Qua.
’1.

.1. Jr
ad: PBS
5; .91 v
.ymcfifl? :
. Jun... :43}
.l. u. .-

, 1,...

1
\a

u. I: u:
341“. Eur“ n.
,1 ‘,.IP
i... .1
x- 5“!

. ..
- 5
hfim: $512.9;
r: r!
ham.
w (it

#:fifi.

Mi} .

3mm.
.9...

a. .. (.
‘ I
'31. outfit!»
I

s.“
4 c “war...

 

 

\. < 44. 3 us. up ‘
i UH. . u. .¢.0.HMnn.1..\le‘ .- . It)»
M . 3:. r

This is to certify that the

dissertation entitled

CHEMO-ENZYMATIC SYNTHESIS OF AROMATICS VIA NON—
SHIKIMATE PATHWAY INTERMEDIATES

presented by

Chad A. Hansen

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Chemis try

 

w W/

/, Major professor

Date 5/4/02-

 

MSUt's an Affirmative Action/Equal Opportunity Institution 0-12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p1/CIRC/DateDue,indd-p,1

 

CHEMO-ENZYMATIC SYNTHESIS OF AROMATICS VIA NON-

SHIKIMATE PATHWAY IN TERMEDIATES
By

Chad A. Hansen

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Chemistry

2002

ABSTRACT

CHEMO-ENZYMATIC SYNTHESIS OF AROMATICS VIA NON-

SHIKIMATE PATHWAY INTERMEDIATES
By

Chad A. Hansen

The manufacture of aromatic molecules by the chemical industry relies on the use
of benzene, toluene, and o, m, or p-xylenes as the starting material. These aromatic
starting materials are obtained from the BTX cut of fossil fuel refining of which benzene
is most important for aromatic manufacture. Benzene exposure has been linked to acute
leukemia and non-Hodgkin’s lymphoma. The environmental release of benzene must be
reduced as mandated by the Chemical Manufacturing Rule issued by the U. S.
Environmental Protection Agency.

Establishing the connectivity between the carbohydrates D-glucose, D-xylose and
L-arabinose and aromatic molecules is the first step towards development of alternative
starting materials for the manufacture of aromatic molecules by the chemical industry.
Carbohydrates offer the advantage of being nontoxic, nonvolatile, and derived from
renewable resources by the depolymerization of the biopolymers starch and cellulose.

The conversion of carbohydrates to the aromatics catechol and hydroquinone
along with the polyhydroxybenzene pyrogallol have been established from the
intermediates of the shikimate pathway for aromatic amino acid biosynthesis. An

Escherechia coli construct has been developed for the conversion of D-glucose to the

hydroaromatic myo-inositol in 20 g/L. Oxidation of the axial alcohol of myo-inositol by
Gluconobacter oxydans ATCC 621 affords myo—2-inosose. Acid-catalyzed dehydration
of myo-2-inosose affords 1,2,3,4-tetrahydroxybenzene in 66% yield which has been

established as a starting material for aurantiogliocladin and coenzyme Q3. Attempts to

increase the titer of myo-inositol by fed-batch fermentation or the direct conversion of D-
glucose to myo—Z-inosose in one microbial host will be discussed. 2-Deoxy-scyllo-
inosose, which could be obtained from butirosin biosynthesis in Bacillus circulans, has
been aromatized to hydroxyhydroquinone in 39% yield under acid-catalyzed dehydration
conditions.

Triacetic acid lactone, obtained from an inadequate supply of NADPH during the
biosynthesis of fatty acids or the polyketide 6-methyl salicylic acid, has been converted to
phloroglucinol and resorcinol. Efforts towards the biosynthesis of triacetic acid lactone
either by fatty acid biosynthesis or polyketide biosynthesis will be discussed. Resorcinol
can be derived from phloroglucinol methyl ether in 80% yield by the development of a
novel deoxygenation methodology for polyhydroxybenzenes. The general utility of the
deoxygenation methodology has been established by the conversion of phloroglucinol to
resorcinol in 82% yield, 1,2,3,4-tetrahydroxybenzene to pyrogallol in 44% yield and the

conversion of hydroxyhydroquinone to hydroquinone in 53% yield.

Copyright by
Chad A. Hansen

2002

To my family

For their love and support

In memory of

Marie Hansen and Doris Ann Hansen

ACKNOWLEDGEMENTS

To start with, I wish to thank John W. Frost for pointing me to the correct
direction to take in my scientific career. John taught methe key to success is more about
hard work and perseverance than intelligence. John’s belief in me and his continued
patience helped me to overcome my fears and have lead to the completion of this degree.
I would also like to thank the members of my committee Milton R. Smith, III, John L.
McCracken and Robert E. Maleczka, Jr. for their guidance and valuable suggestions
during the course of this work. I would especially like to thank Robert E. Maleczka, Jr.
for his effort in making corrections to this body of work.

I am indebted to Dr. Karen Frost for helping to learn all of the biological
techniques used to carry out this work. She is certainly the lifeblood of the lab and is
invaluable to this group for her expertise. I am also thankful for having had the
opportunity to work with the likes of Feng Tian, Kai Li, Spiros Kombourakis, Sunil
Chandrin, Jessica Barker and John Arthur who once informed me “Purity is like virginity.
You either have it or you don’t.” I am also thankful to the current group members
Padmesh Venkitasubramanian, Jian Yi, Ningqing Ran, Jianto Guo, and Wei Niu who
offered many helpful suggestions throughout my time in the group.

I would like to thank my mother and father who were also patient and
understanding when I was not always able to find time to call or come home to visit.
Their belief in me carried me through each day. And finally, to the love of my life, I
would like to thank Aimee Brazil for always believing in me and continuing to believe in
us when I hit some of the lowest points in my life. I am fortunate to be spending the rest

of my life with you.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. xii

LIST OF FIGURES... ...................................................................................................... xiii

LIST ABBREVIATIONS ................................................................................................ xvii

Chapter 1 .............................................................................................................................. 1

INTRODUCTION ............................................................................................................ 1

Hydroxylated Aromatics ............................................................................................ 3

(a) Catechol .......................................................................................................... 5

(b) Resorcinol ....................................................................................................... 5

(c) Hydroquinone ................................................................................................. 6

(d) Pyrogallol ....................................................................................................... 7

(e) Phloroglucinol ................................................................................................ 8

Biocatalysis ............................................................................................................. 1 1

(a) Aspartame ..................................................................................................... 13

(b) Acrylamide ................................................................................................... 15

(c) L-Ascorbic Acid ........................................................................................... 16

Aromatic Products Derived from the Shikimate Pathway ....................................... 18

(a) 3-Dehydroshikimic Acid .............................................................................. 21

(b) Catechol ........................................................................................................ 21

(c) Adipic Acid .................................................................................................. 22

(d) Vanillin ......................................................................................................... 24

(e) Gallic Acid and Pyrogallol...........................................; .............................. 26

(f) Shikimic Acid, Phenol and p-Hydroxybenzoic Acid ................................... 27

(g) Quinic Acid and Hydroquinone .................................................................. 29

References ............................................................................................................... 32

CHAPTER 2 ...................................................................................................................... 37
BIOSYNTHESIS OF M YO-INOSITOL, M Y0-2-INOSOSE AND

SCYLLO-2,5-DIKETOINTOSITOL ................................................................................ 37

Introduction .............................................................................................................. 37

Production of myo-Inositol by Fed-Batch Fermentation ......................................... 39

(a) Background ................................................................................................... 39

(b) Fed-Batch Fermentor Conditions ................................................................. 40

(0) Expression of [NO] Under the tac Promoter ................................................ 44

(d) Sequencing of [NO] ..................................................................................... 46

(e) Codon Usage ................................................................................................ 48

(f) Resynthesized INOI : S YNINOI .................................................................. 50

(g) T7 Promoted INOI ...................................................................................... 57

(h) Coexpression of INOI With groESL ............................................................ 64

Production of myo-2-Inosose by Fed-Batch Fermentation ...................................... 67

(a) Background ................................................................................................... 67

vii

(b) Synthesis of myo-2-Inosose In Escherichia coli .......................................... 69
(0) Synthesis of myo-2-Inosose From D-Glucose With Gluconobacter

 

 

oxydans ....................................................................................................... 8 l
(d) Oxidation of myo-Inositol and neo-Inositol by Gluconobacter
oxydans ....................................................................................................... 86
Discussion .............................................................. ' ................................................. 89
References ............................................................................................................... 95
CHAPTER 3 ...................................................................................................................... 99
AROMATIZATION OF M YO-INOSITOL, 2-DEOXY-SCYLLO-INOSOSE AND
TRIACETIC ACID LACTONE ...................................................................................... 99
Introduction ............................................................................................................. 99
myo-Inositol Chemistry .......................................................................................... 100
(a) Background ................................................................................................. 100
(b) Nitric Acid Oxidation ................................................................................. 101
(c) Chemical Synthesis of myo-2-Inosose ........................................................ 104
(d) Aromatization of myo-2-Inosose ................................................................ 105
(e) Chemical Synthesis of l,2,3,4-Tetrahydroxybenzene ................................ 108
(f) l,2,3,4-Tetrahydroxybenzene as a Synthetic Building Block .................... 110
Synthesis and Aromatization of 2-Deoxy- scyllo-inosose ...................................... 118
(a) Synthesis of 2-Deoxy- scyllo-inosose ......................................................... 119
(b) Aromatization of 2-Deoxy- scyllo-inosose ................................................. 120
Aromatization of Triacetic Acid Lactone .............................................................. 122
Polyhydroxybenzene Deoxygenation Methodology .............................................. 126
Discussion ............................................................................................................. 133
References ............................................................................................................. 137
CHAPTER 4 .................................................................................................................... 140
EXPLORATION OF POLYKETIDE AND FATTY ACID BIOSYNTHESIS
FOR BIOSYNTHESIS OF TRIACETIC ACID LACTONE ........................................ 140
Introduction ........................................................................................................... 140
Evaluation of Polyketide Biosynthesis .................................................................. 141
Evaluation of Fatty Acid Biosynthesis .................................................................. 146
(a) Triacetic Acid Lactone by Escherichia coli Fatty Acid Biosynthesis ........ 149
(b) Triacetic Acid Lactone by Brevibacterium ammoniagenes Fatty
Acid Biosynthesis ...................................................................................... 152
(c) Site-directed Mutagenesis of fasB .............................................................. 155
(d) fasB Under the T7 Promoter ....................................................................... 165
Discussion .............................................................................................................. 168
References .............................................................................................................. 172
CHAPTER .................................................................................................................... 174
EXPERIMENTALS ....................................................................................................... 174
General Methods .................................................................................................... 174
(a) General Chemistry ...................................................................................... 174
(b) Reagents and Solvents ................................................................................ 174

viii

(c) Chromatography ......................................................................................... 175

(d) Spectroscopic and Analytical Measurements ............................................. 175
Protein Purification ................................................................................................ 176
(a) General Information ................................................................................... 176
(b) MIP synthase .............................................................................................. 176
(c) FAS-B .......................................................... ' ............................................... 178
Enzyme Assays ...................................................................................................... 179
(a) MIP synthase Assay ................................................................................... 179
(b) Inositol Dehydrogenase Assay ................................................................... 180
(c) Fatty Acid Ketoreductase Assay ................................................................ 181
(d) In vitro Analysis of Triacetic Acid Lactone ............................................... 181
Microbial Strains and Plasmids ............................................................... . .............. 1 82
Storage of Bacterial Strains and Plasmids ............................................................. 183
Culture Medium ..................................................................................................... 183
General Fed-Batch Fermentation Conditions ........................................................ 185
Analysis of Fermentation Broth ............................................................................. 186
Genetic Manipulations ........................................................................................... 188
(a) General Information ................................................................................... 188
(b) Large Scale Purification of Plasmid DNA ................................................. 189
(c) Small Scale Purification of Plasmid DNA ................................................. 191
((1) Restriction Enzyme Digestion of DNA ...................................................... 192
(e) Agarose Gel Electrophoresis ...................................................................... 192
(f) Isolation of DNA from Agarose ................................................................ 193
(g) Treatment of DNA with Klenow Fragment .............................................. 193
(h) Treatment of Vector DNA with Calf Intestinal Alkaline Phosphatase ..... 194
(i) Ligation of DNA ........................................................................................ 194
0) Preparation and Transformation of Competent Cells (E. coli) .................. 194
(k) Preparation of Electrocompetent Cells (G. oxydans) ................................. 196
(1) Preparation of Electrocompetent Cells (S. cerevisiae) ............................... 197
(m) ADEB Lysogeny ........................................................................................ 198
Chapter 2 ................................................................................................................. 199
Strain Construction ................................................................................................. 199
(a) Strain JWF1(DE3) ...................................................................................... 199
Plasmid Constructions ........................................................................................... 199
(a) Plasmid pAD1.45A ..................................................................................... 199
(b) Plasmid pAD1.88A .................................................................................... 200
(c) Plasmid pCH6.112A ................................................................................... 200
(d) Plasmid pCH6.123A ................................................................................... 200
(e) Plasmid pCH7.41 ........................................................................................ 201
(f) Plasmid pCH7.61B ..................................................................................... 201
(g) Plasmid pCH7.66A .................................................................................... 201
(h) Plasmid pCH7. 170 ..................................................................................... 202
(i) Plasmid pCH7.l94A ................................................................................... 202
(j) Plasmid pCH6.254A ................................................................................... 202
Culture Conditions for Codon Usage with MIP synthase ...................................... 203
Culture Conditions for groESL Heat Shock Proteins with MIP synthase ............. 203

ix

Oxidation of myo-Inositol by Gluconobacter oxydans ATCC 621 ....................... 204

Oxidation of neo-Inositol by Gluconobacter oxydans ATCC 621 ........................ 205
Fermentation with Gluconobacter oxydans ATCC 621/pCH6.254A .................... 208
Chapter 3 ................................................................................................................ 208
Synthetic Procedures .............................................................................................. 208
Synthesis of myo—2-Inosose.................................... ............................................... 208
(a) myo-2-inosose (air oxidation) ..................................................................... 208
(b) myo-Inositol acetinide l4 ........................................................................... 209
(c) Tetrabenzyl inositol 15 ............................................................................... 209
(d) Pentabenzyl inositol 16 .............................................................................. 210
(e) Pentabenzyl inosose 17 ............................................................................... 210
(f) myo-2-Inosose via 17 ................................................................................. 211
Synthesis of l,2,3,4—Tetrahydroxybenzene ............................................................ 211
(a) 1,2,3,4-Tetrahydroxybenzene (aromatization) ........................................... 211
(b) Tribenzyloxypyrogallol 25 ......................................................................... 212
(c) 2,3-Dibenzyloxy-1,4-benzoquinone 26 ...................................................... 212
(d) l,2,3,4-Tetrahydroxybenzene (synthesis #1) ............................................. 213
(e) 2,3,4-Tribenzyloxybenzaldehyde 27 .......................................................... 213
(f) 2,3,4-tribenzyloxyphenol 28 ...................................................................... 214
(g) l,2,3,4-Tetrahydroxybenzene (synthesis #2) ............................................. 214
Synthesis of Aurantiogliocladin ............................................................................. 215
(a) l,2,3,4-Tetramethoxybenzene 29 ............................................................... 215
(b) 2,3,4,5—Tetramethoxytoluene 40 ................................................................ 215
(c) 1,2,3,4-Tetramethoxy-5,6-dimethylbenzene 30 ......................................... 216
(d) Aurantiogliocladin ..................................................................................... 216

Synthesis of Coenzyme Q3
(a) (2’E, 6’ E)- 2- (3, 7, 11 -trimethyldodeca-2, 6, 10-trienyl)- 6- methyl- 2, 3, 4, 5-

tetramethoxybenzene 41 ............................................................................ 216
(b) Coenzyme Q 3 ............................................................................................. 217
Synthesis of 2-Deoxy- scyllo-inosose ..................................................................... 218
(a) Tetrabenzyl epoxide 42 .............................................................................. 218
(b) Tetrabenzyl alcohols 43a and 43b ............................................................. 219
(c) Tetrabenyl-2-deoxy inosose 44 .................................................................. 219
(d) 2-Deoxy- scyllo-inosose ............................................................................. 220
Aromatization of 2-Deoxy- scyllo-inosose ............................................................. 221
(a) 1,2,4-Trihydroxybenzene ........................................................................... 221
Aromatization of Triacetic Acid Lactone .............................................................. 221
(a) 4-Methoxy—6-methyl-2-pyrone 51 via dimethyl sulfate ............................. 221
(b) 4—Methoxy-6-methyl—2-pyrone 51 via addition of methanol ..................... 221
(c) 4-Methoxy-6-methyl-2-pyrone 51 via trimethyl phosphate ....................... 222
(d) Phloroglucinol methyl ether 52 .................................................................. 222
(e) Phloroglucinol ............................................................................................ 223
Deoxygenation Methodology ................................................................................. 223
(a) Resorcinol via phloroglucinol .................................................................... 223
(b) Resorcinol via phloroglucinol methyl ether 52 .......................................... 224
(c) Pyrogallol ................................................................................................... 224

(d) Hydroquinone ............................................................................................. 224

Chapter 4 ................................................................................................................ 225
Plasmid Construction ............................................................................................. 225
(a) Plasmid pCH4.l84A ................................................................................... 225
(b) Plasmid pCH4.264 ..................................................................................... 225
(c) Plasmid pCH4.267A .................................... ‘ ............................................... 2 26
(d) Plasmid pCH5.148 ..................................................................................... 226
(e) Plasmid pCH5.l74A ................................................................................... 226
(f) Plasmid pCH5.183B ................................................................................... 227
(g) Plasmid pCH5.263A ................................................................................... 227
(h) Plasmid pCH5 .213 ..................................................................................... 228
(i) Plasmid pCH5.287A .................................................................................. 228
(j) Plasmid pCH5.303 ..................................................................................... 228
Mutagenesis .......................................................................................................... 229
(a) Site-directed mutagenesis of FAS-B .......................................................... 229
(b) Homologous Recombination ...................................................................... 231
(c) Fermentations with mutated FAS—B .......................................................... 232
(d) RB791 serAzzaroB/pCH5.28M1 ................................................................. 232
(e) JWF1(DE3)/pCH5.303 ............................................................................... 234
(f) Triacetic acid lactone via YZ166 .............................................................. 234
Polyketide Biosynthesis ......................................................................................... 235
(a) Fermentation with yeast ............................................................................. 235
References .............................................................................................................. 237

xi

LIST OF TABLES

Table 1. Specific Activities for MIP synthase ................................................................... 45
Table 2. Primers for sequencing INOI ............................ A .................................................. 47
Table 3. Codon Usage ....................................................................................................... 48
Table 4. MIP synthase activities vs. tRN A expression ..................................................... 50
Table 5. Specific Activity of MIP synthase with GroES and GroEL ............................... 66
Table 6. E. coli strains for expressing inositol dehydrogenase ......................................... 73
Table 7. Specific Activities for inositol dehydrogenase expression in E. coli ................. 77
Table 8. Inositol dehydrogenase Specific Activities from fermentations ........................ 81
Table 9. Catalytic oxidations of myo-inositol .................................................................. 104
Table 10. Aromatization conditions of my0-2-inosose ................................................... 107
Table 11. Aromatization of 2-deoxy-scyllo-inosose ....................................................... 121

Table 12. Reaction of triacetic acid lactone with magnesium or sodium methoxide..124
Table 13. Reductions of polyhydroxyaromatics .............................................................. 132

Table 14. Primers for mutation of fasB. Mutations are underlined ................................. 159

xii

LIST OF FIGURES

Figure 1. The hydroxylated benzenes .................................................................................. 4
Figure 2. Industrial synthesis of catechol......................... ................................................... 5
Figure 3. Industrial synthesis of resorcinol .......................................................................... 6
Figure 4. Industrial syntheses of hydroquinone ................................................................... 7
Figure 5. Industrial synthesis of pyrogallol ......................................................................... 8
Figure 6. Syntheses of phloroglucinol ............................................................................... 10
Figure 7. Biosynthesis of L-sodium glutamate and L-lysine by genetically engineered C.
glutamicum ......................................................................................................................... 13
Figure 8. Conventional chemical synthesis of aspartame .................................................. 14
Figure 9. Enzyme catalyzed production of aspartame ....................................................... 15
Figure 10. Synthesis of acrylamide from acrylonitrile ...................................................... 16

Figure 11. Reichstein synthesis of L-ascorbic acid (steps a-f) and a fermentation bypass

(Steps g and h) ................................................................................................................... 17
Figure 12. The Shikimate pathway for aromatic amino acid biosynthesis ......................... 20
Figure 13. Synthesis of catechol and adipic acid from D-glucose ..................................... 22
Figure 14. Synthesis of adipic acid via benzene and D-glucose ....................................... 23
Figure 15. Synthesis of vanillin via benzene or D-glucose ............................................... 25
Figure 16. Synthesis of gallic acid and pyrogallol via benzene and D-glucose ................. 27

, Figure 17. Synthesis of phenol and p-hydroxybenzoic acid via intermediacy of Shikimic

acid ..................................................................................................................................... 28
Figure 18. Synthesis of hydroquinone via quinic acid ..................................................... 30
Figure 19. The inositol family of molecules ..................................................................... 39
Figure 20. Biosythesis of myo-inositol ............................................................................. 40

xiii

Figure 21.

Figure 22.
Figure 23.
Figure 24.
Figure 25.
Figure 26.
Figure 27.
Figure 28.
Figure 29.
Figure 30.
Figure 31.

Figure 32.
Figure 33.

Figure 34.

Figure 35.
IPTG .........

Figure 36.
Figure 37.

Figure 38.

IPTG .........

Figure 39.
Figure 40.

Figure 41.

Construction of pAD1.45a ............................................................................. 42

Construction of pAD1.88a ............................................................................... 43
Fermentation of JWFl/pADl.88a: PmcINOI, sent with 5 mg IPT G ............. 45
Construction of pCH6.112a ............................................................................. 52
Construction of pCH6.123a ............................................................................. 53
Fermentation of JWFl/pCH6.123a: PmCSYNINOI, serA with 5 mg IPT G ..... 54
Fermentation of JWF 1/pCH6.123a: PmcSYNINOI, serA with 10 mg IPTG...55
Fermentation of JWFl/pCH6. 123a: PmcSYNINOI, serA with 20 mg IPTG...56
Construction of pCH7.41 ................................................................................. 59
Construction of pCH7.61b ............................................................................... 60
Construction of pCH7.66a ............................................................................... 61
Fermentation of JW F l(DE3)/pCH7.66a: P171N01, sent with 1 mg IPTG....63
Fermentation of JWFl(DE3)/pAD1.88a: PmcINOI, sent with 5 mg IPTG...64
Proposed route for myo-inositol catabolism .................................................... 68
Fermentation of JWFl/pAD2.28a: PtacINOI , PlaciolG, serA with 10 mg
.......................................................................................................................... 70
Construction of pCH7 .170 .............................................................................. 74
Construction of pCH7.l94a ............................................................................. 78

Fermentation of JWFl/pCH7. 194a. PtaclNOI, PmciolG, serA with 10 mg
.......................................................................................................................... 80
Metabolic map of carbohydrate metabolism in G. oxydans ............................ 83
Construction of pCH6.254a (G. oxydans vector region in black) ................... 85
Cyclitol oxidations carried out by G. oxydans ATCC 621 .............................. 86

xiv

Figure 42. Conversion of D-mannose to neo—inositol ....................................................... 87
Figure 43. The hydrates of scyllo-2,5-diketoinositol ........................................................ 88
Figure 44. Oxidation of myo-inositol .............................................................................. 100
Figure 45. Oxidation of myo-inositol to leuconic acid ................................................... 102
Figure 46. Chemical synthesis of myo-2-inosose ........................................................... 105
Figure 47. Results of treating myo-2—inosose with acid or base ...................................... 106
Figure 48. Original synthesis of l,2,3,4-tetrahydroxybenzene ....................................... 109
Figure 49. Alternative syntheses of l,2,3,4-tetrahydroxybenzene .................................. 110
Figure 50. Derivatives of l,2,3,4-tetrahydroxybenzene displaying biological activity..l 11
Figure 51. Synthesis of aurantiogliocladin ...................................................................... 112
Figure 52. Coenzyme Q10 mode of action as an antioxidant .......................................... 113
Figure 53. Synthesis of coenzyme Q10 from vanillin ..................................................... 114
Figure 54. Synthesis of coenzyme Q10 from p-cresol .................................................... 116
Figure 55. Synthesis of coenzyme Q10 from l,2,3,4-tetrahydroxybenzene .................... 117
Figure 56. Conversion of D-glucose 6-phosphate to 2-deoxy-scyllo-inosose by the
enzyme BtrC .................................................................................................................... 118
Figure 57. Synthesis of racernic 2-deoxy-scyllo-inosose ................................................ 119
Figure 58. Aromatization of triacetic acid lactone derivative via Claisen (a) or Aldol (b)
Condensations .................................................................................................................. 122
Figure 59. Treatment of dehydroacetic acid with base ................................................... 123
Figure 60. Aromatization of triacetic acid lactone ........................................................... 125
Figure 61. Reactions of phloroglucinol .......................................................................... 127
Figure 62. Tautomers of phloroglucinol ......................................................................... 128
Figure 63. Conversion of phloroglucinol to resorcinol ................................................... 129

XV

Figure 64

. Reduction intermdiate of phloroglucinol methyl ether 52 ............................ 131

Figure 65. Conversion of D-glucose to pyrogallol via the Shikimic acid pathway or myo-
inositol biosynthesis ......................................................................................................... 135
Figure 66. Conversion of D-glucose to pyrogallol via the Shikimic acid pathway or myo-
inositol biosynthesis ......................................................................................................... 136
Figure 67. Industrial synthesis of triacetic acid lactone .................................................. 141
Figure 68. Biosynthesis of 6-methylsalicylic acid (6-MSA) .......................................... 142
Figure 69. Mutations made of wild type 6-MSAS .......................................................... 145
Figure 70. Carboxylation of acetyl-CoA by accABCD .................................................. 147
Figure 71. Three mechanisms for fatty acid biosynthesis initiation ............................... 148
Figure 72. Chain propagation in fatty acid biosynthesis ................................................. 149
Figure 73. X-ray film exposed to radioactive extracts from fatty acid synthases ........... 153
Figure 74. Amino acid sequence comparisons of B-ketoreductase sites for fatty acid
biosynthesis and polyketide biosynthesis ........................................................................ 155
Figure 75. Overlap extension PCR ................................................................................. 157
Figure 76. Construction of pCH4.267A .......................................................................... 158
Figure 77. Mutant M1 of fasB from B. ammoniagenes .................................................. 160
Figure 78. Mutant M2 of fasB from B. ammoniagenes ................................................... 161
Figure 79. Mutant M3 of fasB from B. ammoniagenes ................................................... 161
Figure 80. Construction of pCHS .213 ............................................................................. 163
Figure 81. Construction of pCH5.287A .......................................................................... 164
Figure 82. Construction of pCH5.303 ............................................................................. 165

xvi

Ac

ATP
Ap
ApR

CA
Cbz
CIAP
Cm
CmR
COMT
DAHP
DCU
DEAE
DHQ
DHS
DO
D'I'l‘
E4P
EPSP
FBR
GA

HPLC

LIST OF ABBREVIATIONS

acetyl

adenosine diphosphate

adenosine triphosphate

ampicillin

ampicillin resistance gene

base pair

chorismic acid

benzyloxycarbonyl (carbobenzoxy)
calf intestinal alkaline phosphatase
chloramphenicol

chloramphenicol resistance gene
catechol-0-methyltransferase
3-deoxy-D-arabino-heptulosonic acid 7-phosphate
digital control unit
diethylaminoethyl

3-dehydroquinic acid
3-dehydroshikimic acid

dissolved oxygen

dithiothreitol

D-erythrose 4-phosphate
5—enolpyruvoylshikimate 3-phosphate
feedback resistant

gallic acid

hour

high pressure liquid chromatography

xvii

IPT G
Kan
KanR
kb
kg

LB

M9

min
11L

11M
mRN A
NAD
NADH
NADP

NADPH
NMR
OD
ORF
PCA
PEG
PEP

pfu

isopropyl B-D-thiogalactopyranoside

kanamycin

kanamycin resistance gene.

kilobase

kilogram

Michaelis constant

luria broth

molar

minimal salts

minute

milliliter

rnicroliter

millimolar

rnicromolar

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

phosphoenolpyruvic acid

plaque forming units

xviii

PHB
PID
PCR
Phe
PMSF

psi

SAM
SDS
S3P
Spec
Tc
TCA
TMS

TSP
Tyr
UDP
UV

Vmax

yr

p-hydroxybenzoic acid
proportional-integral-derivative
polymerase chain reaction.
L-phenylalanine
phenylmethylsufonyl fluride
pounds per square inch
phosphotransferase system
quinic acid

ribosome binding site
rotations per minute
Shikimic acid
S-adenosylmethionine
sodium dodecyl sulfate
Shikimate 3-phosphate
spectinomycin

tetracycline

tricarboxylic acid
trimethylsilyl

L—tryptophan

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

uridine diphosphate
ultraviolet

maximal velocity

year

xix

Shawl

INTRODUCTION

Arguably the most important raw material obtained from fossil fuels, aromatics
are used in the production of plastics, synthetic rubber, polymeric fibers and in several
molecules displaying biological activity.1 A majority of aromatics are derived from
benzene, toluene or from the o, m, or p-xylenes obtained from the BTX cut of fossil fuel
refining.2 With an annual production of approximately 8 x 109 kg in the US,3 benzene
is the most important starting material in the manufacture of aromatic and cycloaliphatic
chemicals.4 Exposure to benzene has been linked to both acute leukemia and non-
Hodgkin’s lymphoma.5 Besides the fact that all aromatic starting materials are derived
from nonrenewable resources, benzene, toluene, and xylenes are highly volatile and
require high cost measures to reduce exposure to the environment.6 Benzene is a
hazardous organic air pollutant whose emission must be reduced as mandated by the
Chemical and Manufacturing Rule issued by the U. S. Environmental Protection
Agency.7

Perhaps the most effective way to address human health risks from exposure to
volatile aromatics is the development of pathways to aromatic products derived from
non-toxic, non-volatile and renewable starting materials. Unlocking and reconfiguring
the carbon locked in carbohydrates such as D-glucose, D-xylose and L-arabinose offers a
unique opportunity for the production of industrially relevant commodity and fine
chemicals from non-toxic and renewable resources. D-Glucose is readily available from

the depolymerization of starch8 while D-xylose and L-arabinose can be obtained in a

similar fashion from hemicellulose.9 Representing the most abundant biopolymer,
cellulose will likely replace starch as the primary source of D-glucose when improved
depolymerization technology is available.10

The work carried out for completion of this thesis focused on expanding the
spectrum of aromatic molecules available from D-glucose using biocatalysis followed by
environmentally benign chemical synthesis. Previous work in the Frost group explored
production of aromatics via intermediates of the aromatic amino acid biosynthetic
pathway in Escherichia c0li.11’12913 As a significant departure from previous results,
three alternative biocatalytic intermediates were investigated with no direct ties to
forming aromatic intermediates in nature. Chapter 2 presents the results of the
biosynthesis of my0-2-inosose via intermediacy of myo-inositol and scyllo-2,5-
diketoinositol via intermediacy of neo-inositol. The biosynthesis of my0-2-inosose was
further investigated as a two-host system versus direct conversion of D-glucose to myo-2-
inosose in a single host. Chapter 4 provides the initial results for the biosynthesis of
triacetic acid lactone via polyketide biosynthesis or fatty acid biosynthesis. Concurrent
with development of the biocatalytic routes to myo-2-inosose, scyllo-2,5-diketoinositol
and triacetic acid lactone, Chapter 3 investigates conditions to establish the conversion of
these biosynthetic intermediates into polyhydroxyaromatics by benign chemical
reactions. Also found in Chapter 3 is aromatization of chemically synthesized 2-deoxy-
scyllo-inosose. The elaboration of the available aromatics from this biosynthetic
intermediate will justify recombinant expression of 2-deoxy-scyllo-inosose synthase from
Bacillus circulans in E. coli. Convenient catalytic reduction methodology has been

developed for the selective deoxygenation of polyhydroxyaromatics formed by acid

catalyzed dehydration. The methodology further expands the number of
polyhydroxylated aromatics from myo-2-inosose, 2-deoxy-scyllo-inosose and triacetic
acid lactone. By combining biocatalysis and conventional organic synthesis, l,2,3,4-
tetrahydroxybenzene, hydroxyhydroquinone, phloroglucinol, pyrogallol, hydroquinone,
and resorcinol can now be synthesized from D-glucose via the intermediacy of my0-2-
inosose, 2-deoxy-scyllo-inosose, and triacetic acid lactone.
Hydroxylated Aromatics

The Shikimate pathway for aromatic amino acid biosynthesis has provided
convenient access to catechol,11 hydroquinone12 and pyrogallol.13 Although biocatalytic
pathways have now been elaborated for the conversion of D-glucose to catechol11 and
pyrogallol”, both routes must contend with the toxicity of the aromatic product towards
growing microbial biocatalysts. The direct conversion of D-glucose to catechol was
carried out by initial growth of the microbial biocatalyst to stationary phase followed by
resuspension and culturing in minimal salts medium where synthesis of catechol
occurred. Synthesis of pyrogallol entailed addition of microbially synthesized gallic acid
to pregrown cells. An alternative strategy is to employ microbial catalysis to synthesize
an intermediate, which is not toxic to the microbe, followed by chemical conversion of
the intermediate to the desired aromatic product. Synthesis of phenol14 and
hydroquinone12 from D—glucose provide examples of this strategy. This thesis will
explore the latter strategy involving microbial synthesis of non-toxic intermediates,
which are subsequently chemically converted to aromatics. A key distinction of this
research will be the use of biosynthetic pathways that are distinct from the Shikimate

pathway. Aromatics targeted for synthesis will include catechol, resorcinol, and

hydroquinone along with polyhydroxybenzenes, which include pyrogallol,

phloroglucinol, hydroxyhydroquinone, the tetrahydroxybenzenes, pentahydroxybenzene,

and hexahydroxybenzene (Figure 1).

HO ;
OH

catechol

HO ; OH

OH
pyrogallol

OH
HO

HO
OH

1 ,2, 3,4-tetra-
hydroxybenzene

phenol

HO : OH

resorcinol

OH

HO

3
9‘0
O
I
4?

hydroquinone

HO OH

a.

OH
1 ,2 ,3,5-tetra-
hydroxybenzene

HO OH

8.

HO
OH
pentahydroxy-
benzene

Figure l. The hydroxylated benzenes.

0°”
HO

hydroquinone

OH
HO i OH

phloroglucinol

HO OH

Ci

HO OH

1 ,2, 4, 5-tetra-
hydroxybenzene

OH
HO OH

HO OH
OH
hexahydroxy-
benzene

Catechol

With an annual production of over 2.2 x 107 kg, catechol finds use as a starting
material in the manufacture of pharmaceuticals (L-DOPA, adrenaline), flavors (vanillin,
eugenol, isoeugenol), agrochemicals (carbofuran, propoxur), and polymerization
inhibitors and antioxidants (4—tert-butylcatechol, veratrol).15 Industrial manufacture of
catechol starts with Friedel-Crafts alkylation of petroleum-derived benzene with
propylene to afford cumene (Figure 2).15 Hock oxidation of the alkyl substituent to
phenol followed by subsequent oxidation with 70% hydrogen peroxide affords a mixture
of catechol and hydroquinone. Separation of the two products is carried out by
distillation. The process for producing catechol is dependent on petroleum-derived and

carcinogenic benzene and the use of highly energetic hydrogen peroxide.

“6 a 63°“

benzene cumene phenol:> catechol

0°”
HO

hydroquinone

 

 

 

    

  

Figure 2. Industrial synthesis of catechol. Key: a) propylene, solid H3PO4 catalyst,
200-260 °C, 400-600 psi.; b) 02, 80-130 °C then 802, 60-100 °C; c) 70% H202, EDTA,
PC“ or C0”, 70-80 °C.

Resorcinol
Annual production of resorcinol reaches 3.5 x 107 kg.15b Formulations of
resorcinol with formaldehyde are used as tackifiers in the manufacture of tires and as

wood adhesives.15b Resorcinol also finds use as the active ingredient in throat lozenges

and has been used as a starting material in the synthesis of UV blockers and other
molecules displaying biological activity.15b Industrial synthesis of resorcinol is carried
out by two methods utilizing carcinogenic benzene derived from nonrenewable fossil
fuels as the starting material (Figure 3).15b’l6 The formation of the disulfate ester 1
followed by alkali fusion generates large quantities of salt solutions as a waste stream.
Resorcinol is also synthesized by Hock-type oxidation of 1,3-diisopropylbenzene, 2. The

peroxide intermediates formed during resorcinol manufacture are an explosion hazard.

H038 SO3H
81/ ‘ \b.
O ,0...

H
benzene c\* /d( resorcmol

2

Figure 3. Industrial synthesis of resorcinol. Key: a) 803, Na2SO4, 150 °C; b) NaOH,
350 °C; 0) 2-propene, HZSM-l2; d) 02, NaOH, 90-100 °C.

Hydroquinone

Hydroquinone finds use in photography development and as an intermediate in
the synthesis of antioxidants and polymerization retardants.16 Annual global production
of hydroquinone is 4.5 x 107 kg which is manufactured by several routes utilizing
benzene as a common feature (Figures 2 and 4)}5",l6 Reaction of phenol with H202 in
the presence of strong acids leads to a mixture of catechol and hydroquinone (Figure 2).

The oldest process for hydroquinone synthesis is oxidation of aniline with MnOz and

H2SO4 to form benzoquinone (Figure 4). Reduction of benzoquinone with Fe0 or by

catalytic hydrogenation affords hydroquinone. This manufacturing technique generates
large quantities of MnSO4, (NH4)2SO4, and iron oxide salts. 1513,16

Over half of the hydroquinone production is manufactured by use of Hock-type
oxidation conditional“),l6 The p-diisopropylbenzene 3 is synthesized by Friedel-Crafts
alkylation of benzene with either propylene or isopropanol (Figure 4). Air oxidation of 3
proceeds at 90-100 °C in an aqueous solution also containing organic bases along with
cobalt or copper salts. As discussed previously in the manufacture of catechol and
resorcinol, Hock-type oxidation conditions form explosive organic peroxide

intermediates which present a safety hazard.16

—-“ 0’”
HO

3 hydroquinone

(3 lb
benzene 0 ”Hz a O
O

aniline benzoquinone

// \°

Figure 4. Industrial syntheses of hydroquinone. Key: a) MnOz, H2804; b) FeO; c) 2-
propene, HZSM-l2; d) 02, NaOH, 90-100 °C.

Pyrogallol

Pyrogallol is readily available by the decarboxylation of gallic acid in copper
lined kettles under 12 atm pressure and 175 °C (Figure 5).17 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
which has led to the development of chemical routes to pyrogallol. Synthesis of
pyrogallol starts from perchlorination of benzene derived cyclohexanone (Figure 5).17
Tetrachlorinated cyclohexanone 4 is hydrolyzed with base to pyrogallol. Gallic acid and
pyrogallol find use in molecules displaying biological activity such as the antibiotic

trimethoprim, the muscle relaxant gallamine triethiodide, and the insecticide

bendiocarb. 17

 

  

COzH
HO OH HO OH
OH OH
gallic acid pyrogallol cyclohexanone benzene

 

  

Figure 5. Industrial synthesis of pyrogallol.
Key: a) 12 atm, 175 °C; b) C12; c) NaOAc.

Phloroglucinol

Phloroglucinol is used in diazodyes, as a crosslinker in the manufacture of
polymers, and as a starting material in the synthesis of pharmaceutical agents.17 The
significance of phloroglucinol as a useful intermediate is impeded by the complications
of its synthesis. Toluene-derived trinitrotoluene (TNT) is the starting material employed
in the current industrial manufacture of phloroglucinol (Figure 6).17’183 Chromium
oxidation of TNT affords the carboxylic acid 5. Reduction in HCI using FeO affords
triamino benzene 6 which is hydrolyzed in aqueous HCl to afford phloroglucinol. The
manufacture of phloroglucinol requires the disposal of highly acidic, aqueous waste

streams of chromium, iron, and ammonium salts reaching 5.8 x 104 gallons/day.18a

Phloroglucinol is no longer manufactured in the U. S. due to the explosive nature of the
TNT used as the starting material.17 Several alternative processes have been developed
to replace TNT as the starting material, although the extent to which these alternate
routes are used is difficult to ascertain (Figure 6).18

The expense for the synthesis of hydroxyhydroquinone and the remaining
polyhydroxyaromatics precludes their widespread use in chemical manufacture.17 The
orientation of oxygens in hydroxyhydroquinone is found in several molecules displaying

insecticidal activit .17 The antioxidant coenz me Q10 uards low—densit Ii 0 roteins
y Y g y P P

from atherosclerosis-related oxidative modification19 while fumagatin and
aurontiogliocladin display antibiotic activity.20 Each of these molecules is a
functionalized 1,2,3,4-tetrahydroxybenzene. Other polyhydroxyaromatics such as
1,2,3,5-tetrahydroxybenzene, 1,2,4,5-tetrahydroxybenzene, pentahydroxybenzene and
hexahydroxybenzene also display antioxidant and biological characteristics which may
become industrially significant provided cheap routes to their manufacture can be

developed. 17

O NH2

<— ——>
H2N NH2 ‘— _’

0 CH3

 

 

O o benzene
19 i i
o NHCI Br
CIHN NHCI Br/[ja
o 0 ii
in
OCH3
ii
HN NH2 H300 OCH3
° (it it 1"
HZNJLN N NH2 i n
H H \ OH /

   
   

HO OH
phloroglucinol

00Ha
° CL
——>
02N No2 H2N NH2 HO OH

 

COzH

5 6

in ie

N02

0 o H co
<———a M 3 V

OZN NO2 9' 9' CH3

CH3 CH3

TNT toluene

Figure 6. Syntheses of phloroglucinol. Key: a) HNO3, H2804; b) NazCrzO7, H2804;
c) FeO, HCl; d) HCl, H20; e) EtzO, -74 °C; f) HCl; g) HCl, C12; h) NH3; i) BC], 180 °C;
j) NaOMe, DMF, CuI; k) HCl; 1) NH20H; m) CF3C02H; n) HCI.

10

Biocatalysis

Biocatalysis employs the use of enzymes for the production of commodity,
pseudo-commodity, fine and ultra-fine chemicals either for use as building blocks in the
chemical industry or as end products. Biocatalysis embodies both one-step bioconversion
of intermediates or fermentation technology.” A bioconversion is defined by in vitro
chemical modification brought about by an enzyme or the use of a microbial host to
direct the one-step chemical modification of an intermediate added to the cell broth.”
Fermentation harnesses a single enzyme or a pathway of several enzymes in a microbial
host capable of converting a carbohydrate to a biosynthetic intermediate in an
environment under stringent control of pH, dissolved oxygen, temperature and
pressure.”

In vitro biocatalysis is carried out by enzymes. The enzymes can be used either
from crude cell extracts, partially purified, or homogeneous forms. When using soluble
protein, the expense for the catalytic process increases as the purity of the enzyme
increases due to costly purification steps.22 Recombinant technology can lower the
requirement for purification by increasing the percentage of the desired protein relative to
the total protein in crude extracts.” Immobilization of homogenous enzymes can offer
the advantage of higher enzyme loadings, prolonged lifetime of enzyme catalytic activity,
the and the ease of product purification by filtration of the immobilized enzymes from the
reaction supernatant which allows the enzyme to be reused.22 Many enzymes require the
use of expensive cofactors. Although cofactor recycling systems have been developed,”
the use of whole cells to direct biocatalytic conversions is more practical if the

intermediate is non-toxic and can be transported into the cell.22 The use of whole cells

11

offer the advantage of not requiring cell disruption for the release of the enzyme and can
harness multiple enzymes to carry out a multi-step conversion. Technology is also
available for the immobilization of cells which offers the same advantages as discussed
for immobilized enzymes.

Biocatalysis can play an important role in the ultimate incorporation of
carbohydrates as building blocks for the manufacture of organic chemicals. Exchanging
chemical synthesis with biocatalysis can also benefit the environment. As evidenced by
the industrial syntheses of catechol, resorcinol, hydroquinone and the
polyhydroxybenzenes pyrogallol and phloroglucinol, chemical reactions oftentimes use
toxic reagents, volatile organic solvents and accumulate large quantities of waste salt
streams. On the other hand, biocatalysis utilizes aqueous solutions consisting of low salt
concentrations and are oftentimes conducted at ambient temperatures and pressures while
waste streams are typically innocuous and biodegradable.

The value of fermentation technology is exemplified by the production of amino
acids. The market for amino acids is constantly growing with demand increases of 10%
each year.24 Total amino acid production doubled from 4.25 x 105 tons in 1982 to 8.0 x
105 tons in 1991.2421 The largest-volume amino acids produced are L-glutamate, L-lysine
and D,L-methionine. Used in food preparations, L-sodium glutamate is a flavor enhancer
while L-lysine and D,L-methionine are used as feed additives.24a Because poultry contain

an enzyme capable of converting D-methionine to L-methionine, the cost of making

enantiomerically pure L-methionine by fermentation is not justified.25

12

CD;—

+ -— +
Na oz,c’\)‘r\iH3

/ L-sodium glutamate

.“OH
C. glutamicum

_,_ OH
HO OH \
D-glucose

 

coz-

+W +
H3N NH3
L-lysine

Figure 7. Biosynthesis of L-sodium glutamate and L-Iysine by genetically engineered
C. glutamicum.

Annual production of L-sodium glutamate is 4.24 x 105 tons while for L-lysine is
1.52 x 105 tons.24a An L-glutamate accumulating soil bacterium Corynebacterium
glutamicum was discovered and has become the primary organism for production of L-
glutarnate and L-lysine (Figure 7). Two mutants were genetically engineered to produce
L-glutamate and L-lysine from molasses (cane or sugar-beet), sucrose, or starch
hydrosylates as the carbon sources. Microbe-synthesized L-lysine can reach a titer of 170
g/L in 48 h resulting in a 54% yield based on carbohydrate starting material. The
remainder of this section on biocatalysis will provide examples of bioconversions and
fermentations developed to directly compete or replace the use of fossil fuels and
conventional organic synthesis in the production of commodity or fine chemicals.
Aspartame

Aspartame is a low calorie sweetener used in several foods and beverages and is
claimed to be 200 times sweeter than sucrose.26 Sold under the name Nutrasweet,

aspartame is a dipeptide of L-aspartic acid and the methyl ester of L-phenylalanine.

l3

Conventional chemical coupling of L-aspartic acid and the methyl ester of L-
phenylalanine of aspartame starts with the synthesis of Cbz-L-aspartic anhydride and
enantiomerically pure L-phenylalanine methyl ester. Upon coupling and hydrogenolysis,
the a-anomer aspartame is afforded in 60-80% yield. along with formation of the B—
anomer in 20-40% yield (Figure 8).27 The undesired B-anomer imparts a bitter

component and must be separated from the a-anomer.

O O
/U\O + O

 

Ham/”x OCH3
CBZHN CY _o _ u o
o o Tl/
CBZ—L-aspartic O
anhydride 1- ft. EtOAc : aspartame
0 2. hydrogenolysis 0

O
H
H2N\/U\mH3 _OJJ\_;_/\[r N\/U\OCH3

U 33

L-phenylalanine B-anomer
methyl ester

Figure 8. Conventional chemical synthesis of aspartame.

Because aspartame is a food additive, the expense incurred for protection and
deprotection of the B carboxylic acid and the use of enantiomerically pure L-
phenylalanine methyl ester would make the final cost of aspartame unattractive to
consumers. The biocatalytic development of aspartame is a good example of an in vitro
bioconversion using thermolysin obtained from the thermotroph Bacillus
thennoproteolyticus Rokko (Figure 9).28 Immobilized thermolysin catalyzes the key
coupling step without protection of the B carboxylic acid of L-aspartic acid. Therrnolysin

also exhibits selectivity for the L enantiomer of phenylalanine methyl ester allowing the

14

methyl ester to be added as a racernic mixture and thus reducing the input cost. After
extractive removal of aspartame with ethyl acetate, leftover D-phenylalanine methyl ester

is recycled by acid catalyzed racemization.

O O

 

+
H0 H3N\£/u\u OCH3
CBZHN OH ‘ of 0
O O aspartame
+ 2. hydrogenolysis .,.
H3” OCH3 H3” OCH3

D ,L-phenylalanine D-phenylalanine

methyl ester methyl ester

1 H‘, MeOH |

 

Figure 9. Enzyme catalyzed production of aspartame.
Acrylamide

With an annual production of 4.3 x 109 kg/yr, acrylamide is an important starting
material used in polymers for flocculents, synthetic fibers, petroleum recovery, and
additives for paint.29 Industrially, acrylamide is manufactured by hydration of the nitrile
of acrylonitrile with a copper salt catalyst (Figure 10).30 The isobutyronitrile-utilizing
bacterial strain Pseudomonas chlororaphis B23 has been isolated from soil and shown to
exhibit a nitrile hydratase activity.31 Culture conditions have now been developed to
induce the nitrile hydratase activity by growing the cells in a medium containing
methacrylamide. Immobilized P. chlororaphis 323 at 10 °C cleanly hydrates

acrylonitrile to acrylamide at concentrations reaching 400 g/L by a whole-cell

15

bioconversion process. Employment of biocatalysis for the hydration of acrylonitrile to
acrylamide avoids the use of Cu2+ as catalyst and the expense of recovering and
regenerating the catalyst. The low temperature conversion avoids acrylamide
polymerization which constitutes the major byproduct of the conventional chemical

hydration of acrylonitrile.

 

 

2+
pCN CU ’ H20! A > Ar NH2
0
P. h h' B
”ON c Iororap IS 23: ATM-'2
O
acrylonitrile acrylamide

Figure 10. Synthesis of acrylamide from acrylonitrile.
L-Ascorbic Acid

L-Ascorbic acid is an antioxidant and is increasingly being used in vitamin-
fortified foods such as cereals and juices. The industrial synthesis of L-ascorbic acid
relies on the Reichstein method introduced in 1933 (Figure 11).32 The Reichstein
method starts with hydrogenation of D-glucose to D-sorbitol followed by oxidation to L-
sorbose with Gluconobacter oxydans. L-Sorbose is then protected by acetonization
followed by chemically oxidizing the remaining primary alcohol then deprotecting to
afford 2-keto-L-gulonic acid (2-KLGA). Cyclization of 2-KLGA affords L-ascorbic acid.
Although the method utilizes D-glucose as the starting material and incorporates a
selective oxidation step with G. oxydans, the current industrial synthesis of L-ascorbic

acid is inefficient due to a chemical protection/deprotection scheme.

16

OH

 

,.\OH b HO .,\OH
——->
. 5. : 0'1
HO OH OH OH
D'Q'UCOSB D-sorbitol y L-sorbose
10
OH
“OH HO
OH ' \,. O
O O
O 5 OH /(Ow O
0.. \‘<
L-sorbosone diacetone-L-sorbose
lh 1d

HO

H020
\OH 0
HO O O fHZZQOH‘i

HO OH H2O

L-ascorbic 2- ketoo -L-gulonic
acid acid

diacetone- 2-keto-
L-gulonic acid

Figure 11. Reichstein synthesis of L-ascorbic acid (Steps a-f) and a fermentation
bypass (Steps g and h). Key: a) H2, Pt/C; b) D-sorbitol dehydrogenase; c) acetone, 11*;

d) KMnO4, 'OH; e) H3O+; f) MeOH, H+; g) L-sorbose dehydrogenase; h) L-sorbosone
dehydrogenase.

The key intermediate 2-KGLA can now be made biocatalytically from D-sorbitol
by the genetically modified strain of Gluconobacter oxydans G624 known to accumulate
the Reichstein method intermediate ll.-sorbose.33 The plasmid pSDH-tufBl containing
the genes for L-sorbose dehydrogenase (SDH) and L-sorbosone dehydrogenase (SNDH)
was made to provide a pathway for L-sorbose to 2-KGLA (Figure 11). G. oxydans G624
was then chemically mutated to eliminate the activity of 2-KLGA reductase to stop the
conversion of 2-KGLA to L-iodonic acid. The G. oxydans mutant NB6939, harboring the

plasmid pSDH-tufBl, was capable of converting D-sorbitol to 2-KGLA at titers of 130

17

g/L. The direct conversion of D-glucose to 2-KGLA in a genetically engineered strain of
Erwinia herbicola was only capable of making 2-KGLA at titers of 15 g/L.32b
Aromatic Products Derived from the Shikimate Pathway

The previous examples showed how biocatalysis could be used in the form of
one-step conversions with an enzyme or by genetically engineering a microbial host to
convert a carbohydrate feedstock into a desired intermediate. The benzene-free synthesis
of aromatic molecules from carbohydrates represents an intellectual abstraction from
conventional organic synthesis used by the chemical industry. The use of non-toxic and
nonvolatile carbohydrate feedstocks avoids the problems and expense associated with the
use of benzene. Harnessing the Shikimate pathway for aromatic amino acid biosynthesis
found in bacteria and plants is a logical first step in deriving aromatics from
carbohydrates by unlocking the potential of the pathway intermediates towards formation
of aromatic compounds.

The condensation of phosphoenolpyruvate with D-erythrose 4—phosphate to form
3-deoxy-D-arabino-heptulosonic acid 7-phosphate, catalyzed by the enzyme 3-deoxy-D-
arabino-heptulosonic acid 7-phosphate synthase, is the first committed step towards the
synthesis of aromatic amino acids (Figure 12).34 3-Deoxy-D-arabino-heptulosonic acid
7-phosphate synthase is represented by the isozymes AroF, AroG and AroH. 3-
Dehydroquinate synthase (AroB) cyclizes 3-deoxy-D-arabino-heptulosonic acid 7-
phosphate to 3-dehydroquinate to result in the first carbocyclic ring. 3-Dehydroquinate
dehydratase (AroD) catalyzes the elimination of the tertiary alcohol of 3-dehydroquinate
resulting in 3-dehydroshikimic acid. Reduction of the ketone by the activity of Shikimate

dehydrodenase (AroE) results in Shikimic acid. The isozymes AroK and AroL,

18

representing Shikimate kinase I and Shikimate kinase II, respectively, catalyze the
condensation of the newly formed secondary alcohol with ATP to afford Shikimate 3-
phosphate. A second equivalent of phosphoenolpyruvate is condensed with Shikimate 3-
phosphate to afford 5-enolpyruvylshikimate 3-phosphate by the activity of 5-
enolpyruvylshikimate 3-phosphate synthase (AroA). Isomerization and elimination of
the phosphate by chorismate synthase (AroC) converts 5-enolpyruvylshikimate 3-
phosphate into chorismic acid. Chorismic acid represents the common intermediate from
which L-phenylalanine, L-tyrosine, and L-tryptophan are derived.

Taking advantage of intermediates of the Shikimate pathway requires siphoning
additional carbon through the pathway. The isozymes AroF, AroG, and AroH, catalyzing
the first committed step into the Shikimate pathway, are heavily regulated by feedback
inhibition by the aromatic amino acids L-phenylalanine, L-tyrosine, and L-
tryptophan.3‘"aaC Mutation of AroF to become feedback resistant (AroFFBR) and
introducing the mutated gene on a multicopy plasmid fulfilled the requirement of
siphoning additional carbon into the Shikimate pathway.35 The remainder of this section
will discuss the latest advances made in the overproduction of 3-dehydroshikimic acid,
Shikimic acid, and the Shikimic acid pathway side product quinic acid. In turn, each of
these intermediates have shown value in the production of industrially relevant

chemicals.

l9

       

 

 

 

 

 

OH COZH
HO co H
.txOH I"'2C)3PO’§ ' 2
—> PEP _.a
s OH OH O ; OH
HO OH H203PO H H203PO OH
D-glucose DAHP\E
Hg c0214
<—-
o 5 OH
OH
DHQ
1d
Coat-I HqcozH
H203P0‘“ , Ho“'©‘ OH
OH OH
$3P A

 

 

(it

a o COZH
OH
chorismic acid

/1\

L-phenylalanine L-tryptophan
L-tyrosine

Figure 12. The Shikimate pathway for aromatic amino acid biosynthesis.

Key: Abbreviations: PEP, phosphoenolpyruvate; E4P, D-erythrose 4-phosphate; DAHP,
3-deoxy-D-arabino-heptulosonate 7-phosphate; DHQ, 3-dehydroquinate; DHS, 3-
dehydroshikimic acid; SA, Shikimic acid; S3P, Shikimate 3-phosphate; EPSP, 5-
enolpyruvylshikimate 3-phosphate; QA, quinic acid. Enzymes: a. 3-deoxy-D-arabino-
heptulosonate 7-phosphate synthase (AroF, AroG, AroH); b. 3-dehydroquinate synthase
(AroB); c. 3-dehydroquinate dehydratase (AroD); d. Shikimate dehydrogenase (AroE); e.
Shikimate kinase I (AroK), Shikimate kinase II (AroL); f. 5-enolpyruvylshikimate 3-

phosphate (AroA); g. chorismate synthase (AroC).

20

3-Dehydroshikimic acid

3-Dehydroshikimic acid is a powerful antioxidant.36 The aromatization of 3-
dehydroshikimic acid to protocatechuic acid is crucial to the value of 3-dehydroshikimic
acid as an intermediate towards the synthesis of several commodity and fine chemicals.
Protocatechuic acid is a branch point in the biosynthesis of catechol, cis,cis—muconic
acid, vanillin, gallic acid and pyrogallol. Hydrogenation of cis,cis-muconic acid affords
adipic acid. The E. coli strain KL3, containing a mutation of aroE, was incapable of
converting 3-dehydroshikimic acid to Shikimic acid.35 A serAzzaroB cassette was
introduced by homologous recombination to provide a second genomic copy of aroB to
prevent the accumulation of 3-deoxy-D-arabino-heptulosonic acid in the culture
supernatant and to remove native L-serine biosynthesis. Enhancing the production of 3-
dehydroshikimic acid required transforming KL3 with the multicopy plasmid pKL4.130B
containing PamparoFFBR, the locus serA to reestablish serine biosynthesis in KL3 and
for plasmid maintenance by nutritional pressure, and a copy of tktA37 to increase the
specific activity of transketolase. Because transketolase was found to increase the in vivo
availability of E4P, AroFFBR was capable of condensing PEP with E4P resulting in
increasing the carbon flow into the Shikimic acid pathway. Under fed-batch fermentor
conditions with D-glucose limitation developed by Konstantinov,38 a titer of 69 g/L 3-
dehydroshikimic acid was achieved with a 30% (monol) yield from D-glucose.
Catechol

As an alternative to benzene as a starting material, D-glucose can be converted in
an aroE auxotroph to 3-dehydroshikimic acid by plasmid-based expression of AroFFBR,

AroB, and TktA. A second plasmid incorporates a copy of aroZ and aroY isolated from a

21

genomic library of Klebsiella pneumoniae (Figure 13).11 3-Dehydroshikimic acid
dehydratase is expressed by aroZ and is responsible for the aromatization of 3-
dehydroshikimic acid to protocatechuic acid while aroY expresses protocatechuic acid
decarboxylase for the conversion of protocatechuic acid to catechol. Currently,
biocatalytic production with the construct AB2834/pKDl36/pKD9.069A affords catechol
at titers of 2 g/L after resuspending mature cells in minimal salts with D-glucose. The
production of catechol represents a 33% yield from D-glucose and allows catechol to be

ultimately derived by a renewable, nontoxic resource under environmentally benign

 

  

 

conditions.
OH COZH COzH OH
. fl 3 a 6°“
—> —> —>
5 OH 0 5 OH Ho
HO OH OH OH catGChO'
D-glucose 3-dehydroshikim'c protocatechuic

acid acid

Figure 13. Synthesis of catechol and adipic acid from D-glucose.
Key: a) E. coli AB2834/pKD136/pKD9.096A, 37 °C.

Adipic Acid

Polymerization of adipic acid with 1,6-hexanediamine affords the useful synthetic
fiber nylon-6,639 The global demand for adipic acid exceeds 1.9 x 109 kg/yr.39
Manufacture of adipic acid starts with hydrogenation of benzene to cyclohexane followed
by air oxidation to cyclohexanol and cyclohexanone (Figure 14).39~40 Exhaustive
oxidation of the cyclohexyl intermediates with nitric acid affords adipic acid. Adipic acid

production accounts for 10% of the atmospheric nitrous oxide levels which contributes to

22

ozone depletion and global warming."'1 The process also requires high temperatures and

pressures adding to the danger involved with its production.

 

 

  

COzH 002H OH
“°” . fl . ‘ 3 6°”
-——> -——> ——>
. OH o 5 OH HO
HO OH OH OH
D-glucose 3-dehydroshikimic protocatechuic catechol
acid acid
OH la
HOZC
c,d cyclohexanol e b / COZH
—> —> +—
0 H020 /
benzene COZH cis,cis-muconic
adipic acid acnd

 

  

cyclohexanone
Figure 14. Synthesis of adipic acid via benzene and D-glucose.
Key: a) E. coli WNl/pWN2.248, 37 °C; b) 10% Pt/C, H2, 50 psi, rt; c) Ni-A1203, H2,
370-800 psi, 150-250 °C; (1) Co, 02, 120-140 psi., 150-160 °C; e) Cu, NH4VO3, 60%
HNO3, 60-80 °C.

The enzyme catechol l,2-dioxygenase carries out the oxidative ring cleavage of
catechol to afford cis,cis-muconic acid and was obtained by isolating catA from
Acinetobacter calcoaceticus.42 The E. coli aroE auxotroph WN l was made by
homologous recombination of an aroBaroZ cassette into the genomic copy of serA and
by homologous recombination of a tktAaroZ cassette into the genomic lacZ locus.43 The
genomic insertions of aroB and tktA, when coupled with a plasmid copy of AroFFBR, aid
in channeling carbon towards production of 3-dehydroshikimic acid. The double copy of

aroZ-encoded 3-dehydroshikimate dehydratase converts 3-dehydroshikimic acid to

protocatechuic acid. The plasmid pWN2.248 carried Pam}: aroFFBR, PmccatA, aroY, and

23

serA inserts.43 The construct WNl/pWN2.248 afforded 36.8 g/L cis,cis-muconic acid

from D-glucose in 23% yield (mol/mol) (Figure 14).“3 Hydrogenation of clarified cell

broth with 10% Pt/C (5% mol/mol) and 50 psi H2 afforded adipic acid in 97% yield from

cis,cis-muconic acid.43’44 The coupling of biocatalysis and anienvironmentally benign,
catalytic reaction allows adipic acid to be produced without formation of large amounts
of nitrous oxide.

Vanillin

Vanillin is a flavoring used in the food and beverage industry and also finds use in
formulations for perfumes."'5 Natural vanillin obtained from the vanilla beans of the
orchid Vanilla planifolia accounts for only 2 x 104 kg/yr of the worlds 1.2 x 107 kg/yr
demand for vanillin.46 The difference is made up by synthetic vanillin starting from
petroleum-derived guaiacol.“5 Condensation of guaiacolqwith glyoxylic acid affords
mandelic acid (Figure 15). Oxidation of mandelic acid and subsequent decarboxylation
results in vanillin.

Synthetic vanillin sells for $12/kg while natural vanilla flavoring extracted from
vanilla bean containing 2% vanillin sells for $30 - 120/kg.“7 The high price for natural
vanilla flavoring reflects the labor-intensive cultivation, pollination, harvesting and
curing of vanilla beans. The demand for natural flavorings has, in turn, prompted the

development of biocatalytic routes to vanillin.

24

002H
Hsco’ WE‘L HSCO’ :9 C

 

 

 

 

benzene
mandelic guaiacol
acid
lb
COCOZH CHO COZH
H3CO H3CO H300
OH OH
phenylglyoxilic vanillin vanillic
acid acid
id
CO2 H
HO OH
o—glucose 3-dehydcioshikimic protocatHechuic
acid acid

Figure 15. Synthesis of vanillin via benzene or D-glucose. Key: a) HCOCOzH; b) 02;
c) H+; d) KL7/pKL5.26A; e) N. crassa aryl aldehyde dehydrogenase.

Biocatalytic conversion of D-glucose to vanillin once again passes through the
intermediacy of 3-dehydroshikimic acid. The E. coli aroE auxotroph KL7 was developed
by homologous recombination of an aroBaroZ cassette into the locus of serA":8 The
plasmid pKL5.26A was constructed with copies of Pomp aroFFBR, Ptac COMT, and
serA. The enzyme catechol-0-methyltransferase (COMT) catalyzes the methylation of
protocatechuic acid to vanillic acid and isovanillic acid. The construct KL7/pKL5.26A
afforded 4.9 g/L vanillic acid by fed-batch fermentation from D-glucose when the

construct was supplemented with L-methionine (Figure 15). The in vitro reduction of

25

vanillic acid to vanillin was carried out by aryl aldehyde dehydrogenase purified from the
fungus Neurospora crassa.49 The two-step biocatalytic synthesis of vanillin from D-
glucose falls under the guidelines of natural vanillin. Although the process is low-
yielding and inefficient, this route is the only biocatalytic synthesis of vanillin using a
carbohydrate as a starting material.
Gallic Acid and pyrogallol

Gallic acid can be made either by chemical oxidation of 3-dehydroshikimic acid
or by fed-batch fermentation from D-glucose (Figure 16). In the chemical oxidation,
treatment of microbially synthesized 3—dehydroshikimic acid with Cu2+ in the presence
of Zn2+ affords gallic acid.50 The fed-batch fermentation required the expression of a
mutant of p-hydroxybenzoate hydroxylase, encoded by pobA”‘51 from Pseudomonas
aeruginosa, to direct the hydroxylation of protocatechuic acid. The construct
KL7/pSK6.161, harboring a genomic copy of aroZ and a plasmid-localized copy of
p0bA* in a 3-dehydroshikimic acid producer, afforded gallic acid in titers of 20 g/L from
D-glucose.52 Decarboxylation of gallic acid to pyrogallol was conducted by the construct
E. coli RB791serA::aroB/pSK6.234 expressing a copy of aroY-encoded protocatechuic
acid decarboxylase. Addition of gallic acid to a batch culture of E. coli
RB791serAzzaroB/pSK6.234 at stationary phase and at 33 °C and neutral pH afforded
pyrogallol in 97% yield. The toxicity of pyrogallol towards growing E. coli cells
precluded the direct synthesis of pyrogallol from D-glucose using a single microbial

CODSII’UCI.

26

OH co,H

COZH
—-»a £1 —-~a
OH O OH HO

OH

HO OH OH
D-glucose 3-dehydroshikimic protocatechuic
acid acid

b\ /a

   

 

 

   

   
    

HO ; OH

OH
pyrogallol

HO OH
OH
gallic acid

 

 

Figure 16. Synthesis of gallic acid and pyrogallol via benzene and D-glucose.
Key: a) E. coli KL7/pSK6.161; b) Cu”, Zn”, AcOH; c) E. coli
RB79lserAzzaroB/pSK6234.
Shikimic Acid, Phenol and p-Hydroxybenzoic Acid

The current source of shikimic acid is by extraction from the fruit of the Illicium
plant at a cost of approximately $10,000/kg.53 The expense of shikimic acid makes it
unattractive for the synthesis of the neuraminidase inhibitor GS-41014, which is used to
treat influenza infections and sold under the name TamifluTM.54 Like the high-yielding,
high-titer synthesis of 3-dehydroshikimic acid, synthesis of shikimic acid required
channeling additional carbon flow through the shikimic acid pathway (see Figure 12).
The E. coli host SP1.1 was created by successive P1 phage-mediated transductions to
interrupt AroL and AroK by insertion of tetracycline resistance and chloramphenicol,
respectively, into each gene thus stopping aromatic amino acid biosynthesis at shikimic
acid.55 The serA locus was interrupted with the insertion of an additional genomic copy

of aroB to prevent accumulation of 3-deoxy-D-arabino-heptulosonic acid in the culture

supernatant. The plasmid pKDlZ.138 containing Pam]: aroFFBR, Pmc aroE, tktA and a

27

copy of the serA locus was transformed into SP1.1. The construct SP1.1/pKD12.138
synthesized 52 g/L shikimic acid under fed-batch fermentation conditions with a yield of

18% from D-glucose. Running the fermentation under glucose-rich conditions was vital

OOH

phenol

to inhibit formation of quinic acid.

OH COzH COzH

“OH
—-—a—->
5 OH Ho“ OH
HO OH

D-glucose shIkiHmic \p-hydroxybenzoic
acid acid

  

CO2 Et

3C)“. :0 - IINI"'|2°I"13F’O4

NHAc
GS- 41014

Figure 17. Synthesis of phenol and p-hydroxybenzoic acid via intermediacy of
shikimic acid. Key: a) E. coli SP1.1/pKD12.138; b) H20, 350 °C; c) 1 M H2SO4 in
AcOH.

As seen in the chemical synthesis of catechol, phenol is derived from benzene by
alkylation followed by Hock oxidation (see Figure 2).56 Approximately 20% of the
global production of benzene is used to make 5 x 109 kg of phenol each year.56 Reaction
of shikimic acid with near-critical water at 350 °C for 30 min afforded phenol in 45%
isolated yield (Figure 17).14 Heating shikimic acid at 120 °C in 1 M H2804 in acetic

acid at atmospheric pressure resulted in p-hydroxybenzoic acid in 57% yield.14 As the

yield and titer of shikimic acid synthesized from D-glucose continue to improve, the

28

hydroaromatic may become an example of disposable chirality. Not only can shikimic
acid be valued as a chiral synthon, but conditions are now available for its aromatization
to commodity and fine chemicals.
Quinic Acid and Hydroquinone

Quinic acid, which is isolated from Cinchona bark, is another example of a
molecule derived from an undependable source.57 Quinic acid results from reduction of
3-dehydroquinic acid. Mutating the serA locus of the aroD auxotroph AB2848 by
homologous recombination with the serAzzaroB cassette provided E. coli strain QP1.1.12
The construct QP1.1/pKD.138 under D-glucose limited, fed-batch fermentation
conditions synthesized 49 g/L quinic acid from D-glucose in 20% (monol) yield (Figure
18).12 As opposed to reduction of 3-dehydroshikimic acid in route to shikimic acid,
aroB-encoded Shikimate dehydrogenase carried by pKD1.138 catalyzes the reduction of
3-dehydroquinic acid to quinic acid.

After removal of ammonium ions by passing clarified fermentor broth through a
strong cation-exchange resin (Dowex-H+), quinic acid can be oxidized by household
bleach followed by heating at reflux to afford hydroquinone in 87% isolated yield by way
of intermediate 7 (Figure 18).12 Reaction with catalytic amounts of Ag3PO4 (10 mol%)

and K28208 as the co-oxidant at 50 °C followed by heating to reflux provides a chlorine-

free method for converting quinic acid to hydroquinone in 85% yield. 12

29

 

HO OH OH d
D-glucose quinic \ 0“
acid

OH
hydroquinone

Figure 18. Synthesis of hydroquinone via quinic acid.
Key: a) E. coli QP1.1/pKDlZ.138; b) i. NaOCl, ii. isopropanol, W“; C) A; d) i. Ag3PO4
(10 mol %), K2S203, 50 °C, ii.reflux.

Harnessing the intermediates of the Shikimate pathway for aromatic amino acid
biosynthesis provides access to phenol, catechol, hydroquinone and the
polyhydroxybenzene pyrogallol all derived from D-glucose as an alternative feedstock to
benzene. The aromatics of interest also include resorcinol, phloroglucinol,
hydroxyhydroquinone, the tetrahydroxybenzenes, pentahydroxybenzene and
hexahydroxybenzene. The limitation of the available aromatics from the Shikimate
pathway intermediates necessitates exploration of alternative biological pathways. As a
significant departure from the Shikimate pathway, this thesis will explore myo-inositol
biosynthesis and the biosynthesis of triacetic acid lactone from fatty acid biosynthesis and
polyketide biosynthesis as biosynthetic intermediates derived from D-glucose to be used
towards the synthesis of aromatics. Reaction conditions will be presented to direct

aromatization of myo-inositol via the intermediacy of my0-2-inosose, triacetic acid

lactone via its methyl ether and 2-deoxy-scyllo-inosose. Although the biosynthesis of

30

triacetic acid lactone is still under development, this thesis establishes the connectivity of
D-glucose to resorcinol, hydroquinone, and the polyhydroxyaromatics pyrogallol,
phloroglucinol, hydroxyhydroquinone and 1,2,3,4-tetrahydroxybenzene by coupling
biosynthetic pathways not associated with aromatics in nature with conventional

chemical synthesis.

31

References

1 Weissermel, K.; Arpe, H.-J. In Industrial Organic Chemistry, 3rd ed.; VCH: New
York, 1997: p. 310.

2 Weissermel, K.; Arpe, H.-J. In Industrial Organic Chemistry, 3rd ed.; VCH: New
York, 1997: p. 312-314.

3 Chem. Eng. News 2000, 78 (26), 51.

4 Weissermel, K.; Arpe, H.-J. In Industrial Organic Chemistry, 3rd ed.; VCH: New
York, 1997: p. 335.

5 (a) O’ Connor, 8. R.; Farmer, P. B.; Lauder, I. J. Pathol. 1999, I89, 448. (b)
Farris, G. M.; Everitt, J. I.; Irons, R.; Popp, J-. A. Fundam. Appl. Toxicol. 1993, 20, 503.
(c) Huff, J. E.; Haseman, J. K.; DeMarini, D. M. Environ. Health Perspect. 1989, 82,
125.

6 Trade Secrets, A Moyer Report; http://www.pbs.org/tradesecrets/transcript.html.

7 (a) Noah, T. Wall Street J. 1994, B2 (March 2); (b) Washington Post 1994, A3
(March 2); (c) Ember, L. Chem. Eng. News 1994, 72, 4 (March 7).

8 Com: Chemistry and Technology; Watson, S. A.; Ramstad, P. E., Eds.; American
Association of Cereal Chemists: St. Paul, MN, 1987.

9 Carlson, T. “Industrial utilization of corn starch and corn fiber for increased
ethanol yield” Abstracts of Papers, Corn Utilization Conference V, St. Louis, MO,
National Corn Growers Association.

10 (a) Bozell, J. J .; Landucci, R. Alternate Feedstocks Program Technical and
Economic Assessment; U. S. Department of Energy, Office of Industrial Technologies,
1993. (b) Lynd, L. R.; Cushman, J. H.; Nichols, R. J .; Wyman, C. E. Science 1991, 251,
1318.

‘1 Draths, K. M.; Frost, J. W. J. Am. Chem. Soc. 1995, 117, 2395.

12 Ran, N.; Knop, D. R.; Draths, K. M.; Frost, J.W. J. Am. Chem. Soc. 2001, 123,
10927.

‘3 Kambourakis, S.; Draths, K. M.; Frost, J. W. J. Am. Chem. Soc. 2000, 122, 9042.

14 Gibson, J. M.; Thomas, P. 8.; Thomas, J. D.; Barker, J. L.; Chandran, S. S.;
Harrup, M. K.; Frost, J. W. Angew. Chem, Int. Ed. 2001, 40, 1945-1948.

32

15 (a) Franck, H.-G.; Stadelhofer, J. W. In Industrial Aromatic Chemistry; Springer-
Verlag: New York, 1988; pp. 183-190. (b) Varagnat, J. In Kirk-Othmer Encyclopedia of
Chemical Technology, 3rd ed.; Grayson, M. Ed.; Wiley: New York, 1981; Vol. 13, pp.
39-69. (c) Szmant, H. H. Organic Building Blocks of the Chemical Industry; Wiley: New
York, 1989; pp. 512-519.

16 Krumenacker, L.; Costantini, M.; Pontal, P.; Sentenac, J. In Kirk-Othmer
Encyclopedia of Chemical Technology; Kroschwitz, J. I.; Howe-Grant, M., Eds.; Wiley:
New York, 1995; Vol. 13, p. 996.

17 Leston, G. In Kirk-Othmer Encyclopedia of Chemical Technology; Kroschwitz, J.
I.; Howe-Grant, M., Eds; Wiley: New York, 1996; Vol. 19, p. 778-812.

18 (a) McKillop, A.; Howarth, B. D.; Kobylecki, R. J. Synth. Commun. 1974, 4, 35.
(b) Toyonaka, T. C.; Ibaraki, M. N.; Otsma, S. 1., US Patent 4,463,199; 1984. (c)
Kleinwallstadt, H.-G. Z.; Erlenbach, M. B., US. Patent 4,057,588; 1977. (d)
Kleinwallstadt, H.-G. Z.; Erlenbach, M. 3., US. Patent 4,115,451; 1977. (e) Effenberger,
F.; Schonwalder, K.-H. Chem. Ber. 1974, 117, 3270.

19 (a) Ingold, K. U.; Bowry, V. W.; Stocker, R.; Walling, C. Proc. Natl. Acad. Sci.
U.S.A. 1993, 90, 45. (b) Stocker, R.; Bowry, V. W.; Frei, B. Proc. Natl. Acad. Sci. U.S.A.
1991, 88, 1646. (c) Steinberg, D. Circulation 1991, 84, 1420.

20 (a) Vischer, E. B. J. Chem. Soc. 1953, 815. (b) Baker, W.; McOmie, J. F. W.;
Miles, D. J. Chem. Soc. 1953, 820. (c) Baker, W.; Raistrick, H. J. Chem. Soc. 1941, 670.

” Held, M.; Schmid, A.; van Beilen, .l. B.; Witholt, B. Pure Appl. Chem. 2000, 72,
1337.

22 Jenkins, R. O. In Biotechnological Innovations in Chemical Synthesis; van Dam- '
Mieras, M. C. E., de Jeu, W. H., de Vries, J., Currell, B. R., James, J. W., Leach, C. K.,
Patmare, R. A., Eds; Butterworth Heinemann: Oxford 1997: pp. 9-34.

23 Wandrey, C.; Liese, A.; Kihumbu, D. Org. Proc. Res. Dev. 2000, 4, 286.

24 (a) Eggling, L.; Sahm, H. Appl. Microbiol. Biotechnol. 1999, 52, 146. (b)
Hodgson, J. Biotechnology 1994, 12, 152.

25 Sinskey, A. J. In Organic Chemicals from Biomass; Wise, D. L. Ed.; The
Benjamin/Cummings: Menlo Park 1983: p. 48.

26 Mazur, R. H.; Schlatter, J. M.; Goldkamp, A. H. J. Am. Chem. Soc. 1969, 91,
2684.

33

27 (a) U. S. Patent 3,786,039, 1974. (b) U. S. Patent 3,833,553, 1974. (c) U. S.
Patent 3,933,781, 1976.

28 Oyama, K.; Irino, S.; Hagi, N. Methods Enzymol. 1987, 136, 503.

29 (a) Weissermel, K.; Arpe, H.-J. In Industrial Organic Chemistry, 3rd ed.; VCH:
New York, 1997: p. 302. (b) Weissermel, K.; Arpe, H.-J. In Industrial Organic
Chemistry, 3rd ed.; VCH: New York, 1997: p. 310.

30 Weissermel, K.; Arpe, H.-J. Industrial Organic Chemistry, 3rd ed.; VCH: New
York, 1997: p. 310.

31 (a) Yamada, H.; Ryuno, K.; Nagasawa, T.; Enomoto, K.; Watanabe, I. Agric. Biol.
Chem. 1986, 50, 2859. (b) Nagasawa, T.; Ryuno, K.; Yamada, H. Experientia 1989, 45,
1066.

32 (a) Reichstein, T.; Grussner, A.; Oppenauer, R. Helv. Chim. Acta 1933, I6, 1019.
(b) Lazarus, R.; Seymour, J. L.; Stafford, R. K.; Dennis, M. S.; Lazarus, M. G.; Marks, C.
B.; Anderson, S. In Biocatalysis; Abramowicz, D. Ed.; Van Nostrand Reinhold: New
York, 1990: p. 135-155.

33 (a) Saito, Y.; Ishii, Y.; Hayashi, H.; Imao, Y.; Akashi, T.; Yoshikawa, K.;
Noguchi, Y.; Soeda, S.; Yoshida, M.; Niwa, M.; Hosoda, J .; Shimoura, K. Appl. Environ.
Microbiol. 1997, 63, 454. (b) Saito, Y.; Ishii, Y.; Hayashi, H.; Yoshikawa, K.; Noguchi,
Y.; Yoshida, S.; Soeda, S.; Yoshida, M. Biotechnol. Bioeng. 1998, 58, 309.

34 (a) Haslam, E. In The Shikimate Pathway; Wiley: New York, 1974. (b) Bently, R.
Crit. Rev. Biochem. Mol. Biol. 1990, 25, 307. (c) Herrman, K. M. In Amino Acids:
Biosynthesis and Genetic Regulation; Hermann, K. M., Somerville, R. L., Ed.; Addison-
Wesley: Reading, 1983: p. 301.

35 Li, K.; Mikola, M. R.; Draths, K. M.; Worden, R. M.; Frost, J. W. Biotechnol.
Bioeng. 1999, 64, 61.

36 Richman, J. E.; Chang, Y.-C.; Kambourakis, S.; Draths, K. M.; Almy, E.; Snell,
K. D.; Strasburg, G. M.; Frost, J. W. J. Am. Chem. Soc. 1996, 118, 11587.

37 Draths, K. M.; Pompliano, D. L.; Conley, D. L.; Frost, J. W.; Berry, A.; Disbrow,
G. L.; Staversky, R. J.; Lievense, J. C. J. Am. Chem. Soc. 1992, 114, 3956.

33 (a) Konstantinov, K. B.; Nishio, N.; Seki, T.; Yoshida, T. J. Ferment. Bioeng.

1991, 71, 350. (b) Konstantinov, K. B.; Nishio, N.; Yoshida, T. J. Ferment. Bioeng. 1990,
70, 253.

34

39 Davis, D. D.; Kemp, D. R. In Kirk-Othmer Encyclopedia of Chemical
Technology, 4th ed.; Kroschwitz, J. 1.; Howe-Grant, M., Eds.; Wiley: New York, 1991;
Vol. 1, pp. 466-493.

40 (a) Franck, H.-G.; Stadelhofer, J. W. In Industrial Aromatic Chemistry; Springer-
Verlag: New York, 1988. (b) Szmant, H. H. In Organic Building Blocks of the Chemical
Industry; Wiley: New York, 1989.

41 (a) Thiemens, M. H.; Trogler, W. C. Science 1991, 251 , 932. (b) Dickinson, R. E.;
Cicerone, R. J. Nature 1986, 319, 109.

42 Neidle, E. L.; Ornston, L. N. J. Bacteriol. 1986, 168, 815.

43 Niu, W.; Draths, K. M.; Frost, J. W. J. Am. Chem. Soc., submitted.

44 Draths, K. M.; Frost, J. W. J. Am. Chem. Soc. 1994, 116, 399.

45 Esposito, L.; Formanek, K.; Kientz, G.; Mauger, F.; Maureux, V.; Robert, G.;
Truchet, F. In Kirk-Othmer Encyclopedia of Chemical Technology, 4th ed.; Kroschwitz,
J. 1., Howe-Grant, M., Eds.; Wiley: New York, 1997; Vol. 24, p. 812.

46 Clark, G. S. Perfum. Flavor. 1990, I5, 45.

47 Priefert, H.; Rabenhorst, J.; Steinbuchel, A. Appl. Microbiol. Biotechnol. 2001,
56, 296.

48 Li, K.; Frost, J. W. J. Am. Chem. Soc. 1998, 120, 10545.

49 (a) Gross, G. G.; Bolkart, K. H.; Zenk, M. H. Biochem. Biophys. Res. Commun.
1968, 32, 173. (b) Gross, G. G.; Zenk, M. H. Eur. J. Biochem. 1969, 8, 413. (0) Gross, G.
G. Eur. J. Biochem. 1972, 11, 585. (d) Zenk, M. H.; Gross, G. G. Recent. Adv.
Phytochem. 1972, 4, 87.

50 Kambourakis, 8.; Frost, J. W. J. Org. Chem. 2000, 65, 6904.

51 (a) Entsch, B.; Palfey, B. A.; Ballou, D. P.; Massey, V. J. Biol. Chem, 1991, 266,
17341. (b) Eschrich, K.; van der Bolt, F. J. T.; de Kok, A.; van Berkel, W. J. H. Eur. J.
Biochem. 1993, 216, 137.

52 Kambourakis, S.; Draths, K. M.; Frost, J. W. J. Am. Chem. Soc. 2000, 122, 9042.

53 Haslam, E. In Shikimic Acid: Metabolism and Metabolites; Wiley: New York,
1993;p.40.

35

54 (a) Federspiel, M.; Fischer, R.; Henning, M.; Mair, H.-J.; Oberhauser, T.;
Rimmler, G.; Albiez, T.; Bruhin, J .; Esterrnann, H.; Gandert, C.; Gockel, V.; Gotzo, S.;
Hoffmann, U.; Huber, G.; Janatsch, G.; Lauper, S.; Odette, T.-S.; Trussardi, R.; Zwahlen,
A. G. Org. Process Res. Dev. 1999, 3, 266. (b) Kim, C. U.; Lew, W.; Williams, M. A.;
Liu, H.; Zhang, L.; Swaminathan, S.; Bischofberger, N.; Chen, M. S.; Mendel, D. 8.; Tai,
C. Y.; Laver, W. G.; Stevens, R. C. J. Am. Chem. Soc. 1997, 119, 681. (c) Rohloff, J. C.;
Kent, K. M.; Postich, M. J.; Becker, M. W.; Chapman, H. H.; Kelly, D. E.; Lew, W.;
Louie, M. S.; McGee, L. R.; Prisbe, E. J.; Schultze, L. M.; Yu, R. H.; Zhang, L. J. Org.
Chem. 1998, 63, 4545.

55 (a) Draths, K. M.; Knop, D. R.; Frost, J. W. J. Am. Chem. Soc. 1999, 121, 1603.
(b) Knop, D. R.; Draths, K. M.; Chandran, S. S.; Barker, J. L.; von Daeniken, R.; Weber,
W.; Frost, J. W. J. Am. Chem. Soc. 2001, 123, 10173.

56 Wallace, J. In Kirk-Othmer Encyclopedia of Chemical Technology, 4th ed., Vol.
18; Kroschwitz, J. I.; Howe-Grant, M., eds.; Wiley: New York, 1992, pp. 592-602. (b)
Franck, H.-G.; Stadelhofer, J. W. In Industrial Aromatic Chemistry; Springer-Verlag:
New York, 1988:1313. 146-183.

57 Haslam, E. In Shikimic Acid: Metabolism and Metabolites; Wiley & Sons: New
York, 1993; p. 56.

36

h r2

BIOSYNTHESIS OF M YO-INOSITOL, M YO-2-INOSOSE AND
S C YLLO-2,5-DIKETOINOSITOL

Introduction

The abundance of aromatics in nature can be tied to three primary sources. The
microbial and plant kingdom harness the Shikimate pathway for providing their own
tyrosine, tryptophan, and phenylalanine.1 Prevalent amongst fungi, but also found in a
limited number of bacteria, polyketide biosynthesis couples equivalents of acetate and
malonate to make a diverse family of polyketides oftentimes containing aromatic
functionality for use in biological warfare against environmental competition.2
Isoprenoid biosynthesis couples isoprene units to make a class of molecules associated
with cell wall biosynthesis, antioxidants and electron transfer agents used for cell
signaling and respiration.3

The available enzyme pathways pose the greatest limitation for the bioconversion
of carbohydrates to aromatic building blocks for the chemical industry. By combining
biocatalysis with chemical conversions, enzyme pathways not associated with the
biosynthesis of aromatics can, in theory, be used to synthesize aromatics from
carbohydrates. Inositols and inososes are uniquely situated in their chemistry and
biosynthesis to serve as intermediates linking carbohydrate starting materials and
aromatic products. Inositol biosynthesis and catabolism constitute a significant departure
from established routes to aromatics in nature.

The inositols are distinguished as having the highest density of chiral centers

available in nature. Of the nine members of the inositol family, two cyclitols are

37

optically active and the other seven are meso-forms (Figure 19).4 Scherer is credited with
discovering the first inositol in 1850 when myo-inositol was isolated from muscle tissue
extracts.5 Eukaryotes incorporate myo-inositol in the phospholipids of cell tissue.6 The
constant synthesis and metabolism of inositol phospholipids is part of the complex signal
transmission system used by neurotransmitters, hormones, and growth factors by
signaling the flux of Ca2+ across cell membranes. Plants, especially cereal grains, use
myo-inositol as the hexaphosphate, phytic acid, for the storage of phosphate in seeds.7
Plants also accumulate myo-inositol and various methylated inositols in response to
osmotic stress and drought.8

This chapter will discuss the bioconversion of D-glucose to myo-inositol and myo-
2-inosose in Escherichia coli. The first section will focus on the microbial synthesis of
myo-inositol by fed-batch fermentation. While carrying out this work, the goal of
maximizing myo-inositol titer was addressed by attempting to increase the specific
activity of myo-inositol 1-phosphate synthase expressed from the INOI locus of
Saccharomyces cerevisiae in E. coli. Improving the specific activity of L-myo-inositol 1-
phosphate synthase was the first goal and included consideration of codon usage,
promoter strength and protein folding. Oxidation of myo-inositol to myo-2-inosose is
then examined. Oxidation of myo-inositol with Gluconobacter oxydans ATCC 621 in a
two-host system, heterologous expression of INOI-encoded MIP synthase from S.
cerevisiae in G. oxydans ATCC 621 or heterologous expression of iolG-encoded inositol
dehydrogenase from Bacillus subtilis in the myo-inositol producing construct were
examined. In the latter case, the lac and tac promoters were compared for expression of

inositol dehydrogenase. Finally, oxidation of neo-inositol to scyllo-2,5-diketoinositol by

38

G. oxydans ATCC 621 was examined to determine the selectivity of G. oxydans ATCC

621 in oxidizing axial alcohols.

OH OH
HO,“ ,,\OH HO,“ 0R1 HO», OH
Ho“' OH Ho“‘ "’OH Ho“‘ , OH
OH 0R2 OH
Fl = H: D-chiro-inositol R1,R2 = H: L-chiro-inositol neo-inositd
R = CH3: D-pinitol R1 = H.122 = CH3: L-pirl'tol
Fl, = CH3,R2 = H: L-quebrachitol
OH OH OH
HO OH HO,“ .“on HO,“ OH
HO OH Ho‘“ "’OH Ho‘“ i "’OH
OH OH OH
cis-inositol Fl = H: muco-inosibl alloinosltol
Fl = CH3: o-1-Omehyl
muco-inositol
OH OFl1 OH
HO”, ,,\OH F160,, ,.\OR2 HO”, ,,\OH
Ho“' i "’OH H50“ 3 0R3 HO i OH
OH OH, OH
apt-inositol Fifi-16 = H: myo-inositol scyllo-inositol

12,-H6 = 1303sz phytic acid

R1,R3-R6= H; R2 =CH3: sequoyitol

R1‘R3,R5,R6 = H, R4 = CH3: L‘mmSlIOl

R1 = CH3; th-Fle = H: ononitol

R1-R5 = H; n, = CH3: D-bomesitol

R1'Ra,R5 = H, R4,R5 = CH3: llfiwmdbl
R1,R2,R4,1R5 = H; R3,R6 = CH3: dambonitol

Figure 19. The inositol family of molecules.

Production of myo-Inositol by Fed-Batch Fermentation

Background

Biosynthesis of myo-inositol starts with uptake of D-glucose and conversion to D-
glucose 6-phosphate catalyzed by the E. coli phosphotransferase system9 where

phosphoenolpyruvate provides the energy driving the transport and is the source of the

39

transferred phosphoryl group (Figure 20). The acyclic form of D-glucose 6-phosphate is
cyclized to L-myo-inositol l-phosphate (MIP) in a reaction catalyzed by L-myo-inositol 1-
phosphate synthase (MIP synthase). This enzyme is encoded by the S. cerevisiae INOI
locus. N AD+ but no divalent cation or Schiff base is required for the enzyme-catalyzed
reaction. A mono phosphatase activity catalyzes the phosphoester hydrolysis of MIP

resulting in accumulation of myo-inositol in culture supematants.

OH PEP OH 0'1 0H

.\ O H b H203P0h. .\\OH C HO,“ “sol
—'——> ———>

i OH é OH HO“ 5 OH HO“ 5 OH

HO OH H203PO OH OH OH
D-glucose D-glucose-6- L-myo-inositol- myoinositol

phosphate 1 phosphate

  
   

   

Figure'20. Biosythesis of myo-inositol. Key: (a) phosphotransferase system; (b) L-myo-
inositol l-phosphate synthase; (c) monophosphatase.

MIP synthase, encoded by the S. cerevisiae locus INOI exists as a tetramer of

244,000 daltons with subunits of 62,000 daltons“),ll Lineweaver-Burk kinetic analysis

of purified enzyme showed a Km = 1.18 x 10'3 M and a Vmax = 167 nmol/h for D-glucose

6-phosphate.ll The sequence for INOI was first reported9 in 1989. The plasmid
pJH31812 harboring a copy of the INOI locus was obtained from S. A. Henry.
Fed-Batch Fermentor Conditions

Fed-batch fermentations were performed in a 2.0 L capacity Biostat MD B-Braun
fermentor equipped with a DCU system and a Dell Optiplex Gs+ 5166M personal
computer utilizing B-Braun MFCS/win software for data acquisition and automatic
process monitoring. Ferrnentations were run at 33 °C, pH 7.0 and dissolved oxygen

(D.O.) level was maintained at 10%. All parameters were controlled by proportional-

40

integral-derivative (PID) control loops. A Metler-Toledo 12 mm sterilizable 02 sensor
fitted with an Ingold A-type permeable membrane was used for monitoring D.O. levels.
Inoculants were initially grown in 5 mL of complete. M9 medium with 50 ug/mL
Ap for 24 h at 37 °C and 250 rpm. This culture was transferred to 100 mL of the same
medium and grown under identical conditions for 10 h before transferring to the
fermentor. The initial D-glucose concentration was kept at 18 g/L. The entire
fermentation can be divided into three stages, each of which corresponds to a different
procedure for controlling DC. at 10%. In the first stage, the airflow was kept constant at
0.06 LIL/min and the stirrer was ramped up from 50 rpm to 940 rpm at a rate sufficient to
maintain the DC. at 10%. Once the stirrer reached its preset maximum of 940 rpm, the
mass flow controller increased the airflow from 0.06 L/L/min to 1.0 L/L/rnin. These two
stages take anywhere from 8 h to 10 h for completion. As soon as airflow reached 1.0
L/L/min at a constant impeller speed of 940 rpm, the D-glucose pump was switched on
and thereafter the DO. level was maintained at 10% by the addition of a 60% (w/v)
solution of D-glucose. The rate of addition of D-glucose was cOntrolled by the PID
setting gain (Kc). The PID setting of Kc = 0.1 was employed for all runs. At the
beginning of the third stage of the fermentation and for every timepoint thereafter, 5 mg
of IPTG was added to the fermentor for induction of the tac promoter. Ferrnentations
were run for 54 h and fermentor samples were taken at 12 h and every 6 h thereafter to
measure cell density and accumulation of myo-inositol. Enzyme samples for MIP
synthase were taken at 18 h, 30 h, 42 h and 54 h. Enzyme samples were prepared for

storage at —80 °C by harvesting cells from 30 mL of fermentor broth at 3 700g for 5 min,

41

4 °C then resuspending the cell pellet in 5 mL of 20 mM Tris0Cl, pH 7.7 containing 10

mM NH4CI, 10 mM B-mercapto ethanol, 1.0 mM EDTA and 0.5 mM PMSF.

pJH31 8
i) PCR
ii) EcoR1 digest
EcoR1 EcoFi1
1 .7-kb

 

pJF118EH INO1

i) EcoH1 digest
ii) CIAP treatment

 

EcoFl1 Sma1

pAD1 .45A
7.0-kb

 

Figure 21. Construction of pADl.45a.

42

pD2625

digest with Dra1

 

 

 

and EcoRV
Dra1 EcoRV
1 .9-kb >1
pAD1 .45A serA

Iigate

EcoFl1 [Sma1]

pAD1.88A Ap
8.9-kb

 

Figure 22. Construction of pADl.88a.

43

Expression of IN 01 Under the tac Promoter

The first biocatalyst which was tested as a myo-inositol producer was
JWFl/pADl.88a. The plasmid pAD1.88a carries a copy of the INOl locus under the
control of the tac promoter, a copy of the serA gene and the B-lac gene conferring Ap
resistance. Construction of pAD1.88a started with amplifying the 1.7-kb open reading
frame of the INOI locus from pJH318 by PCR, digested with EcoRI and ligated with
pJF118EH which had been linearized with EcoRl. This resulted in the 7.0-kb plasmid
pAD1.45a (Figure 21). The vector pJF118EH is a 5.3-kb plasmid carrying the ColEl
origin of replication and has a copy number of approximately 10-15 per cell.13 The
vector also contains a tac promoter, a copy of mm for promoter tunability, and the
genetic marker encoding for Ap resistance. The serA locus was obtained by digestion of
pD262514 with EcoRV and Dra1. Blunt end ligation of the 1.9-kb fragment into Smal
digested pAD1.45A afforded the 8.9-kb pAD1.88a with the serA locus transcribed in a
direction opposite to that of INOI (Figure 22).

JWFl/pADl.88a was examined under fed-batch fermentor conditions. The
addition of 5 mg of IPTG was optimized by Dean.15 After 54 h, myo-inositol production
reached 20 g/L with a yield of 9% (mol of myo-inositol produced/ mol of D-glucose
consumed) (Figure 23). The most rapid production occurred between 36 h and 42 h,
where production occurred at a rate of approximately 1.5 g/h. Between these two time
points, there was a corresponding increase in cell dry weight of approximately 27 g/L
over the six hours. The enzyme activity for MIP synthase, which was taken at 18 h, 30 h,
42 h, and 54 h, held steady at approximately 0.1 nmol min'lmg‘1 for the first three time

points and then dropped slightly over the last 12 h (Table 1).

N
01

 

metabolite concentration (g/L)
dry cell weight (g/L)

 

 

 

12 18

24 30 36

time (h)

42 48 54

Figure 23. Fermentation of JWFl/pADl.88a: PmINOI, sent with 5 mg IPTG.
myo-inositol: I, acetate: 1, formate: . , dry cell weight: ‘ .

Table 1. Specific Activities for MIP synthase.

g
‘

 

 

 

Ml P syngase activity“ myo-inositol IPTGb

construct 18 h 30 h 42 h 54 h titer (g/L) (mg/6 h)
JWFl/pAD1.88& 0.113 0.123 0.102 0.080 20.0 5
JWF1/pCH6.1233 0.036 0.100 0.133 0.116 11.5 5
JWF1/pCH6.1238 0.067 0.130 0.127 0.112 14.6 10
JWF1 /pCH6.1 233 0.130 0.122 0.094 0.086 16.0 20
JWF1 (DE3)/pCH7.66a <0.01 0.010 0.010 0.030 4.6 1

 

 

a umol rnin‘l mg'l.

b isopropyl B—D-thiogalactopyranoside.

The maximum titer of 20 g/L suggested that MIP synthase was not succesfully
competing with other enzymes for intracellular concentrations of D-glucose 6-phosphate.

Raising the specific activity of the enzyme by increasing the level of expression of MIP

45

synthase in E. coli was then explored. A problem often encountered during heterologous
expression of foreign proteins in E. coli is low level protein production. This
phenomenon occurs most often when expressing eukaryotic. proteins in prokaryotic hosts.
Oftentimes, the low expression level can be attributed to the differences in codon usage
between the recombinant E. coli host and the host from which the gene of interest was
isolated.16b An insufficient supply of tRNA to recognize rare codons during protein
synthesis can result in low expression levels, frame shifts, truncation or misincorporation
of amino acids during the translation process.16 The fact that INOI was obtained from S.
cerevisiae, a eukaryote, and expressed in E. coli, a prokaryote, made the issue of codon
usage a logical consideration in the pursuit of enhanced expression of MIP synthase for
improved production of myo-inositol.
Sequencing of INOI

The sequence of the [NO] locus of S. cerevisiae was riddled with several
discrepancies in the literature. The original sequence reported by Henry in 198910 was
later revised by Klig in 1994.17 In this revision, adenines were omitted at positions 1437,
1496, 1506 and one in the series 1583-1587 relative to the start codon. These revisions
caused a major frame-shift resulting in truncation of the gene by 48-bp. The encoded
protein was thus shortened from 553 amino acids to 537 amino acids in length.
Unfortunately, the story does not end there. In 1996, Alexandraki reported18 a new set of
revisions in addition to the changes made by Klig.l7 Changes made to the Henry revision
started with exchanging the adenine at 40 for a guanine. The guanine at 41, adenine at 42
and thymine at 43 were removed resulting in no frame-shift. The following nucleotides

were also removed from the revised Klig sequence: cytosines at 83, 235, 293, 302;

46

adenines at 85, 202, 304; thymines at 233, 234, 303; and a guanine at 84. To the revised
sequence was added a cytosine after cytosine 246 and a guanine after adenine 252. The
sum of all of the additions and deletions shifted the stop codon resulting in the removal of
an additional 12 nucleotides from the gene length and reducing the encoded protein
length to 533 amino acids. The final changes within the revised sequence from Klig were
juxtapositions of guanine 432 with cytosine 433 and cytosine 1341 with guanine 1342.
The second juxtaposition resulted in changing an alanine to a proline in the translated
protein which proved to be key in resolving the active site in the x-ray crystal structure of
the protein. 19 A majority of the changes between the translated proteins occur within the
first 100 amino acids.

Table 2. Primers for sequencing INOI .

 

 

 

sequence INO1 sequencing lab
primer (5'93') region , reference
JWF309 GAGCGGATAACAATI'I'CACACA Pm JF57
JWF324 GCACAATGTGGAGTI'TCAAACTAAGG 275-301 JF61
JWF31 1 A‘lTTGG ATGAAAAAGG CAACGT 623-644 JF44
JWF312 GAGGGTACATTCATI’GCGGGAG 934-967 JF45

JWF31 6 GGCAATG GACG AGTATTACAGTG AG 1 251 -1 2 75 J F51

 

 

The discrepancies in the literature for the sequence of the S. cerevisiae INOI
locus required an extensive effort in confirming the correct gene sequence. Table 2 lists
the primers used to sequence INOI from the plasmid pAD1.88a obtained by amplifying
the INOI locus from pJH318 by PCR. The primer positions were chosen to obtain a
minimum of 100-bp overlap between sequence experiments to ensure the correct

sequence was verified. The dye terminator technique used by the DNA Sequencing Lab

47

in the Department of Plant Biology at Michigan State University verified the sequence by
Alexandraki.l8 As a second line of evidence, the sequence of the INOI locus on pJH318
also agreed with the Alexandraki sequence indicating no mutations occurred during the
amplification process used to construct pAD1.88a.
Codon Usage

The codons that have proven to be a problem for heterologous expression of
proteins in E. coli are shown in Table 3.20 The largest disparity in codon usage between
S. cerevisiae and E. coli is for the AGA and AGG codons for arginine. The AGA and
AGG code for 69% of the arginine codons used in S. cerevisiae but only represent 6% of
the arginine codons used in E. coli. Isoleucine and leucine, represented by ATA and
CTA, respectively, are also used more frequently in S. cerevisiae than in E. coli. Proline,
on the other hand, shows a minor difference in use for the CCC codon between the two

hosts.

Table 3. Codon usage.

 

 

 

, % occurrence“ incidence"
amino acid codon S. cerevisiae E. coli in INO1
arginine AGA 48.0 3.8 9
AGG 20.9 2.2 3
isoleucine ATA 27.3 7.4 1
Ieucine CT A 14.1 3.7 3
proline CCC 15.5 12.5 6

 

 

‘1 % occurence relative to all other codons encoding the same amino acid within each
host.
9 The number of times each codon occurs within the INOI open reading frame.

48

There are two options available to address issues concerning codon usage for
heterologous expression of proteins.” The first option is to replace rare codons by silent
mutations with codons more commonly used in the recombinant host. Introducing silent
mutations by resynthesizing the gene can be a lengthy and cumbersome process with no
guarantee of improved expression. A faster approach is overexpression of the
corresponding tRNA for each rare codon by plasmid-based introduction of additional
copies of the corresponding loci encoding each tRN A sythetase in the recombinant host.

Stratagene has developed the Codon PlusTM system for rapid evaluation of codon
usage issues associated with heterologous protein expression. The strain BL-21-
CodonPlus-RIL harbors a ColEl compatible plasmid containing copies of the argU, ile Y,
and leuW tRNA genes. These genes encode tRNA synthetases for tRNAs that recognize
the arginine codons AGA and AGG, the isoleucine codon ATA, and the leucine codon
CTA, respectively. The strain BL-21-CodonPlus-RP harbors a ,ColEl compatible
plasmid containing copies of argU for arginine and proL recognizing the proline codon
CCC. Each plasmid also contains a copy of the genetic marker conferring Cm resistance
for plasmid maintenance.

The constructs BL-21/pAD1.88a, BL-2l-CodonPlus-RIL/pADl.88a and BL-21-
CodonPlus-RP/pADl.88a were grown by rotary agitation in LB at 37 °C using the
appropriate antibiotics for plasmid maintenance. Production of MIP synthase was
induced by the addition of [RT G to a final concentration of 1 mM. After growing the
cultures to stationary phase, the cells were harvested and checked for MIP synthase
activity. Both BL-2l-CodonPlus constructs enhanced expression of MIP synthase by

approximately two times in comparison to the BL-21 control construct (Table 4). The

49

control construct exhibited MIP synthase activity of 0.027 umol min'l mg“l while the
BL-21-CodonPlus-RIL and BL-21-CodonPlus-RP constructs exhibitied activities of
0.045 umol min'1 mg‘1 and 0.042 limo] min'1 mg], respectively. The common link
between the CodonPlus constructs was the increased expression of argU associated with
the tRNA recognizing the AGG and AGA codons for arginine. The experiment
demonstrated the probability that codon usage was an issue for the expression of MIP
synthase in E. coli.

Table 4. MIP synthase activities vs. tRNA expression.

 

 

 

MIP-synthase"
Host/C0nstruct(s) tRNA Gene Specific Activity
BL-21/pAD1 .88A control 0. 027
BL-21 RIL/pAD1.88A argU, ileY, leuW 0.045
BL-21 RP/pAD1.88A argU, proL 0.042

 

 

a umol rnin'l mg‘l.
Resynthesized INOI: S YNINOI .

Based on enhancement of MIP synthase activity with the Codon PlusTM system,
the INOI locus was mutated to replace rare codons with their common counterparts. The
sequence for the INOI locus was provided to David Rozzell at Biocatalytics. Using a
proprietary algorithm developed by Biocatalytics, the INOI locus was subjected to a
series of gene sequence mutations for the purpose of increasing expression levels in E.
coli but not affecting the translated protein sequence. The first objective was to exchange
all rare codons with more commonly used codons by E. coli. The second objective was

to minimize the negative value of AG of the transcribed mRNA. Minimizing the negative

value for the free energy minimizes the secondary and tertiary structure of the

50

corresponding mRNA making the template more freely accessible for ribosomal
synthesis. Translation of the resynthesized gene S YNINOI encodes a protein with an
identical sequence to the protein encoded by INOI .

The second biocatalyst which was tested as a myo-inositol producer was
JWFl/pCH6.123a. The plasmid pCH6.123a carries a copy of the resynthesized
S YNINOI locus synthesized by Biocatalytics under the control of the tac promoter, a
copy of the serA gene and the fi-lac gene conferring Ap resistance. Construction of
pCH6.123a started with digestion of pGEM-3z/SYNINOI with EcoRI liberating the
S YNINOI open reading frame as a 1.6-kb fragment followed by ligation with pJF118EH
which had been linearized with EcoRI. This resulted in the 6.9-kb plasmid pCH6.112a
with S YNINOI positioned such that its transcription was under the control of the tac
promoter (Figure 24). The serA locus was obtained by digestion of pRCl.55b22 with
EcoRV. Blunt end ligation of the 1.7-kb fragment into BamHl digested and Klenow-
treated pCH6.112a afforded the 8.6-kb pCH6.123a with the serA locus oriented such that

its transcription was in the opposite direction as SYNINOI (Figure 25).

51

pGEM-Sz/SYNINO1

EcoR1 digest

EcoR1 EcoRt

 

pJF1 18EH SYNINO1

EcoR1 digest

 

EcoFl1 Sma1

 
    
     

ApH

pCH6112a
6.9-kb

Figure 24. Construction of pCH6.112a.

52

pFiCl .550

 

 

 

EcoRV digestion
EcoRV EcoRV
1.7-kb >1
pCH6.112a serA
i) BamH1 digestion
ii) Klenow
iii) CIAP treatment

 

EcoR1 Smat

  
    
 

[BamH 1]

pCH6.123a

Figure 25. Construction of pCH6.123a.
JWFl/pCH6.123a was examined under fed-batch fermentor conditions as
described for JWF 1/pAD1.88a. For the initial run, 5 mg IPTG was added to induce

protein synthesis to compare with the level of expression achieved with JWF 1/pAD1.88a.

53

After 54 h, myo-inositol production reached 11.5 g/L with a yield of 5% (mol of myo-
inositol produced/ mol of D-glucose consumed) (Figure 26). Similar to what was seen
with JW F1/pAD1.88a, the most rapid rate of myo-inositol synthesis occurred between 36
h and 42 h. The rate of production was 0.6 g/h, or less than half of what was achieved
with the wild-type INOI. By the 36 h time point, cell density had peaked with only
marginal increases for the remaining 18 h of the run. The enzyme activity for MIP
synthase at 18 h was 0.035 umol min'lmg'l, which was about one third of the activity
measured at 18 h for the control construct JWF 1/pAD1.88a (Table 1). At 30 h, the

activity increased to 0.1 umol min'lmg'l and continued to increase to 0.13 umol min“

1mg"l at 42 h. The activity dropped very slightly to 0.12 umol rnin'lmg'1 at 54 h.

 

 

 

 

14 40
$12.. ‘ ..35
‘ g ‘ A ‘ * "30 Q
:1: 10"- ‘ 3
g E
E ‘ ..25 p)
o 8.1- ‘ 6
g --20 3
o 6.. Q)
g d! 15 3
§ 4" A l --10
115
E 2.. ll II5
o. I -O
12 18 24 30 36 42 48 54

time (h)

Figure 26. Fermentation of JWFl/pCH6.123a: PmSYNINOI, serA with 5 mg IPTG.
myo-inositol: I, acetate: . ., formate: I , dry cell weight: ‘ .

54

The low activity for MIP synthase at 18 h for PmcSYNINOI led to a series of

experiments to determine the concentrations and addition regimen of IPTG that resulted

in optimal MIP synthase expression. The enzyme activity at 18 h doubled to 0.068 mol

rnin'lmg'l when adding 10 mg (Table 1). The initial increase in enzyme activity
translated to a subsequent increase at 30 h. After 30 h, activities at 42 h and 54 h
paralleled what was observed for 5 mg IPTG induction. Surprisingly, the initial boost in
enzyme activity did not translate into an increase in myo-inositol accumulation until 36 h.
At 36 h, the level of myo-inositol accumulation was 2.5 g greater using 10 mg IPTG than
the 36 h time point using 5 mg IPTG. The final titer for myo-inositol was 14.6 g at 54 h

resulting in a 6% yield from D-glucose (Figure 27).

_.L
a)
.h
0"

 

 

314.. ‘ ‘ 40
g 12.. ‘ A 35 go,
as ‘ v
b 10... ‘ 1 30 E
§ A 25 .g’
c 84- l 3
8 A 20 =
a) 6" 8
E 15 2‘
,9 4" A 10 ‘3
“é 2-- l l . 5

0 I I , l._ 1:1.“ 0

12 18 24 3O 36 42 48 54

time (h)

Figure 27. Fermentation of JWFl/pCH6.123a: PtacSYNIN01,serA with 10 mg
IPTG. myo-inositol: I, acetate! I, formate: I, dry cell weight: A.

55

The use of 20 mg IPTG for protein induction achieved a specific activity of 0.130
umol min'lmg‘l for MIP synthase at 18 h (Table 1). This activity was maintained at 30 h
then decreased slightly for the remainder of the fermentation. At 54 h, accumulation of
myo-inositol reached 16.0 g/L, which was slightly less than what was observed for the
wild-type INOI under optimized IPTG addition. While accumulation of myo-inositol
plateus at 54 h for wild-type INOI, the S YNINOI construct continued to accumulate myo-
inositol for an additional 12 h. The addition of 20 mg IPT G resulted in a maximum of

20.1 g/L of myo-inositol after 66 h resulting in a 7% yield from D-glucose (Figure 28).

25

 

20-

15-

10-

metabolite concentration (g/L)
dry cell weight (g/L)

 

 

 

12 18 24 30 36 42 48 54 60 66 72
time(h)

Figure 28. Fermentation of JWFl/pCH6.123a: PtacS YNIN01 , serA with 20 mg
IPTG. myo-inositol: l, acetate: 1 1, formate: a, dry cell weight: A.

The expression of S YNINOI under Pmc failed to increase the specific activity of

MIP synthase. The achieved specific activities obtained by fed-batch fermentation were

no better than those obtained from the wild-type INOI. Further evaluation of the

56

JWF 1/pCH6.123a construct was terminated due to a lack of increased myo-inositol V
accumulation in the supernatant from the resynthesized gene product.
T7 Promoted INOI

Transcription of mRNA by T7 RNA polymerase from the bacteriophage T7 can
elongate chains five times faster than the native E. coli RNA polymerase.23 E. coli
transcriptional machinery is often unable to compete with the T7 RNA polymerase
resulting in transcription of exclusively those genes under the T7 promoter.23 In these
next experiments, the T7 promoter was evaluated for increasing the level of expression of
MIP synthase. Increased MIP synthase activities might, in turn, be reasonably expected
to increase myo-inositol titers microbially synthesized from D-glucose under fed-batch
fermentor conditions.

The third biocatalyst which was tested as a myo-inositol producer was
JWFl/pCH7.66a. The plasmid pCH7.66a carries a copy of the INOI locus under the

control of the T7 promoter, a copy of the serA locus, a copy of [ac] and the B-lac gene

conferring Ap resistance. Construction of pCH7.66a started with digestion of pMIP24
with Ndel and EcoRI liberating the INOI open reading frame as a 1.8-kb fragment with
the Ndel site on the 5’ end and the EcoRl site on the 3’ end. The fragment was ligated
with pET—22b that had been linearized with Ndel and EcoRI. The vector pET-22b25 is a
5.5-kb plasmid carrying the ColEl origin of replication and has a copy number of
approximately 10-15 per cell. The vector also contains a T7 promoter, a copy of lac]
with a lad repressor protein binding site at the T7 promoter region for promoter
tunability, and the genetic marker B—lac for resistance to Ap. This resulted in the 7.3-kb

sub-clone pCH7.41 with INOI positioned such that its transcription was under the control

57

of the T7 promoter (Figure 29). The serA locus was obtained by digestion of pRCl.55b
with EcoRV. Blunt end ligation of the 1.7-kb fragment into PshAl of pET-22b afforded
the 7.2-kb pCH7.61b subclone with the serA locus transcribed in the same direction as
lac] (Figure 30). To make the final plasmid, both pCH7.4l and pCH7.6lb were digested
with Ndel and Scal. The pCH7.41 digest liberated a 3.0-kb fragment containing the
entire INOI open reading frame just after the T7 promoter and 200-bp of the 5’ end of
ApR which was gel purified from the remaining 4.3-kb fragment. The pCH7.61b digest
liberated a 6.0-kb fragment containing the remainder of ApR, serA, [ac] and the T7
promoter which was gel purified from the 1.2-kb remaining fragment. The 3.0-kb
fragment from pCH7.41 and the 6.0-kb fragment from pCH7.6lb were ligated resulting
in rebuilding ApR and placing INOI positioned such that its transcription was under the

control of the T7 promoter to afford the plasmid pCH7.66a (Figure 31).

58

leP pET-22b

N 1”:- R1d' t i)Nob1,EcoR1 digest
d9 00 Iges i0 ClAP treatment

Nde1 EcoR1
1 .8-kb

INO1

RF

N091 Lacl repressor binding site

    

 

 

 

Figure 29. Construction of pCH7 .41.

59

EcoR1 Nd91

 
 
      
   
   

Sca1
pRC1 .55b
pET-22b Iacl ECORV “9°51
EcoRV EcoRV
I 1 .7-kb I
serA

PshA1

 
    

i) PshA1 digest
ii) CIAP treatment

EcoR1 Nde1

pCH7.61b

72*” [PshA1, EcoRV]

 

[EcoRV,PshA1]

Figure 30. Construction of pCH7.61b.

60

pCH7.41 pCH7.61b

i) Nde1, Sca1 digest i) Nde1, Sca1 digest
ii) gel purifiy ii) gel purify
3.0-kb fragment 6.0-kb fragment
ligate

Nde1 Lacl repressor binding site

[EcoRV, PshA1]

Sca1

 

[PshA1, EcoRV]

Figure 31. Construction of pCH7.66a.

Wild—type E. coli RNA polymerase does not recognize the T7 promoter.

Transcription of genes under the control of the T7 promoter in E. coli requires a copy of

the T7 RNA polymerase gene. Typically, this is accomplished by lysogenization of the

T7 RNA polymerase gene onto the E. coli genome under the control of the lacUV5

61

promoter resulting in a DE3 strain. For transcription of the INOI locus, the serA
auxotroph JWFl was lysogenzyed with the T7 RNA polymerase with a kit purchased
from Novagen. Successful lysogenization was established by measuring the activity of
MIP synthase transcribed from leP and by SDS/Page gel of protein extracts. The
construct JWFl/pMIP, which was incapable of recognizing the T7 promoter, was directly
compared to JWFl(DE3)/pMIP.

The construct JWFl(DE3)/pCH7.66a was examined under fed-batch fermentor
conditions. The first fermentation incorporated the addition of 5 mg of [PT G every six
hours. 1H NMR failed to detect accumulation of myo-inositol in fermentor supematants
after 48 h. The cell density reached 19.1 g/L which was significantly lower than the 50
g/L achieved for JWFl/pADl.88a. Raising the IPT G addition to 10 mg resulted in an
increase in the cell density to 29.3 g/L at 48 h, but failed to accumulate myo-inositol.
Enzyme activities for MIP synthase in this run at 18 h and 30 h were not measurable.
Raising the IPTG addition to 100 mg again only resulted in increasing the cell density to
32.8 g/L.

Minimizing the IPTG addition to 1 mg for every six hours resulted in
accumulation of myo-inositol (Figure 32). After 54 h, myo-inositol production reached
4.6 g/L with a yield of 4% (mol of myo-inositol produced/ mol of D-glucose consumed).
The most rapid synthesis of myo-inositol occurred between 48 h and 54 h, when the myo-
inositol titer nearly doubled. The enzyme activity for MIP synthase taken at 18 h was

negligible, but at 30 h and 42 h, the activity reached 0.01 umol min‘lmg‘1 (Table 1). At
54 h hours, the activity increased again to 0.03 umol min'lmg'1 which may explain the

sudden increase in titer for myo-inositol.

62

U'l

 

 

 

 

 

Q
9 4
.5 2
a c»
2 L ‘ E
8 3' a,
O as
3
.9 2 .. g
E a
S 1.. A '0
o
E

O : — ; g

12 18 24 30 36 42 48 54
time (h)

Figure 32. Fermentation of JWFl(DE3)/pCH7.66a: P171N01, serA with 1 mg
IPTG. myo-inositol: I, acetate: 1 l, formate: I, dry cell weight: A .

Poor expression of MIP synthase and very low cell density required a control
experiment with the construct JWFl(DE3)/pADl.883. The control experiment was
meant to determine if overexpression of the T7 RNA polymerase may be responsible for
the low yield of overall cell mass obtained by fed-batch fermentation. The construct
JWFl(DE3)/pADl.88a was examined under fed-batch fermentor conditions using 5 mg
IPTG for induction of protein transcription as optimized for the construct
JWFl/pAD1.88a. As observed for the construct JWFl(DE3)/pCH7.66a, cell density was
much lower for the construct JWF 1(DE3)/pAD1.88a than what is typically observed for
JWFl/pADl.88a (Figure 33). Under normal conditions for JWFl/pAD1.88a, cell
densities can reach approximately 50 g/L. Lysogenized JWFl seemed to hit a ceiling at
30 g/L for dry cell weight. The construct JWF 1(DE3)/pAD1.88a synthesized 10.6 g/L

myo-inositol, which is about half of what is usually observed for the inositol producing

63

construct. Clearly, there was a significant price paid for using the strong T7 promoter.
Appropriation of carbon flow to make the T7 RNA polymerase may have contributed to
lower cell density and lower myo-inositol titer. Cell density and the titer of myo-inositol
were significantly lower using the T7 promoter system making this strategy for
increasing MIP synthase specific activity unsuccessful for the synthesis of myo-inositol

by E. coli under fed-batch fermentor conditions.

12 35
11 --
104-

9... ‘ ‘
8:-
7+' ‘
6.-
5-. ‘ I . 'p15

4..
3.. ‘ l

 

   

metabolite concentration (g/L)
dry cell mass (g/L)

2 II ‘ ‘ II 5
1 .. l
O - —- ' - — - - O
12 18 24 30 36 42 48 54
time (h)

 

 

Figure 33. Fermentation of JWFl(DE3)/pADl.88a: PmINOI, serA with 5 mg
IPTG. myo-inositol: I, acetateu ‘, formate: I, dry cell weight: A.

Coexpression of INOI With ngSL

The third factor affecting enzyme activity is post-translational folding of the
protein leading to catalytically active enzyme. In many cases, the biologically active
form of the enzyme is an oligomeric structure of several subunits. Some heterologously
expressed proteins lack the capability to correctly assemble into their catalytically active

forms. During heterologous expression, the failure to assemble folds correctly is

reflected by the formation of insoluble aggregates known as inclusion bodies. Although
the aggregates can often be solublized in cell extracts by the addition of urea to obtain
catalytically active enzymes, this method can only be done in vitro and is not feasible for
in vivo protein folding. A ubiquitous class of proteins called “chaperonins” have been
discovered, which aid in the folding process and facilitate the formation of the soluble,
catalytically-active oligomeric structures.26 The groEL and groES gene products from E.
coli have been identified to exhibit chaperonin characteristics. The groELS-encoded
chaperonin proteins are induced in E. coli as a response to heat-shock when the growth
temperature is suddenly increased. Several examples have shown the overexpression of
GroELS proteins work together to aid in the correct folding of heterologously expressed
proteins.26b The correct folding can lead to the formation of the soluble, catalytically-
active oligomeric structures and reduction in the formation of inclusion bodies.

The plasmid pGroESL,27 containing copies of groEL and groES, was generously
provided by Du Pont. The plasmid contained the chloramphenicol resistance gene for
plasmid maintenance and utilized the p15A replicon allowing for transformation with the
myo-inositol producing plasmid pAD1.88a as a two plasmid system in the E. coli B strain
BL-21. The constructs BL-21/pAD1.88a and BL-21/pADl.88a/pGroESL were directly
compared for specific activity of MIP synthase. Cell cultures were grown in LB medium
with the appropriate antibiotics for plasmid maintenance and protein production was
induced by the addition of IPTG to a final concentration of 0.4 mM. IPT G was added
when undiluted cell broth reached an absorbance range of 0.4 — 0.6 at 600 nm.

The first experiment was conducted at 37 °C. The cell cultures were shaken until

reaching an OD600 of 3.0 — 3.5 for a 1:10 dilution of cell broth. The cells were then

65

harvested and checked for the specific activity for MIP synthase. The control construct
BL-21/pADl.88a exhibited a specific activity of 0.077 umol min'l mg'1 for MIP
synthase after partially purifying the enzyme extracts by eluting through a DEAE column
using a gradient of buffered ammonium chloride (Table 5, entries 2 and 5). The construct
BL-21/pAD1.88a/pGroESL reached a specific activity of 0.084 umol min'l mg'1 for
partially purified MIP synthase. The coexpression of the chaperonins GroEL and GroES
had a negligible affect on the specific activity of MIP synthase at 37 °C.

Table 5. Specific Activity of MIP synthase with GroES and GroEL.

 

 

 

culture
Construct temperature MIP-synthase Activitya
1. JWF1/pAD1.88a 25 °cb <0.001
2. JWF1/pAD1.88a 37 °C° 0.077
3. JWF1/pAD1.88a 42 °C° <0.001
4. JWF1/pAD1.88a/pgroESL 25 °Cb <0. 001
5. JWF1/pAD1.883/pgroESL 37 °cb 0.084
6. JWF1/pAD1.88a/pgroESL 42 °C° <0.001

 

 

a umol min“l mg"l
b The experiment in its entirety was carried out at this temperature.
C The inoculum was grown at 30 °C prior to adding to the growth medium pre—incubated
at 42 °C.

Temperature can have profound affects on the formation of inclusion bodies
during heterologous expression of proteins. The GroESL proteins can be naturally
induced by sudden exposure of growing cells to elevated temperatures.28 The induced

chaperonins are believed to protect the oligomeric structure from denaturing into

inclusion bodies due to exposure to heat. As a second experiment in an attempt to use the

66

GroESL proteins to enhance the specific activity of MIP synthase, inoculums of both
constructs were grown at 30 °C then added to culture medium pre-warmed at 42 °C. The
same promoter induction procedure was followed as carried out for cells grown at 37 °C.
Heat shocking the cells resulted in no measurable MIP synthase activity for either
construct (Table 5, entries 3 and 6). Lower temperatures during cell growth have also
been shown to alleviate the formation of inclusion bodies. Cultures of either construct
grown at 25 °C also failed to express active MIP synthase (Table 5, entries 1 and 4).
Production of myo-Z-Inosose by Fed-Batch Fermentation

Background

Conversion of myo-inositol to an aromatic will require oxidation of the secondary
alcohols on the carbocyclic ring. One direction to take is the selective oxidation of the
axial alcohol leading to formation of myo-2-inosose. This selective oxidation is known to
exist in nature in microorganisms capable of growth on myo-inositol (Figure 34).
Bacillus subtili329 and Klebsiella aerogenes3O use the oxidation of myo-inositol to my0-2-
inosose catalyzed by inositol dehydrogenase as the first committed step in myo-inositol
catabolism. After dehydration to D-2,3-diketo-4-deoxy-epi-inositol 8, the carbocycle is
cleaved between carbons 2 and 3 to afford 2-deoxy-5-keto-D-gluconic acid 10.29C
Phosphorylation of the terminal alcohol provides 2-deoxy-5-keto-D-gluconic acid 6-
phosphate 11 which is cleaved to dihydroxyacetone phosphate 12 and malonic
semialdehyde 13. The malonic semialdehyde is further processed to acetyl-CoA and
carbon dioxide. Curiously, an enzyme catalyzing the oxidation of myo-inositol also
exists in the bacterial strain Gluconobacter oxydans ATCC 621.31 In this case, inositol

dehydrogenase allows G. oxydans ATCC 621 to produce the required reducing

67

equivalents for conducting cellular processes without utilizing myo-inositol as a source of

 

carbon.
H20
Ho,, Q Ho. H.014 2 Ho,, H_,.0H
——>
V ‘—
Ho‘
OH O
myo-inositol myo-2-inosose 8
OH
Q .\\OH-—> HOOhA “\OH O .‘\OH
Ho2 c “203“) 1102 c
O
9 10 1 1
co2
O + O O 2
'203PO\/u\/OH HOJKJL H fi
12 13 AcetyI-CoA

Figure 34. Proposed route for myo-inositol catabolism.29c

This section will explore the coexpression of inositol dehydrogenase locus iolG of
B. subtilis with the INOI locus from S. cerevisiae in E. coli for the biocatalytic
conversion of D-glucose to my0-2-inosose in one bacterial host. An alternative to E. coli
will be the expression of the INOI locus in G. oxydans ATCC 621 to provide a second
construct for the direct synthesis of my0-2-inosose from D-glucose using a single
microbial host. The oxidation capacity of G. oxydans ATCC 621 will be explored with
the substrates myo—inositol and neo-inositol by a one step bioconversion to establish the
availability of inososes by oxidation of the corresponding inositol with G. oxydans ATCC

621.

68

Synthesis of myo-2-Inosose In E. coli

Although the bioconversion of myo-inositol to myo-2-inosose has been
established in the literature,31 neither the enzyme nor the gene encoding inositol
dehydrogenase in G. oxydans ATCC 621 has been identified. The biocatalytic
conversion of myo-inositol to myo-2-inosose in E. coli required employment of the iolG
locus encoding inositol dehydrogenase isolated from Bacillus subtilis. Inositol
dehydrogenase from B. subtilis has already been successfully overexpressed in E. coli.”b
The active form of inositol dehydrogenase is a 155,000 - 160,000 dalton tetramer
composed of subunits of 39,000 daltons.29a The enzyme requires NAD+ as a cofactor

and has a measured Km of 0.23 mM and 18 mM for NAD+ and myo-inositol,

respectively.2981 The construct MV1184/pIOLOSd15,29b an E. coli strain harboring a
plasmid carrying a segment of the B. subtilis genome including the region containing
iolG in front of a lac promoter, was provided by Y. Fujita.

The first biocatalyst which was tested as a myo-2-inosose producer under fed-
batch fermentation conditions was JWF1/pAD2.28a (Figure 35).32 The plasmid
pAD2.28a is a derivative of pAD1.88a with the addition of a copy of the iolG open
reading frame including the native Shine-Delgamo region under the control of the lac
promoter. JWFl/pAD2.28a was examined under fed-batch fermentor conditions using
the same conditions as used for the JWF1/pAD1.88a construct but with the addition of 10
mg of IPTG to induce both the tac and lac promoters. After 54 h, myo-inositol
production reached 18.2 g/L with a yield 8.2% (mol of myo-inositol produced/ mol of D-
glucose consumed). The accumulation of myo-2-inosose was not observed until the 36 h

time point. A maximum of 1.0 g/L was obtained at 54 h. The fermentation also showed

69

accumulation of myo-inositol l-phosphate (MIP) in the broth. Accumulation of MIP
occurred concurrently with accumulation of myo-inositol and reached a maximum titer of
4.1 g/L at 42 h. During the course of the fermentation, enzyme samples were taken at 18
h, 30 h, 42 h, and 54 h to measure the specific activity of inositol dehydrogenase.
Although Fujita reported29b an activity of 12 umol min'l mg“1 for inositol dehydrogenase
activity from MV1184/pIOL05d15 grown in shake flasks, inositol dehydrogenase
activity from fermentor samples were measured to be 0.037 - 0.045 umol min'l mg'1
(Table 8). Both systems incorporated iolG under a lac promoter. With the activities for
inositol dehydrogenase being much lower by several orders of magnitude relative to the
reported specific activities, efforts focused on raising inositol dehydrogenase expression

levels under fed-batch fermentation conditions.

 

20 ~ ‘ 50
.. - :2-
E 14 , A 35 E
5 12 ‘ A 30 .g’
0 l

a

g 10 I ‘ 25 3
g 8 20 o
= 6 15 .S‘
5 .
6'5 4 , 10
E 2 . . l 1 5

o — o

1 2 1 8 24 30 36 42 48 54
time (h)
Figure 35. Fermentation of JWFl/pAD2.28a: PmINOI, PlaciolG, serA with 10 mg

IPTG. acetate:| I; formate: I; myo-inositol:. ;'myo-inositol l-phosphate: I;
dry cell weight: A.

70

The large difference in inositol dehydrogenase activities required close
examination of the plasmids pIOL05d1529b and pAD1.240a,32 a plasmid containing only
the iolG locus under control of the lac promoter and without a copy of the INOI locus.
The plasmid pIOLOSdlS was identified from a plasmid library of the B. subtilis genome
and carries a copy of iolG and 240 additional nucleotides on the 5’ end of the start codon
positioned such that its transcription was under the control of the lac promoter. Included
in the 240 nucleotides is the Shine-Delgamo region from B. subtilis. A series of deletions
in the region upstream from the start codon of iolG revealed the highest specific activity
was conferred by pIOLOSdlS with the additional 240 nucleotides. Although the
deletions were made to narrow the search for identifying the open reading frame for iolG,
no plasmid was available from this published work?-9b where only the open reading frame
and the Shine-Delgamo from B. subtilis were positioned such that its transcription was
under the control of the lac promoter. The plasmid pAD1.240a was a sub-clone carrying
only the £010 open reading frame with the B. subtilis Shine-Delgarno region under the lac
promoter with the gene conferring resistance to chloramphenicol. A third plasmid,
pCH7.l70, was constructed with iolG and its Shine-Delgamo region positioned such that
its transcription was under the control of the tac promoter for comparison of enzyme
expression under the stronger promoter (Figure 36). These plasmids were examined with
JM83, the E. coli strain from which MV1184 was derived, and JWF], a derivative of
RB791, and both hosts were compared to the specific activity of the construct
MV1184/pIOL05d15 reported by Fujita29b (Table 6).

The E. coli strain MV1184 was developed for the propagation of phagemids for

the construction of plasmid libraries of microbial genomes.33 The strain was created

71

from E. coli strain JM83 by knocking out recA activity by phage insertion of tetracycline
resistance and by the introduction of traD which acts to stabilize the host as a phage
carrier. The introduction of tetracycline resistance intorecA results in the inability to
insert DNA sequences onto the genome of MV1184 by homologous recombination.
Plasmid maintenance by nutritional pressure requires the capability to knock out serA by
homologous recombination. This inability to eliminate genomic expression of serA by
the procedure used for creation of JWFl precluded the use of MV1184 in the fermentor.
Because it is not understood why the inositol dehydrogenase activity reported by Y.
Fujita29b is much higher than what was achieved by fed-batch fermentation, contributions
to the enzyme activity from the mutations found in the E. coli host could not be excluded.
To evaluate the affect of common mutations on the specific activity of inositol
dehydrogenase, JM83 was obtained from ATCC. However, none of the mutations in
either JM83 or MV1184 were suspected to have any affect on inositol dehydrogenase

activity.

72

Table 6. E. coli strains for expressing inositol dehydrogenase.

 

 

 

E. coli strain genotype
JM83 ara A(lac-proAB) rpsL thi (¢ 80 IacZAM15)
MV1184 ara A(Iac-proAB) rpsL thi (4) 80 lacZAM15)

A(sr1-recA)306::Tn 1 (Xtef‘)
F' [traD36 proAB* Iacf' IacZAM15]

RB791 W3110 Iaclq L8
JWF1 W3110 Iaclq L8 serA'

 

 

Construction of pCH7.l70 started with amplification of the iolG locus from
plOLOSdlS with B. subtilis Shine-Delgamo region by PCR using primers which
incorporated a Smal site on the 5’ end and a Pstl site on the 3’ end of the amplified
product. Initially, the PCR product was to be ligated into Smal and Pstl digested
pJF118EH vector DNA. Several attempts at the double digestion of the vector always
resulted in the failure of one or the other enzyme from making the appropriate cut.
Ligation experiments resulted in recircularizing pJF118EH with both out sites
unadultered. To circumvent the problem, pPV3.20a, carrying the E. coli dxr locus with
Smal ends on the vector pJF118EH, was used. The advantage was digestion with Pstl
could be monitored for completion before removing the 1.2-kb dxr fragment by Smal
digestion to relinquish linearized pJF118EH with Smal and Pstl ends. Digestion of the
amplified DNA with Smal and P511 followed by ligation into linearized pJF118EH
afforded the 6.4-kb plasmid pCH7.l70 with iolG positioned such that its transcription

was under the control of the tac promoter (Figure 36).

73

Sma1

 

 

Sma1
Pst1
plOLO5d15
lpcn
Sma1 Pst1
l 1.1-kb I
i) Pst1 digest ,
in) Sma1 digest ’O’G
iii) CIAP treatment
iv) gel purify
ll
Pst1 Sma1 Sma1/Pst1
I 5.3-kb I digest
Ap" Iacf' Pm
r

 

Pst1

 

Figure 36. Construction of pCH7.l70.

74

As the control, the activity for inositol dehydrogenase from the construct
MV1184/pIOL05d15 reported29b by Y. Fujita was confirmed while carrying out the
inositol dehydrogenase assay for each construct listedin Table 7. Inoculants were
initially grown in 5 mL of LB medium with 50 ug/mL Ap for 12 - 16 h at 37 °C and 250
rpm. This culture was transferred to 500 mL of the same medium and grown under
identical conditions until reaching an absorption of 0.4 - 0.6 in undiluted culture broth at
600 nm. Production of inositol dehydrogenase was initiated by the addition of IPT G to 1
mM. The culture was then grown an additional 4 h. Cells from each culture were
harvested by centrifugation, washed with 100 mM Tris-Cl buffer at pH 8.5 buffer, then
resuspended in the same buffer prior to lysing by French Press. Specific activities for
inositol dehydrogenase were determined by following the formation of NADH at 340
nm.293 The enzyme reaction was initiated by the addition of an appropriate amount of
crude lysate to a solution of 40 mM myo-inositol and 0.5 mM NAD+ in 100 mM Tris-Cl
buffer at pH 9.0. The specific activities of inositol dehydrogenase were also determined
by expression of the iolG locus from plasmids pIOL05d15, pADl.240A and pCH7.l70
in E. coli strains JM83 and JWF1 under identical culture conditions with the appropriate
antibiotic for plasmid maintenance (Table 7).

The measured specific activity of inositol dehydrogenase from the construct
MV1184/pIOL05d15 was 18.5 umol min'1 mg'1 (Table 7, entry 2). This result agreed
with the activity of 12.0 umol min‘l mg'1 reported by Fujita29b (Table 7, entry 1).
Expressing iolG from pIOL05d15 in JM83 resulted in a measured specific activity of
15.0 umol min"l mg'l. This was a negligible loss from the MV1184/pIOL05d15

construct indicating the additional mutations in MV1184 were not essential. By

75

comparison, the specific activity for inositol dehydrogenase from the construct
JM83/pAD1.240a was 0.52 umol min‘l mg'l. This result further confirmed the low
inositol dehydrogenase activities observed in cells obtained by fed-batch fermentation
conditions. A specific activity of 0.17 umol min'l mg'1 from the construct
JWF1/pIOL05d15 was a surprise. Comparison of the mutations of JM83 and JWF1 do
not offer any reason for why inositol dehydrogenase activity is much better in JM83 over
JWF1. Expression of inositol dehydrogenase from the plasmid pAD1.240a, where only
the iolG ORF and the B. subtilis Shine-Delgarno are transcribed from the lac promoter,
consistently gave lower activity in either JM83 or JWF 1. This indicates the 240
additional nucleotides upstream from the iolG start codon expressed from the lac
promoter of plasmid pIOL05d15 may contain necessary information for higher
expression levels. Expressing iolG with its native Shine-Delgarno region from the tac
promoter in pCH7.l70 in JM83 or JWF1 resulted in a specific. activity for inositol
dehydrogenase as observed for pIOL05d15 in MV1184 or JM83 indicating the additional
240 nucleotides upstream from the iolG start codon were unnecessary. The specific
activity of inositol dehydrogenase from the construct JM83/pCH7.l70 was 9.8 umol min‘
1 mg'1 (Table 7, entry 7). The specific activity of 10.9 umol rnin'l mg‘1 for the construct
JWF1/pCH7.l70 (Table 7, entry 8) confirmed the mutations in JM83 were not essential
and the 240 nucleotides additional DNA on pIOL05d15 were unnecessary for enhanced
expression of inositol dehydrogenase. These experiments pointed to the direction to take
to increase inositol dehydrogenase activity in the fermentor and established JWF1 was

capable of expressing inositol dehydrogenase to a high specific activity.

76

Table 7. Specific activities for inositol dehydrogenase expression in E. coli.

 

 

 

 

inositol dehydrogenase
construct promoter specific activity“
1. MV1184/pIOL05d15 axe/c; 12.0b
2. MV1184/pIOL05d15 PmiolG 18.5
3. JM83/plOLO5d15 Bach/G 15.0
4. JM83/pAD1.240A Bach/G 0.52
5. JWF1/pIOL05d15 Basia/G 0.17
6. JWF1/pAD1.240A PmioIG 0.94
7. JM83/pCH7.170 P,acioIG 9.5
8. JWF1/pCH7.170 PmioIG 1 0.9

 

 

a umol min'l mg'l.
b Ref. 28b.

77

pCH7.170
6.4-kb

   

Ssp1 digest
Ssp1 Ssp1
I 1.7-kb I
Pm iolG
i) Nru1 digest
ii) CIAP treatment

 

'IG pCH9.1943
'0 10.7-kb serA

[Nru1,Ssp1 ]

Firgure 37. Construction of pCH7.l94a.
Based on the enzyme activities measured in the previous experiments, it was

reasonable to pursue the construction of a plasmid incorporating iolG expression under

78

the tac promoter. The second biocatalyst which was tested as a my0-2-inosose producer
in E. coli was JWF 1/pCH7.194a. The plasmid CH7 .194a carried a copy of the INOI
locus under the control of the tac promoter, a copy of the iolG locus with the B. subtilis
Shine-Delgamo region under the control of the tac promoter, a copy of the serA locus and
the fi-lac gene conferring Ap resistance. Construction of pCH7.l94a started with
liberating the 1.7-kb PtaciolG fragment from pCH7.l70 by digestion with Ssp1 and
ligating into pAD1.88a linearized with Nru1. This resulted in the 10.7-kb plasmid
pCH7.l94a with PtaciolG transcribed in the same direction as mm from pAD1.88a
(Figure 37).

JWF1/pCH7.l94a was examined under fed-batch fermentor conditions exactly as
conducted with JWF1/pAD1.88a. The fermentation incorporated the addition of 10 mg
IPTG at phase-change and at each 6h time point thereafter. After 54 h, myo-inositol
production reached a maximum of 11.6 g/L with a yield of 4.2%. (mol of myo-inositol
produced/ mol of 1D -glucose consumed) (Figure 38). As observed with the
JWF1/pAD2.28a myo-2-inosose producing construct, MIP accumulated in the fermentor
broth for JWF1/pCH7.l94a. MIP accumulation was first observed at 24 h and reached a
maximum of 6.4 g/L at 36 h. By 48 h, MIP had disappeared from the fermentation broth.
Between 60 h and 66 h, the myo—inositol titer decreased from 11.5 g/L at 60 h to 9.6 g/L
at 66 h. Production of my0-2—inosose could not be confirmed by 1H NMR of the cell-free

broth at 66 h.

79

14 60

 

 

 

E? 12" v- 50
8 ‘ 5 ‘ ‘ A A A
E 1O" ‘ --4o 3
U)
5 3.. ‘ a
5 ‘ v30 E
o 6.. ‘ E
8’ --20 ;
§ 4.. ‘ U
H - -1o
“5’ 2' l
0. . 7 _ _7 .1 _; -.. L0
12 18 24 30 36 42 48 54 60 66
time (h)

Figure 38. Fermentation of JWF1/pCH7.l94a: Pun-INOI, PmciolG, serA with 10 mg
IPTG. acetate: 1 ; formate: I; myo-inositolm ; myo-inositol l-phosphate: a;
dry cell weight: A-

Taking advantage of the stronger tac promoter made a significant difference in the
specific activity of inositol dehydrogenase in the fermentor. At'each time point, the
activity increased by over 100 times from that measured for the lac promoted iolG. Use
of the tac promoter met the goal of increasing the specific activity of inositol

dehydrogenase but failed to translate to accumulation myo-2-inosose in the fermentor

broth.

80

Table 8. Inositol dehydrogenase specific activities from fermentations.

 

 

inositol dehydrogenase
specific activitiesa
Construct 18 h 30 h 42 h 54 h

JWF1/pADZ.28A 0.037 0.037 0.033 0.045
JWF1/pCH7.194a 4.9 9.0 9.3 8.4

 

 

 

a umolrnin'1 mg'l.
Synthesis of myo-Z-Inosose From D-Glucose With Gluconobacter oxydans.

The use of Gluconobacter oxydans strain ATCC 621 provides two avenues to
explore for the conversion of D-glucose to myo-inositol. The ATCC 621 strain has been
widely shown to selectively oxidize the axial alcohol of myo-inositol (Figure 41).31 The
production of myo-inositol from D-glucose by fermentation in E. coli followed by
oxidation with G. oxydans ATCC 621 would provide a two step route to myo-2-inosose.
Ideally, G. oxydans ATCC 621 could be engineered to make myo-2-inosose directly from
D-glucose. This approach would require plasmid based expression of the INOI locus
from S. cerevisiae in G. oxydans ATCC 621. Although the use of G. oxydans as a host
for heterologous expression of plasmid-localized genes is not as well documented as E.
coli, promoters recognized by E. coli have been shown to work in G. oxydans.33 The G.
oxydans/E. coli shuttle vector pGE-135 was obtained from A. Fujiwara at Nippon Roche
in Japan.

The preferred carbohydrates for growth by G. oxydans are D-mannitol or D-
sorbitol. Good growth can be observed for D-glucose if pH is controlled. At pH 3.5 -

4.0, D-glucose is nearly quantitatively oxidized to gluconic acid with only a minute

portion being oxidized via the pentose phosphate pathway.36 At pH values below 3.5-

81

4.0, G. oxydans cannot grow on chemically defined medium further indicating the loss of
activity for the pentose phosphate pathway.36 Assimilation of carbohydrates by G.
oxydans is solely dependent upon the pentose phosphate pathway and at pH < 3.5, G.
oxydans requires complex medium for growth.3637 When the pH is controlled above
5.5, direct oxidation of D-glucose to gluconic acid and phosphorylation of D-glucose to D-
glucose 6-phosphate coexist for the first step towards assimilation of D-glucose into the
pentose phosphate pathway.36 Phosphorylation of gluconic acid or oxidation of D-
glucose 6-phosphate provides the intermediate 6-phospho—gluconate as the common entry
point for D-glucose into the pentose phosphate pathway.37 There is evidence supporting
the occurrence of the Entner-Doudoroff pathway in G. oxydans which may act as an
efficient bypass to the pentose phosphate pathway for the conversion of 6-phospho-
gluconate to glyceraldehyde 3-phosphate. For production of myo-inositol, it is essential
that kinase activity is sufficient to provide an in vivo concentration of D-glucose 6-
phosphate as the substrate for MIP synthase. Assimilation of D-glucose by G. oxydans

into the pentose phosphate pathway is shown in Figure 39.

82

   

  
    

..0H 0 OH 9H
—>_ ’ 0H
_ OH 0 OH 0
' H
”0 0” ' acetyl-CoA
D-glucose gluconate I
OH I
.-°H 0 OH 9H a o
____> 2
. o. ~0W°P°ZH° —’ “Yon-l.
H203PO 6H OH OH OH
D-glucose-6— 6-phosphogluconate glyceraldehyde
3-phosphate
phosphate

Figure 39. Metabolic map of D-glucose in G. oxydans.37
Key: a. pentose phosphate pathway or Entner-Doudoroff Pathway.

The biocatalyst which was tested as a myo-2-inosose producer from D-glucose in
G. oxydans was ATCC 621/pCH6.254a. The plasmid pCH6.254a carries a copy of the
INOI locus under the control of the tac promoter, a copy of laclq for promoter tunability,
and a copy of kanR conferring resistance to kanamycin. Construction of pCH6.254a
started by isolating the 4.1-kb fragment containing PtacINOI and lach by digesting
pAD1.45a with Nru1 and Dra1. Blunt end ligation of the fragment into
pGE-l35 linearized by Scal digestion afforded the 16.0-kb plasmid pCH6.254a. The
PmCINOI was transcribed in the same direction as kanR (Figure 40). The shuttle—vector
pGE-l is an 11.9-kb plasmid carrying the origin of replications for G. oxydans and E. coli

and the genetic marker kanR.

Inoculants were initially grown in 5 mL of YPG medium with 50 pg/mL Kan for
40 h at 30 °C and 200 rpm. This culture was transferred to 100 mL of the same medium

and grown under identical conditions for 17 h before transferring to the fermentor

83

containing 900 mL YPG medium with 50 ug/mL Kan. The initial D-glucose
concentration was 25 g/L. The fermentation was conducted at pH 6.0 and maintained by
either the addition of 4 M NaOH or 10% H3PO4 with the temperature set at 28 °C. The
three stage fermentation condition used for myo-inositol production was also used for
ATCC 621/pCH6.254a with the dissolved oxygen set at 20%.

During the course of the run, the stirring rate reached a maximum of 600 rpm by
12 h. The stirring rate then dropped to 460 rpm and the rate was maintained for the
remainder of the fermentation. Analysis of D-glucose concentration showed 23 g/L was
present at 12 h indicating the drop in stirring rate was not due to starvation of the G.

oxydans. At this stage, 5 mg IPT G was added for induction of the PtacINOI for synthesis

of MIP synthase. The fermentation was continued for an additional 18 h. Analysis of D-
glucose concentration at 30 h showed 22 g/L remained. No myo-inositol or my0-2-

inosose was detected by 1H NMR at 30 h and D-glucose consumption ceased.

84

pAD1 .45a
Sca1

Nru1, Dra1
digest

Nru1 Dra1

I 4.1-kb I

lac/q Ptac ”V01

 

i) Sca1 digest
ii) CIAP treatment

 

[Scat , Nru1]

 

[Dra1, Sca1]

Figure 40. Construction of pCH6.254a (G. oxydans vector region in black).

85

Oxidation of myo-Inositol and neo-Inositol by Gluconobacter oxydans

The bioconversion of myo-inositol to myo-2-inosose utilizes G. oxydans ATCC
621 as resting cells. G. oxydans contains a membrane-bound inositol dehydrogenase.
The oxidition of myo-inositol provides the cell with the required reducing equivalents to

maintain cellular functions. Because G. oxydans lacks an enzyme pathway to utilize

myo-2-inosose, the oxidation product accumulates in the culture supernatant.

 

 

OH OH
HOthH G. oxydans H00, .\\OH
OH OH
myo-inositol myo-2-inosose
OH OH
HO!“ O H G. oxyda ’78 H 0],. O
OH OH
neo-inositol scyIIo-2,5-diketoinositol

Figure 41. Cyclitol oxidations carried out by G. oxydans ATCC 621.

For the bioconversion of myo-inositol to myo-2-inosose, 10 mL cultures of G.
oxydans ATCC 621 were inoculated by introducing cells from an agar slant to a medium
containing 1 g D-sorbitol and 50 mg yeast extract.31b The cultures were shaken at 200
rpm and 30 °C until turbid. The entire culture was then added to 400 mL of medium
containing 400 mg D-sorbitol, l g yeast extract and 12 g myo-inositol. The cultures were
grown as previously described for the starter culture. The oxidation reaction was initiated
upon depletion of D-sorbitol and typical conversions were complete within 48 h. After

harvesting the cells by centrifugation and concentrating the supernatant to one-half

86

volume by reduced pressure, my0-2-inosose was precipitated from culture supematants
by the addition of MeOH to afford the oxidation product in 95% yield (Figure 41).

The use of G. oxydans ATCC 621 has not (been limited to myo-inositol.
Magasanik and Chargaff38 investigated oxidation of the known cyclitols with G. oxydans
ATCC 621 and formulated three rules to predict the structures of the alcohols oxidized by
this microbe.

1. The alcohol to be oxidized must be axial in the most stable chair conformation.

2. Placing the axial alcohol facing up, oxidation can only occur if there is an
equatorial alcohol at the meta- position in a counterclockwise sense.

3. Oxidation of the axial alcohol requires an equatorial alcohol at the para-

     

position.
OH OH OH OH
.OH 3 OH b 2031301.“ G014 C HO», G014
———> —> —->
5 OH , OH 140" , OH Ho“' . OH
HO OH '203PO OH OH OH
D-mannose D-mannose-6- neo-inositol- neo-inositol

phosphate 1-phosphate

Figure 42. Conversion of D-mannose to nee-inositol.
Key: a) kinase; b) L-myo-inositol l-phosphate synthase (bovine); c) phosphatase

An alternative cyclitol of interest for oxidation is neo-inositol. As a constituent of
soil, neo-inositol hexaphosphate has been identified with scyllo-inositol hexaphosphate
and phytic acid.39 neo-Inositol has also been identified in brain and testis of bovine and
in brain, heart, kidney, testis and spleen of rat.40 Mammalian L-myo-inositol l-phosphate
synthase isolated from bovine testis has been shown to convert mannose-6—phosphate into

neo-inositol l-phosphate (Figure 42).40 The enzyme L-myo-inositol l-phosphate

synthase from various organisms have an unusually high level of homology for the

87

protein sequence. Is it possible the L-myo-inositol l-phosphate synthase from S.
cerevisiae could catalyze the same reaction?

According to rule #1 of cyclitol oxidation formulated by Magasanik and Chargaff,
neo—inositol should be oxidized twice. However, as soon as one of the axial alcohols is
oxidized, the mono-ketone product no longer fits the criteria of rule #3. It has been
shown that rule #3 does not always apply by the oxidation of nee-inositol with G.
oxydans ATCC 621.41 neo-Inositol was synthesized from myo-inositol following the
published procedure of Potter“2 with no modification and investigated as a substrate with
G. oxydans ATCC 621.

Cultures for the oxidation of neo-inositol were started by adding G. oxydans
ATCC 621 to 5 mL of media composed of 10 g/L yeast extract and 50 g/L D-sorbitol.
After shaking at 200 rpm and 30 °C for 72 h, the entire sub-culture was added to 100 mL
of media containing 5 g/L yeast extract, 1 g/L D-sorbitol and 2 g/L neo-inositol. The
oxidation was carried with the same conditions for the sub-culture for seven days. After
harvesting the cells, reducing the aqueous volume by 80% under reduced pressure

resulted in precipitation of scyll0-2,5-diketoinositol in 50% isolated yield (Figure 41).

HO, OH 0 H20 HO, OH OH H20 HO, DHQ-l

it 4? rim—~7— HO

O 5 OH H O o 5 OH O HO 5 OH'
OH 2 OH H2 OH
scyllo2,5-diketoinosito| monohydrate dihydrate

Figure 43. The hydrates of scyllo-2,S-diketoinositol.
The solubility of neo-inositol is much lower than myo—inositol which‘required
adding a smaller amount to the oxidation media. However, during several oxidation

reactions, a precipitate forms within the first 24 h then disappears towards the end of the

88

bioconversion. Although it is possible the precipitate is neo-inositol, 1H NMR of the

scyllo-2,5-diketoinositol in D20 indicates it exists in the diketo, monohydrate and

dihydrate forms (Figure 43). During the course of oxidation, the precipitate could be the

mono-oxidation product which becomes more soluble as the second oxidation occurs.

Discussion

Efforts to increase the titer of myo—inositol in fed-batch fermentations of E. coli
expressing the INOI locus of S. cerevisiae have been unsuccessful. The major culprits
are the poor Km and Vmx of the enzyme L-myo-inositol l-phosphate synthase. Although
the substrate D-glucose 6-phosphate should be readily available, it is apparent that MIP
synthase is not competitive with other pathways for D-glucose 6-phosphate utilization.
The poor Km and Vmx of the enzyme required increasing the in viva expression levels of
MIP synthase.

The proprietary algorithm developed by Biocatalytics introduced a series of
mutations in the INOI locus to remove the rare codons in this gene relative to codon
usage in E. coli and making the AG of the corresponding mRNA less negative. Making
the AG of the transcribed mRNA less negative reduces the level of folding and allows
translational machinery to bind the mRNA and translate the corresponding protein at an
increased rate and thus increasing the overall cellular concentration of the protein. The
mutations made to the gene do not change the amino acid sequence of the translated
protein.

Unfortunately, the algorithm failed to make any improvement in the specific

activity of MIP synthase heterologously expressed in E. coli. The failure to increase the

89

specific activity of MIP synthase by increasing the rate of protein translation may be the
result of increased protease activity. Cellular machinery can also degrade unnecessary
mRNA as part of its arsenal of degradation enzymes- The minimized folding of the
mRNA could result in easier access for RNAse to bind and degrade the mRNA. Making
the free energy of the mRN A less negative may enhance protein expression, but sets up a
paradoxical situation for its degradation. MIP synthase may thus be a victim of an
unfavorable equilibrium between mRNA synthesis and degradation.

The second problem with over-expressing MIP synthase may be the inability for
E. coli to efficiently fold the protein into the active form of the enzyme. Codon usage
can have dramatic affects on the specific activity of heterologously expressed enzymes.
Increasing the corresponding tRNA for rare codons resulted in increasing the specific
activity of MIP synthase by a factor of two. However, exchanging rare codons for more
common codons in the gene sequence failed to show any effect onthe specific activity of
MIP synthase under fed-batch fermentor conditions. Coexpressing the GroESL proteins
also failed to increase the specific activity of MIP synthase. Although many experiments
have shown GroESL proteins increase enzyme activity by aiding the protein folding
process, there are many cases where GroESL exhibits no affect at all. One option that
remains to be explored involves the laborious task of generating mutants of MIP synthase
and screening for increased titers of myo-inositol in E. coli expressing the mutant INOI
isozymes.

The goal of increasing the specific activity of inositol dehydrogenase in the
fermentor was met by expressing iolG from a tac promoter. Although the specific

activity was increased by at least two orders of magnitude from what was achieved with

90

iolG under the lac promoter, there was no corresponding improvement in myo-2-inosose
titer. There are two issues to consider that may affect the ability for E. coli to accumulate
myo—Z-inosose. The first issue is the enzyme assay for measuring the activity of inositol
dehydrogenase requires the buffer to be at pH 9.0. At pH 7.0, the equilibrium constant
favors reduction of myo-2-inosose to myo-inositol.29’43 The second issue is availability
of myo-inositol as a substrate for the enzyme. If inositol dehydrogenase does not oxidize
myo-inositol prior to its export into the culture supernatant, then an import mechanism is
required for uptake of myo-inositol from the culture supernatant to carry out the oxidation
to myo-2-inosose. The myo-inositol accumulated in the fermentor broth may not have a
pathway to cross the cell membrane and re-enter the cytoplasmic space where the enzyme
resides.

Microbial hosts such as Bacillus subtilis, Klebsiella aerogenes and Pseudomonas
spp. are capable of utilizing myo-inositol as a carbon source. The [first committed step in
each instance of microbial catabolism of myo-inositol entails oxidation to myo-2-inosose.
Each microbe expresses an inositol dehydrogenase that exhibits an in vitro sensitivity to
pH. In all cases, cells were grown in an environment of pH S 7.5 to exhibit growth on
myo—inositol as the sole carbon source. It is unlikely an internal pH 9.0 is maintained in
the aforementioned bacteria during myo-inositol catabolism and equally unlikely the
enzyme equilibrium constant favors reduction within the cytoplasm.

A probable explanation is the in vivo availability of myo-inositol as a substrate for
inositol dehydrogenase. MIP is not a substrate for inositol dehydrogenase and the rate of
myo-inositol export from the cytoplasm may be more rapid than the rate of myo-inositol

oxidation to my0-2—inosose. The suhB-encoded E. coli gene product has been shown to

91

have a high degree of amino acid sequence homology with known mammalian inositol
monophosphatases.44 In vitro experiments have established the suhB-encoded protein
has an activity for the dephosphorylation of L-myo—inositol l-phosphate similar to inositol
monophosphatases from human and bovine brain.44 The suhB-encoded gene product is
believed to be involved with controlling RNase activity in E. colz',"""t”c but may also be
involved with the dephosphorylation of L—myo-inositol l-phosphate to myo-inositol. The
dephosphorylation of L-myo-inositol l-phosphate to myo-inositol may also occur upon
exiting the cell by an unidentified periplasmic phosphatase activity. Once outside the
cell, it is not known whether a transport system is available for the transport of myo-
inositol back into the microbial cytoplasm.

Microbes such as B. subtilis that utilize myo-inositol as a sole source of carbon
express transport proteins for uptake of myo-inositol from the environment. Improving
the yields and titers of myo-2-inosose may require expression inE. coli of the protein
from a microbe such as B. subtilis that mediates myo-inositol transport into the
cytoplasm. For uptake of myo-inositol from the culture media, S. cerevisiae expresses the
proteins ITRl and ITR2.45 The mechanism employed by these proteins for myo-inositol
transport is unknown. In the absence of D-glucose, whole cells of S. cerevisiae were
unable to transport tritium labelled myo-inositol.46 The corresponding genes, [TR] and
ITR2, have been identified and could be candidates for expression in E. coli.45 Similar
Na+-independent transport systems for myo-inositol have also been identified in
Klebsiella aerogenes47 and Pseudomonas putida.“8

A second mechanism is also available for myo-inositol transport across cell

membranes. The protozoan Leishmania donovani has been identified to contain a proton-

92

dependent transport system that is specific for myo-inositol.49 Should the energy
dependent mechanism used by S. cerevisiae fail to import myo-inositol into E. coli, the
proton dependent mechanism from L. donovani provides a second option. The gene
coding for the myo-inositol/proton symporter has been identified.493

G. oxydans ATCC 621 successfully oxidized myo-inositol to myo-2-inosose in
near quantitative yield. Although oxidation of myo-inositol synthesized by E. coli using
G. oxydans ATCC 621 is a two-step process, the microbial oxidation provides easy
access to myo-2-inosose over chemical synthesis as will be described in Chapter 3. The
oxidation of neo-inositol to scyllo-2,5-diketoinositol was achieved by G. oxydans ATCC
621 in 50% yield.

The attempt at converting D-glucose directly to myo-Z-inosose by expressing the
INOI locus in ATCC 621 failed to yield the desired product. A previous report indicated
the importance of growing G. oxydans at pH 2 5.5 for maintaining the activity of the
pentose phosphate pathway thereby limiting the accumulation of gluconic acid in the
culture supernatant.36 Running the fermentation at pH 6.0 should have maintained an
active pentose phosphate pathway. The decision to start with a high concentration of D-
glucose in the fermentor may have been detrimental. A concentration of D-glucose in the
range of 5 - 15 mM results in direct oxidation of D-glucose to gluconic acid and
phosphorylation of D-glucose to D-glucose 6-phosphate to work simultaneously for the
assimilation of D-glucose into the pentose phosphate pathway via 6—phospho-gluconate.36
The extent to which either pathway to 6-phospho-gluconate is operating is not known.

Enhancing the direct phosphorylation pathway may require inactivating the direct

93

oxidation pathway. Tn5 mutagenesis has already been used to knock-out the direct

oxidation pathway in another strain of G. oxydans.50

94

References

1 Haslam, E. In Shikimic Acid: Metabolism and Metabolites; Wiley: New York,
1993.

2 (a) Hopwood, D. A.; Sherman, D. H. Annu. Rev. Genet. 1990, 24, 37. (b) Bentley,
R.; Bennett, J. W. Annu. Rev. Microbiol. 1999, 53, 411.

3 Biosynthesis: Aromatic Polyketides, Isoprenoids, Alkaloids. Topics in Current

Chemistry; Leeper, F. J.; Vederas, J. C., Eds.; Springer-Verlag: Berlin, Heidelberg, New
York; 2000, Vol. 209.

4 (a) Angyal, S. J. Quart. Revs. (London) 1957, 11, 212. (b) Posternak, T. The
Cyclitols; Holden-Day: San Francisco, 1965. (c) Hoffmann-Ostenhof, 0.; Pittner, F. Can.
J. Chem. 1982, 60, 1863.

5 Scherer, J. Liebigs Annalen 1850, 73, 322.

6 Billington, D. C. In The Inositol Phosphates; VCH: New York, 1993.

7 Posternak, T. In The Cyclitols; Holden-Day: San Francisco, 1965; pp. 68-70.

8 (a) Rammesmayer, G.; Pichomer, H.; Adams, P.; Jensen, R. G.; Bohnert, H. J.
Arch. Biochem. Biophys. 1995, 322, 183. (b) Ishitani, M.; Majumder, A. L.; Bomhouser,
A.; Michalowski, C. B.; Jensen, R. G.; Bohnert, H. J. Plant J. 1996, 9, 537. (c)
Sheveleva, E.; Chmara, W.; Bohnert, H. J .; Jensen, R. G. Plant Physiol. 1997, 115, 1211.
(d) Wanek, W.; Richter, A. Physiol. Plant 1997, 101, 416. (e) Nelson, D. E.;
Rammesmayer, G.; Bohnert, H. J. Plant Cell 1998, 10, 753.

9 Postma, P. W.; Lengeler, J. W.; Jacobson, G. R. In Escherichia coli and
Salmonella, 2nd ed.; Neidhardt, F. C., Curtiss, R., 111, Ingrahm, J. L., Lin, B. C. C., Low,

K. B., Magasanik, B., Reznikoff, W. S., Riley, M., Schaechter, M., Umbarger, H. E.,
Eds.; ASM: Washington, 1996; Vol. 1, p. 1149.

10 Dean-Johnson, M.; Henry, S. A. J. Biol. Chem. 1989, 264, 1274.
11 Donahue, T. F.; Henry, S. A. J. Biol. Chem. 1981, 256, 7077.
12 Hirsch, J. P.; Henry, S. A. Mol. Cell. Biol. 1986, 6, 3320.

13 Furst, J. P.; Pansegrau, W.; Frank, R.; Blocker, H.; Scholz, P.; Bagdasarian, M.;
Lanka, E. Gene 1986, 48, 119.

14 Snell, K. D.; Draths, K. M.; Frost, J. W. J. Am. Chem. Soc. 1996, 118, 5605.

95

‘5 Hansen, C. A.; Dean, A. B.; Draths, K. M.; Frost, J. W. J. Am. Chem. Soc. 1999,
121, 3799.

16 (a) Seetharam, R.; Heeren, R. A.; Wong, E. Y.; Bradford, 8. R.; Klein, B. K.;
Aykent, S.; Kotts, C. E.; Mathis, K. J.; Bishop, B. F.; Jennings, M. J.; Smith, C. E.;
Siegel, N. R. Biochem. Biophys. Res. Commun. 1988, 155, 518. (b) Brinkmann, U.;
Mattes, R. E.; Buckel, P. Gene 1989, 85, 109. (c) Spanjaard, R. A.; Chen, K.; Walker, J.
R.; van Duin, J. Nucleic Acids Res. 1990, 18, 5031. (d) Rosenberg, A. H.; Goldman, E.;
Dunn, J. J.; Studier, F. W.; Zubay, G. J. Bacteriol. 1993, 175, 716. (e) Calderone, T. L.;
Stevens, R. D.; Oas, T. G. J. Mol. Biol. 1996, 262, 407.

17 Klig, L. S.; Zobel, P. A.; Devry, C. G.; Losberger, C. Yeast 1994, 10, 789.

18 Katsoulou, C.; Tzermia, M.; Tavemarakis, N .; Alexandraki, D. Yeast 1996, 12,
787.

‘9 Communication with Dr. James Geiger, Department of Chemistry, Michigan State
University.

20 Nakamura, Y.; Gojobori, T.; Ikemura, T. Nuc. Acids Res. 1998, 26, 334.
21 Makrides, S. C. Micro. Biol. Rev. 1996, 60, 512.

22 Unpublished plasmid, Rachel Crist, Dept. Chemistry, Michigan State University,
2000. '

23 Studier, F. W.; Rosenberg, A. H.; Dunn, J. J.; Dubendorff, J. W. Methods Enz.
1990, I39, 60.

24 pMIP: P771N01, Blue.

25 pET-22b purchased from N ovagen

26 (a) Hemmingsen, S. Nature 1988, 333, 330. (b) Thomas, J. G.; Ayling, A.;
Baneyx, F. Appl. Biochem. Biotechnol. 1997, 66, 197.

27 Goluobinoff, P.; Gatenby, A. A.; Lorimer, G. H. Nature 1989, 357, 44.
28 Cowing, D. W. Proc. Natl. Acad. Sci. U. S. A. 1986, 82, 2679.
29 (a) Ramaley, R.; Fujita, Y.; Freese, E. J. Biol. Chem. 1979, 254, 7684. (b) Fujita,

Y.; Shindo, K.; Miwa, Y.; Yoshida, K. Gene 1991, 108, 121. (c) Yoshida, K.; Aoyama,
D.; Ishio, 1.; Shibayama, T.; Fujita, Y. J. Bacteriol. 1997, I 79, 4591.

96

30 (a) Berman, T.; Magasanik, B. J. Biol. Chem. 1966, 241, 800. (b) Berman, T.;
Magasanik, B. J. Biol. Chem. 1966, 241, 807.

31 (a) Kluyver, A. J .; Boezaardt, A. Rec. Trav. Chim. 1939, 58, 956. (b) Posternak,
T. Biochem. Preparations 1952, 2, 57. (c) Carter, H. B.; Belinsky, C.; Clark, R. K., In;
Flynn, E. H.; Lytle, B.; McCasland, G.; Robbins, M. J. Biol. Chem. 1948, 174, 415.

32 Dean, A. B., unpublished results, Dept. of Chemistry, Michigan State University,
1999.

33 MV1 184: araA(lac-praAB) rpsL thi(¢80 lacZAMlS) A(srl-recA)306::Tn10(tetR)
F ’ [traD3 6 proAB+ lach lacZAM 15].

34 Saito, Y.; Ishii, Y.; Hayashi, H.; Imao, Y.; Akashi, T.; Yoshikawa, K.; Noguchi,
Y.; Soeda, S.; Yoshida, M.; Niwa, M.; Hosoda, J.; Shimomura, K. Appl. Environ.
Microbial. 1997, 63, 454.

35 (a) Shinjoh, M.; Hoshino, T. J. Ferment. Bioeng. 1995, 79, 95. (b) Fujiwara, A.;
Hoshino, T.; Shinjoh, M. US. Patent 5,399,496, 1995.

36 (a) Olijve, W.; Kok, J. J. Arch. Microbial. 1979, 121, 283. (b) Macauley, S.;
McNeil, B.; Harvey, L. M. Crit. Rev. Biotech. 2001, 21, l.

37 Asai, T. In Acetic Acid Bacteria; University of Tokyo Press: Tokyo, 1968; pp.
246-262.

38 (a) Magasanik, B.; Chargaff, E. J. Biol. Chem. 1948, 174, 173. (b) Pasternak, T.;
Rapin, A.; Haenni, A.-L. Helv. Chim. Acta 1957, 30, 1594.

39 (a) Cosgrove, D. J. Aust. J. Soil Res. 1963, I, 203. (b) Cosgrove, D. J .; Tate, M. E.
Nature 1963, 200, 568.

40 Sherman, W. R.; Goodwin, S. L.; Gunnell, K. D. Biochemistry 1971, 10, 3491.

41 Anderson, L.; Takeda, R.; Angyal, S. J.; McHugh, D. J. Arch. Biochem. Biophys.
1958, 78, 518.

42 Riley, A. M.; Jenkins, D. J .; Potter, B. V. L. Carbohydr. Res. 1998, 314, 277.

43 Lamer, J .; Jackson, W. T.; Graves, D. J .; Stamer, J. R. Arch. Biochem. Biophys.
1956, 60, 352.

44 (a) Matsuhisa, A.; Suzuki, N.; Noda, T.; Shiba, K. J. Bacterial. 1995, 177, 200.

(b) Inada, T.; Nakamura, Y. Biachimie 1995, 77, 294. (c) Inada, T.; Nakamura, Y.
Biachimie 1996, 78, 209. ((1) Chen, L.; Roberts, M. F. Biochemistry 2000, 39, 4145.

97

45 Nikawa, J .; Tsukagoshi, Y.; Yamashita, S. J. Biol. Chem. 1991, 266, 11184.
46 Nikawa, J.; Nagumo, T.; Yamashita, S. J. Bacterial. 1982, 150, 441.

47 Deshusses, J .; Reber, G. Biachim. Biophys. Acta 1972, 274, 598.

48 Reber, G.; Belet, M.; Deshusses, J. J. Bacterial. 1977, 131, 872.

49 (a) Langford, C. K.; Ewbank, S. A.; Hanson, S. 8.; Ullman, B.; Landfear, S. M.
Mol. Biochem. Parasitol. 1992, 55, 51-64. (b) Seyfang, A.; Kavanaugh, M. P.; Landfear,
S. M. J. Biol. Chem. 1997, 272, 24210.

50 Gupta, A.; Verma, V.; Qazi, G. N. FEMS Microbial. Lett. 1997, 147, 181.

98

chapter;

AROMATIZATION OF M YO-INOSITOL, 2-DEOXY-SCYLLO-
INOSOSE AND TRIACETIC ACID LACTONE

Introduction

Biosynthesis of mya-inositol, 2-deoxy-scylla—inosose and triacetic acid lactone
represent the first step towards synthesis of aromatics from D-glucose via biocatalytic
routes other than the Shikimate pathway. Chemical conversion of these intermediates to
an aromatic molecule, preferably under catalytic conditions, is the next step. Among the
goals for these conversions are the avoidance of protective groups and minimizing waste
streams.

Conditions have been established for converting mya-inositol to l,2,3,4-
tetrahydroxybenzene via the intermediacy of mya-2-inosose using acid-catalyzed
dehydration followed by aromatization. The l,2,3,4 orientation of oxygens on a six-
member ring carbocycle is the core structure of several molecules exhibiting biological
activity. The syntheses of aurantioglyocladin and coenzyme Q3 will be presented which
employ l,2,3,4-tetrahydroxybenzene as the starting material. Acid-catalyzed dehydration
and aromatization of 2-deoxy-scylla-inosose provided access to hydroxyhydroquinone.
Triacetic acid lactone has been converted to phloroglucinol and resorcinol. During the
course of developing a catalytic route for converting phloroglucinol to resorcinol, a
hydrogenation methodology was developed, which has demonstrated general

applicability to the deoxygenation of polyhydroxybenzenes.

99

mya-Inositol Chemistry
Background
The literature provides only one condition for the direct conversion of ' myo-
inositol to an aromatic. Treating mya-inositol with DMSO and acetic anhydride afforded
pentahydroxybenzene pentaacetate as the oxidative/elimination product.l Although the
molecule could be deprotected to pentahydroxybenzene, five moles of acetate waste is
formed for every mole of pentahydroxybenzene produced. The other option is to oxidize

one or more of the alcohols on the myo-inositol ring to promote dehydration and

aromatization.
OH O
HO,“ .,\OH H00, ..\OH
\‘ + \'
0H 3 Ho‘ _,_ O HO‘ i OH
HO,,, _,.OH / OH OH
D,L-epi-inosose
Ho“' 3 OH \ OH
.0”. ”0'0 HO... ..\OH o OH OH
myo-mosutol + OH
O OH HO
OH OH OH O
myo-2-inosose hexaric acid

Figure 44. Oxidation of mya-inositol.
Key: a) concentrated HNO3; b) G. axydans ATCC 621; c) air, 10% Pt on C, H20, 80 °C.

The conversion of myo-inositol to a corresponding inosose has been carried out by

reaction with HNO3, air oxidation catalyzed with Pt on C, and oxidation using G.

oxydans ATCC 621. Treating mya-inositol with concentrated HNO3 was first shown to

afford a mixture of tetrahydroxybenzoquinone and rhodizonic acid.2 Much later, the

pathway from myo-inositol to the quinone and acid was determined to go through the

100

intermediacy of D,L-epi-inosose (Figure 44).2c.g The selective axial oxidation of myo-
inositol to mya-2-inosose with air over Pt on C, was also evaluated.3 Conversion of myo-
inositol to myo-2-inosose by treatment with the G. ‘axydans strain ATCC 621 was
explored in Chapter 2 and will not be covered in this chapter.
Nitric Acid Oxidation

Reports on the oxidation of mya-inositol by boiling to near dryness in HNO3 were
shrouded with contradictions and uncertainty from the initial report in 188723 through
196428 when the last of the intermediates formed during this oxidation were finally
identified. At least two manuscripts”!e opened with a comment concerning efforts to
clear up the uncertainty in what was known and what was speculated for the conversion
of mya-inositol to leuconic acid (Figure 45). Posternak was the first to report the
isolation of the initial oxidation intermediate D,L-epi-inosose in 16% yield from myo-
inositol?-C Isolation required formation of the phenylhydrazone intermediate followed by
recrystallization and deprotection to release the inosose. Fatiadi reported the isolation of
D,L-xyla-4,5,6-trihydroxycyclohexenediolic acid as the corresponding monohydrate of
the potassium salt in 14% yield.28 The formation of hexahydroxybenzene can only be
suggested by formation of the hexaacetate from xyla-4,5,6-t1ihydroxycyclohexane-1,2,3-
trione.2f Under the strong oxidation conditions, the half-life of hexahydroxybenzene is
believed to be too short and is immediately oxidized to tetrahydroxybenzoquinone and

rhodizonic acid.

101

H01, H,\OH HO], H,\OH
OH HO O“ O
H0,“ “‘04 a . [Q] 0 ,‘\OH [0]

5H [0]

Ho‘“ , O , OH
°” £1201“; 5“
m - it . - ' -
yo inos 01 HO, HO i OH xon-4,5,6 tnhydroxy

cyclohexane-1,2,3-trione

D,L- epi-inosose D,L-xyIa-4,5,6-trihydroxy-
cyclohexane diolic acid

0 OH
O OH
OH o
HO OH[O] rhodizonic [O] O COZZfiOq [O] 0&0
—" acid —"
HO OH O o
OH HO OH HcroconicH leuconic
hexahydroxy- triquinoyl acid acid
benzene HO OH
O
tetrahydroxy-

benzoquinone

Figure 45. Oxidation of mya-inositol to leuconic acid.
Key: a) HNO3

Three intermediates of the oxidation pathway were of interest as candidates for
aromatization or for deoxygenation to polyhydroxyaromatics. The intermediates D,L-
xyla-4,5,6-trihydroxycyclohexenediolic acid and D,L-epi-inosose could be dehydrated
once or twice, respectively, prior to aromatization. In the case of D,L-xyla-4,5,6-
trihydroxycyclohexenediolic acid, one dehydration would afford pentahydroxybenzene.
Two eliminations of water from D,L-epi-inosose would result in a mixture of
tetrahydroxybenzenes. The successful isolation of hexahydroxybenzene was also

attractive due to conditions already established for its deoxygenation to phloroglucinol.4

102

Unfortunately, the chemistry associated with oxidation of mya-ino’sitol using
nitric acid was complicated by a variety of factors. myo-Inositol likely forms high energy
nitrate esters upon reaction with nitric acid, which presents an explosion hazard.
Isolation of D,L-epi-inosose would likely require the formation of corresponding
phenylhydrazone unless the percent composition of the mixture could be substantially
increased in favor of the inosose. Attempts at circumventing this problem by
recrystallization were unsuccessful. Oxidation with nitric acid also results in the
formation of N20, which is both an ozone depleter and contributes to the greenhouse
effect.

Oxidation of the axial alcohol of mya-inositol has also been affected by air with
platinum metal. A sub-surface feed of air into an aqueous solution of mya-inositol at 90
°C afforded myo-2-inosose in 60% yield along with 30% unreacted myo-inositol and up
to 10% of hexaric acid, which is a carbon-carbon cleavage product.3b As shown in Table
9, reducing the temperature of the reaction impeded the formation of hexaric acid, but
also resulted in less oxidation of the starting material to the desired product. Additional
catalyst offered no benefit for conversion at any of the temperatures. A major roadblock
would once again be separation of the inosose from the reaction. Anderson3a reported
the use of a cellulose column, which would be just as impractical on large-scale as
forming the hydrazone intermediate. A 10 mol% loading of Pt on C catalyst for an

incomplete conversion of myo-inositol to mya-2-inosose was also unattractive.

103

Table 9. Catalytic oxidations of mya-inositol.

OH OH
HO,,,©:OH Hofiw
——>
Ho“' 5 OH O 5 OH
OH OH
mya-inositol mya-2-inosose

 

 

Catalyst ratioa oxidant temperature yieldb

 

10% Pt/C 11:1 air 90 °C 60 %
10% Pt/C 11:1 air 85 °C 12 °/o
10% Pt/C 11:1 air 75 °C 35 %
10% Pt/C 11:1 air 50 °C 0 °/o
ATCC 621c 30 °C 95%

 

 

a molzmol, mya-inositolth.
b mya-2-inosose.
C G. oxydans ATCC 621.
Chemical Synthesis of mya-2-Inosose

While the oxidation of mya-inositol to mya-2-inosose catalyzed by G. oxydans
ATCC 621 was being assessed, a pure sample of mya-2-inosose was prepared by
conventional chemical synthesis for evaluation of conditions to affect its aromatization.
Starting from mya-inositol, a racemic mixture of acetinides 14 were made with 2,2-
dimethoxypropane (Figure 46).5 Protection of the four remaining alcohols as benzyl
ethers was followed by acid-catalyzed removal of the acetinide to give diol 15 in 77%
overall yield.5 Treatment diol 15 with KOH 6 and benzyl chloride gave a mixture of the
pentabenzylated product 16 with the axial alcohol left unprotected in 45% yield.

Exchanging KOH for NaH as base improved the yield of 16 to 60%. The undesired

hexabenzylated product was removed by washing the solid with petroleum ether.

104

Oxidation by Swern conditions or with TPAP failed to afford pentabenzylated mya-2-
inosose, 17. Jones oxidation conditions afforded mya—2-inosose in 17% yield after
recrystallization from EtOAc/hexanes. Oxidation with Dess-Martin periodinone7
provided the inosose 17 in 92%. Hydrogenation of pentabenzyl mya-2-inosose 17 over

Pd on C cleanly afforded mya-2-inosose in quantitative yield (Figure 46).

OH OBn
HO,“ .\\OH O I @014 H0,“ @0811
a '- b
——> >< —>
Ho“ 5 OH 0" 5 OH HO‘“ 5 OBn
OH OH OBn
mya-inositol 14 15
10
OH OBn
H00, ,,\OH 9 BnO,,, “\030 d 31101,. .003"
4— <—
o a OH o 5 OBn Ho“' 5 OBn
OH OBn (Bn
mon-inosose 17 ~ 16

Figure 46. Chemical synthesis of mya-2-inosose.

Key: a) 2,2-dimethoxypropane, DMSO, stOH, 75%; b) i. BnBr, NaH, DMF, 0-25 °C;
ii. 6N HCl/MeOH, 77%; c) BnCl, NaH, benzene, 60%; d) Dess-Martin, CHzClzzAcOH,
4:3, 92%; e) H2, 10% Pd on C, THFszO, 10:1, 100%.

Aromatization of mya-2-Inosose

Angyal had reported that treatment of mya-2-inosose in an aqueous solution of
NaOH ultimately resulted in the formation of the keto-enediol tautomers 19 and 20
shown in Figure 47.8 Heating the keto-enediols in acetic anhydride with NaOAc resulted
in aromatization to l,2,3,4- and 1,2,3,5-tetraacetoxybenzene in a ratio of 1:2. The
l,2,3,4- orientation can be directly attributed to the keto-enediol tautomers l9 and 20.
Formation of the 1,2,3,5- orientation of acetates presumably occurs by acetylation of the

B-elimination intermediate 18 not isolated by Angyal. Heating mya-2-inosose in acetic

105

anhydride with pyridine or NaOAc exclusively afforded 1,2,3,5-tetraacetoxybenzene.2°
Attempts at treating mya-2-inosose with base and avoiding a protected intermediate failed

to afford an aromatic.

OH OH
HO,, ,.OH a HO,, H,.OH=: H,.OH= HO ,..OH
——>
O . OH ‘—HO
OH O
my02- inosose 20

lb x 1: K of

OH OAc
HO AcO
HO AcO OAc AcO
OH OAc OAc
1,2,3 ,4-tetrahydroxy- 1,2,3,5-tetraacetoxy- 1 ,2,3,4-tetraacetoxy-
benzene benzene benzene

Figure 47. Results of treating mya-Z-inosose with acid or base.
Key: a) NaOH (aq), 25 °C; b) 0.5 M H2804, A, 66 %; c) AczO, NaOAc, A.

Aromatization of myo-2-inosose under acid-catalyzed dehydration conditions
exclusively afforded 1,2,3,4-tetrahydroxybenzene in one step with no detectable
formation of the 1,2,3,5- orientation of alcohols as observed under basic acetolysis
conditions. A dilute solution of mya-2-inosose in acetic acid at 25 - 60 °C failed to
produce any aromatic products (Table 10, entries 1 and 2). The inability of acetic acid to
catalyze aromatization of mya-2—inosose was due in part to the low solubility of the
inosose in acetic acid at the temperature examined. In acetic acid at reflux, mya-2-
inosose was converted to 1,2,3,4-tetrahydroxybenzene in 60% yield quantified by HPLC
(Table 10, entry 3). Unfortunately, even at the refluxing temperature, the solubility of

mya-2-inosose was still low. The reaction formed a considerable amount of

106

unidentifiable polymeric impurities that could not be removed without column
chromatography.

Table 10. Aromatization conditions of mya-Z-inosose.

OH OH
HOnOH H. HO
O 5 OH HO
OH OH
myo-2-inosose 1 ,2,3,4-tetrahydr0xy-
benzene

 

 

solvent substrate conc. temperature yield

 

1. AcOH 0.06 M 25 °c N. R.

2. AcOH 0.06 M 60 °C N. R.

3. AcOH 0.06 M 120°C 60 %a
4. 1.0 M H2804 0.06 M 120°C 43 %°
5. 1.0 M H2304 0.1 M 120°C 49 %b
6. 1.0 M H2804 0.5 M 120°C 53 %°
7. 0.5 M H2804 0.2 M 120°C 66 %b
8. cat. st04 0.2 M 120°C ‘ 18 %°

 

 

3 HPLC quantification.
b Isolated by recrystallization.
C Incomplete reaction.

Changing from an organic solvent to water solved the problem of low solubility.
A dilute solution of mya-2-inosose in 1.0 M sulfuric acid led to a 43% yield of l,2,3,4-
tetrahydroxybenzene (Table 10, entry 4). Extraction of the product from the aqueous
solution with t-butyl methyl ether followed by recrystallization from EtOAc/hexane
allowed the pure product to be isolated without chromatography. Increasing the

concentration of mya-2-inosose in 1.0 M sulfuric acid to 0.5 M resulted in an additional

10% yield of l,2,3,4-tetrahydroxybenzene without complicating the purification (Table

107

10, entry 6). The highest yield of 1,2,3,4-tetrahydroxybenzene was obtained with a 0.2 M
solution of mya-2-inosose in 0.5 M sulfuric acid. After heating for 9 h at 120 °C, the
reaction cleanly afforded 1,2,3,4-tetrahydroxybenzene in 66% isolated yield (Table 10,
entry 7). An attempt to reduce the amount of acid catalyst resulted in incomplete
conversion of myo-2-inosose to l,2,3,4-tetrahydroxybenzene (Table 10, entry 8).

Based upon the results of aromatization of myo-2-inosose under basic acetolysis
conditions, the use of acidic conditions was expected to make either 1,2,3,5-

tetrahydroxybenzene by two B-eliminations followed by aromatization or a mixture of

1,2,3,4- and 1,2,3,5-tetrahydroxybenzene similar to what was observed by Angyal.8
Because neither molecule was readily available, no sprectral techniques could be
definitively used to identify the product except 13C NMR. The isolated product gave
three resonances by 13C NMR. T o confirm the orientation of alcohols on the aromatic
ring, 1,2,3,4-tetrahydroxybenzene was made by two independent syntheses.
Chemical Synthesis of 1,2,3,4-Tetrahydroxybenzene

The only literature synthesis of 1,2,3,4-tetrahydroxybenzene was reported by
Pfieffer and Cobliner in 1904 (Figure 48).9 Pyrogallol was protected as the carbonate 21
with phosgene. Nitration of carbonate 21 afforded a mixture of the ortho and para nitro
carbonates 223 and 22b with respect to the free alcohol. Deprotection of the nitration
mixture with KOH afforded triol 23. The nitro substituent was reduced to the amine 24
by Zn reduction in HCl followed by hydration to afford 1,2,3,4-tetrahydroxybenzene.
Although the synthesis of l,2,3,4-tetrahydroxybenzene could be accomplished in five

steps from pyrogallol, the tedious nature of handling phosgene and the nitration

108

conditions precluded the ease of making large quantities of l,2,3,4—tetrahydroxybenzene

for evaluation as a building block for synthesis.

0 O O
OH )\~o o o
O O
”babe +b“
HO HO HO HO
NO2
pyrogallol 21 22a 22!)

OH OH + _ OH
Hojfij/ OH 9 Hojfij/ NH3C| d How NO2
HO HO HO

1,2, 3,4-tetrahydr0xy- 24 23
benzene

Figure 48. Original synthesis of l,2,3,4-tetrahydroxybenzene.
Key: a) phosgene; b) HNO3/HzSO4; c) KOH; d) Zn, HCl; e) H20.

Two independent syntheses of 1,2,3,4-tetrahydroxybenzene were developed to
confirm the conversion of the polyhydroxybenzene from mya-2-inosose (Figure 49). The
shortest synthesis was in three steps from pyrogallol. Treating pyrogallol with benzyl
bromide and K2CO3 in acetone afforded tribenzyl ether 25 in 83% yield. Oxidation with
30% 11202 and 0.5 equivalents K3Fe(CN)6 provided 2,3-tribenzyloxy-1,4-benzoquinone
26 in 11% yield.10 Hydrogenation of the 26 with 50 psi H2 and 10% Pd on C affected

both deprotection of the benzyl groups and reduction of the ring to l,2,3,4-
tetrahydroxybenzene in one step and in quantitative yield. The low yield for direct
hydroxylation of intermediate 25 and problems associated with separating 26 from the
benzyl alcohol byproduct made this route impractical to scale up.

An alternative and higher overall yielding route to l,2,3,4-tetrahydroxybenzene

required four steps (Figure 49). From the tribenzyl ether 25, the ortho aldehyde 27 was

109

obtained by Vilsmeier-Haack conditions with N-methylformanilide in 90% yield.11
Formylation with N,N-dimethylformamide was sluggish. Attempts to enhance the
conditions with N,N-dimethylformamide by increasing the reaction temperature or
extending the reaction time resulted in debenzylation side reactions. Baeyer—Villigerll
oxidation with performic acid followed by deprotection using NaOMe afforded the
tribenzyl alcohol 28 in a combined yield of 97%. Deprotection by hydrogenation
afforded 1,2,3,4-tetrahydroxybenzene in 80% yield. The l,2,3,4-tetrahydroxybenzene
obtained from either synthesis was identical by 1H and 13C NMR to the material obtained

from the aromatization of mya-2-inosose under acid-catalyzed conditions.

OH OBn
“’15 _,. ”Ii —-'--” HE;
HO BnO
pyrogallol 25\1,2,3,4-tetrahydroxy-

f/ benzene

3:93;? 233%

Figure 49. Alternative syntheses of l,2,3,4-tetrahydroxybenzene.

Key: a) BnBr, K2CO3, acetone, reflux, 83%; b) 30 % H202, K3Fe(CN)6, AcOH, 11%; c)
H2, 10% Pd on C, EtOH, 99%; d) N-methylformanilide, POCl3, 60 °C, 90%; e) i. formic
acid, 30% H202, CHzClz; ii. NaOMe, MeOH, reflux, 97%; f) H2, 10% Pd on C, EtOH,
80%.

l,2,3,4-Tetrahydroxybenzene as a Synthetic Building Block
Polyhydroxybenzenes and polyhydroxyquinones possessing the oxygenation

pattern of 1,2,3,4-tetrahydroxybenzene often display biological activity.

110

Aurantiogliocladin,12 obtained from the species Gliacladium, and fumigatin,l3 found in
the cell broth of Aspergillus fumigatus, are antibiotics (Figure 50). Dillapiol is a
pyrethrin synergist and is responsible for the sedative effect of Perilla frutescens

leaves.l4 Coenzyme Q10 is an essential antioxidant in humans protecting low density

lipoproteins from atherosclerosis-related oxidative modification.15 Coupling biocatalysis
with catalytic chemical modification has provided convenient access to 1,2,3,4-
tetrahydroxybenzene via myo-inositol intermediacy. The general utility of 1,2,3,4-
tetrahydroxybenzene was demonstrated by a concise synthesis of aurantiogliocladin and

coenzyme Q3.

0 o o—-\
H3CO CH3 HO CH3 H3CO 0
H3CO CH3 H3CO H3CO
O O /
aurantiogliocladin fumigatin dillapiol
O
H3CO CH3
Haco \ \ \ \ \ \ \ \ \ \
O

coenzyme 010
Figure 50. Derivatives of l,2,3,4-tetrahydroxybenzene displaying biological activity.
Variations in strategies for hydroxyl protection combined with the ease of
metallation and alkylation of the aromatic nucleus makes 1,2,3,4-tetrahydroxybenzene a
versatile intermediate for a wide spectrum of naturally-occurring 1,2,3,4-
tetrahydroxybenzene derivatives. The simplest use of l,2,3,4-tetrahydroxybenzene was
in the synthesis of aurantiogliocladin. Synthesis of aurantiogliocladin started with the

perrnethylation of l,2,3,4-tetrahydroxybenzene with dimethyl sulfate to afford 29 in 69%

111

yield (Figure 51). Intermediate 29 was sequentially lithiated followed by methylation
with methyl iodide twice to afford 30. Subsequent oxidative demethylation with
(NH4)2Ce(NO3)6 afforded aurantiogliocladin in four steps from 1,2,3,4-
tetrahydroxybenzene. Attempts at permethylating the ring carbons without isolating the
mono ring methylated intermediate failed. Irregardless of using excess n-butyl lithium to
make the highly reactive dianion on the aromatic ring or a sequence of deprotonation and
methylation without isolating the mono ring methylated intermediate, acquisition of

intermediate 30 from 29 was not possible in one step.

OH OCH 3 OCH3CH
HO 8 H3CO H3CO H300
HO H300 H300 H300
OH
1 ,2, 3, 4tetra- aurantiogliocladin

hydroxybenzene
Figure 51. Synthesis of aurantiogliocladin.
Key: a) (CH3O)2S02, NaOH, EtOH, 69%; b) i. n-BuLi, TMEDA, hexanes, THF, 0 °C; ii.

CH3I, 0 0C, 83 %; c) i. n-BuLi, TMEDA, hexanes, THF, 0 °C; ii. CH3I, 0 °C, 54 %; (1)
CAN, pyridine—2,6-dicarboxylate, CH3CN/l-120, 80%.

Homologues of coenzyme Q differ from organism to organism in the length of the
isoprenoid substituent attached to the 1,2,3,4-tetrahydroxylated benzene ring. Coenzyme
Q is an important antioxidant, which has been studied for the treatment of
atherosclerosis.” Studies suggest that the free-radical oxidation of low density
lipoproteins (LDL) within the arterial walls is the first step towards the development of
atherosclerosis. Once oxidized, the LDL is taken up by macrophages. These
macrophages are then deposited along the inner arterial wall as a foam cell.15° Over
time, the collection of foam cells forms an arterial plack leading to blockage of blood

flow and heart failure.

112

It is believed that arterial deposits can be stopped by protecting LDL’s from
oxidation. Recognizing that oxidation is carried out by a free-radical process suggests
that lipid soluble antioxidants could be employed as an atherosclerosis prophylactic. In
comparison to OL-tocopherol, the reduced form of coenzyme Q10 (Qlon) is more
effective at inhibiting free-radical oxidation of LDL’s.15‘i"b Whereas the oxidative free-
radical of a-tocopherol remains in the lipid bilayer to participate in chain propagation
leading to the oxidation of LDL’s, the oxidative free-radical of Q10H2 becomes water
soluble and leaves the lipid bilayer (Figure 52). When combined with L-ascorbic acid, no
LDL oxidation is detected until all L-ascorbic acid then Q10H2 disappears. The interest
in coenzyme Q10 as a potential therapeutic agent has heightened interest in the synthesis
of this molecule from easily accessible starting materials. Current synthetic routes utilize

pyrogallol, gallic acid, or vanillin as starting materials.16

 

OH OH
H300 CH3 H3CO CH3
0 + R00- —» + FiOOH
HSCO \ H H3CO \ H
OH 10 O, 10
Q10H2 Q10H'
Lipid Phase
Aqueous Phase
OH OH
H300 CH3 2 ascorbate H300 CH3
H300 \ H H300 \ H
OH 10 O, 10
Q10H2 010""

Figure 52. Coenzyme Q10 mode of action as an antioxidant.

113

After the potential utility of coenzyme Q10 in drug therapy was realized, the
Nisshin Flour Milling Company became the first industrial producer of coenzyme Q10 in
1974.17 The process required six steps (Figure 53) to- synthesize the six carbon quinone
core for coupling with the isoprene sidechain (Figure 53). Coupling the aromatic core
with the isoprene sidechain by a Lewis acid catalyzed condensation followed by

oxidation to the quinone afforded coenzyme Q10 in eight steps from vanillin. As

discussed in Chapter 1, vanillin is derived from benzene in seven steps. A total of fifteen

synthetic steps (Figure 53) are required to convert benzene into coenzyme Q10 via the

intermediacy of vanillin.

N02 0 NH2 0

© H3COUE1H¢ ab H3CO H d H300 H
——"
H300
32

benzene vanillin 31
i e
OH H300 ‘_H300
WT
CH3 H300 ‘_HaCO

35
H3
H300 \ H3CO
OH 1°

coenzyme 010
Figure 53. Synthesis of coenzyme Q10 from vanillin.

Key: a) Ac20; b) HNO3; c) i. NaOH; ii. (CH3)2SO4; (1) H2, 10 % Pd on C; e)
NO(SO3K)2; f) i. Zn; ii. Ac20; iii. NaOH; g) BF3°Et20; h) NaOH, 02.

114

Efforts to reduce the number of sythetic steps led Keinen to consider p-cresol as a
starting material.18 Perbromination of p-Cresol provided 39 in 76% yield (Figure 54).
Copper(I)-catalyzed nucleophilic aromatic substitution of the bromines with methoxides
followed by methylation of the remaining alcohol with dimethyl sulfate afforded 40 in
94% yield. Coupling of 40 with farnesyl bromide was carried out by lithiation of 40,
formation of the organcuprate with copper cyanide, and then nucleophilic attack on
farnesyl bromide to afford 41 in 73% overall yield from 40. Oxidative demethylation
with (N H4)2Ce(NO3)6 afforded coenzyme Q3.

Keinen reported this route to be the shortest to the coenzyme Qn family of
molecules.18 Including the steps toward making p-cresol from toluene, the coenzyme Q,
family of molecules can now be assembled in six steps using the appropriate sized
isoprene sidechain. The manufacture of p—Cresol is a two step procedure starting from
toluene (Figure 54).19

The decaprenyl sidechain of coenzyme Q10 was not readily available from natural
resources or for purchase from chemical suppliers. Solanesol contains nine of the ten
isoprene units to make coenzyme Q10 and can be obtained by extraction from the leaves
of tobacco, potato plants, and mulberry bushes. Although solanesol could be used for
synthesis of the decaprenyl sidechain, the expense of purchasing solanesol or the
cumbersome extraction and purification of solanesol from natural sources precluded
attempts at synthesis of the decaprenyl sidechain. The difficulties confronting synthesis

of coenzyme Q10 led us to focus on synthesis of coenzyme Q3 due to the availability of

the precursor needed for attachment of the farnesyl sidechain.

115

toluene CH pcresol Br
I_c_. YO 3_d_l 39
38
OCH3 OCH, 0
H300 CH3 9 H3CO CH3 h H3CO CH3
——> —>
H300 H300 \ 3H H300 \ 3H
CCH3 OCH3 o
40 41 coenzyme 03

Figure 54. Synthesis of coenzyme Q10 from p-cresol.
Key: a) C12, FeCl3; b) NaOH, fusion; c) propene, AlCl3; d) 02; e) Brz, Fe, CHC13, 76%;
f) i. Na, MeOH, DME, dimethyl carbonate, CuCN; ii. (CH3)2SO4, 94%; g) i. n-BuLi,
hexanes, TMEDA, 0 °C; ii. THF, CuCN; iii. farnesyl bromide, THF, 73%; h) 2,6-
pyridine dicarboxylate, CAN, CH3CNzH20, 0°C, 48%.

Borrowing from the synthesis of aurantiogliocladin, synthesis of coenzyme Q3
started with perrnethylation of l,2,3,4-tetrahydroxybenzene to afford 29 (Figure 55).
Methylation of the aromatic ring was again accomplished by treatment of 29 with n-butyl
lithium followed by reaction with methyl iodide. This two-step procedure led to the same
intermediate, 40, exploited by Keinen.18 The three-step, one pot procedure to couple the
farnesyl side Chain was carried out with a slight modification to afford coenzyme Q3.
Changing the polarity of the solvent system by the addition of diethyl ether during the

formation of the organocuprate prevented the formation of the undesired SN2’ product.

Again, oxidative demethylation of 41 with (NH4)2Ce(NO3)6 afforded coenzyme Q3.

116

The synthesis of coenzyme Q3 reported by Keinen18 requires six synthetic steps
using toluene as the starting material. The synthesis of coenzyme Q3 from 1,2,3,4-
tetrahydroxybenzene via intermediacy of myo-inositol requires seven steps using D-
glucose as the starting material. Meeting the ultimate goal of producing myo-2-inosose
using one microbial biocatalyst would reduce the synthesis of coenzyme Q3 to six steps
from D-glucose by avoiding the isolation and subsequent oxidation of myo-inositol. In
either case, both procedures offer significant advantages over the route employed by the
Nisshin Flour Company in Japan.17 Reducing the number of synthetic steps from fifteen
to six or seven lowers the cost, time, and generated waste required for synthesis of

coenzyme Q“.

0 H 0C H3 m H3
HO a H300 b H300 CH3 C
—> —> ——--
HO H3CO H3CO
OH OCH3 - OCH3
1 ,2,3,4»tetra- 29 40
hydroxybenzen e
OCH, 0
H300 CH3 d H300 CH 3
—>
OCH3 3 O
41 coenzyme 03

Figure 55. Synthesis of coenzyme Q10 from l,2,3,4-tetrahydroxybenzene.

Key: a) (CH3O)2802, NaOH, EtOH, 69%; b) i. n-BuLi, TMEDA, hexanes, THF, 0 °C; ii.
CH3I, 0 °C, 83 %; C) i. n-BuLi, TMEDA, hexanes, 0 °C, ii. CuCN, T HF, Et20, 0 °C, iii.
farnesyl bromide, -78 °C, 57%; (1) CAN, pyridine-2,6-dicarboxylate, CH3CN/I-120, 0 °C,
46%.

117

'203PO

 

NADH 1

 

HO'“ OH

HO ZOH
2-deoxy-scyio-
inosose

Figure S6. Conversion of D-glucose 6-phosphate to 2-deoxy-scyllo-inosose by the
enzyme BtrC.
Synthesis and Aromatization of 2-Deoxy-scyllo-inosose

The success of converting myo-2-inosose into 1,2,3,4-tetrahydroxybenzene by
acid-catalyzed dehydration followed by aromatization prompted our interest in
aromatization of 2-deoxy-scyllo-inosose. The antibiotics butirosin and neomycin
produced by Bacillus circulans and Streptomyces fradiae, respectively, require 2-
deoxystreptamine as a common intermediate. The intermediate 2-deoxystreptamine
originates from D-glucose through the intermediacy of 2-deoxy-scyllo-inosose.20 The
enzyme BtrC was identified to isomerise D-glucose 6-phosphate to 2-deoxy-scyllo-
inosose via a mechanism similar to that elaborated for dehydroquinate synthase, an
enzyme in the common pathway of aromatic amino acid biosynthesis (Figure 56). The

gene btrC has been identified from B. circulans SANK 72073 and successfully Cloned in

118

E. coli.20d Before attempting to produce 2-deoxy-scyllo-inosose by fermentation in E.
coli, the aromatization of 2-deoxy-scyllo—inosose was evaluated.
Synthesis of 2-Deoxy-scyllo-inosose

A synthesis of enantiomerically pure 2-deoxy-scyllo-inosose in nine steps has
been reported by Kakinuma.21 For evaluation of the aromatization of 2-deoxy-scyllo-
inosose, enantiomerically pure 2-deoxy-scyllo-inosose was not required. A seven-step
synthesis of racemic 2-deoxy-scyllo-inosose from myo—inositol using a route similar to
that developed for synthesis of myo-Z-inosose allowed quick access to the inosose.
Starting from the racemic tetrabenzylated intermediate of myo—inositol 15, the epoxide 42
was prepared by a method developed by Sharpless in 75% yield (Figure 57).22 The
epoxide was opened with lithium aluminum hydride. The unseparated mixture of
alcohols 43a and 43b were oxidized by Dess-Martin periodinone7 followed by

hydrogenation of the protected inosose 44 to afford racemic 2-deoxy-scyllo-inosose.

OBn OBn OBn
HO,,,l:'j:OBn a "6:08" b dorm
—> O), _ —>
HO“' , OBn , OBn HO“' 3 OBn
QBn OBn OBn
15 42 13a
0H OBn OBn
QOH d mOBn C 1'10”.de
<—-—- +——

O ; OH O 5 OBn 3 (En
OH OBn OBn
2-deoxy- 44 43b
scylloinosose
(racemic)

Figure 57. Synthesis of racemic 2-deoxy-scyllo-inosose.

Key: a) i. trimethyl orthoacetate, p-TsOl—l, benzene, rt; ii. (CH3)3SiCl, CH2C12,reflux; iii.
MeOH, K2CO3, rt, 75%; b) LiAlH4, 0 °C, 73%; c) Dess-Martin, CHzClz, 91%; (1) H2,
10% Pd on C, THF:H20, 100%.

119

Aromatization of 2-Deoxy-scyllo-inosose

Aromatization of 2-deoxy-scyllo-inosose drew heavily from the work with myo-
inosose. As observed with my0-2-inosose, aromatizatiOn of 2-deoxy-scyllo-inosose in 0.5
M H2804 at room temperature resulted in no reaction after 24 h (Table 11, entry 2). It
was no surprise that heating the deoxy-inosose to 120 °C in in 0.5 M H2804 resulted in
the necessary dehydrations to afford hydroxyhydroquinone in 33% yield. Although
hydroxyhydroquinone was the predominant product, a streak of several spots was

observed by TLC. 1H NMR of the reaction showed several peaks between 2-8 ppm.
Lowering the concentration of acid from 0.5 M to 0.05M decreased the
streakiness observed by TLC, but only netted a 24% yield of hydroxyhydroquinone

(Table 11, entries 1 and 3). A concentration of 0.005 M H2804 only gave a trace

amount of hydroxyhydroquinone after 24 h at 120 °C (Table 11, entry 4). The reaction

resulted in recovery of most of the starting material. Changing the acid from 0.5 M

H2804 to 0.5 M H3PO4 and heating at 120 °C resulted in a 39% yield of
hydroxyhydroquinone. Although not a significant yield improvement over 0.5 M H2804,
the use of H3PO4 resulted in much lower levels of side product formation.

Unfortunately, the remaining 61% of the starting material was unaccounted for.

120

Table 11. Aromatization of 2-deoxy-scyllo-inosose.

OH OH

LEE“ °”
0 . OH

OH OH

 

 

acid substrate
concentration molarity temp. yield

. 0.5 M H2804 0.2 M 120 °C 33 °/o
0.5 M 112804 0.2 M RT N. R.
0.05 M H2804 0.2 M 120 °C 24 %
0.005 M H2804 0.2 M 120 °C trace
0.5 M H3PO4 0.2 M 120 °C 39 °/o

 

WPP’N‘

 

 

Six months following the completion of this work, Kakinuma reported the
aromatization of 2-deoxy-scyllo—inosose.23 Refluxing 2-deoxy-scyllo-inosose in acetic
acid with acetic anhydride afforded hydroxyhydroquinone triacetate in 29% yield. As
with the work of Posternak and Angyal with my0-2-inosose, acetOlysis of 2-deoxy-scyllo-
inosose leads to an aromatic product. However, while the use of mineral acid directly
affords a polyhydroxyaromatic, acetolysis requires removal Of acetates to obtain the
desired product. Reductive elimination of water from 2-deoxy-scyllo-inosose was also
affected by concentrated H1 in acetic acid which resulted in the formation of catechol in
59% yield. Catechol was not observed to be formed under the acid-catalyzed conditions
developed in 0.5 M H3PO4. Although the yield of catechol is intriguing, the high salt
stream caused by using concentrated HI and the sodium sulfite used to quench iodine
fails to meet the goal of reducing waste associated with polyhydroxybenzene production

in industry.

121

Aromatization of Triacetic Acid Lactone

Evaluation of Claisen condensation conditions was the primary objective for
converting triacetic acid lactone into an aromatic molecule.24 The precedent most
relevent to the conversion Of triacetic acid lactone into phloroglucinol relates to the
conversion of a triacetic acid lactone derivative 45 into phloroglucinol derivative 46 and
benzoate derivative 47 (Figure 58).25 The phloroglucinol derivative 46 results from an
intramolecular Claisen condensation catalyzed by Mg(OCH3)2 while the benzoate
derivative 47 is formed by an intramolecular aldol condensation catalyzed by KOH in
MeOH. The different products formed upon reaction of B-polyketo species to either
Mg(OCH3)2 or KOH in MeOH warranted investigation of both conditions for the

aromatization of triacetic acid lactone.

OH OH O OH

3 O |\ b meow
HO OH P O O P 0'1

Ph 0
46 45 47

Figure 58. Aromatization of triacetic acid lactone derivative via Claisen (a) or Aldo]
(b) Condensations.
Key: a) Mg(OCH3)2, MeOH; b) KOH, MeOH.

Evidence indicating uncertainty in the conversion of triacetic acid lactone to an
aromatic relates to the precedented treatment of dehydroacetic acid with Mg(OCH3)2 or
NaOMe in MeOH (Figure 59). Reaction of dehydroacetic acid with ten equivalents of
Mg(OCH2CH3)2 efficiently produces the tri-B-keto ester 48 in 82% yield. Replacing

Mg(OCH2CH3)2 with NaOMe results in the tri-B-keto ester 49 (18%), methyl

122

acetoacetate 50 (16%), and triacetic acid lactone (7%) with no indication of forming an

aromatic.

0H 0 fix 82%

H30 0 O OH

\

dehydroacetic b ('1 O 0 0 0 0
acid I + +
H3C O 0 WOW Mom
7% 18% 16%
triacetic acid 49 50
lactone

Figure 59. Treatment of dehydroacetic acid with base.
Key: a) Mg(OCH2CH3)2; b) NaOCH3, CH3OH.

Reaction of triacetic acid lactone with varying amounts of Mg(OCH3)2 and
NaOMe was subsequently evaluated. Treating triacetic acid lactone with either five or
ten equivalents of Mg(OCH3)2 at reflux in MeOH resulted in no reaction after 24 h
(Table 12, entries 1 and 2). Formation of the tri-B-keto ester 49 was observed only after
using 0.5 equivalents of Mg(OCH3)2 (Table 12, entry 3). Similar results were obtained
with NaOMe as base. Ten equivalents of NaOMe gave no reaction (Table 12, entry 4),
while one equivalent showed formation of the tri-B-keto ester 49. At this point, no

condition was successful at converting triacetic acid lactone directly to an aromatic. The

pyrone ring was found to be quite resilient to base and the tri-B-keto ester appeared to be

a dead-end product.

123

Table 12. Reaction of triacetic acid lactone with magnesium or sodium methoxide.

 

 

 

OH
4
2| \3 ——>A'bas° mom
H30 (1) O CH30H 3
triacetic 49
acid lactone
base eq. product
1. Mg(OCH3)2 10 NR.
2. Mg(OCH3)2 5 NB.
3- M9(OCH3)2 0.5 49
4. NaOCH3 10 NR.
5. NaOCHa 1 49
6. NaOCH3, 5 MR.

 

 

Aromatization of Triacetic Acid Lactone

Aromatization of triacetic acid lactone required protection of the free alcohol at
carbon 4 (Figure 60). Converting the alcohol to a methyl ether 51 followed by
concentrating the ether with five equivalents of NaOMe in MeOH at 185 °C afforded
phloroglucinol methyl ether 52 in 85% yield after Kugelrohr distillation.27 Work up of
the crude reaction mixture was critical to whether pure phloroglucinol methyl ether was
obtained after distillation. The residue obtained after distilling away the MeOH at 185 °C
was dissolved in water followed by the addition of 50% H2804 to pH 7.5. Extracting the
ether at pH 7.5 afforded relatively clean product with only minor impurities that were
removed upon distillation of product. At pH 7.0 or below, a red syrup was extracted
from the aqueous phase, which contained an impurity not separable from phloroglucinol

methyl ether by distillation.

124

OH OH OH
”H K)
OH OH HO OH
OH OCH3 OCH3 }, phloroglucunol
\ aorborc \ d e f
| _. | ___.
H3C O 0 H30 O O HO OHQ\K

triacetic acid 51 52

lactone O
HO OH

resorcinol

Figure 60. Aromatization of triacetic acid lactone.

Key: a) (CH3)2SO4, Na2C03, acetone, A, 85%; b) MeOH, Dowex-50 (H+), A, 43%; c)
(CH3O)3PO, K2CO3, 140 °C, 79%; (1) Na, MeOH, 185 °C, 85%; e) 12 N HCl, rt, 56%; f)
i. 50 psi. H2, 5% Rh on A1203, 1 N NaOH (aq.); ii. 0.5 M H2804, 82%; g) i. 50 psi H2,
5% Rh on A1203, 1 N NaOH (aq.); ii. 0.5 M H2504, 80 %.

The methyl ether of triacetic acid lactone 51 was readily available in 85% yield by
treating the lactone with NazCO3 and dimethyl sulfate in acetone at reflux (Figure 60).
Although a high yield was achieved by this procedure, the high toxicity of dimethyl
sulfate28 as a methylating agent warranted exploring other options for methylation. A
1,4-addition of MeOI-l followed by elimination of water was catalyzed by Dowex-50
(H+). Triacetic acid lactone methyl ether was obtained in 43% yield with significant
quantities of the tri-B-keto ester 49. The use of Dowex-SO (H+) offered the advantage of
being able to remove catalyst by filtration. After concentrating, the triacetic acid lactone
methyl ether was easily purified by recrystallization. A neat solution of triacetic acid
lactone and K2C03 in trimethyl phosphate at 140 °C afforded the triacetic acid lactone
methyl ether in 79% yield following recrystallization. Trimethyl phosphate is a
methylating agent exhibiting much lower toxicity than dimethyl sulfate.29 The use of

trimethyl phosphate as a neat solution offered the advantage of using a much less toxic

125

methylating agent while minimizing waste by removing acetone from the reaction
condition.

Removal Of the methyl moiety of 52 to make phloroglucinol was conducted most
successfully by stirring the phloroglucinol methyl ether 52 at room temperature in 12 N
HCl for 36 h (Figure 60).”,30 The reaction was quenched by the addition Na2C03
followed by continuous extraction of the aqueous phase for 24 h with t-butylmethyl ether
to obtain a mixture of phloroglucinol and its dimer. Pure phloroglucinol was obtained by
Kugelrohr distillation in 56% yield leaving behind the dimer in 10% yield. The reaction
condition required for removing the methyl ether caused phloroglucinol to self-condense
into a dimer.30 Attempts to minimize dimerization in concentrated HCl were
unsuccessful. Stirring the phloroglucinol methyl ether in 12 N HCl at 4 °C resulted in
incomplete conversion to phloroglucinol while heating the reaction to 60 °C altered the
ratio of products in favor of dimer formation. Minimizing the formation of the dimer by
sacrificing complete conversion of phloroglucinol methyl ether to phloroglucinol was
essential to obtain pure phloroglucinol by distillation. Unfortunately, recrystallization
from water or from combinations of EtOAc/hexanes afforded phloroglucinol in
unpredictable quality and color. Attempts to remove the methyl ether using 48% HBr31

or with A11332 in CH3CN did not improve the yield of phloroglucinol.

Polyhydroxybenzene Deoxygenation Methodology
The key to obtaining resorcinol via triacetic acid lactone hinged upon
development of a catalytic reduction methodology for removing a hydroxyl substituent
from phloroglucinol. Over the years, several reaction conditions have been identified

where the hydroxyl groups Of phloroglucinol undergo tautomerization between the keto

126

and enol forms.33 The formation of a trioxime33a from phloroglucinol treated with
hydroxylamine lead Baeyer to predict phloroglucinol existed in the triketo form (Figure
61). Treatment of phloroglucinol with base and excess methyl iodide or with excess
bromine or chlorine results in the hexamethylated or hexahalogenated triketone,

respectively, from an analogous triketo form (Figure 61).33b

PflDH

HON‘ i LNOH
OCH3 T
a

H3 CH3
0H H30 CH3
H300 OCH3 e b O O
\ {it /
HO OH

phloroglucinol
0A d C O
C
/ \ x
X X
ACO OAC O O
X X
X = Cl or Br

Figure 61. Reactions of phloroglucinol.
Key: a) hydroxylamine; b) CH3I, KOH; c) C12 or Brz; d) ACCl; e) CH2N2.

Reactions of phloroglucinol have also resulted in the formation of products from
the enol form. Acylation of the hydroxyl substituents with acetyl chloride results in only
the formation of the triacetate.33b Treatment of phloroglucinol with diazomethane results
in the formation of the trimethyl ether of phloroglucinol.33b

The use of Raman, ultraviolet and infrared spectroscopy have failed to show any
evidence of tautomer-form of phloroglucinol in the solid state or in solution at pH 7.331)

The common link among reaction conditions apparently involving the keto tautomer of

127

phloroglucinol require the use of base. Indeed, the 1H NMR of phloroglucinol at neutral
pH shows a single peak in the aromatic region in D2034 Upon addition of two mole
equivalents of base, the aromatic peak shifts upfield 'to the alkene region while a second
peak develops for the formation of a methylene proton. The use of 13C and UV
spectroscopy have now provided further evidence for the complex equilibrium that exists

when phloroglucinol is dissolved in an aqueous solution of base.35 At a pH range of 9.2 -
14, the phloroglucinol dianion exists predominately as the 3,5-dihydroxy-2,5-

cyclohexadienone structure 54a shown in Figure 62.

OOH pK_,___= 8.0 Hog:— pK2=9.2 0&7 pK_3__=14 0
HO _O O—

phloroglucinol 54a

ll ll
Ki... .0.

Figure 62. Tautomers of phloroglucinol.

 

It had been reported that reduction of phloroglucinol with sodium borohydride
cleanly afforded resorcinol in 90% yield.36 Fray had presumed the reduction went
through a dihydrophloroglucinol intermediate shown in Figure 63. After treating
phloroglucinol with sodium borohydride, an equivalent of water was eliminated by
refluxing the borate residue in benzene resulting in the formation of resorcinol. In an
attempt to eliminate the borate waste formed using sodium borohydride, alternative

reduction conditions were explored for the conversion of phloroglucinol to resorcinol.

128

OH

O . O
HO OH HO OH

 

phloroglucinol I resorcinol
b C
\ OH A
O _ C)
Na“
dihydrophloroglucinol
sodium salt

Figure 63. Conversion of phloroglucinol to resorcinol.
Key: a) NaBH4; b) 50 psi. H2, 5% Rh on A1203, 1 N NaOH (aq.); c) 0.5 M H2504, A.

Catalytic hydrogenation of either aqueous or alcoholic solutions of phloroglucinol
with PtOz, Pt on C or Pd on C at neutral pH failed to form resorcinol.37 The use of
atmospheric H2 at either rt or reflux in either aqueous or alcoholic solution, or 50 psi H2
at rt on a Parr apparatus resulted in either quantitative recovery of phloroglucinol or a
mixture of phloroglucinol and the fully reduced 1,3,5-cyclohexanetriol. Basic conditions
using either sodium dithionite38 or Zn39 only formed trace amounts of resorcinol even
upon addition of several equivalents of reagent.

Catalytic transfer hydrogenation using formic acid was also employed for the
reduction of phloroglucinol to resorcinol.4O The use of a neat solution of phloroglucinol
in formic acid with 10% Pd on C failed to yield resorcinol at reflux. The more active
potassium salt of formic acid in water only gave trace amounts of resorcinol.41 The in
situ formation of ammonium formate by adding formic acid to a solution of
phloroglucinol in 30% ammonium hydroxide with 10% Pd on C afforded resorcinol in

12% yield after flash chromatography.42 Attempts to improve the conversion by adding

129

additional formic acid, ammonium hydroxide or catalyst failed to drive the reaction to
completion. The fact that only basic conditions gave detectable formation of resorcinol
was further evidence of the importance of forming the dianion 54a (Figure 62) for the
reduction of phloroglucinol to resorcinol.

Smissman and coworkers reported the use of dihydrophloroglucinol sodium salt
as an intermediate for the synthesis of the euphoria-inducing drug dihydrokavain (Figure
63).“3 Using a slightly modified version of their reported procedure, a l M solution of
phloroglucinol in 1 N NaOH was shaken under 50 psi H2 in the presence of a 1.2 mol%
loading of 5% Rh on A1203. After filtering away the catalyst through a plug Of Celite,
the aqueous solution was acidified to pH 6.0 with 10% HCl followed by concentrating to
a yellow oil. Heating the oil at reflux in a solution of 0.5 M H2SO4 afforded resorcinol in
82% yield after Kugelrohr distillation.

The conversion of phloroglucinol to resorcinol using catalytic hydrogenation
under basic conditions followed by acid-catalyzed dehydration is a new methodology for
removing a hydroxyl substituent from an aromatic ring. The methodology avoids
modifying the hydroxyl group to a sulfonate ester,44 tosylate,“5 phosphate ester,46 or
tetrazol47 to enhance deoxygenation of the aromatic ring. Although the equilibrium for
the dianionic species favors structure 54a (Figure 62), an undetectable quantity of dianion
54b must also be present. Whereas sodium borohydride reduces the carbonyl of the
dianion of 54a, hydrogenation is favored at the double bond of the protonated enol of 54b
to afford 56 (Figure 64). The high electron density at the to other two double bonds of

the dianion 54b makes them less reactive towards hydrogenation. Neutralization of 56

130

results in the 1,3-diketo intermediate dihydrophloroglucinol (Figure 63), which is
dehydrated under acid-catalyzed conditions to afford resorcinol.

Because demethylation of the methyl ether 52 required concentrated HCl and the
yield of phloroglucinol was low, it was essential to find a condition for the direct
conversion of 52 to resorcinol. Subjecting 52 to the two-step deoxygenation reduction
methodology directly afforded resorcinol in 80% yield (Figure 60). Hydrogenation is
favored for the double bond at the methyl ether position of the carbocyclic ring 57 as
opposed to the other two double bonds due to their high electron density (Figure 64).
Reduction of the double bond allowed the methyl ether to be eliminated as methanol
under acid catalyzed conditions. The advantage to this procedure was the high yielding,
direct formation of resorcinol from the phloroglucinol methyl ether. The harsh
conditions required for demethylation to phloroglucinol and attendant dimer formation
can now be avoided.

HOH

OH OH
ii iii”
HO OH — O 0 "' _ O O _ \<
phloroglucinol 54b 55 :
HO OH

OCH3 OCH3 H3CO H resorcinol
O Q £1” 4
HO OH _' O O — _ O O " CH30H
52 57 58

Figure 64. Reduction intermdiate of phloroglucinol methyl ether 52.

131

Table 13. Reductions of polyhydroxyaromatics.

 

 

polyhydroxy- areaction

 

benzene catalyst product t’yield
0H Rh/Al203 32
Rh/C 74
Pth O 60
HO .OH Pd /C HO . OH 32
phloroglucnnol resorCInOl
OH
Rh/Al203 44
0“ Rh/C 0“ 43
OH PVC OH 42
OH Pd/C OH 41
1,2, 3,4—tetra- pyrogallol
hydroxybenzene
0“ Rh/Al203 0“ 53
Rh/C 47
W0 46
HO
OH Pd/C OH 18
hydroxyhydro- hydroquinone
qurnone

 

 

a All catalysts were 5% on designated support.
b Isolation by distillation.

The general utility of the deoxygenation methodology was then explored with
other tri- and tetrahydroxylated polyphenols along with the use of other hydrogenation
catalysts (Table 13). deoxygenation of l,2,3,4-tetrahydroxybenzene resulted in a yield of
44% of pyrogallol when 5% Rh on A1203 was used for the initial hydrogenation.
Hydroxyhydroquinone derived from 2-deoxy-scyllo-inosose afforded hydroquinone in
53% yield upon hydrogenation using 5% Rh on A1203 followed by acid-catalyzed
dehydration. The use of 5% Pd of C resulted in significantly lower yields for

deoxygenation of phloroglucinol and hydroxyhydroquinone. The methodology was

132

further explored by examining deoxygenation of l,2,3,4-tetrahydroxybenzene under
varying pH. At pH 7.0 in H20 or pH 2.5 with H2804 in H20, no deoxygenation was
observed using the methodology. Besides Changes in the overall yield observed using
each hydrogenation catalyst, no catalyst changed the product obtained from the
deoxygenation sequence.
Discussion

The aromatization of myo-inositol to l,2,3,4-tetrahydroxybenzene via the
intermediacy of myo-2-inosose is a significant departure from using intermediates in the
common pathway of aromatic amino acid biosynthesis for deriving aromatics from D-
glucose. Under the current procedure, D-glucose is converted to myo-2-inosose in two
steps utilizing four enzymes. As presented in Chapter 2, the E. coli construct
JWF1/pAD1.88a accumulates 20 g/L myo-inositol in 54 h under glucose-limited, fed-
batch fermentation conditions. G. oxydans ATCC 621 elicits. a selective axial alcohol
oxidation of myo-inositol to myo-2-inosose in near quantitative yield by a one-step, whole
cell biocatalytic conversion. The acid-catalyzed dehydration and subsequent
aromatization of an aqueous solution of myo-2-inosose yielding l,2,3,4-
tetrahydroxybenzene is the last step in the synthesis of a polyhydroxyaromatic from D-
glucose. The procedure starts with nontoxic and renewable D-glucose as the starting
material and avoids the use of protecting groups. This contrasts with the synthesis of
1,2,3,4-tetrahydroxybenzene from benzene-derived pyrogallol. Historically, the number
of steps and the harsh conditions required for oxygenating the aromatic ring in the
synthesis of l,2,3,4-tetrahydroxybenzene made use Of this polyhydroxyaromatic

impractical as a building block for chemical manufacture. Future conversion of D-

133

glucose in one fermentation step to myo-2-inosose will make l,2,3,4-
tetrahydroxybenzene even more readily available. Such a convenient Chemo-enzymatic
synthesis of l,2,3,4-tetrahydroxybenzene may Change the outlook for this molecule’s use
as an intermediate in the manufacture of polymers, antioxidants, and pharmaceuticals.
The utility of l,2,3,4-tetrahydroxybenzene was demonstrated in the four-step
synthesis of the antibiotic aurantiogliocladin and in the four-step synthesis of coenzyme

Q3. The four oxygens atoms and six carbon atoms of l,2,3,4-tefiahydroxybenzene are

directly derived from the oxygen and carbon atoms of a single molecule of D-glucose. In
the biosynthesis of coenzyme Q in E. coli via the Shikimate pathway, only one oxygen

atom is derived from D-glucose. While the synthesis of coenzyme Q3 from l,2,3,4-

tetrahydroxybenzene is no shorter than the route reported by Keinen18 from p-cresol,
synthesis of aurantiogliocladin was completed in significantly fewer steps than the Baker
synthesis.12

Acid catalyzed dehydration 2-deoxy-scyllo-inosose afforded
hydroxyhydroquinone. The practicality of this conversion is contingent on development
of a biocatalytic route for the biosynthesis of 2-deoxy-scyllo-inosose from D-glucose.
With only two enzymes required for the conversion of D-glucose to 2-deoxy-scyllo-
inosose and one chemical step for acid-catalyzed dehydration and aromatization,
synthesis of hydroxyhydroquinone via 2-deoxy-scyllo-inosose constitutes the shortest
synthesis yet achieved of an aromatic from D-glucose.

Triacetic acid lactone has been shown to be a useful intermediate for the synthesis
of phloroglucinol and resorcinol. Unfortunately, triacetic acid lactone could not be

directly converted to phloroglucinol. Aromatization catalyzed by sodium methoxide

134

required protecting the alcohol on the lactone as the corresponding methyl ether. A
biocatalytic route for synthesis of triacetic acid lactone may ultimately be realized using
fatty acid biosynthesis or polyketide biosynthesis.

Carrying out the conversion of triacetic acid lactone to phloroglucinol and
resorcinol led to the development of novel methodology for the deoxygenation of
polyhydroxyaromatics. The most important conversion is the selective removal of the
methyl ether from phloroglucinol methyl ether 52 to afford resorcinol. Because
resorcinol is derived directly from 52, isolation of phloroglucinol as an intermediate is no
longer necessary. This avoids the strongly acidic, low-yielding deprotection step

required for deprotection of the methyl ether of phloroglucinol.

O OH
HOfiOH 6 OH
——>
HO 5 OH OH
/7 OH OH

\OH a myo-2-inosose 1,2,3,4-tetra

OH

   
   

hydroxybenzene
3 OH d CO H c
2
HO OH \ _ OH
D-glucose e (I
HO OH OH
OH OH
gallic pyrogallol
acid

Figure 65. Conversion of D-glucose to pyrogallol via the shikimic acid pathway or
myo-inositol biosynthesis.
Key: a) 4 enzyme steps; b) 0.5M H2804, 120 °C; c) i. 5% Rh on A1203, 1 N NaOH, 50
psi. H2; ii. 0.5 M H2804, 120 °C; (1) 21 enzyme steps; e) AroY.

Extension of the deoxygenation methodology revealed that l,2,3,4-

tetrahydroxybenzene could be converted to pyrogallol in approximately 44% yield. With

four enzyme steps for the conversion of D-glucose to myo-2-inosose followed by two

135

chemical steps for the conversion of myo-2-inosose to pyrogallol, D-glucose can now be
converted tO pyrogallol in six steps overall (Figure 65). By comparison, the previous
biocatalytic conversion of D-glucose to pyrogallol through the shikimic acid pathway
required 22 enzyme-catalyzed steps. By combining myo-inositol biosynthesis with two

Chemical conversions, the synthesis of pyrogallol from D-glucose is reduced by sixteen

O OH
”(it b
—>
HO 3 OH OH

steps.

 

2-deoxy-scyflo hydroxy-

inosose hydroquinone
a 10
HO OH HO, COZH
D—glucose \dlk.<j‘ofle__> OOH
HO“
quinHic hydroquinone
acid

Figure 66. Conversion of D-glucose to pyrogallol via the shikimic acid pathway or
myo-inositol biosynthesis.

Key: a) two enzyme steps; b) 0.5M H3PO4, 120 °C; c) i. 5% Rh on A1203, 1 N NaOH, 50
psi H2; ii. 0.5 M H2804, 120 °C; (1) 20 enzyme steps; e) i. NaOCl; ii. A.

Synthesis of hydroquinone from D-glucose via the Shikimate pathway
intermediate quinic acid requires 20 enzyme-catalyzed steps and one chemical step. By
contrast, taking advantage of 2-deoxy-scyll0-inosose intermediacy may allow
hydroquinone to be synthesized using two enzyme catalyzed steps and two chemical

steps (Figure 66). Utilizing the intermediacy of 2-deoxy-scyllo-inosose makes

hydroquinone available from D-glucose in seventeen fewer steps.

136

References
1 Fatiadi, A. J. Chem. Commun. 1967, 441.

2 (a) Maquenne, L. Ann. Chem. 1887, 12, 112. (b) Gelormini, O.; Artz, N. E. J. Am.
Chem. Soc. 1930, 52, 2483. (C) Posternak, T. Helv. Chim. Acta 1936, I9, 1333. (d)
Hoglan, F. A.; Bartow, E. J. Am. Chem. Soc. 1940, 62, 2397. (e) Preisler, P. W.; Berger,
L. J. Am. Chem. Soc. 1942, 64, 67. (f) Fatiadi, A. J.; Isbell, H. S.; Sager, W. F. J. Res.
NBS 1963, 67A, 153. (g) Fatiadi, A. J .; Isbell, H. S. J. Res. NBS 1964, 68A, 287.

3 (a) Post, G. G.; Anderson, L. J. Am. Chem. Soc. 1962, 84, 471. (b) McCasland, G.
E. In Methods in Carbohydrate Chemistry; Whistler, R. C., Wolfram, M. C. Eds.;
Academic Press: New York, 1962; pp. 291-294.

4 Kuhn, V. R.; Quadbeck, G.; Rohm, E. Ann. 1949, 565, 1.

5 Grigg, R.; Warren, C. D. J. Chem. Soc. C 1969, 2367.

O\

Angyal, S. J .; Tate, M. E. J. Chem. Soc. 1965, 6949.

7 (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4156. (b) Ireland, R. E.;
Liu, L. J. Org. Chem. 1993, 58, 2899.

3 Angyal, S. J .; Range, D. Carbohydr. Res. 1979, 76, 121.

9 Einhom, A; Cobliner, J .; Pfeiffer, H. Ber. Dtsch. Chem. Ges. 1904, 37, 110.
10 Matsumoto, M.; Kobayashi, H.; Hotta, Y. J. Org. Chem. 1985, 50, 1766.

11 Kolontis, P.; Major, A.; NOgradi, M. Acta. Chim. Hung. 1983, 113, 367.

12 (a) Vischer, E. B. J. Chem. Soc. 1953, 815. (b) Baker, W.; McOmie, J. F. W.;
Miles, D. J. Chem. Soc. 1953, 820.

13 Baker, W.; Raistrick, H. J. Chem. Soc. 1941, 670.

14 (a) Honda, 6.; Koezuka, Y.; Tabata, M. Chem. Pharm. Bull. 1988, 36, 3153. (b)
Tomar, S. S.; Saxena, V. S. Agric. Biol. Chem. 1986, 50, 2115.

15 (a) Ingold, K. U.; Bowry, V. W.; Stocker, R.; Walling, C. Proc. Natl. Acad. Sci.
U. S. A. 1993, 90, 45. (b) Stocker, R.; Bowry, V. W.; Frei, B. Proc. Natl. Acad. Sci. U. S.
A. 1991, 88, 1646. (c) Steinberg, D. Circulation 1991, 84,1420.

16 (a) Syper, L.; Kloc, K.; Mlochowski, J. Tetrahedron 1980, 36, 123. (b) Sugihara,
H.; Watanabe, M.; Kawamatsu, Y.; Morimoto, H. Liebigs Ann. Chem. 1972, 763, 109. (C)

137

For an overview of earlier syntheses, see: Mayer, H.; Isler, 0. Methods Enzymol. 1971,
I8, 182.

17 Yamamura, Y.; Folkers, K. Ito, Y. In Biochemical and Clinical Aspects of
Coenzyme Q; Elsevier: Amsterdam, 1981, vol. H1.

18 Keinen, B.; Eren, D. J. Org. Chem. 1987,52, 3872.

19 Franck, H. G.; Stadelhofer, J. W. In Industrial Aromatic Chemistry; Springer-
Verlag: New York, 1988.

20 (a) Kudo, F.; Yamauchi, N.; Suzuki, R.; Kakinuma, K. J. Antibiotics 1997, 50,
424. (b) Iwase, N.; Kudo, F.; Yamauchi, N.; Kakinuma, K. Biosci. Biotechnol. Biochem.
1998, 62, 2396. (c) Kudo, F.; Hosomi, Y.; Tamegai, H.; Kakinuma, K. J. Antibiotics
1999, 52, 81. (d) Kudo, F.; Tamegai, H.; Fujiwara, T.; Tagami, U.; Hirayama, K.;
Kakinuma, K. J. Antibiotics 1999, 52, 559.

21 Yamauchi, N .; Kakinuma, K. J. Antibiotics 1992, 45, 756.

22 Kolk, H. C.; Sharpless, K. B. Tetrahedron 1992, 48, 10515.

23 Kakinuma, K.; Nango, B.; Kudo, F.; Matsushima, Y.; Eguchi, T. Tetrahedron
Lett. 2000,41, 1935.

24 Money, T. Chem. Rev. 1970, 70, 553.

25 ' (a) Harris, T. M.; Wachter, M. P; Wiseman, G. A. Chem. Commun. 1969, 177.
(b) Harris, T. M.; Carney, R. L. J. Am. Chem. Soc. 1967, 89, 6734.

26 Crombie, L.; Games, D. B.; James, A. W. G. J. Chem. Soc. Perkin Trans I 1996,
2715.

27 Huber, U., U. S. Patent 4,112,003, 1978.

28 (a) The Merck Index, 11th ed.; Budavar, 8., Ed.; Merck: Rahway, NJ, 1989; p
3247. (b) Review of carcinogenicity studies: IARC Monographs 1974, 4, 271.

29 Nelson, R. B. U. S. Patent 4,453,017.
30 McKillop, A.; Howarth, B. D.; Kobylecki, R. J. Synth. Commun. 1974, 4, 35.
31 Hwang, K.; Park, S. Synth. Commun. 1993, 23, 2845.

32 Bhatt, M. V.; Babu, J. R. Tetrahedron Lett. 1984, 25, 3497.

138

33 (a) Baeyer, A. Ber. Dtsch. Chem. Ges. 1886, I9, 159. (b) For a review on various
reactions with phloroglucinol: Ershov, V. V.; Nikiforov, G. A. Russ. Chem. Rev. 1966,
35, 817.

34 Highet, R.; Batterham, T. J. Org. Chem. 1964, 29, 475.

35 (3) Wang, D.; Hildenbrand, K.; Leitich, J.; Schuchmann, H.-P.; Sonntag, C. v.
Zeits. Fur Natur. B 1993, 48, 478. (b) Lohrie, M.; Knoche, W. J. Am. Chem. Soc. 1993,
115, 919.

36 Fray, G. I. Tetrahedron 1958, 3, 316.

37 Smith, H. A.; Stump, B. L. J. Am. Chem. Soc. 1961, 83, 2739.

38 (a) Vries, J. G.; Bergen, T. J.; Kellogg, R. M. Synthesis 1977, 246. (b) Castaldi,
G.; Perdoncin, G.; Giordano, C. Tetrahedron Lett. 1983, 24, 2487. (c) Krapcho, A. P.;
Seidman, D. A. Tetrahedron Lett. 1981, 22, 179. (d) Louis-Andre, 0.; Gelbard, G.
Tetrahedron Lett. 1985, 26, 831. (e) Camps, R; Coll, J .; Guitart, J. Tetrahedron 1986, 42,
4603.

39 (a) Yamamura, S.; Hirata, Y. J. Chem. Soc (C) 1968, 2887. (b) Sangaiah, R.;
Gold, A. J. Org. Chem. 1987, 52, 3205.

40 (a) Brieger, G.; Nestrick, T. Chem. Rev. 1974, 74, 567. (b) Johnstone, R. A. W.;
Wibly, A. H.; Entwistle, I. D. Chem. Rev. 1985, 85, 129.

41 Rajagopal, S.; Spatola, A. F. J. Org. Chem. 1995, 60, 1347.

42 (a) Ram, S.; Spicer, L. D. Tetrahedron Lett. 1988, 29, 3741. (b) Ram, S.; Spicer,
L. D. Synth. Commun. 1992, 22, 2673.

43 Smissman, E. E.; Voldeng, A. N.; McCarthy, J. F. J. Pharm. Sci. 1966, 55, 1101.
44 Cacchi, S.; Ciattini, P. G.; Morera, B.; Ortar, G. Tetrahedron Lett. 1986, 27, 5541.

45 (a) Kenner, B.; Murray, C. J. Chem. Soc. 1949, 178. (b) Rottendorf, H.; Stemhell,
S. Aust. J. Chem. 1963, I6, 647.

46 Welch, S. C.; Waters, M. E. J. Org. Chem. 1978, 43, 4797.
47 (a) Musliner, W. J.; Gates, J. W., Jr. J. Am. Chem. Soc. 1966, 88, 4271. (b)

Hussey, B. J.; Johnstone, R. A.; Entwistle, I. D. Tetrahedron 1982, 38, 3775. (c)
Johnstone, R. A. W.; Price, P. J. J. Chem. Soc., Chem. Commun. 1984, 845.

139

h r4

EXPLORATION OF POLYKETIDE AND FATTY ACID
BIOSYNTHESIS FOR BIOSYNTHESIS OF TRIACETICACID
LACTONE

Introduction

As discussed in Chapter 3, a chemical route has been developed for the
conversion of triacetic acid lactone to phloroglucinol and resorcinol (Figure 60, Chapter
3). The success of the Chemical modification of triacetic acid lactone to phloroglucinol
and resorcinol is contingent on the development of a biocatalytic route to triacetic acid
lactone from D—glucose. Current manufacture of triacetic acid lactone starts with the
pyrolysis of acetic acid to ketene (Figure 67).1 Ketene then dimerizes to diketene upon
cooling followed by opening of the lactone with ethanol to afford ethyl acetoacetate.
Condensation with a second equivalent of ethyl acetoacetate over FeO affords the lactone
product dehydracetic acid.2 Treatment of dehydracetic acid with strong acid affords the
deacylated product triacetic acid lactone.3 Although acetic acid is relatively inexpensive,
the synthesis of triacetic acid lactone in five chemical steps make triacetic acid lactone an
unattractive building block for making aromatics.

Chapter 4 will focus on the synthesis of triacetic acid lactone through biocatalysis
using either polyketide biosynthesis or fatty acid biosynthesis. The polyketide
biosynthetic enzyme 6-methylsalicylic acid synthase required mutation of the
ketoreductase module in the enzyme to remove the reduction step. For fatty acid
biosynthesis, two approaches were explored. In E. coli, fabG codes for the ketoreductase
step used in fatty acid biosynthesis. The knockout of this gene was anticipated to make

triacetic acid lactone. A second approach was to express the fasB gene from

140

Brevebacterium ammoniagenes in E. coli. Mutation of the ketoreductase step will allow

the FAS-B protein to make triacetic acid lactone.

0 a b
—> ='=O —> =< >=o
OH .

acetic acid ketene diketene

10

OH OH 0
m A .\ A M

O O O 0 CB
triacetic acid dehydracetic ethyl acetoacetate
lactone acid

Figure 67 . Industrial synthesis of triacetic acid lactone.
Key: (a) (CH3CH20)3PO, 700 °C; (b) 3540 °C; (c) CH3CH20H, H+; (d) FeO, A; (e) 94-
99% H2804.
Evaluation of Polyketide Biosynthesis

Triacetic acid lactone has already been recognized as a derailment product of the
fully functional 6-methylsalicylic acid synthase (6-MSAS) even in the presence of the
optimum concentration of NADPH required for the formation of 6-methylsalicylic acid
(6-MSA).4 When NADPH is removed from in vitro enzyme reactions, triacetic acid
lactone is the exclusive product synthesized by 6-MSAS.4 There are several examples of
polyketide synthetic platforms that are capable of adding a third equivalent of malonyl-
COA to the enzyme bound triketide intermediate to make an enzyme-bound tetraketide
intermediate prior to the first modifying step.5 However, in 6-MSAS, there is no
ketosynthase domain that recognizes the enzyme-bound triketide intermediate. This

results in cyclization of the enzyme-bound triketide intermediate followed by release of

triacetic acid lactone.

141

 

 

 

O O
0°? °’ ° U
2 ESH Es HO SCoA _
a ACPSW ACPSH
O O
O O CO2

 

 

m

m
K‘
v

 

 

 

”SW 3
O o
ESH ESH +
ACPSW -——> ACPSH “0
O O O 0 0”
6-MSA

Figure 68. Biosynthesis of 6-methylsalicylic acid (6-MSA).
Key: (a) ketosynthase; (b) ketoreductase; (c) dehydrase.

The enzyme 6-MSAS is classified as a Type I polyketide biosynthetic route with
all enzymatic domains contained on one protein. 6-MSAS catalyzes a minimum of 11

steps in the synthesis of 6-MSA (Figure 68).6 Formation of 6-MSA starts with the

142

acyltransferase mediated acetylation and malonylation of the ketosynthase’s active-site
cysteinyl residue and acyl-carrier protein (ACP) by, respectively, acetyl-COA and
malonyl-COA. Ketosynthase then mediates the decarboxylation followed by
condensation to form the acetoacetylated ACP in the ketosynthase domain. After
transacylation to the cysteinal residue of the ketosynthase domain, a second round of
ACP malonylation followed by ketosynthase-catalyzed decarboxylation/condensation
results in 3,5-diketohexanoyl-ACP. With NADPH providing the reducing equivalents,
ketoreductase reduces the C-3 carbonyl. Dehydrase then catalyzes elimination of H20 to
form 5-ketohex-3-enoate, which undergoes transacetylation from ACP to the
ketosynthase active site cysteinal residue. A third and final round of ACP malonylation
and ketosynthase-mediated decarboxylation/condensation forms the 3,7-diketo-5-enoyl-
ACP that undergoes cyclization and aromatization to form 6—methylsalicylyl-ACP. 6-
MSA is then released from 6—MSAS.

Although the fungal enzyme from Penicillium patulum was known for at least
thirty years, the sequence of the gene encoding 6-MSAS was not determined until 1990.
Two exons of 29 and 1,745 base pairs are separated by a 69 base pair intron.6a
Expression Of 6-MSAS in Streptomyces coelicolor A3, a bacterial host, required
reconstructing the open reading frame of the gene to remove the intron.7 However,
production of 70 mg/L of 6-MSA was disappointingly low. Expression of catalytically
active 6-MSAS in S. coelicolor indicates that the bacterial host is capable of correctly
folding the tetramer into the active form and providing the 4’ phosphopantetheine

cofactor required to convert the apo-ACP into the active holo-ACP form.

143

S. coelicolor has not been widely used for heterologous gene expression. This
contrasts with S. cerevisiae and E. coli, which have been widely employed as hosts for
expression of recombinant proteins. Expression of 6-MSAS in either S. cerevisiae or E.
coli resulted in inactive enzyme due to these hosts’ inability to convert the apo-ACP into
active holo-ACP using the native phosphopantetheinyl transferase.8 Coexpression of 6-
MSAS with the sfp gene encoding the phosphopantetheinyl transferase from B. subtilis
resulted in expression of active 6—MSAS in both S. cerevisiae and E. coli.8’9

The S. cerevisiae strain INVScl (his3DI, leu2, trp1-289, ura3-52) was used for
the biocatalytic production of 6-methylsalicylic acid. 6-MSAS was expressed from the
alcohol dehydrogenase promoter ADH2 with copies of leu2 and ura3-52 for plasmid
maintenance on pKOSlZ-102d while 5}? was expressed using the ADH2 promoter and a
copy Of trpI-289 on pKOS 12-128a. The construct INVScl/pKOS12-102d/pKOS12-128a
was initially grown in synthetic medium lacking leucine, tryptophan or uracil
supplementation while using D-glucose as carbon source. Cells Were harvested and then
transferred to a rich medium with additional D-glucose. The ADH2 promoter is repressed
in the presence of D-glucose and derepressed by ethanol upon depletion of D-glucose.
The construct produced 1.7 g/L 6-MSA in 150 h.

Production of triacetic acid lactone by 6-MSAS was only shown to occur in vitro
and resulted in two conclusions: (1) the reduction step occurred at the triketide stage and
(2) the ketosynthase domain was incapable of accepting the unreduced triketide as a
substrate. To test if both conclusions were true required inactivation of the ketoreductase
domain in 6-MSAS followed by identification of resulting products. Mutations were

made of the corresponding gene in the conserved GxGxxG sequence associated with the

144

nucleotide binding site of the Rossman fold (Figure 69).“),11 A single amino acid
mutation was made for G1419A. Expression of the mutated gene did not result in the
formation of triacetic acid lactone in cell cultures. After exchanging all three amino acids
of the Rossman fold initiation site, G1419A, G1421P, and Gl424A, triacetic acid lactone
was observed in an unreported yield in cell cultures. .

Wild Type 6-MSAS

5147 5182
CTG ATC ACC GGT GGT CTT GGC GTC CTC GGT CTC GAG

1415Leu lle Tyr Gly 911 Leu 9J1 Val Leu fily Leu Glu1

KR AxGxxG Mutant

426

5147 5182
CTG ATC ACC GGA Gm CTT GGC GTC CT C GGT CTC GAG
Leu lle Tyr GI)l Ala Leu gm Val Leu Gly Leu Glu1

1415 426
KR AxPxxA Mutant
5147 5182

CTG ATC ACC GGA GEE C'I'I'QQA GTA CTC GET CTC GAG

Leu lle Tyr GI)l Ala Leu 13m Val Leu Ala Leu Glu1

1415 426

Figure 69. Mutations made of wild type 6-MSAS.

The plasmid pMR228,10 containing the triple mutant of 6-MSAS under the ADH2
promoter, and plasmid pKOS 12-128a,8 which carries an sfp insert, were obtained from C.
Khosla and J. Kealey. The plasmids were transformed into S. cerevisiae INVScl by
electroporation and plated on synthetic medium with D-glucose and without leucine,
tryptophan or uracil for plasmid selection. Following the procedure developed by Kealey
in the production of 6-MSA,8 single colonies were first grown on synthetic medium with
D-glucose and without leucine, tryptophan or uracil for plasmid maintenance in shake
flasks. The inoculum was then transferred to rich medium with D-glucose and grown for

150 h. Although growth on rich medium facilitates plasmid loss, it has been well

145

established in S. cerevisiae that plasmid loss is relatively insignificant.12 Cell density

reached an OD600 of 16-17 and D-glucose was depleted as observed in culture

supematants by 1H NMR. Unfortunately, no triacetic acid lactone was identified by 1H
NMR. Continuous extraction of the cell free supernatant with ethyl acetate also failed to
reveal triacetic acid lactone. Formation of ethanol was verified by solvent suppression
1H N MR in D20.
Evaluation of Fatty Acid Biosynthesis

Both polyketide biosynthesis and fatty acid biosynthesis draw their precursors
from the acetyl-COA pool. The E. coli fatty acid biosynthetic machinery is classified as a
Type II system where all of the enzyme active sites are located on separate enzymes and
encoded by genes dispersed over the chromosome.13 Polyketide biosynthesis exhibits a
wide array of ketosynthase, ketoreductase, dehydrase and cyclization activities leading to
a diverse family of structures. Fatty acid biosynthesis is much more systematic and

typically only makes long Chain saturated and unsaturated fatty acids.

146

The carbonylation of acetyl-COA to malonyl-COA represents the first committed
step in fatty acid biosynthesis in E. coli (Figure 70).14 The sequence of events starts with
the ATP-dependent carboxylation of the accB-encoded biotin carboxyl carrier protein
(BCCP) by the accC—encoded biotin carboxylase using bicarbonate. The second step is
transfer of the carbonate from BCCP to acetyl-COA by transcarboxylase, a heterodimer of

the accA and ach gene products.

 

 

ATP + H003- ADP + P; ACPSH COASH
> ’
SCOA A AB CD HO SCOA FabD HO SACP

Figure 70. Carboxylation of acetyl-COA by accABCD.
Key: ACCC, biotin carboxylase; ACCB, biotin carboxyl carrier protein; AccA, Ach,
transcarbylase.

Initiation of fatty acid biosynthesis can occur by three mechanisms requiring

transacylation of malonyl-COA to malonyl-ACP (Figure 71). The first pathway couples

malonyl-ACP with acetyl-ACP to give acetoacetyl-ACP and an (equivalent of C02 by B-

ketoacyl synthase 1 (KAS H.143,15 Acetyl-ACP can be made available by first
decarboxylating malonyl-ACP by KAS I followed by coupling with a second equivalent
of malonyl-COA by KAS I as the second mechanism for initiating fatty acid biosynthesis.

As the third mechanism, B-ketoacyl synthase III (KAS 111) can couple malonyl-ACP with

acetyl—COA to make acetoacetyl-ACP.16

147

CO,, ACPSH O O

1.H OUSACP-i- /U‘SACP A MSACP

malonyl- -ACP acetyl- -ACP acetoacetyl- -ACP
CO2 O O 002, ACPSH o O
2.H OUSACP i» )LSACP + HOUSACP % MSACP
malonyl- -ACP acetyl- -ACP malonyl-ACP acetoacetyl-ACP

DU 0 COZ’COASHM
3.H SACP 4' /U\SC OA-l> SACP

malonyl—ACP acetyl-COA b acetoacetyl-ACP

Figure 71. Three mechanisms for fatty acid biosynthesis initiation.
Key: (a) B-ketoacyl-ACP synthase 1, fabB (KAS I); (b) B-ketoacyl-ACP synthase III,
fabH (KAS III).

The irreversible coupling of malonyl—ACP with an acyl-ACP is the first step of
the fatty acid elongation cycle (Figure 72).13 The B—keto group is reduced by the fabG-
encoded B-ketoreductase with NADPH. Dehydration of the resulting alcohol by fabZ-
encoded B-hydroxyacyl-ACP dehydrase followed by reduction of the 0t,B-unsaturated
thioester by enoyl-ACP reductase with NADH encoded by fabI results in the four carbon
acyl chain that can undergo subsequent coupling with malonyl-ACP, carbonyl reduction,
dehydration, and olefin reduction to add two additional carbons. This cycle is repeated

until a length is obtained that is dictated by the fatty acid biosynthetic machinery.

148

 

 

3

C) O 0“
Jk/‘L ACPSH
,ji\/fi: HO SACP:’ji~/fi\/ji _,4 ’> I‘\
SACP a,bor C\ SACP CH O O
C
NADPH C5

NADP+
HO NADH NAD’
OH O 2

j _ 0 \/_ °
’/L\/H\SACP e "//§§/HLSACP f "”“\/JLSACP

Figure 72. Chain propagation in fatty acid biosynthesis.

Key: (a) B-ketoacyl synthase 1, fabB (KAS I); (b) B-ketoacyl synthase II, fabF (KAS II);
(c) B—ketoacyl synthase III, fabH (KAS III); ((1) B-ketorecducatse, fabG; (e) B-
hydroxyacyl-ACP dehydrase, fabZ; (f) enoyl-ACP reductase, fabI.

 

 

Triacetic Acid Lactone by E. coli Fatty Acid Biosynthesis

Similar to polyketide synthesis, triacetic acid lactone has been shown to be a
derailment product of fatty acid biosynthesis. In vitro studies of fatty acid synthases from
baker’s yeast, pigeon liver, and E. coli have shown that in the absence of NADPH,
acetoacetyl-ACP coupled with malonyl-ACP to make a triketide-ACP intermediate that
cyclized to triacetic acid lactone.l7 Although triacetic acid lactone accumulation during
fatty acid biosynthesis is not significant under physiological conditions, these reports
established a direction to take in constructing a triacetic acid lactone-synthesizing
microbial host such as E. coli. An inadequate in vitro supply of reducing equivalents in
the form of NADPH was the common feature for each report of triacetic acid lactone
synthesized by purified fatty acid synthase. Our goal was then to reduce the availability
of NADPH to fatty acid biosynthesis in E. coli. This would be accomplished by reducing

or eliminating the ability of B—ketoreductase to bind NADPH.

149

Transcriptional studies of the fab cluster in E. coli by Cronan had shown that
eliminating transcription of the gene coding for B—ketoreductase activity, fabG, was a
lethal knock-out.l8 In this work, a copy of the ApR gene, a terminator cartridge
containing the CmR gene, and a truncated portion of the E. coli fabG from the 5’ end
were inserted into the E. coli genome. This results in an incomplete copy offabG with its
native promoter and a downstream, complete copy offabG lacking its native promoter
separated by polar allele duplication.19 The process occurs by a homologous
recombination event and blocks synthesis of the fabG gene product at the transcriptional
level. Because polar allele duplication failed to produce any recombinants, a conclusion
was drawn that knocking out expression of FabG was lethal to the cell.18 It was not until
an E. coli construct harboring a plasmid containing fabG from Salmonella typhimurium
for the synthesis of a catalytically functional, non-native FabG could transformation with
the suicide vector successfully block E. colifabG transcription from the genome. E. coli
YZl66 was constructed which harbored the plasmid pYZ7l expressing the S.
typhimurium fabG under the PamBAD promoter and the polar allele duplication of the E.
coli fabG on the genome.

Because fabG from S. typhimurium is under control of ParaBADv growth of YZ166
required L-arabinose for expression of the non-native FabG. It was reported that the
strain was incapable of growth in the absence of L-arabinose or in the presence of D-
glucose.18 E.coli YZl66 was grown in LB medium supplemented with 0.2% L-arabinose
in a shake flask. Upon reaching stationary phase, the cells were harvested, washed with
M9 salts to remove residual L-arabinose, and then resuspended in M9 salts containing

0.4% D-glucose and 0.002% L-fucose. In the absence of L-arabinose, transcription of

150

plasmid based fabG should cease. However, resuspended cells were unable to convert D-
glucose into triacetic acid lactone. Even more surprising was the increase in cell mass
from 1.33 g/L upon resuspension in M9 salts to 2.06 g/L after culturing for 36 h. 1H
NMR analysis of the supernatant indicated that all of the D-glucose had been consumed
and a small amount of acetate was present. It has been well established that D-glucose
inhibits transcription from PamBAD irrespective of the availability of L-arabinose.
However, the continued presence and catalytic activity of ketoreductase resulting from
transcription and translation of fabG prior to exposure of cultures to D-glucose may
explain the failure to accumulate triacetic acid lactone.

When provided with only D-glucose as a carbon source, YZl66 should be
incapable of providing its own fatty acids to make cellular membranes. Cronan had
determined that E. coli could survive when fatty acid auxotrophs were supplemented with
oleic acid.20 YZl66 was grown in LB supplemented with 0.4% D-glucose and 0.4%
oleic acid. Time points taken during the 48 h growth period did not contain triacetic acid
lactone and yet the cell density reached 2.06 g/L. Again, this came as a surprise because
PamBAD should be inactive in the presence of D-glucose. It may be that fatty acid
biosynthesis was repressed or feedback inhibited by oleic acid when cells were
supplemented with oleic acid. When YZl66 was only grown with glycerol, cellular
growth was still observed with no formation of triacetic acid lactone. Growth on

glycerol, which does not activate PamBAD, indicates that a B-ketoreductase activity must

be present.

151

Triacetic Acid Lactone by Brevibacterium ammoniagenes Fatty Acid Biosynthesis

An alternative approach to triacetic acid lactone biosynthesis relies on expression
Of mutated, non-native fatty acid machinery in E. coli. This approach would allow the
native fatty acid machinery to make fatty acids for cell viability while the non-native
machinery could be modified to make triacetic acid lactone. Critical to the success of this
approach is an absence of exchange of biosynthetic intermediates between the native and
non-native fatty acid biosynthetic pathways. Unlike the E. coli Type II fatty acid
synthase in which each of the catalytic domains is on separate enzymes, the Type I fatty
acid synthase from Brevibacterium ammoniagenes is a single, multifunctional enzyme
that catalyzes all of the necessary steps for fatty acid synthesis.21 FAS-A, the gene
product offasA, is capable of producing both saturated and unsaturated fatty acids while
FAS-B, encoded by fasB, is limited to only synthesis of saturated fatty acids. By a
serendipitous discovery, functional expression of either gene in E. coli requires co-
expression with the B. ammoniagenes PPTI gene.22 PPTI encodes for the
phosphopantetheine transferase required for the conversion of the non-functional apo-
protein into the functional holo-protein.

The lack of literature precedent for the formation of triacetic acid lactone using
FAS—B from B. ammoniagenes necessitated establishing that this fatty acid synthase had
capabilities similarly shown with fatty acid synthases from pigeon liver17a, baker’s
yeast17b and E. coli.”C The plasmid pGM44 containing fasB under its native promoter
with the PPTI from B. ammoniagenes was obtained from E. Schweizer and transformed
into E. coli strain DHSOt.21C FAS-B was partially purified to remove the E. coli fatty acid

synthase enzymes in the crude lysate. Purification consisted of reserving the 30—60%

152

(NH4)2SO4 precipitate followed by centrifuging the resuspended pellet at 100 000g.21b
The active form of FAS-B is an 0% homohexamer made up of subunits with an estimated
weight of 327,500 daltons.21a The large size of active FAS-B should be precipitated in
the ultracentn'fugation step and leave the much smaller'E. coli fatty acid synthases carried
over from the (NH4)2SO4 precipitation suspended in the buffer. As a control, protein
from DH50t without pGM44 was subjected to the same purification procedure to compare
if native enzymes carried over during the partial purification of the lysate were capable of

making triacetic acid lactone.

 

triacetic acid lactone —> '

.1 r.

 

l 2 3 4 5
Figure 73. X-ray film exposed to radioactive extracts from fatty acid synthases.
Key: Lane 1, DH50t/pGM44 with NADPH; Lane 2, DHSOt/pGM44 without NADPH;
Lane 3, no enzyme with NADPH; Lane 4, DH50t with NADPH; Lane 5, DHSOL without
NADPH.
The presence of FAS-B after partially purifying the lysate was first verified by
measurement of the B-ketoreductase activity. The assay measured the depletion of

NADPH at 340 nm over time when malonyl—COA was added to a mixture of the cofactor,

acetyl—COA, and enzyme.23 The enzyme sample from DH5a/pGM44 exhibited a

153

measurable activity of 3 mU whereas the enzyme sample from DHSOt exhibited no

activity. Biosynthesis of triacetic acid lactone was detected using 14C radioisotope
labeled [1-14C] acetyl-COA. A mixture of malonyl-COA and 0.5 “Ci of [1-14C] acetyl-
CoA were treated with partially purified enzyme samples from DHSOt/pGM44 and DHSOL
without pGM44.21C For both enzyme samples, experiments were conducted in the
presence or the absence of NADPH. A fifth experiment without enzyme was conducted
to determine the fate of acetyl-COA and malonyl-COA in the assay medium at room
temperature. At the termination of the assay, the solutions were doped with authentic
triacetic acid lactone prior to extraction of the organic soluble products into ethyl acetate.
A TLC of the extracts had indicated that the UV active spot for triacetic acid lactone had
a corresponding spot of the same Rf after exposure to x-ray film for only the experiments
using the enzyme sample from DHSOt/pGM44 (Figure 73). The enzyme sample from
DHSOt and the lane with no enzyme showed no radioactive spot for triacetic acid lactone.
The experiments indicated that like other fatty acid synthases, FAS-B from B.
ammoniagenes was capable of making triacetic acid lactone in vitro and was a reasonable
candidate for making triacetic acid lactone in vivo.

The putative location of the B-ketoreductase site was proposed to include the
region of nucleotides 6425-6440 of fasB based on conserved amino acid sequences from
established B-ketoreductase amino acid sequences isolated from eukaryotic hosts.2121 The
formation of the binding pocket for NADPH in the tertiary structure of an enzyme is
associated with a specific amino sequence known as the Rossman fold characterized by a
GxGxxG amino acid sequence.11 FAS-B from B. ammoniagenes contains a GGxxG

sequence. Although the sequence is not the required GxGxxG sequence of the Rossman

154

fold, the GGxxG sequence is homologous with Rossman fold regions found within the

known B—ketoreductases (Figure 74).21a

1. EVAVVTE§:c_§:_'s::rf:C:§§IASEIVANLLREGATVIATTSRLG

 

2.KYVLITGAMKGSICMEVLQGLLQGGAKVVVTTSRFs

 

 

3.KNVLMTGAMAGSIGMEVLQGLISGGAQVIVTTSRFS

 

 

4.sv-IITGELGGFGh-ELARWLVLRGAQRLVLTSRSG

 

 

5.SYoIITdGLCCFdL-ELAQWLIERGAQKLVLTSRSG

 

 

6.TY-LITdGLGVLflL-EVADFLVEKCARRLLLISRRA

 

Figure 74. Amino acid sequence comparisons of B-ketoreductase sites for fatty acid
biosynthesis and polyketide biosynthesis.

Key: 1. FAS-B, B. ammoniagenes; 2. FAS, S. cerevisiae; 3. FAS, P. patulum; 4. FAS, rat;
5. FAS, chicken, 6. MSAS, P. patulum. Conserved amino acids (boxed) were related to
B. ammoniagenes FAS (solid lines) or to specific subgroups of other fatty acid

synthetases (dashed lines)?”1

Tampering with the putative Rossman fold would require mutation of the GGxxG
enzyme sequence and subsequently alteration Of the NADPH binding pocket leading to
loss of B-ketoreductase activity. With FAS-B being a multifunctional enzyme, simply
adding or deleting nucleotides from the gene may adversely affect the overall folding of
the protein and thus jeopardize key active sites for triacetic acid lactone synthesis to
occur. From the radioisotope experiments, it was now understood that the B—ketoacyl
synthase could couple malonyl-ACP with acetoacetyl-ACP to make the triketide
intermediate required for cyclization to make triacetic acid lactone.
Site-directed Mutagenesis of fasB

Efforts were then focused on mutating the putative Rossman fold portion of the B-
ketoreductase domain. The 9.5-kb fasB gene was too large for site-specific mutagenesis
by common PCR protocols. Mutations were introduced into the fasB gene by overlap

extension PCR (Figure 75).24 The first step requires amplifying two overlapping

155

fragments of DNA, fragments AB and CD, incorporating the nucleotide Changes in the
overlapping region with primers B and C in separate amplification experiments. After
purifying the amplified DNA, fragments AB and CD are mixed together, denatured, and
reannealed. The top strand of the AB fragment can prime with the bottom strand of the
CD fragment and enter into non-exponential amplification of fragment AD. The addition
of primers A and D exponentially amplifies the AD fragment. The resulting AD
fragment containing the nucleotide changes introduced by the overlapping region of
primers B and C is then introduced into the open reading frame to re-establish the
modified gene.

The large size offasB and the inability to conveniently introduce the entire fasB
mutant into a vector precluded conducting the site-directed mutagenesis experiment by
PCR for the entire gene. Because only a few mutations within a small region of the gene
were to be introduced, a fragment offasB could be mutated by PCR then reintroduced
into the gene after excising the non-mutated fragment. Unfortunately, there were no
unique restriction sites to work with a fragment of DNA of a reasonable size. A 2.0 kb
fragment could be excised from fasB by digestion with XbaI and ScaI. A second ScaI
site was located in the A pR gene, which complicated the cloning process. Because
pGM44 required insertion of serA as a selection marker for use in fed-batch
fermentations, a kanRserA cassette was assembled and introduced into the Seal site in the
ApR gene to afford pCH4.267A (Figure 76). By placing the kanRserA cassette in the Seal
site of ApR in pGM44, the nutritional pressure marker serA would now be on the plasmid,

a new antibiotic selection would be available to reduce contamination problems in

156

inoculums, and a 2.0 kb fragment of fasB could be mutated and subsequently re-

introduced into the gene with fewer complications.

 

 

 

 

 

 

 

 

 

 

A C
5'4 4A 3|
3' VT f 5'
D
B
1 PCR3
5| * 3'
3' * 5'
AB 5' * 3|
3' * CD 5
l denature, reannealb
5L AB * 3'
3| * CD 5|
D
1 PCR"
51 a 3'
3' AD * 5

Figure 75. Overlap extension PCR.

3 PCR of AB fragment and CD fragment are carried out separately. The B primer carries
the mutation for the AB fragment and the C primer carries the mutation for the CD

fragment.

b Fragements AB and CD are purified then combined as the new templates.

0 PCR is carried out with AB and CD fragments as template and primers A and D for
amplifying the mutated DNA fragment.

157

pCH4.184A

PCR
EcoRV Eco RV
0'31 I . 3.3-kb I
kan" serA
Rsrl l
EcoRV digest

 

i) Scal partial digest
ii) CIAP treatment

 

 

 

   
   
    

pCH4.267A
16.8-kb

Sca1

fasB

Figure 76. Construction of pCH4.267A.

158

Table 14. Primers for mutation of fasB. Mutations are underlined.

 

 

 

primer sequence
A 5'-CTCG GCGCGTG AAGACCTCGTG
D 5'-AGGCCCAATTCCG CAG CCAAGC
B-1 5'-GAA§CACCGGTG ACAACAGCAA
C-1 5'-‘ITGTCACCGGTGQTTCGCCTGG
B-2 5'-AGAQAGG CGAAQCAACGGTGACAACAG CAAC
C-2 5'-CG1_'TG QT TCGCCTQICTCTATTGCCTCGGAA
B-3 5'-AGAQAGG CGIAQCAACG GTGACAACAGCAAC
C-3 5'-CGII'GQT_A_CGCCT91CTCTA‘ITG CCTCGGAA

 

 

According to Khosla, mutating 6-MSAS to make triacetic acid lactone required
replacing all three glycines in the Rossman fold with alternative amino acids.10 Our
approach was to start with a single mutation as a conservative measure to be sure the
technique worked before attempting to make multiple mutations. The 2.0-kb C-l-D
fragment was amplified using primers C-l and D from Table 14. The primer C-l
exchanged nucleotide (36429 for a C resulting in converting a glycine to an alanine in the
translated protein (Scheme 77). The 187 bp AB-l fragment of the single mutant was
amplified with primers A and B-1. The primer B-l exchanged a C for a G in the
complimentary sequence of fasB. The fragments AB-l and C-l-D, containing the single
nucleotide mutation, was designed to contain a 16 bp complimentary region at their
respective 3’ ends. The 2.2-kb A-l-D fragment was amplified by using fragments AB-l
and C-l-D as the template with primers A and D. The A-l-D fragment was digested with
XbaI and Seal to liberate a 2.0-kb fragment. The mutated A-l-D fragment was then

introduced into pCH4.267A digested with XbaI and Sca1 to afford pCH5.28M1

159

containing the fasB single mutant. The mutated gene sequence, and subsequent

mutations, were confirmed by dye terminator sequencing at the DNA sequencing facility
at Michigan State University using primer A.

c-1
6410 TT GTC ACC GGT GgT TCG CCT GG 6448

5'-GTT GCT GTT GTC ACC GGT GGT TCG CCT GGC TCT ATT GGC—3'
Val Ala Val Val Thr Gly Gly Ser Pro Gly Ser Ile Ala
AA CGA CAA CAG TGG CCA C§A AG

B-1

,1 l

5'—GTT GCT GTT GTC ACC GGT GQT TCG CCT GGC TCT ATT GGC—3'
Val Ala Val Val Thr Gly Ala Ser Pro Gly Ser Ile Ala

 

Figure 77 . Mutant M1 of fasB from B. ammoniagenes.

The second FAS-B mutant incorporated three amino acid changes in the protein
sequence. The 2.0 kb C-2-D fragment was amplified with primers D and C-2 from Table
14. The primer C-2 exchanged G6426 with a T, G6429 with a C, and 664375433 with a C
and a T, respectively. The nucleotide changes resulted in altering GZO63V, G2064A and
G2066L in the translated protein (Figure 78). The mutation was then introduced into

pCH4.267A as carried out previously to afford pCH5.110M2 containing the fasB triple

mutant.

160

C-2

6410 C GIT GCT TCG CCT QIC TCT ATT GCC TCG GAA

5'—GTT GCT GTT GTC ACC GGT GGT TCG CCT GGC TCT ATT GCC TCG GAA-3'
Val Ala Val Val Thr Gly Gly Ser Pro Gly Ser Ile Ala Ser Glu

CAA CGA CAA CAG TGG CéA ch AGC GGA géo A
B-2

.. l

5'-GTT GCT GTT GTC ACC GET GQT TCG CCT 91C TCT ATT GCC TCG GAA—3'
Val Ala Val Val Thr Val Ala Ser Pro Len Ser Ile Ala Ser Glu

 

Figure 78. Mutant M2 of fasB from B. ammoniagenes.

The final FAS-B mutant incorporated four amino acid changes in the protein
sequence. The 2.0 kb C-3-D fragment was amplified using primers D and C-3 from
Table 14. The primer 03 made one additional change from primer C-2 by exchanging

T6431 to an A. The additional nucleotide change altered 82065T (Figure 79). The

mutation was then introduced into pCH4.267A as carried out previously to afford

pCH5.111M3 containing the fasB quadruple mutant.

C-3

6410 C GIT GQT ACG CCT QIC TCT ATT GCC TCG GAA

5'-GTT GCT GTT GTC ACC GGT GGT TCG CCT GGC TCT ATT GCC TCG GAA-3'
Val Ala Val Val Thr Gly Gly Ser Pro Gly Ser Ile Ala Ser Glu

CAA CGA CAA CAG TGG CéA ch goo GGA gas A
B-3

.3 l

5'—GTT GCT GTT GTC ACC GET GQT ACG CCT QIC TCT ATT GCC TCG GAA-3'
Val Ala Val Val Thr Val Ala Thr Pro Lau Ser Ile Ala Ser Glu

 

 

Figure 79. Mutant M3 of fasB from B. ammoniagenes.
The identification of triacetic acid lactone in the radioisotope experiments using

wild-type FAS-B assayed in the absence of NADPH established the procedure for

161

identifying triacetic acid lactone in the supernatant of a fed-batch fermentation using the
plasmid pCH4.267A. The plasmid, also containing a copy of E. coli serA locus for
plasmid maintenance, was transformed into RB791 serAzzaroB and grown in the
fermentor under D-glucose limitation. The cell mass reached 31.1 g/L after 24 h, but
after 48 h, no triacetic acid lactone was observed by 1H NMR. The construct RB791
serA::aroB/pCH5.28Ml, containing the fasB single mutant, also failed to produce
triacetic acid lactone under fed-batch fermentation conditions. The plasmid
pCH5.110M2 and pCH5.111M3, containing fasB mutations encoding for the triple and
quadruple mutant, respectively, would not grow when freshly transformed into RB791
serAzzaroB then plated directly onto M9/D-glucose plates. E. coli strains harboring either
plasmid would only grow on LB using antibiotic pressure for plasmid maintenance. Cells
transformed with pCH5.110M2 or pCH5.111M3 and plated onto LB would subsequently
grow on M9/D-glucose in liquid medium. Attempts to grow RB791 serA::aroB with
pCH5.110M2 or pCH5.111M3 in the fermentor were unsuccessful due to very poor
growth characteristics. SDS-Page analysis of cellular extracts revealed that cells
expressing fasB with multiple mutations failed to synthesize a denatured protein of
327,500. daltons as observed with gene products from pGM44, pCH4.267A and

pCH5.28Ml.

162

pGM44

 

 

 

PCR
Ndel RSI“ NOfl Hindlll
I \/
partial FAS-B
Hindlll
Nde1
i) digestwith Ndel digest with Ndel
and Hindlll and Hindlll

i0 CIAP treatment

 

  
 
   

. Hindlll
Notl

Scal

pCH5.213

Figure 80. Construction of pCH5.213.

163

pCH4.184A

PCR
EcoRV EcoRV
pCH 5 .213 Q=—l4
ken" serA
i) Scal digest
ii) CIAP treatment EcoRV dgest

 

 

Hindlll
Notl

Figure 81. Construction of pCH5.287 A.

164

pCH5.287A pGM44

i) digest with R31“, Non i) digest with Fisrll, Nod
ii) CIAP treatment ii) gel purify
iii) gel purify

 

 

 

Figure 82. Construction of pCH5.303.

The activity of FAS-B from B. ammoniagenes in our hands was two orders of
magnitude lower than what was reported.21C Transcription is initiated by the native
promoter from B. ammoniagenes.21C Although the promoter was found to work in E.
coli, the promoter did not allow any control of expression of the enzyme. The T7

promoter was used to overexpress FAS-B. The fasB gene in pGM44 was not amendable

165

for cloning into the vector pT7-7 for transcription from the T7 promoter. The plasmid
pT7-7 incorporates the restriction site Ndel for transcription of genes with the T7
promoter, which contains ATG as part of its recognition sequence.25 The start codon of
fasB is the rare GTG.21a A 1.5-kb fragment was amplified from pGM44 incorporating a
mutation to change the start codon from GTG to the more common ATG making the
restriction site Ndel available for cloning into pT7-7. The 3’ end of the amplified
fragment incorporated NotI and HindIH recognition sequences. Amplification from the
3’ end was started 500-bp downstream from a unique RerI restriction site in fasB. The
amplified DNA was ligated into the pT7-7 in the NdeI and Hindlll sites, resulting in the
plasmid pCH5.213 (Figure 80). Prior to reassembling fasB on pT7-7, pCH5.213 was
digested with Sca1, located in the A p” gene of pT7-7, followed by insertion of the
kanRserA cassette to afford pCH5.287A (Figure 81). Digestion of pGM44 with RerI and
NotI liberated a 9.0-kb fragment containing the remainder of fasB and PPTl which was
inserted into the RerI and Natl double digested pCH5.287A to give pCH5.303 with fasB

under P77 (Figure 82). Although restriction digestion of pCH5.303 shows fasB was

present, and sequencing of the junction in the region of Rer was correct, no activity was
observed when JWF1(DE3)/pCH5.303 was grown under fed-batch fermentor conditions.
Protein expression was subsequently checked by SDS/Page gel, but the 327,500 Da
protein was not present.

When expression of the triple and quadruple mutants failed, efforts were then
focused towards the strategy of employing non-native fatty acid machinery to sustain cell
viability in E. coli while the native E. coli fatty acid synthase was exploited for triacetic

acid lactone biosynthesis. FAS-B was not capable of synthesizing unsaturated fatty acids

166

making it insufficient for sustaining cellular requirements. The gene fasA from B.
ammoniagenes, encoding FAS-A, is capable of making both saturated and unsaturated
fatty acids.21b’C The gene has been localized in the plasmid pHPS6 with PPTI required
to convert the apo-protein to the holo form.22 ‘ Knock-out experiments with B.
ammoniagenes had shown FAS-A could alone meet the fatty acid requirements of B.
ammoniagenes while FAS—B failed.22

Biosynthesis of triacetic acid lactone by native E. coli fatty acid synthase would
require knocking out fabG, the gene responsible for encoding the B-ketoreductase step of
fatty acid biosynthesis. A derivative of pMAK705, labeled pCH5.263A, was created
containing fabGzzkanR from E. coli.26 The plasmid contained a copy of CmR and the
pSCl temperature sensitive replicon. As Cronan had shown, a knock-out of fabG was a
lethal mutation. To circumvent this issue, pCH5.263A was transformed into the serA
auxotroph JWF1 harboring pHPS6 which allowed FAS-A to provide the essential
saturated and unsaturated fatty acids when the E. coli fatty acid synthetic capability was
lost during the homologous event with pCH5.263A. Several transformations were
attempted followed by plating on LB/Cm/Kan/Ap pre-warmed at 42 °C with no
formation of colonies after three days. When transformations were incubated at 30 °C,
the temperature sensitive pCH5.263A was stable to replication and 37 colonies were
observed within 24 h. Replication of the colonies on LB/Cm/Kan/Ap at 42 °C resulted in
4 homologous recombination events after 24 h. Promotion of the second homologous
recombination event by growing the replicates in LB media with only Ap at 30 °C or a

combination of Ap and Kan did not result in a successful knock-out of the native E. coli

167

fabG. Colonies obtained on plates at 42 °C after several serial dilutions at 30 °C failed to
afford a colony resistant to Ap and Kan while exhibiting sensitivity to Cm.
Discussion

Whole-cell microbial synthesis of triacetic acid lactone remains elusive. One
issue to consider is the ability of cells to export triacetic acid lactone into the supernatant.
When Penicillium stipitatum is exposed to ethionine, an inhibitor of the methylation step
in tropolone biosynthesis, triacetic acid lactone is observed to accumulate in the culture
supernatant.27 This observation establishes that triacetic acid lactone can be exported
into culture supematents and that triacetic acid lactone does not possess any unexpected,
high-level toxicity towards microbial growth and metabolism.

The advantage of using 6-MSAS as the system for triacetic acid lactone
biosynthesis is that 6-methylsalicylic acid is not required for microbial growth or
metabolism. Mutations made to 6-MSAS would not be expected to jeopardize a host’s
ability to survive or require supplementation to counter the mutation. In our hands, the 6-
MSAS mutant failed to make either 6-MSA or triacetic acid lactone. Sequencing of the
region of the gene containing the mutations showed that the correct mutations had been
introduced. It is possible that S. cerevisiae is unable to express an active form of the
mutated 6-MSAS. Either no 6-MSAS is expressed at all, or co-expression of the 31p-
encoded gene product did not convert the apo-ACP to the active holo-ACP. An
SDS/Page gel would verify the presence of the expressed gene products and the extent to
which each enzyme is produced in S. cerevisiae using the ADH2 promoter. If protein

expression is verified by gel electropheresis, then an enzyme assay with 14C radioisotope

168

labeled [1-14C] acetyl-COA could be used to measure the activity of the mutant 6-MSAS
and its capability to make triacetic acid lactone.

If a fabG knock-out is understood as being a lethal mutation to E. coli, then the
construct YZl66/pYZ71, a fabG knock-out harboring a plasmid containing the S.
typhimurium fabG under PamBAD, should not grow in the absence of L-arabinose.
Cronan reported the construct showed no signs of growth in 24 h when the construct was
grown on LB containing 0.4% D-glucose and 0.002% L—fucose.18 Indeed, in our hands, a
similar observation was observed. If the cells are allowed to grow for an additional 12 h,
cell growth was eventually observed and the final cell density reached normal levels.
This calls into question whether the transcription of the native fabG is permanately

blocked or if ParaBA D is a leaky promoter in the absence of L-arabinose. Transcription of

fabG was blocked by a genomic insertion of ApR and CmR between fabG and its native
promoter by a homologous recombination event. Cells deprived of L-arabinose may
respond by reversing the homologous recombination event to re-establish transcription of
the native fabG. To determine if this were the case, genomic DNA would need to be
isolated from the cells and sequencing for ApR and CmR and their position relative to
fabG and its promoter sequence. A more likely explanation is that the repressor protein
does not completely block transcription from ParaBAD- Even without L-arabinose, there
may be enough basal level transcription of the non-native fabG to maintain microbial
growth albeit at a slower rate.

Employment of FAS-B from B. ammoniagenes failed to make triacetic acid
lactone in vivo. The protein sequence designated as the putative site for the B-

ketoreductase domain based on sequence homologies with known B-ketoreductases from

169

eukaryots did not contain the GxGxxG sequence associated with the Rossman fold of the
nucleotide binding site.2181 Attempts at mutating the GGxxG region, a region with good
homology with the established Rossman fold sites in the alternative B-ketoreductases, did
not result in triacetic acid lactone accumulation in culture supematents. The fact that the
triple and quadruple mutants failed to make the 327,500 dalton denatured protein
suggests additional mutations may have occurred during the site directed mutagenesis of
the 2.0-kb fragment of fasB resulting in a truncated protein. Failure of both mutants to
grow on M9 medium when freshly transformed into E. coli may suggest a new protein
was translated that might be toxic to E. coli.

The alternative approach of using the native E. coli fatty acid synthase would
require the expression of a non-native fatty acid synthase to meet saturated and
unsaturated fatty acid requirements. The capability of FAS-A to independently sustain
growth in B. ammoniagenes established a third route to pursue in employing fatty acid
biosynthesis to make triacetic acid lactone”),c Two explanations can be offered for the
inability to make an E. coli fabG knock-out by homologous recombination with
fabGzzkanR. The length of fabG is 0.75-kb which provides only a small region of
homology for recombination into the genome compared to the 1.0-kb kanR fragment that
was to be inserted into the center of the gene. Efforts to include the fabD gene on the 5’
end and the acpP gene on the 3’ end of fabGzzkanR could not be inserted into pMAK705
for subsequent attempts at homologous recombination into the E. coli chromosome. The
second issue concerns the ability of FAS—A to sustain growth in E. coli. There are
temperature sensitive-mutants of E. coli reported by Cronan that exhibit an inability to

grow at 42 °C without unsaturated fatty acid supplementation.20b If FAS-A can provide

170

fatty acids for E. coli, then fatty acid synthase temperature sensitive mutants harboring
pHPS6 containing fasA and PPTI from B. ammoniagenes should not require fatty acid
supplementation at the restrictive temperature. If FAS-A can provide fatty acids for E.
coli, efforts should be directed again at introducing the fabD-fabGzzkanR-acpP fragment

into pMAK705 for homologous recombination.

171

References

1 (a) Abaecherli, C.; Miller, R. J. In Kirk-Othmer Encyclopedia of Chemical
Technology; Kroschwitz, J. I.; Howe-Grant, M., Eds.; Wiley: New York, 1991; Vol. 14,
pp 954-978. (b) Weissermel, K.; Arpe, H.-J. In Industrial Organic Chemistry, 3rd ed.;
VCH: New York, 1997: p. 180-183.

2 Collie, J. N. J. Chem. Soc. 1891, 59, 607.
3 Opie, T. R. Patent EU 0,059,052, 1982.

4 (a) Dimroth, P.; Walter, H.; Lynen, F. Eur. J. Biochem. 1970, I3, 98. (b) Scott, A.
1.; Phillips, G. T.; Kircheis, U. Bioorg. Chem. 1971, I, 380.

5 (a) Gaisser, S.; Trefer, A.; Stockert, S.; Kirschining, A. Bechtold, A. J. Bacterial.
1997, 179, 6271. (b) Hopwood, D. A.; Sherman, D. H. Annu. Rev. Genet. 1990, 24, 37.

6 (3) Beck, J .; Ripka, S.; Siegner, A.; Schiltz, B.; Schwiezer, E. Eur. J. Biochem.
1990, I92, 487. (b) Spencer, J. B.; Jordan, P. M. Biochem. J. 1992, 288, 839. (c) Schorr,
R.; Mittag, M.; Muller, G.; Schweizer, E. J. Plant Physiol. 1994, I43, 407.

7 Bedford, D. J .; Schweizer, B.; Hopwood, D. A.; Khosla, C. J. Bacterial. 1995,
I 77, 4544.

8 Kealey, J. T.; Liu, L.; Santi, D. V.; Betlach, M. C.; Barr, P. J. Proc. Natl. Acad.
Sci. USA 1998, 95, 505.

9 Nakano, M. M.; Corbell, N.; Besson, J .; Zuber, P. Mal. Gen. Genet. 1992, 232,
313.

10 Richardson, M. T.; Pohl, N. L.; Kealey, J. T.; Khosla, C. Metab. Eng. 1999, I,
180.

11 (a) Chen, Z.; Lu, L. Shirley, M.; Lee, W. R.; Chang, S. H. Biochemistry 1990, 29,
1112. (b) Rescigno, M.; Perham, R. N. Biochemistry 1994, 33, 5721.

12 (a) Hardjito, L.; Greenfield, P. F.; Lee, P. L. Enzyme Microb. Technol. 1993, 15,
120. (b) Koo, J. H.; Kim, S.-Y.; Park, Y.-C.; Han, N. S.; Seo, J.-H. J. Microbial.
Biotechnol. 1998, 8, 203.

13 Magnuson, K.; Jackowski, S.; Rock, C. 0.; Cronan, J. B., Jr. Microbial. Rev.
1993, 57, 522.

172

14 (a) Alberts, A. W.; Bell, R. M.; Vagelos, P. R. J. Biol Chem. 1972, 247, 3190. (b)
Gucchait, R. B.; Polakis, B.; Dimroth, P.; Stoll, B.; Moss, J.; Lane, M. D. J. Biol. Chem.
1974, 249, 6633.

15 Rosenfield, I. S.; D’Angelo, G.; Vagelos, P. R. J. Biol. Chem. 1973, 248, 2452.

16 Tsay, J .-T.; Oh, W.; Larson, T. J.; Jackowski,'S.; Rock, C. O. 1992, 267, 6807.

17 (a) Brodie, J. D.; Wasson, G.; Porter, J. W. J. Biol. Chem. 1964, 239, 1346. (b)
Yalpani, M.; Willecke, K.; Lynen, F. Eur. J. Biochem. 1969, 8, 495. (c) Brock, D. J. H.;
Bloch, K. Biochem. Biophys. Res. Comm. 1966, 23, 775.

18 Zhang, Y.; Cronan, J. E., Jr. J. Bacteriol. 1998, 180, 3295.

19 Zhang, Y.; Cronan, J. B., Jr. J. Bacteriol. 1996, 178, 3614.

20 (a) Harder, M. B.; Beacham, I. R.; Cronan, J. E., Jr.; Beacham, K.; Honegger, J.
L.; Silbert, D. F. Proc. Natl. Acad. Sci. USA 1972, 69, 3105. (b) Cronan, J. B., Jr.;
Gelmann, E. P. J. Biol. Chem. 1973, 248, 1188.

21 (a) Meurer, G.; Biermann, G.; Schutz, A.; Harth, S.; Schweizer, E. Mol. Gen.
Genet. 1992, 232, 106. (b) Stuible, H.-P.; Wagner, C.; Andreou, 1.; Huter, G.;

Haselmann, J .; Schweizer, E. J. Bacterial. 1996, 178, 4787. (c) Stuible, H.-P.; Meurer,
G.; Schweizer, E. Eur. J. Biochem. 1997, 247, 268.

22 Stuible, H.-P.; Meier, S.; Schweizer, E. Eur. J. Biochem. 1997, 248, 481.
23 Lynen, F. Methods Enzymal. 1969, 14, 17.

24 Horton, R. M.; Pease, L. R. In Directed Mutagenesis: A Practical Approach;
McPherson, M. J ., Ed.; Oxford University Press: New York, 1991; pp.217-247.

25 Tabor, S.; Richardson, C. C. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 1074.

26 Hamilton, C. M.; Aldea, M.; Washbum, B. K.; Babitzke, P.; Kushner, S. R. J.
Bacteriol. 1989, 171, 4617.

27 Bentley, R.; Zwikowits, P. M. J. Am. Chem. Soc. 1967, 89, 676.

173

Chants);
EXPERIMENTALS

General Methods

General Chemistry

All reactions sensitive to air and moisture were carried out in oven and/or flame
dried glassware under positive argon pressure. Air or moisture sensitive reagents and
solvents were transferred to reaction flasks fitted with rubber septa via syringes or
cannula. Unless otherwise specified, all reactions were carried out at room temperature.
Solvents were removed using either a BUchi rotary evaporator at water aspirator pressure
or under high vacuum (0.5 mm Hg). Hydrogenations were performed on a Parr

hydrogenation apparatus under 50 psi of hydrogen at rt unless otherwise specified.

Reagents and Solvents

DMF, DMSO, hexanes and acetone were dried and stored over activated Linde 4
A molecular sieves under nitrogen. Benzene and CHzClz were distilled from calcium
hydride under nitrogen. MeOH was distilled from sodium metal under argon and stored
over Linde 4 A molecular sieves under argon. THF and diethyl ether were distilled under
nitrogen from sodium benzophenone ketyl. Water used in purifications was glass
distilled and deionized. All reagents and solvents were used as available from
commercial sources or purified according to published procedures. Organic solutions of

products were dried over anhydrous MgSO4.

174

Chromatography

Radial chromatography was carried out with a Harrison Associates Chromatotron
using 1, 2 or 4 mm layers of silica gel 60 PF254 containing gypsum (E. Merck). Silica gel
60 (40-63 pm, E. Merck or Spectrum Chemicals) Was used for flash chromatography.
Analytical thin-layer chromatography (TLC) utilized precoated plates of silica gel 60 F-
254 (0.25 mm, Whatman). TLC plates were visualized by immersion in anisaldehyde
stain (by volume: 93% ethanol, 3.5% sulfuric acid, 1% acetic acid and 2.5%
anisaldehyde) or vanillin stain (by volume: 98% ethanol, 1% sulfuric acid, 1% vanillin)
followed by heating.
Spectroscopic and Analytical Measurements

1H NMR and 13C NMR spectra were recorded on a Varian VX-3OO FT -NMR
spectrometer. Chemical shifts for 1H NMR spectra were reported in parts per million
(ppm) relative to CHDzCOCD3, 5 = 2.04 ppm with CD3COCD3 as solvent, relative to
internal tetramethyl silane (Me4Si, 5 = 0.0 ppm) with CDCl3 as the solvent, and to
sodium 3-(trimethylsilyl) propionate-2,2,3,3-d4 (TSP, 5 = 0.0 ppm) or water (DHO, 8 =
4.67 ppm) when D20 was the solvent. The following abbreviations are used to describe
spin multiplicity: s (singlet), d (doublet), m (unresolved multiplet), dd (doublet of
doublets). Chemical shifts for 13C NMR spectra are reported in ppm relative to
CD3COCD3 (CD3COCD3, 8 = 29.8 ppm), relative to CDC13 (CDCl3, 5 = 77.0 ppm), or
internal standard methanol (CH3OH, 5 = 49.0 ppm) in D20. Fast atom bombardment

(FAB) mass spectra were obtained on a double-focusing Kratos M850 mass spectrometer
employing glycerol as the matrix at the University of South Carolina. Elemental analysis

were performed by Atlantic Microlab Inc. (Norcross, GA). UV and visible spectra were

175

recorded on a Perkin-Elmer Lambda 3B spectrometer or a Hewlett Packard 8452A diode
array spectrometer.
Protein Purification

General Information

Cells were harvested by centrifugation at 3 000g, 6 min, 4 °C unless otherwise
specified. Cell lysis was achieved by two passes through a French pressure cell (SLM
Aminco) at 16 000 psi and the cellular debris was separated from the lysate by
centrifugation (48 000g, 25 min, 4 °C). Protein concentration was quantified using the

Bradford dye-binding procedure1 with assay solution from Bio-Rad.

MIP synthase

Partial purification of cellular lysate was required to quantify myo-inositol 1-
phosphate synthase (MIP synthase) activity over background cellular phosphatase
activity.2 Cells were collected from 30 mL of fermentation broth by centrifugation at 3
700g for 6 min at 4 °C. Cells were resuspended in 10 mL of resuspension buffer
consisting of Tris-HCl (20 mM), pH 7.7, NH4Cl (10 mM), 2-mercaptoethanol (B-ME, 10
mM), phenyl methyl sulphonyl fluoride (PMSF, 0.5 mM), and EDTA (1 mM).
Resuspended cells were frozen at -80 °C for up to 4 days until purification was carried
out. PMSF was prepared by dissolving in 1 mL ethanol prior to adding to the prepared
.buffer.

Thawed cells were lysed by two passages through a French press at 16 000 psi.
Cellular debris was removed by centrifugation at 48 000g for 30 min at 4 °C. Clarified

cellular lysate containing approximately 200 mg of protein was loaded onto a DEAE

176

cellulose column (5 x 25 cm) at 4 °C. The DEAE column was prepared by eltuing with

200 mL each of the following solutions in the order listed: 1) 4M urea (aq), 2) H20, 3)

1M NaOH (aq), 4) 1 M NaCl (aq), 5) H20, 6) Buffer A. The column was eluted with a
gradient of NH4Cl in the following buffer (Buffer A): Tris-HCl (20 mM). pH 7.7, B-ME

(10 mM), PMSF (0.5 mM), EDTA (1 mM), and NH4Cl (10 mM); (Buffer B): Tris-HCl
(20 mM), pH 7.7, B-ME (10 mM), PMSF (0.5 mM), EDTA (1 mM), and NH4C1 (250
mM). The gradient consisted of 200 mL of Buffer A and 200 mL Buffer B. Fractions (9
mL) were collected throughout the gradient. Each column was cleaned between enzyme

samples from the same fermentation by eluting with 200 mL Wash Buffer followed by

200 mL Buffer A. Wash Buffer consisted of the following: Tris-HCl (20 mM), pH 7.7,

B—ME (10 mM), PMSF (0.5 mM), EDTA (1 mM), and NH4Cl (500 mM). The columns

were cleaned more rigourously between fermentation batches by following the protocol
for preparing the columns.

Fractions were checked for MIP synthase activity without background
phosphatase activty. MIP synthase assay buffer was prepared by combining the
following: 10 mL Tris-Cl (50 mM), pH 7.7, NH4Cl (2 mM), DTT (0.2 mM); 1 mL D-
glucose 6-phosphate, sodium salt (100 mM) in H20; 0.2 mL NAD+ (100 mM) in H20.
For determining fractions containing MIP synthase free of background phosphatase, 200
”L assay buffer was combined with 100 uL of collected enzyme fraction. Following
incubation at 37 °C for 1 h, enzyme reactions were stopped by the addition of 50 ILL 20%
trichloroacetic (w/v) acid. Precipitated protein was removed by centrifugation and the

activities were assessed by the presence of released inorganic phosphate using the

colorimetric methodof Ames.3

177

From each enzyme reaction sample, 100 uL was combined with 200 [1L H20 and
a second 100 uL was combined with 100 uL an aqueous solution of 200 mM NaIO4.
The divided samples were incubated at 37 °C for l h. Samples containing NaIO4 were
quenched with 100 uL of aqueous 1M NaZSO3. To each of the divided samples was

added 700 uL Pi assay solution (30 mL of stock solution containing 4.2 g ammonium

molybdate, tetrahydrate, 56 mL conc. H2804 diluted to 1 L with distilled H20; 5 mL
10% L-ascorbic acid (w/v) in H20) followed by incubation for 30 min at 45 °C. Samples
that turned blue in the presence of NaIO4 and remained yellow/orange in the
corresponding H20 sample contained MIP-synthase without measurable background

phosphatase activity. Fractions devoid of phosphatase were combined and concentrated
to less than 5 mL using an Amicon Ultrafiltration Stirred Cell equipped with a PMIO

membrane using N2 pressure. Concentrated protein (1.5-5 .0 mg) was used to measure

MIP synthase activity.

FAS-B

Measurement of FAS—B activity4 expressed in E. coli required partial purification
from background fatty acid synthase enzymes of the E. coli host. A single colony of
DHSOl/pGM44 was grown in 5 mL FAS-B medium with ampicillin. All cultures were
grown at 30 °C and 250 rpm. After 12 h, the 5 mL growth was added to 500 mL of FAS-
B medium containing ampicillin. Cell density was monitored by following the OD600 of
the culture until reaching an absorbance of 3.0-3.5. Cells were harvested by
centrifugation (2 700g, 5 min) and then resuspended in 0.4 M potassium phosphate buffer

(2 mL buffer/g of wet cell weight), pH 7.5, containing 2 mM DTT and 0.001% PMSF.

178

Cellular membranes were disrupted by two passages through a French press (19 000 psi)
and the supernatant was separated from cellular debris by centrifugation (31 000g, 20
min). Enzyme manipulations were carried out at 4 °C. Ammonium sulfate was slowly
added to the culture supernatant to 30% saturation, and the solution was stirred for an
additional 30 min. Precipitated protein was removed by centrifugation (31 000g, 20 min)
and discarded. Ammonium sulfate was added to the supernatant to 65% saturation, and
the solution was stirred an additional 30 min. Following centrifugation (31 000g, 20
min), the precipitated proteins were redissolved in 40 mL of 0.1 M potassium phosphate,
pH 7.3, and then centrifuged for 16 h at 100 000g. Soluble proteins were decanted away
and residual proteins were resuspended in 1 mL of 0.4 M potassium phosphate, pH 7.3,
containing 3 mM DTT.
Enzyme Assays

MIP synthase assay

The specific activity for MIP synthase was measured by the following assay
procedure.2 MIP synthase assay buffer (2 mL) was pre-incubated at 37 °C for 10 min.
The assay was initiated by the addition of 1 mL of concentrated protein sample diluted to
0.6 mg total protein/mL in the assay buffer devoid of D-glucose 6-phosphate, sodium salt
or NAD+. The assay was run for 30 min and 200 uL aliquots were removed every 3 min
and added to 50 ML 20% TCA to stop the enzyme reaction. From each time-point, 100
uL of the enzyme reaction was combined with 200 uL H20 and a second 100 uL of the
enzyme reaction was combined with 100 uL an aqueous solution of 200 mM NaIO4. The

divided samples were incubated at 37 °C for 1 h. Samples containing NaIO4 were

quenched with 100 pL of aqueous 1M Na2S03. To each of the divided samples was

179

added 700 [.tL Pi assay solution (30 mL of stock solution containing 4.2 g ammonium
molybdate, tetrahydrate, 56 mL conc. H2SO4 diluted to l L with distilled H20; 5 mL
10% L—ascorbic acid (w/v) in H20) followed by incubation for 30 min at 45 °C. The
absorbance of each sample at room temperature was measured at 820 nm (E = 26 000 L

mol”l cm'l). MIP synthase activity was measured by the difference in slopes between the

line for NaIO4 and the control with H20. One unit of MIP synthase activity was defined

as the formation of 1 umol of MIP per min per mg protein at 37 °C.
Inositol dehydrogenase assay

The specific activity for inositol dehydrogenase was measured with the following
assay.5 Harvested cells were washed by resuspending in 15 mL of buffer consisting of
Tris-HCl (100 mM), pH 8.5 and phenyl methane sulphonyl fluoride (PMSF, 2 mM). The
cells were centrifuged at 3 700g and cell pellets were resuspended in 5 mL of the same
buffer followed by cell disruption by passage through a French press at 16 000 psi. Cell
lysates were clarified by centrifugation at 48 000g. Crude protein extracts were used to
measure inositol dehydrogenase activity. The inositol dehydrogenase activity was
measured by first combining 700 uL Tris-HCl buffer (100 mM), pH 9.0, with 100 uL
mya-inositol (400 mM) in buffer and 100 ,uL NAD+ (5 mM) in buffer and pre-incubating
at 37 °C. After measuring the background at 340 nm, 100 uL of protein of known
concentration was added and measurements were taken every second for 208. The
increase in absorbance at 340 nm (8 = 6 220 L mol'l cm'l) was measured for formation
of NADH. One unit of inositol dehydrogenase activity was defined as the formation of l

umol of NADH per min per mg protein at 37 °C.

180

Fatty Acid Ketoreductase assay

The assay of the ketoreductase step of FAS-B was conducted as described by
Schweitzer by measuring the loss of NADPH.4 The assay solution (1 mL) contained
potassium phosphate buffer (0.4 M), pH 7.3, DTT (3 mM), NADPH (0.70 mM), acetyl-
CoA (0.15 mM), 0.5 mg bovine serum albumin and a sample of enzyme solution.
Background loss of NADPH was recorded at 334 nm (8 = 6 220 L mol'l cm’l) for 300 s.
The enzyme reaction was started by addition of malonyl-CoA (0.20 mM), and loss of
NADPH was recorded over 300 s at 334 nm. One unit of ketoreductase activity is defined
as one umol of NADPH loss per minute per mg of protein at 25 °C.

In vitro analysis of triacetic acid lactone biosynthesis

The assay for triacetic acid lactone biosynthesis4 by FAS-B was conducted by
qualitatively identifying the formation of [14C]-labeled TAL from [1-14C] acetyl-COA.
The assay solution (1 mL) contained potassium phosphate buffer (0.4 M), pH 7.3, DTT (3
mM), NADPH (0.70 mM), acetyl-COA containing 0.5 “Ci of [1-"’C] acetyl-COA (0.15
mM), 0.5 mg bovine serum albumin and a sample of enzyme solution. The enzyme
reaction was started by addition of malonyl-CoA (0.2 mM). A protein sample of DHSOt
excluding pGM44 was prepared using the same partial purification procedure delineated
for cells expressing FAS-B. Five assays were run concurrently: DHSOL derived protein
with and without NADPH, DHSor/pGM44 derived protein with and without NADPH and
a fifth experiment containing no protein and no NADPH. After running the assay for 14
h at rt, to the 1 mL samples was added 30 uL of a saturated, aqueous solution of TAL.
The enzyme reaction was stopped by first adding 50 11L of 5 N NaOH followed by 60 11L

of 12 N H2SO4. The aqueous portion was extracted with EtOAc (3 x 200 uL), combined

181

organic layers were evaporated, and the resulting residue was dissolved in 60 [LL EtOAc
followed by spotting on a silica gel TLC plate. The five experiments were run on the
same plate and resolved using a mixture of 90:25:5 of toluenezdioxanezAcOH (v/v/v).
The presence of TAL was first marked on the plates after development by using UV light.
The TLC plate was then exposed for 48 h to a sheet of x-ray film. TAL biosynthesis was
confirmed by comparing the Rf of the UV active spot for TAL against the x-ray film

detected position on the plate indicating radioactivity.

Microbial Strains and Plasmids
E. coli DHSa [F ’ endAI hst17(r'Km+K) supE44 thi-1 recA] gyrA relA]
a801acZAM15 A(lacZYA-argF)U169], RB791 (W3110 lacL819), RB79lserA::aroB and
JWF1 (RB791 serA) were obtained previously by this laboratory. E. coli strain JM83
[ara A(lac-praAB) IPSL thi(¢ 801acZAM15)] was purchased from ATCC. E. coli BL-21,
BL-21 RIL and BL-21 RP were purchased from Stratagene. G. oxydans ATCC 621 was

purchased from the American Type Culture Collection. Saccharomyces cerevisiae strain
INVScl (MATOL his3AI leu2 trpI—289 ura3-52) was purchased from Invitrogen. Plasmid
pJH318 was obtained from the laboratory of Susan Henry. Plasmid pIOL05d15 was
obtained from the laboratory of Yasutaro Fujita. Plasmids pGM44 and pHPS6 were
obtained from the laboratory of Eckhart Schweizer. Plasmid pMR228 was obtained from
the laboratory of Chaitan Khosla. Plasmid pKOSl2-128a was obtained from James
Kealey of Kosan Biosciences. Plasmid pGE-l was obtained from A. Fujisawa. Plasmid
pGroESL was obtained from DuPont. Plasmid pGEM-3z/SYNIN01 was obtained from

David Rozzell at Biocatalytics. E. coli strain YZl66/pYZ71 was obtained from J. E.

182

Cronan at the University of Illinois. The plasmid pMIP harboring P171N01 was made in
the Frost lab by David Spears. The plasmid pRCl.55b harboring the E. coli locus serA

with EcoRV ends was made by Rachel Christ. The plasmid pET-22b was purchased

from Novagen.

Storage of Bacterial Strains and Plasmids
All bacterial strains were stored at -78 °C in glycerol. Plasmids were transformed
into DHSa 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 medium6 (1 L)
contained Bacto tryptone (10 g), Bacto yeast extract (5 g), and NaCl (10 g) with L-
arabinose (2 g), D-glucose (4 g), and L-fucose (250 mg) added when indicated. FAS-B
medium4c (1 L) contained casamino acids (20 g), yeast extract (10 g) and NaCl (10 g)
and was adjusted to pH 7.5. M9 salts6 (1 L) contained NazHPO4 (6 g), KHzPO4 (3 g),
NH4C1 (1 g), and NaCl (0.5 g). M9 minimal medium6 contained D-glucose (10 g),
MgSO4 (0.12 g), and thiamine hydrochloride (0.001 g) in l L M9 salts with supplements

added as noted. Solid medium was prepared by addition of Bacto agar to the liquid
medium to a final concentration of 1.5 % (w/v). Antibiotics were added as noted to the
following final concentrations: chloramphenicol (Cm), 20 ug/mL; ampicillin (Ap), 50

ug/mL; kanamycin (Kan), 50 ug/mL; and spectinomycin (Sp), 50 ug/mL. Inorganic salts,

183

L-arabinose, D-glucose, L-fucose and MgSO4 solutions were autoclaved separately while

thiamine hydrochloride and antibiotic stock solutions were sterilized through 0.22-um
membranes.

Gluconobacter oxydans ATCC 621 was groWn in YPG medium7 which contains
(1 L) Bacto peptone (3 g), Bacto yeast extract (5 g) followed by the addition of D-glucose
(25 g) after autoclaving. D-glucose was exchanged with D-sorbitol where indicated.

Saccharomyces cerevisiae cultures were grown in YPD medium8 which contained
(1 L) Bacto yeast extract (10 g), Bacto peptone (20 g) followed by the addition of D-
glucose (20 g) after autoclaving. SC minimal medium9 (1 L) contained Bacto yeast
nitrogen base without amino acids (6.7 g), D—glucose (20 g), and the following
supplements: adenine (0.1 g), arginine (0.1 g), cysteine (0.1 g), leucine (0.1 g), lysine (0.1
g), threonine (0.1 g), aspartic acid (0.05 g), histidine (0.05 g), isoleucine (0.05 g),
methionine (0.05 g), phenylalanine (0.05 g), proline (0.05 g), serine (0.05 g), tyrosine
(0.05 g), and valine (0.05 g). The amino acids were combined and autoclaved in 1/10
volume of the prepared medium in H20 prior to adding to medium. Solid YPD medium
and SC medium was prepared by addition of 2.0% (w/v) Bacto agar to liquid medium.
Following electroporation procedure, S. cerevisiae were plated onto SC minimal medium
agar plates containing 18% (w/v) sorbitol“) D—glucose, sorbitol and amino acids were
autoclaved separately.

Fermentation medium (1 L) contained KzHPO4 (7.5 g), citric acid monohydrate
(2.1 g), ammonium iron (III) citrate (0.3 g), and concentrated H2SO4 (1.2 mL). The
fermentation medium was adjusted to pH 7.0 with concentrated NH4OH prior to

autoclaving. Before inoculation, the following supplements were added to the

184

fermentation medium (1 L): D-glucose, MgSO4 (0.24 g), and trace minerals including
(NH4)6(M07024) - 4H20 (0.0037 g), ZnSO4' 7H20 (0.0029 g), H3BO4 (0.0247 g),

CuSO4 - 5H20 (0.0025 g), and MnClz- 4H20 (0.0158 g). Solutions of D-glucose and
MgSO4 were autoclaved separately while the trace minerals and aromatic vitamins were

filtered through sterile 0.22-um membranes.

General Fed-Batch Fermentation Conditions

Fermentations were conducted in a B. Braun M2 culture vessel with a 2 L
working capacity. Environmental conditions were supplied by a B. Braun Biostat MD
controlled by a DCU-1. Data was acquired on a Dell Optiplex Gs+ 5166M personal
computer utilizing B. Braun MFCS/Win software. PID control loops were used to
control temperature, pH, and glucose addition. The temperature was maintained at 33 °C
and the pH was maintained at 7.0 by addition of NH4OH 6r 2 N H2SO4. Glucose was
added as a 60% (w/v) solution. Dissolved oxygen (D.O.) was monitored using a Mettler-
Toledo 12 mm sterilizable Oz sensor fitted with an Ingold A-type O2 permeable
membrane. D.O. was maintained at 10% air saturation throughout the course of the
fermentations unless otherwise specified. Antifoam (Sigma 204) was manually pumped
into the vessel as needed.

Inoculants were initiated by introduction of a single colony into 5 mL of M9
medium and grown at 37 °C with agitation for 12-36 h until the culture was turbid. After
this time, the starter cultures were transferred to 100 mL of M9 medium and grown for an

additional 12 to 24 h at 37 °C and 250 rpm. After an appropriate OD600 was reached

185

(1.0-3.0), the inoculant was transferred to the fermentation vessel. The initial glucose
concentration in the fermentation medium was 18 g/L unless otherwise specified.

Three staged methods were used to maintain D.O. levels at 10% air saturation
during the course of the run. 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 its preset maximum of 940 rpm. When the impeller reached its preset
maximum, the mass flow controller then maintained D.O. levels by increasing the airflow
rate from 0.06 LIL/min to a preset maximum of 1.0 LIL/min At constant impeller speed
and constant airflow rate, D.O. level was finally maintained at 10% air saturation for the
remainder of the fermentation by oxygen sensor-controlled glucose feeding. PID control
parameter were set to 0.0 (off) for the derivative control (to), and 999.9 3 (minimum

control action) for integral control (11). Xp was set to 950.0% to achieve a Kc of 0.1. All

strains were cultured in the fermentation vessel for a total of 54 h. Isopropyl B-D-
thiogalactopyranoside (IPT G) was added every six hours beginning at the third phase of
growth for fermentations at concentrations specified for each fermentation. IPI‘ G was

sterilized as an aqueous solution through 0.22-um membranes prior to addition.

Analysis of Fermentation Broth
For strains being evaluated in shake flasks, samples (3-5 mL) of the culture were
taken at timed intervals and the cells were removed by centrifugation using a Beckman
microcentrifuge. Samples (5 mL) of fermentation broth were taken at indicated intervals.
Cell densities were determined by dilution of fermentation broth with water (1:100)

followed by measurement of absorption at 600 nm (OD600). Dry cell weight for E. coli

186

(g/L) was obtained using a conversion coefficient of 0.43 g/L/ OD600. The remaining
fermentation broth was centrifuged using a Beckman microcentrifuge to obtain cell-free
broth.

Solute concentrations in the cell-free culture supernatant were determined by 1H
NMR. For by 1H NMR quantitation of solute concentrations, solutions were
concentrated to dryness under reduced pressure, concentrated to dryness on 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). Concentrations were
determined by comparison of the integrals corresponding to each compound with the
integral corresponding to TSP (8: 0.00 ppm) in the 1H NMR. Compounds were
quantitated using the following resonances: acetate (1.92 ppm, s, 3 H), formate (8.51
ppm, s, 1 H), mya—inositol (4.07 ppm, m, 1H), mya-inositol l-phosphate (4.13 ppm, m,
1H) and mya—2-inosose (4.25 ppm, (1, 2H). Concentration of myo-inositol derived from
its 1H NMR integral values tended to be overestimated and the precise concentration of
mya-inositol was calculated by application of the following formula: [mya-inositol]acmal
= 0.8184 x [mya-inositol]NMR. The equation was obtained by dissolving various known
amounts of mya-inositol in 1 mL of D20 containing 10 mM TSP and recording their 1H
NMR spectra. The concentration of mya-inositol in each sample that was estimated for
1H NMR was plotted against the known concentration of myo-inositol for that sample
resulting in the calibration curve. All 1H NMR spectra were recorded on a Varian VXR-

300 FT-NMR spectrometer (300 MHz).

187

Genetic Manipulations

General Information

Standard protocols were used for construction, purification, and analysis of
plasmid DNA.6 E. coli DHSa was used as the host'strain for plasmid manipulations. T4
DNA ligase, Large Fragment of DNA polymerase I (Klenow fragment) and agarose
(electrophoresis grade) were purchased from either Gibco BRL Products. All gel purified
DNA was extracted using the Zymoclean DNA extraction kit obtained from Zymo.
Wizard PCR prep DNA purification kit was obtained from Promega. Kits were used as
described by the manufacturer. Restriction enzymes were purchased from Gibco BRL
Products or New England Biolabs. Calf intestinal alkaline phosphatase was purchased
from Roche Molecular Biochemicals. The ADE3 lysogenization kit was purchased from
Novagen. PCR amplifications were performed as described by Sambrook.6 Each
reaction (0.1 mL) contained 10 mM KC], 20 mM Tris-HCl (pH 8.8), 10 mM (NH4)2SO4,

2 mM MgSO4, 0.1% Triton X-100, dATP (0.2 mM), dCT P (0.2 mM), dGTP (0.2 mM),

dTTP (0.2 mM), template DNA (0.02 ug or 1 ug), 0.5 uM of each primer, and 2 units of
Vent polymerase. Primers were synthesized by the Macromolecular Structure Facility at
Michigan State University.

Phenol was prepared by addition of 0.1% (w/v) 8-hydroxyquinoline to distilled,
liquefied phenol.6 Two extractions with an equal volume of 1 M Tris-HCl (pH 8.0) 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 chloroformand isoamyl alcohol (24:1 v/v).

TE buffer contained 10 mM Tris-HCI (pH 8.0) and 1 mM disodium EDTA (pH 8.0).

188

Endostop solution (10X concentration) contained 50% glycerol (v/v), 0.1 M disodium
EDTA, 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
DNAase-free RNAase was added to 1 mL of 10X Endostop solution. DNAase-free
RNAase (10 mg/mL) was prepared by dissolving RNAase in 10 mM Tris-Cl (pH 7.5) and
15 mM NaCl.6 DNAase activity was inactivated by heating the solution at 100 °C for 15

min. Aliquots were stored at -20 °C.

Large Scale Purification of Plasmid DNA

Plasmid DNA was purified on a large scale using a modified lysis method
described by Sambrook co-workers.6 In a 2 L Erlenmeyer flask, 500 mL of LB medium
containing the appropriate antibiotic was inoculated from a single colony, and the culture
was incubated at 37 °C or 30 °C for approximately 15 h with agitation at 250 rpm. Cells
were harvested by centrifugation (4 000g, 5 min, 4 °C) and then resuspended in 10 mL of
cold solution 1 (50 mM glucose, 20 M Tris-HCl, pH 8.0, 10 mM EDTA, pH 8.0) into
which lysozyme (5 mg/mL) had been added immediately before use. The suspension was
stored at room temperature for 5 min. Addition of 20 mL of solution 2 (1% SDS (w/v) in
0.2 N NaOH) was followed by gentle mixing and storage on ice for 15 min. Fifteen
milliliters of ice cold solution 3 (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

189

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 centrifugation (20 000g, 20
min, 4 °C). The DNA pellet was then rinsed with 70% ethanol and dried.

The isolated 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 RNAase, the solution was
transferred to a 1.5 mL microcentrifuge tube and stored at rt for 30 min. DNA was
precipitated from solution upon addition of 500 pL of 1.6 M NaCl containing 13%
polyethylene glycol (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 uL of TE. The sample was extracted
sequentially with phenol (400 p.L), phenol and SEVAG (400 nL each), and finally
SEVAG (400 uL). Ammonium acetate (10 M, 100 uL) 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 250-500 uL of TE.

The concentration of DNA in the sample was determined as follows: an aliquot

(10 uL) of the DNA solution was diluted to 1 mL in TE and the absorbance at 260 nm

190

was measured relative to the absorbance of TE. The DNA concentration was calculated
based on the fact that the absorbance at 260 nm of 50 ug/mL of double stranded DNA is

1.0.

Small Scale Purification of Plasmid DNA

An overnight culture (5 mL) of the plasmid-containing strain was grown in LB
medium containing the appropriate antibiotics.6 Cells from 3 mL of the culture were
collected in a 1.5 mL microcentrifuge tube by centrifugation. The harvested cells were
resuspended in 0.1 mL of cold solution 1 into which lysozyme (5 mg/mL) had been
added immediately before use and the solution was stored on ice for 10 min. Addition of
solution 2 (0.2 mL) was followed by gentle mixing and storage on ice for 5-10 min.
Solution 3 (0.15 mL) was added to the sample and shaken vigorously. The sample was
stored on ice for 5 min and the cellular debris was removed by centrifugation
(microcentrifuge, 15 min, 4 °C). The supernatant was transferred to another
microcentrifuge tube and extracted with equal volumes of phenol and SEVAG (0.2 mL
each). The aqueous phase (approximately 0.5 mL) was transferred to a fresh microfuge
tube and the 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, rt) to isolate the
DNA. The DNA pellet was rinsed with 70% ethanol, dried, and redissolved in 50-100 uL
TE. DNA isolated from this method was used for restriction enzyme analysis and the

concentration was not determined by spectroscopic methods.

191

Restriction Enzyme Digestion of DNA

Restriction enzyme digests were performed in buffers provided by Gibco BRL or
New England Biolabs. A typical restriction enzyme digest contained 0.8 ug of DNA in 8
pL of TE, 2 “L of restriction enzyme buffer (10X concentration), 1 uL of bovine serum
albumin, 1 uL of restriction enzyme and 8 pL TE. Reactions were incubated at 37 °C for
1 h, terminated by addition of 2.2 pL of 10X Endostop and analyzed by agarose gel
electrophoresis. For cloning experiments, the reaction was terminated by addition of l
pL of 0.5 M NazEDTA (pH 8.0), followed by extraction with phenol and SEVAG (0.1
mL each). 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
mixed and kept at -78 °C for 3 h. Precipitated DNA was recovered by centrifugation (15
min, 4 °C), treated with 0.1 mL 70% ethanol and again centrifuged (15 min, 4 °C). DNA

was dried and redissolved in TE.

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. Ethidium
bromide (0.5 rig/ml) was added to the agarose to allow visualization of DNA fragments
under an 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—Rb, 2.0-Rb, 1.9-kb, 1.6-kb, 1.4-kb, 0.9-Rb, 0.8-kb, and 0.6-kb).

192

Isolation of DNA from Agarose

Two methods were used for isolating DNA from agarose gels. The first method
involved cutting out the band of agarose containing the required DNA from the gel and
chopping it thoroughly with a razor. The agarose was then transferred to a 0.5 mL
microfuge tube packed tightly with glass wool and having an 18 gauge hole at the
bottom. The tube was centrifuged for 5 min using a Beckman microfuge to extrude the
DNA solution from the agarose into a 1.5 mL microfuge tube. The DNA was
precipitated out using 3 M NaOAc and 95% ethanol as previously described.

The second method involved use of the Zymoclean DNA isolation kit purchased

from Zymogen.

Treatment of DNA with Klenow Fragment

DNA fragments with recessed 3’ termini were modified to DNA fragments with
blunt ends by treatment with the Klenow fragment of E. coli DNA polymerase I. After
restriction digestion (20 uL) of the DNA (0.8-2 pg) was complete, a solution (1 uL)
containing each of the four dNTP’s was added to a final concentration of 1 mM.
Addition of 1-2 units of Klenow fragment was followed by incubation at rt for 30 min.
Since the Klenow fragment works well in the common restriction enzyme buffers, 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 precipitation as described previously and

redissolved in TB.

193

Treatment of Vector DNA with Calf Intestinal Alkaline Phosphatase

Following restriction enzyme digestion, plasmid vectors were dephosphorylated
to prevent self-ligation. Digested vector DNA was dissolved in TE (88 uL). To this
sample was added 10 mL of dephosphorylation buffer (10X concentration) and 2 uL of
calf intestinal alkaline phosphatase (2 units). The reaction was incubated at 37 °C for l h.
The phosphatase was inactivated by the addition of l uL of 0.5 M EDTA (pH 8.0)
followed by heat treatment (65 °C, 20 min). The sample was extracted with phenol and
SEVAG (100 uL each) to remove the protein, and the DNA was precipitated as

previously described and redissolved in TE.

Ligation of DNA

Molar ratios of insert to vector were typically maintained at 3 to 1 during
ligations. A typical reaction contained 0.1 ug vector DNA, 0.05 to 2.0 ug insert in a total
volume of 7 pL. To this was added 2 uL of T4 ligation buffer (5X concentrations) and l
pL of T4 DNA ligase (2 units). The reaction was incubated at 16 °C for at least 10 h and

then used to transform competent cells.

Preparation and Transformation of Competent Cells (E. coli)

Competent cells were prepared using a procedure modified from Miller.ll An
aliquot (1 mL) from an overnight culture (5 mL) was used to inoculate 100 mL of LB
containing the appropriate antibiotics. The cells were cultured at 37 °C with shaking at

250 rpm until the OD600 was between 0.4 and 0.6. The culture was transferred to a

centrifuge bottle that had been sterilized with bleach and rinsed exhaustively with sterile

194

water. The cells were harvested by centrifugation (4 000g, 5 min, 4 °C) and the culture
medium was discarded. All manipulations were carried out on ice during the remainder
of the procedure. The harvested cells were resuspended in ice cold 0.9% NaCl (100 mL)
and the cells were reisolated by centrifugation (4 000g, 5 min, 4 °C). The cells were
resuspended in ice cold 100 mM CaClz (50 mL) and stored on ice for 30 min. After
centrifugation (4 000g, 5 min, 4 °C), the cells were resuspended in 4 mL of ice cold 100
mM CaClz containing 15% glycerol (v/v). Aliquots (0.25 mL) of competent cells in 1.5
mL microfuge tubes were frozen in liquid nitrogen and stored at -78 °C.

Frozen competent cells were thawed on ice for 5 min before transformation. A
small aliquot (l to 10 uL) 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 (1.5 min).
LB (0.5 mL, no antibiotics) was added to the cells, and the sample was incubated at 37
°C (no agitation) for at least 1 h. Cells were collected by centrifugation in a Beckman
microcentrifuge. If the transformation was to be plated onto LB plates, 0.5 mL of the
culture supernatant was removed and the cells resuspended in the remaining 0.1 mL of
LB, and spread onto plates containing the appropriate antibiotics. If the transformation
was to be plated onto minimal medium plates, the cells were washed once with a solution
of M9 inorganic salts. After resuspension in fresh solution of M9 inorganic salts (0.1
mL), the cells were spread onto the plates. A sample of competent cells with no DNA
added was also carried through the transformation procedure as a control. These cells
were used to check the viability of the competent cells and to verify the absence of

growth on selective medium.

195

Transformation was also performed by electroporation using electrocompetent
cells. In order to prepare electrocompetent cells a 100 mL culture was grown. Once an
absorbance of 0.5-0.7 was observed at 600 nm, the cells were kept on ice for 15 min and
harvested. The cells were gently washed twice with cold water (100 mL then 50 mL) and
then resuspended in 5 mL cold aqueous 10% glycerol (v/v). The cells were centrifuged
(4 000g, 5 min, 4 °C) and the pellet was slowly resuspended with 1 mL of cold aqueous
10% glycerol. The cells were transferred into microfuge tubes in 0.25 mL aliquots,
immediately frozen using liquid nitrogen and stored at -78 °C.

The electroporation was performed in Bio-Rad gene pulser cuvettes with an
electrode gap of 0.2 cm. The cuvettes were chilled on ice for 5 min prior to use.
Electrocompetent cells were thawed in ice for 5 min and 0.1 mL thawed cells was added
to the chilled cuvette. To this was added 1-2 pL of plasmid DNA (0.1 mg mL‘l) and the
mixture was gently shaken. The Bio-Rad gene pulser was set at 2.5 Kvolts, 25 uF and
200-400 Ohms. The outside surface of the cuvette was wiped clean and it was placed in
the sample chamber. A single pulse was applied, the cuvette was removed and 0.5 mL of
LB was added to it. The contents of the cuvette were transferred to a 1.5 mL microfuge
tube. The cells were incubated at 37 °C for l h. The transformed cells were plated in the

same manner as in a normal transformation with the same controls.

Preparation of electrocompetent cells (G. oxydans ATCC 621)

G. oxydans ATCC 621 was prepared for electroporation by the following
method.12 A culture of G. oxydans was shaken for 36 h in 5 mL YPG at 30 °C. To 100

mL YPG was added 1 mL of the sub-culture followed by shaking until reaching an

196

absorbance of 0.3504 at 540 nm. The cells were harvested (3 700g, 10 min, 4 °C) then
resuspended in 100 mL cold M9 salt solution. The cells were pelleted (3 700g, 10 min, 4
°C) followed by resuspending in 100 mL cold 300 mM sucrose. The cells were pelleted
(3 700g, 10 min, 4 °C) followed by resuspending in 0.5 mL cold 300 mM sucrose
containing 10% glycerol (w/v).

The electroporation was performed in Bio-Rad gene pulser cuvettes with an
electrode gap of 0.2 cm. The cuvettes were chilled on ice for 5 min prior to use.
Electrocompetent cells (60 pL) were added to the chilled cuvette. To this was added 1.5
pg of plasmid DNA and the mixture was gently shaken. The Bio-Rad gene pulser was
set at 2.5 Kvolts, 25 uF and 200 Ohms. The outside surface of the cuvette was wiped
clean and it was placed in the sample chamber. A single pulse was applied, the cuvette
was removed and 1.0 mL of YPG was added to it. The contents of the cuvette were
transferred to a 15 mL Falcon tube then shaken for 2 h at 30 °C. The culture was
transferred to a 1.5 mL microfuge tube then microfuged. for 30 s. The broth was
discarded and the cells were resuspended in 0.1 mL YPG followed by plating on YPG
agar plates containing the appropriated anitbiotic(s) for plasmid maintenance followed by

incubation at 30 °C.

Preparation of electrocompetent cells (S. cerevisiae)

S. cerevisiae was prepared for electroporation with the following procedure.10
S. cerevisiae strain INVScl was initially grown from a single colony in 5 mL YPD and
shaken for 16 h at 30 °C. From the 5 mL culture, 1 mL was added to 500 mL YPD which

was grown until reaching an absorbance of 1.3-1.5 at 600 nm. The cells were chilled to 4

197

°C then harvested by centrifugation (3 700g, 8 min, 4 °C). The supernatant was
discarded, and the cell pellet was resuspended in 500 mL ice cold water. The cells were
harvested by centrifugation (3 700g, 8 min, 4 °C) then resuspended in 250 mL ice cold
water. Following harvesting by the same conditions, the cells were resuspended in 40
mL of ice cold, aqueous l M sorbitol. The cells were once again harvested followed by
resuspending in 0.5 mL cold 1 M sorbitol.

For transformation, 40 uL of cells were combined with 1 [lg plasmid DNA
followed by incubation for 5 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 1.5 kV,
25 rtF, and 200 Ohms. Following electroporation, 1 mL cold 1M sorbitol was
immediately added to the cuvette and the entire cell solution was plated with the

appropriate selection followed by incubation for 3-5 days at 30 °C.

ADE3 Lysogeny

The ADE3 lysogeny was performed according to the protocol provided by
Novagen. The ADE3 lysate (2.5 x 1010 pfu/mL), Helper phage lysate (3.6 x1010 pfu/mL),
and Selection phage lysate (5.6 x 1010 pfu/mL) were stored at -78 °C and thawed on ice
immediately before use. A colony of the strain to be lysogenized was inoculated into 5

mL of LB containing the appropriate antibiotics, 0.2% maltose and 10 mM MgSO4. The

maltose stock solution was sterilized by passage through 0.22-um membranes and the

MgSO4 solution was separately autoclaved. The culture was grown at 37 °C with
agitation until the OD600 reached 0.5. Various amounts (1, 3, 5, 7, 10 uL) of the culture

were transferred to individual microfuge tubes. lDE3 lysate (4 uL), Helper phage lysate

198

(2.78 uL), and Selection phage lysate (1.79 uL) were added to each tube and mixed
gently. The host/phage mixture was incubated at 37 °C for 20 min. The mixture was
plated onto LB plates with appropriate antibiotics and incubated overnight at 37 °C.
Several of the resulting colonies were selected, transformed with a plasmid containing an
assayable gene under a T7 promoter, and the specific activity of the enzyme expressed
from the 77 promoter was measured. The host strain providing the highest specific

activity was chosen and named as the (DE3) strain.

CHARTER}
Strain Constructions
Strain JWFl(DE3) .
JWFl(DE3) was prepared using the ADE3 Lysogenization Kit according to the

manufacturer’s protocol.

Plasmid constructions

Plasmid pAD1.45A

Construction of the 7.0-kb plasmid started with PCR amplification of the INOI
locus from pJH318 using the following primers incorporating EcaRI recognition sites: 5’-
CGAA’I‘TCATGACAGAAGATAATATTGCT and 5’-CGAATT(_ZI'CGCTCTCCTC-
AACT. The 1.7-kb PCR fragment was digested with EcoRI then ligated into pJF118EH
linearized by digestion with EcoRI. The INOI locus was placed such that transcription
was from the toe promoter and in the same direction as the genetic marker bla conferring

resistance to Ap.

199

Plasmid pAD1.88A

Construction of the 8.9-kb plasmid started with excision of the serA locus from
pD2625 by digestion with Dra1 and EcoRV. The 1.9-kb fragment was isolated by gel
purification followed by ligation into pAD1.45A linearized by treatment with Sma1.

Orientation of the serA locus was such that transcription was in the opposite direction of

INOI .

Plasmid pCH6.112A

Construction of the 6.9-kb plasmid started with digestion of pGEM-3z/SYNIN01
obtained from Biocatalytics with EcoRI to liberate the 1.6-kb S YNINOI locus. The 1.6-
kb fragment was ligated with pJF118EH linearized by treatment with EcoRI. The
S YNINOI locus was such that transcription was from the tac promoter and oriented in the

same direction as the genetic marker bla conferring Ap resistance.

Plasmid pCH6.123A

Construction of the 8.6-kb plasmid started with excising the serA locus from
pRCl.55B by digestion with EcoRV. After isolation the 1.7-kb fragment by gel
purification, the fragment was ligated into pCH6.112A digested with BamHI followed by
filling in the 5’ overhangs by treatment with Klenow fragment. The serA locus was

oriented such that transcription was in the opposite direction as the S YNINOI locus.

200

Plasmid pCH7.41

Construction of the 7.3-kb plasmid started with digestion of leP with Ndel and
EcoRI to liberate INOI as a 1.8-kb fragment with Nde1 on the 5’ end and EcoRI on the 3’
end. After gel purification, the 1.8-kb fragment was ligated into pET-22b digested with
Ndel and EcoRI. The INOI locus was such that transcription was from the T7 promoter

and oriented in the same direction as the genetic marker bla conferring resistance to Ap.

Plasmid pCH7.61B

The 7.2-kb plasmid was constructed by first excising the serA locus from
pRCl.55B by digestion with EcoRV. The gel purified, 1.7-kb fragment was then blunt-
end ligated into pET-22b linearized by treatment with PshA1. The serA locus was

oriented such that transcription was in the same direction as lacI.

Plasmid pCH7.66A

Construction of the 9.0-kb plasmid started with digestion of pCH7.41, containing
the locus P771N01, with Ndel and Sca1. The 3.0-kb containing the [NO] locus in its
entirety and the 5’ end of bla was gel purified. The plasmid pCH7.6lB was digested with
Ndel and Sca1 to liberate a 6.0-kb fragment containing the remaining 3’ end of bla, the
serA locus and [ad which was gel purified. The two fragments were ligated to
reassemble bla conferring resistance to Ap and reuniting the [NO] locus with the T7

promoter.

201

Plasmid pCH7.l70

Construction of the 6.4-kb plasmid started with removing dxr locus from
pPV3.20a by digestion first with PstI followed by Smal. The 5.3-kb fragment containing
the lac promoter, lach and bla was isolated by gel purification. The iolG locus was
amplified from pIOL05d15 using the following PCR primers containing the Smal
terminal recognition sequence on the 5’ end and the PstI terminal recognition sequence
on the 3’ end: 5’-TCCCCCGGGAACAGAAGGAGTGGCTGTCAATG and 5’-
GCGQTGCAGGGATCCAATGCTAACTI‘CATA. After digestion and gel purification,
the 1.1-kb fragment was ligated with the 5.3-kb fragment from pPV3.20a. The iolG locus
was oriented such that transcription was from the toe promoter and in the same direction

as the genetic marker bla conferring resistance to Ap.

Plasmid pCH7.l94A
Construction of the 10.7-kb plasmid started with digestion of pCH7.l70 with Ssp1

to liberate a 1.7-kb fragment containing PmcialG with blunt ends. After gel purification,

the 1.7-kb fragment was blunt-end ligated with pAD1.88A linearized by treatment with

Nru1. The orientation of PmcialG was such that transcription was in the same direction as

laclq and in the opposite direction of PmCINOI .

Plasmid pCH6.254A
Construction of the 16-kb plasmid started with digestion of pAD1.45A with Nru1

and Dra1 to liberate a 4.1-kb fragment containing lach and PmcINOI . The fragment was

ligated into the G. subaxydans/E. coli shuttle vector pGE-ll3 containing the genetic

202

marker for kanamycin resistance. The PtacINOI locus was oriented such that

transcription was in the same direction as the kanamycin resistance gene.

Culture Conditions for Codon Usage with MIP synthase

E. coli B strains BL21-CodonPlus-RIL[argU ileY leuW Cami], BL21-CodonPlus-
RP [argU praL Cam'] and BL-2l Gold were acquired from Stratagene and transformed
with pAD1.88A following the protocol with the cells. The constructs BL-21—
RIL/pAD1.88A and BL-2l-RP/pAD1.88A were plated on LB/Cm/Ap and BL-
21/pAD1.88A was plated on LB/Ap at 37 0C. A single colony was used to start 5 mL
sub-cultures of LB containing the appropriate antibiotic at 37 °C. After shaking for 12 h,
each sub-culture was added to 500 mL LB containing the appropriate antibiotics in 2 L
flasks followed by shaking at 37 °C. When the absorbance at 600 nm reached 0.6-0.8,
0.5 mL of 400 mM IPTG was added followed by continued shaking at 37 °C until
reaching an absorbance of 3.5 at 600 nm. The cells were harvested by centrifugation (3
700g, 5 min, 4 °C) then resuspended in Buffer A (2 mL/ g wet cell weight) for partially
purifying MIP synthase. The cells were disrupted by French press and the lysates were
clarified by centrifugation (32 000g, 25 min, 4 °C). The lysates were applied to DEAE
columns (5 x 2.5 cm) and the assay was carried out as described in the beginning of
Chapter 5.

Culture Conditions for groESL heat shock protein with MIP synthase

The constructs BL-21/pGroESL/pAD1.88A and BL-21/pADl.88A were grown as
previously described for the Codon Usage strains. The culture temperatures were

maintained according to the temperatures described in Chapter 2.

203

Oxidation of mya-Inositol by G. oxydans ATCC 621

A culture medium was prepared with D-sorbitol (1.0 g), yeast extract (50 mg) in
H20 (10 mL, distilled).14 After autoclaving for" 25 min and cooling to rt, the culture
medium was inoculated with Gluconobacter oxydans ATCC 621 then placed in a shaker
thermostatted at 30 °C for 24 h. Growth medium was prepared with inositol (12.0 g, 66.7
mmol), sorbitol (0.4 g), yeast extract (2.0 g) in H20 (400 mL, distilled). After
autoclaving 25 min and cooling to rt, the growth medium was inoculated with the culture
medium then placed in a shaker thermostatted at 30 °C for 48 h. The solution was
centrifuged, decanted from cells, then concentrated in vacuo to 75 mL, diluted with
MeOH (400 mL), then chilled for 12 h at —20 °C. The resulting precipitate was filtered
and washed with methanol and dried to afford myo-2-inosose (8.17 g, 69%) as a white
powder contaminated with Z 5% sorbitol. A second crop was obtained after chilling the
filtrate at -20 °C for 12 h affording myo-2-inosose (3.09 g', 11.87 g theor., 17.4 mmol,

26%) as a white powder contaminated with S 5% sorbitol. 1H NMR (D20) 6 4.25 (d, J =
10.2 Hz, 2 H), 3.66 (dd, J = 934,934 Hz, 1 H), 3.26 (m, 2 H). 13C NMR (D20) 5 206.0,

94.3, 76.2, 74.5, 74.1, 73.3.

204

Oxidation of neo-Inositol by G. oxydans ATCC 621
A sub-culture medium was prepared with D-sorbitol (250 mg), yeast extract (50

mg) in H20 (5 mL, distilled) and inoculated with Gluconobacter oxydans ATCC 621

then placed in a shaker thermostatted at 30 °C for '72 h.15 Growth medium was prepared

with nea-inositol (400 mg, 2.2 mmol), D-sorbitol (200 mg), yeast extract (1.0 g) in H20

(200 mL, distilled). After autoclaving 25 min and cooling to rt, the entire sub-culture was
added to the growth medium then placed in a shaker thermostatted at 30 °C for 168 h.
The solution was centrifuged, decanted from cells, then concentrated in vacuo to 20 mL
followed by chilling for 48 h at 4 °C. The resulting precipitate was filtered and washed
with methanol and dried to afford scyllo-2,5-diketoinositol (190 mg, 1.08 mmol, 49%) as

a tan powder. 1H NMR (D20) 5 4.55 (s, 4 H, diketo), 4.43 (d, J = 10.1 Hz, 2 H, mono-

hydrate), 3.57 (d, J = 10.1 Hz, 2 H, mono-hydrate), 3.49 (s, 4 H, dihydrate). 13C NMR

(D20) 5 206.5, 94.2, 93.9, 75.9 (2), 75.0, 73.9.

205

 

 

rl l

 

Figure 83. 1H NMR of scyllo-2,5-diketoinositol in D20.

206

1

I l T 17 I’ I I T T I I 7

TT

T

777 I l

 

 

 

 

Figure 84. 13C NMR of scylla-2,S-diketoinositol in D20.

207

 

IIIIIfllllrflllll1ll1illlljIIIIITIIT)

IIIIIJITllllI‘llllillllllIllllllllllll‘rllIIiiI'IIIIII'JIIIIIIlIlIII

 

IT—I’rljllrlg

20

40

60

100

120

140

160

180

200

Fermentations with G. oxydans ATCC 621/pCH6.254A

Fermentations employed a 2.0 L working capacity B. Braun MD2 culture vessel.
Utilities were supplied by a B. Braun Biostat MD controlled by a Dell Otiplex Gs+ 5166
personal computer equipped with B. Braun MFCS/W in software. PID control loops were
used to control temperature, pH, and glucose addition. The temperature was maintained
at 28 °C, and the pH was maintained at 6.0 by addition of 4M NaOH or 10% H3PO4.
Dissolved oxygen (D.O.) was measured using a Mettler-Toledo 12 mm sterilized 02
sensor fitted with an Ingold A-type 02 permeable membrane. D.O. was maintained at
20% air saturation. Antifoam (Sigma 204) was added manually as needed.

G. subaxydans ATCC621/pCH6.254a was grown to an absorbance of 1.4-1.6 at
650 nm in 100 mL YPG containing 50 ug/mL Kan at 30 °C. The entire culture was
added to the fermentor containing 900 ml. YPG with 25 g D-glucose and Kan (50
ltg/mL).

HAPTER
Synthetic Procedures

Synthesis of myo-2-inosose

mya-Z-inosose (air oxidation).16 To a solution of mya-inositol (1.0 g, 5.6 mmol)
in H20 (100 mL) was added 10% Pt/C (1.0 g, 10 mol%). With vigorous stirring, the
reaction was heated to 90 °C while sparging air through the solution. After 4.5 h, the
reaction was cooled and the catalyst was filtered through Celite followed by washing

with 30 mL H20. The aqueous solution was concentrated in vacuo to a tan solid (0.62 g,

60 % yield) containing myo-2-inosose that was not purified further. 1H NMR (D20) 8

208

4.25 (d, J = 10.2 Hz, 2 H), 3.66 (dd, J = 9.34,9.34 Hz, 1 H), 3.26 (m, 2 H). 13‘C NMR

(D20) 5 206.0, 94.3, 76.2, 74.5, 74.1, 73.3.

myo-inositol acetinide 14.17 To a stirred suspension of myo-inositol (75 g, 416
mmol) and p-toluene sulfonic acid (0.9 g) in dry DMSO was added 2,2-
dimethoxypropane (180 mL, 1.46 mol, 3.5 eq). The suspension was stirred under an
atmosphere of N2 at 110 °C for 6h. After cooling to rt, the brown solution was
concentrated in vacuo to a dark brown oil. The acetinide was triturated with 750 mL
EtOAc then left for 12h at 0 °C. The resulting tan powder was filtered and washed with
cold EtOAc. The filtrate was left at 0 °C for an additional 24h resulting in a second crop
of powder. The filter cakes were combined to give 78.5 g of acetinide 14 (75% yield).
1H NMR (D20) 5 4.31 (dd, J = 4.67, 4.67 Hz, 1 H), 3.89 (m, 1 H), 3.71 (dd, J = 4.40,
4.12, 1 H), 3.42 (m, 2 H), 3.11 (dd, J = 9.62, 9.62, 1 H) 1.39 (s, 3 H), 1.25 (s, 3 H). 13C
NMR (D20) 8 110.5, 78.7, 76.2, 74.8, 74.5, 72.7, 71.2, 69.5, 27.5, 25.3.

tetrabenzyl inositol 15.17 The inositol acetinide 14 (20 g, 91 mmol) was stirred
as a suspension in dry lDMF (363 mL, 0.25 M) with tetrabutyl ammonium iodide (1.67 g,
4.5 mmol) while under an atmosphere of Ar at 0 °C. To the suspension was added benzyl
bromide (70 g, 409 mmol, 4.5 eq) followed by NaH as a 55% dispersion in parafin (17.8
g, 409 mmol, 4.5 eq). To expose the NaH, the suspension was washed with 30 mL dry
hexanes. After stirring for 2h at 0 °C, the ice bath was removed and the reaction was
allowed to warm to rt then stirred for an additional 18h at rt. The reaction was slowly
quenched with 500 mL saturated NH4Cl followed by extraction with CH2C12 (3 x 250
mL). The combined CH2C12 was washed with saturated NH4Cl (200 mL) then dried over

MgSO4, filtered and concentrated in vacuo to a brown syrup.

209

The syrup was dissolved in a solution of MeOH (450 mL), H20 (75 mL) and

concentrated HCl (75 mL) then stirred at reflux. After 2h, a white solid precipitated. The
reaction was cooled and MeOH was removed in vacuo. The remaining aqueous portion

was extracted with CHzClz (3 x 150 mL). The combined CH2C12 were dried over

MgSO4 then concentrated in vacuo to a waxy solid. Recrystallization from petroleum

ether resulted in the tetrabenzyl inositol 15 (37.8 g, 77%) as a white solid. 1H NMR

(CDC13): 5 7.2-7.4 (m, 20H), 4.7—5.0 (m, 8H), 4.18 (dd, J = 2.7, 2.7 Hz, 1H), 3.97 (dd, J
= 9.45, 9.45 Hz, 1H), 3.84 (dd, J = 9.6, 9.6 Hz, 1H), 3.4-3.5 (m, 3H); 13C NMR (CDC13):
5 138.5, 138.4, 137.7, 128.5 (2), 128.3, 127.9, 127.8 (2), 127.6, 83.2, 81.6, 81.3, 79.9,

75.9, 75.7, 75.6, 72.7, 71.7, 69.1.

pentabenzyl inositol 16.18 To solution of 15 (20 g, 37 mmol) and benzyl
chloride (4.4 mL, 37 mmol) in benzene (400 mL) blanketed with argon was added a 60%
dispersion of NaH (12 g, 300 mmol). The reaction was heated to reflux for 1 h then
cooled on ice prior to slowly adding H20 (500 mL) to quench the NaH. The organic
layer was separated and the aqueous layer was extracted with benzene (2 x 150 mL). The
combined organics were washed with brine (100 mL) then dried and concentrated to a
waxy solid. The solid was washed of hexabenzylated inositol with light petroleum ether,
filtered, then recrystallized from MeOH to afford 16 (14.4 g, 62%) as a white solid. mp
122-123 °C. 1H NMR (CDC13) 5 7.33-7.24 (m, 25H), 4.92-4.71 (m, 10H), 4.22 (m, 1H),
4.00 (dd, J = 9.34, 9.62 Hz, 2H), 3.49-3.37 (m, 3H); 13C NMR (CDC13) 5 138.7, 138.6,

137.9, 128.4, 128.3, 128.0, 127.8, 127.5, 83.1, 81.1, 79.7, 75.9, 72.7, 67.4.

pentabenzyl inosose 17. To a solution of Dess-Martin Reagent (5.3 g, 12.4
mmol) in AcOH (60 mL) and CH2C12 (42 mL) was added by dropwise addition a

210

solution of 16 (6.0 g, 9.5 mmol) in CH2C12 (42 mL) over 5 min. The reaction stirred at rt
for 12 h. The oxidant was dried by addition of 1.2 N NaOH (100 mL) followed by

stirring for 20 min. The organic layer was separated then washed with 1.2 N NaOH (50

mL), brine (50 mL) then dried and concentrated to give 17 (5.5 g, 92%) as a white solid
without any further purification. mp 146-148 °C. 1H NMR (CDCl3) 5 7.43-7.24 (m,

25H), 4.92-4.83 (m, 10H), 4.76 (d, J = 10.5 Hz, 1H), 4.55 (d, J = 11.5 Hz, 1H), 4.15 (d, J

= 10 Hz, 1H), 3.87 (dd, J = 8.8, 9.0 Hz, 1H) 3.62 (dd, J = 9.8, 9.5 Hz, 1H); 13C NMR
(CDCl3) 5 202.2, 138.1 (2), 137.3, 128.4(2), 128.1 (2), 128.0, 127.9, 127.8, 83.7, 82.2,

81.4, 76.2, 76.0, 73.4.

myo-Z-inosose via 17. To a solution of pentabenzyl inosose 17 (3.0 g, 4.8 mmol)

in THF (100 mL) and H20 (10 mL) blanketed with argon was added 10% Pd/C (2.0 g).
The stirred suspension was evacuated then placed under an atomsphere of H2. After 3 h,
the catalyst was filtered through Celite then washed with H20 (20 mL). The combined
solvents were concentrated in vacuo to a light gray solid of myo-2-inosose (840 mg,

99%). 1H NMR (D20) 5 4.25 (d, J = 10.2 Hz, 2 H), 3.66 (dd, J = 9.34, 9.34 Hz), 3.26

(m, 2 H). 13C NMR (D20) 5 206.0, 94.3, 76.2, 74.5, 74.1, 73.3.

Synthesis of 1,2,3,4-tetrahydroxybenzene.
1,2,3,4-tetrahydroxybenzene (aromatization). myo-2-Inosose (11.0 g, 61.2

mmol) treated with degassed 0.5 M H2304 (310 mL, 200 mM) and warmed to reflux

under Ar. After 9h, the solution was cooled to 4 °C and the pH was adjusted to 4.0 with

sat. aq. NaHCO3 then concentrated in vacuo to 100 mL. Product was obtained by method

of continuous extraction with t-butyl methyl ether (500 mL) for 18 h. The organic phase

211

was concentrated to 100 mL in vacuo and the resulting precipitate was filtered and the
filtrate reserved. The solid was washed with cold hexanes and dried affording 1,2,3,4-
tetrahydroxybenzene (4.72 g, 54%) as a tan powder. The reserved filtrate was treated
with 300 mL hexanes and the resulting precipitate was filtered, washed with cold hexanes

and dried to afford additional 1,2,3,4-tetrahydroxybenzene (1.08 g, 12%) as a dark brown
solid. mp 162-164 °C. 1H NMR (d6-acetone) 5 7.24 (s, OH), 6.20 (s, 2 H). 13C NMR
(d6-acetone) 5 139.7, 134.7, 106.2. Anal. Calcd for C6H604: C, 50.71; H, 4.23. Found:

C, 50.63; H, 4.32.

tribenzyloxypyrogallol 25.19 A solution of pyrogallol (20 g, 159 mmol) in dry,

degassed acetone (200 mL, 0.8 M) was treated with benzyl bromide (57 mL, 481 mmol)
then K2CO3 (100 g, 725 mol) at 22 °C under Ar. After 30 min, the reaction was

warmed to reflux. After 24 h the reaction was diluted with a solution of NaOH (1.6 g) in
MeOH (32 mL) and stirred for 30 min at reflux. After cooling to 22 °C, the solids were
filtered and washed with acetone. The filtrate was concentrated in vacuo.
Recrystallization of the residue from MeOH afforded 1,2,3-tribenzyloxypyrogallol 25 (52
g, 83%) as a cream powder. mp 67-68 °C. 1H NMR (CDCl3) 5 7.4 - 7.2 (m, 15 H), 6.9
(t, J = 8.5 Hz, 1H), 6.6 (d, J = 8.2 Hz, 2H), 5.09 (s, 4H), 5.07 (s, 2H); 13C NMR (CDCl3)
5 153.1, 138.6, 137.9, 137.2, 128.5, 128.4, 128.1, 127.8, 127.7, 127.3, 123.6, 108.0, 75.1,

71.1.

2,3-dibenzyloxy-1,4-benzoquinone 26.20 To a solution of 25 (2.0 g, 5.0 mmol)

in HOAC (30 mL), was added K3Fe(CN)6 (0.82 g, 2.5 mmol) and 30% H202 (1.3 g, 11.5

mmol) followed by stirring the solution at rt for 18 h. The solution was diluted with

CH2C12 (50 mL) and the organic layer subsequently washed with H20, saturated

NaHCO3, and brine. Drying and concentration resulted in a red oil. Purification by

212

radial chromatography (2 mm thickness, EtOAc/hexane, 1:19, v/v) afforded 26 as a red
oil. 1H NMR (CDC13) 5 7.36-7.32 (m, 10 H), 6.58 (s, 2 H) 5.20 (s, 4 H). 13C NMR
(CDCl3) 5 184.1, 145.2, 136.1, 134.6, 128.5, 128.4, 128.1, 75.1. Anal. Calcd for
C20H1604: C, 74.99; H, 5.03. Found: c, 75.04; H, 5.06. HRMS (FAB) calcd for
C20H16O4 (M+): 320.1049. Found: 320.1059.

1,2,3,4-tetrahydroxybenzene (synthesis #1). A solution of 2,3-dibenzyloxy-1,4-
benzoquinone 26 (179 mg, 0.56 mmol) in EtOH (7.0 mL) with 10% Pd/C (50 mg) was
stirred at rt under an atmosphere of H2 for 3 h. The solution was filtered through Celite
and concentrated to afford 1,2,3,4-tretahydroxybenzene (79.0 mg, 99%) as a tan solid.
mp 162-164 °C. 1H NMR (d6-acetone) 5 7.24 (s, OH), 6.20 (s, 2 H). 13C NMR (d6-
acetone) 5 139.7, 134.7, 106.2.

2,3,4-tribenzyloxybenzaldehyde 27.21 N-methylformanilide (175 mL, 1.4 mol)
was treated with POCl3 (155 mL, 1.66 mol) at 22 °C under Ar resulting in a yellow solid.
After 2 h, the solid was treated with a solution of 1,2,3-tribenzyloxypyrogallol 25 (20 g,
51 mmol) in dry DMF (40 mL) and heated to 60 °C. After 3 h, the resulting crimson
solution was cooled to 22 °C then poured into ice water (3 L) with vigorous stirring for
12 h. The resulting brown precipitate was filtered, washed with hexanes (3 x 100 mL)
then recrystallized from MeOH to afford 2,3,4—tribenzyloxybenzaldehyde 27 (19.8 g,
93%) as a white powder. mp 73-74 °C. 1H NMR (CDCl3) 5 10.1 (s, 1H), 7.57 (d, J =
8.8 Hz, 1H), 7.4-7.3 (m, 15H), 6.82 (d, J = 8.8 Hz, 1H), 5.21(s, 2H), 5.16 (s, 2H), 5.08 (s,
2H); 13C NMR (CDC13)5 188.8, 158.5, 155.9, 141.0, 136.9, 136.2, 135.8, 128.7, 128.5,

128.5, 128.3, 128.2, 127.4, 124.0, 124.0, 109.1, 76.8, 75.5, 70.9. Anal. Calcd for

213

C23H24O4: C, 79.22; H, 5.70. Found: C, 79.17; H, 5.80. HRMS (FAB) calcd for
C23H2404 (M+): 424.1675. Found: 424.1669.

2,3,4-tribenzyloxyphenol 28.21 A solution of 2,3,4-tribenzyloxybenzaldehyde
27 (9.8 g, 23.1 mmol) in CH2C12 (50 mL) was treated with a solution of 30% H202 (6

mL, 57.8 mmol) and 85% formic acid (32 mL, 600 mmol) dropwise over 30 min at 0 °C.
After 1 h, the reaction was warmed to 22 °C. After 24 h, the reaction was cooled to 4 °C

and diluted with aqueous 10% NaZSO3 (50 mL). The CH2C12 was separated and the
aqueous phase was washed with CH2C12 (3 x 40 mL). The combined organic extracts
were dried (MgSO4) and concentrated in vacuo. The residue was dissolved in a

methanolic solution of NaOMe (30 mL, 0.1 N) and refluxed. After 10 min, the solution
was cooled to 4 °C and acidified with 6 N HCl. MeOH was removed in vacuo. The

mixture was diluted with H20 (15 mL) and the aqueous phase was extracted with
benzene (3 x 40 mL). The combined organic extracts were dried and concentrated in
vacuo to give 2,3,4-tribenzyloxyphenol 28 (9.0 g, 95%) as a brown oil. 1H NMR
(CDC13) 5 7.45-7.31 (m, 15 H), 6.65 (d, J = 9 Hz, 1H), 6.58 (d, J = 9 Hz, 1H), 5.28 (s, 1
H), 5.12 (s, 2H), 5.11 (s, 2H) 5.04 (s, 2 H); 13C NMR (CDCl3) 5 146.0, 144.0, 142.0,

139.6, 137.3, 137.1, 136.6, 128.4, 128.3 (2), 128.2, 127.9, 127.7, 127.4, 110.4, 109.0,

75.6, 75.3, 71.7. Anal. Calcd for C27H24O4: C, 78.62; H, 5.87. Found: C, 78.71; H,
5.86. HRMS (FAB) calcd for C27H24O4 (M+): 412.1675. Found: 412.1673.

1,2,3,4-tetrahydroxybenzene (synthesis #2). A solution of 2,3,4-
tribenzyloxyphenol 28 (5.8 g, 14.1 mmol) in EtOH (100 mL) was treated with 10% Pd/C

(1 g) under 1 atmosphere of H2 at 22 °C. After 2 h, the solution was filtered through

Celite® and the filtrate was concentrated in vacuo. The residue was purified by flash

214

chromatography (10% MeOH/CHzClz) affording 1,2,3,4-tetrahydroxybenzene (1.6 g,
80%) as a tan powder. mp 162-164 °C. 1H NMR (D20) 5 6.20 (s); 13C NMR (D20) 5

139.7, 134.7, 106.1.

Synthesis of Aurantiogliocladin
l,2,3,4-tetramethoxybenzene 29. A solution of 1,2,3,4-tetrahydroxybenzene
(2.0 g, 59 mmol) and dimethyl sulfate (37.5 mL, 396 mmol) in EtOH (21 mL) was added

dropwise to an 8.5 M aqueous solution of NaOH (42 mL) over 20 min at 22 °C. After 2
h, the rxn was diluted with H20 (300 mL) and cooled to -20 °C for 12 h. The resulting

precipitate was filtered, washed with H20 then recrystallized from hexanes to afford
1,2,3,4-tetramethoxybenzene 29 (8.12g, 69%) as colorless needles. mp 84-85 °C. 1H
NMR (CDC13) 5 6.58 (s, 2H), 3.90 (s, 6H) 3.82 (s, 6H); 13C NMR (CDC13) 5 147.7,
143.3, 106.3, 61.1, 56.3. Anal. Calcd for C10H1404: C, 60.59; H, 7.12. Found: C, 60.44;
H, 7.07.

2,3,4,5-tetramethoxytoluene 40. A solution of 1,2,3,4-tetramethoxybenzene
29 (4.0 g, 20.2 mmol) and TMEDA (6 mL, 38.0 mmol, 1.9 eq.) in hexanes (44 mL) and
THF (80 mL) was treated with 1.6 M n-BuLi in hexanes (16 mL, 25.6 mmol, 1.3 eq.)
dropwise over 10 min at 0 °C under Ar. After 30 min, at 0 °C under Ar, the reaction was
treated with methyl iodide (20 mL, 160 mmol) dropwise over 8 min. After 3 h, the
reaction was diluted with aqueous ammonium chloride (10 mL) and ether (20 mL). The
organic phase was washed with aqueous, concentrated ammonium hydroxide, H20, brine
then dried and concentrated. The residue was purified by flash chromatography (hexanes
then 5% EtOAc/hexanes) to afford 2,3,4,5-tetramethoxytoluene 40 as a clear oil (3.6g,
83%). 1H NMR (CDC13) 5 6.45 (s, 1H), 3.93 (s, 3H), 3.87 (s, 3H), 3.82 (s, 3H), 3.79 (s,
3H), 2.23 (s, 3H); 13C NMR (CDC13) 5 149.1, 146.9, 145.3, 140.7, 125.7, 108.2, 61.0,
60.9, 60.5, 55.9, 15.7.

215

l,2,3,4-tetramethoxy-5,6-dimethylbenzene 30. A solution of 1,2,3,4-
tetramethoxytoluene 40 (500 mg, 2.4 mmol) and TMEDA (0.7 mL, 4.5 mmol.) in
hexanes (5 mL) was treated with 1.6 M n-BuLi in hexanes (2 mL, 3.3 mmol) dropwise
over 20 min at 0 °C under Ar. After 30 min, at 0 °C under Ar, the reaction was treated
with methyl iodide (1 mL, 16 mmol) dropwise over 8 min. After 2.5 h, the reaction was
diluted with aqueous ammonium chloride (10 mL) and ether (20 mL). The organic phase

was washed with aqueous, concentrated ammonium hydroxide, H20, brine then dried and

concentrated. The residue was purified by Chromatotron (1 mm plate, hexanes then 5%

EtOAc/hexanes) to afford 30 as a clear oil (285 mg, 54%). 1H NMR (CDC13) 5 3.90 (s,
6H), 3.78 (S, 6H), 2.14 (S, 6H); 13C NMR (CDC13) 5 147.7, 144.6, 125.4, 61.1, 60.7,

12.1.

aurantiogliocladin. A solution of 30 (194 mg, 0.86 mmol) in H20 (1.7 mL) and
CH3CN (4 mL) was cooled to 0 °C followed by the addition of pyridine-2,6-
dicarboxylate (360 mg, 2.2 mmol). A chilled solution of (NH4)2Ce(NO3)6 (1.18 g, 2.2
mmol) in a H20 (1.2 mL) and CH3CN (1.2 mL) was added dropwise over 6 min to the
solution of 30. The reaction was stirred for 1 h at 0 °C followed by 20 min at rt. The
reaction was diluted with H20 (10 mL) then extracted CH2C12 (3 x 30 mL). The
combined organics were dried and concentrated to an orange solid. The solid was
purified by Chromatotron (1 mm, 15% EtOAc/Hexanes) to afford aurantiogliocladin (135
mg, 80%). 1H NMR (CDC13) 5 3.99 (s, 6H), 2.01 (s, 6H); 13C NMR (CDC13) 5 184.2,
144.3, 138.8, 61.1,12.1.

(2'E,6'E)-2-(3,7,ll-trimethyldodeca-2,6,l0-trienyl)-6-methyl-2,3,4,5-tetra-
methoxybenzene 41.22 n-BuLi (1.6 M, 0.9 mL) was added dropwise over 15 min to a
solution of 40 (0.212 g, 1.0 mmol) and TMEDA (0.3 mL, 1.9 mmol) in hexane (2.2 mL)

at 0 °C under Ar. This yellow precipitate-containing reaction mixture was then stirred at

216

0 °C under Ar for 30 min, diluted with THF (4 mL) and ether (11 mL), followed by
addition of CuCN (0.125 g, 1.4 mmol). After stirring for 30 min at 0 °C under Ar, the
temperature was reduced to -78 °C, and a solution of farnesyl bromide (0.285 g, 1.0

mmol) in hexane (2 mL) was added dropwise over 30 min. Further reaction for 3 h at -78

°C and subsequent slow warming to rt was followed by addition of saturated NH4Cl (10

mL) and ether (20 mL). Washing the organic phase with concentrated NH40H, water

and brine was followed by drying and concentrating. Purification of the residue by radial

chromatography (2 mm thickness, hexane/EtOAc, 9:1, v/v) afforded 41 as a clear oil
(0.236 g, 57%). 1H NMR (CDC13) 5 5.12-5.01 (m, 3H), 3.90 (s, 6H), 3.78 (s, 6H), 3.32

(d, J = 7.0 Hz, 2H), 2.14 (s, 3H), 2.08-1.91 (m, 8H), 1.77 (s, 3H), 1.58 (s, 6H). 13C NMR
(CDC13) 5 147.8, 147.6, 144.9, 144.6, 135.0, 134.9, 131.2, 129.2, 154.4, 124.3, 124.1,
122.8, 61.1, 60.6, 39.7, 26.7, 26.5, 25.7, 25.6, 17.6, 16.2, 15.9, 11.7.

Synthesis of Coenzyme Q3

(2'E,6'E)-2-(3,7,l1-Trimethyldodeca-2,6,10-trienyl)-3-methyl-5,6-dimethoxy-
1,4-benzoquinone coenzyme Q3.22 A solution of the 41 (125 mg, 0.3 mmol) in
acetonitrile (1.4 mL) and H20 (0.6 mL) was treated with pyridine-2,6-dicarboxylate (125

mg, 0.75 mol) at 0 °C resulting in a suspension. A solution of ceric ammonium nitrate

(411 mg, 0.75 mmol, 2.5 eq.) in acetonitrile (0.4 mL) and H20 (0.4 mL) previously
cooled to 0 °C was added dropwise to the suspension of 41 over 10 min. After 40 rrrin,
the reaction was warmed to 22 °C for 20 min. The reaction mixture was diluted with
H20 (10 mL) then extracted with CH2C12 (3 x 100 mL). The combined organic phases
were dried (MgSO4) and concentrated in vacuo. The residue was purified by radial

chromatography (1 mm, 5% EtOAc/hexanes) to afford coenzyme Q3 (53 mg, 46%) as an

217

orange oil. 1H NMR (CDC13) 5 5.07 (dd, J = 7, 7 Hz, 1H), 5.05 (dd, J = 7, 7 Hz, 1H),
4.94 (dd, J = 7, 7 Hz, 1H), 3.99 (s, 3H), 3.98 (s, 3H), 3.18 (d, J = 6.8 Hz, 2H), 2.08-1.91
(m, 8H), 2.01 (s, 3H), 1.74 (s, 3H), 1.67 (s, 3H), 1.59 (s, 3H), 1.58 (s, 3H); 13C NMR
(CDC13)5184.7, 183.9, 144.3, 144.2, 141.6, 138.8, 137.6, 135.2, 131.3, 124.3, 123.8.
118.8, 61.1, 39.7, 26.7, 26.4, 25.7, 25.3, 17.6, 16.3, 16.0, 11.9; HRMS (FAB) calcd for

C24H34O4 (M+): 386.2457. Found: 386.2461.

Synthesis of 2-deoxy-scyllo-inosose.

tetrabenzyl epoxide 42. Tetrabenzyl inositol 15 (25 g, 46.3 mmol) was dissolved
in dry benzene (625 mL) while under an atmosphere of Ar. To the solution was added p-
toluene sulfonic acid (150 mg) then trimethyl orthoacetate (20 mL, 157.2 mmol).23 The
solution was stirred for 12h at rt. The solution was concentrated in vacuo to a yellow oil.

To the oil was added dry CHzClz (375 mL) then trimethyl silyl chloride (7.6 mL,
60.2 mmol). The solution was stirred for 1h at rt followed by heating to reflux for 12h
under an atmosphere of Ar. The solution was cooled then concentrated in vacuo to a
yellow oil which was subsequently dissolved in MeOH (375 mL). To the solution was
added K2C03 (19.3 g) then the suspension was stirred for 12h at rt under an atmosphere
of Ar. To the suspension was added CH2C12 (80 mL) followed filtering then
concentrated in vacuo to a white solid. The epoxide was purified by flash
chromatography (10%-30% EtOAc/hexanes) to afford the epoxide 42 as a white solid
(18.2 g, 75%). mp 113-114 °C. 1H NMR (CDC13): 5 7.2-7.5 (m, 20H), 4.7-5.0 (m, 8H),
3.95 (d, J = 8.2 Hz, 2H), 3.68 (dd, J = 10.4, 8.5 Hz, 1H), 3.53 (dd, J = 8.8 Hz, 1H), 3.37

(br, 1H), 3.24 (d, J = 3.8 Hz, 1H); 13C NMR (CDC13): 5 138.5 (2), 138.1, 137.5, 128.5,

218

128.4, 128.3, 128.0, 127.9, 127.8 (2), 127.6, 127.5, 83.4, 79.2, 79.1, 78.9, 75.9, 75.5,
73.2, 73.0, 55.1, 53.8. Anal. Calcd for C34H3405: C, 78.14; H, 6.56. Found: C, 77.86;
H, 6.56. HRMS (FAB) calcd for C34H3405 (M+) : 522.2406. Found: 522.2394.
tetrabenzyl alcohol 43a and 43b. To a solution of the tetrabenzyl epoxide 42
(2.6 g, 5 mmol) in dry THF (62 mL) under an atmosphere of Ar at -78 °c was added
lithium aluminum hydride (380 mg, 10 mmol). After stirring at -78 °C for 15 min, the
temperature was raised to 0 °C. Once equilibrated at 0 °C, the temperature was allowed
to slowly reach rt over 12h. The reaction was quenched with saturated NH4C1 with
external cooling followed by dilution with H20 (20 mL). The aqueous was extracted
with diethyl ether (3 x 50 mL) and the combined organics were dried over NaZSO4 then
concentrated in vacuo to a yellow oil. Purification by flash chromatography (10%-30%
EtOAc/hexanes) to afford the alcohols 43a and 43b as a regio-isomeric mixture (1.91g,
73% combine). 433: 1H NMR (CDC13): 5 7.2-7.4 (m, 20H), 4.8-5.0 (m, 4H), 4.6-4.8 (m,
4H), 4.1 (d, J = 3.0 Hz, 1H), 3.94 (m, 1H), 3.84 (dd, J = 9.6, 9.3 Hz, 1H), 3.49 (m, 2H),
2.36 (ddd, J = 4.4 Hz, 1H), 1.38 (t, J = 13.5 Hz, 1H); 13C NMR (CDC13): 5 138.8, 138.7,
138.6, 137.9, 128.5, 128.3, 127.9(2), 127.7(2), 127.5, 127.4, 85.7, 82.7, 81.5, 76.0, 75.7,
72.9, 72.8, 65.9, 32.4. 43b: 1H NMR (CDC13): 5 7.2-7.4 (m, 20H), 4.9-5.1 (m, 3H), 4.7-
4.9 (m, 2H), 4.6-4.7 (m, 3H), 3.4-3.65 (m, 4H), 3.34 (t, J = 9.1, 1H), 2.33 (ddd, J = 4.7,
4.4 Hz, 1H), 1.46 (dd, J: 11.8, 12.4 Hz, 1H); 13C NMR (CDC13): 5 138.6, 138.4, 138.2,
128.6, 128.4, 128.3, 127.9 (2), 127.7, 127.6 (2), 85.5 (2), 83.3, 75.8, 75.4, 72.3, 68.3,

33.9.
tetrabenzyl-2-deoxy inosose 44. The regio-isomeric mixture of tetrabenzyl

alcohols 43a and 43b (10.8 g, 20.6 mmol) was dissolved in CHzClz (340 mL). To the

219

solution was added Dess-Martin periodinone (15.8 g, 37.1 mmol). After 2.5h, the
reaction was quenched by the addition of 1N NaOH (45 mL). After stirring vigorously
for 20 min, the organic layer was separated then washed with H20 (100 mL) then brine
(100 mL). After drying and concentrating in vacuo, the resulting yellow oil was purified
by flash chromatography (20% EtOAc/hexanes) to afford the protected deoxy-inosose 44
as a white solid (9.8 g, 91%). mp 68-70 °C. 1H NMR (CDC13): 5 725-75 (m, 20H),
4.94-4.53 (m, 8H), 4.24 (d, J = 9.5 Hz, 1H), 3.92 (t, J = 8.3 Hz, 1H), 3.6-3.7 (m, 2H),
2.82 (dd, J = 5.1, 4.9 Hz, 1H), 2.55 (dd, J = 10.7, 11.5 Hz, 1H); 13C NMR (CDC13): 5
203.6, 138.2, 137.7, 137.5, 128.5, 128.4, 128.1, 128.0, 127.9, 127.8, 127.7(2), 85.3, 84.0,
81.8, 76.1, 75.3, 73.5, 72.6, 42.8. Anal. Calcd for C34H3405: C, 78.14; H, 6.56. Found:
C, 78.09; H, 6.47. HRMS (FAB) calcd for C34H3405 (M+) : 522.2406. Found:

522.2489.
2-deoxy-scyllo-inosose. Tetrabenzyl-2-deoxy inosose 44 (17.8 g, 34.1 mmol)

was dissolved in a mixture of H20 (150 mL) in THF (650 mL). After blanketing with

Ar, 10% Pd/C (1.5 g) was added to the solution followed by stirring vigorously under an

atmosphere of H2. After 24h, the reaction was filtered through Celite followed by
washing with H20 (2 x 50 mL). The solution was concentrated in vacuo to afford 2-
deoxy-scylla-inosose as a white solid (5.5 g, 99%). 1H NMR (d6-DMSO): 5 4.00 (dd, J =
9.1, 9.1 Hz, 1H), 3.93-3.84 (m, 1H), 3.61 (dd, J = 9.3, 9,6 Hz, 1H), 3.14-3.02 (m, 2H),
2.93 (dd, J = 5.2, 5.2 Hz, 1H); 13C NMR (d6-DMSO): 5 206.9, 78.6, 77.0, 75.1, 69.0,

45.7.

220

Aromatization of 2-deoxy-scyllo-inosose.

l,2,4-trihydroxybenzene. To a degassed, aqueous solution of 0.5 M H3PO4 (6
mL) was added 2-deoxy-scylla-inosose (200 mg, 1.23 mmol). The solution was stirred at
125 0C for 12 h under an atmosphere of Ar. The resulting brown solution was extracted
with 3 x 20 ml t-butyl methyl ether, dried over MgSO4 and concentrated in vacuo. The
resulting brown oil was purified by column chromatography (10% EtOAc/hexane to
30%) to afford l,2,4-trihydroxybenzene (61 mg, 39%) as a white, crystalline solid. 1H
NMR (D20): 5 6.85 (d, J = 8.2 Hz, 1H), 6.56 (m, 1H), 6.41 (dd, J = 6.0, 1.9 Hz, 1H); 13C
NMR (D20): 5 149.6, 145.0, 137.5, 117.0, 106.9, 103.9.

Aromatization of triacetic acid lactone.

4-methoxy-6-methyl-2-pyrone 51 via dimethyl sulfate.24 Na2C03 (26.0 g, 206
mmol) was flame dried in a 1 L flask under vacuum. After cooling the Na2C03 to rt
under Ar, dry acetone (315 mL) was added followed by triacetic acid lactone (25.0 g, 198
mmol). To the stirred suspension was added dimethyl sulfate (18.8 mL, 198 mmol), and
the suspension was heated to reflux for 12 h. The mixture was cooled, and salts were
filtered away from the solution. The filtrate was concentrated in vacuo, and the residue
was recrystallized from EtOAc/hexanes to afford 4-methoxy-6-methyl-2-pyrone 51 (23.6
g, 169 mmol, 85%) as yellow crystals. mp 86-88 °C. 1H NMR (CDC13): 5 5.79 (s, 1 H),
5.41 (s, 1 H), 3.80 (s, 3 H), 2.21 (s, 3 H). 13C NMR (CDC13): 5 171.2, 164.9, 161.9,
100.3, 87.2, 55.7, 19.7.

4-methoxy-6-methyl-2-pyrone 51 via addition of methanol. Dowex-SO (5 .5 g)
and triacetic acid lactone (2.0 g, 16 mmol) were combined and MeOH (50 mL) was

added. The mixture was heated to reflux for 12 h, cooled to rt, and the Dowex-SO was

221

filtered. The filtrate was concentrated in vacuo and the residue recrystallized from
EtOAc/hexanes to afford 4-methoxy-6-methyl-2-pyrone 51 (1.0 g, 7.1 mmol, 43%) as
yellow crystals. mp 86-88 °C. 1H NMR (CDC13): 5 5.79 (s, 1 H), 5.41 (s, 1 H), 3.80 (s,

3 H), 2.21 (s, 3 H). 13C NMR (CDC13): 5 17121649, 161.9, 100.3, 87.2, 55.7, 19.7.

4-methoxy-6-methyl-2-pyrone 51 via trimethyl phosphate.25 Trimethyl
phosphate (9.7 mL, 83.3 mmol) was added to a mixture of 4-hydroxy-6-methyl-2-pyrone
(5 g, 39.7 mmol) and K2CO3 (6.6 g, 47.8 mmol). The suspension was mechanically
stirred under Ar at 140 °C for 1 h. After cooling to rt, the resulting solid was dissolved in
100 mL H20 and extracted with 4 x 50 mL EtOAc. The combined organics were dried
and concentrated in vacuo. The resulting yellow solid was recrystallized from
EtOAc/hexanes, filtered and washed with hexanes to afford 4-methoxy-6-methyl-2-
pyrone 51 (4.37 g, 31.2 mmol, 79%) as yellow crystals. mp 86-88 °C. 1H NMR
(CDC13): 5 5.79 (s, 1H), 5.41 (s, 1H), 3.80 (s, 3H), 2.21 (s, 3H); 13C NMR (CDC13): 5
171.2, 164.9, 161.9, 100.3, 87.2, 55.7, 19.7. Anal. Calcd for C7H303: C, 59.99; H, 5.75.
Found: C, 59.73; H, 5.75.

Methoxy-phloroglucinol 52.24 To MeOH (120 mL) cooled in ice was slowly
added small pieces of sodium (4.6 g, 200 mmol). After dissolving, the solution was
added via cannula to a solution of 4-methoxy-6-methyl-2-pyrone 51 (5 g, 35.7 mmol) in
MeOH (20 mL). The MeOH was distilled from the reaction under a blanket of Ar. The
resulting tan solid was heated to 185 °C for 30 min then cooled in ice while slowly
dissolving in 200 mL ice water. The pH was adjusted to 7.5 with 50% H2804 followed
by extraction with 5 x 100 mL t-butyl methyl ether. The combined organics were dried

then concentrated in vacuo to an orange oil. Purification by Kugelrohr distillation in

222

vacuo at 0.5 mm Hg and 120 °C afforded methoxy-phloroglucinol 52 (4.34 g, 31 mmol,
87%) as a yellow solid. mp 78-80 °C. 1H NMR (d-6-acetone): 5 8.17 (s, OH) 5.96-5.97
(m, 1H), 5.92-5.93 (m, 1H), 3.67 (s, 3H), 3.68; 13C NMR (CDC13): 5 162.6, 159.9, 96.2,
93.9, 55.2. Anal. Calcd for C7H303: C, 59.99; H, 5.75. Found: C, 60.09; H, 5.78.
phloroglucinol. Methoxy-phloroglucinol 52 (2 g, 14.3 mmol) was dissolved in
12 N HCl (100 mL).26 The orange solution was stirred for 36 h then quenched by slowly
adding solid Na2C03 (40 g). The precipitated NaCl was filtered and the aqueous solution

was extracted continuously for 24 h with t-butyl methyl ether. The organic was dried
then concentrated in vacuo to afford a brown, gummy foam. Kugelrohr distillation in
vacuo at 0.5 mm Hg and 110 °C afforded phloroglucinol (1.01 g, 8.0 mmol, 56%) as a
white powder. 1H NMR (d-6-acetone): 5 5.79 (s, 3H); 13C NMR (d—6-acetone): 5 160.1,
95.3.
Reduction methodology

resorcinol (via phloroglucinol). Phloroglucinol (1.0 g, 8.0 mmol) was dissolved
in degassed 1.0 N aqueous NaOH (8 mL). To the solution was 1.2 mol% of 5% Rh on
alumina catalyst. The suspension was shaken 12 h under 50 psi. H2. After filtering the
catalyst through Celite, pH was adjusted to 6.0 with 10% HCl. The solution was
concentrated in vacuo to a yellow oil which was subsequently dissolved in 0.5 M H2804
(20 mL) then refluxed under Ar for 9 hours. The cooled solution was extracted with
5x20 mL ether, then dried over MgSO4, filtered, then concentrated in vacuo. The
resulting brown oil was purified by Kugel-Rohr distillation in vacuo at 120 °C affording

resorcinol as white crystals. 1H NMR (d-6-acetone): 5 6.97 (dd, J = 8.0, 8.0 Hz, 1H), 6.35

— 6.30 (m, 3 H) ; 13C NMR (d-6-acetone): 5 159.4, 130.7, 107.4, 103.4.

223

resorcinol (via methoxy-phloroglucin0152). Methoxy-phloroglucinol 52 (1.4 g,
10.0 mmol) was dissolved in degassed 1.0 N aqueous NaOH (10 mL). To the solution
was added 1.2 mol % of 5% Rh on alumina catalyst. The suspension was shaken 12 h
under 50 psi. H2. After filtering the catalyst through Celite, pH was adjusted to 6.0 with
10% HCl. The solution was concentrated in vacuo to a yellow oil which was
subsequently dissolved in 0.5 M H2804 (20 mL) then refluxed under Ar for 9 hours. The
cooled solution was extracted with 5x20 mL ether, then dried and concentrated in vacuo.
The resulting brown oil was purified by Kugelrohr distillation in vacuo at 120 °C
affording resorcinol as white crystals. 1H NMR (d-6-acetone): 5 6.97 (dd, J = 8.0, 8.0 Hz,
1H), 6.35 — 6.30 (m, 3 H) ; 13C NMR (d-6-acetone): 5 159.4, 130.7, 107.4, 103.4.

pyrogallol. 1,2,3,4-tetrahydroxybenzene (0.7 g, 5 mmol) was dissolved in
degassed 1.0 M aqueous NaOH (5 mL). To the solution was added 1.2 mol% of 5% Rh
on alumina catalyst. The suspension was shaken 12 h under 50 psi. H2. After filtering
the catalyst through Celite, pH was adjusted to 6.0 with 10% HCl. The brown solution
was concentrated in vacuo to an oil which was subsequently dissolved in 0.5 M H2S04
(20 mL) then heated to reflux under Ar for 12 hours. The cooled solution was extracted
with 4x50 mL ether then dried and concentrated to afford a brown oil. Purification by
Kugel-Rohr distillation in vacuo at 0.5 mm Hg and 90 °C afforded pyrogallol as white
crystals. 1H NMR (d-6-acetone): 5 6.51 (dd, J = 7.4, 7.4 Hz, 1H), 6.36 (d, J = 7.4 Hz,
2H); 13C NMR (d—6-acetone): 5 146.7, 133.7, 119.9, 108.0.

hydroquinone. 1,2,4-tetrahydroxybenzene (500 mg, 4 mmol) was dissolved in
degassed 1.0 M aqueous NaOH (4 mL). To the solution was added 1.2 mol % catalyst.

The suspension was shaken 12 h under 50 psi. H2. After filtering the catalyst through

224

Celite, pH was adjusted to 6.0 with 10% HCl. The brown solution was concentrated in
vacuo to an oil which was subsequently dissolved in 0.5 M H2804 (16 mL). The solution
was refluxed under Ar for 24 hours then cooled and extracted with 5x20 mL ether. After
drying and concentrating, the resulting brown oil was purified by Kugelrohr distillation in
vacuo at 0.5 mm Hg and 110 °C affording hydroquinone (233 mg, 53%) as white
crystals. 1H NMR (d-6-acetone): 5 6.65 (s, 4H); 13C NMR (d-6-acetone): 5 151.1, 116.5.

Reductions carried out with 5% Rh/C, 5% Pt/C and 5% Pd/C were conducted under the

same conditions as described for 5% Rh on alumina.

Chapter 4
Plasmid constructions

Plasmid pCH4.184A.

This 7.4-kb plasmid was constructed by ligating the serA locus into pKAD.62A.
A 1.9-kb E. coli serA locus was liberated from pD2625 by digestion with Dra1 and
EcoRV. Plasmid pKAD.62A was digested with EcoRI, and the overhanging ends were
eliminated with Klenow fragment. Subsequent ligation of serA to pKAD.62A afforded
plasmid pCH4.184A with serA at the 3’ end of kanR gene and transcribed in the same
orientation as kanR.
Plasmid pCH4.264.

This 7.2-kb plasmid was constructed by inserting the kanRserA cassette into
pTrc.99A. The kanRserA cassette was amplified from pCH4.184A using the following
primers containing EcoRV terminal recognition sequences: 5’-GGATATCCG-

TTGTGTCTCAAAATCTCTG and 5’-GGATATCTTAGTACAGCAGACGGGCGC.

225

The amplified 3.0-kb PCR fragment was ligated into the EcoRV site of pTrc.99A to

create plasmid pCH4.264.

Plasmid pCH4.267A.

This 16.4-kb plasmid was created by inserting the kanRserA cassette into the Seal
site located in the ApR gene of pGM44. The 3.0-kb kanRserA cassette was liberated from
pCH4.264 by digestion with EcoRV. The plasmid pGM44 was partially digested with
Sca1 followed by gel purification. The linearized DNA was excised from a 0.4% agarose
gel and isolated with the Zymoclean DNA extraction kit. The kanRserA cassette was
ligated into Sca1 partially digested pGM44. Transformants were selected for resistance to
Kan and sensitivity to Ap affording pCH4.267A. The kanRserA cassette was oriented

such that transcription was in the same direction as the ApR gene.

Plasmid pCH5.148.

This 4.2-kb plasmid was constructed by removing the NcoI site in the multiple
cloning site of pTrc.99A. The vector pTrc.99A was digested with NcoI, and treated with
Klenow fragment. Ligation of the plasmid afforded pCH5.148, which lacked a NcaI

restriction site.

Plasmid pCH5.174A.

This 4.9-kb plasmid was constucted by ligation of a 0.75-kb fragment encoding
the E. coli fabG gene into pCH5.148. The fabG gene was amplified from RB791

genomic DNA with primers obtained from Genosys Sequences (fabG-A and fabG-C)

226

which contained SapI terminal recognition sequences. The resulting 0.75-kb PCR
fragment was digested with SapI and treated with Klenow fragment. Subsequent ligation
into pCH5.148 which had been previously digested with Sma1 afforded pCH5.174A. The

fabG gene was in the same orientation as ApR of pCH5.148.

Plasmid pCH5.183B.

This 6.0-kb plasmid was constructed by inserting the 1.1-kb kanR gene into
pCH5.174A. The kanR gene was amplified from pKAD.62A using the following primers
containing terminal Ne 01 recognition sequences: 5’-GTCCATGQCG-
TTGTGTCTCAAAATCTCTG and 5’-GGCQATQG'I'I‘GATGAGAGCTI‘TG'I'TGTAG.
The amplified 1.0-kb fragment was ligated into the NcoI site internal to the fabG gene of
pCH5.174A to afford pCH5.183B. The kanR gene was oriented such that transcription

was in the same direction as fabG.

Plasmid pCH5.263A.

This 8.5-kb plasmid was created by inserting the fabGzzkanR cassette into the
multiple cloning site of pMAK705.27 Digestion of pCH5.183B with HpaI and Sca1
liberated a 3.0-kb fragment containing the fabGzzkanR cassette. The cassette was excised
from a 0.4% agarose gel and isolated with the Zymoclean DNA extraction kit. The

cassette was ligated into the HincII site of pMAK705 to afford pCH5.263A.

227

Plasmid pCH5.213.

This 4.0-kb plasmid was created by inserting the initial 1.5-kb of fasB amplified
from pGM44 into the Ndel and Hindlll sites of pT7-728 behind the P17 promoter. The
initial 1.5-kb of fasB was amplified from pGM44 with a single amino acid change to
convert the less common GTG start codon to an ATG start codon. The 5’ primer
contained a terminal Ndel recognition sequence incorporating the single mutation: 5’-
GCTGGACATATGACTATTGGCATCTCTAACC. The 3’ primer contained Natl and
H indIII terminal recognition sequences to facilitate reconstruction of fasB: 5’-
GTCAAGCTTGCGGQCGCATATTGTC'I'I‘CCTGGAG'IT. The 1.5-kb fragment was
ligated into pT7-7 which had previously been digested with Ndel and Hindlll to afford

pCH5.213.

Plasmid pCH5.287A.

This 7.0-kb plasmid was constructed by ligation of the kanRserA cassette into the
ApR gene of pCH5.213. The 3.0-kb kanRserA cassette was liberated from pCH4.264 by
digestion with EcoRV. The cassette was excised from a 0.4% high purity agarose gel and
isolated with the Zymoclean DNA extraction. The cassette was inserted into Sca1
digested pCH5.213 to afford pCH5.287A. The kanRserA cassette was oriented in the

same direction as the ApR gene.

Plasmid pCH5.303.
This 16.0-kb plasmid was constructed by inserting a 9.5-kb fragment from

pGM44 encoding the remainder of fasB and PPTI into pCH5.287A. The 9.5-kb

228

fragment was liberated from pGM44 by digestion with Rer and Natl followed by gel
purification using the Zymoclean DNA purification kit. Plasmid pCH5.287A was
digested with Rer and NotI followed by gel purification of the 6.5-kb fragment using the
Zymoclean DNA purification kit. The 9.5-kb fragment was ligated into the 6.5-kb

fragment of pCH5.287A to afford pCH5.303.

Mutagenesis
Site-directed mutagenesis of FAS-B. Site-directed mutagenesis was performed using
the procedure of Horton and Pease.29 Plasmid pGM44 was linearized with Natl and
subsequently purified using the Wizard PCR Prep DNA Purification System (Promega)
according to the manufacturer’s protocol. A 187-bp fragment AB-l was amplified from
linearized pGM44 using the following primers: 5’-CTCGGCGCGTGAA-GACCTCGTG
and 5’-GAAGCACCGGTGACAACAGCAA. A 2.06-kb C-l-D fragment was amplified
from linearized pGM44 using the following primers: 5’-TTGTC-
ACCGGTGCTTCGCCTGG and 5’-AGGCCCAA'I'I‘CCGCAGCCAAGC. Amplified
DNA fragments were separated by gel electropheresis (0.4% agar) and purified using the
Zymoclean gel extraction kit (Zymo) according to the manufacturer’s protocol. The 2.2-
kb A-l-D fragment containing the site-directed mutation was amplified using 500 ng of
fragments AB-l and C-l-D as template with the following primers: 5’-CTCGGCGCGT-
GAAGACCTCGTG and 5’-AGGCCCAA'ITCCGCAGCCAAGC. The 2.2-kb fragment
was purified by gel electropheresis (0.4% agar) and then extracted by the Zymoclean Gel
Extraction Kit (Zymo) according to the manufacture’s protocol. Digestion with XbaI and

Seal liberated a 2.0 kb fragment. The gel purified DNA was ligated into the XbaI and

229

Sca1 digested pCH4.267 to give pCH5.28M1, a plasmid containing the fasB gene coding
for FAS-B with a single mutation.

The triple mutant of FAS-B was created by amplifying a 187-bp AB-2 fragment
from linearized pGM44 using the following primers: 5’-CTCGGCGCGTGAAGACCTC-
GTG and 5’-AGAGAGGCGAAGCAACGGTGACAACAGCAAC. A 2.06-kb C-2-D
fragment was amplified from linearized pGM44 using the following primers: 5’-CGTTG-
CIT CGCCTCTCTCTATTGCCTCGGAA and 5’-AGGCCCAA'ITCCGCAGCCAAGC.
Amplified DNA fragments were separated by gel electropheresis (0.4% agar) and
purified using the Zymoclean Gel Extraction Kit (Zymo) according to the manufacture’s
protocol. The 2.2-kb A-2-D fragment containing the site-directed mutations was
amplified using 500 ng of fragments AB-2 and C-2-D as template with the following: 5’-
CTCGGCGCGTGAAGACCTCGTG and 5’-AGGCCCAA'ITCCGCAGCCAAGC. The
2.2-kb fragment was purified by gel electropheresis (0.4% agar) and then extracted by the
Zymoclean Gel Extraction Kit (Zymo) according to the manufacture’s protocol.
Digestion with XbaI and Seal liberated a 2.0-kb fragment. The gel purified DNA was
ligated into the XbaI and Sca1 digested pCH4.267 to give pCH5.110M2, a plasmid
containing the fasB gene coding for FAS-B with three mutations.

The quadruple mutant of FAS-B was created by amplifying a 187-bp AB-3
fragment from linearized pGM44 using the following primers: 5’-CTCGGCGCGTGAA-
GACCTCGTG and 5’-AGAGAGGCGTAGCAACGGTGACAACAGCAAC. A 2.06-kb
C-3-D fragment was amplified from linearized pGM44 using the following primers: 5’-
CGTI‘GCTACGCCTCTCTCTATTGCCTCGGAA and 5’-AGGCCCAA'ITCCGCAGC-

CAAGC. Amplified DNA fragments were separated by gel electropheresis (0.4% agar)

230

and purified using the Zymoclean Gel Extraction Kit (Zymo) according to the
manufacture’s protocol. The 2.2-kb A-3-D fragment containing the site-directed
mutations was amplified using 500 ng of fragments AB-3 and C-3-D as template with the
following primers: 5’-CTCGGCGCGTGAAGACCTCGTG and 5’-AGGCCCAATTCC-
GCAGCCAAGC. The 2.2-kb fragment was purified by gel electropheresis (0.4% agar)
and then extracted by the Zymoclean Gel Extraction Kit (Zymo) according to the
manufacture’s protocol. Digestion with Xbal and Sca1 liberated a 2.0-kb fragment. The
purified DNA was ligated into the X bal and Sca1 digested pCH4.267 to give

pCH5.111M3, a plasmid containing the fasB gene coding for FAS—B with four mutations.

Homologous Recombination. Conditions for homologous recombination were based on
those previously described. The E. coli strain JWF1 was first transformed with the
plasmid pHPS630, harboring fasA and PPTI from B. ammoniagenes, followed by plating

on LB supplemented with Ap. Competent JWF1/pHPS6 cells were freshly prepared and

transformed with pCH5.263A, a pMAK70510 derivative containing fabG::kanR from E.
coli. Following heat-shock treatment, cells were incubated in LB at 37 °C for 1.5 h and
subsequently plated on LB containing Cm, Kan, and Ap. Plates were first incubated at 30
°C for 36 h. Colonies were replicated onto plates of LB containing Cm, Kan, and Ap
followed by incubation for 24 h at 42 °C. Replicates that grew at 42 °C were inoculated
into LB containing Ap for retention of pHPS6, and cells were grown at 30 °C for 24 h to
allow excision of plasmid from the genome. Cultures were diluted (1:10,000) in LB
containing Ap, and two more cycles of growth at 30 °C for 24 h were carried out to

enrich cultures for more rapidly growing cells that had lost the temperature-sensitive

231

replicon. Cultures were then diluted (1:10,000) in LB containing Ap at 42 °C for 24 h to
promote loss of pCH5.263A from the cells. Serial dilutions of each culture were spread
onto LB plates containing Kan and Ap followed by incubation at 30 °C for 24 h. The
resulting colonies were screened for the correct phenotype by showing resistance to Kan
and Ap and sensitivity to Cm. No colony exhibited sensitivity to Cm while retaining Kan

resistance.

Fermentations with mutated FAS-B. Fermentations employed a 2.0 L working
capacity B. Braun MD2 culture vessel. Utilities were supplied by a B. Braun Biostat MD
controlled by a Dell Otiplex Gs+ 5166 personal computer equipped with B. Braun
MFCS/Win software. PID control loops were used to control temperature, pH, and
glucose addition. The temperature was maintained at 33 °C, and the pH was maintained
at 7.0 by addition of concentrated NH40H or 2N H2804. Dissolved oxygen (D.O.) was
measured using a Mettler-Toledo 12 mm sterilized 02 sensor fitted with an Ingold A-type

02 permeable membrane. D.0. was maintained at 20% air saturation. Antifoam (Sigma

204) was added manually as needed.

RB791 serA::aroB/pCH5.28Ml. Each inoculant was initiated by introduction of a
single colony of RB791 serA::araB/pCH5.28M1 into 5 mL of M9 medium containing
kanamycin and grown at 37 °C with agitation for 12 to 24 h until the culture was turbid.
After this time, 100 uL of the starter culture was transferred to 100 mL of M9 medium
containing kanamycin and grown for an additional 12 to 24 h at 37 °C and 250 rpm.

After an appropriate OD600 was reached (1.0 — 3.0), the inoculant was transferred to the

232

fermentation vessel. The initial glucose concentration in the fermentation medium was
20 g/L. Three staged methods were used to maintain D.O. levels at 20% air saturation
during the course of each fermentor run. 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 its preset maximum of 940 rpm. Approximately 8 h was
required for the impeller speed to increase to 940 rpm. With the impeller constant at 940
rpm, the mass flow controller then maintained D.O. levels by increasing the airflow rate
from 0.06 L/L/min to its preset maximum of 1.0 LIL/min over approximately 2.0 h. At
constant impeller speed and constant airflow rate, D.O. levels were maintained at 20%
saturation for the remainder of the fermentation by oxygen sensor-controlled glucose
feeding. The PID control parameters were set to 0.0 (off) for the derivative control (1.1))
and 999.9 3 (minimum control action) for the integral control (11). Xp was set to 950% to
achieve a Kc of 0.1.

Samples (6 mL) of fermentation broth were taken at 6h intervals starting at 12h.
Cell densities were determined by dilution of the fermentation broth with water (1:10)
followed by measurement of absorption at 600 nm (OD600). Dry cell weight (g/L) was
obtained using a conversion coefficient of 0.43 g/L/ OD600. Fermentation broth was
centrifuged to remove cells. Solute concentrations in cell-free broth were determined by
1H NMR. Compounds were quantified using the following resonances: triacetic acid

lactone (52.24, s, 3H).

233

JWFl(DE3)/pCH5.303. The procedure for evaluating FAS-B with the T7 promoter
under fed-batch fermentor conditions was carried out as described above. No triacetic

acid lactone was observed by 1H NMR.

TAL via YZ166

A. A single colony of YZ16631 was used to inoculate 5 mL of LB containing Ap,
Tc, Cm, Sp and 0.2% L-arabinose. The cell were grown for 12 h at 37 °C and 250 rpm.
This inoculant was then transferred to 500 mL of LB also containing 0.2% L-arabinose
and antibiotics which was grown at 37 °C and 250 rpm. Cell growth was monitored by
following the OD600. When the 0D600 reached approximately 3.0, the cells were
collected by centrifuging at 4200g at 6 °C for 6 min. The cells were washed with 100 mL
M9 medium, centrifuged at 4200g at 6 °C for 6 min, then resuspended in 500 mL M9
medium supplemented with thiamine (5 mg), methionine (50 mg) with 0.4 % D-glucose
and 0.025% L-fucose. The cells were shaken at 37 °C and 250 rpm while monitoring
OD600 and synthesis of TAL by 1H NMR. After 36 h, all D-glucose was consumed. The

cells reached a final 0D600 of 4.8 with no production of TAL.

B. A single colony of YZl66 was used to inoculate 5 mL of LB containing Ap,
Tc, Cm, Sp and 0.2% L-arabinose. The cells were grown for 12 h at 37 °C and 250 rpm.
This inoculant was then transferred to 250 mL of LB with 0.2% L-arabinose and
antibiotics which was grown at 37 °C and 250 rpm. Cell growth was monitored by
0D600- When the OD600 reached approximately 3.0, D-glucose was added to the flask to

a final concentration of 0.4 %. The flask was shaken 24 h at 37 °C and 250 rpm prior to

checking the supematent by 1H NMR. No triacetic acid lactone was observed.

234

C. A single colony of YZ166 was used to inoculate 5 mL of LB containing Ap,
Tc, Cm, Sp, 0.4% D-glucose, and oleic acid (0.4 mL/L). The cells were grown for 12 h at
37 °C and 250 rpm. This inoculant was then tranSferred to 500 mL of LB with 0.4% D-
glucose, oleic acid and antibiotics which was grown at 37 °C and 250 rpm. Cell growth
was monitored every 24 h by OD600. Supematents were checked by 1H NMR to monitor

accumulation of TAL. After 48 h, no TAL was observed with a final OD600 of 4.8.

D. A single colony of YZl66 was used to inoculate 5 mL of LB containing
containing Ap, Tc, Cm, Sp, thiamine (5 mg), methionine (50 mg) and glycerol (4g/L).
The cells were grown for 24 h at 37 °C and 250 rpm. This inoculant was transferred to
250 mL of M9 medium containing glycerol, thiamine, methionine and antibiotics which

was grown at 37 °C and 250 rpm Cell growth was monitored by following the 0D600.

When the OD600 reached approximately 1.7, glycerol was depleted as indicated by 1H

NMR. To the medium was added 5 mL of 0.4% glycerol. After 60 h, no triacetic acid

lactone was observed by 1H NMR.

Polyketide Biosynthesis
Fermentation with yeast. Saccharomyces cerevisiae strain Inchl (MATOL his3AI leu2
trp1-289 ura3-52) was obtained from Invitrogen. The plasmid pKOS12-128a32,
containing the sfp gene with the T rp selection marker, was obtained from J. Kealey of
Kosan Biosciences. The plasmid pMR22833, containing the mutated 6-MSAS with the
selection markers for Leu and Ura, was obtained from C. Khosla of Stanford University.
Both plasmids were co-transforrned into Inchl by electroporation with the Bio-Rad

Gene Pulser 11 using protocols available from the manufacturer. The transformed cells

235

were immediately plated on SC minimal medium supplemented with sorbitol and
excluding tryptophan or uracil. The plates were incubated at 30 °C for 96h.

A single colony of Inchl/pKOS12-128a/pMR228 was used to inoculate 10 mL
of SC minimal medium without tryptophan or uracil supplementation. The inoculant was
grown with agitiation for 48 h at 30 °C and 220 rpm. A 100 uL aliquot of this growth
was transferred to 10 mL of YPD which was grown 18 h at 30 °C and 220 rpm. From
this growth, 0.5 mL was transferred to 50 mL of YPD which was grown at 30 °C and 220
rpm. Samples (1-2 mL) were taken every 24 h for a duration of 144 h. Cells were
removed by centrifugation and the cell-free supernatant was checked for triacetic acid

lactone by 1H NMR in D20. No triacetic acid lactone was observed by 1H NMR.

236

References
1 Bradford, M. M. Anal. Biochem. 1976, 72, 248.
2 Chen, P. S.; Toribara, T. Y.; Warner, H. Anal. Chem. 1956, 28, 1756.
3 Ames, B. N. Methods Enzymol. 1966, 8, 115-118.
4 (a) Meurer, G.; Biermann, G.; Schutz, A.; Harth, S.; Schweizer, E. Mol. Gen.
Genet. 1992, 232, 106. (b) Stuible, H.-P.; Wagner, C.; Andreou, I.; Huter, G.;
Haselmann, J .; Schweizer, E. J. Bacteriol. 1996, 178, 4787.
5 Ramaley, R.; Fujita, Y.; Freese, E. J. Biol. Chem. 1979, 254, 7684.

6 Sambrook, J.; Fritsch, E. F.; Maniatis, T. In Molecular Cloning: A Laboratory
Manual; Cold-Spring Harbor Laboratory: Plainview, NY, 1990.

7 De Ley, J .; Swings, J. In Bergey’s Manual of Systematic Bacteriology; Krieg, N.
R.; Holt, J. G., Eds.; Williams and Wilkins: Baltimore; 1984, Vol. 1, pp. 275-278.

8 Sherman, F. In Guide to Yeast Genetics and Molecular Biology; Guthrie, C.; Fink,
G. R., Eds.; Academic Press: San Diego; 1991, Vol. 194, p. 13.

9 (a) Sherman, F. In Guide to Yeast Genetics and Molecular Biology; Guthrie, C.;
Fink, G. R., Eds.; Academic Press: San Diego; 1991, Vol. 194, p. 14-15. (b) Ausubel, F.
M. In Current Protocols in Molecular Biology; Green and Wiley: New York; 1987.

‘0 Becker, D. M.; Guarente, L. In Guide to Yeast Genetics and Molecular Biology;
Guthrie, C.; Fink, G. R., Eds.; Academic Press: San Diego; 1991, Vol. 194, pp. 182-187.

11 Miller, J. H. Experiments in Molecular Genetics; Cold Spring Harbor Laboratory:
Plainview, NY, 1972.

12 Crevan, M.; Fitzgerald, R. J .; O’ Gara, F. Can. J. Microbiol. 1994, 40, 491.

13 (a) Shinjoh, M.; Hoshino, T. J. Ferment. Bioeng. 1995, 79, 95. (b) Fujiwara, A.;
Hoshino, T.; Shinjoh, M., U. S. Patent 5,399,496, 1995.

14 Posternak, T. Biochem. Preparations 1952, 2, 57.

15 Anderson, L.; Takeda, R.; Angyal, S. J .; McHugh, D. J. Arch. Biochem. Biophys.
1958, 78, 518.

237

16 (a) Post, G. G.; Anderson, L. J. Am. Chem. Soc. 1962, 84, 471. (b) McCasland, G.
E. In Methods in Carbohydrate Chemistry; Whistler, R. C., Wolfram, M. C. Eds.;
Academic Press: New York, 1962; pp. 291-294.

17 Grigg, R.; Warren, C. D. J. Chem. Soc. C 1969, 2367.

‘8 Angyal, S. J .; Tate, M. E. J. Chem. Soc. 1965, 6949.

19 Kolontis, P.; Major, A.; Négradi, M. Acta. Chim. Hung. 1983, 113, 367.

20 Y Matsumoto, M.; Kobayashi, H.; Hotta, Y. J. Org. Chem. 1985, 50, 1766.

21 Kolontis, 1).; Major, A.; Nogradi, M. Acta. Chim. Hung. 1983, 113, 367.

22 Keinen, E.; Eren, D. J. Org. Chem. 1987, 52, 3872.

23 Kolk, H. C.; Sharpless, K. B. Tetrahedron 1992, 48, 10515.

24 Huber, U., U. S. Patent 4,112,003, 1978.

25 Nelson, R. B. U. S. Patent 4,453,017, 1984.

26 (a) Huber, U., U. S. Patent 4,112,003, 1978. (b) McKillop, A.; Howarth, B. D.;
Kobylecki, R. J. Synth. Commun. 1974, 4, 35.

27 Hamiloton, C. M.; Aldea, M.; Washbum, B. K.; Babitzke, P.; Kushner, S. R. J.
Bacteriol. 1989, 171, 4617.

28 Tabor, S.; Richardson, C. C. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 1074.

29 Horton, R. M.; Pease, L. R. In Directed Mutagenesis: A Practical Approach;
McPherson, M. J ., Ed.; Oxford University Press: New York, 1991; pp.217-247.

30 Stuible, H.-P.; Meier, S.; Schweizer, E. Eur. J. Biochem. 1997, 248, 481.
31 Zhang, Y.; Cronan, J. E., Jr. J. Bacterial. 1998, 180, 3295.

32 Kealey, J. T.; Liu, L.; Santi, D. V.; Betlach, M. C.; Barr, P. J. Proc. Natl. Acad.
Sci. USA 1998, 95, 505.

33 Richardson, M. T.; Pohl, N. L.; Kealey, J. T.; Khosla, C. Metab. Eng. 1999, 1,
180.

238

          

 
 

   

. ' ” l'lllll‘ljllllgl