. Shikimic acid producer E. coli SP1.1 was not suitable for P1 transduction
because, of the chloramphenicol marker in the genome. The E. coli KL3 transductants were
evaluated for DHS production after transformation with plasmid pJY 1.216A under glucose-
rich fermentor conditions. The major difference between those two methods (shake flask
conditions and fed-batch fementation conditions) is that a much higher concentration of
DHS is accumulated extracellularly in the second method than the concentration of shikimic

acid accumulated in the first method.

Determination of intracellular shikimate concentrations of E. coli SP1.4 mutants
under shake flask conditions

E. coli SP1.4 mutants were grown in 100 mL LB medium for 12 h, washed with
100 mL M9 salts medium and resuspended in 100 rnL M9 salts medium supplement with

L-serine and glucose as the sole carbon source. After cultivation for 24 h at 37°C, the

150

intracellular and extracellular concentrations of shikimic acid were determined using silicon
oil centrifugation and chromogenic assay as described in the previous section. Table 27
shows the results of the 16 mutants. Some E. coli SP1.4 mutants (ilvl, ydaQ, tch, ksgA,
argE, argH, pdxA) retained a 2-3 fold more intracelluar shikimic acid concentration than the
control E. coli SP1.4. However, accumulation of shikimic acid did not reach levels
comparable to those reported during export of L—lysine,l 6 L-isoleucinel 7 and L-threonine.l 4

Table 27. Intracellular and extracellular shikimic acid accumulation of SP1.4
mutants under shake flask conditions.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mutant ID Locfiggé‘n‘ggnlo SA,mm (mM) SAEm (mM)

SP1.4 0.8 0.5
1 ile 2.4 0.4

2 ile 2.4 0.4

3 narU 1.1 0.4

4 ile 1.5 0.4

5 ydaQ 2.2 0.3

6 tch 1.8 0.4

7 ksgA 2.0 0.4

8 ykgH 1 .2 0.3

9 argE 1 .7 0.4
10 argG 1.4 0.4

1 1 argH 2.4 0.3
12 cysN 0.8 0.4
13 purH 1.0 0.4
14 metE 0.7 0.4
15 cysN 1.0 0.4
16 pdxA 1.6 0.5

 

 

 

 

151

 

Table 28. Production of DHS by KL3 derivatives/pJYl.216A under glucose-rich
conditions.

 

 

 

 

 

 

 

 

 

Construct DHS Titer DHS Yield Total Yield
(g/L) (‘70) (‘70)
KL3/pJYl.216A 69 35 51
KL3tch/pJY1216A 71 38 53
KL3ilv1/pJY1.216A 52 28 37
KL3narU/pJY1.216A 65 35 49
KL3ydaQ/pJY1.216A 69 36 51
KL3ksgA/pJY1.2l6A 46 34 43
KL3ykgH/pJY1.216A 61 35 49

 

 

 

 

DHS titers and yields by KL3 derivatives

The specific gene knockouts in E. coli KL3 were obtained by phage P1 transduction
using E. coli SP1.4 mutantsl-8 (see Table 26) as the donor strain. The E. coli KL3
derivatives with specific gene knockouts were transformed with plasmid pJYl.216A and
tested for the production of DHS under glucose-rich fed-batch fermentor conditions with
0.5 mL 100 mM IPTG added as described previously:28

The results (Table 28) indicated that except E. coli KL3ilv1 and KL3ksgA, other
KL3 mutants produced almost the same amount of DHS as the control strain E. coli
KL3/pJYl.216A. E. coli KL3ksgA/pJ Y1.216A grew poorly and the maximum dry cell
weight during the course of the fermentation run of E. coli KL3ksgA/pJY1.216A was only
two thirds of the maximum dry cell weight observed during the course of the fermentation
of E. coli KL3/pJY 1.216A (data not shown). E. coli KL3ilv1/pJYl.216A grew normally and
produced significantly less DHS with lower yield relative to the control strain E. coli
KL3/pJYl.216A (Table 28). However, the reason for reduced DHS production by E. coli
KL3ilv1/pJYl.216A is probably due to reduction of PEP synthase activity rather than the
disruption of shikimate export machinery. The PEP synthase activities for E. coli

KL3ilv1/pJY1.216A were determined to be much less throughout the fermentation relative to

152

 

the control fermentation of E. coli KL3/pl Y1.216A (Table 29). To prove this hypothesis,
plasmid pKL5.17A that didn’t overexpress the PEP synthase was transformed to KL3ile
and this time the ilvl gene knockout had no effect on the production of DHS relative to the
control strain E. coli KL3/pKL5.17A (data not shown).

Table 29. PEP synthase specific activities for fed-batch fermentation of

KL3ilv1/pJY1.216A and KL3/pJY 1.216A with 0.5 mL 100 mM IPTG addition under
glucose-rich conditions.

 

 

 

 

 

PEP synthase specific activity (U/mg)
Construct
24 h 36 h 42 h
KL3ilv1/pJY1.216A 0.02 0.02 0.02
KL3/pJY 1.216A 0.07 0.06 0.05

 

 

 

 

The above results raised the hypothesis that there might exist multiple export
carriers for shikimic acid (NarU may be one of them). The existence of multiple export
carriers is reasonable because in contrast to the export of toxic compounds and of
fermentative end products, there seems to be no need for the evolution of specific systems

for export of hydroaromatics under normal physiological conditions.

Second approach: resistant to fluoroshikimic acid

In the second approach, a plasmid library of E. coli genomic DNA was constructed
and probed for enhanced resistance towards the antibiotic (6S)-6-fluoroshikimic acid.29
The basis of this selection is the hypothesis that the shikimic acid export system can also
export the structurally similar molecule (6S)-6-fluoroshikirnic acid. The shikimate transport
system (ShiA) has been shown to also transport (6S)-6-fluoroshikimic acid.30 The
overexpression of a gene encoding the shikimic acid export system in a plasmid should
increase the rate of (6S)-6-fluoroshikimic acid export from the E. coli cytoplasm. Increased
export of (6S)-6-fluoroshikimic acid should result in growth in the presence of

concentrations of this antibiotic that completely prevent growth of wild—type E. coli. (6S)-6-

153

 

Fluoroshikimic acid was synthesized from shikimic acid by a modification3| of literature

procedures.32

Table 30. Identified genes to resist (6S)-6-fluoroshikimic acid toxicity.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

FSA in the plate
ID Responsible genes in plamsid
2ug/mL 10 ug/mL 50 ug/mL

1 yibA, yibJ + - -
2 b1497, yddA + - -
3 entC + +
4 ampC, frdBCD + -
5 b2432, b2433, b2434 + - _
6 b2420, cysM + - -
7 entC + + +
8 pepA, halC + - -
9 yidS + - -
10 b2353 + - -
11 halC + - _
12 entC + + +
13 yfiP. yfiQ + - -
14 fadL + + -
15 glvB, gva + - _
16 yifB, ile + - -
17 ppdABC + - -
18 entC + + +
19 yafO, yafN, dinP + - -
20 None - - -

 

 

 

 

 

A genomic library prepared from strain RB791 was constructed as

follows.

Chromosomal DNA was purified from a cell lysate.33 The purified DNA was partially

digested with Sau3A, and DNA fragments estimated to range in size from 3 to 10 kb were

isolated from an agarose gel. The resulting DNA was inserted into pBR322, which has

been digested with BamHI and dephosphorylated with calf intestinal alkaline phosphatase.

154

 

The resulting ligation mixture was introduced by transformation into XLlO-Gold
ultracompetent cells from Stratagene, and the resulting transformants were selected at 37 °C
on M9 minimal medium plates containing ampicillin and fluoroshikimic acid (0.5 ug/ mL).
The 0.5 ug/ mL fluoroshikimic acid is adequate to prevent the growth of wild-type XL10-
Gold cells. The plasmids that suppressed fluoroshikimic acid toxicity were isolated from
19 colonies that grew on the selection plate. The colonies were also replicate plated on solid
medium with increasing concentrations (2 ug/ mL, 10 ug/ mL and 50 11 g/ mL) of
fluoroshikimic acid (Table 30).

The plasmids that suppressed fluoroshikimic acid toxicity were sequenced in the
Genomics Technology Support Facility in Michigan State University and compared in the
National Center for Biotechnology Information (NCBI) database to determine the
responsible genes in the plasmid (Table 30). The only gene leading to resistance to 50
ug/mL fluoroshikimic acid is entC which encodes isochorismate synthase. In addition to
entC, overexpression of fadL, which encodes the transport protein for long-chain fatty acids,
or ampC/frdBCD, which encodes B-lactamase resistance and fumarate reductase in the host
strain led to resistance towards 10 ug/mL fluoroshikimic acid. There were a number of
genes leading to resistance towards 2 ug/mL fluoroshikimic acid. Some of those genes
encode unidentified proteins, but none of them are transport proteins based on sequence
homology. Some genes encode functional proteins such as DNA polymerase III (halC),
damage inducible protein (dinP), PTS transport protein (glvBC), or proteins involved in the
amino acid biosynthesis (ile and cysM). The difficulty of identifying a shikimate export
protein by the selection of fluoroshikimic acid resistance is probably because of the low
accumulation of fluoroshikimic acid inside the cells. If the export machinery inside the cells
is already functioning well, increased expression will not likely lead to a selectable
phenotype.

Fluoroshikimic acid appears to enter the cytoplasm of E. coli via the transport

155

HZO3PO

002H COZH COzH

phosphoenolpyruvate d1? AroK GR AroA
-—-> . —> \‘ —>
OH O HO“ i OH AroL HZOSPO‘ .=. OH
OH OH
. H
= R = H, shikimate R = H, S3P
“203130 OH R = F, fluoroshikimate R = F, fluoro sap

erythrose 4-phosphate

 

 

 

 

   
 

COZH COZH COZH
@121 AroC R UbiC © . .
JL —» JL ——> —>ub1qu1nones
Hzospo‘“ i o COZH i O COzH
OH OH OH
R = H, EPSP R = H, chorismate hydroxybenzoate
R = F, fluoro EPSP R = F, fluorochorismate
trpDE TyrA, PheA

   
   

ntC

 
 

C02H CO2H COZH HOZC’“ CO H
2
G NH2 .01 ©
0 CO2H 5-
OH
2-aminobenzoate isochorismate NH2
4-aminobenzoate prephenate
tryptophan menaquinones folic acid tyrosine

enterobactins phenylalanine

Figure 34. Summary of the aromatic biosynthetic pathway. Genetic loci are as
follows: AroK, shikimate kinase; AroL, shikimate kinase; AroA, EPSP synthase; AroC,
chorismate synthase; EntC, isochorismate synthase; PabAB, PABA synthase; TyrA,
chorismate mutase; PheA, chorismate mutase; TrpDE, anthranilate synthase; UbiC,
chorismate-pyruvate 1yase.

system (ShiA) for shikimate.30 It then becomes a substrate for shikimate kinase (AroK,

AroL) and 5-enol-pyruvyl-shikimate-3-phosphate synthase (AroA) and chorismate

156

synthase34 (AroC) thereby being transformed to 6-fluorochorismate (see Figure 34).35
Thereafter, although the individual steps have not been elucidated, it has been suggested that
6-fluorochorismate inhibits the synthesis of 4-aminobenzoate (Figure 34), which is a key
intermediate in the production of the folate coenzymes for one-carbon metabolism.29
Davies et a1. observed that addition of 4-aminobenzoic acid was sufficient to overcome the
growth inhibition of fluoroshikimic acid.29 Addition of aromatic amino acids did not
overcome fluoroshikirrric acid growth inhibition. In our study, the overexpression of entC
gene encoding isochorismate synthase led to resistance to the highest concentration of
fluoroshikimic acid (50 rig/mL). Our result suggests that the 6-flurochorimsate inhibited
the isochorismate synthase responsible for the production of menaquinones and
enterobactins. Reconciling this observation with the previously reported ability of 4-
aminobenzoic acid supplementation to reverse fluoroshikimate growth inhibition requires

further study.

Conclusion

Two approaches to identify the shikimate export protein were attempted without
success. The first approach attempted to identify a mutant deficient in shikimate export
while the second approach was based on the overexpression of a shikimate export protein.
The failure of both approaches is consistent with the possibility that multiple proteins exist
for the export of shikimic acid. Thus the first approach failed to identify the shikimate
export protein because mini-TnlO mutagenesis can only knock out one gene at a time and
the second approach failed because the intracellular concentration of (6S)-6-fluoroshikimic
acid that accumulated is below the Km of the exporter. Furthermore, the determination that
intracellular shikimic acid concentrations are much lower than extracellular concentrations
suggests that export of this hydroaromatic is not a limiting factor during microbial

synthesis.

157

REFERENCE

1 Blattner, F. R.; Plunkett, G.; Bloch, C. A.; Pema, N. T.; Burland, V.; Biley, M.; et al.
The complete genome sequence of Escherichia coli K-12. Science 1997, 277, 1453-1474.

2 Zgurskaya, H. 1.; Krishnamoorthy, G.; Tikhonova, E. B.; Lau, S. Y.; Stratton, K. L.
Mechanism of antibiotic efflux in Gram-negative bacteria. Frontiers in Biascience 2003, 8,
$862-$873.

3 Nies, D. H. Efflux-mediated heavy metal resistance in prokaryotes. FEMS
Microbial. Rev. 2003, 27, 313-339.

4 Hvorup, R. N.; Winnen, B.; Chang, A. B.; Jiang, Y.; Zhou, X.; Saier, M. H. The
multidrug/oligosaccharidyl-lipid/polysaccharide (MOP) exporter superfamily. Eur. J.
Biochem. 2003, 270, 799-813.

5 (a) Bost, S.; Silva, F.; Belin, D. Transcriptional activation of ydeA, which encodes a
member of the major facilitator superfamily, interferes with arabinose accumulation and
induction of the Escherichia coli arabinose PBAD promoter. J. Bacterial. 1999, 181, 2185-
2191. (b) Liu, J. Y.; Miller, P. F.; Willard, J.; Olson, E. R. Functional and biochemical
characterization of Escherichia coli sugar efflux transporters. J. Biol. Chem. 1999, 274,
22977-22984.

6 (a) Aleshin, V. V.; Zakataeva, N. P.; Livshits, V. A. A new family of amino-acid-
efflux proteins. Trends Biochem. Sci. 1999, 24, 133-135. (b) Dabler, T.; Maier, W;
Winterhalter, C.; Bock, A. Identification of a major facilitator protein from Escherichia coli
involved in efflux of metabolites of the cysteine pathway. Mal. Microbial. 2000, 36, 1101-
1112. (c)Vrljic, M.; Sahm, H.; Eggeling, L. The LysE superfamily: topology of the lysine
exporter LysE of Carynebacterium glutamicum, a paradigm for a novel superfarrrily of
transmembrane solute translocators. J. Mal. Microbial. Biotechnol. 1999, I, 327-336. ((1)
Eggeling, L.; Sahm, H. New ubiquitous translocators: amino acid export by
Carynebacterium glutamicum and Escherichia coli. Arch. Microbial. 2003, I80, 155-
160.

7 Bellmann, A. M.; Vrljic, M.; Patek, M.; Sahm, H.; Kramer, R.; Eggeling, L.
Regulation and specificity of the lysE-mediated export of amino acids by Carynebacterium
glutamicum. Microbiology 2001, 147, 1765-1774.

8 Vrljic, M.; Sahm, H.; Eggeling, L. A new type of transporter with a new type of
cellular function: L-lysine export from Carynebacterium glutamicum. Mal. Microbial.
1996, 22, 815-826.

9 Nisbet, T. M.; Payne, J. W. The characteristics of peptide uptake in Streptococcus
faecalis: studies on the transport of natural peptides and antibacterial phosphonopeptides. J.
Gen. Microbial. 1982, 128, 1357-1364.

10 Payne, J. W.; Bell, G. Direct determination of the properties of peptide transport
systems in Escherichia coli, using a fluorescent-labeling procedure. J. Bacterial. 1979, 137,
447-455.

11 Juillard, V.; Le Bars, D.; Kunji, E. R. S.; Konings, W. N.; Gripon, J.; Richard, J.

Oligopeptides are the main source of nitrogen for Lactacaccus lactis during growth on
milk. Appl. Environ. Microbial. 1995, 61, 3024-3030.

158

12 Kunji, E. R. S.; Mierau, 1.; Poolman, B.; Konings, W. N.; Venema, G.; Kok, J. Fate
of peptides in peptidase mutants of Lactacacc us lactis. Mal. Microbiol. 1996, 21, 123-131.

13 Kramer, R. Secretion of amino acid by bacteria: physiology and mechanism. FEMS
Microbial. Rev. 1994, 13, 75-94.

14 Simic, P.; Sahm, H.; Eggeling, L. L—Threonine Export: Use of Peptides To Identify a
New Translocator from Corynebacterium glutamicum. J. Bacteriol. 2001, 183, 5317-5324.

15 Zakataeva, N. P.; Aleshin, V. V.; Tokmakova, I. L.; Troshin, P. V.; Livshits, V. A.
The Novel Transmembrane Escherichia coli Proteins Involved in the Amino Acid Efflux.
FEBS Lett. 1999, 452, 228-232.

16 Vrljic, M.; Sahm, H.; Eggeling, L. A new type of transporter with a new type of
cellular function: L-lysine export from Corynebacterium glutamicum. Mal. Microbial.
1996, 22, 815-826.

17 Morbach, S.; Sahm, H.; Eggeling, L. L-Isoleucine Production with Corynebacterium
glutamicum: Further Flux Increase and Limitation of Export. Appl. Environ. Microbial.
1996, 62, 4345-4351.

18 (a) Kruse, D.; Kramer, R.; Eggeling, L.; Rieping, M.; Pfefferle, W.; Tchieu, J. H.;
Chung, Y. J .; Jr Saier, M. H.; Burkovski, A. Influence of threonine exporters on threonine
production in Escherichia coli. Appl. Microbial. Biotechnol. 2002, 59, 205-210. (b) Simic,
P.; Willuhn, J .; Sahm, H.; Eggeling, L. Identification of glyA (Encoding Serine
Hydroxymethyltransferase) and Its Use Together with the Exporter ThrE To Increase L-
Threonine Accumulation by Carynebacterium glutamicum. Appl. Environ. Microbial.
2002, 68, 3321-3327.

19 (a) Rottenberg, H. The measurement of membrane potential and pH in cells,
organelles, and vesicles. Methods Enzymal. 1979, 55, 547-569. (b) Kashket, E. R. The
proton motive force in bacteria: A critical assessment of methods. Annu. Rev. Microbial.
1985,39, 219-242. (c) Ebbighausen, H.; Weil, B.; Kramer, R. Isoleucine excretion in
Carynebacterium glutamcum: evidence for a specific efflux carrier system. Appl.
Microbial. Biotechnol. 1989,31, 184-190. (d) Zakataeva, N .; Aleshin, V. V.; Tokmakova, I.
L.; Troshin, P. V.; Livshits, V. A. FEBS Lett. 1999, 452, 228-232.

20 Cromartie, T. H.; Polge, N. D. In US; (Syngenta Limited, UK). Method of
Detecting Shikimic Acid. Int. Patent WOOO/49399, August 24, 2000.

21 Chandran, S. S.; Yi, J .; Draths, K. M.; von Daeniken, R.; Weber, W.; Frost, J. W.
Phosphoenolpyruvic Acid Availability and the Biosynthesis of Shikimic Acid. Biatechal.
Prag. 2003, 19, 808-814.

22 (a) Klingenberg, M.; Pfaff, E. Means of terminating reactions. Methods Enzymal.
1967, 10, 680—684. (b) Ebbighausen, H.; Weil, B.; Kramer, R. Transport of branched-chain
amino acids in Carynebacterium glutamicum. Arch. Microbial. 1989, 151, 238-244.

23 Draths, K. M. Unpublished results.

24 Kleckner, N.; Bender, J.; Gottesman, S. Uses of Transposons with Emphasis on
Tn10. Meth. Enzymal. 1991, 204, 139-180.

159

25 Miller, J. H. A short course in bacterial genetics; Cold Spring Harbor Laboratory:
Plainview, NY, 1992.

26 Kwon, Y. M.; Ricke, S. C. Efficient Amplification of Multiple Transposon-Flanking
Sequences. J. Microbial. Methods 2000, 41, 195-199.

27 Clegg, S.; Yu, F.; Griffiths, L.; Cole, J. A. The roles of the polytopic membrane
proteins NarK, NarU and NirC in Escherichia coli K-12: two nitrate and three nitrite
transporters. Mal. Microbial. 2002, 44, 143-155.

28 Yi, J.; Li, K.; Draths, K. M.; Frost, J. W. Modulation of Phosphoenolpyruvate
Synthase Expression Increases Shikimate Pathway Product Yields in E. coli. Biotechnol.
Prag. 2002, 18, 1141-1148.

29 Davies, G. M.; Barrett-Bee, K. J .; Jude, D. A.; Lehan, M.; Nichols, W. W.; Pinder,
P. E.; Thain, J. L.; Watkins, W. J.; Wilson, R. G. (6S)-6-Fluoroshikimic Acid, an
Antibacterial Agent Acting on the Aromatic Biosynthetic Pathway. Antimicrob. Agents and
Chemo. 1994, 38, 403—406.

30 Ewart, C. D. C.; Jude, D. A.; Thain, J. L.; Nichols, W. W. Frequency and
Mechanism of Resistance to Antibacterial Action of ZM240401, (6S)-6-Fluoro-Shikimic
Acid. Antimicrob. Agents and Chemo. 1995, 39, 87-93.

31 Tian, F. Mechanism-Based Inhibition of 3-Dehydroquinate Synthase and myo-
Inositol 1-Phosphate Synthase. Ph. D. Thesis, Michgan State University, 1998.

32 (a) Sutherland, J. K. Watkins, W. J. Bailey, J. P.; Chapman, A. K. Davies, G. M.
The Synthesis of 6a- and 6|3—Fluoroshikimic Acids. J. Chem. Soc. Chem. Commun. 1989,
18, 1386-1387. (b) Bowles, S.; Campbell, M. M.; Sainsbury, M.; Davies, G. M. Reactivity
Studies in the Shikimic Acid Series. Tetrahedron Lett. 1989, 30, 3711-3714. (c) Bowles, S.
A.; Campbell, M. M.; Sainsbury, M.; Davies, G. M. Reactivity Studies in the Shikimic Acid
Series: The Synthesis of Racemic Methyl 6a-Fluoroshikimate. Tetrahedron 1990, 46,
3981-3992.

33 Pitcher, D. G.; Saunders, N. A.; Owen, R. J. Rapid Extraction of Bacterial Genomic
DNA with Guanidium Thiocyanate. Letters in Applied Microbiology 1989, 8, 151-156.

34 Bomemann, S.; Ramjee, M. K.; Balasubramnian, S.; Abell, C.; Coggins, J. R.;
Lowes, D. J .; Thorneley, R. N. F. Escherichia coli Chorismate Synthase Catalyzes the
Conversion of (6S)-6-Fluoro-5-enolpyruvylshikimate-3-phosphate to 6-Fluorochorismate.
J. Biol. Chem. 1995, 270, 22811-22815.

35 Bomemann, S.; Ramjee, M. K.; Lowe, D. J.; Thorneley, R. N. F.; Coggins, J. R.;
Abell, C.; Balasubramanian, S.; Hawkes, T. R.; Nichols, W. W.; Davies, G. M. Studies on
Escherichia coli chorismate synthase. p. 843-846. In K. Yagi (ed.), Flavins and
flavoproteins 1993. Walter de Gruyter, Berlin.

160

CHAPTER 5

Experimental

General methods

Spectroscopic measurements

1H NMR spectra were recorded on a Varian 300 MHz VX-300 FT-NMR
spectrometer. Chemical shifts were reported in parts per million (ppm) downfield from
internal sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4 (TSP, 6 = 0.00) with D20 as
solvent. TSP was purchased from Lancaster. UV and visible measurements were
recorded on a Perkin-Elmer Lambda 3b UV-vis spectrophotometer or on a Hewlett
Packard 8452A Diode Array Spectrophotometer equipped with HP 89532A UV-Visible

Operating Software.

Bacteria strains and plasmids

All the strains and plasmids used are shown in Table 31. E. coli K-12 strain
RB791 was obtained from the American Type Culture Collection (ATCC strain 53622).
E. coli LE392 was obtained from the American Type Culture Collection (ATCC number
53010). E. coli AB2834,l JM2053, AB2847l were obtained from the E. coli Genetic
Stock Center at Yale University. E. coli KL3),2 SP1.1,3 SP3] and SP1.4 were
constructed in the lab previously. E. coli AL0807 was provided by Professor M. G.
Marinus (University of Massachusetts). E. coli UE794 was provided by Professor W.

Boos (University of Konstanz). E. coli CB7345 was provided by Professor Akowski.

161

Table 31. Bacterial strains and plasmids.

 

 

Strain/Plasmid Relevant Characteristics Source
Strain

RB791 W3110 lacL8l" ATCC
LE392 F supE44 supF58 lacYI galKZ galT22 metBl trpR55 hst514 mch A ATCC
AB2834 tsx-352 gan42 A‘ araE353 Ina/T352 CGSC
JM2053 ptsGO manY0(Am) uhp-0(Canst) fruAO thyAO galP77::TnIO ilvD kgaO CGSC

nagO mglPO umgCO pe[0( Am )
AB2847 tsx-354 gan42 A‘ araB351 Ina/T354 CGSC
KL3 AB2834 serA::araB Ref. 2
SP1.1 RB791 serA::araB araL478::Tn/0 araK I 7::C 111” Ref. 3
SP3.1 KL3 aral/I78::Tn10 araKl 7::ka Lab
SP1.4 RB791 serA::araFFBRaroB araL::TnlO tyrR3 Lab
AL0807 F araL478::TnIO araK17::CmR Marinus
UE79 galR glK thi AptsHIcrr zfc-706::Tn10 Ref. 4
CB734 C600 leu thiI A(gal-araG-nadA)50 araF.'car (Cm’) AaraH::Kan’ recAl Ref. 5
XL-lO Gold {A(mcrA)183 A(mchB-hstMR-mrr)l73 endAl supE44 thi-I recAl Stratagene

gyrA96 relAI lac Hte [F ’ praAB lacl"ZAM15 Tn10 Amy Cam’]
DH501 F ¢801acZAM 15 A(lacZYA-argF) U169 recAl endAI hst] 7(r,,_, m“) Invitrogen

phaA A” supE44 thi-I gyrA96 relAI

Plasmid

pJFl 18EH ApR, PM. lac!" Ref. 6
pTC325 nglfglk Ref. 7
pD2625 serA source GCI
pSUl8 CmR, P,m.lacZ', p 15A replicon Ref. 8
pBR322 ApR, TetR, rep replicon Ref. 9
p34H ApR Ref. 10
pKL1.87A CmR, ppsA Lab
pKL4.66B CmR, araFFBR, araFFBR, serA in pSUl8 Ref. 2
pKDl 1.291A CmR, serA, araFFBR, Pump Ref. 2
pKL5.17A CmR, tktA in pKDl 1.29lA Ref. 2
pMFS 1A ApR, tktA Lab
pMF63A ApR, araFFBR Lab
pKD12.138 ApR, araFFBR, tktA, PmaraE, serA Ref. 11
pKDlS.07lB ppsA, araFFBR, tktA, PmaroE. serA Ref. 11

 

E. coli XLlO-Gold was obtained from Stratagene. Plasmid constructions were carried out
in E. coli DH5, which is available from Invitrogen.

Plasmid pJF118EH6 was provided by Professor M. Bagdasarian (Michigan State
University). Plasmid pTC3257 was provided by Professor Lonnie 0. Ingram (University

Plasmids

of Florida). Plasmid pD2625 was obtained from Genencor International.

pSU18,8 pBR322,9 and p34Hl0 were obtained previously by this lab. Plasmids
pKL1.87A, pKL4.66B,2 pKD11.291A,2 pKL5.17A} pMF51A,“ pMF63A,”

pKD12.138,12 pKD15.07lBl2 were constructed in the lab previously.

162

Storage of bacterial strains and plasmids

All bacterial strains were stored at -78 °C in glycerol. Plasmids were transformed
into DHSOL for long-term storage. Glycerol samples were prepared by adding 0.75 mL of
an overnight culture to a sterile vial containing 0.25 mL of 80% (v/v) glycerol. The

solution was mixed, left at room temperature for 2 h and then stored at -78 °C.

Culture medium

All solutions were prepared in distilled, deionized water. LB medium” (1 L)
contained Bacto tryptone (10 g), Bacto yeast extract (5 g), and NaCl (10 g). L-Brothl3 (l
L) contained Bacto tryptone (10 g), Bacto yeast extract (5 g), NaCl (5 g), glucose (1 g)
and CaCl2 (2.5 mM). Soft agar” (100 mL) contained Bacto tryptone (1 g), Difco agar
(0.55 g), and NaCl (0.5 g). TB medium” (1 L) contained tryptone (10 g) and NaCl (5 g).
After autoclaving and directly before use, MgSO4 (10 mL of 1 M stock per L) was added
to the TB medium. M9 salts” (l L) contained NazHPO4 (6 g), KH2PO4 (3 g), NH4C1 (1
g), and NaCl (0.5 g). M9 medium contained carbon sources (D-glucose, D-xylose, D-
maltose or D-mannitol, 10 g), MgSO4 (0.12 g), and thiamine (0.001 g) in 1 L of M9 salts.
Solutions of inorganic salts, magnesium salts, and carbon sources were autoclaved
separately and then mixed. Antibiotics were added where appropriate to the following
final. concentrations unless noted otherwise: chloramphenicol, 20 ug/mL; ampicillin, 50
ug/mL; kanamycin, 50 ug/mL; tetracycline, 12.5 ug/mL; and fosfomycin, 20 rig/mL.
Stock solution of antibiotics were prepared in water with the exception of
chloramphenicol which was prepared in 95% ethanol and tetracycline which was

prepared in 50% aqueous ethanol. L-Phenylalanine, L-tyrosine, L-tryptophan, and L-

163

serine were added to M9 medium where indicated to a final concentration of 0.04 g/L.
Antibiotics, isopropyl B-D-thioglucopyranoside (IPTG), thiamine, CAMP and amino acid
supplementations were sterilized through 0.22-um membranes prior to addition to M9
medium. Solid medium was prepared by addition of 1.5% (w/v) Difco agar to the
medium. Fermentation medium (1 L) contained KZHPO4 (7.5 g), ammonium iron (III)
citrate (0.3 g), citric acid monohydrate (2.1 g), L-phenylalanine (0.7 g), L-tyrosine (0.7 g),
and L-tryptophan (0.35 g), and concentrated H2304 (1.2 mL). The culture medium was
adjusted to pH 7.0 by addition of concentrated NH4OH before autoclaving. The following
supplementations were added immediately prior to initiation of the fermentation: glucose
(19-24 g under glucose-limited conditions or 30 g under glucose-rich conditions), MgSO4
(0.24 g), aromatic vitamins p-aminobenzoic acid (0.01 g), 2,3-dihydroxybenzoic acid
(0.01 g), and p-hydroxybenzoic acid (0.01 g), and trace minerals (NH4)6(Mo7Oz4)-4HZO
(0.0037 g), ZnSO4-7HZO (0.0029 g), H3BO3 (0.0247 g), CuSO4-5HZO (0.0025 g), and
MnC12-4HZO (0.0158 g). D-Glucose and MgSO4 were autoclaved separately while
aromatic vitamins and trace minerals were sterilized through 0.22-um membranes prior

to addition to the medium.

Fed-batch fermentation (general)

Fermentationsl4 employed a 2.0 L working capacity B. Braun M2 culture vessel
fitted with a stainless steel baffle cage consisting of four 1/2” x 5” baffles. Utilities were
supplied by a B. Braun Biostat MD controlled by a DCU-1 or DCU-3. Data acquisition
utilized a Dell Optiplex Gs+ 5166M personal computer (PC) equipped with B. Braun

MFCS/Win software (v2.0). Temperature, pH, and dissolved oxygen (D.O) were

164

controlled with PID control loops. Temperature was maintained at 36 °C, and pH was
maintained at 7.0 by addition of concentrated NH4OH or 2 N H2804. Dissolved oxygen
was measured using a Mettler-Toledo 12 mm sterilizable O2 sensor fitted with an Ingold
A-type O2 permeable membrane. D.O. was maintained at 20% air saturation. Antifoam
(Sigma 204) was added as needed.

Inoculants were prepared by introduction of a single colony into 5 mL of M9
medium. The culture was grown at 37 °C with agitation at 250 rpm until they were
turbid (~l8-30 h) and subsequently transferred to 100 mL of M9 medium. Cultures were
grown at 37 °C for an additional 12 h. The inoculant (OD600 = 1.0-2.0) was then

transferred into the fermentor vessel and the batch fermentation was initiated (t = 0 h).

Glucose-rich fermentor conditions

The initial glucose concentration in the fermentation medium was 30 g/L. Three
staged methods were used to maintain D.O. levels at 20% air saturation during the course
of the fermentations. With the airflow at an initial setting of 0.06 LIL/min, D.O.
concentration was maintained by increasing the impeller speed from its initial set point of
50 rpm to a preset maximum of 750 rpm. With the impeller rate constant at 750 rpm, the
mass flow controller then maintained D.O. levels by increasing the airflow rate from 0.06
L/L/min to a preset maximum of 1.0 L/L/min. After the preset maxima of 750 rpm and
1.0 L/L/min were reached, the third stage of the fermentation was initiated in which
glucose (65% w/v) was added to the vessel at a rate sufficient to maintain a glucose
concentration in the range of 5 to 30 g/L for the remainder of the run. Airflow was

maintained at 1.0 L/L/min, and the impeller was allowed to vary in order to maintain the

165

D.O. concentration at 20% air saturation. The impeller speed typically varied from 750
rpm to 1400 rpm during the remainder of the run. A solution of IPTG (100 mM; 0, 0.25,
0.50, 0.75, 1.0, and 2.0 mL) was added at timed intervals after initiation of the run to
achieve reported IPTG concentrations respectively of 0, 6.0, 12, 18, 24, and 48 mg/L in

the fermentation medium.

Glucose-limited fermentor conditions

The initial glucose concentration in the fermentation medium was 19-24 g/L,
depending on the strain being examined. Three staged methods were used to maintain
D.O. levels at 20% air saturation, with the first two stages identical to those described for
the glucose-rich conditions. After the preset maxima of 750 rpm and 1.0 L/L/min of
airflow were reached, the third stage of the fermentation was initiated in which the DO.
concentration was maintained at 20% air saturation for the remainder of the run by
oxygen sensor-controlled glucose feeding. At the beginning of this stage, the DO.
concentration initially fell below 20% air saturation due to residual initial glucose in the
medium. This lasted for up to 30 min before glucose (65% w/v) feeding commenced.
The glucose feed PID control parameters were set to 0.0 8 (off) for the derivative control
(13D) and 999.9 3 (minimum control action) for the integral control (171). XP was set to
950% to achieve a KC of 0.1. A solution of IPTG (100 mM; 0, 0.25, 0.50, 0.75, 1.0, and
2.0 mL) was added at timed intervals after initiation of the run to achieve reported IPTG
concentrations respectively of 0, 6.0, 12, 18, 24, and 48 mg/L in the fermentation

medium.

166

Analysis of fermentation broths

Samples (5 mL) of fermentation broth were taken at the indicated timed intervals.
Cell densities were determined by dilution of fermentation broth with water (1:100)
followed by measurement of absorption at 600 nm (ODbm). Dry cell weight (g/L) was
calculated using a conversion coefficient of 0.43 g/L/ODOOO. The remaining fermentation
broth was centrifuged to obtain cell-free broth.

Glucose concentrations in cell-free broth were measured using the Glucose
Diagnostic Kit purchased from Sigma. Solute concentrations in the cell-free broth were
quantified by 1H NMR. Solutions were concentrated to dryness under reduced pressure,
concentrated to dryness one additional time from D20, and then redissolved in D20
containing a known concentration of the sodium salt of 3-(trimethylsilyl)propionic-
2,2,3,3-d4 acid (TSP). 1H NMR spectra were recorded and concentrations were
determined by comparison of integrals corresponding to each compound with the integral
corresponding to TSP (6 = 0.00 ppm). A standard concentration curve was determined
for each metabolite using solutions of authentic, purified metabolites. The following
resonances were used to quantify each compound: shikimic acid (6 4.57, d, 1 H); 3-
dehydroshikimic acid (6 4.28, d, 1 H); 3-dehydroquinic acid (6 4.38, d, l H); 3-deoxy-D-
arabina-heptulosonic acid (6 1.81, t, l H); and gallic acid (6 7.02, s, 2 H). The following
response factor was used for each molecule: shikimic acid, 0.70; 3-dehydroshikimic acid,
0.95; 3-dehydroquinic acid, 0.89; 3-deoxy-D-arabina-heptulosonic acid, 1.22; gallic acid,

1.36.

167

Genetic manipulations

General

Recombinant DNA manipulations generally followed methods described by
Sambrook.'5 Restriction enzymes were purchased from Invitrogen or New England
Biolabs. T4 DNA ligase was obtained from Invitrogen. Fast-LinkTM DNA Ligation Kit
was obtained from Epicentre. Zymoclean Gel DNA Recovery Kit and DNA Clean &
Concentrator Kit was obtained from Zymo Research Company. Maxi and Midi Plasmid
Purification Kits were obtained from Qiagen. Calf intestinal alkaline phosphatase was
obtained from Boehringer Mannheim. Agarose (electrophoresis grade) was obtained
from Invitrogen. Phenol was prepared by addition of 0.1 % (w/v) 8-hydroxyquinoline to
distilled, liquefied phenol. Extraction with an equal volume of 1 M Tris-HCl (pH 8.0)
two times was followed by extraction with 0.1 M Tris-HCl (pH 8.0) until the pH of the
aqueous layer was greater than 7.6. Phenol was stored at 4 °C under an equal volume of
0.1 M Tris-HCl (pH 8.0). SEVAG was a mixture of chloroform and isoamyl alcohol
(24:1 v/v). TE buffer contained 10 mM Tris-HCl (pH 8.0) and 1 mM NazEDTA (pH
8.0). Endostop solution (10X concentration) contained 50% glycerol (v/v), 0.1 M
NazEDTA, pH 7.5, 1% sodium dodecyl sulfate (SDS) (w/v), 0.1% bromophenol blue
(w/v), and 0.1% xylene cyanole FF (w/v) and was stored at 4 °C. Prior to use, 0.12 mL of
DNase-free RNase was added to 1 mL of 10X Endostop solution. DNase-free RNase (10
mg mL'l) was prepared by dissolving RNase in 10 mM Tris-Cl (pH 7.5) and 15 mM
NaCl. DNase activity was inactivated by heating the solution at 100 °C for 15 min.

Aliquots were stored at -20 °C. PCR amplifications were carried out as described by

Sambrook.ls Each reaction (0.1 mL) contained 10 mM KCl, 20 mM Tris-Cl (pH 8.8), 10

168

mM (NH4)ZSO4, 2 mM MgSO4, 0.1% Triton X-100, dATP (0.2 mM), dCTP (0.2 mM),
dGTP (0.2 mM), dTTP (0.2 mM), template DNA, 0.5 11M of each primer, and 2 units of
Vent. Taq or Pfu polymerase also have been used for PCR reaction with the reaction

buffers provided. Initial template concentrations varied from 0.02 rig to 1.0 ug.

Large scale purification of plasmid DNA

Plasmid DNA was purified on a large scale using a modified alkaline lysis
method described by Sambrook.” In a 2 L Erlenmeyer flask, LB (500 mL) containing
the appropriate antibiotics was inoculated from a single colony, and the culture was
incubated in a gyratory shaker (250 rpm) for 14 h at 37 °C. Cells were harvested by
centrifugation (4 000g, 5 min, 4 °C) and then resuspended in 10 mL of cold GETL

solution [50 mM glucose, 20 mM Tris-HCl (pH 8.0), 10 mM NazEDTA (pH 8.0)] into

which lysozyme (5 mg mL'l) had been added immediately before use. The suspension
was stored at room temperature for 5 min. Addition of 20 mL of 1% sodium dodecyl
sulfate (w/v) in 0.2 N NaOH was followed by gentle mixing and storage on ice for 15
min. Fifteen milliliters of an ice cold solution containing 3 M KOAC (prepared by
combining 60 mL of 5 M potassium acetate, 11.5 mL of glacial acetic acid, and 28.5 mL
of H20) was added. Vigorous shaking resulted in formation of a white precipitate. After
the suspension was stored on ice for 10 min, the cellular debris was removed by
centrifugation (48 000g, 20 min, 4 °C). The supernatant was transferred to two clean
centrifuge bottles and isopropanol (0.6 volumes) was added to precipitate the DNA.

After the samples were left at room temperature for 15 min, the DNA was recovered by

169

centrifugation (20 000g, 20 min, 4 °C). The DNA pellet was then rinsed with 70%
ethanol and dried.

Further purification of the DNA sample involved precipitation with polyethylene
glycol (PEG). The DNA was dissolved in TE (3 mL) and transferred to a Corex tube.
Cold 5 M LiCl (3 mL) was added and the solution was gently mixed. The sample was
then centrifuged (12 000g, 10 min, 4 °C) to remove high molecular weight RNA. The
clear supernatant was transferred to a clean Corex tube and isopropanol (6 mL) was
added followed by gentle mixing. The precipitated DNA was collected by centrifugation
(12 000g, 10 min, 4 °C). The DNA was then rinsed with 70% ethanol and dried. After
redissolving the DNA in 0.5 mL of TE containing 20 ug/mL of RNase, the solution was
transferred to a 1.5 mL microcentrifuge tube and stored at room temperature for 30 min.
To this sample was added 500 11L of 1.6 M NaCl containing 13% PEG-8000 (w/v)
(Sigma). The solution was mixed and centrifuged (microcentrifuge, 10 min, 4 °C) to
recover the precipitated DNA. The supernatant was removed and the DNA was then
redissolved in 400 11L of TE. The sample was extracted sequentially with phenol (400
11L), phenol and SEVAG (400 1.1L each), and finally SEVAG (400 ML). Ammonium
acetate (10 M, 100 ML) was added to the aqueous DNA solution. After thorough mixing,
95% ethanol (1 mL) was added to precipitate the DNA. The sample was left at room
temperature for 5 min and then centrifuged (microcentrifuge, 5 min, 4 °C). The DNA
was rinsed with 70% ethanol, dried, and then redissolved in 200-500 11L of TE.

Alternatively, DNA was purified using a Qiagen Maxi Kit or Midi Kit as
described by the manufacturer. The purity of DNA isolated by this method was adequate

for DNA sequencing.

170

Small scale purification of plasmid DNA

An overnight culture (5 mL) of the plasmid-containing strain was grown in LB
containing the appropriate antibiotics.” Cells from 3 mL of the culture were collected in
a 1.5 mL microcentrifuge tube by centrifugation. The resulting cell pellet was liquefied
by vortexing (30 sec) and then resuspended in 0.1 mL of cold GETL solution into which
lysozyme (5 mg mL‘l) had been added immediately before use. The solution was stored
on ice for 10 min. Addition of 0.2 mL of 1% sodium dodecyl sulfate (w/v) in 0.2 N
NaOH was followed by gentle mixing and storage on ice for 5-10 min. To the sample
was added 0.15 mL of cold KOAc solution. The solution was shaken vigorously and
stored on ice for 5 min before centrifugation (15 min, 4 °C). The supernatant was
transferred to another microcentrifuge tube and extracted with equal volumes of phenol
and SEVAG (0.2 mL). The aqueous phase (approximately 0.5 mL) was transferred to a
fresh microfuge tube, and DNA was precipitated by the addition of 95% ethanol (1 mL).
The sample was left at room temperature for 5 min before centrifugation (15 min, room
temperature) to collect the DNA. The DNA pellet was rinsed with 70% ethanol, dried,
and redissolved. in 50 -100 ML TE. DNA isolated from this method was used for
restriction enzyme analysis, and the concentration was not determined by spectroscopic

methods.

Determination of DNA concentration
The concentration of DNA in the sample was determined as follows. An aliquot

(10 ML) of the DNA was diluted to 1 mL in TE and the absorbance at 260 nm was

171

measured relative to the absorbance of TE. The DNA concentration was calculated based

on the fact that the absorbance at 260 nm of a 50 ug mL‘1 of plasmid DNA is 1.0.

DNA precipitation

DNA was precipitated by addition of 0.1 volume of 3 M NaOAc (pH 5.2)
followed by thorough mixing and addition of 3 volumes of 95% ethanol. Samples were
stored for at least 2 h at -78 °C. Precipitated DNA was recovered by centrifugation (15
min, 4 °C). To the DNA pellet was added 70% ethanol (100 ML), and the sample was

centrifuged again (15 min, 4 °C). DNA was dried and redissolved in TE.

Restriction enzyme digestion of DNA

Restriction enzyme digests were performed using restriction enzyme buffers
supplied by Invitrogen or New England Biolabs. A typical digest contained
approximately 0.8 ug of DNA in 8 ML TE, 2 11L of restriction enzyme buffer (10X
concentration), 1 ML of restriction enzyme, and TE to a final volume of 20 uL. Reactions
were incubated at 37 °C for 1 h. Digests were terminated by addition of 2.2 uL of
Endostop solution (10X concentration) and subsequently analyzed by agarose gel
electrophoresis. When DNA was required for subsequent cloning, restriction digests
were terminated by addition of 1 11L of 0.5 M N agEDTA (pH 8.0) followed by extraction

of the DNA with equal volumes of phenol and SEVAG and precipitation of the DNA.

172

 

Agarose gel electrophoresis

Agarose gels were run in TAE buffer containing 40 mM Tris-acetate and 2 mM
EDTA (pH 8.0). Gels typically contained 0.7% agarose (w/v) in TAE buffer. Higher
concentrations of agarose (1%-2%) were used to resolve DNA fragments smaller than 1
kb. Lower concentrations of agarose (0.35%) were used to resolve DNA fragments
larger than 10 kb. Ethidium bromide (0.5 ug mL'l) was added to the agarose to allow
visualization of DNA fragments over a UV lamp. The size of the DNA fragments were
determined by using two sets of DNA standards: A DNA digested with Hindlll (23.1-kb,
9.4-kb, 6.6-kb, 4.4-kb, 2.3-kb, 2.0-kb, and 0.6-kb) and A DNA digested with EcoRI and
Hindlll (21.2-kb, 5.1-kb, 5.0-kb, 4.3-kb, 3.5-kb, 2.0-kb, 1.9-kb. 1.6-kb, 1.4-kb, 0.9-kb,
0.8-kb, and 0.6-kb). Also 100 bp DNA Ladder (Invitrogen) was used to determine the
size of small DNA fragements. The ladder consists of 15 blunt-ended fragments ranging

in length from 100 to 1500 bp, at 100 bp increments, and an additional fragment at 2,072

bp.

Isolation of DNA from agarose

The band of agarose containing DNA of interest was excised from the gel while
visualized with high wavelength UV and chopped thoroughly with a razor in a plastic
weighing tray. The agarose was then transferred to a spin column consisting of a 500 11L
microfuge tube packed tightly with glass wool and with an 18 gauge hole in its bottom.
The spin column was then centrifuged for 5 min using a microcentrifuge to separate the

DNA solution from the agarose. The DNA-containing aqueous phase collected after

173

centrifugation were mixed with 3 M NaOAc and 95% ethanol. The DNA was
precipitated as described previously and redissolved in TE.

Alternatively, the band of agarose containing DNA of interest was excised from
the gel while visualized with high wavelength UV. Zymoclean Gel DNA Recovery Kit
was used to isolate DNA from the agarose gel according to the protocol provided by the

Zymo Research Company.

Treatment of vector DNA with calf intestinal alkaline phosphatase

Plasmid vectors digested with a single restriction enzyme were dephosphorylated
to prevent self-ligation. Vector DNA after digestion was immediately combined in a total
volume of 60 11L. To this sample was added 7 11L of dephosphorylation buffer (10X
concentration, provided by enzyme supplier) and 3 ML of calf intestinal alkaline
phosphatase (3 units). The reaction was incubated at 37 °C for 1 h. The phosphatase was
inactivated by addition of 1 ML of 0.5 M N azEDT A (pH 8.0) followed by heat treatment
(65 °C, 20 min). The sample was extracted with phenol and SEVAG (100 11L each) to
remove protein, and the DNA was precipitated as previously described and redissolved in

TE.

Treatment of DNA with Klenow fragment

DNA with recessed 3' termini was modified to blunt-ended fragment by treatment
with the Klenow fragment of E. coli DNA polymerase 1. After the DNA (0.8-2 rig)
restriction digestion was completed in a 20 11L reaction, a solution (1 uL) containing each

of the desired dNTPs was added to provide a final concentration of 1 mM for each dNTP.

174

Addition of 1-2 units of the Klenow fragment to the reaction was followed by incubation
of the mixture at room temperature for 20—30 min. Since the Klenow fragment works
well in the common buffers used for restriction digestion of DNA, there was no need to
purify the DNA after restriction digestion and prior to filling recessed 3’ termini. Klenow
reactions were quenched by extraction with equal volumes of phenol and SEVAG. DNA

was recovered by DNA precipitation.

Ligation of DNA

DNA ligations were designed so that the molar ratio of insert to vector was 3 to 1.
A typical ligation reaction contained 0.03 to 0.1 ug of vector and 0.05 to 0.2 ug of insert
in a total volume of 7 11L. To this sample was added 2 11L of ligation buffer (5X
concentration, provided by the enzyme supplier) and 1 1.1L of T4 DNA ligase (2 units).
The reaction was incubated at 16 °C for at least 4 h and then used to transform competent
cells.

Alternatively, Fast-Link DNA Ligation Kit (Epicentre) was used for ligation of
insert DNA with cohesive or blunt ends into vectors with compatible cohevsive ends

according to the protocol provided by the manufacturer.

Preparation and transformation of competent cells

Competent cells were prepared using a procedure modified from Sambrook.l5 An
aliquot (1 mL) from an overnight culture (5 mL) was used to inoculate 100 mL of LB
(500 mL Erlenmeyer flask) containing the appropriate antibiotics. The cells were

cultured in a gyratory shaker (37 °C, 250 rpm) until they reached the mid-log phase of

175

growth (judged from the absorbance at 600 nm reaching 0.4-0.6). The culture was
poured into a large centrifuge bottle that had been previously sterilized with bleach and
rinsed with sterile water. The cells were collected by centrifugation (4 000g, 5 min, 4 °C)
and the culture medium was discarded. All manipulations were carried out on ice during
the remaining portion of the procedure. The cell pellet was washed with 100 mL of cold
0.9% NaCl (w/v) and then resuspended in 50 mL of cold 100 mM CaClz. The suspension
was stored on ice for a minimum of 30 min and then centrifuged (4 000g, 5 min, 4 °C).
The cell pellet was resuspended in 4 mL of cold 100 mM CaClz containing 15% glycerol
(v/v). Aliquots (0.25 mL) were dispensed into 1.5 mL microcentrifuge tubes and
immediately frozen in liquid nitrogen. Competent cells were stored at -78 °C with no
significant decrease in transformation efficiency over a period of six months.

Frozen competent cells were thawed on ice for 5 min before transformation. A
small aliquot (1 to 10 11L) of plasmid DNA or a ligation reaction was added to the thawed
competent cells (0.1 mL). The solution was gently mixed and. stored on ice for 30 min.
The cells were then heat shocked at 42 °C for 2 min and placed on ice briefly (1 min).
LB (0.5 mL, no antibiotics) was added to the cells, and the sample was incubated at 37 °C
(no agitation) for 1 h. Cells were collected in a microcentrifuge (30 s). If the
transformation was to be plated onto LB plates, the cells were resuspended in a small
volume of LB medium (0.1 mL), and then spread onto plates containing the appropriate
antibiotics. If the transformation was to be plated onto minimal medium plates, the cells
was washed once with the same minimal medium. After resuspension in fresh minimal
medium (0.1 mL), the cells was spread onto the plates. A sample of competent cells with

no DNA added was also carried through the transformation procedure as a control. These

176

cells were used to check the viability of the competent cells and to verify the absence of

growth on selective medium.

Purification of genomic DNA

Genomic DNA was purified using a modified method described by Silhavy.l6 A
single colony of the strain was inoculated into 100 mL of TB medium (500 mL
Erlenmeyer flask). The cells were cultured in a gyratory shaker (37 °C, 250 rpm) for
12 h. Centrifugation (4 000g, 5 min, 4 °C) of the culture was followed by resuspension of
the cell pellet in 5 mL of buffer [50 mM Tris-HCl (pH 8.0), 50 mM EDTA (pH 8.0)] and
storage at -20 °C for 20 min to freeze the suspension. To the frozen cells was added 0.5
mL of 0.25 M Tris-HCl (pH 8.0) that contained 5 mg of lysozyme. The suspension was
thawed at room temperature in a water bath with gentle mixing and then stored on ice for
45 min. The sample was then transferred to a Corex tube. After addition of 1 mL of
STEP solution [25 mM Tris-HCl (pH 7.4), 200 mM EDTA (pH 8.0), 0.5% SDS (w/v),
and proteinase K (1 mg mL‘l, Sigma), prepared just before use], the mixture was
incubated at 50 °C for at least 1 h with gentle, periodic mixing. The solution was then
divided into two Corex tubes, and the contents of each tube were extracted with phenol (4
mL). The organic and aqueous layers were separated by centrifugation (1 000g, 15 min,
room temperature), and the aqueous layer was transferred to a fresh Corex tube. All
transfers of the aqueous layer were carried out using wide bore pipette tips to minimize
shearing of the genomic DNA. The contents of each tube were extracted again with a
mixture of phenol (3 mL) and SEVAG (3 mL). Extractions with phenol/SEVAG were

repeated (approximately 6 times) until the aqueous layer was clear.

177

Genomic DNA was precipitated by addition of 0.1 volume of 3 M NaOAc (pH
5.2), gentle mixing, and addition of 2 volumes of 95% ethanol. Threads of DNA were
spooled onto a sealed Pasteur pipette and transferred to a Corex tube that contained 5 mL
of 50 mM Tris-HCl (pH 7.5), 1 mM EDTA (pH 8.0), and 1 mg of RNase. The mixture
was stored at 4 °C overnight to allow the DNA to dissolve completely. The solution was
then extracted with SEVAG (5 mL) and centrifuged (l 000 g, 15 min, room temperature).
The aqueous layer was transferred to a fresh Corex tube and the genomic DNA was
precipitated as described above. The threads of DNA were spooled onto a Pasteur pipette
and redissolved in 2 mL of 50 mM Tris-HCl (pH 7.5) and 1 mM EDTA (pH 8.0).
Genomic DNA was stored at 4 °C.

Alternatively, genomic DNA was purified using a method described by Pitcher.l7
Broth cultures (20 mL) were harvested at the end of the exponential growth phase by
centrifugation (1 000g, 15 min, room temperature). A small cell pellet was obtained.
The cells of Gram-positive species were resuspended in 100 11L of fresh lysozyme (50
mg/mL) in TE buffer and incubated at 37°C for 30 min. The Gram-negative species were
resuspended in 100 111 of TE buffer without enzyme treatment and incubated at 37°C for
30 min. Cells were lysed with 0.5 mL 5 M guanidium thiocyanate (Sigma), 100 mM
EDTA and 0.5% v/v sarkosyl (GES reagent), which was prepared as follows. Guanidium
thiocyanate (60 g), 0.5 M EDTA at pH 8 (20 mL) and deionized water (20 mL) were
heated at 65°C with mixing until dissolved. After cooling, 5 mL of 10% v/v sarkosyl
were added, the solution was made up to 100 mL with deionized water, filtered through a

022-er membrane and stored at room temperature.

178

Cell suspensions were vortexed briefly and checked for lysis (clear solution) after
5-10 min. The lysates were cooled on ice and 0.25 mL cold 7.5 M ammonium acetate
was added with mixing on ice for 10 min. To this sample, 0.5 mL SEVAG was added,
and the solution was mixed thoroughly. After centrifugation in a 1.5 mL Eppendorf tube
(25 000g, 10 min, room temperature), supernatant fluids were transferred to Eppendorf
tubes and 0.54 volumes of cold 2-propanol was added. The tubes were inverted for l min
to mix the solutions and the fibrous DNA precipitate was deposited by centrifugation (6
500g, 20 s, room temperature). Pellets of DNA were washed five times in 70% ethanol

and dried at room temperature for 20 min. Genomic DNA was redissolved in 100 1.1L TE.

Pl-mediated transduction

Transduction with P1 phage was carried out using a method modified from
Miller.18 P1 phage lysate was prepared by propagation of phage in the donor strain using
the following procedure. Serial dilutions of P1 phage stock (0.1 mL, 10'1 to 10's) in LB
were prepared in sterile test tubes (13 x 100 mm). An aliquot (0.1 mL, approximately 5 x
108 cells) of an overnight culture of the donor strain was added to each tube. Sterile,
molten soft agar (45 °C) was added to each tube. The contents of each tube were mixed
and poured immediately onto a pre-warmed (37 0C) L plate, swirling gently to achieve
uniform coverage of the plate. After the agar had solidified, the plates were incubated at
37 °C until confluent lysis had occurred (approximately 8 h). Because the multiplicity of
infection is critical to phage generation, confluent lysis occurred on only one or two of
the plates. L-Broth (4 mL) was added to these plates, which were then stored overnight

at 4 °C to allow the phage particles to diffuse into the broth. The L-broth was collected

179

from the plate and vortexed with several milliliters of CHCl3 to make certain that all of
the cells had lysed. The solution was centrifuged (2 000g, 5 min, room temperature) to
separate the layers. Aqueous phage lysate was stored in 1.5 mL microfuge tubes over
several drops of CHCl3 at 4 0C.

Infection of the recipient strain with phage lysate proceeded as follows.
Overnight culture (2 mL) of the recipient strain was centrifuged (microfuge, 30 s, 4 oC)
and the growth medium discarded. The cells were resuspended in 1 mL of 5 mM CaCl2
and 100 mM MgSO4 and shaken (200 rpm) at 37 °C for 15 min to promote aeration of the
cells. In the meantime, 0.1 mL serial dilutions (100 to 10”) of phage lysate in LB were
prepared in sterile microfuge tubes. An aliquot (0.1 mL) of aerated recipient cells was
added to each of the phage dilutions, the samples were gently mixed and then incubated
at 37 °C for 20 min without shaking. Sodium citrate (1 M, 0.2 mL) was added to each
sample, and the cells were harvested (microfuge, 30s, room temperature) and
resuspended in 0.2 mL of LB containing 100 mM sodium citrate. After incubation at 30
0C for 30 min, cells were again harvested (microfuge, 30 s, room temperature),

resuspended in 0.1 mL of growth medium, and plated out onto appropriate agar plates.

Enzyme assays

After collected. and resuspended in proper resuspension buffer, the cells were
disrupted by two passages through a French pressure cell (SLM Aminco) at 16000 psi.
Cellular debris was removed from the lysate by centrifugation (48 000g, 20 min, 4 °C).

19

Protein was quantified using the Bradford dye-binding procedure. A standard curve

180

was prepared using bovine serum albumin. The protein assay solution was purchased

from Bio-Rad.

DAHP synthase assay

DAHP synthase was assayed according to the procedure described by Schoner.20
Harvested cells were resuspended in 50 mM potassium phosphate (pH 6.5) that contained
10 mM PEP and 0.05 mM CoClz. The cells were disrupted using a French press as
described above. Cellular lysate was diluted in a solution of potassium phosphate (50
mM), PEP (0.5 mM), and 1,3-propanediol (250 mM), pH 7.0. A dilute solution of E4P
was first concentrated to 12 mM by rotary evaporation and neutralized with 5 N KOH.
Two different solutions were prepared and incubated separately at 37 °C for 5 min. The
first solution (1 mL) contained E4P (6 mM), PEP (12 mM), ovalbumin (1 mg/mL), and
potassium phosphate (25 mM), pH 7.0. The second solution (0.5 mL) consisted of the
diluted lysate. After the two solutions were mixed (time = 0), aliquots (0.15 mL) were
removed at timed intervals and quenched with 0.1 mL of 10% trichloroacetic acid (w/v).
Precipitated protein was removed by centrifugation, and the DAHP in each sample was
quantified using thiobarbituric acid assay2| as described below.

An aliquot (0.1 mL) of DAHP containing sample was reacted with 0.1 mL of 0.2
M NaIO4 in 8.2 M H3PO4 at 37 °C for 5 min. The reaction was quenched by addition of
0.8 M NaA502 in 0.5 M NazSO4 and 0.1 M H2SO4 (0.5 mL) and vortexed until a dark
brown color disappeared. Upon addition of 3 mL of 0.04 M thiobarbituric acid in 0.5 M
NagSO4 (pH 7), the sample was heated at 100 °C for 15 min. Samples were cooled (2

min), and the pink chromophore was then extracted into distilled cyclohexanone (4 mL).

181

The aqueous and organic layers were separated by centrifugation (2 000g, 15 min, room

temperature). The absorbance of the organic layer was recorded at 549 nm (e = 68000 L
mol"1 cm'l). One unit of DAHP synthase activity was defined as the formation of 1 nmol

of DAHP per min at 37 °C.

Phosphoenolpyruvate synthase assay

PEP synthase activity was assayed at 30 °C according to the procedure described
by Cooper.22 The reaction (1 mL) contained 100 mM Tris-HCl (pH 8.0), 10 mM MgC12,
10 mM ATP, 1.25 mM sodium pyruvate and was initiated by addition of cell-free lysate.
Aliquots (100 111) of reaction mixture were removed at l min intervals and immediately
added to a microcentrifuge tube containing 0.33 mL of a 0.1% aqueous solution of 2, 4-
dinitrophenylhydrazine and 0.9 mL H20. The resulting mixture was incubated at 30 °C
for 10 min. Following addition of 1.67 mL of 10% (w/v) NaOH, the mixture was further
incubated at 30 °C for 10 min. The disappearance of pyruvate was quantified by
measuring the absorbance at 445 nm. A molar extinction coefficient of 18,000 L mol'l
cm’I was used to quantify pyruvate. One unit of PEP synthase activity was defined as the
amount of enzyme that catalyzed the consumption of 1 nmol of pyruvate per min at 30

°C.

Glucokinase assay
Glucokinase was assayed by coupling with glucose-6-phosphate dehydrogenase
according to the method of Fraenkel.23 The assay solution contained 50 mM Tris-HCl

(pH 7.6), 0.5 mM glucose, 2 mM MgC12, 2 mM ATP, 1 ug of glucose 6-phosphate

182

dehydrogenase (3 units), cell-free lysate and was initialized by addition of 0.4 mM
NADP. The formation rate of glucose 6-phosphate was equal to the formation rate of
NADPH which was measured by the increase in absorbance at 340 nm (8 = 6220 L mol‘l
cm"). One unit of glucokinase activity was defined as the formation of 1 nmol of

glucose 6-phosphate per min at 25 °C.

Transketolase assay

Transketolase was assayed using a coupled enzyme procedure described by
Paoletti.24 The assay solution (1 mL) contained triethanolamine buffer (150 mM, pH
7.6), MgCl2 (5 mM), thiamine pyrophosphate (0.1 mM), NADP (0.4 mM), [3-
hydroxypyruvate (0.4 mM), D-erythrose 4-phosphate (0.1 mM), glucose 6-phosphate
dehydrogenase (3 units), and phosphoglucose isomerase (10 units). The solution was
incubated at room temperature and the absorbance at 340 nm was monitored for several
minutes. After all the unreacted D-glucose-6-phosphate from the D-erythrose 4-
phosphate synthesis had reacted, an aliquot of transketolase solution was added and the
reaction monitored at 340 nm for ‘10-20 min. One unit of transketolase activity was

defined as the formation of 1 nmol of NADPH (8 = 6220 L mol'I cm") per min.

Chapter 2
E. coli JYl (KL3 AptsHIcrr)
E. coli JYl (KL3 AptsHIcrr) was made from E. coli KL3 by P1 transduction
using E. coli UE79 as the donor strain. Colonies were plated on M9 maltose plates

containing fosfomycin with aromatic and serine supplementation. The resulting colonies

183

 

were screened for loss of PTS activity by checking the growth on the following plates: no
growth on M9 glucose plates with aromatic, serine supplementation; no growth on M9
xylose plates with aromatic and serine supplementation; growth on M9 xylose plates with
aromatic, serine and CAMP (1 mM final concentration on the plate) supplementation; no

growth on M9 mannitol plates with aromatic and serine supplementation.

E. coli JY1.2 (KL3 AptsHIcrr galP*)

Plasmid pRC55B containing serA gene was transformed into E. coli JYl. The
resulting E. coli JY1/pRC55B was used to inoculate 1 L of fermentation medium
containing 20 g glucose.l4 Doubling times for growth under fermentor-controlled
conditions were determined” by dilution of fermentation broth (1:100) followed by
measurement of absorption at 600 nm (OD600). After cultivation under fermentor-
controlled conditions for 3.5 days, the culture’s doubling time decreased from 12 h to 3.7
h. When the culture reached OD600 = 20, fermentor broth (900 mL) was discarded and
replaced with fresh fermentation medium containing 20 g glucose. Repetition of this
dilution process two more times led to a culture with a doubling time of 2.7 h. Serial
dilutions of the final culture were spread onto M9 glucose plates with aromatics
supplementation and incubated at 37 °C. Single colonies were subsequently obtained
after growth on M9 glucose plates. These colonies were screened for growth on M9
maltose plates containing fosfomycin with aromatic supplementation and an absence of
growth on M9 mannitol plates with aromatic supplementation. Colonies possessing the
proper phenotype were also screened on M9 glucose plates with aromatic

supplementation for the most rapid growth. One colony with most rapid growth was

184

inoculated into 5 mL LB culture and grown at 37°C for 12 h. Cultures were diluted
(1210000) in LB, and two more cycles of growth at 37 °C for 12 h were carried out to
promote plasmid loss from the cells. Serial dilutions of the final culture were spread onto
LB plates and the resulting colonies were screened for resistance to Cm. A CmS colony
designated as E. coli JY1.2 was isolated that retained the phenotype for the absence of
PTS-mediated glucose transport as characterized by growth on M9 maltose plates
containing fosfomycin with aromatic and serine supplementation and an absence of

growth on M9 mannitol plates with aromatic and serine supplementation.

E. coli JY1.2 galP::Tn10

E. coli JY1.2 was made galP::Tn10 by P1 transduction from E. coli JM2053.
Colonies were plated onto LB plates containing Tc to select for colonies which had been
transduced by galP::Tn10. The resulting E. coli JY1.2 galP::Tn10 didn’t grow on M9

glucose plates with aromatic and serine supplementation.

E. coli JY1.3 (KL3 AptsHIcrr gall”)

E. coli JY1.2/pRC55B was used to inoculate a 5 mL liquid M9 glucose medium
with aromatic supplementation and grown at 37°C until the culture turned turbid.
Cultures were diluted (1210000) in M9, and four more cycles of growth at 37 °C were
carried out to obtain fast growing mutants of E. coli JY1.2/pRC55B. Serial dilutions of
the final culture were spread onto M9 glucose plates with aromatic supplementation and
incubated at 37 °C for 24 h. The biggest colony on the M9 glucose plate was inoculated

into 5 mL LB culture and grown at 37°C for 12 h. Cultures were diluted (1:10000) in LB,

185

and two more cycles of growth at 37 °C for 12 h were carried out to promote plasmid loss
from the cells. Serial dilutions of the final culture were spread onto LB plates and the
resulting colonies were screened for resistance to Cm. A CmS colony designated as E.
coli JY1.3 was isolated that retained the phenotype for the absence of PTS-mediated
glucose transport as characterized by growth on M9 maltose plates containing fosfomycin
with aromatic and serine supplementation and an absence of growth on M9 mannitol

plates with aromatic and serine supplementation.

E. coli JY1.3 galP::Tn10

E. coli JY1.3 was made galP::Tn10 by P1 transduction from E. coli JM2053.
Colonies were plated onto LB plates containing Tc to select for colonies which had been
transduced by galP::Tn10. The resulting E. coli JY1.3 galP::Tn10 didn’t grow on M9

glucose plates with aromatic and serine supplementation.

Plasmid pJY1.131

Digestion of pKL4.66B2 with BamHI and Sall afforded a 1.3-kb fragment of
DNA that encoded araFFBR. Localization of the araFFBR fragment in pJF118EH which
had been previously digested with BamHI and Sall resulted in pJY1.131. Transcription
of the araFFBR gene in pJY1.l31 proceeds in the opposite orientation relative to the

vector-encodedP sequence.

(ac

Plasmid pJY1.143A

186

 

The ppsA gene was originally amplified from RB791 genomic DNA using the
following primers: 5’-AACTGCAGCGATCCAGTTTCATCTCTTGT and 5’-
AACTGCAGTTAT TTCTTCAGTTCAGCCAG. Pstl restriction sequences (underlined
nucleotides) were included to facilitate cloning. Localization of the resulting 3.0-kb ppsA
fragment into the Pstl site of pSU19 afforded pKLl.87A. The ppsA ORF was
subsequently amplified from pKLl.87A using the following primers: 5’-
GGAATTCTTATTTCTTCAGTTCAGCCAG and 5’-
GGAATTCATGTCCAACAATGGCTCGT. Insertion of the resulting ppsA ORF-
encoding fragment into the EcoRI site of pJYl.l31 afforded the 9.0-kb plasmid
pJY1.143A in which the ppsA ORF is transcribed from the vector-encoded P,“ sequence

that immediately precedes the gene.

Plasmid pJY 1.207A
A 0.2-kb sequence of DNA encoding the three operator binding sites associated

with the araF gene, designated P was amplified from pMF63A using the following

aroF,

primers: 5’-GCGGATCCGAATTCAAAGGGAGTGTA and 5’-
GCGGATCCCCTCAGCGAGG ATGACGT. Localization of the Pump fragment of DNA

into the BamHI site of pJY1.143A afforded pJ Y1 .207A.

Plasmid pJYl.211A

Ligation of a 1.9-kb DraI/EcaRV serA-encoding fragment of DNA obtained from

pD2625 into the Smal site of pJYl.207A resulted in isolation of pJYl.211A in which

187

transcription of serA proceeds in the opposite orientation relative to the vector-encoded

P sequence.

tac

Plasmid pJY 1.216A

Following digestion of pMF51A with BamHI, the resulting 2.2-kb tktA fragment
was treated with Klenow fragment and ligated to pJYl.211A which had previously been
incubated with Hindlll followed by Klenow fragment. The 13.3-kb plasmid pJYl.216A
was isolated in which transcription of the tktA gene proceeds in the orientation opposite

to transcription from the vector-encoded P

tac'

Plasmid pKL7.115A

Plasmid pKL7.115A is a pSUl8-derived plasmid that encodes araFFBR, PamF,
serA, nglf, tktA, and CmR. Construction of pKL7.115A began with pKD11.291A, a
5.6-kb plasmid that was described previously.2 A 2.2-kb nglf-encoding fragment was
liberated from pTC325 by digestion with BamHI, Hindlll and XbaI. Insertion of the
nglf fragment into Sall site of pKDl 1.291A yielded pKL6.239A. Following digestion
of pMF51A with BamHI, the resulting 2.2-kb tktA-encoding fragment was inserted into

Hindlll site of pKL6.239A to afford the 10.0-kb plasmid pKL7.115A.

Plasmid pJY 2.182

The 5.0-kb plasmid was created by ligation of a 2.2-kb tktA fragment obtained by

BamHI digestion of pMF51A into the BamHI site of p34H.

188

Plasmid pJY2.183A

Plasmid pJY2.183A is a pSU18-derived plasmid that encodes araFFBR, PamF,
serA, nglfglk, tktA, and CmR. Construction of pJY2.183A began with pKDl 1.291A. A
3.9-kb Pmcglfglk-encoding fragment was liberated from pTC325 by digestion with BamHI
and XbaI. Insertion of the nglfglk fragment into Sall site of pKDl 1.291A yielded
pKL7.85A. Following digestion of pJY2.182 with Sacl, the resulting 2.2-kb tktA-

encoding fragment was inserted into SacI site of pKL7.85A to afford the 11.7-kb plasmid

pJY2.183A.

Plasmid pJY3.29B

Plasmid pJY3.29B is a pSU18-derived plasmid that encodes araFFBR, P serA,

aroF,
glk, tktA, and CmR. Construction of pJY3.29B began with pKDl 1.291A. A 2.2-kb glk-
encoding fragment was liberated from pTC325 by digestion with Hindlll. Insertion of
the glk fragment into Sall site of pKD11.291A yielded pJY3.21B. Following digestion of

pJY2.182 with SacI, the resulting 2.2—kb tktA-encoding fragment was inserted into SacI

site of pJY3.21B to afford the 10.0-kb plasmid pJY3.29B.

Plasmid pRC55B

The 3.9-kb plasmid was created by ligation of a 1.6-kb serA fragment obtained by

PCR from E. coli RB791 genomic DNA into the Smal site of pSU18.

189

Chapter 3

Isolation of total RNA

Total RNA was isolated with hot phenol and used to prepare [a-33P]dCTP-labeled
cDNA as described by Tao.25 All the solutions are 0.1% diethyl pyrocarbonate (DEPC,
Aldrich) treated for 24 h at 37 oC and autoclaved for 25 min to avoid RNase
contamination. All the glass bottles used to store solutions were prepared by filling the
bottles with 0.1% DEPC solutions at 37 °C for 24 h followed by discarding 0.1% DEPC
solutions and autoclaving the empty bottles. Phenol (Ultrapure, Invitrogen) was
equilibrated with an equal volume of 50 mM NaOAc (2.07 g in 500 mL water and adjust
pH to 4.8 using HOAc) three times in a small bottle. Before use, 0.1% 8-
hydroxyquinoline was added to this acidic phenol and stored at 4 °C for up to one week.
Unless specified, all the enzymes were purchased from Invitrogen.

A sample of the E. coli culture from the fermentation (0.1 to 0.2 mL) was rapidly
transferred to a 17 x 100 mm sterilized test tube containing 4 mL of hot lysis buffer (1%
sodium dodecyl sulfate, 30 mM sodium acetate, 3 mM EDTA; pH 5.0) in a boiling water
bath close to the fermentor and held for 5 min with intermittent mixing. The resulting
lysates were extracted with 3 mL acidic phenol (prewarmed to 65 °C) with vigorous
intermittent mixing for 3 min. The use of acidic phenol was essential to minimize
genomic DNA contamination of the RNA sample. After cooling the sample on ice for 3
min, the sample was centrifuged (l2 000g, 10 min, 4 °C). The aqueous layer was
extracted a second time with 3 mL acidic phenol (prewarmed to 65 °C) followed by

extraction with equal volumes (1.5 mL) of phenol and chloroform at room temperature.

190

After the final extraction with 3 mL of chloroform at room temperature, the aqueous
layer was transferred to a fresh tube. Care must be taken to avoid transferring any of the
organic-phase along with the aqueous-phase. The aqueous layer was divided equally
between two fresh tubes and one-tenth volume of 3 M sodium acetate (pH 5.2, ice cold)
was added followed by 2.5 volumes of 100% ethanol (-20 °C). The sample was kept at
—20 °C overnight. Total RNA (two tubes) was pelleted by centrifugation (12 000g, 20
rrrin, 4 oC), washed carefully by addition of 70% ethanol (ice cold), and then pelleted
again by centrifugation (12 000g, 20 min, 4 °C). The RNA pellet (two tubes) was
washed again by addition of 70% ethanol (ice cold) followed by centrifugation (12 000g,
20 min, 4 °C). The RNA pellet (two tubes) at the bottom of the tube was marked outside
the tube with a pen and then dried at room temperature under air for 2-3 h. When the
RNA pellet (two tubes) appeared clear or translucent, RNA was dissolved in 90 11L of
sterile, RNase-free water by repeatedly pipeting. This time the RNA solutions from two
tubes were combined (total volume was about 180 11L) and stored on ice. Contaminating
DNA was removed from RNA by hydrolysis with DNaseI. RNA (total volume
approximately 180 11L), 20 11L DNaseI reaction buffer (10X, provided by the
manufacturer), DNaseI (8 uL, 200 U) and RNase inhibitor (2 ML, 80 U) was incubated at
room temperature for 25 min. After the reaction, Qiagen Rneasy Mini Kit was used to
purify the RNA according to the procedure provided by the manufacturer. The purified
RNA sample (approximately 160 11L) was quantified by measuring the absorbance at 260
nm (usually 10 11L of RNA in 1 mL water). The RNA concentration was calculated
based on the fact that the absorbance at 260 nm of a 40 ug mL’1 of RNA is 1.0. The

integrity of the RNA and the amount of genomic DNA contamination was checked by

191

agarose gel electrophoresis. A non-denaturing agarose gel was used so that any genomic
DNA contamination could be easily observed. The TAE buffer used to run the gel was
pretreated with 0.1% DEPC as described previously. The RNA sample (10 11L) was
loaded with 1 11L RN ase inhibitor (Invitrogen, 40 U) and stained with ethidium bromide.
The 233 and 16S ribosomal RNA bands should be clearly visible at about 2:1 ratio (23S:
16S) of staining intensity. If genomic DNA was present in the RNA sample, it would be
seen as high molecular weight-staining material. The final RNA sample was kept at —80
°C after addition of one-tenth volume 3 M sodium acetate (pH 5.2, ice cold) and 2.5

volumes of 100% ethanol (-20 °C).

Synthesis of radiolabeled cDNA

A solid mixture of primers homologous to the 3’ ends of the predicted 4,290 open
reading frame (ORFs) in E. coli was obtained from Sigma-Genosys (Panorama E. coli
cDNA labeling primers) and was resuspended in 40 1.1L sterile RN ase-free water before
use. The E. coli RNA (5 ug) was pelleted from stored RNA sample (—80 °C) by
centrifugation (microfuge, 20 min, 4 °C), washed carefully by addition of 20 11L 70%
ethanol (ice cold), and then pelleted by centrifugation (microfuge, 20 min, 4 °C). The
RNA was dissolved in 8.5 ML of sterile RNase-free water, mixed with 8 1.11 of the
prepared E. coli cDNA labeling primers, and incubated in a heat block at 90°C for 2 min
to denature the RNA. Then the E. coli cDNA labeling primers were annealed to the RNA
template by removing the block from the heating device and placing on the work bench
to cool the sample slowly to 42°C. To this sample 2 11L deoxynucleoside triphosphate

mix (lacking dCTP, 10 mM each), 3 uL DTT (0.1 M), 7 ML reverse transcriptase buffer,

192

1.5 11L dCTP (0.2 mM), 2 11L Superscript II Rnase H’ reverse transcriptase (400 U), 1 11L
RNase inhibitor (40 U), and 2 11L [or-33P]dCTP (20 u Ci; specific activity, 2000-3000
uCi/mmol; New England. Nuclear) were added to initiate cDNA synthesis (final volume
35 11L). The reaction mixture was incubated at 42 °C for 1 h. To this sample 1.5 11L
EDTA (0.5 M, pH 8.0) and 1.5 1.1L NaOH (5 M) was added to degrade the RNA at 65 °C
for 30 min. The sample was neutralized by adding 4 ML Tris-HCl (1 M, pH 8.0) and 3.5
ML HCl (2 M) followed by addition of 54.5 uL HZO to make the total volume 100 11L.
An aliquot (2 11L) of this sample was withdrawn to serve as a control for measuring total
radioactivity by scintillation counter. In the remaining sample, radioactive cDNA was
separated from other small molecules by gel filtration using a Micro Bio-spin P-30 Tris
Chromatography Column (Bio-rad) according to the procedure provided by the
manufacturer. An aliquot (2 11L) of the purified cDNA was withdrawn to measure the
radioactivity of the synthesized cDNA by scintillation counter. The remaining purified
cDNA was ready for hybridization. Approximately 50-70% of total radioactive label was
incorporated into synthesized cDNA.

To measure the radioactivity by scintillation counter, 2 11L sample was directly
added to a small filter paper. The filter paper was put into a vial filled with 10 mL
3a20TM complete counting cocktail (Research Products International). After shaking the
vial vigorously, the radioactivity of the sample was recorded in a 1414 Winspectral liquid

scintillation counter [PerkinElmer LS (WALLAC)].

Hybridization

193

After preparing the radioactively labeled cDNA, the next step was to perform a
hybridization to the Panorama gene array (Sigma-Genosys) containing duplicate, UV-
cross-linked arrays of 4,290 ORF-specific DNAs. All blots (filters) used in these
experiments were from the same manufacturing lot. Each blot (filter) was stripped and
used a maximum of three times. A total of 24 hybridizations (48 arrays) were analyzed.
The hybridizations were performed in roller bottles at 65 °C in a hybridization incubator
(Fisher Biotech), where minimal volumes of hybridization solutions are employed.
Before use, the temperature control in the hybridization incubator was calibrated
according to the procedure provided by the manufacturer.

The hybridization and washing steps were carried out as described by Tao.25 The
hybridization incubator was prewarmed to 65 °C. The roller bottles were filled with
water and placed into hybridization incubator to check if they were leaking. Sheared
salmon sperm DNA (Invitrogen, 100 ug/mL), which is a blocking agent to reduce the
background in hybridization experiments,” was boiled on 95 °C water bath for 15 min
followed by rapid cooling on ice to denature the DNA. The 10 mL hybridization solution
were prepared by mixing 1 mL 20% sodium dodecyl sulfate (SDS), 2.5 mL 20X SSPE
(3.6 M NaCl, 0.2 M sodium phosphate, 20 mM EDTA; pH 7.4), 6.2 mL distilled water,
0.2 mL 50X Denhardt’s solution (USB, prewarmed to 65 °C) and 0.1 mL sheared salmon
sperm DNA.25 The blot was rinsed twice in 400 mL 2X SSPE (0.36 M NaCl, 20 mM
sodium phosphate, 2 mM EDTA; pH 7.4) for 10 min in a big container. A hybridization
nylon mesh (Hybaid) was cut into certain size (24.5 cm x 12.5 cm) and used to wrap the
blot. The wrapped blot was inserted into the roller bottle filled with 50 mL 2X SSPE

(prewarmed to 65 °C). After washing the blot twice with 50 mL 2X SSPE (prewarmed to

194

65 °C) inside the roller bottle, 5 mL prepared hybridization solution (prewarmed to
65 °C) was added to the roller bottle to initiate prehybridization at 65 °C. After 6 h of
prehybridization, the hybridization solution inside the roller bottle was discarded and the
labeled cDNA (boiled on 95 °C water bath for 15 min followed by rapid cooling on ice
for 5 min to denature) in a fresh 4 mL hybridization solution was added. After
hybridization for 16 h at 65 °C, the blot was washed three times (5 min per wash) by
inverting the roller bottle by hand at room temperature and three times at 65 °C (25 min
per wash) in the hybridization incubator with 80 mL 0.5X SSPE containing 0.2% sodium
dodecyl sulfate. The wet blot was wrapped in Mylar film (1.5 pm, Persuation Marketing
Specialists) and was exposed to a Low Energy Storage Phosphor Screen (Molecular
Dynamics) for 20 h. The imaging screen was scanned at a 50 um pixel size rather than a
100 um or 200 um pixel size, for greater resolution of spots, using a STORM 820
Phosphoimager (Molecular Dynamics) in the Biochemistry Department of Michigan
State University. Each image file was about 50 MB and was backed up immediately.
The blot was stripped at 100 °C with 1% SDS in Tris-EDTA buffer according to the
procedure specified by the manufacturer. The same blot was hybridized, stripped, and
rehybridized for 3 times. The stripped blot was wrapped in a plastic food wrap and was

kept at —20 °C until ready to use.

Quantitation of cDNA as a relative measure of gene expression
The image files were analyzed by determining the pixel density (intensity) for
each spot in the array by using Array Vision software (version 6.1; Imaging Research

Inc.). A grid of individual ellipses corresponding to each of the DNA spots on the blots

195

was laid down on the image to designate each spot to be quantified. Background was
subtracted automatically by the Array Vision software using the “surrounding entire
template” method. The intensities for each spot were exported from Array Vision
software into a Microsoft Excel spreadsheet. The sum of values for all 4,290 ORFs in
each array was normalized to 100 million to allow comparisons of gene expression
between experiments. For example, intensity value of 10000 corresponded to 0.01% of
total cDNA hybridized to the array. For each condition investigated (with or without
IPTG addition, different times during fermentation; 8 different conditions), expression
levels for 6 arrays (3 filters with duplicate arrays) were measured. Results for every time
point represent an average of six determinations (three hybridizations each, two arrays
per filter). A threshold value for reliable detection was established as 0.002% of total
bound cDNA (intensity value 2000).25 An average expression level for all 8 conditions
was calculated for each gene using Microsoft Excel. The 347 genes with an average

expression intensity below 2000 were not included in further statistical analysis.

Statistical analysis

To address the issue of reproducibility of the gene expression data, the coefficient
of variation (standard deviation expressed as a percentage of the mean) was used as a
standard measure of gene expression data quality described by Tao.25 The standard
deviation is a measure of how spread out a distribution in a sample is. The formula for
computing standard deviation (0) in a sample using Microsoft Excel (function STDEV) is

= nzxz -(2x)2

n(n — 1)

 

o where n is the number of scores in a sample. The coefficient of

 

196

variation for each gene under 8 different conditions was calculated (n = 6) and the
average coefficient of variation of total 3943 genes under each condition was calculated.
Student's t test is one of the most commonly used techniques for testing a
hypothesis on the basis of a difference between sample means. Explained in layman's
terms, the t test determines a probability p that two populations are the same (null
hypothesis) with respect to the variable tested. Some researchers suggested that if the
difference is in the predicted direction, only one half (one "tail") of the probability
distribution is considered and thus divide the standard p-level reported with a t-test (a
"two-tailed" probability) by two. Others, however, suggested that the standard, two-
tailed t-test probability should always be reported. The only assumption in the student I
test is that the population follows a normal distribution (small samples have more scatter
and follow what is called a t distribution). Because log values of expression data have
previously been shown to approximate a normal distribution,” log values of expression
data were generated using Microsoft Excel and the significance of gene comparisons
between two samples (with IPTG treatment and without IPTG treatment) was evaluated.
by computing the probability values for the null hypothesis using two-tailed student t test
in Microsoft Excel (function TTEST). The t test can be performed knowing just the
means, standard deviation, and number of data points for the two samples. The two

2:;

 

 

samples t test yields an intermediate statistic t, in which t= . Note that the
J03 0,.
+ —'.
m n

numerator of the formula is the difference between means of two samples (x-bar and y-
bar is the sample mean). The denominator is a measurement of experimental error in the

two samples combined (0x and oy is the sample standard deviation, m and n is the number

197

of scores in the sample). The higher the value of t, the greater the confidence that there is
a difference. A precise probability value p can be attached to that confidence using the t
distribution table originally worked out by W.S. Gossett?6 The following formula was

used to approximate the degrees of freedom used in the r distribution table:

2 2

 

 

m n

tn (4

+
m—l n—l

 

df=

 

Cluster analysis was used to compare expression data for each condition (with or
without IPTG addition, different times during fermentation). Clusting analysis was
performed using hierarchical Clusting methods by Cluster and TreeView software
(download from http://rana.lbl.gov/ in Esien’s lab) as described by Eisen.27 The gene
(condition) similarity used was a form of correlation coefficient. For any two conditions

X and Y observed over a series of N genes, a similarity score was computated as

 

1
follows:S X,Y =—
( > NE

i= LN

X,—§)(Y,—Y

) where X-bar and Y-bar is the mean of
01’

OX

 

observations and 0X and oY is the standard deviation of the observations. For any sets of
11 conditions, an upper-diagonal similarity matrix was computed by using the similarity
formula described above, which contained similarity scores for all pairs of conditions.
The matrix was scanned to identify the highest value (respresenting the most similar pair
of conditions). A node was created joining these two conditions, and a gene expression
profile is computed for the node by averaging observation for the joined elements. The
similarity matrix was updated with this new node replacing the two joined elements, and

the process was repeated n-l times until only a single element remains.

198

Medium for PCR mutagenesis of araG

The selection medium for PCR mutagenesis of araG gene consisted of KHZPO4 (3
g/L), NazHPO4 (6 g/L), NH4C1 (1 g/L), NaCl (0.5 g/L), MgSO4 (1 mM), glucose (4 g/L),
thiamine (6 mg/L), leucine (25 mg/L), nicotinic acid (6 mg/L) and phenylalanine (3.3
g/L).

The SOC medium used for transformation of electrocompetent cells was prepared
as the following. Tryptone (2 g), yeast extract (0.5 g), NaCl (1 mL 1 M NaCl), KCl (0.25
mL 1 M KCl) was added to 97 mL of distilled water. After dissolving and autoclaving, 1
mL 1 M MgCl2 and 1 mL 2 M glucose was added to the solution. The complete medium

was filtered through a 0.22 pm filter.

Plasmid pJY6.119A

This 6.3-kb plasmid was constructed by ligation the open reading frame (ORF) of
araG into the Smal/BamHI site of pJF118EH. The araG ORF was amplified from E. coli
RB791 genomic DNA using the following primers: 5’-
CGGGATCCTTACCCGCGACGCGCTTT and 5’-
TCCCCCGGGATGAATTATCAGAACGACGAT. Localization of the resulting 1.0-kb

fragment into pJF 1 18EH afforded pJY6.1 19A.

Plasmid pJY6.149 (#1-#13)
The araG ORF from plasmid pJY6.119A was subject to PCR mutagenesis using

the following primers: 5’-CGGAATTCATGAATTATCAGAACGACGAT and 5’—

199

CGGGATCCTTACCCGCGACGCGCTTT. The PCR product was purified and digested
with EcaRI/BamHI. The resulting DNA was ligated to pJF118EH, which has been
digested with EcaRI/BamHI. The resulting ligation mixture was introduced by
transformation into CB734 electrocompetent cells and the resulting transformants were

selected at 37 °C on the minimal selection plate describe above. The 13 plasmids

pJY6.149 (#1-#13) were isolated from the colonies that grew on the selection plate.

Plasmid pJY 6.186

A 12.2-kb DNA fragment was amplified by Long PCR (described below) from
plasmid pJYl.216A using the following primers: 5’-
GAAGATCTGATCTGAACGGGCAGCTGAC and 5’-
GAAGATCTCGATCCTGTTTATGCTCGTTTG. The PCR product was digested with

BglII and religated by itself to afford 12.2-kb plasmid pJY6.186.

Plasmid pJY6.231A

This 13.2-kb plasmid was constructed by ligation the ORF of araGFBR (digested
with BamHI) into the BglII site of pJY6.186. The araGFBR ORF was amplified by PCR
from plasmid pJY6.149#4 using the following primers: 5’-
CGGGATCCATGAATTATCAGAACGACGAT and 5’-
CGGGATCCTTACCCGCGACGCGCTTT. The ORF of araGFBR is transcribed in the

same orientations as the araF promoter in pJY6.186.

PCR mutagenesis of araG gene

200

 

The protocol for mutagenic PCR is derived from standard PCR conditions as
described by Cadwell.” Plasmid pJY6.119A was digested with Smal/BamHI and the
araG ORF DNA was purified by gel electrophoresis. Mutagenic PCR buffer (10X)
contained 70 mM MgC12, 500 mM Kcl, 100 mM Tris (pH 8.3), and 0.1% (w/v) gelatin.
Deoxynucleoside triphosphate mix (10X) contained 2 mM dGTP, 2 mM dATP, 10 mM
dCTP, and 10 mM dTTP. The mutagenic PCR reaction was prepared by mixing 10 uL
10X mutagenic PCR buffer, 10 1.1L 10X dNTP mix prepared above, 14 uL 50 mM MgC12,
30 pmoles of each primer, 20 ng of input DNA, and an amount of water that brought the
total volume to 88 1.11. To this sample was added 10 ML 5 mM MnCl2 (make sure a
precipitate was not formed) and 2 ul Taq polymerase (5 Unit), bringing the final volume
to 100 11L. The PCR reaction was performed in a thermal cycler (Eppendorf,
Mastercycler Gradient). The sample was incubated for 30 cycles of 94°C for 45 s, 45°C
for 1 min, and 72°C for 45 s. A “hot start” procedure or a prolonged extension time at

the end of the last cycle was not employed.

Preparation of electrocompetent cells (E. coli)

E. coli CB734 was prepared for electroporation with the following procedure. E.
coli strain CB734 was initially grown from a single colony in 5 mL LB/Cm and shaken
for 16 h at 37°C . From the 5 mL culture, 2 mL was added to 500 mL 2 x YT medium
(Bacto-tryptone 16 g, yeast extract 5 g, NaCl 5 g per liter) which was grown until
reaching an absorbance of 0.6 - 0.8 at 600 nm. The cells were chilled to 4°C then
harvested by centrifugation (3 000g, 5 min, 4 °C). The supernatant was discarded, and

the cell pellet was resuspended in 500 mL ice cold water. The cells were harvested by

201

 

centrifugation (3 000g, 10 min, 4 °C) then resuspended in 250 mL ices cold water.
Following harvesting by the same conditions, the cells were resuspended in 100 mL ice
cold 10% glycerol solution (Sterile filtered, not autoclaved). The cells were once again
harvested by the same conditions and resuspended in 1.5 mL ice cold 10% glycerol
solution. 0.1 mL aliquots were fast frozen in liquid nitrogen and store at —80 °C.

For transformation, 40 uL of cells were combined with plasmid DNA followed by
incubation for 2 min at 0 °C. The cells were then transferred to a pre-chilled
electroporation cuvette with a 0.2 cm gap width. The cells were electroporated at 2.5 kV,
25 uF, and 200 Ohms using Gene Pulser II (Bio-rad). Following electroporation, 1 mL
room temperature SOC medium was immediately added to the cuvette and incubated at

37 °C for 1 h. The entire cell solution was plated with the appropriate selection plate.

Long PCR

The long PCR (12.2 kb) from plasmid pJYl.216A was performed according to the
modified protocol of Cheng.29 Efficient Long PCR results from the use of two
polymerases: a non-proofreading polymerase is the main polymerase in the reaction, and
a proofreading polymerase is present at a lower concentration. The Tth DNA polymerase
(Epicentre Technologies) was used as the main component and Vent DNA polymerase
(Invitrogen) as the fractional-component polymerase. Each reaction contained 5 U Tth
DNA polymerase, 0.2 U Vent DNA polymerase, 1X Long PCR buffer [(85 mM KOAc,
25 mM tricine (pH 8.7), 8% glycerol, 1% DMSO, 1.2 mM Mg(OAc)2], 0.2 mM each
dNTP, 50 pmoles each primer and 40 ng template DNA. The target sequences were

amplified through 15 cycles of 94 °C for 10 s, 64 °C for 1 min and 72 °C for12 min 18 s,

202

 

followed by another 15 cycles of 94 °C for 10 s, 64 °C for l min and 72 °C forl2 min 18
s + 30 s/cycle. A hot start method was employed for Long PCR. The reaction was
divided into two parts: a template/primer fraction which was 4/5 of the reaction volume,
and a polymerase fraction which constituted the remaining 1/5 of the reaction. Each
fraction was 1X buffer concentration. The polymerase fraction contained only
polymerase, buffer and water. All other components were included in the
template/primer fraction. The template/primer fraction was incubated in the PCR
machine to 94 °C for 10 s to denature. After denaturing, an 80 °C step was used for
adding the polymerase fraction. The following guidelines were used for picking primers
for Long PCR: primers are 20 to 23 bases in length; G + C = 12 bases; A + T = 8 to 11
bases; ideal Tm = 60 to 68 °C; avoid primer hairpins; avoid primers with 3’
complementarity (results in primer-dimers). The primers were designed with the aid of

Oligo software.

Chapter 4

Silicon oil centrifugation

Cells were separated from the medium and inactivated by silica oil centrifugation
with the further working up procedure described by Ebbighausen.30 Cells from
fermentation experiments were harvested by centrifugation (microfuge, 30 s, 4 °C)
followed by one wash cycle in a buffer contain 50 mM Tris-HCl, pH 7.5, 10 mM NaCl,
20 mM KCl. The final pellet was suspended in the same buffer at a cell density of about

109 cells/mL (OD,>00 = 5-10). An aliquot (200-400 1.1L) was transferred to a 1.5 mL

203

 

 

microfuge tube containing a layer of 260 ML silicone oil (AR200, Fluka) floating on a
layer of 120 11L 20% perchloric acid. Cells (800 11L) from shake-flask experiments were
directly transferred to the above microfuge tube. The tube was immediately centrifuged
(microfuge, l min, 4 °C) followed by removing the supernatant and silicone oil layer.
After homogenization the sediment by sonication (six 30 s bursts and 30 s intermittent
cooling in ice), the perchloric extracts were neutralized with a solution containing 5 M
KOH/l M triethanolamine and chilled on ice for 10 min to precipitate the KClO4 formed
on neutralization. Together with the denatured protein and residual oil the precipitated
salt was then sedimented by centrifugation (microfuge, 5 min, room temperature) and the

supernatant was used for shikimic acid analysis.

Shikimic acid quantitation

A recently reported chromogenic assay that facilitates rapid detection and
quantitation of shikimic acid31 was used to analyze the above supernatant. Supernatant
(10 11L — 50 11L) was added with water to make final volume 250 11L. Then 250 1.1L of
periodate reagent (0.5% periodic acid and 0.5% sodium meta-periodate) was added.
After 30 minutes at 37 °C, the reactions were quenched with 0.3 mL of 1 N NaOH
followed by 0.2 mL of 0.056 M sodium sulfite. Absorbance of the solution at 382 nm
was used to quantitate the amount of shikimic acid in the solution. A standard
concentration curve was determined for shikimic acid using solutions of authentic,

purified shikimic acid.

204

Shake flask experiment of E. coli SP1.4
E. coli SP1.4 cells were grown in 100 mL LB medium for 12 h, washed with 100
mL M9 and resuspended in 100 mL M9 minimal medium with glucose and serine

addition. After cultivation of cells for 36 h at 37 °C, the supernatant was analyzed by 1H

NMR.

Mini-Tn10 mutagenesis of E. coli SP1.4

The mini-Tn10 phage vehicle ANK1324 was obtained from the American Type
Culture Collection (ATCC number 77347). This mini-Tn10 construct is useful for
general transposon mutagenesis in E. coli and is marked with a chloramphenicol
resistance fragment from Tn9. The procedures for generating transpositions from A into
E. coli chromosome were described by Kleckner.32 A phage lysate was prepared by
propagation of phage in E. coli LE392 using the following procedure. Serial dilutions of
A phage stock (0.1 mL, 10‘l to 10°) in TMG buffer (1.2 g/L Tris base, 2.46 g/L
MgSO4.7HzO, 0.1 g/L gelatin, pH 7.4) were prepared in sterile test tubes (13 x 100 mm).
An aliquot (0.1 mL, approximately 5 x 108 cells) of an overnight culture of the E. coli
LE392 was added to each tube. To each tube 2.5 mL sterile, molten top agar (10 g/L
tryptone, 5 g/L NaCl, 7 g/L agar, 45 °C) was added. The contents of each tube were
mixed and poured immediately onto a pre-warmed (37 °C) TBl plate (10 g/L tryptone, 5
g/L NaCl, 11 g/L agar), swirling gently to achieve uniform coverage of the plate. After
the agar had solidified, the plates were incubated at 37 °C overnight. A single plaque
was picked up with a micropipette and transferred to 10 mL LB plus 10 mM MgSO4 and

0.1 mL of a fresh overnight LB culture of the E. coli LE392. After shaking the sample at

205

37 °C for 4 - 5 h, the culture should gradually become somewhat cloudy, then clear. A
few drops of chloroform were added to the culture to make certain that all of the cells had
lysed and the culture was then centrifuged (20 000g, 10 min, room temperature).
Aqueous phage lysate was stored in 1.5 mL microfuge tubes over several drops of CHCl3
at 4 °C.

The titer of the phage lysate was checked by the following way. The phage lysate
was diluted 10‘3 and 10‘6 respectively using LB solution. The diluted phage lysate and 0.1
mL E. coli LE392 overnight cells was mixed with different ratios and plate in 4 mL top
agar on a LB agar plate. The plates were incubated for 8 h at 37 °C. The number of
plaque formed was counted and the titer of the phage lysate obtained was normally about
1-2 x 10'0 pfu/mL.

Transpositions from A delivery vehicles were isolated by infecting SP1.4 cells
with the phage. E. coli SP1.4 was grown overnight on LB medium with 0.2% maltose.
Cells were concentrated and resuspended in the LB medium with IPTG (2.5 mM). Cells
(0.1 mL, approximately 108-109 cells) in the concentrated culture were mixed with certain
amount of the prepared phage lysate to give a ratio of 0.12 to 0.3 phage per cell. After
incubating the cells at room temperature for 30 min, and 37 °C for 90 min, 200 pl 1 M
sodium citrate was added. After centrifugation (microfuge, l min, room temperature)
and resuspension in 200 ill LB with sodium citrate (50 mM), the infected cells were
plated on LB plate with chloramphenicol to select for transposition mutants. Normally
100 to 200 colonies per plate were obtained by this procedure. Streaking techniques were

then applied to obtain single isolated mutants.”

206

Y-linker PCR technique

Construction of the Y-linker was described by Kwon” using linker 1 (5’-
TTTCTGCTCGAATTCAAGCTTCTAACGATGTACGGGGACACATG) and linker 2
(5’-TGTCCCCGTACATCGTTAGAACTACTCGTACCATCCACAT). Linker 2 (9 ML,
350 ng/uL) was first phosphorylated at the 5’ end using T4 polynucleotide kinase
(Invitrogen) according to the procedure provided by the manufacturer. After heat
denaturation of polynucleotide kinase at 65 °C for 20 min, 9 ML linker l (350 ng/uL) was
added to a final volume of 29 11L. This mixture was heated at 95°C for 2 min and slowly
cooled to room temperature, resulting in the formation of the Y linker at a final
concentration of 200 ng/uL.

The genomic DNA was isolated from mini-Tn10 mutants of E. coli SP1.4 by the
method of Pitcher” and completely digested with NlaIII (New England Biolabs).
Approximately 40 ng of the digested DNA was ligated to 1 ug of the Y linker with l ptl
T4 DNA ligase (1 Unit, Invitrogen) in a final volume of 20 11L. After overnight
incubation at room temperature, the reaction mixture was diluted with double distilled
water to a final volume of 200 11L and heated at 65 °C for 10 min to denature T4 DNA
ligase. 4 ML aliquots were used as templates in the PCR amplification. All PCR
reactions were performed using a primer specific to mini-Tn10 (5’-
TCCCAGAGCCTGATAAAAAC) and a primer specific to Y-linker (5’-
CTGCTCGAATTCAAGCTTCT). Taq DNA polymerase was used to perform the PCR.
The target sequences were amplified through 30 cycles of 94 °C for 1 min, 55 °C for l
min, and 72 °C for 1 min, followed by a final cycle of 72 °C for 10 min. The PCR

products were analyzed on a 1.5% agarose gel and purified using Zymoclean Gel DNA

207

Recovery Kit from Zymo Research Company. Sequences are determined on an ABI

PRISM®3100 Genetic Analyzer by biochemistry facility in Michigan State University.

E. coli KL3 derivatives

E. coli KL3 was made tchssTnIO, ile::Tn10, narU::Tn10, ydaQ::TnlO,
ksgA.-:Tn10, ykngTnIO by P1 transduction from respective E. coli SP1.4 mutants.
Colonies were plated onto LB plates containing Cm to select for colonies which had been

transduced by mini-Tn10 insertion.

208

REFERENCE

1 Pittard, J .; Wallace, B. J. Distribution and Function of Genes Concerned with
Aromatic Biosynthesis in Escherichia coli. J. Bacterial. 1966, 91, 1494-1508.

2 Li, K.; Mikola, M. R.; Draths, K. M.; Worden, R. M.; Frost, J. W. Fed-Batch
Fermentor Synthesis of 3-Dehydroshikimic Acid Using Recombinant Escherichia coli.
Biotechnol. Bioeng. 1999, 64, 61-73.

3 Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von Daeniken, R.;
Weber, W.; Frost, J. W. Hydroaromatic Equilibration During Biosynthesis of Shikimic
Acid. J. Am. Chem. Soc. 2001, 123, 10173-10182.

4 Meyer, D.; Schneider-Fresenius, C.; Horlacher, R.; Peist, R.; Boos, W. Molecular
Characterization of Glucokinase from Escherichia coli K-12. J. Bacterial. 1997, 179,

1298-1306.

5 Akowski, J. P.; Bauerle, R. Steady-State Kinetics and Inhibitor Binding of 3-
Deoxy-D-arabina-heptulonsonate-7-Phosphate Synthase (Trptophan sensitive) from
Escherichia coli. Biochemistry 1997, 36, 15817-15822.

6 Furste, J. P.; Pansegrau, W.; Frank, R.; Blocker, H.; Scholz, P.; Bagdasarian, M.;
Lanka, E. Molecular Cloning of the Plasmid Rp4 Primase Region in a Multi-Host-Range
tacP Expression Vector. Gene 1986, 48, 119-131.

7 Parker, C.; Barnell, W. O.; Snoep, J. L.; Ingram, L. O.; Conway, T.
Characterization of the Zymamanas mobilis glucose facilitator gene product (glf) in
recombinant Escherichia coli: Examination of transport mechanism, kinetics and the role
of glucokinase in glucose transport. Mol. Microbiol. 1995, 15, 795-802.

8 Bartolome, B.; Jubete, Y.; Martinez, E.; de la Cruz, F. Construction and
Properties of a Family of pACYC184-derived Cloning Vectors Compatible with pBR322
and its Derivatives. Gene, 1991, 102, 75-78.

9 Balbas, P.; Soberon, X.; Merino, E.; Zurita, M.; Lomeli, H.; Valle, F.; Flores, N .;
Bolivar, F. Plasmid vector pBR322 and its special-purpose derivatives - a review.
Gene 1986, 50, 3-40.

10 Tsang, T.; Copeland, V.; Bowden, G. T. A Set of Cassette Cloning Vectors for
Rapid and Versatile Adaptation of Restriction Fragments. Biatechniques 1991, 10, 330.

11 Farabaugh, M. A. Biocatalytic Production of Aromatics from D-Glucose. MS.
Thesis, Michigan State University, 1996.

209

12 Chandran, S. S.; Yi, J.; Draths, K. M.; von Daeniken, R.; Weber, W.; Frost, J. W.
Phosphoenolpyruvic Acid Availability and the Biosynthesis of Shikimic Acid.
Biotechnol. Prag. 2003, 19, 808-814.

13 Miller, J. H. Experiments in Molecular Genetics; Cold Spring Harbor Laboratory:
Plainview, NY, 1972.

14 Yi, J .; Li, K.; Draths, K. M.; Frost, J. W. Modulation of Phosphoenolpyruvate
Synthase Expression Increases Shikimate Pathway Product Yields in E. coli. Biotechnol.
Prag.2002,18, 1141-1148.

15 Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning: A Laboratory
Manual; Cold Spring Harbor Laboratory: Plainview, NY, 1990.

16 Silhavy, T. J .; Berman, M. L.; Enquist, L. W. Experiments with Gene Fusions;
Cold Spring Harbor Laboratory: Plainview, NY, 1984.

17 Pitcher, D. G.; Saunders, N. A.; Owen, R. J. Rapid Extraction of Bacterial
Genomic DNA with Guanidium Thiocyanate. Letters in Applied Microbiology 1989, 8,
151-156.

18 Miller, J. H. A short course in bacterial genetics; Cold Spring Harbor Laboratory:
Plainview, NY, 1992.

19 Bradford, M. M. A Rapid and Sensitive Method for the Quantitation of
Microgram Quantities of Protein Utilizing the Principle of Protein-Dye Binding. Anal.
Biochem. 1976, 72, 248-254.

20 Schoner, R.; Herrmann, K. M. 3—Deoxy-D-arabinase-heptulosonate 7-phosphate
synthase. J. Biol. Chem. 1976, 251, 5440.

21 Gollub, E.; Zalkin, H.; Sprinson, D. B. Assay for 3-deoxy-D-arabinase-
heptulosonate acid 7-phosphate synthase. Meth. Enzymal. 1971, 17A, 349.

22 Cooper, R. A. Phosphoenolpyruvate Synthetase. Methods Enzymal. 1969, 13, 309-
314.

23 Fraenkel, D. G.; Horecker, B. L. Pathways of D-Glucose Metabolism in
Salmonella typhimurium. J. Biol. Biochem. 1964, 239, 2765-2771.

24 Paoletti, F.; Williams, J. F.; Horecker, B. L. An enyzmic method for the analysis
of D-erythrose 4-phosphate. Anal. Biochem. 1979, 95 , 250-253.

25 (a) Tao, H.; Gonzalez, R.; Martinez, A.; Rodriguez, M.; Ingram, L. 0.; Preston, J.
F; Shanmugam, K. T. Engineering a homo-ethanol pathway in Escherichia coli:

210

increased glycolytic flux and levels of expression of glycolytic genes during xylose
fermentation. J. Bacterial. 2001, 183, 2979-2988. (b) Gonzalez, R.; Tao, H.; Purvis, J.
E.; York, S. W.; Shanmugam, K. T.; Ingram, L. 0. Gene Array-Based Identification of
Changes That Contribute to Ethanol Tolerance in Ethanologenic Escherichia coli:
Comparison of KOll (Parent) to LY01 (Resistant Mutant) Biotechnol. Prag. 2003, 19,
612-623.

26 Hocking, R. R. Methods and Applications of Linear Models. Regression and the
Analysis of Variance. 1996, New York: Wiley.

27 Eisen, M. B.; Spellman, P. T.; Brown, P. O.; Botstein, D. Cluster analysis and
display of genome-wide expression patterns. Prac. Natl. Acad. Sci. USA 1998, 95, 14863-
14868.

28 Cadwell, R. C.; Joyce, G. F. Mutagenic PCR. PCR Methods Applic. 1994, 4,
$136-$139.

29 Cheng, S.; Fockler, C.; Barnes, W.; Higuchi, R. Effective amplification of long
targets from cloned inserts and human genomic DNA. Proc. Natl. Acad. Sci. 1994, 91,
5695-5699.

30 Ebbighausen, H.; Weil, B.; Kraemer, R. Isoleucine Excretion in Corynebacterium
glutamicum: Evidence for a Specific Efflux Carrier System. Appl. Microbial. Biotechnol.
1989, 31, 184-190.

31 Cromartie, T. H.; Polge, N. D. In US; (Syngenta Limited, UK). Method of
Detecting Shikimic Acid. Int. Patent WOOD/49399, August 24, 2000.

32 Kleckner, N.; Bender, J .; Gottesman, S. Uses of Transposons with Emphasis on
Tn10. Methods in Enzymalagy 1991, 204, 139-180.

33 Kwon, Y. M.; Ricke, S. C. Efficient Amplification of Multiple Transposon-
Flanking Sequences. J. Microbial. Methods 2000, 41, 195-199.

211

 

 

Aug: £1.23:

i. .;

 

 

 

 

 

 

 

 

 

 

 

 

an“;