THE POLYKETIDE ORIGINS
OF CANNABINOIDS IN
CANNABIS SATIVA
A Thesis Submitted to the College of
Graduate Studies and Research
In Partial Fulfillment of the Requirements
For the Degree of Doctor of Philosophy
In the Department of Biology
University of Saskatchewan
Saskatoon
By
STEVE JOSEPH GAGNE
Copyright Steve J. Gagne, 2013. All rights reserved.
PERMISSION TO USE
In presenting this thesis in partial fulfilment of the requirements for a Postgraduate degree from
the University of Saskatchewan, I agree that the Libraries of this University may make it freely
available for inspection. I further agree that permission for copying of this thesis in any manner,
in whole or in part, for scholarly purposes may be granted by the professor or professors who
supervised my thesis work or, in their absence, by the Head of the Department or the Dean of the
College in which my thesis work was done. It is understood that any copying or publication or
use of this thesis or parts thereof for financial gain shall not be allowed without my written
permission. It is also understood that due recognition shall be given to me and to the University
of Saskatchewan in any scholarly use which may be made of any material in my thesis.
Requests for permission to copy or to make other use of material in this thesis in
whole or part should be addressed to:
Head of the Department of Biology
University of Saskatchewan
Saskatoon, Saskatchewan, S7N 5E2
i
ABSTRACT
Phytocannabinoids are the active substances responsible for the medicinal and psychotropic
effects of Cannabis sativa. Although the bioactivity of cannabis and its preparations have been
known for millennia, several steps in the biosynthetic pathway leading to phytocannabinoids
remain unclear. Phytocannabinoids are prenylated resorcylic acids which are formed in
specialized plant organs called glandular trichomes. Following the analysis of a pre-generated
cannabis trichome cDNA library, a type III polyketide synthase (tetraketide synthase; TKS) was
identified and assayed, yielding three major compounds, hexanoyl triacetic acid lactone (HTAL),
pentyl diacetic acid lactone (PDAL), and olivetol, yet no resorcylic acid was detected. This lack
of resorcylic acid in enzyme assays has instigated the characterization of TKS and a search for
putative cyclases in the cannabis trichome cDNA library, and involved protein pulldown, coimmunoprecipitation, and co-assay experiments. These experiments led to the discovery of a
novel polyketide cyclase protein named olivetolic acid cyclase (OAC) responsible for the proper
cyclization of a polyketide intermediate produced by TKS. This thesis shows that TKS assays
conducted with OAC produce olivetolic acid (OA), an intermediate required during the
biosynthesis of cannabinoids. The TKS/OAC spatial relationship was also investigated following
the creation of fluorescent fusion proteins which show that the enzymes co-localized in vivo
when viewed with confocal microscopy. Furthermore, yeast two-hybrid assays using TKS and
OAC were performed to establish whether the enzymes physically interact. Finally, an attempt to
determine the responsible amino acids involved in OAC’s mechanism was conducted by
comparing the activity of single point OAC mutants with the wild-type OAC. Based on the
available data, mechanisms for the production of HTAL, PDAL, olivetol, and OA are proposed.
ii
ACKNOWLEDGEMENTS
Although the research presented in this dissertation is my work, this project would not have been
possible without the involvement of many people who have ceaselessly provided their valuable
support. I would primarily like to thank my supervisor, Dr. Jonathan Page, who offered me the
opportunity to tackle the cannabinoid conundrum, and continued to provide encouragement and
guidance throughout each process of this work. I’d like to also thank Jon for the important bits of
advice he has passed on to me with regard to professionalism and work flow, which will
continue to assist me in my future career, whether it is in industry, academia, or government
settings. I also warmly thank each member of my committee who have provided important
comments that helped to shape this thesis to its current state; each have generously provided
flexibility with their time and personally contributed important input during the course of this
project; thank you Dr. Gordon Gray, Dr. Chris Todd, and Dr. David Palmer. I also thank the
external examiner, Dr. Reinhart Jetter, who has also contributed a lot of significant input during
the final stages of my dissertation preparation.
I would like to express my sincere gratitude to each lab member, whether past or present,
which I had the pleasure to work with during these last few years. This includes Dr. Jake Stout,
Dr. Shawn Clark, Enwu Liu, David Konkin, Sandra Polvi, Dr. Zakia Boubakir, Shawn Whitfield,
and Nikki Theaker. Aside from providing their invaluable technical expertise, exemplary
teaching and guidance, and stellar participation and inspiration, I am deeply appreciative for their
friendship, openness, and for being the best lab mates one could ask for.
I would also like to thank each and every individual that I had the opportunity to meet at
the University of Saskatchewan and the Plant Biotechnology Institute. The environment that
iii
these people have collectively created is one of highest integrity and good cheer. A special
thanks to Dr. Patrick Covello, Darwin Reed, Stephen Ambrose, Dr. Shawn Gibson, Sharla
Lozinsky, Dr. Julien Cotelesage, Deidre Wasyliw, Brenda Haug, Dr. Leonid Akhov, Ioannis
Mavraganis, and Aaron Karapinka; your precious words have coaxed me to the finish line.
I acknowledge the financial support provided to me by the Natural Sciences and
Engineering Research Council of Canada (NSERC), the Department of Biology, and the
University of Saskatchewan.
Lastly, but most importantly, I would like to thank my best friend and wife, Christine, as
well as my children, Spring and Sethia Om, for I know well enough that without their enduring
love, kindness, and patience, I would not have completed one word of this dissertation.
iv
TABLE OF CONTENTS
PERMISSION TO USE ................................................................................................................... i
ABSTRACT .................................................................................................................................... ii
ACKNOWLEDGEMENTS ........................................................................................................... iii
TABLE OF CONTENTS ................................................................................................................ v
LIST OF TABLES ....................................................................................................................... xiii
LIST OF FIGURES ..................................................................................................................... xiv
LIST OF ABBREVIATIONS .................................................................................................... xviii
1. INTRODUCTION ...................................................................................................................... 1
1.1 Research Goals and Objectives .............................................................................................. 2
2. LITERATURE REVIEW ........................................................................................................... 3
2.1 Overview of Cannabis sativa and Cannabinoids ................................................................... 3
2.1.1 The Biology of Cannabis .................................................................................................. 5
2.1.2 Bioactivity of Cannabinoids............................................................................................ 11
2.1.3 The Current Understanding of Cannabinoid Biosynthesis.............................................. 11
2.2 Olivetolic Acid Biosynthesis ................................................................................................ 14
2.2.1 Type III PKSs: Diversity, Structure and Mechanism ..................................................... 14
2.2.2 Instances of Olivetolic Acid Biosynthesis ...................................................................... 21
2.2.3 The Biosynthesis of Olivetolic Acid Requires a Polyketide Synthase ........................... 21
2.2.4 Type III PKSs from Cannabis ......................................................................................... 22
2.3 Polyketide Cyclization and DABB Proteins ......................................................................... 25
2.3.1 Polyketide Cyclases ........................................................................................................ 25
v
2.3.1.1 Secondary Polyketide Cyclases ................................................................................. 28
2.3.1.2 A Possible Role for Polyketide Cyclases in Cannabinoid Biosynthesis.................... 29
2.3.2 DABB Proteins ............................................................................................................... 30
2.3.2.1 The Structure and Occurrence of DABB Proteins ..................................................... 31
2.3.2.2 The Function of DABB Proteins ............................................................................... 33
2.3.3 Aldolases ......................................................................................................................... 36
3. THE TKS REACTION ............................................................................................................. 37
3.1 Introduction .......................................................................................................................... 37
3.2 Materials and Methods ......................................................................................................... 39
3.2.1 Tetraketide Synthase Assay Methods ............................................................................. 39
3.2.2 The Expression and Purification of TKS ........................................................................ 39
3.2.3 TKS Assays ..................................................................................................................... 41
3.2.3.1 In the Presence of 14C Malonyl-CoA and Hexanoyl-CoA......................................... 42
3.2.3.2 In the Presence of Malonyl-CoA and Hexanoyl-CoA ............................................... 43
3.2.3.3 In the Presence of Different Buffers and Ph .............................................................. 43
3.2.3.4 In the Presence of N-(2-acetamido)iminodiacetic acid (ADA) Buffer ...................... 43
3.2.3.5 With Variable Temperatures...................................................................................... 44
3.2.3.6 With Variable Substrate Concentrations ................................................................... 44
3.2.3.7 In the Presence of DTT .............................................................................................. 44
3.2.4 Cannabis Trichome Preparation and Trichome Extract Assays ...................................... 44
3.2.4.1 Trichome Isolation ..................................................................................................... 45
3.2.4.2 Cannabis Trichome Extracts ...................................................................................... 45
vi
3.2.4.3 The Effect of Recombinant TKS on Trichome Extract Activity ............................... 46
3.2.4.4 Rabbit Antibodies Directed Against TKS ................................................................. 47
3.2.4.5 Ammonium Sulfate Precipitation of Trichome Proteins ........................................... 47
3.2.4.6 Activity of Ammonium Sulfate Trichome Protein Fractions .................................... 48
3.2.4.7 Co-Immunoprecipitation of TKS ............................................................................... 49
3.2.4.8 Protein Pull-Downs of Trichome Protein Extracts .................................................... 50
3.2.5 Cloning and Assaying of Candidate Cyclases ................................................................ 51
3.2.5.1 Cloning and Expression of Candidate Cyclases ........................................................ 51
3.2.5.2 Assaying TKS in the Presence of Candidate Cyclases .............................................. 53
3.3 Results .................................................................................................................................. 53
3.3.1 Tetraketide Synthase (TKS) ............................................................................................ 53
3.3.2 Expression and Purification of TKS ............................................................................... 56
3.3.3 TKS produces HTAL, PDAL, and olivetol ..................................................................... 57
3.3.3.1 Quantification of TKS products................................................................................. 61
3.3.3.2 Buffers and pH ........................................................................................................... 61
3.3.3.3 Temperature ............................................................................................................... 64
3.3.3.4 Substrate Availabilty ................................................................................................. 65
3.3.3.5 TKS Activity in the Presence of DTT ....................................................................... 68
3.3.4 OA Production by Trichome Protein Extracts ................................................................ 70
3.3.4.1 Ammonium Sulfate Precipitation of Drug-Type Trichome Protein Extracts ............ 73
3.3.4.2 The Co-Immunoprecipitation of Trichome Protein Extracts ..................................... 75
3.3.4.3 Protein Pull-Down Assays of Trichome Protein Extracts ......................................... 76
vii
3.3.5 The Analysis of cDNA Cyclase Candidates ................................................................... 77
3.4 Discussion............................................................................................................................. 81
4. THE CHARACTERIZATION OF OAC .................................................................................. 88
4.1 Introduction .......................................................................................................................... 88
4.2 Materials and Methods ......................................................................................................... 89
4.2.1 Phylogenetic Analysis of OAC ....................................................................................... 89
4.2.2 Biochemical Characterization of OAC ........................................................................... 89
4.2.2.1 TKS/OAC Ratios ....................................................................................................... 89
4.2.2.2 Temperature ............................................................................................................... 90
4.2.2.3 Buffers ....................................................................................................................... 90
4.2.2.4 Concentration of Buffer ............................................................................................. 90
4.2.2.5 pH .............................................................................................................................. 90
4.2.2.6 Metals ........................................................................................................................ 90
4.2.2.7 EDTA ......................................................................................................................... 91
4.2.2.8 Malonyl-CoA Synthetase........................................................................................... 91
4.2.2.9 Alternative Substrate Assays ..................................................................................... 91
4.2.2.9.1 Free CoA .............................................................................................................. 91
4.2.2.9.2 TKS products........................................................................................................ 91
4.2.2.9.3 Butyryl-CoA, Octanoyl-CoA, and Acetyl-CoA ................................................... 91
4.3 Results .................................................................................................................................. 93
4.3.1 Olivetolic Acid Cyclase .................................................................................................. 93
4.3.1.1 Expression and Purification of OAC ......................................................................... 94
viii
4.3.2 The Evolutionary Relationship of OAC .......................................................................... 95
4.3.3 The TKS/OAC Reaction Produces Olivetolic Acid ...................................................... 100
4.3.4 Optimum Ratio of TKS and OAC ................................................................................. 104
4.3.5 OAC Activity as a Function of Temperature ................................................................ 105
4.3.6 OAC Activity in Different Buffer Systems ................................................................... 106
4.3.6.1 OAC Activity in Varying Concentration of HEPES pH 7.0 ................................... 107
4.3.7 OAC Activity as a Function of pH ................................................................................ 108
4.3.7.1 OAC Activity in HEPES pH 6.8 to 8.2 ................................................................... 109
4.3.7.2 OAC Activity in Phosphate Buffer pH 6.0 to 8.0 .................................................... 110
4.3.7.3 in ADA Buffer pH 5.7 to 7.8 ................................................................................... 111
4.3.7.4 in Citrate Buffer pH 5.0 to 6.0 ................................................................................. 112
4.3.8 OAC Activity in the Presence of Metals ....................................................................... 113
4.3.8.1 TKS/OAC Activity in the Presence of EDTA ......................................................... 114
4.3.9 TKS/OAC Activity in the Presence of MCS ................................................................ 115
4.3.10 Substrate Utilization in the TKS/OAC Reaction ........................................................ 116
4.3.10.1 The Effect of Free CoA on TKS/OAC Activity .................................................... 117
4.3.10.2 TKS Products as Substrates for OAC .................................................................... 118
4.3.10.3 With Butyryl-CoA ................................................................................................. 120
4.3.10.4 With Octanoyl-CoA ............................................................................................... 120
4.3.10.5 With Acetyl-CoA ................................................................................................... 121
4.4 Discussion........................................................................................................................... 125
4.4.1 Phylogenetic Relationship of OAC ............................................................................... 125
ix
4.4.2 The Optimum TKS/OAC Ratio .................................................................................... 125
4.4.3 The Optimum Temperature for In Vitro OA Production .............................................. 126
4.4.4 The Optimum Buffer and pH for OAC ......................................................................... 127
4.4.5 OAC Does Not Require a Metal for Activity ............................................................... 128
4.4.6 The use of MCS in assays abolishes the need for malonyl-CoA .................................. 129
4.4.7 Neither HTAL, PDAL, nor olivetol are substrates for OAC ........................................ 130
4.4.8 Polyketide Analogs are Cyclized to their Respective Resorcylic Acid by OAC .......... 131
5. ANALYSIS OF PHYSICAL INTERACTION BETWEEN TKS AND OAC ...................... 134
5.1 Introduction ........................................................................................................................ 134
5.2 Materials and Methods ....................................................................................................... 137
5.2.1 Subcellular Localization of TKS and OAC .................................................................. 137
5.2.2 Yeast Two-Hybrid Methods.......................................................................................... 139
5.2.3 Plasmid Construction .................................................................................................... 140
5.2.4 Yeast Transformants ..................................................................................................... 141
5.2.5 Confirmation of DNA Inserts........................................................................................ 142
5.2.5.1 Determination of 3-AT Concentration used in Histidine Dropout Media ............... 143
5.2.5.2 Characterization of Transformants .......................................................................... 143
5.2.6 TKS Magnetic Pulldown Methods ................................................................................ 144
5.2.6.1 Magnetic Talon Beads ............................................................................................. 146
5.2.6.2 TKS Pulldown Reactions ......................................................................................... 146
5.2.7 Equilibrium Enzyme Assays ......................................................................................... 147
5.3 Results ................................................................................................................................ 149
x
5.3.1 Sub-Cellular Localization of TKS and OAC ................................................................ 149
5.3.2 Yeast Two-Hybrid Approach ........................................................................................ 150
5.3.3 The Expression and Purification of OAC in pJE60174 ................................................ 152
5.3.4 TKS Magnetic Pulldowns ............................................................................................. 155
5.3.5 The TKS/OAC Dialysis Reaction ................................................................................. 159
5.4 Discussion........................................................................................................................... 161
6. THE STRUCTURE AND MECHANISM OF OAC .............................................................. 164
6.1 Introduction ........................................................................................................................ 164
6.2 Materials and Methods ....................................................................................................... 167
6.2.1 Primary and Secondary Structure Analysis................................................................... 167
6.2.2 Cloning, Expression and Activity of OAC Mutants ..................................................... 167
6.2.3 Creation of K4A, K38A, and D45A by SDM ............................................................... 168
6.2.3.1 Expression and Purification of OAC and its Mutants ............................................. 169
6.2.3.2 Enzyme Assays ........................................................................................................ 170
6.2.4 MS/MS Monitoring of the TKS/OAC Reaction ........................................................... 171
6.3 Results ................................................................................................................................ 172
6.3.1 Primary and Secondary Structure Analysis................................................................... 172
6.3.2 The Expression and Activity of OAC Mutants ............................................................. 178
6.3.2.1 Experiment A ........................................................................................................... 180
6.3.2.2 Experiment B ........................................................................................................... 182
6.3.3 MS/MS Monitoring of the TKS and TKS/OAC Reactions .......................................... 185
6.4 Discussion........................................................................................................................... 189
xi
6.4.1 Primary and Secondary Structure.................................................................................. 189
6.4.2 Site-Directed Mutagenesis ............................................................................................ 190
6.4.3 CoA analysis of the TKS reaction ................................................................................. 193
7. GENERAL DISCUSSION ..................................................................................................... 194
7.1 Introduction ........................................................................................................................ 194
7.2 A Mechanistic View on the Synthesis of a Linear Tetraketide .......................................... 195
7.3 The Biosynthesis of Olivetol and OA ................................................................................ 198
7.3.1 The Biosynthesis of Olivetol......................................................................................... 198
7.3.2 The Biosynthesis of OA ................................................................................................ 200
7.4 Cannabinoid Biosynthesis .................................................................................................. 202
7.5 Metabolic Engineering of Cannabinoid Biosynthesis ........................................................ 202
7.6 DABB Cyclases May Play a Large Role in Plants ............................................................. 204
7.7 Conclusion .......................................................................................................................... 205
7.8 Future Directions ................................................................................................................ 205
8. REFERENCES ....................................................................................................................... 207
xii
LIST OF TABLES
Table 3.1 Transcripts detected in a C. sativa glandular trichome cDNA library.......................... 38
Table 3.2 Primers used during cloning of CHI-like, MLP-like, and OAC constructs .................. 52
Table 3.3 Details regarding the creation of CHI-like, MLP-like, and OAC constructs ............... 52
Table 3.4 Ratios of product formed during TKS assays. .............................................................. 67
Table 5.1 Primers used during the creation of TKS:CFP and OAC:YFP ................................... 138
Table 5.2 Details regarding the creation TKS:CFP and OAC:YFP constructs .......................... 138
Table 5.3 Plasmids used to transform MaV203 yeast used in Y2H experiments ....................... 140
Table 5.4 Sequencing primers for pDEST22 and pDEST32 based vectors ............................... 140
Table 5.5 Yeast transformants used in Y2H experiments........................................................... 151
Table 6.1 Primers used for SDM and sequencing of pJR60174 ................................................. 168
Table 6.2 OAC constructs and their yields ................................................................................. 179
Table 6.3 Successful OA biosynthesis by OAC point mutants .................................................. 184
Table 6.4 Area response obtained during the MRM analysis of TKS assays ............................. 187
xiii
LIST OF FIGURES
Figure 2.1 Structures for olivetol, OA, PDAL, HTAL and select cannabinoids ........................... 6
Figure 2.2 A female cannabis plant. .............................................................................................. 7
Figure 2.3 Trichomes are the site of cannabinoid biosynthesis. .................................................. 10
Figure 2.4 Understanding of cannabinoid biosynthesis based on current literature. ................... 13
Figure 2.5 Type III PKS reactions. .............................................................................................. 16
Figure 2.6 Polyketides are composed of acetate units. ................................................................ 18
Figure 2.7 Type III PKS reactions can yield unanticipated pyrones. .......................................... 19
Figure 2.8 A PKS from cannabis is capable of producing PDAL, HTAL, and olivetol.............. 24
Figure 2.9 The F and S modes of polyketide cyclization. ........................................................... 27
Figure 2.10 The structure of S. coelicolor ActVa-Orf6. ............................................................... 33
Figure 2.11 Examples of DABB protein activity.......................................................................... 35
Figure 2.12 Type I and II aldolase reactions................................................................................. 36
Figure 3.1 Alignment of TKS and other cannabis PKS homologs. ............................................. 55
Figure 3.2 Expression and purification of TKS. .......................................................................... 56
Figure 3.3 Reverse phase radio TLC analysis of TKS assay products. ....................................... 58
Figure 3.4 LCMS analysis of standards and TKS products. ........................................................ 59
Figure 3.5 Selected ion monitoring of TKS assay products. ....................................................... 60
Figure 3.6 TKS products in different pH and buffers. ................................................................. 62
Figure 3.7 TKS product formation in ADA buffer. ..................................................................... 63
Figure 3.8 Temperature affects TKS product formation. ............................................................ 64
Figure 3.9 TKS product formation in varying malonyl-CoA concentrations. ............................. 66
xiv
Figure 3.10 TKS product formation in varying hexanoyl-CoA concentrations. .......................... 66
Figure 3.11 TKS in the presence and absence of DTT and βME during purification. ................. 69
Figure 3.12 Assays of cannabis trichome protein extracts ........................................................... 71
Figure 3.13 Drug-type trichome extracts assayed with recombinant TKS. .................................. 72
Figure 3.14 Activity of trichome protein extracts in ammonium sulfate. ..................................... 74
Figure 3.15 Co-immunoprecipitation of TKS and trichome protein extracts ............................... 75
Figure 3.16 Protein pulldown using TKS and cannabis drug-type trichome extracts. ................. 76
Figure 3.17 SDS-PAGE of TKS and candidate cyclases. ............................................................. 78
Figure 3.18 TKS activity in the presence of candidate cyclases................................................... 79
Figure 3.19 TKS assays products when assayed in with putative cyclases. ................................. 80
Figure 4.1 SDS-PAGE of Talon purified TKS and OAC. ........................................................... 95
Figure 4.2 Phylogenetic relationship of plant OAC homologs. ................................................... 97
Figure 4.3 Phylogenetic relationship of OAC plant homologs in Clade A. ................................ 98
Figure 4.4 Phylogenetic relationship of OAC plant homologs in Clade B. ................................. 99
Figure 4.5 LCMS of standards and TKS/OAC assay products. ................................................ 101
Figure 4.6 Selected ion monitoring of TKS/OAC assay products. ............................................ 102
Figure 4.7 A comparison of TKS and TKS/OAC assay products. ............................................ 103
Figure 4.8 TKS assays conducted with variable amounts of OAC............................................ 104
Figure 4.9 The activity of TKS/OAC as a function of temperature........................................... 105
Figure 4.10 TKS/OAC assays conducted in different buffers. ................................................... 106
Figure 4.11 Activity of TKS/OAC in variable concentrations of HEPES. ................................. 107
Figure 4.12 Activity of TKS/OAC in HEPES buffer pH 6.8 to 8.2. .......................................... 109
xv
Figure 4.13 Activity of TKS/OAC in phosphate buffer pH 6.0 to 8.0. ...................................... 110
Figure 4.14 Activity of TKS/OAC in ADA buffer pH 5.7 to 7.8. .............................................. 111
Figure 4.15 Activity of TKS/OAC in citrate buffer pH 5.0 to 6.0. ............................................. 112
Figure 4.16 The effect of metal on the TKS/OAC reaction. ....................................................... 113
Figure 4.17 EDTA affects the synthesis of TKS/OAC products. ............................................... 114
Figure 4.18 TKS/OAC assays conducted in presence and absence of MCS. ............................. 115
Figure 4.19 TKS/OAC product accumulation in the presence of free CoA. .............................. 117
Figure 4.20 LCMS of OAC assays conducted with HTAL. ....................................................... 119
Figure 4.21 LCMS of TKS/OAC assays conducted with butyryl-CoA...................................... 122
Figure 4.22 LCMS of TKS/OAC assays conducted with octanoyl-CoA. .................................. 123
Figure 4.23 LCMS of TKS/OAC assays conducted with acetyl-CoA........................................ 124
Figure 5.1 Flowchart for TKS pulldown assays. ....................................................................... 145
Figure 5.2 The Harvard Apparatus Dialyzer.............................................................................. 148
Figure 5.3 Co-localization of TKS and OAC. ........................................................................... 150
Figure 5.4 Determining 3-AT concentrations used in Y2H assays. .......................................... 151
Figure 5.5 Histidine biosynthesis and β-Galactosidase activity in Y2H screens....................... 152
Figure 5.6 SDS-PAGE of OAC expressed from the pJexpress 411 vector. .............................. 154
Figure 5.7 SDS-PAGE of Thrombin cut OAC. ......................................................................... 154
Figure 5.8 Products accumulating during TKS pulldown assays. ............................................. 156
Figure 5.9 The production of OA by OAC following TKS pulldown. ...................................... 157
Figure 5.10 Monitoring OAC substrate availability following TKS pulldowns. ....................... 159
Figure 5.11 The micro-dialysis of TKS assay products. ............................................................. 160
xvi
Figure 6.1 CoA fragmentation during MS analysis. ................................................................... 166
Figure 6.2 Sequence alignment of OAC and homologs. ........................................................... 173
Figure 6.3 Prediction of OAC Secondary Structure. ................................................................. 174
Figure 6.4 Swiss Model template retrieval of OAC. ................................................................. 175
Figure 6.5 Templates and models for OAC. .............................................................................. 177
Figure 6.6 Space filled model of OAC. ..................................................................................... 177
Figure 6.7 SDS-PAGE of wtOAC and mutants. ........................................................................ 179
Figure 6.8 LC results for OAC mutant assays - Experiment A ................................................. 181
Figure 6.9 LC results for OAC mutant assays - Experiment B. ................................................ 183
Figure 6.10 Location of SDM target sites mapped on the OAC model. .................................... 184
Figure 6.11 Parent ion scan of TKS and TKS/OAC reactions. .................................................. 186
Figure 6.12 Multiple Reaction Monitoring of TKS and TKS/OAC reactions. ........................... 187
Figure 6.13 MRM for the neutral loss of CoA during a TKS reaction. ...................................... 188
Figure 6.14 The proposed active site for OAC. .......................................................................... 192
Figure 6.15 The structure of SnoaL and TcmI. ........................................................................... 192
Figure 7.1 A proposed mechanism for TKS. ............................................................................. 197
Figure 7.2 A proposed reaction mechanism for the synthesis of olivetol. ................................. 199
Figure 7.3 A proposed reaction mechanism for the biosynthesis of OA. .................................. 201
Figure 7.4 A proposed biosynthesis of cannabinoids. ............................................................... 203
xvii
LIST OF ABBREVIATIONS
3-AT
3-amino-1,2,4-triazole
6-DDHK
6-deoxydihydrokalafungin
AA
amino acid
Ab
Antibody
ABM
antibiotic biosynthetic monooxygenase
ACP
acyl carrier protein
AD
GAL4 Activating Domain
ADA
N-(2-Acetamido)iminodiacetic acid
Amp
Ampicillin
AmpR
Ampicillin resistance marker
AtHS1
Arabidopsis thaliana heat stable protein
BDAL
butyryl diacetic acid lactone
BP
before present
BRI
Biotechnology Research Institute
BTAL
butyryl triacetic acid lactone
B-TKS
Talon-bead affixed recombinant TKS
C4H
cinnamate 4-hydroxylase
CBCA
cannabichromenic acid
CBD
cannabidiol
CBDA
cannabidiolic acid
CBG
cannabigerol
xviii
CBGA
cannabigerolic acid
CBN
cannabinol
CDD
Conserved Domain Database
CHI
chalcone isomerase
CHR
chalcone reductase
CHS
chalcone synthase
CNBR
cyanogen bromide activated Sepharose 4B
CoA
Coenzyme A
D4R
dihydroflavonol 4-reductase
DABB
dimeric α + β barrel
DBD
GAL4 DNA binding domain
DHK
dihydrokalafungin
DMAPP
dimethylallyldiphosphate
DTT
dithiothreitol
EDTA
ethylenediaminetetraacetic acid
Gen
gentamycin
GenR
gentamycin resistance marker
GOT
geranylpyrophosphate:olivetolate transferase
GPP
geranyl diphosphate
HDAL
heptyl diacetic acid lactone
HEPES
2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid
HPLC
high performance liquid chromatography
xix
HTAL
hexanoyl triacetic acid lactone
IMAC
immobilized metal chromatography
IPP
isopentenyl diphosphate
JTT
Jones-Taylor-Thornton
Kan
Kanamycin
KanR
Kanamycin resistance marker
LCMS
liquid chromatography mass spectrometry
MCS
malonyl-CoA synthethase
MLMI
methylmuconolactone methyl isomerase
MLP
Betv1-like (major latex) protein
MOPSO
3-(N-morpholinyl)-2-hydroxypropanesulfonic acid
MRM
multiple reaction monitoring
MW
molecular weight
MWCO
molecular weight cut-off
NNI
Nearest-Neighbor-Interchange
NRC
National Research Council Canada
OA
olivetolic acid
OAC
olivetolic acid cyclase
OAC:YFP
OAC-yellow fluorescent protein
OLS
olivetol synthase
OMT-1
O-methyltransferase
OTAL
octanoyl triacetic acid lactone
xx
PAL
phenylalanine ammonia-lyase
PBI
Plant Biotechnology Institute
PDA
photo array detector
PDAL
pentyl diacetic acid lactone
PDB
protein databank
PKC
polyketide cyclase
PKS
polyketide synthase
PtSP1
Populus tremula stable protein
PVPP
polyvinylpolypyrrolidone
Rmsd
root mean square deviation
RT
room temperature
RT-PCR
reverse transcriptase-polymerase chain reaction
SCOP
Structural Classification of Proteins database
SDM
site-directed mutagenesis
Spc
spectinomycin
SpcR
spectinomycin resistance marker
SRABB
stress responsive alpha beta barrel protein
STCS
stilbene carboxylate synthase
S-TKS
soluble recombinant TKS
STS
stilbene synthase
TAL
triacetic acid lactone
THC
∆9-tetrahydrocannabinol
xxi
THCA
tetrahydrocannabinolic acid
TKS
tetraketide synthase
TKS:CFP
TKS-cyan fluorescent protein
TLC
thin layer chromatography
TRIS
2-amino-2-hydroxymethyl-propane-1,3-diol
TTAL
tetraacetic acid lactone
Uniprot
Universal Protein Resource database
VPS
valerophenone synthase
wtOAC
wild-type OAC
X-gal
5-bromo-4-chloro-3-indolyl- beta-D-galactopyranoside
Y2H
yeast two-hybrid
βME
β-mercaptoethanol
xxii
1. INTRODUCTION
Cannabis sativa is one of the most recognizable plants in our society having been harvested
since antiquity as food and fibre, and due to its unique pharmaceutical properties. Due to its
medicinal and psychotropic effects, the plant has attracted substantial attention, and is the source
of controversy and close scrutiny by the international community. Yet, despite its importance,
the legal status of cannabis has impeded its research, and the biosynthesis of the compounds
known as cannabinoids, which are exclusively produced in cannabis, has remained unclear. It is
the purpose of this dissertation to provide insight into a key step in the cannabinoid biosynthetic
pathway, and to provide evidence that a type III polyketide synthase (PKS) and a dimeric α + β
barrel (DABB) polyketide cyclase are both responsible and required for olivetolic acid (OA)
biosynthesis. Tetraketide synthase (TKS) is a type III PKS initially identified from a cannabis
trichome-specific cDNA library. Although it was anticipated that TKS was responsible for the
biosynthesis of OA, a precursor involved in cannabinoid biosynthesis, the enzyme cannot
independently produce OA, and instead yields a decarboxylated analog. This lack of OA-forming
activity prompted an investigation into TKS, which eventually led to the discovery of a DABB
protein called olivetolic acid cyclase (OAC).
Formally, this dissertation describes the initial characterization of TKS and the steps that
resulted in the identification of OAC (Chapter 3), followed by a report on the in vitro
optimization of the OA biosynthetic reaction (Chapter 4). Subcellular localization experiments
and yeast two-hybrid assays were employed to determine if TKS and OAC physically interact,
and progressed to the use of protein pulldowns and micro-dialysis experiments to further probe
1
the nature of the TKS/OAC reaction (Chapter 5). Furthermore, an attempt to determine which
amino acids are involved in the OAC mechanism was conducted by monitoring and comparing
the activity of 15 single point OAC mutants alongside the wild-type OAC (Chapter 6).
Observations made within the course of this work support the hypothesis that OA biosynthesis
requires 1) TKS to generate a labile tetraketide intermediate and 2) OAC for the intermediate’s
proper cyclization. This led to the MS/MS screening of TKS/OAC reactions for tetraketo-CoA
linked analytes, including a theoretical short-lived polyketide intermediate (Chapter 6).
1.1 Research Goals and Objectives
The initial goal of this research was to determine why OA was not being produced from TKS.
This required the characterization of TKS, the testing of alternative assay conditions to induce
TKS to produce OA, and to identify putative cyclases identified in a cannabis trichome cDNA
library (Stout et al. 2012). Once the polyketide cyclase OAC was identified, it was characterized
and an attempt to delineate its underlying mechanism was conducted. This involved determining
if TKS and OAC co-localized, physically interacted, and could produce OA when physically
separated. Structure/function studies were also conducted in an attempt to discern amino acids
involved in OAC’s mechanism. It is also a goal of this research to implicate TKS and OAC as
two principle enzymes required for the biosynthesis of OA, which are therefore components of
the cannabinoid biosynthetic pathway. The information resulting from this thesis supports the
hypothesis in favor of the existence of a transient linear tetraketide.
2
2. LITERATURE REVIEW
2.1 Overview of Cannabis sativa and Cannabinoids
Cannabis has been a multipurpose resource for thousands of years. Humans and other hominids
have ingested its seeds since prehistoric times (Callaway, 2004). Archaeological evidence
indicates that it was cultivated for its use as a textile in Neolithic China (i.e. 6500 before present;
BP) (Fleming and Clarke, 1998). It is unknown when the psychoactive properties of cannabis
were first discovered, although physical evidence suggests they were known as early as 2700 BP
(Jiang et al., 2006; Russo et al., 2008). The earliest record of medicinal cannabis appeared in the
first century in the pen-ts'ao ching, where it was recognized as a treatment for numerous ailments
including rheumatism, constipation and malaria (Touwn, 1981). Cannabis was also used in
western medicine during the early 20th century, before it was outlawed and subsequently
removed from the British and American pharmacopoeias in 1932 and 1944, respectively. The
history of cannabis and its uses has been well reviewed by several authors (Abel, 1980; Fleming
and Clarke, 1998; Russo et al., 2008; Zuardi, 2006).
The psychoactive properties of cannabis have long been the cause for scientific
examination of the responsible biological compounds. The bioactive fraction from cannabis,
which is now known to consist of many cannabinoids, was initially distilled from hemp resin and
described as a red oil in 1896; the acetate crystals obtained from this oil were named cannabinol
(CBN) and at the time were believed to be the psychoactive agent in cannabis (Wood et al.,
1896). It was not until 35 years later that similar experiments were successfully reproduced, and
that a partial structural elucidation of CBN was published (Cahn, 1932). In 1940, a second
3
cannabinoid named cannabidiol (CBD) was also isolated from red oil, but both CBN and CBD
were found to be pharmacologically inert and lacked ‘marihuana activity’ (Adams et al., 1940a;
Loewe, 1945). The psychoactive cannabinoid ∆9-tetrahydrocannabinol (THC) was later
discovered in 1940 as a product of in vitro CBD isomerisation, and was not isolated from
cannabis, albeit in an impure form, until the following year (Adams et al., 1940b; Wollner et al.,
1942). The isolation of pure THC from cannabis was eventually achieved in 1964, and its
structure was determined in 1971 (Gaoni and Mechoulam, 1964; Gaoni and Mechoulam, 1971)
(Figure 2.1).
Since the pioneering days of cannabis research, more than 100 cannabinoids exclusively
produced in cannabis have been identified (Mechoulam and Hanus, 2000; Russo, 2011).
Cannabinoids are a unique class of chemicals that are composed of terpene and phenolic moieties
(i.e. terpenophenolics, or prenylated polyketides). They do not contain any nitrogen, although
cannabis alkaloids have also been reported (Mechoulam, 1988). The structural elucidation and
synthesis of THC allowed for the identification of their endogenous targets, the CB1 and CB2
receptors (Devane et al., 1988; Munro et al., 1993). In turn, the discovery of the cannabinoid
receptors led to the identification of their endogenous ligands, arachidonoylethanolamide
(anandamide) and 2-arachidonoylglycerol, and have since been recognized as components of the
endocannabinoid system (Devane et al., 1992; Mechoulam et al., 1995) (Figure 2.1).
Cannabinoids that are produced in planta are sometimes referred to as phytocannabinoids to
distinguish them from endogenously produced CB1 and CB2 ligands, which are collectively
known as endocannabinoids (Russo, 2011). Phytocannabinoids are produced as acids in planta
4
but are decarboxylated to their neutral form upon heating and storage (Kimura and Okamoto,
1970; Turner et al., 1980; Yamauchi et al., 1967).
2.1.1 The Biology of Cannabis
Cannabis is a dioecious annual herb that produces either male staminate or female pistillate
flowers, although monoecious varieties are also known to exist (Figure 2.2). Staminate flowers
develop five stamens with near sessile anthers that are surrounded by a calyx of five sepals. The
male inflorescences are indeterminate racemes that form into a panicle arrangement of flowers
that tend to bloom slightly earlier than their female counterparts. The pollen is released during
the bloom and is normally dependent on wind to fertilize nearby female plants. Soon after
staminate plants shed their pollen, they enter senescence and die. Female plants bear pistillate
flowers which initially appear late during the long-day vegetative cycle as two elongate stigmas
at branching nodes. During the short-day flowering cycle, the nodal bracts blossom and
accumulate more stigmas, each surrounded by short lobed calyces that protect the developing
fruits, known as achenes, or more commonly as seeds (Stearn, 1970). Although the existence of
sexual dimorphism in cannabis is evident, the determining mechanism is not completely
understood nor is it completely agreed upon. Some models for sexual determination have implied
x/autosomal dosage mechanisms, although a growing body of evidence, including RAPD and
cDNA-AFLP analysis suggests that a heterogametic X/Y system exists (Ainsworth, 2000;
Mandolino et al., 2002; Moliterni et al., 2004).
Prior to 1998, cannabis belonged to the Urticales order, but since then this order has been
embedded into the Rosales order (Sytsma et al., 2002). Cannabis is now circumscribed as an
urticalean rosid (Rosales) and belongs to the Cannabaceae family (Soltis et al., 2011). The
5
Cannabaceae family is now recognized as sensu lato to include the Celtis genus (hackberry), or
as sensu stricto as to only include the Humulus genus (hops). Analysis of cpDNA restriction sites
support the relationship between Cannabaceae and Celtidaceae families (Song et al., 2001).
Figure 2.1 Structures for olivetol, OA, PDAL, HTAL and select cannabinoids
Resorcylates and pyrones are the main focus of this study; they are in vitro products from
enzymes involved in the biosynthesis of phytocannabinoids. Phytocannabinoids are produced in
plants, whereas endocannabinoids occur in animals.
6
Figure 2.2 A female cannabis plant.
Cannabis trichomes used in this study were obtained from plants such as this Cannabis sativa
cultivar Finola. This plant has reached maturity evident by the developed inflorescence.
Some confusion also exists as to whether Cannabis is a monotypic or polytypic taxon;
some authors believe only one taxon exists, Cannabis sativa L. (Small and Beckstead, 1973),
whereas other authors claim that three distinct species exist: C. sativa, Cannabis indica, and
Cannabis ruderalis (Schultes et al., 1974). This work will consider cannabis as an inclusive
7
species called Cannabis sativa and will not discriminate between sub-species and/or cultivars.
Perhaps more importantly is the classification of the different chemical phenotypes. Varieties of
cannabis grown as industrial hemp for the production of fibre and seeds generally contain lower
amounts of tetrahydrocannabinolic acid (THCA), which is the precursor to the psychoactive
THC, and higher amounts of the non-psychoactive compound cannabidiolic acid (CBDA). Drugtype cannabis, which is grown both for recreational and medicinal use, contains significantly
more THCA than CBDA (Fetterman et al., 1971; Small and Beckstead, 1973; Toffoli et al.,
1968).
Cannabinoids are primarily biosynthesized in specialized structures known as glandular
trichomes (Kim and Mahlberg, 1997; Potter, 2009). Trichomes are differentiated epidermal
outgrowths which are specialized in form and function in many plants (Schilmiller et al., 2008).
Four types of trichomes are found on cannabis plants; non-glandular hair-type trichomes, and
three secretory-type trichomes: bulbous, capitate-sessile, and capitate-stalked glandular
trichomes (Dayanandan and Kaufman, 1976; Hammond and Mahlberg, 1973) (Figure 2.3). Nonglandular trichomes are non-secretory, unicellular, and are prominently cystolith hairs containing
high concentrations of silicon. Amongst the glandular trichomes, the capitate-stalked trichomes
(150 to 500 µm) are the largest accumulators of cannabinoids, developing during the mature
flowering stage. They are similar to capitate-sessile trichomes, but include a stalk in addition to a
secretory disc and a subcuticular cavity. The secretory disc, which is a convex rosette of 8 – 13
cells, is the site of synthesis for fatty acids, terpenes, and cannabinoids which accumulate in the
subcuticular cavity, giving the capitate gland a globular appearance (Dayanandan and Kaufman,
1976). Several studies have implicated disc cells in glandular trichomes to be responsible for the
8
production and secretion of secondary metabolites (Wagner, 1991). Sirikantaramas et al. (2005)
found that the presence of mRNA for the enzyme THCA synthase was exclusively detected by
RT-PCR in cannabis trichome disc cells, and furthermore, when heterologously expressed in
tobacco, a THCA synthase-green fluorescent protein localized to the plant’s trichomes (Figure
2.3).
The glandular trichome is understood as being the site of unique gene expression related
to the production of secondary metabolites. Some evidence was provided by Keene and Wagner
who purified glands from tobacco that were capable of efficiently producing diterpenes and
sucrose esters, although surrounding tissues (epidermal and subepidermal cells) could not (Keene
and Wagner, 1985). The genomic analysis of glandular trichomes, which typically involves the
construction of a cDNA library derived from isolated trichomes, has been successful in
elucidating specialized metabolic pathways and identifying enzymes specific to the trichome
(Fridman and Pichersky, 2005). The specialized peltate glands from sweet basil were
successfully analyzed this way, leading to the identification of a methyltransferase responsible
for the biosynthesis of methylchavicol (Gang et al., 2001; Iijima et al., 2004). Sesquiterpene
biosynthesis in wild tomato Solanum habrochaites was also determined by evaluating trichomespecific EST data, leading to the identification of two terpene synthases (Sallaud et al., 2009).
The lupulin glands from Humulus lupulus were also separated and used to create a large set of
ESTs which were then used to identify an O-methyltransferase (OMT-1) required in the
conversion of desmethylxanthohumol to xanthohumol (Nagel et al., 2008). Trichome-specific
cDNA libraries have also been used for gene discovery in cannabis, resulting in the detection of
9
a type III PKS capable of synthesizing olivetol and an acyl-activating enzyme responsible for the
generation of hexanoyl-CoA (Marks et al., 2009; Stout et al., 2012).
Figure 2.3 Trichomes are the site of cannabinoid biosynthesis.
The cannabinoids of Cannabis sativa are synthesized and stored in glandular trichomes found on
the female flower. The inset shows the localization (green) of an enzyme of the cannabinoid
biosynthetic pathway, tetrahydrocannabinolic acid synthase, using a green fluorescent proteinlabeled probe to the head of a glandular trichome. Reproduced with permission from Jirschitzka
et al. (2013).
10
2.1.2 Bioactivity of Cannabinoids
One of the earliest evaluations of cannabis in western medicine was conducted by William
O’Shaughnessy in 1838, wherein he describes the anaesthetic, orexigenic, antiemetic, and
anticonvulsant properties of the drug (O'Shaughnessy, 1843). Cannabinoids are now recognized
as therapeutic agents for a number of debilitating ailments including chemotherapy-associated
nausea, epilepsy, glaucoma, Parkinson’s disease, dystonia, multiple sclerosis, and pain (Amar,
2006). The antineoplastic properties of certain cannabinoids, namely THC, have been reported as
early as 1975, and although still requiring conclusive study, cannabinoids show great potential as
antiproliferative, tumor-selective apoptotic agents (Brown et al., 2012; Munson et al., 1975).
Generally, the mechanism of cannabinoids are related to their binding to the endogenous CB1
and CB2 receptors, but has also been shown to involve the ankyrin, melastatin, and vanilloid
transient potential receptors (Pertwee et al., 2010). Cannabinoids acting on the classical
cannabinoid receptors, which are G-protein coupled receptors, inhibit adenylcyclase and result in
the modulation of the MAPK/ERK pathway.
2.1.3 The Current Understanding of Cannabinoid Biosynthesis
In the last 80 years several groups have proposed different cannabinoid biosynthesis pathways.
The earliest proposal was offered by Simonsen and Todd (1942), after Adams (1940) showed
that CBD could be converted to THC and CBN in vitro; they proposed that a cannabidiol-like
compound formed by the condensation of menthatriene (2-methyl-5-prop-1-en-2-ylcyclohexa1,3-diene) and olivetol cyclized to THC and later to CBN(Adams, 1940; Simonsen and Todd,
1942). Following the Birch (1957) proposal for phenol biosynthesis, Farmilo and Davis (1961)
suggested that hexanoic acid condenses with three molecules of acetic acid yielding a
11
hexenedionic acid that could enolize to OA. Farmilo and Davis (1961) were the first to depict
cannabinoids in their natural acidic form. These earlier accounts for cannabinoid biosynthesis
were presented without the knowledge of cannabigerolic acid (CBGA) and were based on the
belief that THC production was a result of CBD isomerization. Gaoni and Mechoulam (1964)
realized that cannabigerol (CBG) was a missing intermediate in the cannabinoid biosynthetic
pathway and suggested that two alternative pathways existed; one pathway each dedicated to the
biosynthesis of either the acid or neutral cannabinoids, but both these pathways were believed to
require the condensation of geranyl diphosphate (GPP) with olivetol (or OA in the case of acidic
forms) (Gaoni and Mechoulam, 1964). Shoyama et al. (1970) later proposed a model for the
cannabinoid biosynthesis pathway accounting for alternate forms of cannabinoids, namely
cannabichromenic acid (CBCA) and CBGA. Shoyama et al. (1975) later published data from
experiments with
14
C-labelled malonate/mevalonate and 3H-labelled geraniol, indicating that
CBGA is formed by the condensation of geraniol and a polyketide derived from malonate. These
findings have provided the basis for some of the current understanding of the cannabinoid
biosynthetic pathway.
Cannabinoids are terpenophenolic compounds that originate from the prenylation of the
polyketide-derived OA with GPP, converting it to the first cannabinoid in the cannabinoid
biosynthesis pathway, CBGA (reviewed by (Taura et al., 2007) (Figure 2.4).
A
geranylpyrophosphate:olivetolate transferase (GOT) was proposed to prenylate OA after
observing that CBGA is produced by cannabis cell extracts when provided with GPP and OA
(Fellermeier and Zenk, 1998). Importantly, this study concluded that OA served as the phenolic
substrate for GOT and not its decarboxylated derivative, olivetol. The Zenk group provided
12
Figure 2.4 Understanding of cannabinoid biosynthesis based on current literature.
The intermediate CBGA is produced by the prenylation of OA by GOT. Whereas GPP is derived
from the MEP pathway, OA has been proposed to be synthesized by a PKS capable of accepting
hexanoyl-CoA esters once they are activated by an acyl-CoA synthetase. CBGA is converted to
CBDA, THCA, CBCA by CBDA synthase, THCA synthase, and CBCA synthase, respectively.
13
evidence that the terpenoid substrates responsible for the generation of GPP, isopentenyl
diphosphate (IPP) and dimethylallyldiphosphate (DMAPP), are derived from the deoxyxylulose
phosphate pathway (Fellermeier et al., 2001). Their work strongly suggested that the phenolic
moiety is a polyketide-derived compound.
The prenylation reaction resulting in the biosynthesis of CBGA is followed by the
conversion into CBDA or THCA catalyzed by the enzymes, CBDA synthase and THCA
synthase, respectively (Taura et al., 1996; Taura et al., 1995). It is important to note that the
neutral form of CBGA, CBG is not accepted by THCA synthase or CBDA synthase as a
substrate. CBGA also acts as the substrate for other cannabinoid synthases such as CBCA
synthase (Morimoto et al., 1997; Morimoto et al., 1998) (Figure 2.4). These cannabinoids, their
analogs, and a variety of others can be further classified into ten distinct classes of cannabinoids
based on their structural backbone (Sanchez and Verpoorte, 2008).
2.2 Olivetolic Acid Biosynthesis
The concept that OA originates from a polyketide is based on the biosynthetic speculations made
by both Collie and Birch, who independently proposed that phenolic natural products are derived
from β-polyketones (Hill and Staunton, 2010). In this Section, an overview of type III PKSs and
the early accounts of experiments suggesting the involvement of a PKS during OA biosynthesis
is presented. Past attempts to detect and isolate a PKS capable of OA biosynthesis are reviewed.
2.2.1 Type III PKSs: Diversity, Structure and Mechanism
PKSs are enzymes that are distributed in the bacterial, fungal, animal and plant kingdoms. They
are responsible for the generation of acetate-based β-polyketo compounds known as polyketides
and are involved in the biosynthesis of many secondary metabolites. Many polyketides are
14
bioactive and play an important role in medicine, where they are used as antibiotics,
immunosuppressants, statins, and cytostatics (Shen, 2003; Staunton and Weissman, 2001;
Weissman, 2009). PKSs conduct the repeated condensation of malonyl esters with an appropriate
starter molecule, resulting in an elongated polyketide, a process reminiscent of fatty acid
biosynthesis. Not surprisingly, these enzymes are thought to have evolved from fatty acid
synthases, which in addition to their similar activity share a highly conserved structural fold
(Austin and Noel, 2003; Jenke-Kodama et al., 2005). PKSs are categorized as type I, II, or III:
type I PKSs are large modular enzymes found in fungi, bacteria, and animals (Castoe et al.,
2007; Keatinge-Clay, 2012); type II PKSs are part of multi-enzyme aggregates found in bacteria
(Zhan, 2009); type III PKSs are small homodimeric proteins found in plants, bacteria, and more
recently in fungi (Austin and Noel, 2003; Muggia and Grube, 2010).
Type III PKSs constitute their own family of enzymes that are believed to have evolved
from fatty acid synthases, although they are structurally different than type I and II PKSs. The
type III PKS family, also known as the chalcone/stilbene synthase family, is ubiquitous in higher
plants and is involved in the biosynthesis of chalcones and stilbenes, which are themselves
precursors for more complex secondary metabolites, including flavonoids and anthocyanins
(Austin and Noel, 2003). Whereas type I and II PKSs require an acyl carrier protein (ACP) to
activate malonyl extender units, type III PKSs utilize malonyl-CoA esters. Also, type III PKSs
have greater diversity in their use of starter substrates compared to type I and type II PKSs, as is
the case in the chalcone-forming reaction (Figure 2.5). The chalcone synthase (CHS) reaction
performed by type III PKSs iteratively adds acetate units to a p-coumaroyl-CoA starter by the
15
decarboxylative condensation of malonyl-CoA. The resultant tetraketide intermediate is
transformed into naringenin chalcone by an intramolecular Claisen condensation.
Figure 2.5 Type III PKS reactions.
Reactions include (A) the biosynthesis of chalcones such as chalcone naringenin by CHSs, and
(B) the biosynthesis of stilbenes such as resveratrol by STSs. Both CHSs and STSs synthesize
the same tetraketide intermediate, although (C) the cyclization of the reactive intermediate
proceeds in either (blue arrow) a Claisen condensation or (red arrow) an aldol reaction. The R
substituent indicates attachement to an enzyme’s active site.
Type III PKSs do not solely synthesize naringenin chalcone, neither do they all involve a
terminating Claisen condensation. Stilbene synthases (STS) are also type III PKSs which
catalyze the synthesis of stilbenes, including the well known compound resveratrol (Figure 2.5)
(Hathway and Seakin, 1959; Schröder et al., 1988; Takaoka, 1939). The initial steps of PKS
reaction are identical in CHS and STS, where both form a tetraketide intermediate, although a
different cyclization reaction takes place in the final steps of product formation. The CHS
reaction depends on a C6 C1 Claisen condensation, whereas STSs catalyze a C2 C7 aldol
reaction followed by a C1 decarboxylation (Austin and Noel, 2003). The cyclization appears to
16
be controlled by electronic effects of specific AAs in respective enzymes. This was shown by
Austin et al. (2004) after converting the CHS from Medicago sativa into a functional STS, which
they engineered using Pinus sylvestris STS as a model. A third alternative reaction also
conducted by these enzymes is a C5O C1 cyclization resulting in the generation of pyrone
products (Figure 2.6). Pyrone products such as CTAL are commonly referred to as derailment
products, and are regarded as incomplete reaction products. Austin et al. (2004) suggested that
pyrone products may slowly and spontaneously cyclize when tetraketide intermediates are
released by the enzyme, or that an unidentified enzyme may catalyze the final cyclization of
released intermediates (Austin et al., 2004). It is important to note that some plants, such as
Gerbera hybrida and Piper methysticum, do accumulate pyrone products (Eckermann et al.,
1998; Seigler, 1998).
Stilbene carboxylate synthases (STCS) are a hypothetical subset of type III PKSs which
possess a similar activity to STSs, but differ in that they retain the C1 carboxyl moiety following
a C2 C7 cyclization (Figure 2.6) Although there is less support for the existence of STCSs,
they are assumed to biosynthesize phenolic acid metabolites in planta. One STCS, HmSTCS1a,
has been reported to occur in Hydrangea macrophylla, an ornamental that produces the rare
hydrangeic and lunularic acids (Figure 2.7). HmSTCS1a was isolated from Hydrangea cDNA
and shares 65-70% sequence identity with other chalcone and stilbene synthases; however, the
enzyme contains unusual residues in positions conserved in other CHS-like proteins. Although
Eckermann et al. (2003) cloned HmSTCS1a, the hypothesized final product, 5-hydroxyhydrangeic acid was never detected during assays, and instead the derailment product CTAL
accumulated (Figure 2.7). The formation of the CTAL pyrone was regarded as an artefact
17
Figure 2.6 Polyketides are composed of acetate units.
Polyketide synthases use an array of different starter molecules (R) and iteratively couple them
to acetate residues by the decarboxylative condensation of malonyl-CoA. The cyclizations of
longer polyketide intermediates result in the formation of pyrones, whereas the cyclization of
tetraketides can result in pyrones, chalcones, stilbenes, and stilbene carboxylates. Figure adapted
from Schröder (2009).
18
Figure 2.7 Type III PKS reactions can yield unanticipated pyrones.
Certain recombinant CHS-like PKSs putatively involved in the biosynthesis of natural products
do not always yield the anticipated products when assayed in vitro. Examples include (A) the
biosynthesis of 5-hydroxy-hydrangeic acid by H. macrophylla HmSTCS1, (B) the biosynthesis
of chrysophanol anthrone by A. arborescens OKS, and (C) the biosynthesis of emodin anthrone
by H. perforatum HpPKS2. Instead these reactions synthesized pyrone products.
19
of sub-optimal in vitro conditions that did not allow for the appropriate cyclization to occur.
Eckermann et al. (2003) also described the in vitro biosynthesis of 5-hydroxylunularic acid, but
only after a different starter substrate was used (i.e. dihydro-4-coumaroyl-CoA instead of 4coumaroyl-CoA) and after the permethylation of assay products was performed.
Other examples of CHS-like enzymes incapable of cyclizing polyketide intermediates to
the anticipated final products have also been reported in other plants. Aloe arborescens
octaketide synthase is a type III PKS thought to be involved in the biosynthesis of chrysophanol
anthrone, but which instead accumulates the pyrones SEK4 and SEK4b when the recombinant
enzyme is assayed (Figure 2.7) (Abe et al., 2005). Similarly, HpPKS1 and HpPKS2 are two
CHS-like enzymes putatively involved in the biosynthesis of the biologically active hyperforin
and hypericin in Hypericum perforatum L. (St. John’s Wort) (Karppinen and Hohtola, 2008).
Although the production of emodin anthrone was anticipated, the in vitro analysis of the
HpPKS2 enzyme produced tri- to octaketide products, as well as the octaketide pyrones SEK4
and SEK4b, and the heptaketide aloesone, yet no emodin anthrone was detected (Figure 2.7).
Whether these enzymes are involved in the respective reactions, or another putative enzyme or
co-factor is required for the proper cyclization of the polyketide chain is not known.
20
2.2.2 Instances of Olivetolic Acid Biosynthesis
The occurrence of OA is rare in nature, but has been reported to accumulate in the fungal, plant,
and animal kingdoms. OA is found in Primula obconica, where it is used as a substrate for the
natural products primin and miconidin (Horper and Marner, 1995). In animals, the release of OA
from metapleural glands in some Crematogaster ants has been reported (Jones et al., 2005). OA
is also recognized as a component of depsides in numerous lichens (Armaleo et al., 2011; Elix,
1974; Elix, 1994). Additionally, OA analogs that incorporate longer alkyl chains than the pentyl
chain in OA, also occur in nature, and could presumably be the result of similar cyclization
reactions since the carboxylate moiety is retained in the final products. These analogs are found
in numerous fungi and plants including Evernia prunastri (oakmoss), Merulius incarnates
(mushroom), Grevillea robusta (silver oak), and several cereals and grains (Baerson et al., 2010;
Hassani and Razzoul, 2005; Jin and Zjawiony, 2006; Yamashita et al., 2008; Yamashita et al.,
2010; Yao, 2011). More recently the first example of a type III PKS capable of forming
alkylresorcylic acids was reported in Oryza sativa (rice) (Matsuzawa et al., 2010).
2.2.3 The Biosynthesis of Olivetolic Acid Requires a Polyketide Synthase
Most of the data suggesting that OA is synthesized via an acetate polyketide pathway was
provided by heavy isotope labeling experiments conducted in cannabis. The earliest account for
PKS involvement was provided by Shoyama et al. (1975), after
14
C-malonate was incorporated
into cannabinoids during isotope feeding experiments (Shoyama et al., 1975). The involvement
of a PKS was also implied after Fellermeier et al. (2001) used
13
C-glucose to discern the
positions of the isotopes incorporated into cannabinoids. Similarly, after conducting 3H-acetate
labeling studies in Primula obconica, Horper et al. (1996) suggested that a PKS was involved in
21
the biosynthesis of OA, as well as predicting that the starter moiety was hexanoic acid. Results
from these groups indicate that an acetate-derived polyketide corresponds well to the NMR
labelling patterns they obtained. By the year 2000, the search for an OA-producing enzyme had
begun, with the focus that a type III PKS was responsible for the generation of the necessary
substrate.
2.2.4 Type III PKSs from Cannabis
One of the earliest attempts to detect polyketide synthase activity in cannabis was conducted by
assaying the crude extracts from various parts of the plant (Raharjo et al., 2004a). The extracts
were incubated with hexanoyl-CoA and malonyl-CoA esters, and although the authors were able
to detect olivetol, they found no OA. These experiments provided the basis for further rigorous
examination into the enzyme(s) responsible for OA biosynthesis. Raharjo et al. (2004c)
continued their search using a proteomics approach that compared the differential expression of
proteins and mRNA within tissues capable and incapable of cannabinoid biosynthesis. The
authors were unable to determine any of the cannabinoid biosynthetic enzymes during their
proteomic analysis of leaves, flowers, and trichomes. Furthermore, Raharjo et al. (2004c) also
attempted to identify cannabis PKSs using a Pinus sylvestris chalcone synthase antibody, and
were partially successful in that they detected a PKS similar to the H. lupulus chalcone and
phloroisovalerophenone synthases (Matousek et al., 2005; Paniego et al., 1999). Determined to
isolate the recently detected PKS, Raharjo et al. (2004b) conducted degenerate reverse
transcriptase-polymerase chain reaction (RT-PCR) using primers modeled after the H. lupulus
CHS and phloroisovalerophenone synthase, and did successfully clone the first cannabis PKS.
This enzyme (deposited in Genbank AAL92879) showed chalcone and phloroisovalerophenone
22
synthase activity, but was incapable of olivetol and OA biosynthesis, and was assigned as simply
a cannabis CHS.
In 2009, a new cannabis PKS named olivetol synthase (OLS; annotated BAG14339) was
isolated from cannabis leaves using degenerate primers, and reported to biosynthesize olivetol
from hexanoyl-CoA and malonyl-CoA (Taguchi et al., 2008; Taura et al., 2009). OLS also
synthesizes two other products, namely a triketide pyrone called pentyl diacetic acid lactone
(PDAL) and a tetraketide pyrone named hexanoyl triacetic acid lactone (HTAL) (Figure 2.8).
Almost simultaneously with the Taura et al. (2009) publication, Marks et al. (2009) also
reported the identification of a type III PKS possessing olivetol synthase activity. This discovery
was made after a trichome-specific cDNA library was generated in an attempt to identify
cannabinoid biosynthetic genes (Marks et al., 2009). Marks et al. (2009) reported a cDNA library
of 1075 unigenes, of which three were putative type III PKSs named CAN24, CAN383, and
CAN1069. Whereas CAN1069 was identical to the cannabis CHS previously reported by
Raharjo et al. (2004b), CAN24 and CAN383 were reported as novel PKSs due to their divergent
sequence. The authors characterized CAN24 and speculated that it produced pyrones and
olivetol, since their results were inconclusive due to unsatisfactory HPLC analysis. It is now
clear that CAN24 is identical to OLS based on sequence comparisons. Despite the successful
identification of the novel type III PKS in cannabis, the characterization of OLS/CAN24
indicated that OA was not synthesized, and instead only its decarboxylated analog, olivetol,
could be produced.
Another noteworthy analysis of cannabis PKSs was reported by Flores-Sanchez et al.
(2010) after isolating a variety of PKS cDNAs from cannabis glandular trichomes by RT-PCR.
23
The authors described five different PKSs, PKSG1 (annotated EU551163), PKSG2 (annotated
EU551164), PKSF3 (annotated EU551162), PKSG4 (annotated EU551165), and PKSG5
(annotated EU551166), that are 97% homologous over their 385 amino acid span. The largest
difference between these PKSs appears in the size of their untranslated regions, which FloresSanchez et al. (2010) attributed to alternative transcription initiation and polyadenylation sites.
None of the PKSs identified were identical to OLS/CAN24, although they all shared greater than
97% sequence similarity at the amino acid level with it. This small difference in sequence is
believed to be due to the varieties of cannabis used in each study, although these individual
isozymes have not been fully characterized. Flores-Sanchez et al. (2010) noted that the
biochemical characterization of these PKSs should be conducted to better understand their
function and diversity.
Figure 2.8 A PKS from cannabis is capable of producing PDAL, HTAL, and olivetol.
Although this type III PKS can produce olivetol, it cannot produce OA, the suspected
intermediate required for further cannabinoid biosynthesis.
24
2.3 Polyketide Cyclization and DABB Proteins
2.3.1 Polyketide Cyclases
Generally, in all types of PKS systems, the biosynthesis of aromatic polyketides is a
process that not only depends on the iterative condensation of acetate units, but also on the
proper cyclization and tailoring of intermediates. A polyketide intermediate may be cyclized into
several different products, for which only one specific enzyme may be capable of achieving.
Cyclization reactions, with limited exception, differ substantially between fungal, bacterial, and
plant systems, and have each been shown to employ different strategies to achieve the diverse
product biosynthesis observed in nature. The biosynthesis of fungal polyketides is accomplished
by megasynthases that incorporate a folding pattern (F mode) completely different than the
folding pattern used by the actinobacterial species (S mode), which use dissociable multi-enzyme
complexes during polyketide biosynthesis (Thomas, 2001). The S and F modes proposed by
Thomas (2001) involve the cyclization of the initial ring in polyaromatic polyketides, and have
been observed during many 13C-acetate labeling experiments allowing the comparison of fungal
and streptomycete polyketide biosynthesis, such as viridicatumtoxin biosynthesis in Penicillium
viridicatum with oxytetracycline biosynthesis in Streptomyces rimosus (de Jesus et al., 1982;
Thomas and Williams, 1983) (Figure 2.9). Based on these observations, Thomas (2001) designed
a general set of guidelines for S and F modes, which indicated that F-mode cyclization results in
the synthesis of an initial cyclohexane ring containing two intact 13C-acetate moieties attached to
two adjacent odd-numbered chains. Conversely, S-mode cyclization results in a cyclohexane
containing three intact
13
C-acetate units with two even-numbered adjacent chains. Furthermore,
the cyclization of polyketides is not simply a matter of S or F designation, but is also dependent
25
on several influencing factors including the oxidative state of the polyketide (especially around
C9), the priming-compound used, the size of the substrate binding cavity or product template
domain, the fusion of other polyketide intermediates, and the presence of other polyketide
cyclases (Zhou et al., 2010). In their conceptual review, Zhou et al. (2010) have further
categorized type II PKS cyclization based on the initial ring-folding pattern commonly occurring
at the C7-C12 position, which generates the tetracycline, aureolic acid, anthracycline, and
benzoisochromanequinone backbones, or at the C9-C14 position, which gives rise to the
tetracenomycin scaffold. Zhou et al. (2010) suggest that the folding pattern mostly governed by
the chain length factor, which does not only determine the size of a nascent polyketide, but also
coordinates the polyketide after it buckles back, due to the limited space in the substrate-binding
cavity. Additionally, these authors indicate that bacterial cyclases directly influence the initial
ring cyclization pattern, such as seen with ZhuI and TcmN aromatases, which are able to
coordinate C7-C12 and C9-C14 cyclizations, respectively. Although ZhuI and TcmN share
structural similarity, a comparison of their structures indicates that ZhuI contains a smaller
substrate-binding pocket which forces the nascent polyketide to bend at C7 instead of C9 (Ames
et al., 2011). Ketoreductases, such as Streptomyces coelicolor ActIII, are involved in the
reduction of polyketides at the C9 position, and have also been shown to influence C7-C12
cyclization in systems incapable of such cyclizations, such as in Gibberella fujikuroi PKS4
(Hadfield et al., 2004; Ma et al., 2007).
In plants, the folding of polyaromatic polyketides is accomplished by the simultaneous
use of both S and F modes (Thomas, 2001). Aside from the CHS/STS folding pattern previously
discussed (Section 2.2.4), Thomas (2001) refers to 2-methylanthraquinone chrysophanol and its
26
Figure 2.9 The F and S modes of polyketide cyclization.
(A) Fungal and streptomycete folding patterns differ during initial ring cyclization, where in F
mode a C6-C11 bond forms, and in S mode a C7-C12 bond is formed. (B) The incorporation of
13
C-acetate during the biosynthesis of viridicatumtoxin and oxytetracycline is one example that
illustrates F and S modes, respectively. (C) It has been suggested that the biosynthesis of
chrysophanol and aloesponarin in Aloe occurs via a bimodal folding process. Chrysophanol can
also be biosynthesized by P. islandicum, whereas aloesaponarin can be produced in
streptomycetes. Figure adapted from Thomas (2001).
27
isomer, 1-methylanthraquinone aloesaponarin II, to exemplify the bimodal process in plants
(Figure 2.9). Whereas both compounds are produced in Aloe species, chrysophanol is also
produced in Penicillium islandicum through a decarboxylative F-mode cyclization, and
aloesaponarin I can be biosynthesized in recombinant streptomycetes via an S-mode folding
pattern. The occurrences of both these compounds in Aloe suggest an F/S bimodal cyclization
process is responsible.
2.3.1.1 Secondary Polyketide Cyclases
Following primary ring cyclization, type II PKS systems use dissociable cyclases to fold the
remaining rings during polyketide biosynthesis. These cyclases may act independently or may
require other cyclases to synergistically achieve the correct folding pattern resulting in either
linear or angular fused polyketides. In their review, Zhou et al. (2010) have divided these
polyketides by their primary ring cyclization, but also by whether or not they form linear or
angular fused rings, as well as by their secondary, tertiary, and in some cases quaternary
cyclizations which result in several families of compounds that share a similar scaffold. For
example, tetracyclines and aureolic acids are classified together as linear fused aromatics that are
the result of C7–C12, C5–C14, C3–C16, and C1–C18 regiospecific cyclizations, whereas
tetracenomycins are also linearly fused aromatics, but form their own class of compounds since
they are generated following C9–C14, C7–C16, C5–C18, and C2–C19 cyclizations. The
involvement of regioselective cyclases is required regardless of which class of fusion the
polyketides belong to. In the case of tetracycline biosynthesis, it has been demonstrated that, in
S. coelicolor, the cyclization of the four rings in the naphthacene backbone are achieved by only
two cyclases; OxyK is responsible for the initial C7-C12 cyclization and OxyN is responsible for
28
the second ring C5-C14 cyclization (Petkovic et al., 1999; Zhang et al., 2007). The third ring is
believed to result from spontaneous cyclization, but no enzyme has been identified as being
responsible for the fourth ring cyclization. Similarly, the biosynthesis of tetracenomycin C
requires a primary ring, and in this case secondary ring, cyclase, but also requires TcmI to
complete the C2-C19 cyclization of the fourth ring. The mechanism for the third ring cyclization
is still not well understood (Zhou et al. 2010).
2.3.1.2 A Possible Role for Polyketide Cyclases in Cannabinoid Biosynthesis
In an attempt to better understand the cyclization to OA, a general review of similar mechanisms
incorporating the cyclization of polyketide intermediates was conducted. This overview of
cyclization mechanisms capable of retaining a carboxylate moiety in the final or intermediate
products led to the identification of similar phenomena occurring in Actinobacteria.
Actinobacteria are a rich source of diverse secondary metabolites from which hundreds of
antibiotics have been isolated. Particularly, the biosynthesis of tetracenomycin in Streptomyces
glaucescens played an important role in the identification of enzymes putatively involved during
the cyclization of OA. The biosynthesis of tetracenomycin is a multistep reaction that recruits a
type II PKS responsible for the generation of a nonaketide intermediate (via TcmJKLM), that is
initially cyclized to TcmF2 (via TcmN), and is cyclized once more, in a step that retains the
carboxyl moiety by a type II polyketide cyclase known as TcmI (Thompson et al., 2004). The
similarity of this reaction led to the exploration of similar proteins in cannabis, which eventually
led to the successful progress described in this thesis. TcmI is classified as a polyketide cyclase
(PKC), a family of proteins that harbor a specific dimeric α + β barrel (DABB) domain.
29
2.3.2 DABB Proteins
The DABB domain is a common structural motif found in many proteins distributed throughout
nature that form a superfamily of structurally related proteins with some very common features.
Some of these proteins play critical roles during important processes involving drug
biosynthesis, environmental detoxification, and pathogenicity. Although DABB proteins share
striking structural similarity, they are not necessarily derived from a common evolutionary origin
and may not share strong sequence similarity. The dissimilarity in protein sequence is often seen
at the family level where members sharing less than 20% sequence identity may have structures
that superimpose to a root mean square deviation (rmsd) of ≤ 2.0Å. Proteins are typically
categorized based on their similarities in sequence, function, and structure; traits that are
classified differently by the available bioinformatic tools which lead to varying degrees of
clustering by accessible databases (Martinez, 2011). DABB proteins are not excluded from this
ambiguity and are represented differently depending on the database used. For example, the
Structural Classification of Proteins database (SCOP) recognizes DABB proteins as a
superfamily based on their three-dimensional domain, and partitions the superfamily into 23 subfamilies (annotated 54909), whereas the PFAM database recognizes the signature motif for the
DABB clan (annotated CL0032) which is further separated into 15 distinct families (Bateman et
al., 2004; Murzin et al., 1995). As well, the two popular web-based resources, Interpro and the
Conserved Domain Database (CDD), which similarly integrate information from several
databases, distinguish the DABB superfamily very differently; Interpro recognizes DABB
proteins as the parent of several sub-groups/families (annotated IPR011008), one of which the
CDD narrowly defines as the DABB superfamily (Marchler-Bauer et al., 2011; Mulder et al.,
30
2005). In this study, the superfamily of DABB proteins is recognized as a larger superfamily
represented by both SCOP and PFAM, although it focuses more on some functionally similar
families, namely the antibiotic biosynthetic monooxygenases (ABMs), the PKCs, and the stressresponsive α + β barrel proteins (SRABBs).
2.3.2.1 The Structure and Occurrence of DABB Proteins
There are many protein structures available for DABB proteins that display common
fundamental features. DABBs are α + β - classed structures displaying a repeated ferredoxin-like
βαβ fold (Figure 2.10). These structurally analogous domains are characterized by independent
segments of α-helices pressed against a 4-5 strand β-sheet, sometimes referred to as α-β
sandwiches due to the distinct layering of the α-helices and β-sheet (Orengo and Thornton,
1993). DABB proteins typically dimerize at the β-sheet interface and form an anti-parallel βbarrel structure (Figure 2.10). The α-helices found in the repeated βαβ elements are sometimes
segmented or kinked and form 1-2 arches which protrude away from the β-sheet, shaping a
crevice between the secondary structures. This crevice is purported as the active site in several
DABB members (Goblirsch et al., 2010; Sciara et al., 2003; Sugano et al., 2007). Although many
DABB proteins display higher-ordered quaternary structure, they typically form dimers which
may or may not continue to oligomerize. The formation of dimers is partly due to hydrophobic
interactions between the residues of two subunit β-sheets that are not necessarily conserved yet
result in similar β-barrel structures. Sciara et al. (2003) pointed out two highly conserved
residues, Tyr63 and His52, in Streptomyces coelicolor ActVa-Orf6 homologs, which facilitate
dimer stabilization through tyrosine-tyrosine aromatic stacking and tyrosine-histidine hydrogen
bonds and demonstrated that replacement of His52 by site-directed mutagenesis (SDM) led to a
31
severe loss of activity due to the disruption of the dimer interface. Additionally, in ActVa-Orf6
each monomeric C-terminus was shown to contribute to dimer formation by swapping strands
with the opposing β-sheet, also contributing to the bottom portion of the active site. But these
features are not universal in DABB proteins. The Staphylococcus aureus IsdG, which shares
only 15% identity at the amino acid level with ActVa-Orf6, yet can be superimposed to a rmsd of
1.92 Å, does not use the same residues as ActVa-Orf6 to achieve dimerization, and relies on
different hydrophobic interactions to achieve its similar structure (Wu et al., 2005). Essentially,
the involvement of hydrophobic residues is necessary for dimer interaction normally seen at the
β-sheet interface. As well, strand swapping is not always observed in DABB members as it is in
ActVa-Orf6. In Escherichia coli YgiN, there is an extended loop between the β2 and β3 strands
which helps to form the opposing monomer’s β-sheet instead of the strand swapping observed in
ActVa-Orf6, yet structural similarity between ActVa-Orf6 and YgiN is maintained to an overall
rmsd of 1.9 Å (Adams and Jia, 2005).
32
Figure 2.10 The structure of S. coelicolor ActVa-Orf6.
(A) ActVa-Orf6 is pictured as a dimer. The (blue) monomer carboxyl terminus is seen strand
swapping with the opposing (green) monomer β-sheet strengthening the dimeric interaction and
completing the β-barrel structure. The (red) substrate (Nanaomycin D) is shown in the putative
active site found between the α-helices and β-pleats. (B) ActVa-Orf6 is shown at a different
angle allowing the β-barrel structure to be seen. Figure adapted from Sciara et al. (2003).
2.3.2.2 The Function of DABB Proteins
DABB proteins are involved in catalyzing a diverse set of reactions including the oxygenation of
polyaromatic compounds, isomerization of methyl muconolactone, and the cyclization of
polyketides (Figure 2.11). The functionality of the different DABB proteins has been
instrumental in the assignment of subfamilies within the superfamily of proteins. As the name
implies, the ABM family is involved in the oxygenation of cyclic polyketide-derived antibiotic
compounds. The S. coelicolor ActVa-Orf6 serves as the prototypical monooxygenase
representative of the ABM family from which the ABM family name was initially derived.
ActVa-Orf6
catalyzes
the
oxidation
of
6-deoxydihydrokalafungin
(6-DDHK)
to
dihydrokalafungin (DHK), an intermediate in the actinorhodin biosynthetic pathway (Kendrew et
33
al., 1997) (Figure 2.11). Other members of this family include Streptomyces galilaeus AknX and
Streptomyces nogalater SnoaB involved in aclarubicin and nogalamycin biosynthesis
respectively (Raty et al., 2002; Ylihonko et al., 1996).
Methylmuconolactone methyl isomerases (MLMIs) are involved in the degradation of
methyl and chloro-substituted aromatics following ortho-(intradiol) cleavage. The cleavage of 4methylcatechols form 4-methylmuconolactones which cannot be further processed by the 3oxoadipate pathway unless they are initially isomerized to 3-methylmuconolactone by MLMIs
(Figure 2.11). These enzymes have been described in Cupriavidus necator JMP132 (Pieper et al.,
1990), Rhodococcus rhodochronous N75 (Bruce et al., 1989), and Pseudomonas reinekei MT1
(Camara et al., 2007). In addition to 4-methylmuconolactone isomerization, MLMIs have been
shown to synthesize 3-methylmuconolactone using 1-methylbislactone as substrate indicative of
MLMI’s ability to open and close lactone rings (Pieper et al., 1990).
The PKCs form a separate family of DABB enzymes that are found exclusively in
Actinobacteria. This family is exemplified by TcmI, the only functionally and structurally
characterized PKC from this family to date. TcmI is part of the tetracenomycin biosynthesis gene
cluster found in S. glaucescens and was shown to cyclize Tcm F2 to Tcm F1, both intermediates
involved in the biosynthesis of the antitumor antibiotic Tcm C, without requiring a cofactor or
metal ion (Figure 2.11) (Shen and Hutchinson, 1993a). It is important to note that this cyclization
occurs while retaining the carboxyl moiety in the intermediates. Although the functionalities of
these DABB proteins are quite diverse, the ability of the enzymes to bind and modify larger
cyclic compounds appears universally conserved.
34
Figure 2.11 Examples of DABB protein activity.
(A) The oxidation of 6-DDHK to DHK is catalyzed by ActVa-Orf6. (B) TcmF2 is cyclized to
TcmF1 by the TcmI cyclase. (C) Cyclization reactions which retain the carboxyl moietiy
throughout are conducted by TcmI and are reminiscent of those coordinated by SnoaL enzymes
which are sometimes fused to DABB motifs,. (D) MLMIs convert 4-methylcatechol to 3methylmuconolactone.
35
2.3.3 Aldolases
Aldolases are enzymes capable of making or breaking an aldol bond, which are catoregized as
being either class I or class II. Whereas class I aldolases, occurring primarily in animals and
higher plants, form Schiff bases with their substrates via a lysine residue in the enzyme active
site, the class II aldolases, which occur in bacteria and fungi, use a Zn2+ cofactor that acts as a
Lewis acid during the reaction. The aldolases are important enzymes to consider in the context of
this study, and are later discussed with regard to their distinct mechanisms; see Gijsen (1996) and
Fessner (1996). Aldolase reactions are notable since they result in a donor retaining its carboxyl
moiety, a phenomena which also occurs during the biosynthesis of OA.
Figure 2.12 Type I and II Aldolase Reactions
In the case of type I aldolases, the donor compound is attached to the enzyme by a Schiff base
prior to the aldol reaction, whereas type II aldolase reactions require a metal to coordinate the
donor and the acceptor compounds. The substituent (R) shown in the acceptor compound may be
a variety of monosaccaryl or oligosaccharyl chains.
36
3. THE TKS REACTION
3.1 Introduction
This study was preceded by experiments on trichome-specific cDNA library containing over
4,000 unigenes generated from high-THCA cannabis (Stout et al., 2012). This library was
created by sequencing cDNA synthesized from glandular trichome mRNA and was used as a tool
for the identification of enzymes involved in the biosynthesis of cannabinoids, one of which
Stout et al. (2012) described as a hexanoyl-CoA synthetase. Displayed in Table 3.1 are the most
highly expressed transcripts identified in the cDNA library along with their descriptors
determined by BLAST searches against the Universal Protein Resource database (Uniprot). The
most highly expressed transcript in this cDNA library was CL1Contig3, a sequence with 74%
protein identity with Humulus lupus CHS (annotated BAB47196). The involvement of a type III
PKS in the biosynthesis of OA had been previously suggested (Fellermeier et al., 2001; Horper
and Marner, 1996; Shoyama et al., 1975), and soon after the identification of the type III PKS
candidate by the Page laboratory (unpublished), several cannabis PKSs were reported in the
literature (Flores-Sanchez et al., 2010; Marks et al., 2009; Taura et al., 2009). These reports
included a similar trichome-derived cDNA library (Marks et al., 2009) and the characterization
of olivetol synthase (OLS), which shares 99% protein identity to our PKS candidate (Taguchi et
al., 2008; Taura et al., 2009). Some of the work conducted on the PKS candidate presented in
this thesis occurred prior to any of these reports on type III PKSs were publushed, and at the time
this work was started, the PKS candidate had already been successfully subcloned,
heterologously expressed, and purified by Ryan Taschuk, a member of the Page laboratory.
37
Table 3.1 Transcripts detected in a C. sativa glandular trichome cDNA library
Descriptions are from BLASTX comparison against the Uniprot database. Candidate cyclases assayed in this work
are highlighted in yellow and include CHI (Q6V7U9), DABB/OAC (Probable protein POP3) (Q9LUV2), and MLP
(B5KVN9). The most abundant transcript was TKS (B1Q2B6), initially described as OLS/polyketide synthase, is
highlighted in pink. THCA synthase, another protein involved in cannabinoid biosynthesis, is also highlighted in
pink. Adapted from an earlier publication that used the same cDNA library as in this study (Stout et al., 2012).
Description
UniProt # Organism
Contig size % ESTs
Olivetol synthase / polyketide synthase
B1Q2B6
Cannabis sativa
339
3.70
BURP domain-containing protein
B2ZPK5
Solanum tuberosum
192
2.10
4-hydroxy-3-methylbut-2-enyl diphosphate reductase
B9RZD3
Ricinus communis
185
2.02
Malic enzyme
B9RQE8
Ricinus communis
182
1.99
PAR-1c protein
Q43589
Nicotiana tabacum
87
0.95
Limonene synthase
A7IZZ1
Cannabis sativa
65
0.71
Delta12-oleic acid desaturase
Q6RXX0
Euphorbia lagascae
62
0.68
D12 oleate desaturase
Q41305
Solanum commersonii
53
0.58
Putative chalcone isomerase
Q6V7U9
Solanum lycopersicum
51
0.56
Non-specific lipid-transfer protein
Q850K8
Vitis hybrid
49
0.54
Tetrahydrocannabinolic acid synthase
Q8GTB6
Cannabis sativa
48
0.52
Isopentenyl pyrophosphate isomerase
Q6EJD1
Pueraria lobata
48
0.52
Elongation factor 1-alpha
B9SPV9
Ricinus communis
46
0.50
Cystathionine gamma-synthase
B9RYU1
Ricinus communis
44
0.48
Probable protein Pop3
Q9LUV2
Arabidopsis thaliana
41
0.45
Cytochrome P450
B9SG22
Ricinus communis
39
0.43
Alpha-pinene synthase
A7IZZ2
Cannabis sativa
38
0.41
Monoterpene synthase
B6SCF4
Humulus lupulus
35
0.38
4-hydroxy-3-methylbut-2-en-1-yl diphosphate synthase A9ZN14
Hevea brasiliensis
33
0.36
ATP-citrate synthase
B9SHC9
Ricinus communis
31
0.34
Sesquiterpene synthase
B6SCF5
Humulus lupulus
31
0.34
Acetoacetyl-CoA thiolase
Q2L8A7
Picrorhiza kurrooa
28
0.31
Fructose-bisphosphate aldolase
Q9M4M9
Persea americana
28
0.31
Translationally-controlled tumor protein
B9RT61
Ricinus communis
28
0.31
Sesquiterpene synthase
B6SCF5
Humulus lupulus
27
0.29
Isoflavone reductase related protein
O81355
Pyrus communis
26
0.28
Short-chain dehydrogenase
B9RXD8
Ricinus communis
26
0.28
Salutaridinol 7-O-acetyltransferase
B9RML2
Ricinus communis
26
0.28
Lipoxygenase
C4NZX3
Camellia sinensis
25
0.27
Cysteine protease
A9UFX8
Vitis vinifera
25
0.27
Dessication-related protein
Q8LAU8
Arabidopsis thaliana
25
0.27
Anthranilate N-benzoyltransferase protein
B9S035
Ricinus communis
25
0.27
Hevein-like protein
Q944B8
Sambucus nigra
23
0.25
Pathogenesis related protein PR10
B5KVN9
Prunus domestica
23
0.25
Delta12-oleic acid desaturase
Q6RXX0
Euphorbia lagascae
22
0.24
Acetyl-CoA synthetase
B9RE68
Ricinus communis
22
0.24
Pyruvate decarboxylase
B9SWY1 Ricinus communis
21
0.23
38
In this chapter, experiments were designed to characterize the type III PKS, which was
named tetraketide synthase (TKS), and set out to determine if an unknown chemical or biological
factor played an accessory role in the biosynthesis of OA. Another goal of these experiemtns was
to develop analytical techniques capable of detecting products from TKS assays performed in
vitro. Recombinant TKS was assayed in a variety of buffers, pH ranges, and substrate
concentrations to determine the optimal conditions for tetraketide formation. Rabbit anti-TKS
polyclonal antibodies were also
generated from purified TKS
and used in co-
immunoprecipitation and protein pull-down screens in an attempt to discover accessory proteins
putatively involved in OA cyclization. These screening methods required a protein solution
derived from isolated trichomes, which was assayed prior to any co-immunoprecipitation
attempts. Recombinant TKS was also assayed in the presence of candidate proteins that were
potentially involved in the cyclization of linear tetraketides. This chapter focuses on TKS and
trichome assay results that collectively led to the discovery of OAC.
3.2 Materials and Methods
3.2.1 Tetraketide Synthase Assay Methods
All compounds used throughout this work were purchased (Sigma), except for PDAL, HTAL,
and OA which were synthesized by Ms. Nikki Theaker using previously published methods
(Horper and Marner, 1996).
3.2.2 The Expression and Purification of TKS
An E. coli BL21 (DE3) transformant 472-28A harboring pRT31A6 (i.e. TKS in pHIS8GW) was
provided by Dr. Jonathan Page. The culture was streaked onto an LB plate, grown overnight at
37°C, and a single colony used to inoculate 5 mL of LB. Following overnight culture at 37°C, a
39
1 mL aliquot was transferred to start a larger overnight 50 mL culture. Frozen stocks were
prepared by adding 100 µL of culture to 500 µL of 80% glycerol (v/v) and storing them at -80°C.
The 50 mL culture was added to 1 L of AUTOINDUCTION MEDIUM (1.2% tryptone, 2.4%
yeast extract, 0.0125% glucose, 0.05% α-lactose monohydrate, 50 mM glycerol, 50 mM
Na2HPO4. 50 mM KH2PO4, 25 mM (NH4)2SO4, 1 mM MgSO4), and grown overnight at 30°C.
The overnight culture was further induced by adding IPTG to a final concentration of 0.5 mM,
and was grown for an additional 4 h at 30°C. The culture was pelleted at 10,000 g for 20 min at
4°C, and then transferred to a 50 mL tube before it was stored overnight at -80°C. All the media
described contained 50 µg/mL kanamycin as a selective agent. Cultures were shaken at 200 rpm.
Frozen pellets were thawed on ice for 30 min in the presence of 50 mL LYSIS BUFFER (50 mM
TRIS-HCl [pH 7], 500 mM NaCl, 2.5 mM imidazole, 10% glycerol (v/v), 10 mM βmercaptoethanol, 750 µg/mL lysozyme (Sigma)). The solution was homogenized using a
vortexer, then sonicated on ice for 2 x 2 min bursts using a LabsonicU sonicator (B. Braun
Biotech International). The cell suspension was centrifuged for 20 min at 12,000 g at 4°C and the
supernatant was transferred to a 50 mL tube. In preparation of immobilized metal affinity
chromatography (IMAC), a 1 mL suspension of Talon resin (Clontech) equilibrated in WASH
BUFFER (50 mM TRIS-HCl [pH 7], 500 mM NaCl, 10 mM imidazole, 10% glycerol (v/v), 10
mM β-mercaptoethanol) was added to the cell suspension and then tumbled at 4°C for 40 min.
Talon resin was pelleted by centrifugation for 1 min at 1,000 g, and the supernatant was
discarded. Talon resin was washed five times using 10 mL of wash buffer each wash, where each
wash was removed by centrifugation for 1 min at 1,000 g. The washed Talon was transferred to a
10 mL gravity flow column and rinsed with 10 mL wash buffer. Purified TKS was eluted using
40
10 mL ELUTION BUFFER (50 mM TRIS-HCl [pH 7], 150 mM NaCl, 100 mM imidazole, 10%
glycerol (v/v), 10 mM β-mercaptoethanol), and concentrated to 1 mL using a YM10 centrifugal
filter device (Millipore). The protein solution was desalted and the buffer exchanged by
suspending it in 5 mL of STORAGE BUFFER (20 mM HEPES [pH 7.5] containing 25 mM
NaCl, 10% glycerol (v/v), and 5 mM dithiothreitol (DTT)) and centrifuging the filtration device
until the solution was concentrated to 1 mL. This dilution process was repeated two more times.
Purification work was conducted at 4°C. Enzyme preparations were stored at -20°C.
3.2.3 TKS Assays
Unless stated otherwise, 50 µL reactions were conducted using 25 µg of TKS, 0.2 mM hexanoylCoA, 0.6 mM malonyl-CoA, and 5 mM DTT. Reactions were normally extracted twice using
250 µL ethyl acetate. Extracts were dried under vacuum and dissolved in 30 µL
RESUSPENSION SOLUTION consisting of (v/v) 70% water, 30% acetonitrile, and 0.05%
formic acid. Product analysis was generally conducted by liquid chromatography coupled to
mass spectrometry (LCMS) and was normally performed using 10 µL injections of sample in a
Waters Alliance 2695 separation module (Waters) attached to a 2996 photodiode array (PDA)
detector (Waters), a 3100 single quad-mass detector (Waters), and a Symmetry C18 3.5 µm (2.1
x 100 mm) column (Waters). Products were monitored over 13 min isocratic runs using LCMS
RUNNING SOLVENT consisting of (v/v) 63% water, 37% acetonitrile, and 0.05% formic acid.
Flow rate was maintained at 0.2 mL/min with a 30°C column temperature. Results were
processed using the Empower2 software suite (Waters). Assay products (HTAL, PDAL, OA, and
olivetol) were detected by both LC-PDA and MS methods. Products detected by LC-PDA at 270
nm. MS was set to ES+ mode where three selective ion monitoring (SIR) channels were used to
41
detect ions of m/z 181 (cone voltage = 25 V, dwell = 0.010 s), m/z 183 (cone voltage = 40 V,
dwell = 0.010s), and m/z 225 (cone voltage = 25 V, dwell = 0.010 s). Ions were selected for the
specific detection of olivetol (180 g/mol), PDAL (182 g/mol), and both HTAL and olivetolic
acid (224 g/mol).
Some analyses were conducted solely by high performance liquid chromatography
(HPLC) that used a Waters Alliance 2695 separation module (Waters) and 2996 photo array
detector (PDA) (Waters) and a SunFire C18 10 µm (4.6 mm x 150 mm) (Waters). During HPLC
analysis, samples were resuspended in 30 µL methanol, where 10 µL injections were monitored
over 15 min gradient runs starting with (v/v) 59.9% water, 40% acetonitrile, and 0.1%
trifluoroacetic acid and ending with 29.9% water, 60% acetonitrile and 0.1% trifluoroacetic acid.
Flow rate was maintained at 1 mL/min with a 30°C column temperature. Reaction products were
monitored at 270 nm, and results were processed using the Empower2 software suite. TKS
reaction products were quantified using standard curves generated from 100 μM stock solutions
prepared in LCMS running buffer. Solutions were diluted and 5 μL injections were used to
construct 25 to 500 pmol standard curves following LCMS analysis (Appendix A). All curves
presented a R2 value 0.998 when analyzed by linear regression.
3.2.3.1 In the Presence of 14C Malonyl-CoA and Hexanoyl-CoA
Aliquots of 25 µg TKS were assayed in 50 µL solutions containing 70 mM HEPES [pH 7.0], 20
µM malonyl-CoA, 20 µM hexanoyl-CoA and 1 µL of malonyl(2-14C)-CoA (American
Radiolabeled Chemicals Inc., 0.1 mCi/ml, specific activity 55 mCi/mmol) for 30 min at 30°C.
Reactions were stopped and extracted using 200 µL ethyl acetate. Extracts were dried under
vacuum, then resuspended in 10 µL of ethyl acetate before being spotted on a HPTLC-RP18F
42
plate (Analtech). The plate was developed over 60 min using a solvent consisting of 75%
methanol, 24% water, and 1% acetic acid. The plate was analyzed by radio-TLC scanning using
a LS6500 scintillation counter (Beckman Coulter) and by phosphorimaging using a Storm 840
imaging station (Amersham Biosciences).
3.2.3.2 In the Presence of Malonyl-CoA and Hexanoyl-CoA
Reactions were conducted using 20 mM HEPES [pH 7.0] at 20°C for 60 min. Extracts were
dried under vacuum and then resuspended in 20 µL of resuspension solution. Assays were
performed in triplicate.
3.2.3.3 In the Presence of Different Buffers and ph
Reactions used 20 µg of TKS and were conducted for 30 min at 25°C in 20 mM buffer, 0.2 mM
malonyl-CoA, 0.2 mM hexanoyl-CoA, and 5 mM DTT. Buffers included 20 mM MES [pH 5.5,
6.0, 6.5], phosphate [pH 6.0, 6.5, 7.0, 7.5, 8.0], HEPES [pH 7.0, 7.5, 8.0], and TRIS [pH 7.0, 7.5,
8.0, 8.5, 9.0]. Assays were performed in triplicate and analyzed by HPLC.
3.2.3.4 In the Presence of N-(2-acetamido)iminodiacetic acid (ADA) Buffer
Reactions used 100 pmol (~5 µg) of TKS and were conducted in 20 mM ADA buffers [pH 5.7,
6.0, 6.2, 6.4, 6.6, 6.8, 7.0, 7.2, 7.4, 7.6, 7.8]. Due to the large number of samples, reaction times
were extended to 90 min at 20°C to ensure that there was time enough to start/stop samples.
Extracts were resuspended in 20 µL resuspension solution and used 5 µL injections during
LCMS analysis. Assays were performed in triplicate.
43
3.2.3.5 With Variable Temperatures
TKS temperature assays used 20 µg TKS and were conducted in 50 mM HEPES [pH 7.5], 0.2
mM malonyl-CoA, and 0.2 mM hexanoyl-CoA. Reactions were incubated for 30 min at 10, 20,
30, 40, or 50°C. Reactions were performed in triplicate and were analyzed by HPLC.
3.2.3.6 With Variable Substrate Concentrations
Assays were performed in 100 mM HEPES [pH 7.5] containing 5 mM DTT, and either 0.2 mM
hexanoyl-CoA and 0.1, 0.2, 0.5 or 1.0 mM malonyl-CoA, or 0.2 mM malonyl-CoA and 0.1, 0.2,
0.5, or 1.0 mM hexanoyl-CoA. Reactions were performed at 40°C for 30 min in triplicate and
were analyzed by HPLC.
3.2.3.7 In the Presence of DTT
Two sources of TKS were used in assays, both of which were derived from the same culture.
Following the expression of TKS, 472-28A strains were divided into two portions so to purify
proteins in either the presence or absence of β-mercaptoethanol (βME) and DTT in purification
buffers. TKS purified in the presence or absence of reducing agents was used to assay the effect
of DTT in assays. Reactions contained 100 mM HEPES [pH 7.5], 0.2 mM malonyl-CoA, 0.2
mM hexanoyl-CoA, and either 0, 5, 10, 20, or 100 mM DTT. Reactions proceeded for 90 min at
40°C before being extracted twice with 200 µL of ethyl acetate. Reactions were performed in
triplicate and were analyzed by HPLC.
3.2.4 Cannabis Trichome Preparation and Trichome Extract Assays
Two different trichome extracts were used during assays; one derived from non drug-type
cannabis (USO-31 hemp), and the other from drug-type cannabis. TRICHOME EXTRACTION
44
BUFFER consisted of 100 mM HEPES [pH 7.5], 3 mM EDTA, 500 mM sucrose, 10 mM DTT,
and 1,000x diluted protease inhibitor mixture set III (Calbiochem).
3.2.4.1 Trichome Isolation
Female cannabis plants provided by Dr. Page had been grown in a controlled environment
chamber at 18 h day lengths. Flowering bracts (60 g) were removed from the plants and
dissected before soaking them for 1 h in ice water. Trichome separation was achieved using an
adapted Beadbeater method (Gershenzon et al., 1992). The flowering material was moved to a
950 mL vessel (BeadBeater, BioSpec Products) containing 120 mL of 0.5 mm glass beads, 35 g
XAD-4 and 700 mL of TRICHOME ISOLATION BUFFER (25 mM HEPES pH 7.0, 200 mM
sorbitol, 10 mM sucrose, 0.6% (w/v) methylcellulose and 1% (w/v) polyvinylpyrrolidone
(40,000 g/mol) (PVPP)). The mixture was agitated using a teflon rotor for 1 min before filtering
it through a 350 µ mesh. The tissue was then washed with 250 mL of trichome isolation buffer
which did not contain methyl cellulose or PVPP. All filtrates were combined and run through a
105 m mesh, and then again through a 20 m mesh filter prior to trichome collection. The
trichome cells were rinsed using trichome isolation buffer and dried using a Kimwipe (KimberlyClark). Trichomes were stored at -80°C. Yield was ca. 0.9 g from 60 g of flowers.
3.2.4.2 Cannabis Trichome Extracts
A 100 mg sample of isolated drug-type trichomes was homogenized in the presence of 10 mg
PVPP and 200 µL trichome extraction buffer on ice, followed by a 20 min centrifugation at
15,000 g at 4°C. The supernatant was removed, and trichomes were extracted again using
another 200 µL of extraction buffer. Protein extracts were combined and desalted using a 2 mL
Zeba desalting spin column (Thermo Scientific) equilibrated with 100 mM HEPES [pH 7.5].
45
Protein concentration of the extracts was determined using an RC/DC protein assay as directed
by the manufacturer (Bio-Rad). Protein extracts were assayed using 25 µg (ca. 45 µL) of protein
for 30 min at 40°C in a final volume of 55 µL containing 0.4 mM hexanoyl-CoA and 0.4 mM
malonyl-CoA or simply 0.4 mM hexanoyl-CoA and water (no malonyl-CoA negative control).
Samples were prepared for HPLC analysis as described in Section 3.1.3. Injections of 20 µL
were analyzed by HPLC.
Scaled-up USO-31 hemp trichome assays used 195 µg of protein (ca. 350 µL) from
protein extracts prepared as described above. Extracts were incubated in the presence of 0.25
mM malonyl-CoA and 0.25 mM hexanoyl-CoA for 2 h at 40°C. The 400 µL solution was
extracted twice using 2 x 1 mL ethyl acetate. A single 20 µL injection was performed during
HPLC analysis.
3.2.4.3 The Effect of Recombinant TKS on Trichome Extract Activity
Trichome extracts were prepared as described in Section 3.1.4.2. A total of five reactions were
conducted including 1) recombinant TKS (rTKS), 2) trichome extract, 3) trichome extract in the
presence of rTKS 4) trichome extract lacking hexanoyl-CoA (negative control), and 5) rTKS
lacking hexanoyl-CoA (negative control). The 55 µL TKS reaction was conducted by incubating
50 µg of rTKS in the presence of 80 mM HEPES [pH 7.5], 0.4 mM malonyl-CoA, and 0.4 mM
hexanoyl-CoA. The trichome extracts were assayed by incubating 200 µg of protein from
trichome extract in the presence of 0.5 mM malonyl-CoA and 0.25 mM hexanoyl-CoA. A
negative control for this reaction replaced hexanoyl-CoA with water. The TKS/trichome assay
included 50 µg of TKS with 200 µg protein from trichome extracts. Reactions were incubated for
2 h at 40°C before being extracted using five volumes of ethyl acetate. Samples were prepared
46
for HPLC analysis as previously described (Section 3.1.3) and used 20 µL injections during
analysis.
3.2.4.4 Rabbit Antibodies Directed Against TKS
Antibodies were prepared by the National Research Council Canada, Biotechnology Research
Institute (NRC-BRI) using their standard immunization protocols for the development of
polyclonal antibodies, which required the inoculation of two rabbits each with 350 µg of purified
recombinant TKS. Recombinant TKS was prepared and purified as outlined in Section 3.1.1. See
Appendix B for details on rabbit injections and bleed times.
3.2.4.5 Ammonium Sulfate Precipitation of Trichome Proteins
Trichome extracts were prepared as previously described in Section 3.1.4.2. A saturated
ammonium sulfate solution was prepared by stirring 7.5 g of ammonium sulfate in 10 mL of
water. A 1 mL aliquot of extract was tumbled in the presence of 20% ammonium sulfate for 1 h
at 4°C, followed by centrifugation at 15,000 g for 20 min. The supernatant was removed and the
20% ammonium sulfate pellet was stored at -20°C. The supernatant was then tumbled in 40%
ammonium sulfate for 1 h at 4°C, centrifuged at 15,000 g for 20 min, the supernatant removed,
and the pellet stored at -20°C. In a similar fashion, the remaining supernatant was precipitated
using 60% and 80% ammonium sulfate. The remaining supernatant (i.e. post-80% ammonium
sulfate treatment) was desalted using a 5 mL PD10 column (GE Healthcare Biosciences) and
stored at -20°C. Pellets were later thawed on ice and resuspended in 100 µL of 100 mM HEPES
[pH 7.5]. Since there was some difficulty in reconstituting the 20% pellet in this volume, an
additional 100 µL of buffer was used (total 200 µL). Solutions were concentrated using a YM10
concentrator to ca. 150 µL. An aliquot of 3 µL from each solution was added to 2 µL SAMPLE
47
LOADING BUFFER (50 mM TRIS [6.8], 2% SDS (w/v), 10% glycerol (v/v), 1% βME (v/v),
0.02% bromophenol blue (w/v)) and incubated for 5 min at 95°C before being analyzed by SDSPAGE and Western blotting using 10% gels run at 200 V for 45 min. One gel was stained in BioSafe Coomassie stain for 1 h, and destained overnight in dH2O. The other gel was
electrotransferred to a PVDF membrane using 15 V for 30 min. The transferred PVDF
membrane was blocked with a 10% BSA/phosphate buffered saline (PBS) (Sigma-Aldrich)
solution for 1 h and washed twice in PBS for 10 min, then 2 more times each time for 10 min in
PBS containing 0.1% Tween 20. A 1:1,000 primary antibody solution was prepared by adding 20
µL of rabbit antibodies (see Section 3.1.4.5) to 20 mL of 1% BSA in PBS, which was used to
treat the PVDF membrane on an orbital shaker for 1 h. A 1:20,000 secondary antibody (goat
anti-rabbit conjugated with alkaline phosphatase) was prepared in 1% BSA/PBS and allowed to
react with the membrane for 40 min on an orbital shaker. The membrane was washed as above
and investigated by chemiluminescence using 1 mL of substrate from the Immun-Star goat antirabbit IgG (H + L) Detection Kit (Bio-Rad). A 30 s exposure on Kodak Biomax XAR film
(Sigma-Aldrich) was used to detect anti-TKS antibodies in protein extract solutions.
3.2.4.6 Activity of Ammonium Sulfate Trichome Protein Fractions
Trichome protein extracts were assayed by incubating 50 µL of the resuspended solutions (0%,
20%, 40%, 60%, 80%, and remaining supernatant) with 0.3 mM malonyl-CoA and 0.3 mM
hexanoyl-CoA. A TKS positive control was used by assaying 25 µg of TKS in the presence of
0.2 mM malonyl-CoA and 0.2 mM hexanoyl-CoA. A 50 µL aliquot from the 20% fraction was
assayed in the absence of hexanoyl-CoA and used as a negative control. Reactions were started
48
by adding the appropriate fraction (or TKS), and proceeded for 60 min at 40°C. Reaction
mixtures were prepared for HPLC as described in Section 3.1.3.
3.2.4.7 Co-Immunoprecipitation of TKS
Experiments were conducted with a Co-immunoprecipitation Kit (ThermoScientific) using 4 mL
of rabbit anti-TKS antibodies (see Section 3.1.4.5) purified as directed by the manufacturer.
Purified IgG was desalted using two 5 mL D-Salt Excellulose Desalting Columns equilibrated
with 100 mM PBS [pH 7.3]. Purified antibodies were bound to a Cyanogen Bromide-Activated
Sepharose 4B (CNBR) column. Purified IgG (50 µg) was used during column preparation. Two
other columns were also prepared but did not include the use of antibodies, and were used as 1) a
negative control, 2) a pre-binding column to minimize binding that could result from nonspecific interactions between the resin and the sample. Resin and antibodies were incubated for
90 min at RT on an orbital shaker. The resin columns were washed six times, and rinsed once
with 1x coupling buffer. Columns were washed three more times using 200 µL of wash buffer
before further use. Initially, the columns were tested for their efficiency in binding recombinant
TKS. Two samples of crudely purified TKS (100 µg) were used to determine the specificity of
the prepared columns. The 0.5 µL samples were diluted to 400x (200 µL) before applying them
to either column 1) containing control resin, or 2) pre-bind column. Following centrifugation, the
flow-through was either 1) saved for further analysis in the case of the negative control, or 2)
passed through the column containing IgG-bound resin which was incubated 2 h on a shaker at
RT. All columns were washed four times with wash buffer, followed by three elutions using 20
µL of elution buffer. All fractions were saved for analysis by SDS-PAGE. 20 µL from each
fraction were incubated for 3 min at 98°C in the presence of 5 µL of the provided SDS-PAGE
49
loading buffer. Samples from each column were loaded onto separate 10% acrylamide gels and
subjected to electrophoresis for 30 min at 200 V. Gels were stained using BioSafe coomassie
blue (Bio-Rad) for 1 h, and washed overnight using dH20.
Trichome extracts were prepared as previously described in Section 3.1.4.2. A total of
500 µL of trichome extract was divided into two equal portions and applied to either 1) negative
control IP column, or 2) pre-bind IP column. The pre-bind sample was incubated for 60 min at
4°C upright on a shaker, and then centrifuged at 1,500 g for 90 s. The flow-through was
collected. The pre-bind column was washed four times, then eluted using 3 x 1 mL elution
buffer. Flow-through from the pre-bind column was applied to the antibody-bound resin IP
column and allowed to incubate overnight at 4°C alongside the negative control upright on a
shaker. Columns were later washed four times using 200 µL of IP wash/lysis buffer, followed by
centrifugation of 1,500 g for 90 s after each wash. A 30 µL volume of elution buffer was added
to each column, and after 5 min columns were centrifuged at 1,500 g for 5 min. Two more
elutions using 20 µL elution buffer were performed, before columns were washed and placed in
1x coupling buffer for storage at 4°C. A total of 20 µL from each column fraction (flow-through,
washes, and eluted fractions) were heated with 5 µL of sample loading buffer for 5 min at 98°C.
Samples were analyzed by SDS-PAGE as described in Section 3.1.4.5.
3.2.4.8 Protein Pull-Downs of Trichome Protein Extracts
This experiment used bacterial lysate that had already undergone TKS purification as previously
described (Section 3.1.1.). A second round of TKS purification was performed using the lysate,
which was tumbled with 1 mL Talon beads at 4°C for 40 min. The beads were washed five times
using 10 mL wash buffer and transferred to a 10 mL gravity flow column. A 200 µL aliquot of
50
trichome extract that had been previously prepared (Section 3.1.4.2) was thawed, transferred to
the 10 mL gravity flow column and shaken for 1 h at RT. The gravity flow column was washed
three times using 10 mL wash buffer each time. Proteins were eluted using 10 mL elution buffer.
The solutions were concentrated to a 200 µL volume using a YM10 centrifugal filter device
(Millipore), and desalted by resuspending the solution in 3 mL of 100 mM HEPES [pH 7.5], and
then concentrating to a 200 µL volume. Eluted proteins were compared by SDS-PAGE to a
sample of purified TKS obtained from the first round of purification, and to untreated trichome
extracts. Samples included (separate) 2 µg TKS, 3 µg of eluted proteins, 6 µg of eluted proteins,
1 µg of trichome protein, and 2 µg of trichome protein, which were each heated to 95°C for 5
min in the presence of sample loading buffer prior to loading them onto a 10% acrylamide gel.
Proteins were analyzed by SDS-PAGE as previously described in Section 3.1.4.5.
3.2.5 Cloning and Assaying of Candidate Cyclases
3.2.5.1 Cloning and Expression of Candidate Cyclases
Chi-like, mlp-like, and oac (Table 3.1) were PCR-amplified using the primers described in Table
3.2 and conditions outlined in Table 3.3. Whereas chi and mlp were cloned using Phusion High
Fidelity DNA Polymerase (Finnzymes), each oac construct was amplified using Pfu Ultra II
DNA Polymerase (Stratagene). Amplicons were A-tailed by incubating with Taq polymerase
(Invitrogen) in the presence of 5 mM dATP for 30 min at 72°C and purified using a QIAquick
Spin PCR Purification Kit (Qiagen). Following electrophoresis in 0.8% agarose, single bands
were excised and cleaned using a MinElute PCR Purification Kit (Qiagen) before being either
ligated into the linear pCR8GW entry vector (Invitrogen) or, in the case of OAC, cloned directly
into pET100 (see Table 3.3). Plasmids were used to transform E. coli One Shot Top10
51
Table 3.2 Primers used during cloning of CHI-like, MLP-like, and OAC constructs
Gene
Forward Primer
Reverse Primer
chi-like
ATGGCAACCTCAGATGGAAGC
TTAGAAGAGAGAAAGCATGGA
ATGGGTGTTTTCACTTACG
TTAGTTGTAAGCTTCAGGGTG
CACCATGGCAGTGAAGCATTTGATTGTATTG
CTTTCGTGGTGTGTAGTCAAAAATGAGAAG
mlp-like
oac/DABB
1
2
oac/DABB
ATGGCAGTGAAGCATTTGATT
1 - Original pET100 based OAC construct.
2 - pHIS8GW based OAC construct.
CTACTTTCGTGGTGTGTAGTC
Table 3.3 Details regarding the creation of CHI-like, MLP-like, and OAC constructs
Gene
Template
Melting1
Annealing1
Elongation1
Destination
Selection2
chi-like
CL8contig13
98°C/30s
60°C/10s
78°C/15s
pHIS8GW
KanR
mlp-like
pJS163,4
98°C/30s
55°C/30s
72°C/30s
pDEST17
AmpR
oac/DABB5
CL2contig13
95°C/20s
55°C/20s
72°C/60s
pET100
AmpR
oac/DABB6 pET100-OAC 95°C/20s
50°C/20s
72°C/18s
pHIS8GW
KanR
1 - The reactions were cycled 30-35 times before ending with 7 min at 72°C.
2 - Antibiotic resistance markers include Kanamycin (KanR) and Ampicillin (AmpR).
3 - Originally derived from cDNA
4 - An out-of-frame mlp-like construct originally cloned by Jake Stout from trichome cDNA template.
5 - Original pET100 based OAC construct.
6 - pHIS8GW based OAC construct
Competent Cells (Invitrogen), and transformants were plated onto LB containing 50 µg/mL
spectinomycin (or 100 µg/mL ampicillin for pET100 construct), and grown overnight at 37°C.
Colonies were selected for each construct and grown overnight under selective pressure at 37°C.
Plasmids were isolated and sequenced as previously described (Section 3.1.1). Following
sequence confirmation, genes were transferred from the entry vector into the appropriate
destination vector by LR reaction following the manufacturer’s directions (Invitrogen) (see Table
3.3), and the resulting plasmids transformed into E. coli Rosetta2 BL21(DE3)pLysS Competent
52
Cells (Novagen Biosciences). Transformants were cultured with the appropriate antibiotic,
cultures were induced, and proteins purified as previously described (Section 3.1.1).
3.2.5.2 Assaying TKS in the Presence of Candidate Cyclases
Reactions used 100 pmol of TKS (ca. 5 µg) and either 500 pmol of CHI-like, MLP-like, OAC, or
in the case of negative controls, no protein, and were conducted for 1 h at 30°C. Samples were
prepared and analyzed by LCMS in triplicate as described in Section 3.1.3.
3.3 Results
3.3.1 Tetraketide Synthase (TKS)
TKS contains 385 amino acids (AAs) and is a 43 kDa protein with a 6.04 pI (47 kDa and 6.58 pI
with histidine tag). A BLASTP homology search of the NCBI non-redundant database reveals
TKS to be 100% identical to PKSG4 (annotated ACD76855) (Flores-Sanchez et al., 2010), and
similar to several Cannabis homologs, including OLS (annotated BAG14339; 385 AAs; 99%
identity to TKS in 385 AA overlap). These enzymes belong to a superfamily of condensing
enzymes (cl09938) that contain a conserved CHS-like domain (cd00831) attributed to plant
specific type III PKSs. TKS belongs to a subset of type III PKSs that do not produce chalcone
and are markedly different than other cannabis CHSs previously reported in the literature
(annotated AAL92879; 389 AA; 68.9% identity over a 379 AA subset with TKS) (Raharjo et
al., 2004b), (annotated ACT10338; 392 AA; 66.0% identity in 377 AA overlap with TKS)
(Marks et al., 2009). An alignment of known cannabis type III PKSs highlights differences in
their amino acid sequence (Figure 3.1).
53
TKS
PKSG4
PKSG1
PKSG5
PKSG2
PKSF3
OLS
CAN383
CHS2004
1
1
1
1
1
1
1
1
1
---------MNHLRAEGPASVLAIGTANPENILIQDEFPDYYFRVTKSEHMTQLKEKFRK
---------MNHLRAEGPASVLAIGTANPENILIQDEFPDYYFRVTKSEHMTQLKEKFRK
---------MNHLRAEGPASVLAIGTANPENILIQDEFPDYYFRVTKSEHMTQLKEKFRK
---------MNHLRAEGPASVLAIGTANPENILIQDEFPDYYFRVTKSEHMTQLKEKFRK
---------MNHLRAEGPASVLAIGTANPENILIQDEFPDYYFRVTKSEHMTQLKEKFRK
---------MNHLRAEGPASVLAIGTANPENILIQDEFPDYYFRVTKSEHMTQLKEKFRK
---------MNHLRAEGPASVLAIGTANPENILLQDEFPDYYFRVTKSEHMTQLKEKFRK
MASISVDQIRKAQRANGPATVLAIGTANPPTSFYQADYPDFYFRVTKNQHMTELKDKFKR
--MVTVEEFRKAQRAEGPATIMAIGTATPANCVLQSEYPDYYFRITNSEHKTELKEKFKR
TKS
PKSG4
PKSG1
PKSG5
PKSG2
PKSF3
OLS
CAN383
CHS2004
52
52
52
52
52
52
52
61
59
ICDKSMIRKRNCFLNEEHLKQNPRLVEHEMQTLDARQDMLVVEVPKLGKDACAKAIKEWG
ICDKSMIRKRNCFLNEEHLKQNPRLVEHEMQTLDARQDMLVVEVPKLGKDACAKAIKEWG
ICDKSMIRKRNIFLNEEHLKQNPKLVEHDVQTLDARQDMLVVEVPKLGKDACAKAIKEWG
ICDKSMIRKRNCFLNEEHLKQNPRLVEHEMQTLDARQDMLVVEVPKLGKDACAKAIKEWG
ICDKSMIRKRNCFLNEEHLKQNPRLVEHEMQTLDARQDMLVVEVPKLGKDACAKAIKEWG
ICDKSMIRKRNCFLNEEHLKQNPRLVEHEMQTLDARQDMLVVEVPKLGKDACAKAIKEWG
ICDKSMIRKRNCFLNEEHLKQNPRLVEHEMQTLDARQDMLVVEVPKLGKDACAKAIKEWG
ICEKTTIKKRHLYLTEDRLNQHPNLLEYMAPSLNTRQDMLVVEIPKLGKEAAMKAIKEWG
MCDKSMIRKRYMHLTEEILKENPNLCAYEAPSLDARQDMVVVEVPKLGKEAATKAIKEWG
TKS
PKSG4
PKSG1
PKSG5
PKSG2
PKSF3
OLS
CAN383
CHS2004
112
112
112
112
112
112
112
121
119
QPKSKITHLIFTSASTTDMPGADYHCAKLLGLSPSVKRVMMYQLGCYGGGTVLRIAKDIA
QPKSKITHLIFTSASTTDMPGADYHCAKLLGLSPSVKRVMMYQLGCYGGGTVLRIAKDIA
QPKSKITHLIFTSASTTDMPGADYHCAKLLGLSPSVKRVMMYQLGCYGGGTVLRIAKDIA
QPKSKITHLIFTSASTTDMPGADYHCAKLLGLSPSVKRVMMYQLGCYGGGTVLRIAKDIA
QPKSKITHLIFTSASTTDMPGADYHCAKLLGLSPSVKRVMMYQLGCYGGGTVLRIAKDIA
QPKSKITHLIFTSASTTDMPGADYHCAKLLGLSPSVKRVMMYQLGCYGGGTVLRIAKDIA
QPKSKITHLIFTSASTTDMPGADYHCAKLLGLSPSVKRVMMYQLGCYGGGTVLRIAKDIA
QPKSRITHLIFCSTNGVDMPGADYECAKLLGLSSSVKRVMLYQQGCHAGGSVLRIAKDLA
QPKSKITHLVFCTTSGVDMPGADYQLTKLLGLRPSVKRLMMYQQGCFAGGTVLRLAKDLA
TKS
PKSG4
PKSG1
PKSG5
PKSG2
PKSF3
OLS
CAN383
CHS2004
172
172
172
172
172
172
172
181
179
ENNKGARVLAVCCDIMACLFRGPSDSDLELLVGQAIFGDGAAAVIVGAEPDESVGERPIF
ENNKGARVLAVCCDIMACLFRGPSDSDLELLVGQAIFGDGAAAVIVGAEPDESVGERPIF
ENNKGARVLAVCCDIMACLFRGPSDSDLELLVGQAIFGDGAAAVIVGAEPDESVGERPIF
ENNKGARVLAVCCDIMACLFRGPSDSDLELLVGQAIFGDGAAAVIVGAEPDESVGERPIF
ENNKGARVLAVCCDMTACLFRGPSDSNLELLVGQAIFGDGAAAVIVGAEPDESVGERPIF
ENNKGARVLAVCCDIMACLFRGPSDSDLELLVGQAIFGDGAAAVIVGAEPDESVGERPIF
ENNKGARVLAVCCDIMACLFRGPSESDLELLVGQAIFGDGAAAVIVGAEPDESVGERPIF
ENNKGARILTINSEITIGIFHSPDETYFDGMVGQALFGDGASATIVGADPDKEIGERPVF
ENNKGARVLVVCSEITAVTFRGPNDTHLDSLVGQALFGDGSAALIVGSDPIPEV-EKPIF
TKS
PKSG4
PKSG1
PKSG5
PKSG2
PKSF3
OLS
CAN383
CHS2004
232
232
232
232
232
232
232
241
238
ELVSTGQTILPNSEGTIGGHIREAGLIFDLHKDVPMLISNNIEKCLIEAFTPIGISDWNS
ELVSTGQTILPNSEGTIGGHIREAGLIFDLHKDVPMLISNNIEKCLIEAFTPIGISDWNS
ELVSTGQTILPNSEGTIGGHIREAGLIFDLHKDVPMLISNNIEKCLIEAFTPIGISDWNS
ELVSTGQTILPNSEGTIGGHIREAGLIFDLHKDVPMLISNNIEKCLIEAFTPIGISDWNS
ELVSTGQTFLPNSEGTIGGHIREAGLMFDLHKDVPMLISNNIEKCLIEAFTPIGISDWNS
ELVSTGQTILPNSEGTIGGHIREAGLIFDLHKDVPMLISNNIEKCLIEAFTPIGISDWNS
ELVSTGQTILPNSEGTIGGHIREAGLIFDLHKDVPMLISNNIEKCLIEAFTPIGISDWNS
EMVSAAQEFIPNSDGAVDGHLTEAGLVYHIHKDVPGLISKNIEKSLVEALNPIGISDWNS
ELVSAAQTILPDSDGAIDGHLREVGLTFHLLKDVPGLISKNIEKSLNEAFKPLGISDWNS
54
TKS
PKSG4
PKSG1
PKSG5
PKSG2
PKSF3
OLS
CAN383
CHS2004
292
292
292
292
292
292
292
301
298
IFWITHPGGKAILDKVEEKLDLKKEKFVDSRHVLSEHGNMSSSTVLFVMDELRKRSLEEG
IFWITHPGGKAILDKVEEKLDLKKEKFVDSRHVLSEHGNMSSSTVLFVMDELRKRSLEEG
IFWITHPGGKAILDKVEEKLHLKSDKFVDSRHVLSEHGNMSSSTVLFVMDELRKRSLEEG
IFWITHPGGKAILDKVEEKLHLKKEKFVDSRHVLSEHGNMSSSTVLFVMDELRKRSLEEG
IFWITHPGGKAILDKVEEKLHLKSDKFVDSRHVLSEHGNMSSSTVLFVMDELRKRSLEEG
IFWITHPGGKAILDKVEEKLHLKSDKFVDSRHVLSEHGNMSSSTVLFVMDELRKRSLEEG
IFWITHPGGKAILDKVEEKLHLKSDKFVDSRHVLSEHGNMSSSTVLFVMDELRKRSLEEG
LFWIVHPGGPAILNAVEAKLHLKKEKMADTRHVLSEYGNMSSVSIFFIMDKLRKRSLEEG
LFWIAHPGGPAILDQVESKLALKTEKLRATRHVLSEYGNMSSACVLFILDEMRRKCVEDG
TKS
PKSG4
PKSG1
PKSG5
PKSG2
PKSF3
OLS
CAN383
CHS2004
352
352
352
352
352
352
352
361
358
KSTTGDGFEWGVLFGFGPGLTVERVVVRSVPIKY
KSTTGDGFEWGVLFGFGPGLTVERVVVRSVPIKY
KSTTGDGFEWGVLFGFGPGLTVERVVVRSVPIKY
KSTTGDGFEWGVLFGFGPGLTVETVVLRSVPINY
KSTTGDGFEWGVLFGFGPGLTVERVVLRSVPINY
KSTTGDGFEWGVLFGFGPGLTVERVVVRSVPIKY
KSTTGDGFEWGVLFGFGPGLTVERVVVRSVPIKY
KSTTGDGFEWGVLFGFGPGLTVETIVLHSLAN-LNTTGEGLEWGVLFGFGPGLTVETVVLHSVAI—
Figure 3.1 Alignment of TKS and other cannabis PKS homologs.
TKS was aligned with other reported cannabis PKSs using Clustal Omega (Sievers et al., 2011)
and Boxshade software. The alignment members are PKSG1 (annotated ACD76853), PKSG2
(annotated ACD76854), PKSF3 (annotated ACD76852), PKSG4 (annotated ACD76855),
PKSG5 (annotated ACD76856) (Flores-Sanchez et al., 2010), OLS (annotated BAG14339)
(Taura et al., 2009), CAN383 (annotated ACT10338) (Marks et al., 2009), and CHS2004
(annotated ACT10338) (Raharjo et al., 2004b). Residues which differ from the consensus are
depicted in blue if they share similarity with the consensus; residues that do not share similarity
are depicted in red. TKS and PKSG4 share 100% identity.
55
3.3.2 Expression and Purification of TKS
TKS was purified to near homogeneity by exploiting a poly-histidine tag by IMAC using Talon.
The expression and purification of TKS gave yields typically reaching 20 mg/L. Purified TKS
was visualized by SDS-PAGE on a 10% acrylamide gel (Figure 3.2).
Figure 3.2 Expression and purification of TKS.
The SDS-PAGE analysis of a (MW) BioRad Dual Color protein marker and samples taken
during the pre-induction (pre) and post-induction (post) stages of E. coli BL21 cell cultures
harboring CsTks, and from 1 mL eluents of IMAC purified TKS from Talon resin (lanes 1 to 7).
Purified TKS migrated at its estimated 46.5 kDa size.
56
3.3.3 TKS produces HTAL, PDAL, and olivetol
Different analytical methods were explored to determine which could best detect TKS assay
products including the use of radio-labeled
14
C-malonyl-CoA and reverse-phase thin-layer
chromatography (Figure 3.3). The analysis of assay products by reverse-phase thin-layer
chromatography indicated the production of three compounds assumed to be HTAL, PDAL, and
olivetol, although it was determined that the use of an HPLC, and later LCMS, proved a better
option for analysis since it is capable of superior product resolution, quantification, and mass
confirmation (Figure 3.4). The use of authentic standards (Sigma) allowed for the identification
and quantification of the three compounds produced during TKS assays: hexanoyl triacetic acid
lactone (HTAL), pentyl-diacetic acid lactone (PDAL), and olivetol, but no OA was produced
(Figure 3.4). These results confirm that olivetol synthase activity as reported by Taura et al.,
(2009) and Marks et al., (2009) is also present in TKS. Standard and product LCMS traces,
along with their UV spectra and masses are shown in Figure 3.4. The MS characteristics of
HTAL, PDAL, OA, and olivetol were also analyzed by SIR is ES+ mode (Section 3.1.3 for
details) (Figure 3.5). Note that the results shown in Figures 3.4 and 3.5 are from assays
conducted once LCMS methods were well developed, and earlier results (such as seen in Figures
3.13and 3.15), which were based only on HPLC analysis, have product retention times that
slightly differ from later LCMS analysis. TKS assays accumulated PDAL, olivetol and HTAL in
a ratio of 10:5:1, respectively. Although the ratio of triketide:tetraketide production varied during
the intial characterization of TKS, the production of PDAL was always greater through all
experiments performed.
57
Figure 3.3 Reverse phase radio TLC analysis of TKS assay products.
Reverse-phase thin layer chromatography (TLC) was used to separate TKS assay products
(HTAL, PDAL, and olivetol) which were analyzed by (A) phosphorimaging and (B) radioactive
scanning. Products detected are assumed to be (1) HTAL, (2) PDAL, and (3) olivetol, where (0)
is the origin. Further developments of these methods were abandoned for HPLC and LCMS
methods which led to a better resolution of the products.
58
A
absorbance at 270 nm
TKS
0.04
olivetolic acid
HTAL
B
0
1
2
4
5
6
6.0 min PDAL
285.8
300
nm
225.05
287.0
300
nm
183.01
0.045
200
0.035
3
5.3 min HTAL
200
absorbance at 270 nm
PDAL
olivetol
0.02
0.00
standards
600
m/z
400 200
1000 200
600
m/z
7
8
time (min)
9
10
200
1000
200
300
nm
225.2
600
m/z
12
13
14
10.2 min olivetol
276.3
9.0 min OA
263.3
400
11
400
200
1000
200
300
nm
180.99
400
600
m/z
1000
0.025
0.015
0.005
4
5
6
7
8
9
10
11
12
time (min)
absorbance at 270 nm
C
0.06
5.9 min PDAL
5.2 min HTAL
284.6
200
300
0.04
nm
225.05
0.02
200
600
m/z
287.0
200
300
nm
182.96
400
200
1000
600
m/z
10.2 min olivetol
276.3
200
300
400
nm
180.99
400
1000
200
600
m/z
1000
0
4
5
6
7
8
time (min)
9
10
11
12
Figure 3.4 LCMS analysis of standards and TKS products.
(A) LC chromatograms for TKS assay products (black) and standards (red) are overlayed to
emphasize the similarity in retention times, except for OA which was absent in TKS assays. (B)
An LCMS chromatogram of the standards used in this study indicate the retention times, UV
spectra and extracted ion profiles for HTAL (5.2 min; UVmax 285 nm; m/z 225), PDAL (5.6 min;
UVmax 287 nm; m/z 183), OA (9.0 min; UVmax 263 nm; m/z 225.), and olivetol (10.2 min; UVmax
276 nm; m/z 181) and used to confirm the identities of (C) TKS assay products shown in a
similar LCMS trace.
59
Figure 3.5 Selected ion monitoring of TKS assay products.
LCMS was used to confirm the identity of HTAL, PDAL and olivetol produced during TKS
assays. (A) PDA results align well with selected ion monitoring in ES+ mode which was capable
of detecting (B) olivetol (10.2 min; m/z 180.91), (C) PDAL (5.9 min; m/z 182.96), and (D)
HTAL (5.2 min; m/z 224.95). Another peak (5.9 min) detected during (D) selected ion
monitoring at 224.95 m/z is attributed to acetonitrile:PDAL adduct formation.
60
3.3.3.1 Quantification of TKS products
Quantification of TKS reactions products (HTAL, PDAL, OA, and olivetol) was performed
using the standard curves presented in Appendix A.
These standard curves were used to
determine product yields (pmol) from assays conducted throughout this study.
3.3.3.2 Buffers and pH
The effect of buffering agents and pH on TKS activity was examined by assaying TKS in a
series of buffers that overlapped in pH ranges (Figure 3.6). Assay results indicate that although
olivetol production increases at a lower pH (ca. 6.0) in phosphate and MES buffers, similar
results could be obtained by using HEPES at pH 7.0. The pH optimum for product formation was
also analyzed by surveying TKS activity in ADA buffer between pH 5.7 – 7.8 allowing for a
better pH range using only one buffer. The maximum production of tetraketides (HTAL and
olivetol) occurred at a pH 6.4, whereas maximum PDAL production was at pH 7.3 (Figure 3.7).
No OA production was observed.
61
500
HTAL
PDAL
olivetol
amount (pmol)
400
300
200
100
TRIS 9.0
TRIS 8.5
TRIS 8.0
HEPES 8.0
HEPES 7.5
HEPES 7.0
phosphate 8.0
phosphate 7.5
phosphate 7.0
phosphate 6.5
phosphate 6.0
MES 6.5
0
Figure 3.6 TKS products in different pH and buffers.
TKS product formation was first assessed in a pH range of 6.5 – 9.0 using a variety of buffers.
HEPES [pH 7.0] was determined to be the most stable buffer with respect to TKS activity based
on this data set. Error bars represent SD where n=3.
62
amount (pmol)
A 2000
amount (pmol)
B
PDAL
1000
0
5.6
5.8
6
6.2
6.4
6.6
6.8
7
7.2
7.4
7.6
200
HTAL
100
0
5.6
5.8
6
6.2
6.4
6.6
6.8
7
7.2
7.4
7.6
C 600
amount (pmol)
7.8
7.8
olivetol
400
200
0
5.6
5.8
6
6.2
6.4
6.6
6.8
pH
7
7.2
7.4
7.6
7.8
Figure 3.7 TKS product formation in ADA buffer.
The accumulation of TKS products was monitored between pH 5.7 and 7.8 in 20 mM ADA
buffer. Product analysis for (A) HTAL, (B) PDAL, and (C) olivetol are shown in the
corresponding panel. Error bars represent SD where n=3.
63
3.3.3.3 Temperature
To determine the effect temperature had on TKS activity, assays were performed between 10°C
and 50°C in 100 mM HEPES buffer at pH 7.5. This pH was chosen since it corresponded to the
maximum tetraketide production. Olivetol production appeared static until 40°C, but decreased
at 50°C (Figure 3.8). Comparatively, HTAL and PDAL synthase activity appeared to increase
until 40°C. PDAL production was slightly reduced 50°C, whereas HTAL production was
approximately the same as seen in assays conducted at 40°C.
amount (pmol)
700
HTAL
PDAL
olivetol
500
300
200
100
50
40
30
20
10
0
temperature (°C)
Figure 3.8 Temperature affects TKS product formation.
The accumulation of TKS products HTAL (red), PDAL (blue) and olivetol (green) was
determined between 10 to 50°C. Assays were conducted in 100 mM HEPES [pH 7.5] in the
presence of 0.2 mM malonyl-CoA and 0.2 mM hexanoyl-CoA. Vertical axis is split to show
smaller amounts of product accumulation. Error bars represent the SD where n=3.
64
3.3.3.4 Substrate Availabilty
TKS was also assayed in the presence of 0.10 to 1.0 mM malonyl-CoA while 0.2 mM hexanoylCoA was maintained, and similarly in the presence of 0.10 – 1.0 mM hexanoyl-CoA while 0.2
mM malonyl-CoA was maintained. The concentration of both malonyl-CoA and hexanoyl-CoA
affected the chemoselectivity of the TKS reaction. No product was detected when 0.10 mM
malonyl-CoA was used in assays, but an increase in malonyl-CoA concentration correlated with
an increase in product formation, except for PDAL which slightly decreased when 1.0 mM
malonyl-CoA was used (Figure 3.9). Maximum PDAL production (938 ± 123 pmol) was
detected when TKS was assayed with of 0.5 mM malonyl-CoA and 0.2 mM hexanoyl-CoA.
Both HTAL (54 ± 8.4 pmol) and olivetol (79.0 ± 26.5 pmol) production were increased when
TKS was assayed with 1.0 mM malonyl-CoA.
The availability of hexanoyl-CoA also affected TKS product formation (Figure 3.10).
Olivetol accumulated to greater amounts as hexanoyl-CoA increased, whereas HTAL and PDAL
production peaked at 0.2 mM hexanoyl-CoA.
Product ratios used to assess TKS triketide:tetraketide chemoselectivity indicate PDAL is
preferentially formed when TKS was assayed with 0.1 mM hexanoyl-CoA and 0.2 mM malonylCoA (Table 3.4). Conversely, increasing either malonyl-CoA or hexanoyl-CoA concentrations to
1.0 mM resulted in the increased synthesis of tetraketides. Product ratios were also used to
evaluate HTAL:olivetol chemoselectivity, which clearly reflect the preferential production of
olivetol when hexanoyl-CoA concentrations were raised (Table 3.4). The triketide:tetraketide
ratio represents total PDAL divided by the sum of HTAL and olivetol.
65
amount (pmol)
2500
1500
500
250
HTAL
PDAL
olivetol
200
150
100
50
1.
0
0.
5
0.
2
0.
1
0
[malonyl-CoA] (mM)
Figure 3.9 TKS product formation in varying malonyl-CoA concentrations.
amount (pmol)
The amount of TKS products HTAL (red), PDAL (blue), and olivetol (green) were determined
after TKS was assayed in the presence of 0.2 mM hexanoyl-CoA and 0.1 to 1.0 mM malonylCoA. Vertical axis is split to show smaller amounts of product. Error bars are the SD where n=3.
1200
900
600
200
HTAL
PDAL
olivetol
150
100
50
1.
0
0.
5
0.
2
0.
1
0
[hexanoyl-CoA] (mM)
Figure 3.10 TKS product formation in varying hexanoyl-CoA concentrations.
The amount of TKS products HTAL (red), PDAL (blue), and olivetol (green) were determined
after TKS was assayed in the presence of 0.2 mM malonyl-CoA and 0.1 to 1.0 mM hexanoylCoA. Vertical axis is split to show smaller amounts of product. Error bars are the SD where n=3.
66
Table 3.4 Ratios of product formed during TKS assays.
0.1 mM malonyl-CoA
0.2 mM malonyl-CoA
0.5 mM malonyl-CoA
1.0 mM malonyl-CoA
triketide : tetraketide*
nd
11.4
11.6
6.2
HTAL : olivetol
nd
0.31
0.59
0.68
0.1 mM hexanoyl-CoA
0.2 mM hexanoyl-CoA
0.5 mM hexanoyl-CoA
1.0 mM hexanoyl-CoA
13.2
11.4
9.3
7.1
0.42
0.31
0.20
0.11
* Ratios for triketide/tetraketide were determined by comparing the amount of PDAL by the sum
amount of HTAL and olivetol derived from Figure 3.9 and 3.10.
nd = not detected.
67
3.3.3.5 TKS Activity in the Presence of DTT
The use of dithiothreitol (DTT) and β-mercaptoethanol (βME) during protein purification, as
well as the use of DTT in TKS assays, required an analysis of the potential deleterious effects
caused by these strong reducing agents. Two solutions of TKS derived from the same culture
were purified in tandem; one in the presence of reducing agents (DTT and βME), the other in
their absence. Both protein solutions (± DTT/ βME) were assayed in increasing concentrations of
DTT and product formation was monitored. TKS purified in the absence of DTT and βME
possessed < 70% activity of TKS purified in the presence of DTT and βME (Figure 3.11). The
addition of 5 mM DTT to TKS assays increased HTAL and PDAL production by 69% and 56%,
respectively, although olivetol formation decreased by 10%. TKS activity was reduced when
DTT concentrations exceeded 5 mM, and resulted in the least amount of product when TKS was
assayed in the presence of 100 mM DTT.
68
A
HTAL
amount (pmol)
30
Presence
Absence
20
10
10
0
20
10
5
0
0
[DTT] (mM)
amount (pmol)
B
PDAL
1500
Presence
Absence
1000
500
10
0
20
10
5
0
0
[DTT] (mM)
amount (pmol)
C
olivetol
150
Presence
Absence
100
50
10
0
20
10
5
0
0
[DTT] (mM)
Figure 3.11 TKS in the presence and absence of DTT and βME during purification.
Two different preparations of recombinant TKS were assayed; one purified with DTT and βME
(black), the other purified without DTT and βME (grey). Assays were conducted in increasing
concentrations of DTT (0 to 100 mM) to determine if TKS product formation would be affected.
Panels show the production of (A) HTAL, (B) PDAL, and (C) olivetol. Error bars are the SD
where n=3.
69
3.3.4 OA Production by Trichome Protein Extracts
Protein extracts from isolated trichomes obtained from drug-type and USO-31 hemp plants were
separately assayed. Trace amounts of OA were detected in both samples by HPLC analysis
implying OA synthase activity (Figure 3.12). Whereas USO-31 hemp trichome extracts
accumulated HTAL, PDAL, OA and olivetol, no olivetol was detected in drug-type trichome
extracts. In an attempt to increase the synthesis of assay products, trichome protein extracts
obtained from USO-31 hemp were assayed in a scaled-up reaction using ten times more the
starting materials than previous assays (Figure 3.12). In this larger reaction, HTAL, PDAL, OA,
and olivetol accumulated to a ratio of 1:4:1:2, respectively. Absolute amounts were determined
to be 167 pmol HTAL, 658 pmol PDAL, 161 pmol OA, and 318 pmol of olivetol.
Drug-type trichome activity was further investigated in the presence and absence of
recombinant TKS. Reactions used the same amount of TKS and/or trichome extracts in each
assay. Trichome extracts, when assayed in the presence of TKS, produced 638% more OA than
trichome extracts alone (Figure 3.13). HTAL and olivetol yields were also greater in the
combined assays than in assays performed using only trichome protein (858% and 616%,
respectively) or recombinant TKS (209% and 180%, respectively). The ratio of
triketides/tetraketides formed was also markedly different when TKS was assayed alone (ratio
8:1), than when trichome extracts were assayed in the absence or presence of TKS (ratio 1:4 and
1:0, respectively). The HTAL:PDAL:OA:olivetol product ratio for: recombinant TKS assays was
1:45:0:4; trichome protein extracts was 1:7:2:5; for trichome protein extracts in the presence of
recombinant TKS was 1:9:2:4.
70
A
B
0.1
OA standard
extracts
PDAL
absorbance at 270 nm
0.08
HTAL
olivetol
OA
0.06
0.04
0.02
0
5
6
7
time (min)
8
9
Figure 3.12 Assays of cannabis trichome protein extracts
(A) HPLC chromatograms of products formed during cannabis trichome extract assays indicate
the accumulation of OA when compared to an OA standard, except for trichomes assayed in the
absence of malonyl-CoA (negative control). (B) An LC chromatogram of products formed
during a larger USO-31 hemp trichome extract assay clearly detected HTAL, PDAL, olivetol,
and OA.
71
HTAL
PDAL
OA
olivetol
amount (pmol)
500
250
250
200
150
100
50
TK
S
+
tri
ch
o
tri
ch
om
m
e
e
ex
tra
TK
S
ct
ex
tra
ct
0
Figure 3.13 Drug-type trichome extracts assayed with recombinant TKS.
Product yields obtained when TKS and trichome extracts were assayed together or alone indicate
the amount of HTAL (red), PDAL (blue), OA (purple), and olivetol (green) detected. Vertical
axis is split to show smaller amounts of product accumulation.
72
3.3.4.1 Ammonium Sulfate Precipitation of Drug-Type Trichome Protein Extracts
Once it was determined that trichomes possess OA synthase activity (Figure 3.12), ammonium
sulfate was used to precipitate trichome proteins, which might have been able to separate olivetol
and OA synthase activity, thus implying that separate proteins are required during OA
biosynthesis. A protein solution obtained from isolated trichomes was precipitated using
increasing amounts of ammonium sulfate (20%, 40%, 60%, and 80% w/v) followed by
centrifugation. The precipitates were resuspended in 100 mM HEPES buffer [pH 7.5]. An aliquot
from each fraction, along with a sample of the remaining supernatant was separated by
electrophoresis before being transferred to a PVDF membrane (Figure 3.14A and 3.14B). A
Western blot was then performed using polyclonal αTKS antibodies, and showed that the
greatest accumulation of TKS occurred in the 80% cut (Figure 3.14). Polyclonal αTKS
antibodies also detected impurities in recombinant TKS samples, and to a lesser degree, in each
of the precipitated fractions.
The precipitated protein fractions, along with the remaining unprecipitated supernatant
and a positive TKS control, were separately assayed in the presence of hexanoyl-CoA and
malonyl-CoA to determine if any of the precipitated protein fractions retained OA synthase
activity. Resuspended protein derived from the 80% salt cut had OA synthase activity, but other
fractions, including the supernatant, had neither TKS nor OA synthase activity, except for what
appeared to be a trace amount of OA detected in the 0% fraction (Figure 3.14).
Due to the lack of PDAL, HTAL, and olivetol, the appearance of what appears to be OA
in 0% ammonium sulfate fractions is believed to be a contaminant.
73
Figure 3.14 Activity of trichome protein extracts in ammonium sulfate.
(A) 10% SDS-PAGE of equal volume of trichome protein precipitates occurring at 0%, 20%,
40%, 60%, and 80% ammonium sulfate fractions were run with Protein Ladder (MW), 1 µg of
recombinant TKS (TKS), and an aliquot of unprecipitated supernatant obtained from post 80%
ammonium sulfate treatments (S). TKS is circled in red. (B) A Western blot obtained after
transferring proteins from (A) to a PVDF membrane indicates the majority of TKS is detected in
80% ammonium sulfate fractions. (C) HPLC traces of assay products produced from
recombinant TKS and trichome extracts obtained from the ammonium sulfate precipitates are
offset for comparison.
74
3.3.4.2 The Co-Immunoprecipitation of Trichome Protein Extracts
Polyclonal αTKS antibodies were used in co-immunoprecipitation experiments designed to
retrieve TKS and potentially interacting proteins from drug-type trichome protein extracts.
Purified rabbit IgG anti-TKS antibody-linked resin were incubated with either 1) a crude
recombinant TKS solution, or 2) a protein solution extracted from isolated trichomes. Control
resin was also used to determine non-specific protein binding to the resin. The antibody-linked
resin was capable of successfully binding TKS from solutions, as seen when SDS-PAGE with
Coomassie stain was used to visualize proteins (Figure 3.15). Another protein of (ca.) 60 kDa
was also purified from both recombinant and trichome extracts, and sent to the PBI mass
spectrometry lab for Q-TOF MS/MS analysis. The unknown protein(s) was identified as an A.
thaliana chaperonin homolog (annotated NP_175945) (Appendix C). Chaperonins are highly
conserved proteins involved in the proper folding of other proteins (Walter and Buchner, 2002).
The presence of TKS was also confirmed by Q-TOF MS/MS analysis (Appendix D).
Figure 3.15 Co-immunoprecipitation of TKS and trichome protein extracts
Co-immunoprecipitation was conducted on (A) recombinant TKS and (B) trichome protein
extracts using polyclonal αTKS antibody-coupled resin. The resin was capable of binding
recombinant TKS (A; lane 7; 45 kDa) and an unknown protein (A; lane 7; ca. 60 kDa), as well as
TKS found in trichome extracts (B; lane 7; 42 kDa) and another protein (B; lane 7; ca. 60 KDa).
Lanes contained (1) Protein Ladder (numbered in kDa), and 20 µL of (2) flowthrough from resin
columns, (3) first wash, (4) second wash, (5) third wash, (6) fourth wash, (7) first elution, (8)
second elution, and (9) third elution. Drawn circles show the eluted proteins.
75
3.3.4.3 Protein Pull-Down Assays of Trichome Protein Extracts
Recombinant TKS was bound to Talon beads and incubated with trichome protein extracts in an
attempt to capture proteins that interact with TKS. Results from this experiment did indicate the
presence of a ca. 30 kDa protein (Figure 3.16A; lanes 3 and 4) that was not present in a fresh
TKS control (Figure 3.16A; lane 2). Although dilute, trichome extracts did not show this band
(Figure 3.16A; lane 5 and 6). The unknown protein is assumed to be degraded and fragmented
TKS as it was also detected in older TKS-only preparations (Figure 3.16B).
Figure 3.16 Protein pulldown using TKS and cannabis drug-type trichome extracts.
(A) SDS-PAGE of proteins used in an attempt to pull down potentially interacting proteins from
trichome extracts. Lanes contained (1) 10 µL Protein ladder (numbered in kDa), (2) 2 µg
recombinant TKS, (3) 3 µg recombinant TKS following incubation with trichome extracts, (4) 6
µg recombinant TKS following incubation with trichome extracts, (5) 1 µg trichome extract, (6)
2 µg trichome extract. (B) A 10% acrylamide gel containing (1) 10 µL Fermentas Prestained
Protein ladder, and (2) 1 µg recombinant TKS. Although a new band appears in (A) lanes 3 and
4 at ca. 30 kDa, this was presumed to be a TKS degradation product after it was also observed in
(B) a TKS-only sample electrophoresed two days later.
76
3.3.5 The Analysis of cDNA Cyclase Candidates
The goal of experiments was to identify, clone and express trichome-specific proteins that were
potentially involved in the synthesis of OA. Candidate transcripts that were both highly abundant
in the cDNA library and showed similarity to proteins with cyclase or cyclase-like activity were
sub-cloned, expressed and purified before being assayed in the presence and absence of TKS
(Figure 3.17). Candidate proteins included a putative chalcone isomerase (CHI; annotated
Q6V7U9), a Betv1-like protein (pathogenesis related protein PR10; annotated B5KVN9) and a
DABB protein (probable protein Pop3; annotated Q9LUV2) (Table 3.1). The inclusion of CHI
was based on it constituting 0.52% of the total transcripts in trichome ESTs and its relationship
with CHS in other plants, where CHI is capable of cyclizing chalcone to chalcone naringenin
(Gensheimer and Mushegian, 2004; Jez et al., 2001). The Betv1-like protein (also known as
major latex protein/MLP) constituted 0.25% of the total trichome ESTs, and shares structural
similarity to Streptomyces TcmN ARO/CYC, a PKC involved in the biosynthesis of
tetracenomycin (Ames et al., 2008). The DABB protein is also well represented in the trichome
ESTs, representing 0.45% of the total library, and is structurally similar to Streptomyces TcmI, a
cyclase also involved in the biosynthesis of tetracenomycin (Thompson et al., 2004).
Candidate proteins were assayed in the presence and absence of TKS to determine if they
produced OA at a ratio of 100 pmol TKS (ca. 2 µg) to 500 pmol of candidate protein (ca. 20 µg).
Assays were performed in triplicate and their products analyzed by HPLC (Figure 3.18). HPLC
chromatograms clearly show that, when TKS is assayed in the presence of the DABB candidate
cyclase, OA is detectable, along with HTAL, PDAL, and olivetol. No OA accumulated when
TKS was assayed alone, or in the presence of MLP or CHI, although the other TKS products (i.e.
77
PDAL, HTAL, and olivetol) were detected. The accumulation of TKS products also varied
between samples (Figure 3.19). The amount of PDAL appeared invariable amongst the samples
tested, with an average of 2357 ±SE 45.6 pmol produced (where n = 3). The amount of HTAL
produced appeared static when TKS was assayed alone, or with the DABB protein, or CHI, with
an average of 193 ±SE 4.3 pmol; less was produced when TKS was assayed in the presence of
MLP (119 ±SD 10 pmol). The production of olivetol was the most varied between samples, with
an average of 389 ±SE 20 pmol when TKS was assayed alone or in the presence of CHI, but less
when TKS was assayed with the DABB protein or with MLP, producing 166 ±SD 32 pmol and
228 ±SD 27 pmol of olivetol respectively. The ratio of HTAL to olivetol produced when TKS
was assayed in the presence of the DABB protein is 1.2, whereas it was 0.5 when TKS was
assayed in the presence or absence of CHI or MLP.
Figure 3.17 SDS-PAGE of TKS and candidate cyclases.
SDS-PAGE was used to analyze recombinant TKS (TKS; 44.5 kDa), and putative tetraketide
cyclases including a DABB protein (18.7 kDa), Betv1-like protein (MLP; 20.7 kDa), and a
chalcone isomerase-like protein (CHI; 27.6 kDa). Each IMAC-purified protein was loaded onto
the 12% acrylamide gel using (first lane) 1.25 µg and (second lane) 3.75 µg of protein and were
run next to a molecular weight marker (MW).
78
Figure 3.18 TKS activity in the presence of candidate cyclases.
HPLC traces of the products formed when TKS was assayed (A) alone, or with (B) MLP-like,
(C) CHI-like, or (D) DABB proteins are overlayed. Whereas HTAL (RT ca. 5.0 min), PDAL
(RT ca. 5.5 min), and olivetol (RT ca. 9.5 min) were detected in all assays, OA (RT ca. 8.3 min)
was only detected when TKS was assayed in the presence of the DABB protein. Standards used
during HPLC analysis are overlayed (red trace).
79
amount (pmol)
3000
HTAL
PDAL
OA
olivetol
2000
500
400
300
200
100
B
I
D
A
B
C
H
+
TK
S
+
TK
S
M
LP
+
TK
S
TK
S
on
ly
0
Figure 3.19 TKS assays products when assayed in with putative cyclases.
The absolute pmol amount of TKS products detected when TKS was assayed in the presence and
absence of CHI, MLP, and DABB proteins is shown. OA was only detected when TKS was
assayed in the presence of the DABB candidate. Vertical axis is split to show smaller amounts of
product accumulation. Error bars are the SD where n=3.
80
3.4 Discussion
Cannabis contains many secondary metabolites, including cannabinoids, stilbenes, anthocyanins,
flavonoids, and spirans, that are synthesized by different PKSs (Flores-Sanchez and Verpoorte,
2008b). The biosynthesis of these compounds requires type III PKSs that accept different starting
substrates and vary in their chain lengthening activity and cyclization modes. The first cannabis
CHS reported in the literature was identified using degenerate primers based on the sequences of
the H. lupulus CHS and phlorisovalerophenone synthase (VPS). This cannabis CHS was
described as a CHS and VPS (Raharjo et al., 2004b). It is capable of accepting p-coumaroyl-CoA
as a reaction initiator, but fails to use hexanoyl-CoA as a substrate, making it incapable of
producing OA. The subsequent identification of different PKSs in cannabis revealed an
evolutionary divergence between PKSs that cluster either with known CHSs, such as the H.
lupulus CHS/VPS (annotated CAC19808), or with CHS-like sequences, which are not involved
with the biosynthesis of naringenin chalcone (Flores-Sanchez et al., 2010). TKS shares 100%
amino acid identity with PKSG4, a CHS-like paralog that has only been characterized using in
silico methods (Flores-Sanchez et al., 2010). TKS also shares 98.7% amino acid identity with
another CHS-like homolog, OLS, which has been biochemically characterized and has had its
crystal structure solved (Taguchi et al., 2008; Taura et al., 2009). Although the characterization
of OLS has been conducted, the discovery of TKS occurred at approximately the same time. The
partial characterization of TKS which is described in this chapter indicates that TKS has similar
activity as OLS. Amino acid differences between OLS and TKS do not appear to affect olivetol
synthase activity, although a future study could investigate all known TKS homologs for further
robust analysis.
81
The characterization of TKS in this study was not conducted under the same conditions
as used for OLS (Taura et al., 2009). In the present work, a variety of buffers were used to
determine the optimum for the production of tetraketides, and although TKS assays conducted in
ADA buffer showed an increased tetraketide production at pH 6.4, a greater absolute amount of
tetraketides, as well as a lower triketide:tetraketide ratio resulted when TKS was assayed in
HEPES buffer at pH 7.0, and therefore further assays were conducted using HEPES. The
observed changes in TKS activity when assayed in different buffers may be attributed to the
enzyme’s conformational stability (Ugwu and Apte, 2004). All in vitro assays resulted in PDAL
and HTAL production, which are suspected to be early derailment products resulting from
artificial in vitro conditions (Austin and Noel, 2003). It is not known if the biosynthesis of
pyrones by TKS has a specific biological function, or if they are even produced in the plant,
although the expression of Arabidopsis CHS-like LAP5 and LAP6, which also produce alkyl
pyrones during in vitro assays, have been shown to be required for pollen exine development
which plays a crucial role in plant reproduction (Dobritsa et al., 2010). Early characterization
work continued using HEPES buffer at pH 7.5, since a greater amount of tetraketides were
produced. Contrastingly, Taura et al. (2009) assayed OLS in citrate buffer at a pH 5.5. TKS and
OAC assays with citrate buffers are shown in Chapter 4.
The temperature at which assays were conducted in earlier work was determined to be
optimum at 40°C based on the overall amount of products formed (Figure 3.8). At 50°C, both
olivetol and PDAL accumulated to a lesser degree than at lower temperatures, whereas the
production of HTAL was slightly increased. An increase in HTAL production, coupled to a
decreased accumulation of PDAL and olivetol, suggest that the overall enzyme activity is
82
decreased at > 40°C. Furthermore, the increase of HTAL at higher temperatures coupled to a loss
of olivetol suggests that HTAL and olivetol share a common tetraketide intermediate pre-cursor.
The lactonization of a tetraketide intermediate to HTAL may be a spontaneous and
thermodynamically-favored process compared to olivetol cyclization, which may require a
slower step so to achieve cyclization. Alternatively, enzyme stability is compromised at
temperatures > 40°C, leading to an increase in enzyme derailment products.
Substrate availability assays indicate that increased malonyl-CoA concentrations result in
a lower triketide:tetraketide ratio (Figure 3.9 and Table 3.4). This intuitively seems correct, since
a growing polyketide chain that is attached to a PKS active site requires malonyl-CoA to produce
a final, longer product, and that a (shorter) derailment product (i.e. triketide leading to PDAL)
would result if less malonyl-CoA was available. Although a greater malonyl-CoA concentration
favors the production of tetraketides, HTAL production is also favored over the production of
olivetol under these conditions. An increase in the availability of hexanoyl-CoA also decreased
the triketide:tetraketide product ratio, but a greater amount of olivetol was synthesized under
these conditions instead of HTAL (Figure 3.10 and Table 3.4). These results show that TKS
chemoselectivity can be altered by varying substrate concentration and suggest that excess
malonyl-CoA, or possibly its breakdown product, acetyl-CoA, either destabilize an enzymetethered tetraketide intermediate causing its early release and cyclization to HTAL, or influences
the cyclization of a linear tetraketide intermediate that has already been released by TKS, to
favor HTAL production. The use of a malonyl-CoA:hexanoyl-CoA substrate ratio of 3:1 in
subsequent assays was used based on the stoichiometry of tetraketide biosynthesis.
83
DTT is a strong reducing agent routinely added to buffers used during the purification
and assaying of enzymes. DTT is used to prevent cysteine residues from forming disulfide
linkages within and between proteins, and to assure the reduction of the cysteine residue in the
TKS active site. Although DTT has been used extensively, there was concern regarding the
potential interfering effects which could present themselves as a problem in TKS reactions. Type
III PKS activity is dependent on the modification of thiol bonds involving CoA esters (Austin
and Noel, 2003). Since a similar reaction mechanism is involved between type III PKSs and
DTT through the use of sulfhydryl bonds, the additive and deleterious effects of DTT were
examined in TKS assays. Results indicate that the addition of DTT and βME in purification
buffers has a beneficial effect on TKS activity, although the use of DTT in enzyme assays
increases the production of pyrones, but not olivetol, at lower concentrations, and negatively
affects all TKS activity at concentrations > 5 mM. The use of DTT in enzyme assays was
deemed beneficial at concentrations of 5 mM.
The assaying of trichome-derived protein solutions provided direct evidence of OA
synthase activity in cannabis (Figure 3.12). Earlier studies that attempted to detect OAS activity
from cannabis tissues were unable to detect OA, and reported the production of olivetol instead
(Flores-Sanchez and Verpoorte, 2008a; Raharjo et al., 2004a). In this study, when a USO-31
hemp trichome protein solution was assayed, HTAL, PDAL, OA and olivetol were detected,
whereas only HTAL, PDAL, and OA were detected when a drug-type trichome protein solution
was assayed. The small amount of drug-type trichomes that were used may have not been
enough for the production of olivetol in detectable amounts. The availability of drug-type
trichomes restricted further analysis. Protein extract assays strongly suggest that TKS is involved
84
in the biosynthesis of OA since the addition of TKS to protein extract assays resulted in a 5-fold
increase of OA (Figure 3.13). Assays confirmed that cannabis trichomes possess OA synthase
activity, and led to further attempts at identifying other proteins required for the proper
cyclization of tetraketide products. The analysis of trichome extracts following treatments with
ammonium sulfate show that the OA synthase activity is retained in the 80% ammonium sulfate
precipitate, indicating that a protein and not a co-factor is involved. Furthermore, the isolation of
the OA synthase activity along with the majority of TKS in the same fraction suggested the
possibility that a physical interaction existed with TKS and the unknown protein. This instigated
a search for potential TKS interaction partners using co-immunoprecipitation and protein pulldown methods.
Co-immunoprecipitation experiments used αTKS polyclonal antibodies that were bound
to a solid-support agarose matrix. The finding of a chaperonin homolog was most likely due to
the cross-reactivity of polyclonal antibodies that were generated using TKS contaminated with
an E.coli chaperonin homolog (Figure 3.15). These experiments were unable to detect any
protein that physically interacts with TKS. In another effort to discover a TKS-interacting
protein, affinity chromatography (i.e. protein pull-down) was attempted after incubating
trichome extracts in the presence of recombinant TKS, then purifying TKS and any interacting
proteins using IMAC. An additional protein was detected during protein pull-down experiments,
but it was concluded to be a TKS degradation product after a similar protein band was detected
following the SDS-PAGE of TKS-only controls (Figure 3.16). Neither co-immunoprecipitation
nor protein pull-down methods were either capable of retrieving any TKS-interacting proteins
from trichome extracts, suggesting that no such protein exists, or that these methods were
85
incapable of detection. Experiments designed to isolate trichome proteins using chromatography
had been planned, and attempts to tease out an accessory protein from trichome extracts were
already underway, but were not continued since TKS assays conducted in the presence of
potential cyclases meanwhile led to the discovery of OAC.
An analysis of the trichome-specific cDNA library resulted in the identification of several
putative polyketide cyclases that could be involved in the biosynthesis of OA (Table 3.1).
Another cannabis cDNA library prepared from isolated trichomes had also been published at the
time of this study by Marks et al. (2009). TKS and its homolog (CAN24/OLS) in the Marks et al.
(2009) library were highly expressed in EST libraries, where TKS is the most abundant transcript
in the Stout et al. (2012) trichome cDNA library and CAN24 is the eighth most abundant
transcript in the Marks et al. (2009) library. Similarly, the DABB protein (homologous to A.
thaliana heat stable 1 protein; annotated
NP_566569) chosen as a potential cyclase was also
detected as a highly abundant transcript in both cDNA libraries, where it is the 15th-most
abundant transcript in the Stout et al. (2012) trichome cDNA library and the fifth-highest in the
Marks et al. (2009) library. The MLP-like and CHI-like transcripts were the 34th- and 9th-most
abundant transcripts detected in the trichome cDNA library published by Stout et al. (2012)
(Table 3.1), whereas they were the 10th- and 108th-most abundant in the Marks et al. (2009)
library, respectively.
Assays containing TKS and the DABB candidate led to the production of OA, whereas
TKS assays with the other candidate cyclases did not (Figure 3.18). Early TKS assays performed
in the presence of DABB protein produced an equimolar amount of tetraketides, with an
HTAL:PDAL:OA:olivetol ratio of approximately 1:10:1:1 (Figure 3.19). TKS assays preformed
86
in the absence of DABB protein produced an HTAL:PDAL:olivetol ratio of approximately
1:10:2, indicating a proportional two-fold increase in olivetol when compared to TKS assays
conduced in the presence of DABB protein. This observation indicates that 1) HTAL production
is static in all TKS assays performed, and that 2) OA and olivetol are synthesized using the same
TKS-produced intermediate. The discovery of the DABB protein’s cyclase activity prompted the
naming of this protein to olivetolic acid cyclase (OAC), the enzyme that is the focus of the next
chapter.
87
4. THE CHARACTERIZATION OF OAC
4.1 Introduction
Many polyketide cyclases have been reported in actinomycetes species, and are required during
the biosynthesis of antibiotics and pigments (Ames et al., 2011; Shen and Hutchinson, 1993b;
Sultana et al., 2004; Thompson et al., 2004). These cyclases are involved in the modification of
products derived from type II polyketide synthases, which do not occur in plants, although
structurally similar proteins possessing either the Betv1 or DABB motifs have been identified
(Bingman et al., 2004; Radauer et al., 2008). These motifs were identified in trichome proteins,
and due to their structural relationship with bacterial polyketide cyclases, I initiated an
investigation into their putative cyclase activity during the biosynthesis of OA. These
experiments led to the discovery of OAC, a DABB protein capable of producing OA when
assayed together with TKS (Gagne et al., 2012). The discovery of OAC warranted further
investigation of this protein. This chapter describes OAC sequence and phylogenetic analysis,
the identification of assay products by LCMS, and experiments designed to characterize OAC.
Since type III PKSs are known to exhibit flexibility in substrate specificity, TKS and OAC were
also assayed with alternative starter compounds to determine if TKS and OAC could produce
alternative resorcylic acids (Wakimoto et al., 2012). Lastly, an addition to the TKS/OAC assays
was made by adding recombinant malonyl-CoA synthetase (MCS) derived from Rhizobium
leguminosarum in hope of augmenting OA biosynthesis (Zhang and Tang, 2009). In summary,
the experiments presented in this chapter have helped optimize the TKS/OAC reaction, and
provide valuable insight into the biosynthesis of OA by the TKS/OAC reaction.
88
4.2 Materials and Methods
4.2.1 Phylogenetic Analysis of OAC
The nucleic acid sequence (derived from sequencing) of OAC and its deduced protein sequence
were used to conduct eight blast searches against a number of plant protein databases (described
in Section 4.3.2). Initially > 150 protein sequences were obtained, but a third of the sequences
were discarded since a limit of four representatives from any one species was applied.
Representative sequences were arbitrarily chosen. A total of 93 OAC-like sequences (including
OAC) were aligned using ClustalW provided in Mega5 (Tamura et al., 2011) with parameters as
described in Appendix E. A Maximum Likelihood Phylogenic Tree was constructed using the
Mega5 software and alignment file using the settings described in Appendix E. A thousand
bootstraps were performed during the analysis.
4.2.2 Biochemical Characterization of OAC
Assays were performed using 100 pmol TKS and 1,000 pmol OAC in 50 µL volumes containing
0.2 mM hexanoyl-CoA, 0.6 mM malonyl-CoA, 50 mM HEPES [pH 7.0], and 5 mM DTT.
Reactions were maintained at 30°C for 60 min before being extracted and analyzed by LCMS as
described in Section 3.3.1. The following subsections note differences in these methods. All
reactions were performed in triplicate.
4.2.2.1 TKS/OAC Ratios
Assays used 10 mM DTT and a range of 250 to 12,500 pmol OAC. Reactions were maintained at
37°C for 60 min. Reactions were analyzed by HPLC as described in Section 3.3.1.
89
4.2.2.2 Temperature
Assays used 0.5 mM malonyl-CoA. Mixtures were maintained at 10, 15, 20, 25 or 30°C for 5
min before starting the reaction. Reactions were held at the appropriate temperature for 60 min.
4.2.2.3 Buffers
Reactions were conducted in 50 mM buffers including HEPES, TRIS, sodium phosphate, and
PIPES. Reactions were incubated for 60 min at 10°C.
4.2.2.4 Concentration of Buffer
The concentration of HEPES used was 12.5, 25, 50, and 100 mM HEPES [pH 7.0].
4.2.2.5 pH
Assays were conducted in: HEPES pH [6.8 to 8.2]; phosphate pH [6.0 to 7.5] which used 0.5
mM malonyl-CoA and were conducted at 10°C; ADA pH [5.7 to 7.8] as described in Section
3.1.3.4 except for the addition of 1000 pmol OAC; and citrate pH [5.0 to 6.0] which used 20 mM
buffer. Assays conducted in citrate were run in parallel to reaction performed in 20 mM HEPES
pH [7.0] and 20 mM MOPSO pH [7.0]. Citrate buffers were prepared according to (Gomori,
1955).
4.2.2.6 Metals
Reactions were performed in 20 mM HEPES [pH 7.0], and 5 mM of MgCl2, MnCl2, ZnCl2,
FeCl2, CoCl2, CaCl2, CuCl2, or water. All metal solutions were freshly prepared from 50 mM
stock solutions. Analysis was performed by HPLC as described in Section 3.3.1.
90
4.2.2.7 EDTA
Reactions were performed in 20 mM HEPES [pH 7.0] in the presence or absence of 10 mM
EDTA. Reactions used 200 pmol TKS and 200 pmol OAC, and were maintained at 20°C for 1 h.
4.2.2.8 Malonyl-CoA Synthetase
A 1 mL aliquot of Talon-purified recombinant malonyl-CoA synthetase (MCS) (10 mg/mL) was
provided by Mr. Enwu Liu (Zhang and Tang, 2009). Assays using MCS were conducted in 50
µL containing 20 mM HEPES [pH 7.0], 0.2 mM CoA, 0.4 mM ATP, 2.5 mM MgCl2, 8 mM
sodium malonate, 5 mM DTT, and 0.2 mM hexanoyl-CoA. Assays were started by adding the
appropriate mixture of 5 µg TKS, 10 µg MCS, and 5 µg OAC. Reactions that did not use MCS
were conducted using 20 mM HEPES [pH 7.0] as described in Section 4.2.2. Assays were
performed at 20°C and were analyzed by LCMS as described in Section 3.3.1.
4.2.2.9 Alternative Substrate Assays
4.2.2.9.1 Free CoA
Assays used 0, 0.25, 0.50, 0.75, or 1.0 mM excess CoA esters. Reactions were started by adding
a mixture of 200 pmol TKS and 1,000 pmol OAC and were incubated at 37°C for 1 h.
4.2.2.9.2 TKS products
Assays were performed using 1.0 nmol of OAC in 50 mM HEPES [pH 7.0] and either 1 mM
HTAL, 1 mM PDAL, 1 mM olivetol, or 1 mM HTAL + 1 mM CoA and proceeded for 1 h at RT.
4.2.2.9.3 Butyryl-CoA, Octanoyl-CoA, and Acetyl-CoA
Assays used 0.2 mM butyryl-CoA, octanoyl-CoA, or acetyl-CoA instead of hexanoyl-CoA as a
starter molecule. Reactions also used 200 pmol TKS, ±1,000 pmol OAC, 0.6 mM malonyl-CoA,
and 20 mM HEPES [pH 7.0] and were incubated for 30 min at 30°C. Reactions were prepared
91
for LCMS analysis as described in Section 3.3.1. Assays using butyryl-CoA were separated over
a 30 min gradient run that started with (v/v) 90% water, 10% acetonitrile, and 0.05% formic acid
and ended with (v/v) 80% water, 20% acetonitrile and 0.05% formic acid. Assays using
octanoyl-CoA were separated over a 25 min run that started with (v/v) 98% water, 2%
acetonitrile, and 0.05% formic acid and ended with (v/v) 2% water, 98% acetonitrile and 0.05%
formic acid. Assays using acetyl-CoA were analyzed by LCMS as described in Section 3.3.1. All
flow rates were maintained at 0.2 mL/min with a 30°C column temperature. Products were
monitored at 270 nm and were analyzed using the Empower2 software suite. SIR was conducted
as described in Section 3.1.3; SIR m/z values are described in Section 4.3.10.
92
4.3 Results
4.3.1 Olivetolic Acid Cyclase
OAC is a 101 amino acid protein with a predicted molecular weight of 12 kDa. OAC was
initially identified in the CL2contig1 ORF of the CSATR1JP unigenes library previously
described by Stout et al. (2012), and cloned into the pET-100 vector resulting in a new construct
with a calculated molecular weight of 18.7 kDa. At the amino acid level, OAC shares 48%
identity with a heat stable protein from Arabidopsis thaliana named AtHS1 (annotated
NP_566569), and 38% identity to Populus tremula PtSP1 (annotated CAC34953), which is
upregulated during stress responses (Park et al., 2007). The crystal structure for AtHS1 has been
solved and confirms the presence of dimeric α + β barrel domains (Bingman et al., 2004).
Although AtHS1 was shown to possess antifungal and antibacterial activity, no specific
physiological function has been attributed to this protein (Park et al., 2007). Similarly, it is not
known why the heat-stable SP1 is upregulated during heat, salt and stress responses, yet A.
thaliana plants expressing transgenic SP1 have been shown to be heat-tolerant (Zhu et al., 2008).
Structurally, the OAC protein sequence was used to query the Interpro web resource, revealing a
plethora of DABB proteins with many different functions. Of specific relevance is the TcmI
protein from Streptomyces glaucescens (annotated AAA67513), which assists in the cyclization
of TcmF2 to TcmF1 during the biosynthesis of tetracenomycin C, an antitumor/antibiotic
aromatic polyketide (Thompson et al., 2004). The TcmI DABB domain belongs to a superfamily
of actinobacterial proteins known as the PKCs (Interpro IPR006765; CDD cl04663). It was
partially this structural relationship that initially implicated OAC as a candidate polyketide
cyclase. Structurally, the OAC DABB domain belongs to a different family of proteins than
93
TcmI, called the stress-responsive alpha beta barrel (SRABB) proteins by Interpro (IPR013097),
or simply as DABB proteins by the CDD (cl15807). Conversely, DABB proteins are recognized
by Interpro as a superfamily of structurally similar proteins (IPR011008) which include SRABBs
and PKCs, as well as many other structurally similar protein families.
4.3.1.1 Expression and Purification of OAC
Originally, OAC was cloned in the pET100 vector, expressed in E. coli BL21, and purified using
IMAC. Although this produced active protein to be processed, low amounts of impure protein
were obtained. The low expression of OAC in the pET100 vector led to the construction of the
pHIS8-OAC plasmid, but this change did not substantially increase protein yields, which were
only between 1-2 mg/L culture. Furthermore, the expression and purification of OAC from the
pHIS8-OAC vector resulted in protein fragmentation attributed to OAC degradation. During
SDS-PAGE, IMAC-purified OAC migrates as a (ca.) 19 kDa band next to a less intense 17 kDa
band (Figure 4.1). Both bands were analyzed by MS/MS, with results indicating that both bands
correspond to OAC, suggesting that the 17 kDa contaminant was degraded OAC. Various
attempts to further purify OAC using chromatographic techniques including size exclusion, ion
exchange and hydrophobic interaction chromatography had little success in removing the
degradation product. Assays that were performed throughout this chapter used OAC that was
purified using minimal amounts of Talon and, with the exception of the degradation product,
resulted in a purified active product (Figure 4.1).
94
Figure 4.1 SDS-PAGE of Talon purified TKS and OAC.
A 15% polyacrylamide gel was used to resolve 1.25 µg TKS (lane 3), 3.75 µg of TKS (lane 5),
1.25 µg OAC (lane 7) and 3.75 µg OAC (lane 9) next to 10 µL Prestained Protein ladder (lane 1)
at 200V for 45 min. OAC did show a contaminating band at (ca.) 17 kDa as seen in lanes 7 and
9. Numbers on the left of the gel are the corresponding protein masses (kDa).
4.3.2 The Evolutionary Relationship of OAC
A search for OAC homologs was conducted using the OAC protein sequence during a variety of
blast queries in a number of plant sequence databases, namely the NCBI Transcriptome Shotgun
Assembly Sequence Database (Benson et al., 2000), the Cannabis Genome database (Van Bakel
et al., 2011), TIGR Plant Transcript Assemblies (Childs et al., 2007), the Solanum lycopersicum
project (Tomato Genome Consortium, 2012), GreenPhylDB (Conte et al., 2008), and Phytozome
(Goodstein et al., 2012). A total of 93 protein sequences with an expected threshold <1x10-10
were aligned using MEGA5 software’s ClustalW implementation, and then used to construct a
Maximum Likelihood phylogenetic tree (Tamura et al., 2011) (Figure 4.2). Names of sequences
are included in Appendix F. A limit of four homologs from any given species to be included in
the alignment was applied when too many sequences from one species were detected. Sequences
used in the alignment were obtained from BLAST searches against higher plant genomes, but
95
also included Cyanidioschyzon merolae, Chlamydomonas reinhardtii, and Ostreococcus tauri
genomes, although no OAC homolog within an expected threshold of 1x10-2 could be detected
within these algal genomes. Phylogenic analysis indicates that OAC homologs form two distinct
clades that further subdivide into monocot and dicot lineages (Figure 4.3 and 4.4). Broader
searches for OAC homologs against GenBank non-redundant sequences clearly detected their
existence in a variety of bacteria including Proteobacteria, Firmicutes, Enterobacteriaceae, and
Sphingobacteria with E-values less than 1x10-14.
96
Figure 4.2 Phylogenetic relationship of plant OAC homologs.
The evolutionary history was inferred using the Maximum Likelihood method based on the JTT
matrix-based model (Jones et al., 1992). The bootstrap consensus tree inferred from 1,000
replicates represents the phylogenetic relationship of OAC (Felsenstein, 1985). Bootstrap values
are shown along branch lengths. The analysis involved 93 amino acid sequences (Appendix F).
There were a total of 58 positions in the final dataset. Analyses were conducted in MEGA5
(Tamura et al., 2011). Two distinct clades (A and B) divide the homologs, which can be further
divided into eudicot and monocot groups. The scale bar indicates the genetic distance of 0.2
substitution/site.
97
Figure 4.3 Phylogenetic relationship of OAC plant homologs in Clade A.
Plant OAC homologs found in clade A are shown. The scale bar indicates the genetic distance of
0.05 substitution/site.
98
Figure 4.4 Phylogenetic relationship of OAC plant homologs in Clade B.
OAC homologs that form the ‘B’ clade division are shown. These homologs can be further
subdivided into monocot and eudicot classifications. The scale bar indicates the genetic distance
of 0.2 substitution/site.
99
4.3.3 The TKS/OAC Reaction Produces Olivetolic Acid
TKS assays conducted in the presence of OAC produced four compounds identifiable by
comparing their retention times, UV spectra, and masses to authentic standards during LCMS
analysis (Figure 4.5). The four products included HTAL, PDAL, olivetol, and OA, which were
also confirmed by single ion monitoring in electrospray positive ion mode at 224.95 m/z for
HTAL and OA, 182.96 m/z for PDAL, and 180.91 m/z for olivetol (Figure 4.6). Product ratios
(i.e. chemoselectivity) were affected by assay conditions including temperature, pH, and
substrate concentrations and are discussed in the appropriate Sections. Product ratios generally
mirrored product accumulation of TKS assays performed in the absence of OAC, with the
exception of olivetol which accumulated to a lesser degree in TKS assays performed in the
presence of OAC, and of course OA which does not accumulate when TKS is assayed alone
(Figure 4.7). Typically, product ratios of HTAL:PDAL:OA:olivetol were in circa 1:10:1:1
throughout this work, but these values did fluctuate dramatically. Once reaction conditions were
optimized for OA production, a product ratio of 1:7:2:1 was obtained.
The LCMS chromatograms of standards and the TKS/OAC assay presented in Figures
4.5 and 4.6 can be directly compared to the chromatograms presented in Chapter 3 (Figure 3.4
and 3.5); they were obtained from experiments conducted in parallel. Note that the
chromatograms shown are derived from reactions conducted once optimized reaction conditions
were determined.
100
A
0.06
absorbance at 270 nm
0.05
TKS & OAC
PDAL
0.04
OA
0.03
HTAL
0.02
1
2
3
5.3 min HTAL
285.8
300
nm
225.05
200
absorbance at 270 nm
olivetol
0.01
B
0.045
200
0.035
600
m/z
4
5
6
7
8
time (min)
6.0 min PDAL
287.0
300
400 200
nm
183.01
1000 200
600
m/z
9
10
9.0 min OA
263.3
400
200
1000
200
300
nm
225.20
400
600
m/z
1000
11
12
13
14
10.2 min olivetol
276.3
200
300
400
nm
180.99
200
600
m/z
1000
0.025
0.015
0.005
4
C
5
6
5.2 min HTAL
284.6
300
nm
225.10
200
absorbance at 270 nm
standards
200
0.06
600
m/z
7
8
time (min)
6.0 min PDAL
285.8
300
nm
183.01
400
200
1000
200
400
600
m/z
1000
9
10
9.0 min OA
263.3
200
300
nm
225.15
200
400
600
m/z
1000
11
12
10.2 min olivetol
275.1
200
300
400
nm
181.08
200
600
m/z
1000
0.04
0.02
0
4
5
6
7
8
time (min)
9
10
11
12
Figure 4.5 LCMS of standards and TKS/OAC assay products.
(A) Overlayed LC chromatograms for TKS/OAC assay products (black) and standards (red)
indicate the similarity in product retention times. (B) An LCMS chromatogram of the standards
used in this study indicate the retention times, UV spectra and extracted ion profiles for HTAL
(5.2 min; UVmax 285 nm; m/z 225), PDAL (5.6 min; UVmax 287 nm; m/z 183), OA (9.0 min;
UVmax 263 nm; m/z 225.), and olivetol (10.2 min; UVmax 276 nm; m/z 181) and were used to
confirm the identities of (C) TKS/OAC assay products shown in a similar LCMS trace.
101
absorbance at 270 nm
A
0.06
PDA 270
0.04
0.02
0
4
ion count
B
5
6
7
8
time (min)
9
10
11
6.0E+04
12
ES+ 180.91
4.0E+04
2.0E+04
0.0E+00
4
ion count
C
5
6
7
8
time (min)
9
10
11
4.0E+07
12
ES+ 182.96
3.0E+07
2.0E+07
1.0E+07
0.0E+00
4
ion count
D
5
6
7
8
time (min)
9
10
11
2.0E+06
12
ES+ 224.95
1.5E+06
acetonitrile:PDAL adduct
1.0E+06
5.0E+05
0.0E+00
4
5
6
7
8
time (min)
9
10
11
12
Figure 4.6 Selected ion monitoring of TKS/OAC assay products.
LCMS was used to confirm the identity of HTAL, PDAL and olivetol produced during TKS
assays. (A) PDA results align well with selected ion monitoring in ES+ mode which was capable
of detecting (B) olivetol (10.2 min; m/z 180.91), (C) PDAL (6.0 min; m/z 182.96), (D) HTAL
(5.2 min; m/z 224.95) as well as olivetolic acid (9.0 min; m/z 224.95). Another peak (6.0 min)
detected during (D) selected ion monitoring at 224.95 m/z is attributed to acetonitrile:PDAL
adduct formation.
102
1400
TKS
TKS & OAC
amount (pmol)
1200
1000
800
600
400
200
0
HTAL
PDAL
OA
olivetol
Figure 4.7 A comparison of TKS and TKS/OAC assay products.
TKS assayed in the presence of OAC resulted in the accumulation of OA apparently at the cost
of olivetol accumulation. Error bars are the SD where n=3.
103
4.3.4 Optimum Ratio of TKS and OAC
Experiments were conducted to establish the ratio of TKS to OAC required to optimize the
production of OA. Additionally, this work tested if a stoichiometric relationship exists between
TKS and OAC, which could suggest if a physical interaction occurs between these enzymes. The
optimum amount of starting enzyme was determined by assaying 200 pmol (ca. 10 µg) TKS in
the presence of 250 to 12,500 pmol (ca. 230 µg) of OAC (Figure 4.8). The largest amount of OA
accumulation (538 pmol) during these assays was observed when 200 pmol TKS was assayed in
the presence of 7,500 pmol (ca. 140 µg) OAC. OAC activity appears to level off when amounts
greater than 7,500 pmol of OAC was assayed. The accumulation of OA is inversely proportional
to the accumulation of olivetol, and has a correlation coefficient of -0.978 (P value of 0.00001).
10000
HTAL
PDAL
OA
olivetol
amount (pmol)
5000
1000
800
600
400
200
0
12
50
0
10
00
75
00
50
00
25
00
10
00
50
0
20
0
0
OAC (pmol)
Figure 4.8 TKS assays conducted with variable amounts of OAC.
Product accumulation during TKS/OAC assays was affected by the concentration of OAC.
Constant amounts of 200 pmol TKS were assayed in the presence of 250 to 12,500 pmol OAC.
Error bars are the SD where n=3. Vertical axis is split to show smaller amounts of products.
104
4.3.5 OAC Activity as a Function of Temperature
These experiments were conducted to examine the effect temperature has on the TKS/OAC
reaction and to determine the optimum temperature at which OA production occurs. Temperature
influenced both the product ratios and total amounts of products from TKS/OAC assays. Results
indicate that OA production increases at lower (10 to 20°C) temperatures, while HTAL, PDAL
and olivetol production decreased (Figure 4.9). Although previous TKS assays indicated an
increase in product formation at (ca.) 40°C (Figure 3.9), the largest amount of OA detected in
TKS/OAC assays occurred at 15°C and was not significant at 10°C (z for 95% CI= 1.96).
A
amount (pmol)
8000
HTAL
PDAL
OA
olivetol
2000
600
400
200
30
25
20
15
10
0
temperature (°C)
B
triketides
tetraketides
amount (pmol)
10000
8000
6000
4000
2000
30
25
20
15
10
0
temperature (°C)
Figure 4.9 The activity of TKS/OAC as a function of temperature.
(A) As temperature is increased, so is the production of HTAL, PDAL, and olivetol. Conversely,
OAC activity increases at lower (10 to 20°C) temperatures. Vertical axis is split to show smaller
amounts of products. (B) The total amount of products is increased at temperatures of 20 to 30°C
compared to lower temperatures. Error bars are the SD where n=3.
105
4.3.6 OAC Activity in Different Buffer Systems
The activity of both TKS and OAC was strongly influenced by the buffer in which assays were
conducted. Assays were performed in triplicate using (final) concentrations of 50 mM HEPES,
TRIS, phosphate, or PIPES at pH 7.0, and product accumulation was compared (Figure 4.10).
HEPES appears a better buffer than TRIS, phosphate, or PIPES, and was capable of reducing the
amount of PDAL production while increasing overall tetraketide (HTAL, OA, and olivetol)
production. Assays conducted in the presence of HEPES buffer compared to other buffering
systems accumulated more than two times the amount of OA (1,100 pmol), and more than three
times the amounts of HTAL (569 pmol) and olivetol (1,061 pmol).
amount (pmol)
3500
HTAL
PDAL
OA
olivetol
2000
1000
500
PI
PE
S
TR
IS
ph
os
ph
at
e
H
EP
E
S
0
Figure 4.10 TKS/OAC assays conducted in different buffers.
When TKS and OAC were assayed in HEPES, an increase in tetraketide (HTAL, olivetol, OA)
production was observed compared to assays conducted in TRIS, phosphate or PIPES buffer. All
assays were conducted in the presence of 50 mM buffer [pH 7.0]. Vertical axis is split to show
smaller amounts of products. Error bars are the SD where n=3.
106
4.3.6.1 OAC Activity in Varying Concentration of HEPES pH 7.0
Product accumulation from TKS/OAC assays was also investigated using different molarities of
HEPES buffer, where final concentration of HEPES at pH 7.0 were 12.5, 25, 50, and 100 mM
(Figure 4.11A). This experiment was conducted to determine if HEPES concentration affected
OA production. The largest amounts of tetraketides were detected in assays that used 12.5 mM
HEPES including HTAL (573 pmol), OA (621 pmol), and olivetol (574 pmol). Total polyketide
production was unaffected by the HEPES concentrations used (Figure 4.11B)
amount (pmol)
A
8000
HTAL
PDAL
OA
olivetol
4000
1000
800
600
400
200
10
0
50
12
.5
25
0
HEPES (mM)
amount (pmol)
B
10000
triketides
tetraketides
8000
6000
4000
2000
10
0
50
25
12
.5
0
HEPES (mM)
Figure 4.11 Activity of TKS/OAC in variable concentrations of HEPES.
(A) A comparison of product accumulation during TKS and OAC assays conducted in the
presence of 12.5 – 100 mM HEPES buffer at pH 7.0. (B) The total amount of PDAL (triketide) is
compared to the sum of tetraketides (HTAL, OA, and olivetol). Vertical axis is split to show
smaller amounts of products. Error bars are the SD where n=3.
107
4.3.7 OAC Activity as a Function of pH
The accumulation of TKS/OAC assay products was monitored over a pH range in a variety of
buffers including HEPES, phosphate, ADA, and citrate. Given that a marked increase in
tetraketide production was observed in assays conducted in HEPES buffer, product accumulation
at different pHs was analyzed using this buffer. Unfortunately, the effective pH range of HEPES
is too narrow (6.8 to 8.2) to accomodate for more acidic environments. Monitoring of reaction
products at pH lower than 6.8 was later achieved using ADA and phosphate as buffering agents.
Taura et al. (2009) used citrate [pH 5.5] as a buffer for OLS, and this led to the further analysis
of TKS/OAC products in this buffer, which was directly compared to TKS/OAC assays
conducted in HEPES and MOPSO buffers.
108
4.3.7.1 OAC Activity in HEPES pH 6.8 to 8.2
Assays conducted in HEPES buffers within its full pH range indicate that maximal production of
OA, as well as HTAL and olivetol occurs at a lower pH than HEPES could accommodate
(Figure 4.12). The largest amount of OA (704 pmol) was detected in assays conducted at pH 6.8.
amount (pmol)
10000
HTAL
PDAL
OA
olivetol
1000
1000
800
600
400
200
8.
2
8.
0
7.
8
7.
6
7.
4
7.
2
7.
0
6.
8
0
pH
Figure 4.12 Activity of TKS/OAC in HEPES buffer pH 6.8 to 8.2.
A pH profile using 50 mM HEPES was developed by assaying 100 pmol TKS and 1,000 pmol
OAC over the full buffering range of HEPES (6.8 to 8.2). Vertical axis is split to show smaller
amounts of products. Error bars are the SD where n=3.
109
4.3.7.2 OAC Activity in Phosphate Buffer pH 6.0 to 8.0
Although it was previously shown that phosphate decreases TKS and OAC activity (Figure
4.10), the buffering range of phosphate allowed for a broader examination of pH and its effect on
activity (Figure 4.13). These assays suggest an optimum pH 6.5 for the production of OA (442
pmol) and HTAL (173 pmol), pH 7.5 for the production of PDAL (2650 pmol), and pH 6.0 for
olivetol (704 pmol). The lower pH at which olivetol production is increased suggests that
olivetol production could be augmented at an even lower pH.
3000
amount (pmol)
2000
HTAL
PDAL
OA
olivetol
1000
500
8.
0
7.
5
7.
0
6.
5
6.
0
0
pH
Figure 4.13 Activity of TKS/OAC in phosphate buffer pH 6.0 to 8.0.
A pH profile using 50 mM phosphate was obtained by assaying 100 pmol TKS and 1,000 pmol
OAC in the presence of malonyl-CoA and hexanoyl-CoA over a broad pH range (6.0 to 8.0).
Vertical axis is split to show smaller amounts of products. Error bars are the SD where n=3.
110
4.3.7.3 in ADA Buffer pH 5.7 to 7.8
The activity of TKS and OAC were also examined using ADA buffer at pH intervals of 0.2
between the range of 5.7 to 7.8 (Figure 4.14). This pH profile is the most comprehensive of all
pH curves obtained for TKS/OAC assays and indicates a similar trend as the other pH profiles.
Maximal HTAL (153 pmol) and OA (550 pmol) production occurred at pH 6.4, whereas olivetol
(83 pmol) accumulation was highest at pH 6.0, and that of PDAL (pmol 1,361) was at pH 7.2.
These ADA pH profiles were obtained during experiments conducted in parallel with those
presented in Section 3.2.2.1, where TKS was assayed alone.
HTAL
PDAL
OA
olivetol
amount (pmol)
1500
1000
500
7.
8
7.
6
7.
4
7.
2
7.
0
6.
8
6.
6
6.
4
6.
2
6.
0
5.
7
0
pH
Figure 4.14 Activity of TKS/OAC in ADA buffer pH 5.7 to 7.8.
A pH profile using 50 mM ADA was obtained by assaying 100 pmol TKS and 1,000 pmol OAC
in the presence of malonyl-CoA and hexanoyl-CoA over a broad pH range (5.7 to 7.8). Vertical
axis is split to show smaller amounts of products. Error bars are the SD where n=3.
111
4.3.7.4 in Citrate Buffer pH 5.0 to 6.0
TKS and OAC were assayed in the presence of 20 mM citrate at pH 5.0, 5.25, 5.5, 5.75, and 6.0
and product accumulation was directly compared to assays conducted in 20 mM HEPES at pH
7.0 and in 20 mM MOPSO at pH 7.0 (Figure 4.15). The largest amount of OA was detected from
assays performed in MOPSO at pH 7.0, where OA accumulated to 754 pmol. The second largest
OA accumulation was from assays conducted in HEPES at pH 7.0 where OA accumulated to 605
pmol. The largest amounts of olivetol (1,508 pmol) and HTAL (523 pmol) were detected in
assays conducted in citrate buffer at pH 5.75, although assays conducted in citrate generated the
amount (pmol)
least amount of OA (456 pmol), which was maximally produced in citrate at pH 6.0.
HTAL
PDAL
OA
olivetol
8000
5000
2000
1800
1500
1200
900
600
300
6.
00
ci
tr
at
e
5.
75
ci
tr
at
e
5.
5
ci
tr
at
e
5.
25
ci
tr
at
e
5.
00
ci
tr
at
e
7.
PS
O
M
O
H
EP
E
S
7.
00
00
0
Figure 4.15 Activity of TKS/OAC in citrate buffer pH 5.0 to 6.0.
A pH curve using 20 mM citrate was obtained by assaying 100 pmol TKS and 1,000 pmol OAC
in a pH range of 5.0 to 6.0 and is directly compared to assays conducted in HEPES [pH 7.0] and
MOPSO [pH 7.0]. Vertical axis is split to show smaller amounts of products. Error bars are the
SD where n=3.
112
4.3.8 OAC Activity in the Presence of Metals
These experiments were conducted to determine if metals had an effect on TKS and OAC
activity. The effect that metals have on OAC activity was determined by assaying the enzymes in
the presence of Mg2+, Ca2+, Mn2+, Fe2+, Co2+, Cu2+ and Zn2+ (Figure 4.16). Whereas alkaline
earth metals had a relatively small effect on TKS/OAC activity, the transition metals strongly
affected the product profile of the TKS/OAC reaction. Notably, Fe2+, Co2+, Zn2+ and Cu2+
reduced the amount of HTAL, PDAL, and OA to an average of 38%, 20%, and 23%,
respectively when compared to TKS/OAC reactions without metals. These same metals also
increased the amount of olivetol by an average of 84% when compared to the amount of olivetol
produced in TKS/OAC reactions without metals. The largest amount of OA was observed in the
absence of metals.
HTAL
PDAL
OA
olivetol
amount (pmol)
10000
3000
3000
2500
2000
1500
1000
500
l2
Zn
C
l2
C
uC
l2
C
oC
2l
Fe
C
l2
M
nC
2l
aC
C
l2
gC
M
no
m
et
al
0
Figure 4.16 The effect of metal on the TKS/OAC reaction.
This figure compares product accumulation in TKS/OAC reactions conducted in the presence
and absence of selected metals. Vertical axis is split to show smaller amounts of products. Error
bars are the SD where n=3.
113
4.3.8.1 TKS/OAC Activity in the Presence of EDTA
In addition to surveying OAC activity in the presence of metals, TKS and TKS/OAC were also
assayed in the presence and absence of 10 mM EDTA, a hexadentate chelating agent capable of
sequestering metal ions (Figure 4.17). These experiments were conducted to examine TKS/OAC
activity in the absence of metals. The addition of EDTA to TKS and TKS/OAC assays increased
the production of PDAL by about 30%, whereas the production of HTAL and olivetol dropped
by 40% and 25%, respectively, when compared to assays performed in the absence of EDTA.
The production of OA also diminished in TKS/OAC assays conducted in the presence of EDTA,
but only by 8% when compared to similar assays performed in the absence of EDTA.
amount (pmol)
2000
TKS
TKS + EDTA
TKS/OAC
TKS/OAC + EDTA
1500
1000
500
0
HTAL
PDAL
OA
olivetol
Figure 4.17 EDTA affects the synthesis of TKS/OAC products.
A comparison of reaction products that accumulated during TKS and TKS/OAC assays in the
presence and absence of EDTA show a reduction in tetraketide production and an increase in
PDAL production in EDTA containing reactions. Error bars are the SD where n=3.
114
4.3.9 TKS/OAC Activity in the Presence of MCS
Malonyl-CoA synthetase (MCS) was used to produce malonyl-CoA in vitro alongside TKS/OAC
assays. These assays were performed to compare the activity of TKS and TKS/OAC in the
presence and absence of MCS. Although the inclusion of MCS in assays allowed for the
production of malonyl-CoA, the reaction required the addition of free CoA, ATP, MgCl2 and
sodium malonate. Assays which did not include MCS required malonyl-CoA, but not free CoA,
ATP, MgCl2 and sodium malonate. In experiments conducted in parallel, TKS assays that used
malonyl-CoA produced an average 32% more HTAL, PDAL and olivetol than TKS assays using
MCS (Figure 4.18). Similarly, TKS/OAC assays that used malonyl-CoA showed increases of
53% HTAL, 36% PDAL, 9% olivetol, and 40% OA when compared to TKS/OAC assays that
employed MCS.
TKS
TKS + MCS
TKS + OAC
TKS + OAC + MCS
amount (pmol)
4000
1000
1000
800
600
400
200
0
HTAL
PDAL
OA
olivetol
Figure 4.18 TKS/OAC assays conducted in presence and absence of MCS.
TKS and TKS/OAC assays were performed with and without MCS allowing for the comparison
of product accumulation. Vertical axis is split to show smaller amounts of products. Error bars
are the SD where n=3.
115
4.3.10 Substrate Utilization in the TKS/OAC Reaction
These experiments were conducted to determine if TKS assay products were suitable substrates
for OAC. Free CoA was added to the TKS/OAC reaction to determine if it influenced OA
production, and to determine if free CoA could be involved in the acquisition and use of HTAL
as a substrate by OAC. OAC was assayed without TKS in the presence of the polyketide
products themselves, which included HTAL, PDAL, and olivetol, to determine if OAC was
capable of using these compounds as substrates for the production of OA. Additionally, to
explore TKS promiscuity, enzyme assays were performed using starter molecules other than
hexanoyl-CoA, which included octanoyl-CoA, butyryl-CoA, and acetyl-CoA. TKS assays were
conducted with and without OAC using alternative starter compounds and product accumulation
was compared.
116
4.3.10.1 The Effect of Free CoA on TKS/OAC Activity
TKS/OAC assays were conducted in the presence and absence of free CoA and product
accumulation was compared (Figure 4.19). When the concentration of free CoA was increased,
PDAL production also appeared to increase, but the amount of HTAL, olivetol and OA was
diminished. The maximal amount of OA (159 pmol) was produced by TKS/OAC assayed
without CoA, whereas the smallest amount of OA (23 pmol) was recovered when TKS/OAC was
assayed in the presence of 0.75 mM CoA, resulting in an 85% reduction in OA production.
amount (pmol)
4000
HTAL
PDAL
OA
olivetol
2000
500
400
300
200
100
1.
0
0.
75
0.
5
0.
25
0
0
Free CoA (mM)
Figure 4.19 TKS/OAC product accumulation in the presence of free CoA.
An increase in PDAL and a decrease in HTAL, OA, and olivetol were detected when TKS/OAC
was assayed in the presence of free CoA, when compared to assays conducted in the absence of
free CoA. Vertical axis is split to show smaller amounts of products. Error bars are the SD where
n=3.
117
4.3.10.2 TKS Products as Substrates for OAC
The TKS/OAC reaction clearly leads to the production of OA, but the question as to which
substrate OAC cyclizes remains unanswered. The unidentified substrate could be a linear
tetraketide that spontaneously cyclizes to either HTAL or olivetol in solution, or as a release
mechanism from TKS, but studies that have tried to produce reactive triketo acids have failed to
report the production of lactones from the cyclization of free acids (Harris and Carney, 1966;
Harris and Carney, 1967; Staunton and Weissman, 2001). The existence of linear polyketide
intermediates has also been questioned during substrate channeling between CHS and
downstream enzymes, such as chalcone reductases (CHRs); whether or not these substrates are
enzyme-bound or exist as free intermediates remains unknown (Austin and Noel, 2003). To
address the issue of OAC substrate utilization, OAC was assayed in the presence of isolated TKS
products (HTAL, PDAL, and olivetol), which could act as substrates if OAC possesses ringopening activity. Additionally, OAC was assayed with 0.5 mM HTAL and either 1) 0.5 mM free
CoA, or 2) 200 pmol TKS. The addition of free CoA to these assays was investigated since ringopening activity could require a cofactor which was present when TKS products are generated in
vitro. The addition of TKS to assays allowed the investigation as to whether or not TKS could
itself be responsible for the potential ring-opening of HTAL, resulting in the production of
triketo acid intermediates. When OAC was assayed with either PDAL or olivetol, no other
compound could be recovered (traces not shown). OAC assays performed in the presence of 1.0
mM HTAL resulted in the detection of trace amounts of OA during LCMS analysis (Figure
4.20). The addition of free CoA or TKS to OAC assays did not lead to an increase of OA.
118
absorbance
absorptionatat270
270nm
nm
A
0.24
285.8
O
0.19
285.8
O
0.14
O
HO
HTAL
0.09
0.04
250 300
250.00
0
1
2
3
4
5
6
7
time (min)
8
9
10
350 nm
300.00
11
350.00
12
B
absorption at
at 270
270nm
nm
absorbance
222.1
OH
0.005
0.004
261.0
261.0
OH
0.003
300.0
HO
300.0
368.9
OA
0.002
0.001
250.00
250
0.000
0
1
2
3
C 2.0E+06
ion count
O
4
5
6
7
time (min)
8
9
10
300.00
350.00
300 350 nm
11
12
11
12
HTAL
1.5E+06
1.0E+06
OA
5.0E+05
0.0E+00
0
1
2
3
4
5
6
7
time (min)
8
9
10
13
Figure 4.20 LCMS of OAC assays conducted with HTAL.
The LCMS analysis of products generated when OAC was assayed with 0.5 mM HTAL. (A)
PDA trace clearly indicates the elution of HTAL at 5.2 min, although no OA is observable at this
scale. The UV spectrum for HTAL is shown to the right. (B) Trace amounts of OA are detectable
at a UV absorbance scale of 0.005 at 8.9 min. The UV spectrum for OA is shown to the right. (C)
Analysis by SIR at m/z of 224.95 in ES+ mode was capable of detecting HTAL at 5.2 min, as
well as trace amounts of OA at 8.9 min.
119
4.3.10.3 With Butyryl-CoA
TKS was assayed in the presence and absence of OAC using butyryl-CoA as a starter molecule.
These assays employed MCS and its necessary substrates instead of using malonyl-CoA (see
Section 4.2.9). Reaction products were analyzed by LCMS with selected ion monitoring in ES+
mode at m/z values of 153.19 corresponding to 5-propylbenzene-1,3-diol (divarin; an olivetol
propyl analog), 155.16 corresponding to 4-hydroxy-6-propyl-pyran-2-one (butyryl diacetic acid
lactone (BDAL); a PDAL analog), and 197.20 corresponding to both 2-hydroxy-6-(2oxopentyl)pyran-4-one (butyryl triacetic acid lactone (BTAL); an HTAL analog), and 2,4dihydroxy-6-propylbenzoic acid (divarinic acid; an OA analog) (Figure 4.21). When TKS was
assayed in the absence of OAC, compounds were detected at 7.4 min (BDAL), 9.0 min (BTAL),
and at 16.9 min (divarin). When TKS and OAC were assayed together, in addition to those peaks
previously mentioned, another peak at 18.0 min was detected which corresponded exactly with
SIR at 197.20 m/z (divarinic acid). An unknown compound eluting at 16.1 min was also
detected by PDA and SIR at m/z 153.19; this peak is not seen when TKS was assayed alone.
4.3.10.4 With Octanoyl-CoA
TKS was assayed in the presence and absence of OAC using octanoyl-CoA as a starter molecule.
These assays also used MCS and its required components instead of malonyl-CoA (See Section
4.2.9). Reaction products were analyzed by LCMS using selected ion monitoring in ES + mode at
m/z values of 209.30 corresponding to 5-heptylbenzene-1,3-diol (heptylresorcinol; an olivetol
heptyl analog), 211.27 corresponding to 6-heptyl-4-hydroxypyran-2-one (heptyl diacetic acid
lactone (HDAL); a PDAL analog), and 253.31 corresponding to both 4-hydroxy-6-(2oxononyl)pyran-2-one (octanoyl triacetic acid lactone (OTAL); an HTAL analog), and 2-heptyl120
4,6-dihydroxybenzoic acid (heptylresorcinolic acid; an OA heptyl analog) (Figure 4.22). When
TKS was assayed in the absence of OAC, compounds were detected at 6.7 min (HDAL), 7.8 min
(OTAL), 9.5 min (unknown), and at 10.6 min (heptylresorcinol). The unknown compound
observed during PDA analysis eluting at 9.5 min is not detectable by SIR at 253.31; this
compound is not heptylresorcinolic acid. The same compounds were detected when TKS and
OAC were assayed, in addition to another compound, seen as a larger peak eluting at 9.4 min
during PDA analysis; this compound (heptylresorcinolic acid) is clearly detectable by SIR at m/z
253.31 (Figure 4.22).
4.3.10.5 With Acetyl-CoA
TKS was assayed in the presence and absence of OAC using acetyl-CoA as a starter molecule.
These assays used MCS and its required components instead of malonyl-CoA (Section 4.2.9).
Reaction products were analyzed by LCMS using selected ion monitoring in ES + mode at m/z
values of 125.14 corresponding to 5-methyl-1,3-benzenediol (orcinol; an olivetol methyl analog),
127.11 corresponding to 6-methyl-4-hydroxypyran-2-one (triacetic acid lactone (TAL); a PDAL
analog), and 169.15 corresponding to both 2-acetonyl-6-hydroxy-pyran-4-one (tetra acetic acid
lactone (TTAL); an HTAL analog), and 2,4-dihydroxy-6-methylbenzoic acid (orsellinic acid; an
OA methyl analog) (Figure 4.23). Whether TKS was assayed with or without OAC, the assumed
compounds were detected at 4.0 min (TAL), 4.8 min (TTAL), 8.1 min (orcinol), and at 10.1 min
(orsellinic acid). An unknown compound was also observed during PDA analysis at 270 nm
eluting at 3.5 min, which is also detectable by SIR at m/z 169.15 (Figure 4.23). There is little
difference between assays conducted with or without OAC suggesting that shorter alkyl
tetraketides may spontaneously cyclize to orsellinic acid.
121
absorbance at 270 nm
A
4.00E-02
TKS
PDA 270 nm
3.00E-02
ion count
divarinic acid
18.0 min
divarin
16.9 min
BDAL
9.0 min
BTAL
7.4 min
2.00E-02
?
16.1 min
1.00E-02
0.00E+00
0
B
TKS + OAC
2
4
6
8
10
12
14
16
18
20
22
24
26
3.00E+05
28
ES+ 153.19
OH
2.00E+05
?
HO
1.00E+05
CH3
divarin
0.00E+00
0
ion count
C
2
4
6
8
10
12
14
16
18
20
22
24
26
8.00E+07
28
30
ES+ 155.16
6.00E+07
O
4.00E+07
O
2.00E+07
HO
CH3
BDAL
0.00E+00
0
D
2
4
6
8
10
12
14
16
18
20
24
26
6.00E+06
28
30
ES+ 197.20
O
ion count
22
OH
4.00E+06
O
2.00E+06
O
O
OH
HO
CH3
BTAL
HO
CH3
divarinic acid
0.00E+00
0
2
4
6
8
10
12
14
16
time (min)
18
20
22
24
26
28
30
Figure 4.21 LCMS of TKS/OAC assays conducted with butyryl-CoA.
LCMS chromatograms from TKS assays performed in the presence (red) and absence (black) of
OAC are overlaid. (A) LC traces align well with selected ion monitoring in ES+ mode capable of
detecting (B) divarin (16.9 min; m/z 153.19), (C) BDAL (9.0 min; m/z 155.16), (D) BTAL (7.4
min; m/z 197.20) and divarinic acid (18.0 min; m/z 197.20). An unknown compound was also
detected using PDA monitoring at 16.1 min (A) and SIR monitoring at m/z of 153.19 (B). No
divarinic acid was detected when TKS was assayed in the absence of OAC.
122
2.00E-02
absorbance at 270 nm
A
TKS
OTAL
7.8 min
1.50E-02
PDA 270
heptylresorcinolic acid
9.4 min
1.00E-02
HDAL
6.7 min
5.00E-03
heptylresorcinol
10.6 min
0.00E+00
0
2
4
6
8
10
12
2.50E+05
B
TKS + OAC
14
16
18
20
OH
22
24
ES+ 209.30
ion count
1.50E+05
HO
5.00E+04
-5.00E+04 0
ion count
C
2
4
6
8
11
8.00E+07
13
CH3
heptylresorcinol
15
17
19
21
ES+ 211.27
O
6.00E+07
23
O
4.00E+07
HO
CH3
HDAL
2.00E+07
0.00E+00
0
ion count
D
2
4
6
8
2.00E+07
11
13
15
17
19
21
23
O
ES+ 253.31
O
1.50E+07
HO
1.00E+07
O
OTAL
OH
CH3
O
OH
HO
5.00E+06
heptylresorcinolic acid
CH3
0.00E+00
0
2
4
6
8
11
13
15
time (min)
17
19
21
23
Figure 4.22 LCMS of TKS/OAC assays conducted with octanoyl-CoA.
LCMS chromatograms from TKS assays performed in the presence (red) and absence (black) of
OAC are overlaid. (A) LC traces align well with selected ion monitoring in ES+ mode capable of
detecting (B) heptylresorcinol (10.6 min; m/z 209.30), (C) HDAL (7.8 min; m/z 211.27), (D)
OTAL (6.7 min; m/z 253.31) and heptylresorcinolic acid (9.4 min; m/z 253.31). The detection of
heptylresorcinolic acid is confirmed by its appearance at 9.4 min during PDA analysis (A) and its
corresponding peak during SIR at m/z 253.31 (D), which was not detected when TKS was
assayed in the absence of OAC.
123
5.00E-02
A
TKS
PDA 270 nm
TTAL
4.8 min
4.00E-02
absorbance at 270 nm
TKS + OAC
3.00E-02
TAL
4.0 min
2.00E-02
orsellinic acid
10.1 min
orcinol
8.1 min
1.00E-02
0.00E+00
0
ion count
B
1
2
3
4
5
1.5E+05
6
7
8
9
10
11
12
13
14
ES+ 125.14
OH
1.0E+05
HO
CH3
orcinol
5.0E+04
0.0E+00
0
ion count
C
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1.5E+06
ES+
O
1.0E+06
127.11
O
HO
TAL
5.0E+05
CH3
0.0E+00
0
D
1
2
3
4
5
6
7
8
9
10
11
12
13
ES+ 169.15
O
1.0E+06
14
OH
O
O
O
ion count
OH
CH3
HO
TTAL
5.0E+05
CH3
HO
orsellinic acid
0.0E+00
0
1
2
3
4
5
6
7
8
time (min)
9
10
11
12
13
14
Figure 4.23 LCMS of TKS/OAC assays conducted with acetyl-CoA.
LCMS chromatograms from TKS assays performed in the presence (red) and absence (black) of
OAC. (A) LC and MS ES+ results are aligned and capable of detecting what is assumed to be (B)
orcinol (8.1 min; m/z 125.14), (C) TAL (4.0 min; m/z 127.11), (D) TTAL (4.8 min; m/z 169.15)
and orsellinic acid (10.1 min; m/z 169.15), which were detected in all assays performed.
124
4.4 Discussion
TKS assays conducted in the presence of candidate polyketide cyclases led to the discovery of
OAC, an enzyme capable of modifying TKS product formation resulting in the synthesis of OA
(Gagne et al., 2012). This discovery warranted the biochemical characterization of OAC, the
main focus of this chapter.
4.4.1 Phylogenetic Relationship of OAC
Blast searches for OAC homologs indicate that similar proteins are ubiquitous in plants. Their
strong conservation indicates a beneficial property in plants, probably related to salt and heat
stress responses, as their family name suggests (i.e. SRABBs). The occurrence of similar
sequences to OAC in bacterial genomes (namely Rhizobium) is an indicator that they may
originate from ancestral prokaryotes, and may have been introduced to plants by early horizontal
gene transfer events, as it has been hypothesized to occur in Nicotiana (Intrieri and Buiatti,
2001). Phylogenetic analysis divides the OAC homologs into two specific clades suggesting that
a primary OAC homolog (clade A) is conserved, and may have duplicated so to form a clade of
paralogs (clade B). The relationship OAC shares with these clades A and B appears remote,
having formed its own branch with clade A, but includes homologs from Mimulus guttatus
(common monkey-flower), and Populus trichocharpa (poplar).
4.4.2 The Optimum TKS/OAC Ratio
Ratio experiments were designed to determine the amount of OAC required for maximal OA
production. The amount of enzyme used in these assays reported values in pmol instead of µg
values to determine if a stoichiometric relationship existed between TKS and OAC; results
suggest that no such relationship exists and that maximal OA production occurred when
125
OAC:TKS ratios were approximately 40:1. The largest amount of OA production was observed
when 140 µg of OAC (i.e. 7,500 pmol) was assayed with 9 µg of TKS (200 pmol). Larger
amounts (>7,500 pmol) of OAC did not increase OA production, indicating that peak OAC
activity had been achieved, or that self-inhibition of OAC (and/or possibly TKS inhibition) was
occurring due to its increased concentration. Based on its structural classification, it is presumed
that OAC exists as a dimer in solution and that at greater concentrations a higher ordered
oligomer could exist, as it does with the OAC homolog, Populus tremula SP1, which is capable
of forming dodecameric quaternary structures (Dgany et al., 2004). This potential
oligomerization could lead to structural changes that result in self-inhibition, which is known to
occur in other enzymes such as 12-oxophytodienoate reductase from Lycopersicon esculentum
(tomato), which exhibit a weak tendency for dimerization correlating with a loss of activity
(Breithaupt et al., 2006). Results themselves do not suggest a loss of activity, and may instead
indicate a maximum limit of products being reached, and that the effects seen when >7500 pmol
OAC were used is simply a leveling-off of OA production. Results suggest that 100 pmol of
TKS should be assayed with 3,500 pmol of OAC for optimum OA production, but further
analyses were conducted using a TKS:OAC ratio of 100:1,000 pmol due to the huge amounts of
recombinant OAC that would otherwise be required (where 1,000 pmol of OAC is approximately
19 µg of enzyme).
4.4.3 The Optimum Temperature for In Vitro OA Production
Assays designed to determine the optimum temperature for OA production gave insights about
the TKS/OAC reaction. The results from TKS assays conducted in the absence of OAC indicate
that maximal product formation occurred at ca. 40°C. When TKS was assayed in the presence of
126
OAC, the largest amount of OA was produced at a lower temperature of 15°C, and OA
production was reduced at temperatures exceeding 25°C. These results suggest that an unstable
intermediate is produced by TKS, which undergoes a spontaneous decarboxylative cyclization at
higher temperatures resulting in the formation of olivetol. Conversely, at lower temperatures the
tetraketide intermediate is stabilized, allowing it to reach OAC, where it is cyclized to OA before
any thermo-dependent spontaneous cyclization occurs. Based on these data, future assays
optimized for OA production were conducted at 20°C.
4.4.4 The Optimum Buffer and pH for OAC
The choice of buffering agents also played a key role in determining the optimum for OA
production. Results from combinatorial assays performed in the presence of sodium phosphate,
PIPES, HEPES, TRIS, and citrate were directly compared and results indicate that HEPES is the
best buffering system for OA production. Furthermore, a lower concentration of HEPES was
deemed beneficial for tetraketide production, and a 20 mM HEPES buffering solution was used
in subsequent assays. The activity of OAC was also compared over a wide pH range using a
number of buffers due to the limited effective pH range of HEPES. Combined results indicate
that maximal OAC activity occurs at a pH of 6.4, but this pH is not possible with HEPES.
Nevertheless, the amounts of OA detected from assays conducted in HEPES buffer at pH 7.0
were greater than those amounting when assays were conducted in phosphate, TRIS, or PIPES at
pH 6.4. It was decided that future assays would use HEPES at pH 7.0. At a later date, citrate and
MOPSO buffer systems were also compared to HEPES, and showed that MOPSO was as good
as HEPES as a buffering agent, if not superior. In retrospect, the use of MOPSO would have
been beneficial during all work that continued, since its pH range is 6.2 to 7.6, and an optimum
127
pH of 6.4 could have been achieved. However, the MOPSO/HEPES buffer comparison was
tested after many assays had already been performed, and in the interest of time and resources,
the use of HEPES as a buffer continued throughout this study. The cause for the differences in
TKS and OAC activity in the various buffer systems is not known.
4.4.5 OAC Does Not Require a Metal for Activity
The TKS/OAC reaction was assayed in the presence of metals to determine if they would affect
TKS and OAC activity, potentially leading to an augmentation in OA production. Additionally,
the possibility that OAC employs a type II aldol mechanism to achieve OA cyclization, which
would require the participation of a metal ion, was also explored (Silverman, 2002). It was also
reasonable to suspect that OAC required a metal, since Mg2+ has been identified in the AtHS1
crystal structure (Bingman et al., 2004). Results clearly indicate that no metal tested could
increase OAC activity. The effect that some transition metals have on the TKS/OAC product
profile is difficult to explain without further experimentation, but the observation that Fe2+, Co2+,
Zn2+ and Cu2+ affect polyketide chemoselectivity is intriguing enough to warrant a directed study
in the future. It is easy enough to explain a loss of activity when OAC is assayed in the presence
of these metals, but the amplification of olivetol production to levels that are three times the
amounts observed in assays conducted in the absence of metals, coupled to a reduction of PDAL
derailment products suggests a possible route to stabilizing TKS, and putatively other type III
PKS in vitro reactions. These findings indicate that there is no metal requirement for OAC.
Future experiments that would examine this chemoselectivity phenomenon should include a
broader spectrum of metals to test and vary the concentrations of these metals when being
assayed with TKS.
128
In support of the notion that neither TKS nor OAC requires any metal for their activity
are the results from assays conducted in the presence of EDTA. The results showed that,
although TKS product accumulation is decreased when TKS and OAC are assayed with EDTA,
enzyme activity is not abolished, and OA accumulation is hardly affected (Figure 4.17). The
findings discount the metal-catalyzed aldol condensation as a plausible mechanism for OAC.
Likewise, no other bacterial cyclase involved with type II PKSs has been reported to require a
metal (Schneider, 2005).
The inclusion of EDTA also caused an increase in PDAL production, suggesting EDTA
destabilizes the TKS reaction by causing the premature release of nascent polyketide
intermediates, although notably, OAC activity remained unaffected. The effects caused by
EDTA and metals may be due to direct interaction on the enzymes themselves, but may also
have been caused by their reactivity to a yet unidentified labile polyketide intermediate that
remains in solution until it cyclizes.
4.4.6 The use of MCS in assays abolishes the need for malonyl-CoA
Assays that utilized MCS bypassed the requirement of malonyl-CoA in TKS/OAC assays, but
required the inclusion of sodium malonate, ATP, Mg2+, and free CoA instead. These assays were
primarily conducted to determine if MCS could augment OA production during TKS/OAC
assays. Assays performed in the presence of MCS were capable of successfully generating
malonyl-CoA required by TKS, although a reduction in product accumulation was observed.
This reduction in overall TKS/OAC activity could be due to the addition of Mg2+ and/or free
CoA, which have each been shown to be destabilizing agents. It should be noted that the
optimization of the MCS reaction with TKS and OAC has not been conducted, and that future
129
work employing MCS could benefit if this were to take place, which would most likely lead to
an increase in product yield. Nevertheless, the inclusion of MCS in the TKS/OAC reaction was
successful in that these reactions could be conducted without exogenous malonyl-CoA. Assays
were also conducted with free CoA to determine how well this compound could be tolerated. As
previously mentioned, a decrease in the overall TKS/OAC reaction output is correlated to the
concentrations of free CoA in enzyme assays, but the concentration of free CoA used in later
analyses (0.20 mM CoA) did not interfere with the quantification of assay products.
4.4.7 Neither HTAL, PDAL, nor olivetol are substrates for OAC
The possibility that OAC acts directly on TKS products was explored by assaying OAC in the
presence of HTAL, PDAL, and olivetol, but no significant OA production was detected. Trace
amounts of OA were detectable when 1 mM HTAL was incubated with OAC, although this
small amount was almost negligible (20 pmol), since typically TKS reaction generates much
smaller amount of HTAL than the amounts used in these assays. If OAC was capable of opening
and re-cyclizing HTAL into OA, it would be anticipated that OA accumulation would be
comparable to a typical TKS/OAC reaction, where OA accumulated to approximately 500 pmol.
It is noted that OA was detected when OAC was assayed with 1 mM HTAL, but that the minute
amount may have been produced following spontaneous HTAL ring opening followed by
spontaneous or possibly OAC-catalyzed cyclization to form OA. The spontaneous opening and
closing of lactones which results in the formation of resorcylates has been previously reported
with starter-chain homologs (i.e. TTAL) (Guilford et al., 1968). I propose that a similar reaction
also occurs, although at a slower rate, with HTAL. The potential involvement of TKS during
HTAL ring opening was also assessed by including TKS during OAC assays performed in the
130
presence of HTAL, but no OA was detected, conclusively discounting this possibility. Similarly,
the putative involvement of free CoA as a cofactor was also investigated during OAC/HTAL
assays, but no OA was detected during LCMS analysis.
4.4.8 Polyketide Analogs are Cyclized to their Respective Resorcylic Acid by OAC
TKS/OAC assays were also performed using alternative CoA-esters besides hexanoyl-CoA,
which included butyryl-CoA, octanoyl-CoA, and acetyl-CoA. Although no standards were used,
the detection of polyketide analogs was made possible by comparing product UV spectra
produced in these assays with those from standard TKS/OAC assay products that used hexanoylCoA as a starter. Additionally, the use of LCMS SIR for masses of the theoretical products was
also capable of confirming the identities of the polyketide analogs produced (Figures 4.21, 4.22,
and 4.23). Acetyl-CoA, butyryl-CoA and octanoyl-CoA were each accepted as starter
compounds by TKS, and appeared to be successfully incorporated into triketide and tetraketide
products, which included PDAL, HTAL, and olivetol analogs. In addition to these analogs, TKS
assays conducted in the presence of OAC were also capable of producing divarinic acid and
heptylresorcylic acid, when butyryl-CoA and octanoyl-CoA were used as alternative starter
compounds, respectively. These results contradict previous reports that TKS cannot produce
HTAL analogs when either butyryl-CoA or octanoyl-CoA replaces hexanoyl-CoA as starter CoA
ester, and that TKS cannot produce heptylresorcylic acid when octanoyl-CoA is used (Taura et
al., 2009). The lack of polyketide analogs in assays conducted by others may be due to
differences in buffering system, the lower pH employed, or subtle differences between TKS and
OLS.
131
Findings also show that TKS is capable of using acetyl-CoA as a starter molecule, and
results in the production of TAL, TTAL, orcinol, and orsellinic acid (Figure 4.23). These assays
differ from assays using butyryl-CoA or octanoyl-CoA, in that TKS assays that use acetyl-CoA
are capable of accumulating a resorcylic acid analog (orsellinic acid in this case) whether they
are performed in the presence or absence of OAC. When butyryl-CoA or octanoyl-CoA were
used as priming agents, the accumulation of their resorcylic acid analogs was only produced
when OAC was present. This difference in the cyclization can be attributed to the variability in
length and bulkiness of the primer’s alkyl groups, and the propensity of the synthesized
tetraketide’s lactone ring to open and close in solution. The ability of TTAL to spontaneously
and easily reform into orcinol and orsellinic acid has been previously reported (Bentley and
Zwitkowits, 1967; Guilford et al., 1968). Guilford et al. (1968) have also demonstrated that
phenacylpyrone, a TTAL lactone analog attached to a bulky phenyl group, is a moderately stable
lactone
capable
of
withstanding
ring
opening
and
subsequent
conversion
to
its
recorcylate/resorcinol analog, whereas TTAL, which contains only a methyl group instead of a
phenyl ring, can open easily. Therefore, it is not surprising to discover orsellinic acid is a product
of TKS assays that use acetyl-CoA when conducted in the absence of OAC.
The ability of TKS and OAC to synthesize different resorcylic acids by using different
starting molecules is in agreement with the occurrence of cannabinoid analogs observed in
nature. These analogs replace the otherwise normal pentyl side chain of cannabinoids with
methyl (Vreet et al., 1972), propyl (Merkus, 1971), butyl (Harvey, 1976), and possibly even
heptyl side chains (Isbell, 1973). The biosynthesis of these cannabinoid analogs would proceed
through the prenylation of short-chain (1-8 carbon) resorcylic acids which would have been
132
generated by TKS and OAC. The plasticity observed in TKS is reminiscent of GCHS2, a nonCHS type III PKS from Gerbera hybrida also capable of accepting acetyl-CoA as a starter
molecule (Eckermann et al., 1998). Unlike the traditional CHSs that synthesize naringenin
chalcone from phenylpropanoid CoA esters, GCHS2 is a pyrone synthase that synthesizes TAL
after two condensations with acetyl-CoA. GCHS2 is also capable of using only malonyl-CoA as
a primer following decarboxylation to acetyl-CoA. Similar activity was not observed with TKS.
133
5. ANALYSIS OF PHYSICAL INTERACTION BETWEEN TKS AND OAC
5.1 Introduction
This chapter focuses on experiments designed to examine the physical relationship between TKS
and OAC, which may also provide insight into the mechanism of OA biosynthesis. Despite the
biochemical evidence presented in Chapter 4, showing that TKS and OAC function together to
produce OA, fundamental questions about to the mechanism of OA biosynthesis still need to be
addressed. One issue was whether these enzymes exist in the same cellular compartment, for if
they do not, it would be difficult to suggest that they participate together in an enzymatic
reaction. The possibility that TKS and OAC physically interact is reasonable, given that OAC
resembles a type II cyclase, and that type II polyketide biosynthesis is accomplished by
multienzyme complexes (Staunton and Weissman, 2001). Multienzyme complexes (metabolons)
have also been shown to exist in plants, as in Calvin cycle complexes (Suss et al., 1993), serine
and cysteine biosynthesis enzyme complexes (Prabhu et al., 1996; Saito et al., 1995), and the
oxidative pentose phosphate metabolon (Debnam et al., 1997). Winkel-Shirley (1999) suggests
that metabolons frequently occur in the cell and constitute a structured organization of cellular
components instead of loosely associated soluble enzymes (Winkel-Shirley, 1999). This makes
sense, considering that cellular activities could be more tightly regulated in this way, as well as
increasing metabolic efficiency, increasing local substrate concentrations, and partitioning
metabolites to similar branching pathways. Earlier reports from Stafford (1974) also suggest that
enzyme complexes exist so to facilitate the biosynthesis of products that are localized in specific
subcellular
compartments
(Stafford,
1974).
134
Interestingly,
enzymes
from
both
the
phenylpropanoid and flavonoid biosynthesis pathways have been shown to interact. WinkelShirley’s (1999) review on enzyme complex association in the phenylpropanoid and flavonoid
biosynthesis pathways suggests that: 1) metabolites are channelled from one enzyme to another
as in the case of phenylalanine ammonia-lyase (PAL) and cinnamate 4-hydroxylase (C4H),
where 3H-phenylalanine is preferentially used as a substrate instead of
14
C-cinnamic acid in the
production of labelled p-coumaric acid (Czichi and Kindl, 1977); 2) enzymes are co-localized at
microsomal and endoplasmic reticulum sites, as is evident by cell fractionation experiments
conducted by Grisebach and Fritsch (1975), and similarly Hrazdina et al. (1978) who showed
that CHS pellets with the microsomal fraction of Tulipa and Hippeastrum tissues, and later by
Wagner and Hrazdina (1984), who demonstrated that PAL, C4H, and CHS all co-fractionated
with NADPH:Cyt c reductase, an endoplasmic reticulum marker; 3) enzymes involved in the
phenylpropanoid/flavonoid biosynthesis pathway are induced in a co-ordinate manner, which has
led to the classification of early and late genes associated with the synthesis of flavonoids, and
the later synthesis of proanthocyanidins, anthocyanidins and flavonols (Pelletier et al., 1997); 4)
direct protein-protein interactions do occur, as reported by Burbulis et al. (1999), who used a
yeast two-hybrid (Y2H) system to show direct interaction between A. thaliana CHS, chalcone
isomerase (CHI) and dihydroflavonol 4-reductase (D4R), and used polyclonal antibodies
generated against CHI to co-immunoprecipitate CHS and D4R.
To determine if TKS and OAC co-localize in vivo, the subcellular localization of
fluorescent tagged TKS and OAC was performed. This allowed the visualization of these
proteins in planta using confocal microscopy following agroinfiltration in Nicotiana
benthamiana leaves. Once the co-localization of the enzymes was confirmed, a Y2H assay was
135
performed to determine if TKS and OAC physically interact. If a physical interaction existed, it
would imply that TKS physically directs a transient intermediate to OAC, as in the case of PAL
and C4H (Czichi and Kindl, 1977), or possibly that TKS activity is modified by OAC through an
allosteric mechanism that alters the TKS active site.
Further experiments attempted to determine if TKS produces and subsequently releases a
polyketide intermediate, which is then bound and cyclized by OAC. TKS/OAC reactions were
conducted by exploiting the poly-histidine tag in recombinant TKS using magnetic Talon beads,
allowing TKS to be removed during enzyme assays and then observing whether or not OA could
be produced in the absence of TKS.
Similar assays, designed to assess the enzymes’ ability to function independently, were
performed using a microdialysis apparatus which separated TKS and OAC. These experiments
have provided important insights into the physical relationship between TKS and OAC, as well
as providing valuable evidence that supports the notion of a linear polyketide intermediate being
produced and released from TKS before it is acquired and cyclized by OAC.
136
5.2 Materials and Methods
5.2.1 Subcellular Localization of TKS and OAC
The TKS-cyan fluorescent fusion protein (TKS:CFP) and OAC-yellow fluorescent fusion protein
(OAC:YFP) constructs were generated by amplifying TKS and OAC without C-terminal stop
codons using Pfu Ultra II DNA Polymerase (Stratagene) and the primers and cycling conditions
shown in Tables 5.1 and 5.2. PCR products were A-tailed by incubation at 72°C for 15 min
along with 50 U of Taq polymerase (Invitrogen) and 0.10 mM dATP before purification using a
Minelute PCR kit (Qiagen). PCR products were ligated into the pCR8/GW/topo entry vector
(Invitrogen), resulting in the pSG054 and pSG055 plasmids harboring TKS and OAC without
stop codons, respectively, that were separately used to transform E. coli DH5α competent cells
(Invitrogen). Plasmids were isolated using a Qiaprep Spin Miniprep Kit (Qiagen) from DH5α
cultures after they were grown in LB in the presence of 50 mg/mL spectinomycin overnight at
37°C. Following sequence verification (PBI DNA Technologies Unit), pSG054 was subcloned
into the pEARLEY102 (Invitrogen), resulting in the TKS:CFP construct. Similarly, pSG055 was
subcloned into pEARLEY101 (Invitrogen), creating the OAC:YFP construct. Following LR
recombination, constructs were sequenced to confirm orientation using the primers described in
Table 5.2.
Expression constructs were transformed by electoporation in competent Agrobacterium
tumefaciens LBA4404 and plated onto selective LB in the presence of 10 µg/mL of rifampacin
and 50 µg/mL of kanamycin (Weigel and Glazebrook, 2006). Freshly cultured transformants
were grown to an OD600 of 0.2 and used to agroinfiltrate two-week old Nicotiana benthamiana
leaves (Sparkes et al., 2006). Following a two-day incubation period, plant tissue near
137
agroinfiltration sites were cut into 1 x 1 cm squares and placed in dH2O on a microscope slide
prior to being examined by confocal microscopy. A Zeiss LSM510 confocal microscope was
used to visualize the TKS:CFP construct after excitation with a diode laser set to 458 nm and to
capture images using a 475-525 nm bandpass filter. The OAC:YFP construct was visualized by
setting the diode laser to 514 nm and using a 530-600 nm bandpass filter. Images were processed
using the Zen software suite (Carl Zeiss Microscopy).
Table 5.1 Primers used during the creation of TKS:CFP and OAC:YFP
Name
Forward Primer
Reverse Primer
TKS:CFP
ATGAATCATCTTCGTGCTGAG
ATATTTGATGGGAACACTACG
OAC:YFP
ATGGCAGTGAAGCATTTGATT
CTACTTTCGTGGTGTGTAGTC
TKS:CFP sequencing
ATGAATCATCTTCGTGCTGAG
CTTTCGTGGTGTGTAGTC
OAC:YFP sequencing
ATGGCAGTGAAGCATTTGATT
CTTTCGTGGTGTGTAGTC
Table 5.2 Details regarding the creation TKS:CFP and OAC:YFP constructs
Melting1
Annealing1
Name
Template
TKS:CFP
pHIS8-TKS
95°C/20s
50°C/20s
OAC:YFP
pET100-OAC
98°C/20s
55°C/20s
Elongation1
Destination
Selection2
72°C/18s
pEARLEY102
KanR
72°C/18s
pEARLEY101
KanR
1 - The reactions were cycled 30-35 times before ending with 7 min at 72°C.
2 - Kanamycin resistance (KanR).
138
5.2.2 Yeast Two-Hybrid Methods
Experiments used a ProQuest Two-Hybrid System (Invitrogen). The system uses two specialized
Gateway plasmids, pDest32 and pDest22, for rapid cloning, expression and analysis of proteins
of interest. The pDest32 plasmid contains a GAL4 DNA-binding domain (DBD), one of two
domains derived from the GAL4 transcription factor, and is responsible for binding an upstream
operator sequence. Proteins of interest fused to this domain are collectively known as bait
protein. The activating domain (AD) is the other half of the GAL4 transcription factor and is
harbored by the pDest22 plasmid. Proteins of interest fused to this domain are referred to as prey
proteins. β-galactosidase assays of fusion proteins allowed for blue/white screening using 5bromo-4-chloro-3-indolyl- beta-D-galactopyranoside (X-gal) as a substrate. If an interaction
existed between bait and prey fusion proteins, X-gal would be cleaved, resulting in an insoluble
blue 5, 5'-dibromo-4,4'-dichloro-indigo which acts as an colorimetric indicator.
A number of plasmids were required for these experiments, necessitating the cloning of
TKS and OAC into the pDest32 and pDest22 vectors. Furthermore, a combination of plasmids
were used to transform the Saccharomyces cerevisiae strain MaV203, resulting in a collection of
different transformants (results Section 5.3.2). Plasmids used in this work are summarized in
Table 5.3.
139
Table 5.3 Plasmids used to transform MaV203 yeast used in Y2H experiments
Name
Function
Derived from
Selection1
pDEST32
Bait vector
n/a
LEU2 and GenR
pDEST22
Prey vector
n/a
TRP1 and AmpR
pEXP32/Krev1
Control bait
pDEST32
LEU2 and GenR
pEXP22/RalGDS-wt
Strong positive control
pDEST22
TRP1 and AmpR
pEXP22/RalGDS-m1
Intermediate positive control
pDEST22
TRP1 and AmpR
pEXP22/RalGDS-m2
Negative control
pDEST22
TRP1 and AmpR
pRT030
TKS (entry vector)
pCR8/GW/TOPO
SpcR
pSC033
OAC (entry vector)
pCR8/GW/TOPO
SpcR
pSG057
TKS in prey vector
pDEST22 + pRT030
TRP1 and AmpR
pSG058
TKS in bait vector
pDEST32 + pRT030
LEU2 and GenR
pSG059
OAC in prey vector
pDEST22 + pSC033
TRP1 and AmpR
pSG060
OAC in bait vector
pDEST32 + pSC033
LEU2 and GenR
1 – Antibiotic resistance to gentamycin (GenR), ampicillin (AmpR), or spectinomycin (SpcR).
5.2.3 Plasmid Construction
Both Gateway-compatible TKS (pSC033) and OAC (pRT030) were subcloned into each
pDEST22 and pDEST32 by LR recombination as directed by the manufacturer’s protocol.
Reagents were supplied as part of the ProQuest Two-Hybrid System (Invitrogen). The reactions
generated the pSG057, pSG058, pSG059, and pSG060 constructs (Table 5.3). Plasmids were
transformed into E. coli Mach 1 cells (Invitrogen) and, following their culturing, were isolated
using a QIAprep Spin Miniprep Kit (Qiagen) before being sequenced using the recommended
ProQuest kit primers (Invitrogen) (Table 5.4).
Table 5.4 Sequencing primers for pDEST22 and pDEST32 based vectors
Function
Sequence
pDEST32 forward sequencing primer (23mer)
AACCGAAGTGCGCCAAGTGTCTG
pDEST22 forward sequencing primer (21mer)
TATAACGCGTTTGGAATCACT
pDEST32/22 reverse sequencing primer (23mer)
AGCCGACAACCTTGATTGGAGAC
140
5.2.4 Yeast Transformants
Once sequences were confirmed, Saccharomyces cerevisiae MaV203 competent cells were
transformed with the generated plasmids (Table 5.3). MaV203 cells were made competent by
initially inoculating 10 mL of liquid YPAD media (1% Bacto-yeast extract, 2% Bacto-peptone,
2% glucose, 0.01% adenine sulfate [pH 6]) and culturing overnight at 30°C. The overnight
culture was diluted to an OD600 of 0.4 in 50 mL of YPAD and incubated for an additional 4 h.
The culture was centrifuged at 1,000 g and resuspended in 40 mL of sterile TE (10 mM TRISHCl [pH 7.5], 1 mM EDTA). Cells were again centrifuged at 1,000 g (supernatant discarded) and
resuspended in 2 mL of sterile 1 x LiAc/0.5x TE (100 mM LiAc, 5 mM TRIS-HCl [pH 7]). Cells
were incubated in this solution for 10 min before proceeding to transformations. A total of 1 µg
of DNA was used in each transformation, where 500 ng of DNA from each plasmid was used.
An example of this is SG100, where 1 µL of 500 ng/µ L pEXP32/Krev1 was used with 1 µL of
500 ng/µL pEXP22/RalGDS-wt resulting in 1 µg of total DNA being used in this transformation.
For each transformation 1 µg of DNA was added with 100 µg denatured sheared salmon sperm
DNA to 100 µL of competent yeast cells. 700 µL of 1 x LiAc/40% PEG-3350/1x TE was then
added and mixed. Reactions were incubated for 30 min at 30°C, followed by the addition of 88
µL DMSO. Cells were then heat-shocked at 42°C for 7 min, followed by a 1,000 g centrifugation
for 10 s. Supernatant was removed, and cells were resuspended in 1 mL of 1 x TE. Cells were
again pelleted at 1,000 g for 10 s and resuspended in 100 µL TE, before being plated onto
selective SC-Leu-Trp plates (0.67% Bacto-yeast nitrogen base without amino acids (Sigma), 2%
glucose (w/v), 2% Bacto agar (w/v), 0.2% dropout mix (w/v) containing equal (w/w) proportions
of adenine sulfate, alanine, arginine, aspartic acid, asparagine, cysteine, glutamic acid,
141
glutamine, glycine, isoleucine, lysine, methionine, phenylalanine, proline, serine, threonine,
tyrosine, and valine [pH 5.9]). Plates were incubated for three days at 30°C. After three days
colonies were selected and restreaked onto fresh SC-Leu-Trp plates and incubated for another 3
days at 30°C. A negative control lacking plasmid was also set up, although no cells were found
to grow.
5.2.5 Confirmation of DNA Inserts
Yeast DNA was prepared using a QIAprep Spin Miniprep Kit (Qiagen), and a protocol adapted
by Jones (2000). MaV203 yeast transformants SG103, SG200, SG201, SG202, and SG203 were
cultured for two days at 30°C. Cells were pelleted (5 min at 5,000 g) and resuspended in 250 mL
P1 buffer containing 0.1 mg/mL RNase A. Approximately 100 mL of acid-washed glass beads
(Sigma) were used to rupture cells over 2 min in a minibeadbeater (Biospec). The remainder of
the DNA extraction was conducted as directed by Jones (2000), although 25 mL of elution buffer
(EB) was used in the final step instead of 50 mL. A 2 µL aliquot of purified DNA was used to
individually transform 50 mL of E. coli Top 10’ electrocompetent cells (Invitrogen), which were
plated onto LB + 100 µg/mL carbenicillin and LB + 10 µg/mL gentamycin and grown overnight
at 37°C. Cultures were successfully transformed, except for SG103 and SG 200 which did not
grow on LB + 10 µg/mL gentamycin. Single colonies were used to inoculate 5 mL liquid LB +
10 µg/mL gentamycin or LB + 100 µg/mL carbenicillin and grown overnight at 37°C. DNA was
purified using a QIAprep Spin Miniprep Kit (Qiagen), quantified using a NanoDrop 8000
Spectrophotometer
(Thermo
Scientific)
and
sequenced
using
ProQuest-recommended
sequencing primers (Invitrogen) (Table 5.4). Although SG200 did not grow on LB + 10 µg/mL
142
gentamycin, the TKS cassette was detected by direct sequencing using either the pDEST32 or
pDEST22 forward sequencing primers.
5.2.5.1 Determination of 3-AT Concentration used in Histidine Dropout Media
Before proceeding with the full characterization of transformants, the amounts of 3aminotriazole (3-AT) needed to inhibit HIS3 self-activation were assessed. Yeast transformants
SG100, SG101, SG102, SG200, SG201, and SG202 were grown overnight in 5 mL liquid SCLeu-Trp medium at 30°C. Cells were pelleted and resuspended in 1 mL sterile dH2O. Cells were
again pelleted and resuspended in 1 mL sterile water. Cultures were diluted with sterile water
until the OD600 of a 10x dilution (i.e. 100 µL of culture + 900 µL water) was 0.1 ± 0.01. This
allowed the standardization of cultures with an estimated starting OD600 of 1.0 ± 10%. These
cultures were serially diluted resulting in OD600 concentrations of 0.5, 0.25, 0.125, 0.0625, and
0.03125 (based on original OD600). A 5 µL aliquot from each of these dilutions was spotted onto
several SC-Leu-Trp-His plates containing either 10, 25, 50, or 100 mM 3-AT, and incubated at
30°C for three days.
5.2.5.2 Characterization of Transformants
Two-day old starter cells that had been grown at 30°C were pelleted (5,000 g for 10 min), the
supernatant was removed, and cells washed with 2 mL sterile H2O (this step removed excess
amino acids). The cells were again pelleted and resuspended into 1 mL sterile water then diluted
until an OD600 of 1.0 ± 0.1 was achieved. These cells were used to spot 5 µL aliquots of 1x, 2x
and 10x dilutions onto SC –Leu –Trp –His + 50 mM 3-AT and onto YPAD plates covered with
an 82 mm diameter nylon membrane specialized for colony and plaque hybridization (Roche).
The SC –Leu –Trp –His + 3-AT plates were incubated for three days at 30°C and then observed
143
for colony formation. The YPAD + membrane plates were incubated for 24 h at 30°C. The nylon
membrane was placed in liquid N2 for 1 min. The membrane was then thawed on two saturated
150 mm diameter filter papers (Whatman) in a 150 mm petri dish. The filter papers were
saturated with 8 mL X-gal solution (10 mg X-gal in 100 µL DMF + 60 µL 2-mercaptoethanol in
10 mL (filter sterile) Z-buffer (0.06 M Na2HPO4-7H2O + 0.04 M NaH2PO4-H2O + 0.01 M KCl +
0.001 M MgSO4-7H2O in H2O [pH 7]). Petri dishes were covered and incubated 24 h at 37°C.
5.2.6 TKS Magnetic Pulldown Methods
The poly-histidine tag in the pHIS8-TKS construct was exploited using Talon magnetic beads
(Clontech) to remove TKS from solutions containing malonyl-CoA and hexanoyl-CoA. Reaction
mixtures were allowed to incubate with TKS for a given time, after which the TKS was removed,
and the supernatant was transferred to a new reaction vessel containing OAC. The accumulation
of OA was then analyzed by LCMS.
A new OAC construct that was codon-optimized for its expression in E. coli (pJE60174)
was used during this work. The pJE60174 OAC construct was expressed, purified (as in Section
3.1.1), and digested in the presence of Thrombin to remove its poly-histidine tag. Protein
solution eluting from Talon beads was concentrated to a 1 mL solution using YM10 centrifugal
filter device (Millipore) and rinsed using a 5 mL solution of 20 mM HEPES, 50 mM NaCl, and 5
mM DTT before being concentrated again to 1 mL and transferred to a Spectra-Por Float-ALyzer with a 5 kDa MWCO (Sigma). A 10 U aliquot of high activity Thrombin (Millipore) was
added to the dialysis tubing. The solution was allowed to digest and dialyze overnight at 4°C.
The solution was removed from the dialysis tubing and incubated at 4°C for 1 h with 1 mL Talon
(Clontech). The Talon was added to a gravity flow column and the wash-through was collected
144
and concentrated using a YM10 centrifugal filter device (Millipore). This solution contained the
cleaved OAC without a His-tag.
Reaction mixtures were incubated with TKS for a short time, after which the TKS was
removed by magnetic aggregation and the supernatant was transferred to a new reaction vessel
containing either OAC or water. The accumulation of OA was then analyzed by LCMS as
described in Section 3.3.1. A flowchart for this experiment is illustrated in Figure 5.1.
Figure 5.1 Flowchart for TKS pulldown assays.
145
5.2.6.1 Magnetic Talon Beads
The binding capacity is 705 µg of protein for each 1 mL suspension of Talon beads, therefore 25
µL of beads per 10 µg of TKS were used, although 25 µL of beads is capable of binding almost
2x this amount of protein (i.e. 18.75 µg of protein). A total of 325 µL Talon suspension (12
reactions) was pelleted (1,000 g for 1 min) and washed twice using 300 µL of 20 mM HEPES
[pH 7.0] containing 5 mM DTT. A total of 25 µL of 5 mg/mL TKS was added to 275 µL of 20
mM HEPES [pH 7.0] containing 5 mM DTT. This volume was added to the washed Talon beads
and was allowed to tumble for 90 min at 4°C to assure proper binding. Beads were pelleted once
more (1,000 g for 1 min) and the supernatant discarded. A fresh 300 µL of buffer was added to
the Talon-TKS beads. This allowed for 10 µg of TKS per 25 µL volume.
5.2.6.2 TKS Pulldown Reactions
All experiments were conducted in triplicate. Reaction mixtures (100 µL) were derived from a
master mix consisting of 0.2 mM hexanoyl-CoA, 0.6 mM malonyl-CoA, 5 mM DTT, and 20
mM HEPES [pH 7.0]. Reactions were started by adding 25 µL of TKS-beads. Samples included:
Sample 1 – TKS-beads alone. Incubated at 10°C for 30 min. This sample was a negative control.
Sample 2 – TKS-beads + OAC. Incubated at 10°C for 30 min. This was a positive control.
Sample 3 – TKS-beads alone at 10°C for 15 min. Beads were pulled down, the supernatant was
transferred to a new reaction vessel containing OAC and allowed to react for an additional 15
min at 10°C. This sample is the experimental sample.
Sample 4 – TKS beads alone incubated at 10°C for 1 min. Beads were removed and supernatant
was transferred to a new tube, then maintained at 10°C for 14 min. OAC was added to the
146
supernatant and allowed to react for an additional 15 min at 10°C. This sample was to assure no
TKS was being released; it represents another negative control.
Reactions were extracted twice using 500 µL ethyl acetate. Extracts were dried under vacuum
then resuspended in 20 µL of resuspension solution. Injections of 10 µL were analyzed by
LCMS as previously described (Section 3.1.1).
5.2.7 Equilibrium Enzyme Assays
Assays were performed using Fast Micro-Equilibrium Dialyzers with 100 µL chambers (Harvard
Apparatus) (Figure 5.2). Each apparatus contained two chambers separated by a 5,000 Da
MWCO cellulose acetate membrane that had been rinsed in dH20, and then equilibrated for 30
min in 20 mM HEPES [pH 7.0]. Each chamber contained 100 µL reaction mixture consisting of
20 mM HEPES [pH 7.0], 5 mM DTT, 200 µM hexanoyl-CoA, and 600 µM malonyl-CoA.
Reactions were initiated by adding TKS (20 µg) into one of the chambers. Reactions included a
negative TKS-only control, a positive TKS and OAC (10 µg) control containing both enzymes in
one chamber, and a TKS and OAC sample containing 20 µg TKS and 10 µg OAC in separate
chambers. After incubating the reactions for 2 h at 10°C, reactions were transferred to 1.5 mL
tubes and extracted using 2 x 500 µL of ethyl acetate. Extracts were dried under vacuum and
resuspended in 20 µL of resuspension solution. 10 µL aliquots were analyzed by LCMS as
described in Section 3.1.1.
147
Figure 5.2 The Harvard Apparatus Dialyzer
This dialyzer, which is similar to the one used during my experiments, separates two 100 µL
chambers with a cellulose acetate membrane allowing for equilibrium to be reached between
both chambers, while partitioning molecules > 5 kDa (i.e. TKS and OAC).
148
5.3 Results
5.3.1 Sub-Cellular Localization of TKS and OAC
To test whether both TKS and OAC co-localize in planta, fluorescent fusion constructs were
made and used to infiltrate N. benthamiana plants where their location could be monitored by
confocal microscopy. Specifically, TKS was subcloned without a stop codon into the Gatewaycompatible pEarley102 vector to generate a pTKS:CFP, and OAC was subcloned without a stop
codon into the pEarley101 vector, resulting in the generation of pOAC:YFP (Earley et al., 2006).
These two constructs were separately used to transform Agrobacterium tumefaciens, which
subsequently were used to infiltrate N. benthamiana plants based on previously reported methods
(Sparkes et al., 2006; Weigel and Glazebrook, 2006). After a two-day incubation period,
epidermal cells were observed using confocal microscopy, where TKS:CFP and OAC:YFP were
visualized following excitation with a diode laser at the appropriate wavelengths. Results
indicate that both constructs co-localized to the same sub-cellular compartment, with some
residual fluorescence detected in the nucleus (Figure 5.3).
149
Figure 5.3 Co-localization of TKS and OAC.
The subcellular localization of (A) TKS:CFP and (C) OAC:YFP in N. benthamiana cells show
both constructs co-localize as seen when images are (B) overlayed. N. benthamiana epidermal
cells were examined by confocal microscopy.
5.3.2 Yeast Two-Hybrid Approach
Once the subcellular co-localization of TKS and OAC was confirmed, Y2H experiments were
carried out to determine if TKS and OAC physically interact. TKS and OAC were cloned into
the pDest32 and pDest22 vectors. A combination of plasmids were then used to transform the
Saccharomyces cerevisiae strain MaV203, resulting in a collection of different yeast
transformants (Table 5.5). Before proceeding with the full characterization of transformants, it
was required to assess the amounts of 3-aminotriazole (3-AT) needed to inhibit HIS3 selfactivation (Ausubel et al., 1994). Yeast transformants SG100, SG101, SG102, SG200, SG201,
and SG202 with an OD600 equal to 0.1 ± 0.01 were diluted to concentrations of 0.5, 0.25, 0.125,
0.0625, and 0.03125 (based on original OD600), before being spotted onto several SC-Leu-TrpHis plates containing 10 mM, 25 mM, 50 mM or 100 mM 3-AT. Plates were incubated at 30°C
for three days (Figure 5.4). A concentration of 50 mM 3-AT was found sufficient to inhibit HIS3
self-activation.
150
Once the appropriate level of 3-AT was determined, MaV203 yeast transformants were
spotted onto SC-Leu-Trp-His plates containing 50 mM 3-AT, and then grown for three days at
30°C. Other than the positive controls, no other transformant was capable of growth (Figure 5.5).
Additionally, the β-galactosidase activity of transformants was also assayed, yet only the yeast in
positive controls were capable of cleaving X-gal, resulting in a characteristic blue color (Figure
5.5). Collectively, these Y2H results indicate that OAC and TKS do not physically interact at the
molecular level.
Table 5.5 Yeast transformants used in Y2H experiments
NAME
SG100
SG101
SG102
SG103
SG104
SG105
SG106
SG107
SG200
SG201
SG202
SG203
BAIT VECTOR
pEXP32/Krev1
pEXP32/Krev1
pEXP32/Krev1
pDEST32
pDEST32
pDEST32
pSG058 (TKS)
pSG060 (OAC)
pSG058 (TKS)
pSG058 (TKS)
pSG060 (OAC)
pSG060 (OAC)
PREY VECTOR
pEXP22/RalGDS-wt
pEXP22/RalGDS-m1
pEXP22/RalGDS-m2
pDEST22
pSG057 (TKS)
pSG059 (OAC)
pDEST22
pDEST22
pSG057 (TKS)
pSG059 (OAC)
pSG057 (TKS)
pSG059 (OAC)
FUNCTION
strong positive control
intermediate positive control
negative control
negative control
negative control
negative control
negative control
negative control
negative control
test sample
test sample
negative control
Figure 5.4 Determining 3-AT concentrations used in Y2H assays.
MaV203 yeast transformants SG100, SG101, SG102, SG200, SG201, and SG202 were spotted
onto SC-Leu-Trp-His plates containing varying concentrations of 3-amino-1,2,4-triazole (3-AT).
151
Figure 5.5 Histidine biosynthesis and β-Galactosidase activity in Y2H screens.
Yeast two-hybrid assays included strong (SG100) and intermediate (SG101) positive controls,
and a negative (SG102) control. Colonies spotted on Histidine drop-out medium are shown in
left panels, whereas colonies transferred to assess β-galactosidase activity are shown on the
right. Numbers below the panels represent the MaV203 yeast transformant tested. Note that
SG101 did appear faint blue, but the color is not clelarly visible in this photo.
5.3.3 The Expression and Purification of OAC in pJE60174
Due to the low expression of OAC in pHIS8 and pET100 vectors, OAC was codon-optimized for
expression in E. coli using the pJexpress 411 vector (DNA2.0), resulting in the pJE60174
construct. Following the transformation of Rosetta2 BL21(DE3) Competent Cells (Novagen) and
the subsequent growth of a starter culture (100 mL LB + 50 µg/mL of kanamycin), 1 L cultures
152
were grown for 16 h in autoinduction medium containing 50 µg/mL of kanamycin at either 25 or
37°C. Cultures were started using 50 mL starter cultures, from which a pre-induction sample for
SDS-PAGE was taken (Figure 5.6). The variation in temperature during culturing was performed
to determine the optimal temperature for the expression of OAC. Post-induction samples were
also collected prior to purification for SDS-PAGE analysis (Figure 5.6). OAC was purified by
IMAC using Talon as previously described (Section 3.1.1), resulting in ca. 1 mL protein
solutions. A 5 µL aliquot of purified OAC from both 25°C and 37°C was used during SDSPAGE (Figure 5.6). Although the Talon purification of OAC used excessive amounts of Talon to
accommodate for excess protein that may or may not have been present, SDS-PAGE results
indicate that the expression of OAC was increased when 25°C temperature was used instead of
37°C (Figure 5.6). Expression of OAC at lower tempertatures resulted in ca. 8.5 mg of protein,
whereas expression at 37°C yielded ca. 1.2 mg of OAC. Thrombin was also used to remove the
poly-histidine tag from OAC (Figure 5.7).
One evident benefit of the inclusion of OAC in the pJexpress 411 vector was the loss of
the dual band seen in previous OAC purifications derived from pHIS8 and pET100 vectors
(Figure 4.1). The purification of OAC from pJexpress-based vectors resulted in the appearance
of a single band during SDS-PAGE (Figure 5.7)
153
Figure 5.6 SDS-PAGE of OAC expressed from the pJexpress 411 vector.
A 15% polyacrylamide gel was used to resolve 5.0 µL of pre-induced (PRE) and post-induced
(POST) cultures expressing OAC, and Talon-purified OAC (PURE). Cultures were initiated
using the same starter culture, but were then grown in parallel at either 25 or 37°C. Equal
volumes were used to load the gel. Proteins were purified in the presence of excess Talon
(Clontech) to ensure maximal amounts of protein were retained. OAC is evidently expressed, as
is seen in the PURE 25°C lane at the 13 kDa mark. OAC is indicated by arrows. Numbers on the
left of the gel correspond to the kDa of molecular weight markers (MW).
Figure 5.7 SDS-PAGE of Thrombin cut OAC.
IMAC purified OAC (PRE) was digested in the presence of Thrombin resulting in the loss of its
polyhistidine tag. Digested OAC (CUT) was isolated by running the solution over Talon beads,
which adhered to uncut OAC and other contaminants. The Talon-bound contaminants were also
eluted and visualized on this gel (ELUTANT). The gel was overloaded in order to detect
contaminating protein. Equal volumes of solution were used to load this gel. OAC is indicated by
arrows. Numbers on the left correspond to the kDa of molecular weight markers (MW).
154
5.3.4 TKS Magnetic Pulldowns
Experiments were designed to address whether or not OAC is capable of cyclizing TKS products
in the absence of TKS. The experiment allowed TKS to catalyze polyketide synthesis for a brief
time before being magnetically removed (pulled down), and removing the supernatant containing
TKS products. The supernatant was then incubated with OAC or water, extracted, and analyzed
by LCMS. Reactions included: 1) a negative control using only TKS-beads (B-TKS) that should
be capable of producing HTAL, PDAL, and olivetol, but not OA; 2) a positive control that
assayed TKS-beads in the presence of OAC and should be capable of producing OA; 3) an
experimental sample that allowed TKS-beads to react for 15 min before using a magnet to pull
the TKS-beads out of solution, then transferring the supernatant to a new reaction vessel
containing OAC and incubating the solution for another 15 min, and 4) a TKS bead leak control
that pulled down the TKS beads prior to transferring the solution to a reaction vessel containing
malonyl-CoA, then allowing the reaction to proceed for another 15 min. These assays were also
performed in parallel with assays that used soluble TKS (S-TKS) to compare the differences in
product accumulation between assays. Initially, to determine the optimum temperature at which
to conduct these experiments, the assays were performed at 10, 20, and 30°C (Figure 5.8).
Results show that TKS-only negative controls only produced HTAL, PDAL, and olivetol,
whereas TKS and OAC positive controls produced HTAL, PDAL, olivetol, and OA. A
difference between S-TKS and B-TKS chemoselectivity was observed, where S-TKS produced
almost an order of magnitude more HTAL and PDAL than did B-TKS, in the absence or
presence of OAC and at all temperatures (Figures 5.8 and 5.9). Notably, the amounts of olivetol
155
amount (pmol)
A
3.0E+03
10°C
HTAL
20°C
30°C
2.0E+03
1.0E+03
0.0E+00
S-TKS only
amount (pmol)
B
1.5E+04
amount (pmol)
amount (pmol)
D
S-TKS + OAC
B-TKS + OAC
B-TKS pulldown +
OAC
10°C
PDAL
20°C
30°C
1.0E+04
5.0E+03
0.0E+00
S-TKS only
C
B-TKS only
B-TKS only
3.0E+03
S-TKS + OAC
B-TKS + OAC
B-TKS pulldown +
OAC
10°C 20°C 30°C
OA
2.0E+03
1.0E+03
0.0E+00
S-TKS only
B-TKS only
S-TKS + OAC
B-TKS + OAC
B-TKS pulldown +
OAC
10°C 20°C 30°C
B-TKS + OAC
B-TKS pulldown +
OAC
olivetol
3.0E+03
2.0E+03
1.0E+03
0.0E+00
S-TKS only
B-TKS only
S-TKS + OAC
Figure 5.8 Products accumulating during TKS pulldown assays.
The product yields from soluble (S-TKS) and Talon-bound (B-TKS) TKS assays performed at
10, 20, and 30°C in the presence or absence of OAC. Each panel corresponds to one reaction
product, where the products were (A) HTAL, (B) PDAL, (C) OA, and (D) olivetol.
156
7.50E+03
HTAL
PDAL
OA
olivetol
amount (pmol)
6.00E+03
4.50E+03
3.00E+03
1.50E+03
0.00E+00
S-TKS only
B-TKS only
S-TKS + OAC
B-TKS + OAC
B-TKS pulldown +
OAC
Figure 5.9 The production of OA by OAC following TKS pulldown.
A comparison of soluble (S-TKS) and Talon-bound (B-TKS) TKS assays conducted in the
presence and absence of OAC show product amounts obtained when B-TKS was pulled out of
solution before adding OAC. All assays that included OAC successfully produced OA. Error
bars are the SD where n=3.
produced by S-TKS and B-TKS were approximately the same, which accumulated to ca. 1,800
pmol when assayed in the absence of OAC, and to ca. 300 pmol when assayed in the presence of
OAC (Figures 5.8 and 5.9). Similarly, OA production was also approximately the same when
assaying S-TKS or B-TKS in the presence of OAC, where the amount of OA accumulated to ca.
1800 pmol (Figures 5.8 and 5.9). Samples where B-TKS was removed prior to the addition of
OAC did accumulate OA. These samples accumulated OA ca. four times less than B-TKS +
OAC positive controls, accompanied by ca. 2.5 times more olivetol than seen in positive
controls. Optimum temperature for OA production differed between S-TKS and B-TKS samples,
where S-TKS showed increased OA activity at 20°C, and B-TKS showed optimum OA activity
at 30°C. Samples that were designated as B-TKS leak controls did not show any activity. Once
an optimal temperature of 20°C was determined (Figure 5.8), experiments were repeated to
confirm results (Figure 5.9).
157
The hypothesis that TKS produces a soluble linear tetraketide intermediate was also
explored by monitoring the TKS-B reactions over time (Figure 5.10). Reactions were incubated
with TKS-B for 15 min before pulling down TKS and transferring the supernatant to a new
reaction vessel. OAC was added to reactions immediately, or after 5, 10, 20 or 40 min had
elapsed. OAC reactions were incubated 15 min before stopping and extracting the reactions.
These reactions also used a negative control consisting of B-TKS only, and a positive control
which assayed B-TKS in the presence of OAC; both assays were incubated for 30 min before
stopping and extracting them, and were conducted in parallel with the other timed assays
described. All assays were conducted in triplicate. The largest amount of OA (1,600 pmol) was
detected in TKS and OAC positive controls, whereas no OA was detected in TKS-only negative
controls. Pulldown assays that added OAC immediately following the removal of TKS produced
the largest amount of OA (490 pmol), followed decreasingly by the 5, 10, 20, and 40 min timed
samples, which accumulated 287, 256, 238 and 200 pmol of OA, respectively. OAC activity, or
more correctly the substrate for OAC, was retained even after 40 min had elapsed between the
time TKS was removed and the time OAC was added. Conversely, amongst the samples that
were allowed to stand for a given time before OAC was added, the largest amount of olivetol
(1,300 pmol) was detected in samples that stood for 40 min before the addition of OAC, and
decreased as the sitting time before the addiction of OAC was shortened. Whereas HTAL
amounts remained relatively static (41 to 56 pmol), the largest amount of PDAL (754 pmol) was
detected in positive controls that assayed TKS in the presence of OAC. The accumulation of
PDAL appeared to increase as sampling time increased, although the initial 0 min sample did
contain an intermediate amount of PDAL (i.e. more than the 5 and 10 min samples, but less than
158
the 20 and 40 min samples). The least amount of PDAL and HTAL were detected in negative
controls.
HTAL
PDAL
OA
olivetol
amount (pmol)
1500
1000
500
0
B-TKS only
B-TKS +
OAC
0 min
5 min
10 min
20 min
40 min
Figure 5.10 Monitoring OAC substrate availability following TKS pulldowns.
Talon bound TKS (B-TKS) was incubated for 15 min in a reaction mixture before being
removed by magnetic pull-down. Following the transfer of the supernatant, OAC was added after
a given time (0 to 40 min), and allowed to react for another 15 min. Error bars are the SD where
n = 3.
5.3.5 The TKS/OAC Dialysis Reaction
To assess whether or not TKS and OAC can function independently, and if OAC’s substrate is a
diffusible intermediate that is initially produced and released by TKS before being cyclized by
OAC, dialysis experiments separated TKS and OAC enzymes using a 5 kDa membrane which
allowed the diffusion of substrates and reaction products between chambers. A negative control
incapable of OA production was accomplished by only adding TKS to one chamber, where the
second chamber contained no enzyme. A positive control was produced by adding both TKS and
OAC to one chamber, and not adding any enzyme to the second chamber. Experimental samples
included TKS in one chamber and OAC in the other chamber. Following 1 h incubation, each
chamber’s solution was separately extracted, and analyzed by LCMS. Whereas TKS-only
159
controls produced only HTAL, PDAL, and olivetol, positive controls and test samples were
capable of producing OA in addition to the other TKS products (Figure 5.11).
Figure 5.11 The micro-dialysis of TKS assay products.
Chromatograms are results from HPLC sample analysis that were monitored at 270 nm. (A) The
use of a micro-dialysis apparatus allowed TKS/OAC assays to be conducted while enzymes were
physically separated by a 5 kDa membrane in two separate chambers. (B) Positive controls
included both enzymes in one chamber and resulted in the production of HTAL, PDAL, OA, and
olivetol. (C) Negative controls only used TKS in one chamber. The elution of OA is highlighted
in pink across samples, whereas the elution of olivetol is highlighted in green.
160
5.4 Discussion
A central focus of this work was to determine if TKS and OAC interact. Results from subcellular
localization experiments indicate that both TKS and OAC co-localize in planta, and are not
spatially constrained, indicating that both proteins could potentially participate in the in vivo
production of OA (Figure 5.3). This observation raised the question as to whether or not TKS
and OAC physically interact, conceivably modifying TKS activity by an allosteric mechanism,
or possibly by directing a transfer of an unstable reaction intermediate. Y2H assays did not
indicate that any physical interaction exists between TKS and OAC (Figure 5.5). The lack of
physical interaction between these enzymes was also observed during i) TKS pulldown
experiments, and ii) TKS/OAC dialysis assays, both of which physically separated TKS and
OAC during assays, yet were capable of producing OA. Although these latter experiments
cannot directly determine if TKS and OAC physically interact, they show that TKS products can
be cyclized by OAC in the absence of TKS, which shows that these enzymes do not require
physical interaction for OA production. These results also indicate that neither an allosteric
mechanism nor the direct transfer of polyketide intermediates is required for OA biosynthesis.
Instead, these results, along with the observation that neither PDAL, HTAL, nor olivetol are
appropriate substrates for OAC (Section 4.2.10.2), and that OA production is augmented at
colder temperatures (Section 4.2.5) suggest that a somewhat thermolabile tetraketide
intermediate is released by TKS and is stable long enough for OAC to cyclize it. These findings
challenge the current understanding of type III polyketide cyclization, which was understood to
occur exclusively within the PKS active site itself (Austin and Noel, 2003). Namely, the Claisen
condensation that occurs in CHS has been described as an intramolecular cyclization that results
161
in the release of the aromatic compound from either a conserved cysteine that tethers the nascent
polyketide chain to the enzyme, or from CoA. Although the mechanism of polyketide release
from the enzyme is currently unknown, the cyclization specificity of chalcone and stilbene
synthases has been attributed to specific amino acid residues found within the type III PKS
active site (Austin et al., 2004). There exists evidence in support of a short-lived polyketide
intermediate, especially in type II PKSs, which produce irregular shunt products when they are
assayed in the absence of the necessary polyketide cyclases (Shen and Hutchinson, 1996; Sultana
et al., 2004). Since TKS does not cluster with traditional CHS or STS, and is instead related to
unique non-chalcone synthase PKSs, it is likely that the mechanism of retention, release and
cyclization of polyketide intermediates by TKS differs from traditional CHS/STS paradigms (a
descriptive analysis of the phylogenetic relationship between PKSs is shown in Appendix G
(Schröder, 2010).
The pathway for OA biosynthesis in cannabis requires the involvement of OAC, a unique
polyketide cyclase, which has not been previously known to occur in plants. Instead, similar
cyclases have been found in bacterial type II PKS systems, where multiple enzymes participate
in the biosynthesis of polyketide products (Shen and Hutchinson, 1996). The involvement of
OAC-like cyclases during polyketide cyclization may not be exclusive to cannabis; it is
conceivable that similar cyclases are involved in other plants where the accumulation of
carboxylated polyketides remains unclarified. Examples of resorcylic acid accumulation in plants
include anacardic acids in cashew and stilbene carboxylates in Hydrangea (Gorham, 1995;
Kozubek and Tyman, 1999), although it should be noted that type III PKSs capable of producing
162
long-chained alkylresorcylic acids without the requirement of a cyclase have been previously
identified in rice and in the mold Neurospora crassa (Funa et al., 2007; Matsuzawa et al., 2010).
An observation indirectly related to this project was the reduction of lactone products
during assays performed using Talon-bound TKS. Assays that compared the activities of
immobilized and soluble TKS showed a decrease in the production of lactone products (HTAL
and PDAL) with almost no change to OA production when immobilized TKS was used (Section
5.2.3). Similar results have also been reported for alfalfa CHS that had been immobilized onto
Ni-NTA agarose, which, in addition to increased tetraketide:triketide yields, resulted in a bound
enzyme with a much more effective half-life than soluble CHS (Kim et al., 2009). Although the
half-life of TKS was never explored in the current study, the observation that both immobilized
PKSs produce higher tetraketide:triketide yields suggests that product derailment occurs easily in
these enzymes, and that in vivo these enzymes may very well be anchored to a naturallyoccurring support matrix, as suggested for phenylpropanoid and flavonoid metabolism (WinkelShirley, 1999). Recent studies have shown that resuspended type II minimal PKS involved in
actinorhodin biosynthesis increases product yield when retained and stabilized within a dense
coacervate matrix (Crosby et al., 2012). The authors attribute their findings to a crowding effect
that allows for the direct transfer of PKS intermediates, but it may also be due to the stabilization
of the PKS itself which leads to its increased efficiency. The matrix used by Crosby et al. (2012)
is also reminiscent of the unique physiology of the trichome cell, which is dominated by
spherical secretory bodies that could potentially be involved in anchoring TKS (Hammond and
Mahlberg, 1978). The chemoselective adaptability observed in TKS could be of practical benefit
in other type III PKS research that hopes to maximize enzyme efficiency.
163
6. THE STRUCTURE AND MECHANISM OF OAC
6.1 Introduction
In the previous chapter, experiments were conducted to examine the physical relationship shared
between TKS and OAC. The observation that TKS and OAC contructs co-localize in vivo in N.
benthamiana suggested that the proteins could potentially interact, which led to yeast two-hybrid
experiments being performed to determine if physical interaction did occur. Yeast two-hybrid
assays indicated that no physical interaction exists between TKS and OAC. In turn, this
observation led to experiments that separated TKS and OAC, including the physical removal of
TKS midway through a reaction by magnetic pull-down methods, as well as the use of a micro
dialysis apparatus that allowed the enzymes to be separated by a permeable membrane. These
experiments indicated that the enzymes perform separate reactions, wherein TKS synthesizes a
theoretical soluble tetraketide intermediate, which is then independently cyclized to OA by OAC.
The observations that TKS and OAC act separately has led to two related and important
questions, which are; what is the mechanism of cyclization used by OAC, and what is the nature
of the theoretical tetraketide intermediate? In the present chapter, experiments that were
conducted to address the mechanism of OA cyclization and the nature of the intermediate are
presented.
The mechanism of chalcone and stilbene biosynthesis involves an intramolecular
cyclization step that can be dramatically influenced by replacing several key amino acids located
in the active site (Austin and Noel, 2003). Comparatively, TKS appears to behave as an STS
when assayed alone. Similarly, during TKS and STS cyclizations, a carboxyl moiety is lost,
164
which suggests that an aldol reaction occurs during the production of olivetol as it does during
the biosynthesis of stilbenes. This is contrasted by Claisen reactions that occur in CHSs, which
retain all carbon moieties during the cyclization of final products. Yet, when TKS is assayed in
the presence of OAC, the two enzymes together behave as an STCS capable of catalyzing an
aldol C2 C7 cyclization while retaining its C1 carboxyl moiety (Figure 2.5). To better
understand the mechanism of cyclization adopted by OAC, traditional aldolase mechanisms,
which include class I and class II reactions, were considered (Gijsen et al., 1996). The role of a
metal as a catalyst could have implicated a class II aldol mechanism in OAC, but this was
dismissed since no metal is required for OA cyclization (Chapter 4), suggesting that a type I
aldolase mechanism could be employed by OAC. It was a goal of this research to examine the
lysine residues putatively involved in a type I aldolase mechanism in OAC, which were
conserved by homologs that could share a similar function, but in addition to simply
investigating conserved lysine residues, several AA residues that are conserved in OAC and its
homologs were also investigated.
The experiments in the present chapter use the primary and secondary structure of OAC
to determine residues putatively involved with enzyme activity. The primary sequence of OAC
was compared with twelve other SRABB proteins, and a homology-based model of OAC was
developed. These analyses led to the identification of conserved AAs shared between SRABB
homologs, some of which were subsequently targeted for alanine substitution. Additionally, to
better understand the putative active site, the OAC structure was modeled using the crystal
structures of A. thaliana HS1 (pdb 1Q53) and P. tremula SP1 (pdb 1TR0) as templates, allowing
165
the closer inspection of the enzyme’s three-dimensional strucutre and the positions of sitedirected mutatgenesis (SDM) sites.
Since previous observations regarding TKS and OAC reactions imply the existence of a
transient tetraketide intermediate, attempts to detect its existence were performed using highly
sensitive analytical MS techniques. MS ion screens were performed during TKS and TKS/OAC
reactions in order to detect CoA-linked compounds, with specific interest in a tetraketide-CoA
intermediate that could exist transiently in solution. Due to the uncertainty about the reaction
mechanism and in determining which intermediate is produced during a TKS reaction, these
techniques were used to find ions possessing CoA fragments with a common positively charged
fragment of 428 m/z (Figure 6.1). Similarly, neutral loss experiments were also used to detect a
loss of 507 amu from CoA-containing molecules.
A
O
O
O
O
CoA
CoA-linked tetraketide intermediate FW 992.8
B
428 m/z
neutral loss 507 m/z
NH2
N
N
N
OH
N
O
O
P
O
OH
HO
O
P
OH
O
P
O
OH
NH
NH
O
O
S
R
O
OH
O
Figure 6.1 CoA fragmentation during MS analysis.
(A) A proposed linear polyketide intermediate with a molecular weight of 992.8 screened for
during MS analysis. (B) Sites of bond cleavage occur along the dashed lines in the CoA-portion
of acyl-CoAs during MS analysis conducted in positive ion mode. Substituents (R) may include
hexanoyl, acetyl, malonyl, and other products, such as seen in (A).
166
6.2 Materials and Methods
6.2.1 Primary and Secondary Structure Analysis
A BLASTP query using the OAC ORF was conducted against the NCBI non-redundant protein
sequence database using a BLOSUM62 scoring matrix, a gap cost of 11 and extension cost of 1,
using the conditional compositional score matrix adjustment (Altschul et al., 1997). Highestranking homologs with the lowest E-value were selected, although only one homolog from a
species was selected. Sequences were aligned using ClustalW engine as previously described
(Section 4.2.1). Analysis was conducted by querying the OAC sequence using the Swiss-Model
Template Identifier (Arnold et al., 2006). The Interpro Domain Scan was conducted against the
HMMPfam, HMMTigr, ProfileScan, SuperFamily, and BlastProDom databases using a gapped
blast query with an E-value cutoff of 1.0E-6 and a BLOSUM62 matrix (Zdobnov and Apweiler,
2001). Secondary Structure prediction and domain assignment were completed using the
PSIPRED and DISOPRED engines provided by the domain annotation feature in the SwissModel Workspace (Jones, 1999). The DISOPRED E-value cutoff was 2%.
6.2.2 Cloning, Expression and Activity of OAC Mutants
This work used an OAC construct (pJ60174) that was optimized for codon usage in E. coli and
was commercially prepared in the pJE411 vector (DNA2.0). This OAC construct also contains
an N-terminal polyhistidine tag for IMAC purification. In addition to the wild-type OAC
(wtOAC), 12 constructs each containing a single point alanine substitution were also ordered
(DNA2.0). Additional single point OAC mutants at K4A, K38A, and D45A were not
commercially produced, but created by the site-directed mutagenesis (SDM) of pJ60174, and are
167
found in pJ60187, pJ60188, and pJ60189, respectively. Primers used during SDM are shown in
Table 6.1.
Table 6.1 Primers used for SDM and sequencing of pJR60174
Name
Sequence
Function
F-K4A
GTAGCATGGCGGTGGCACACTTGATCGTG
(F) primer for pJ60187 K4A
R-K4A
CACGATCAAGTGTGCCACCGCCATGCTAC
(R) primer for pJ60187 K4A
F-K38A
CATTATTCCGGCGATGGCAGACGTTTATTGGGG
(F) primer for pJ60188 K37A
R-K38A
CCCCAATAAACGTCTGCCATCGCCGGAATAATG
(R) primer for pJ60188 K37A
F-D45A
GTTTATTGGGGCAAGGCTGTGACCCAGAAGAAC
(F) primer for pJ60189 D44A
R-D45A
GTTCTTCTGGGTCACAGCCTTGCCCCAATAAAC
(R) primer for pJ60189 D44A
F-pJE
CTCGAAAATAATAAAGGGAAAATCAG
(F) primer for sequencing
R-pJE
TGGTAGTGTGGGGACTC
(R) primer for sequencing
6.2.3 Creation of K4A, K38A, and D45A by SDM
Reactions (50 µL) were performed using 5 µL Pfu II Ultra HS 10x buffer, 1 µL of 50 ng/µL
pJE60174 (template), 1 µL of 125 ng/µL forward (F) primer, 1 µL of 125 ng/µL reverse (R)
primer, 0.5 µL 100 mM dNTP (25 mM of each), 40.5 µL water, 1 µL of 2 U/µL Pfu II ultra HS
polymerase (Stratagene). Forward and reverse primers are described in Table 6.1. Thermocycling
was performed in five steps, where steps 2 - 4 were repeated 16 times before ending with step 5:
step 1 - 95°C for 1 min; step 2 - 95°C for 1 min; step 3 - 58°C for 1 min; step 4 - 68°C for 2 min;
step 5 - 68°C for 3 min. Reactions were incubated at 37°C for 60 min in the presence of 1 µL of
8 U/µL Dpn1 (Sigma).
Plasmids from SDM reactions were isolated using a MinElute Reaction Cleanup Kit
(Qiagen), and each used to transform separate E. coli One Shot BL21 (DE3) competent cells
(Invitrogen). Transformants were plated onto LB + 50 µg/mL kanamycin and cultured overnight
at 37°C. Six colonies from each transformation were selected and grown in 5 mL LB + 50
168
mg/mL kanamycin overnight at 37°C. Plasmids were purified using a QIAprep Spin Miniprep
Kit (Qiagen) and sequenced (PBI DNA Technologies Unit) using the F-pJE and R-pJE
oligonucleotides (Table 6.1) to confirm that the SDM reactions were successful.
6.2.3.1 Expression and Purification of OAC and its Mutants
Plasmids harboring wild-type OAC (wtOAC) and OAC mutants were individually used to
transform E. coli Rosetta (DE3) competent cells (Novagen). Transformation reactions were
plated on LB + 50 µg/mL kanamycin and grown overnight at 37°C. Single colonies were used to
inoculate two 5 mL cultures of LB + kanamycin, where one solution was used as a starter culture
and the other for i) -80°C stocks (as described in Section 3.1.2), and ii) plasmid isolation and
sequencing as described in Section 6.2.3. A 0.5 mL aliquot of starter culture was used to
inoculate 50 mL LB + 50 µg/mL kanamycin, which was grown overnight at 37°C with shaking
(200 RPM). Cultures were used to inoculate 500 mL of LB + 50 µg/mL kanamycin, and grown
to an OD600 = 0.8 before being induced using 0.5 mM IPTG. Cells were grown for an additional
5 h at 37°C before being harvested by centrifugation at 20,000 g. Pellets were stored overnight at
-80°C.
Cell pellets were thawed on ice in the presence of 50 mL lysis buffer. The solution was
divided into 25 mL aliquots and each sonicated for 3 min on ice. The solutions were then
centrifuged for 20 min at 20,000 g at 4°C. The decanted lysates were tumbled in the presence of
500 µL suspension volume of Talon (Clontech) for 40 min at 4°C. The resin was centrifuged (1
min at 1,000 g) and the supernatant removed. The Talon beads were resuspended in 10 mL wash
buffer and tumbled at 4°C for 10 min to wash away protein impurities. The washing step was
conducted four more times before the Talon beads were rinsed in a gravity flow column with a
169
10 mL wash buffer. Each protein was eluted using 10 mL of elution buffer directly into separate
Amicon Ultra 15 centrifugal concentrators (5,000 MWCO) (Millipore), which were later used to
concentrate the solutions to 2.5 mL. The solutions were desalted using a PD10 column (GE
Healthcare Biosciences) equilibrated with storage buffer (20 mM HEPES pH 7.0, 25 mM NaCl,
10% glycerol, 5 mM DTT). The solutions were further concentrated to 100 µL using Microcon
centrifugal filter devices (Millipore). These solutions were quantified using an RC/DC protein
assay kit (Bio-Rad) as directed by the manufacturer.
6.2.3.2 Enzyme Assays
Separate 50 µL reactions containing a final concentration of 0.2 mM CoA, 0.4 mM ATP, 2.5
mM MgCl2, 8 mM sodium malonate, 5 mM DTT, 20 mM HEPES [pH 7.0], and 0.2 mM
hexanoyl-CoA were prepared using a master mix. Each reaction mixture contained 200 pmol of
wtOAC or one of the single point OAC mutants. The first experiments (experiment A) did not
include the K4A, K38A, or D45A mutants, whereas the second set of experiments (experiment
B) did. Experiments A and B were conducted independently with slight differences described in
the methods below. All reactions were conducted in triplicate.
Experiment A
Reactions were started by adding 2 µL of an enzyme cocktail containing 100 pmol TKS and 200
pmol MCS. Reactions were incubated for 30 min at RT before being extracted twice using 250
µL ethyl acetate. Extracts were dried under vacuum and analyzed by LCMS as previously
described (Section 3.1.1).
170
Experiment B
Reactions were started by adding an enzyme cocktail containing 200 pmol of TKS and 200 pmol
of MCS, and were incubated at 20°C for 45 min before being extracted twice using 200 µL ethyl
acetate. Extracts were dried under vacuum and stored overnight at 4°C. The extracts were
resuspended in 20 µL of resuspension solution. 10 µL aliquots were analyzed by LCMS as
previously described (Section 3.1.1).
6.2.4 MS/MS Monitoring of the TKS/OAC Reaction
Reaction mixtures of 50 µL were prepared with 0.6 mM malonyl-CoA, 0.2 mM hexanoyl-CoA,
5 mg/mL enzyme (TKS and/or OAC) in 1 mM HEPES [pH 7.0] and 5 mM DDT, and were
incubated for 1 h at 37°C, before being stored in the autosampler at 8°C. A control sample was
also prepared without enzyme present. Reaction mixtures were diluted by a factor of ten with 1
mM HEPES [pH 7.0] and injected to a stream of 50:50 ACN:H2O with 5 mM triethylamine
(TEA) and 3 mM acetic acid (HOAc), using an Alliance 2695 HPLC coupled to a Quattro
Ultima mass spectrometer (Waters Corporation). Signals were recorded in positive ion mode
electrospray ionization with tandem MS scanning of selected reaction monitoring (SRM) for
each of the acyl coenzyme-A compounds of interest; malonyl-CoA, acetyl-CoA, hexanoyl-CoA,
free CoA, formyl-CoA, and a tetraketide intermediate-CoA (992 m/z) (de Hoffmann 1996).
Since an intermediate was expected, an alternate mode of tandem MS precursor ion scanning
was also used to monitor any possible acyl-CoA that might be produced during the reaction.
Analyses were conducted by S. Ambrose (Applied Genomics and Analytical Technologies,
NRC-PBI).
171
6.3 Results
6.3.1 Primary and Secondary Structure Analysis
Following a BLASTP query of the OAC protein sequence, 12 representative SRABB proteins
were selected and aligned using ClustalW with the MEGA5 software suite (Figure 6.2). Most
SRABB homologs used in this work are annotated in Genbank as POP3 proteins and include
members from: Medicago truncatula (annotated XP_003609973.1), Vitis vinifera (annotated
XP_002272048.1), Glycine max
(annotated
XP_003526484.1), Zea mays
(annotated
NP_001152608.1), Ricinus communis (annotated XP_002510283.1), A. thaliana (pdb 1Q53), P.
trichocarpa (annotated ABK93400.1), Brachypodium distachyon (XP_003579052.1), Oryza
sativa
(annotated
NP_001043194.1),
Sorghum
bicolor
(annotated
XP_002458107.1),
Physcomitrella patens (annotated XP_001766578.1) and P. tremula (pdb 1TR0). A total of 19 of
the 103 amino acids in OAC (i.e. 18%) are identical in all of the selected homologs, and another
11 amino acids share similarity (Figure 6.2).
Furthermore, the OAC protein sequence was analyzed using software provided by the
Swiss Institute of Bioinformatics (Arnold et al., 2006). The secondary structure was deduced
using InterproScan, PSIPRED and DISOPRED capable of domain recognition, secondary
structure prediction, and protein disorder determination, respectively (Figure 6.3). Interpro
identified both the DABB (IPR011008) and SRABB (IPR013097) domains in OAC between
amino acids 3 to 99 (Zdobnov and Apweiler, 2001). PSIPRED results indicated that OAC
possesses multiple α-helices and β-pleats that fit an α + β classification (Jones, 1999). Results
from DISOPRED strongly suggested that OAC is an ordered protein with a stable secondary
structure possessing low flexibility in solution (Ward et al., 2004).
172
173
- - - - - C sativa OAC - - - - - M truncatula - - - - - V vinifera - - - - M E
G max - - - - M E
Z mays - - - - - M
A thaliana - G S H M E
R communis - - - - M E
P trichocarpa - - - - M E
B distachyon - - - - - M
O sativa - M A A E A
S bicolor - - - - - M
P patens - - - - - P tremula - - - M A T
E
E
A
E
E
E
A
G
A
R
-50
- 4 5
11 12
30
38
42 43 45
- - - M A V K H L I V L K F K D E I T E A Q K E E F F K T Y V N L V N I I P A M K D V Y W G K D V T Q K - - N -50
- - M A V K H I V I A K F K D E I S Q E K I D E L I K G Y A N L V N L I P P M K S F H W G T D V S A E - - N -50
A K G L V K H V L L A K F K D S T P P D Q I E E L I K G Y A N L V N L V P P M K A F H W G K D V S I E - - N -55
A K G V V K H V F L A K F K D D V T P E R I E E L I K D Y A N L V N L I P P M K S F H W G K D V S A E - - N -55
G G G V V K H I L L A S F K E E V T Q E R L D E L I R G Y A A L V G V V P S M K A F H W G T D V S I E - - N -54
A K G P V K H V L L A S F K D G V S P E K I E E L I K G Y A N L V N L I E P M K A F H W G K D V S I E - - N -58
A K G V V K H V L L A K F K D E I P T H E I D K L I K G Y A N L V N L I P P M K A F H W G S D V S I E - - N -55
A K G V V K H V L L A K F K E G I P S D E I E K L I K G Y A N L V N L I E P M K A F E W G T D V S I E - - N -55
G S G V V K H I L L A R F K D E V T P E R L D Q L I R G Y A G L V A V V P S M K A F H W G T D V S I E - - N -54
S G G V V K H I L L A R F K E D V A P E R L D Q L I R G Y A G L V D L V P S M K A F H W G T D V S I E - - N -58
S G G V V K H I L L A R F K E D V T Q E R L D E L I R G Y A A L V A A V P S M K A F H W G T D V S I E - - N -54
M S G V V K H I L L L K F K D S I S P D E Q T A L I Q K Y A A L P N S I S A M K G F E W G T D I S V E - - N -52
T P K L V K H T L L T R F K D E I T R E Q I D N Y I N D Y T N L L D L I P S M K S F N W G T D L G M E S A E -58
- - - - - - 56- 57- -
C sativa OAC - K E E
M truncatula - L H Q
V vinifera - M H Q
G max - M H Q
Z mays - M H Q
A thaliana - L H Q
R communis - L H Q
P trichocarpa - M N Q
B distachyon - M H Q
O sativa - M H Q
S bicolor - M H Q
P patens - M H Q
P tremula - L N R
G
G
G
G
G
G
G
G
G
G
G
G
G
Y
F
F
F
F
Y
F
F
F
F
F
F
Y
T
T
T
T
T
T
T
T
T
T
T
T
T
H
H
H
H
H
H
H
H
H
H
H
H
H
I V EVT
VFESN
VFEST
VFEST
VFEST
I FEST
VFESS
I FEST
VFEST
VFEST
VFEST
AF I ST
AFEST
72
75
78
F E S V E T I Q DY I I H P AH VG F GDVYR S F W E K L L I
F E S V E A I A E Y VA H P A H V E YAN L F L S N L E KV I A
F E S V E G I A E Y V S H P A H V E F AN L F L P H L E KV I V
F E S T E G I A E Y T A H P A H V E YA T L F L S Y L DKV I V
F E S T E G I K E Y I E H P A H V E F A S V F L P V L E KVL I
F E S K E A V A E Y I A H P A H V E F A T I F L G S L DKVLV
F E S T E D I A E Y I S H P A H V E F AN L F L R A V DKV I V
F E S K E G I A E Y I G H P A H VD F GG L F L P A L E KV I V
F E S T E GV K E Y I E H P N H VD F A T E F L A L A E KV I I
F E S T E GV K E Y I E H P A H V E F AN E F L P V L E K T L I
F E S T E G I K E Y I E H P K H V E F ANV F L P V L DKV I V
F D S P QG R D E Y L V H P V H D S F AK E L L G A V DNALV
F E S K S G L Q E Y L D S AA H AA F A E G F L P T L S QRLV
- -
96
F
I
F
I
I
I
I
F
I
I
I
F
I
DYT P
D YK P
D YK P
D YK P
D YK P
D YK P
D YK P
D YK P
D YK P
D YK P
D YK P
DFNP
DYF L
R
T
I
T
T
T
T
T
T
T
T
T
Y
K - - - - I VNL - T VHL - I VKL - S AN - - SVSL - I VHL - VAR F - AAT - - I VNNS S VN - - V A L KA N
- - - - - -
S
-
-
-101
-104
-109
-109
-107
-112
-109
-109
-107
-113
-107
-109
-108
Figure 6.2 Sequence alignment of OAC and homologs.
This alignment was used to determine conserved amino acid residues, some of which (boxed) were targeted for SDM. Conserved residues that are identical are
shown in red, whereas amino acids that share similarity are drawn in blue.
173
A
B
C
Figure 6.3 Prediction of OAC Secondary Structure.
The OAC protein sequence was evaluated using: (A) Interpro, (B) PSIPRED, and (C)
DISOPRED. (A) Interpro identified the DABB and SRABB motifs in OAC. (B) PSIPRED,
which is capable of predicting the probablilty of protein regions to exist as a (red) helix, (yellow)
extended β-sheet or as a (green) coil, indicates the α + β components in OAC. (C) DISOPRED
determines the ordered state of a protein; results indicate that OAC is an ordered protein, as is
depicted by the appearance of dashes (for order) instead of asterisks (for disorder).
174
To identify suitable structural homologs from which an OAC structure model could be predicted,
the OAC protein sequence was evaluated using the template identification feature provided by
the Swiss Model Workspace (Altschul et al., 1997; Arnold et al., 2006) (Figure 6.4). Three
protein structures were identified by Gapped Blast and ten protein structures were identified by
HHSeach Template Library searches.
Figure 6.4 Swiss Model template retrieval of OAC.
A template search using the OAC protein sequence identified three suitable matches when a
Gapped Blast was performed and ten templates when the HHSeach Template Library Search was
used. Both 1Q53 and 1TR0 were each used to develop separate OAC models due to their good
representation in each search.
175
Two homology models of OAC were created using the Swiss-Model automated service
that were based on either A. thaliana HS1 (annotated in protein databank (pdb) 1Q53) or P.
tremula SP1 (pdb 1TR0) (Figure 6.5). The first OAC model (OAC Model 1) used 1Q53 (chains
A and B) as a template, since the template identification feature identified AtHS1 as the most
probable candidate in both Gapped Blast and HHSeaches. The resulting QMEAN Z-score, an
estimate of the quality of the created model, has a value of -3.25 (based on a scoring range
between -4 to +4). Therefore, another model (OAC Model 2) was created using 1TR0 as a
template, since it was also well represented in both searches; this second model was a good
prediction according to its QMEAN Z-score of -1.2. Although both models were created using
separate templates, they are visually similar. Models used two chains from each respective
protein to create a homodimeric model. Each monomer contains four β-strands that align with
the opposing monomer’s β-strands resulting in a barrel-shaped structure at the core of the
protein. Two α-helices project away from the β-sheet core in each monomer. Viewing the spacefilling representation of the 1Q53 based OAC model reveals a crevice between the α-helices and
the β-sheets (Figure 6.6). A similar crevice is also detected in the 1TR0 based OAC model.
176
Figure 6.5 Templates and models for OAC.
(A) AtHS1 and (B) PtSP1 were each used to predict OAC models shown to the right of each
template. All protein structures are shown as homodimers with separate monomers displayed in
blue and red. Structures were positioned as AtHS1.
Figure 6.6 Space filled model of OAC.
OAC Model 1 from Figure 6.5A is shown tilted to highlight a crevice found between the αhelices and β-sheet, and proposed to act as the substrate-binding pocket. The model on the left is
the same as the one on the right save for the addition of the space-filling features. Although not
seen, a similar pocket also exists in the other monomer.
177
6.3.2 The Expression and Activity of OAC Mutants
The structure/function relationship of various AAs in OAC was explored by substituting alanine
at AA sites conserved between SRABB homologs, in addition to His75 and His78, which are not
conserved in the P. tremula homolog (Figure 6.2). Experiments were conducted on two
occasions (experiments A and B) and were initially designed to compare the activity of 12 OAC
mutants with the wild-type, but later included another three mutants (i.e. K4A, K38A, and
D45A). The expression level of mutant proteins was determined by comparing their yields
following IMAC purification to that of wtOAC (Table 6.2). Cultures expressing the Y72A,
H78A, and D96A mutants in experiment A, and the D45A, H78A, and D96A mutants in
experiment B produced the greatest amounts of total protein. The smallest amounts of protein
were produced in cultures expressing the F11A, L30A, and G43A mutants during experiment A,
and the F11A, W42A, and T56A mutants during experiment B. Cultures expressing the K12A
and Y72A mutants showed the largest fluctuation in protein yields between experiments A and
B; cultures expressing the K12A mutant yielded 75% less protein in those expressing wtOAC in
experiment A, but no such loss was seen in experiment B; cultures expressing the Y72A mutant
yielded 70% more protein in cultures expressing wtOAC during experiment A, yet yielded 7%
less protein in its wtOAC counterpart during experiment B. Other than these two mutants, the
amounts of protein produced by cultures showed the same trend between experiments A and B.
The SDS-PAGE of wtOAC and mutants was conducted prior to experiment A (Figure 6.7). This
gel does not include the K4A, K38A, and D45A mutants.
178
Table 6.2 OAC constructs and their yields
Experiment A
Experiment B
plasmid
volume concentration yield expression volume concentration yield expression
OAC
version
(µL)
(mg/mL)
(µg)
(µL)
(mg/mL)
(µg)
wt*
pJE60174
200
1.01
202
100%
100
2.50
250
100%
pJE60187
n/a
K4A
n/a
n/a
n/a
100
0.600
60
24%
H5A
pJE60175
100
0.780
78
39%
100
1.94
194
78%
F11A
pJE60176
60.0
0.430
25.8
13%
100
0.760
76
30%
K12A
pJE60177
60.0
0.850
51.0
25%
100
2.49
249
100%
L30A
pJE60178
60.0
0.400
24
12%
100
1.41
141
56%
K38A
pJE60188
n/a
n/a
n/a
n/a
100
2.22
222
89%
W42A
pJE60179
60.0
0.620
37.2
18%
100
0.850
85
34%
G43A
pJE60180
60.0
0.450
27
13%
100
1.39
139
56%
D45A
pJE60189
n/a
n/a
n/a
n/a
100
3.93
393
160%
T56A
pJE60181
100
0.330
33.0
16%
100
1.16
116
46%
H57A
pJE60182
100
0.580
58
29%
100
1.46
146
58%
Y72A
pJE60183
100
3.33
333
170%
100
2.32
232
93%
H75A
pJE60184
60.0
0.850
51
25%
100
1.47
147
59%
H78A
pJE60185
100
2.18
218
110%
100
3.42
342
140%
D96A
pJE60186
500
0.610
305
150%
100
6.39
639
260%
*wt is the wild-type OAC which did not contain any amino acid substitution. Constructs which were not included
during the experiment are labelled as not applicable (n/a).
Figure 6.7 SDS-PAGE of wtOAC and mutants.
Proteins purified by IMAC were loaded onto a 15% acrylamide gel to confirm their presence and
abundance. A 5 µg aliquot of wild-type OAC (wtOAC) and mutant OAC loaded on equal protein
basis (names above the lanes) were run at 200 V for 45 min beside molecular weight standards
(MW). Numbers on the left of the gel correspond to the kDa weight of MW. Proteins of interest
are seen at the 13 kDa mark.
179
Although assays performed in experiments A and B are similar, they were conducted
under different conditions (where the production of OA was further optimized during experiment
B), and absolute amounts are therefore not averaged. Instead, results are presented separately,
and the observable trends are compared. All reaction products were analyzed by LCMS, and
results are reported as the total pmol yields determined by LC at 270 nm (For Experiment A see
Figure 6.8; For Experiment B see Figure 6.9).
6.3.2.1 Experiment A
The activities of OAC mutants were compared to the wild-type enzyme by assaying each protein
with TKS and MCS. The amount of HTAL collected during assays was between 36 ± 6 pmol and
54 ± 24 pmol (Figure 6.8A). A larger amount of variation (± 20 pmol) was detected in assays that
accumulated greater than 45 pmol of HTAL, which included the L30A, W42A and D96A
constructs. The largest amount of HTAL was detected from D96A assays, whereas the lowest
amount was detected from G43A assays. The amount of PDAL synthesized during each assay
appears static, the lowest amount was seen from G43A assays (376 ± 90 pmol); the largest
amount was detected from H78A assays (484 ± 84 pmol), with the exception of the no-OAC
control (580 ± 252 pmol) (Figure 6.8B). Other than wtOAC assays, which produced 167 ± 34
pmol, OA was only detected in assays performed using K12A, G43A, and D96A constructs,
where they accumulated 42 ± 13 pmol, 14 ± 6 pmol, and 32 ± 14 pmol, respectively (Figure
6.8C). The standard deviation of olivetol produced during assays appear the most dynamic,
where the least amount of variation was detected from assays that used wtOAC (45 ± 21 pmol),
and the largest amount from L30A assays (327 ± 169 pmol) (Figure 6.8D).
180
A 100
HTAL
amount (pmol)
80
60
40
20
0
amount (pmol)
B
wild
type
H5A F11A K12A L30A W42A G43A T56A H57A Y72A H75A H78A D96A
no
OAC
PDAL
800
600
400
200
0
wt
amount (pmol)
C
no
OAC
OA
200
150
100
50
0
wt
D 500
amount (pmol)
H5A F11A K12A L30A W42A G43A T56A H57A Y72A H75A H78A D96A
H5A F11A K12A L30A W42A G43A T56A H57A Y72A H75A H78A D96A
no
OAC
olivetol
400
300
200
100
0
wt
H5A F11A K12A L30A W42A G43A T56A H57A Y72A H75A H78A D96A no OAC
Figure 6.8 LC results for OAC mutant assays - Experiment A
Product yields for (A) HTAL, (B) PDAL, (C) OA, and (D) olivetol biosynthesized during
TKS/OAC assays conducted in the presence of wild-type (wt) or mutant OAC was determined
by LCMS analysis. Error bars are the SD where n=3.
181
6.3.2.2 Experiment B
Experimental conditions for this experiment differed from those in experiment A; conditions
were optimized for greater production of OA (see Section 6.2.3.2.). Due to the number of
samples tested and the necessity of processing the assays rapidly, a no-OAC control assay was
not performed. All enzymes were independently expressed and purified for this experiment. The
amount of HTAL collected from the assays ranged between 32 ± 7 pmol and 48 ± 16 pmol, and
appeared static throughout the assays (Figure 6.9A). The largest amount of HTAL was detected
from wtOAC assays, whereas the least amount was detected from both H57A and Y72A assays
(Figure 6.9A). The amount of PDAL produced during each assay also appear static, where the
least amount detected was 474 ± 39 pmol from D96A assays, and the largest amount was 560 ±
70 pmol during H5A assays (Figure 6.9B). Similar to experiment A, OA was detected in assays
using wtOAC, K12A, G43A, and D96A constructs, where they accumulated 360 ± 134 pmol,
179 ± 16 pmol, 35 ± 13 pmol, and 23 ± 2 pmol, respectively. It was additionally detected in
assays using the K4A, K38A, and D45A mutantsm, which produced 22 ± 8 pmol, 13 ± 5 pmol,
258 ± 25 pmol, respectively (Figure 6.9C). A trace amount of OA was also detected from H75A,
which accumulated 4 ± 2 pmol. The dynamic behavior of olivetol production was also evident
during this experiment, where the smallest amount detected was obtained from D45A, which
accumulated 61 ± 18 pmol, and the largest amount was detected from H5A, which accumulated
402 ± 49 pmol (Figure 6.9D).
The amounts of OA detected during experiments A and B, as well as the relative activity
compared to wtOAC are tabulated in Table 6.3. The location of amino acids targeted for SDM
are shown on the OAC model designed from the 1Q53 template (Figures 6.2 and 6.10).
182
amount (pmol)
A
700
600
500
400
300
200
100
0
C
500
amount (pmol)
amount (pmol)
B
70
60
50
40
30
20
10
0
400
HTAL
PDAL
OA
300
200
100
0
amount (pmol)
D 500
400
olivetol
300
200
100
0
Figure 6.9 LC results for OAC mutant assays - Experiment B.
Product yields for (A) HTAL, (B) PDAL, (C) OA, and (D) olivetol biosynthesized during
TKS/OAC assays conducted in the presence of wild-type (wt) and mutant OAC constructs was
determined by LCMS analysis. Error bars are the SD where n=3.
183
Table 6.3 Successful OA biosynthesis by OAC point mutants
OAC
version
wt*
K4A
K12A
K38A
G43A
D45A
H75A
D96A
Experiment A
OA
OAC
(pmol) activity
167
100%
n/a
n/a
42
25.1%
n/a
n/a
14
8.3%
n/a
n/a
0
0%
32
19.2%
Experiment B
OA
OAC
(pmol) activity
360
100%
22
6.1%
179
49.7%
13
3.6%
35
9.7%
258
71.7%
4
1.1%
23
6.4%
Average relative
OAC activity
100%
n/a
37.4%
n/a
9.0%
n/a
0.6%
12.8%
*wt is the unmutated OAC. n/a samples were not tested.
Figure 6.10 Location of SDM target sites mapped on the OAC model.
The 15 amino acids that were targeted in this study have been mapped onto the OAC model
derived from the 1Q53 template. Sites are either mapped on or next to the actual location. The
L30 site shown in grey exists in the α-helix drawn near the back of the figure.
184
6.3.3 MS/MS Monitoring of the TKS and TKS/OAC Reactions
In an attempt to detect the presence of a CoA-linked polyketide intermediate (Figure 6.1), the
TKS and TKS/OAC reactions were monitored using tandem mass spectrometry. Specifically,
both precursor ion scanning and selected reaction monitoring modes were used to screen the
reaction mixture for CoA-linked intermediates (de Hoffmann, 1996; Glish and Vachet, 2003).
Enzyme assays were analyzed using tandem mass spectrometry during multiple reaction
monitoring (MRM) for parent ion and daughter ion pairs resulting from neutral loss of 507 amu,
as well as 428 amu which are attributed to CoA tethered compounds of interest. Using these
methods, the dissipation and generation of reaction substrates and CoA compounds was
observed.
A parent ion scan was used to analyze: 1) a blank sample containing the reaction mixture
but no enzymes, 2) a TKS reaction, and 3) a TKS/OAC reaction (Figure 6.11). All ions of
interest included free CoA (768 m/z), acetyl-CoA (810 m/z), malonyl-CoA (854 m/z), hexanoylCoA (866 m/z), proposed intermediate (992 m/z), and formyl-CoA (796 m/z). Reactions were
also analyzed using spectrometry in MRM mode and resulted in the same trends as those
appearing during parent ion scans (Figure 6.12). Although reaction blanks did not noticeably
change over time, TKS and TKS/OAC reactions accumulated large amounts of free CoA and
acetyl-CoA, associated with a loss of malonyl-CoA and hexanoyl-CoA. Although not visible in
Figure 6.12, trace amounts of the proposed tetraketide intermediate were detected (Table 6.4).
185
Figure 6.11 Parent ion scan of TKS and TKS/OAC reactions.
MS analysis of CoA-linked compounds was conducted on (A) a reaction blank, (B) a 60 min
TKS only reaction, and (C) a 60 min TKS/OAC reaction. Parent ion scans detected fragments of
428 m/z.
186
ion abundance
1.0E+06
9.0E+05
8.0E+05
7.0E+05
6.0E+05
5.0E+05
4.0E+05
3.0E+05
2.0E+05
1.0E+05
0.0E+00
TKS
TKS + OAC
reaction blank
Figure 6.12 Multiple Reaction Monitoring of TKS and TKS/OAC reactions.
TKS and TKS/OAC reactions were analyzed by MS for CoA-linked compounds. Results are
plotted as the ion abundance following 1 h incubation at 37°C.
Table 6.4 Area response obtained during the MRM analysis of TKS assays
AREA
sample
mal-CoA
hex-CoA
CoA (free)
acetyl-CoA
intermediate-CoA
formyl-CoA
TKS
1435
193633
686457
272484
449
9
TKS + OAC
12047
206387
713452
256350
471
11
reaction blank
497697
933147
60587
28518
4
0
A TKS reaction and sample blank were also monitored for the neutral loss of CoA (507
m/z), resulting in the same trends detected previously during MS analyses (Figure 6.13).
Compared to sample blanks, TKS assays showed a 60% reduction in hexanoyl-CoA, and a 99%
reduction in malonyl-CoA associated with an increase of free CoA and acetyl-CoA. Neither a
tetraketide intermediate nor a hypothetical formyl-CoA byproduct were detected during MRM
screens.
187
Figure 6.13 MRM for the neutral loss of CoA during a TKS reaction.
Tandem mass spectrometry was used to analyze (left side) a sample blank containing no enzyme,
and (right side) a TKS reaction by screening for daughter ion pairs resulting from the neutral loss
of CoA fragments (507 m/z). Samples were screened for (A) a tetraketide CoA intermediate, (B)
hexanoyl-CoA, (C) malonyl-CoA, (D) acetyl-CoA, (E) formyl-CoA, and (F) free CoA. Results
are plotted as the % relative intensity relative to time (min).
188
6.4 Discussion
6.4.1 Primary and Secondary Structure
Results indicate that OAC is a DABB protein belonging to the SRABB family. Repeated BLAST
searches using the OAC primary sequence suggest a close relationship to several plant homologs.
These homologs were used to create an alignment that led to the identification of conserved
amino acids constituting an overall 18% coverage of the SRABB homologs. The high degree of
sequence similarity suggests a strong relationship exists between structure and function, likely
due to the smaller size of the proteins, resulting in the retention of their specific shape, which is
required for activity. OAC models predicted from AtHS1 and PtSP1 templates confirm that OAC
is a DABB protein made up of two alternating βαβ motifs. Specifically, the βαβ motifs appear
more like a βαββααβ, but note that the repeated αα helices are really one α helix that have
adopted a kink halfway through its span. This kink provides a necessary space between the αhelices and β-sheet previously proposed and shown to bind substrates in a number of other
DABB proteins including S. coelicolor ActVa-Orf6, S. galilaeus AknX, and S. aureus IsdG
(Chung et al., 2002; Lee et al., 2008; Sciara et al., 2003).
The 1Q53 and 1TR0 structures served as appropriate templates to model an OAC
structure. Lower-quality modeling using the 1Q53 template may have been due to the NMR
method of deduction used to solve its structure, which may have resulted in the exclusion of
some functions that the modelling software requires, since the software was primarily designed
for structures obtained by crystallography. The OAC model developed from the 1TR0 template
resulted in a similar OAC structure obtained from 1Q53, although less error was produced
possibly due to the methods of crystal elucidation used for 1TR0.
189
6.4.2 Site-Directed Mutagenesis
SDM-targeted amino acids, in most cases, resulted in the abolishment of OAC activity. Amino
acid targets themselves were based on their conservation within OAC homologs, but also placed
some priority on establishing whether or not they played a critical role in OAC activity. Some
amino acids were not suspected to play any role in activity, such as G43, yet their substitution
resulted in a dramatic loss of activity. Again, this suggests that a strong relationship exists
between structure and function in these proteins, probably due to its smaller size, resulting in
misfolding of mutant proteins upon their expression. The inclusion of K4, K38, and D45 during
the second round of experiments was an attempt to delineate their potential roles in either Schiff
base formation, which would implicate a class I aldolase mechanism, or through an acid/basecatalyzed reaction, which could involve an aspartate residue. The substitution of these residues,
as well as with K12 and D96, did not result in the total loss of function, which would have
otherwise suggested their direct involvement in the OAC mechanism. Instead, the K4A, K38A,
and D45A mutants, apart from G43A and H75A, were the only mutants that retained any OAC
activity, whereas all other mutants failed to produce OA when assayed with TKS.
The identification of OAC as a putative cyclase was originally based on its structural
similarity with another class of DABB proteins, the PKCs. The PKC family is exemplified by
TcmI, the only characterized PKC to date. TcmI is part of the tetracenomycin biosynthesis gene
cluster found in Streptomyces glaucescens and was shown to cyclize Tcm F2 to Tcm F1, both
intermediates involved in the biosynthesis of the antitumor antibiotic Tcm C, without requiring a
cofactor or metal ion (Shen and Hutchinson, 1993b). Since the structural elucidation of TcmI
lacks an enzyme-substrate complex, the active site was inferred to exist at an analogous α/β
interface as proposed in ActVa-Orf6 (Sciara et al., 2003). In TcmI, the proposed active site is
190
lined with many hydrophobic residues conducive to binding polyaromatic compounds, where a
single sulfate ion was shown bound to R40 and H26 at the back of the pocket. TcmI residues
(H26, D27, R40, and H51) that were putatively involved in catalysis were mutagenized, but
results indicate the residues are most likely involved in hydrogen bonding or polar interactions
and are not critical for enzyme activity. Although no specific mechanism was offered by
Thompson et al. (2004), they suggested that the role of TcmI was to position the substrate in a
way which accommodates for a water-facilitated reaction. I also postulate that the active site for
OAC exists in the analogous α/β interface proposed in other DABB proteins; similar to TcmI,
this interface also contains many hydrophobic residues that could potentially bind and aid in the
cyclization of a tetraketide intermediate (Figure 6.14).
The intramolecular aldol condensation performed by TcmI is also reminiscent of similar
cyclization reactions mediated by SnoaL-like polyketide cyclases (interpro domain IPR009959),
which also require neither Schiff-base formation nor a metal ion (Sultana et al., 2004). Although
not structurally related, S. nogalater SnoaL is a polyketide cyclase involved in the biosynthesis
of nogalaviketone, where a conserved D121 is thought to act as an acid/base catalyst. Notably,
this distinct SnoaL motif is sometimes found fused to ABM members and is also found upstream
from TcmI, where it is involved in an earlier polyketide cyclization. SRABB homologs,
including OAC, also contain a conserved aspartate residue (D96) analogous to the D121 residue
found in SnoaL. The D96A mutant was created to determine whether or not this residue was
involved in an acid/base mechanism critical for OAC activity, however assay results disqualify
this scenario. Based on its location and assay results, this residue may instead be involved in
either attracting the substrate to the active site cavity, or the stabilization of the dimer interface.
191
Figure 6.14 The proposed active site for OAC.
The proposed active site found between the α/β interface is lined with many hydrophobic
residues conducive to the potential binding of tetraketide intermediates. These amino acids also
include two tyrosine residues (oxygen marked in red) and two histidine residues (nitrogen
marked in blue) that could assist in cyclization.
Figure 6.15 The structure of SnoaL and TcmI.
(A) The structure for SnoaL (pdb 1SJW) has been shown to bind substrates between a β-sheet
and its α-helices, which has also been proposed to occur in (B) TcmI (pdb 1TUW), a PKC
belonging to a different family of proteins than SnoaL. Both enzymes perform aldol-like
reactions on polyketide-derived intermediates.
192
6.4.3 CoA analysis of the TKS reaction
The MS/MS MRM analyses were conducted to detect CoA-linked intermediates which could
help resolve TKS and OAC reaction mechanisms. Reactions were specifically monitored for the
presence of TKS substrates, as well as for putative reaction products and intermediates including
free CoA, acetyl-CoA, and a hypothetical tetraketo-CoA ester. Experiments also monitored
formyl-CoA, which could have indicated an exit strategy for CO2 during the formation of
olivetol, but this hypothetical model was abandoned after no formyl-CoA was detected. All
MS/MS analyses indicate that the TKS reactions proceeded until malonyl-CoA, but not
hexanoyl-CoA, was depleted. The analyses also show that acetyl-CoA is produced during TKS
reactions. This is likely due to the escape of enol intermediates from the TKS active site
following the decarboxylation of malonyl-CoA. Normally, this decarboxylation drives the
addition of acetyl groups onto a nascent polyketide (Austin and Noel, 2003), but suboptimal in
vitro conditions could result in the diffusion of labile ethenol-CoA esters into the surrounding
medium. The detection of acetyl-CoA may also raise questions as to why it is not being used to
produce truncated methyl resorcinol/ate analogs such as those observed during TKS assays that
used acetyl-CoA as a starter compound (Section 4.2.10.5), but this growing acetyl-CoA pool
would be in direct competition with hexanoyl-CoA in an ever dwindling pool of malonyl-CoA.
Although MRM for neutral loss of CoA was incapable of detecting any polyketide
intermediate (992 m/z), MS analyses using a SIR of 428 m/z successfully detected trace amounts
of a putative tetraketo-CoA ester during TKS and TKS/OAC reaction monitoring, which notably
was not detected in sample blanks. Although the trace detection of this intermediate using these
methods could be credited as a success, results are inconclusive. The detection of a tetraketo
CoA-ester may be limited due to its lability under harsh conditions created during MS ionization.
193
7. GENERAL DISCUSSION
7.1 Introduction
Throughout this study TKS assays conducted in the presence of malonyl-CoA and hexanoylCoA have repeatedly produced HTAL, PDAL and olivetol. The addition of OAC to this reaction
resulted in the additional biosynthesis of OA. The results have also shown that neither TKS nor
OAC are capable of producing OA without each other’s activity. As well, TKS and OAC do not
appear to physically interact, and OAC retains activity even when TKS was removed from
solution, therefore dismissing the notion that an allosteric relationship is involved between TKS
and OAC (Chapter 5). Additionally, results have indicated that TKS and OAC do not require a
metal, or Schiff base formation, dismissing the traditional class I and class II aldolase
mechanisms. Instead OAC activity is reminiscent of other unrelated polyketide cyclases, namely
SnoaL, which also does not rely on class I or class II aldolase mechanisms to achieve cyclization.
Instead, these results suggest that an intermediate is produced by TKS, which is released into the
medium and then cyclized to OA by OAC. It was demonstrated that, although TKS produces
HTAL, PDAL and olivetol, these compounds are unlikely the substrate for OAC (Chapter 4).
Attempts to detect an intermediate were equivocal, yet trace amounts of intermediate were
detectable in TKS and TKS/OAC assays when they were analyzed by tandem mass spectrometry
(Table 6.4), which were not present in reaction controls (Figure 6.13).
Although my observations do not provide conclusive evidence on either the TKS or OAC
mechanism, this study does provide the first evidence allowing model mechanisms to be
hypothesized. The simple most important question to address is waht is the substrate for OAC; o
sot a CoA-linked intermediate, or a free tetraketoacid? This question will be discussed in the
194
following section, first addressing the TKS reaction and its products, and then the OAC reaction
and how it utilizes the TKS product.
7.2 A Mechanistic View on the Synthesis of a Linear Tetraketide
TKS is involved in the production of not only a resorcinol, but also that of two lactone products,
HTAL and PDAL. The biosynthesis of lactones by type III PKS is not a new observation; many
instances of lactone production have been documented, where PKSs that are putatively involved
in the biosynthesis of natural products yield lactone products instead of the proposed end
products (Figure 2.7). As well, the production of lactones by PKSs is not only exclusive to type
III PKSs. It has been reported that the reconstruction of the modular type I PKS, DEBS, and the
following repositioning of the thioesterase domain from the terminus of module 6 to a preceding
module 2 resulted in the interruption of chain elongation and premature release of triketide
lactones (Staunton and Weissman, 2001). In their review, Staunton and Weissman (2001) also
note that many α-pyrones are produced during experiments using minimal or depleted PKSs, and
remark that the formation of these α-pyrones could provide a favorable mechanism for product
release. In contrast to type I PKSs, type III PKSs do not possess a transesterase domain, and the
issue of product release remains somewhat obscure (Austin and Noel, 2003; Jez et al., 2000). It
is my proposal that with regard to TKS, and in agreement with Staunton and Weismann, lactones
are artifacts of premature off-loading, but can also result during the re-loading of the active site
with a tetraketide that is incapable of further extension due to constraints in the active site.
Specifically, PDAL may be due to the premature off-loading of an enzyme-bound triketide. The
loss of a hydrogen at the C4 position results in enolate formation. This leads to a C5O C1
attack, resulting in the formation of PDAL, as well as its release from the enzyme (Figure 7.1).
On the other hand, HTAL biosynthesis results following the reattachment of a tetraketide to the
195
active site cysteine. This intermediate can either be released by hydrolysis as a free acid, or by
cyclization to HTAL following a C5O C1 attack (Figure 7.1). This proposal has been written
with consideration to current type III PKS models explaining polyketide biosynthetic activity, as
proposed by Jez et al (2000), who successfully elucidated the active site for 2–pyrone synthase in
Gerbera hybrida.
The proposed TKS mechanism involves the release of a CoA-tethered linear tetraketide.
This is an important conclusion which is supported by several observations. Firstly, product ratio
comparison between TKS assays conducted with or without OAC show that OA is produced at
the cost of olivetol production, but that HTAL production remains the same whether OAC is
present or not (Figure 4.7). The constant amount of HTAL synthesized indicates that, although a
common intermediate exists between olivetol and OA, HTAL production appears independent.
This phenomenon is explained using my proposed mechanism, which shows that HTAL results
from the reattachment of a CoA-tethered tetraketide to the TKS active site instead of its diffusion
into the surrounding solution (Figure 7.1). Additionally, studies by Harris and Carney (1966,
1967) that focused on the cyclization of polyketoacids (and not their CoA analogs) failed to
report the detection of any lactones during their experiments, indicating that pyrones cannot be
spontaneously formed from free polyketoacids and that their production results from either CoA
polyketoesters or, as I propose, due to enzyme binding of CoA-esters (Harris and Carney, 1966;
Harris and Carney, 1967). Factors that markedly affected lactone production included starter
compound availability (Chapter 3), which would directly affect competition for TKS active site
attachment and result in less HTAL production, as more hexanoyl-CoA becomes available, and
the adhering of TKS to Talon, which could potentially help to stabilize the movement and
architecture of the TKS active site during in vitro reaction.
196
Figure 7.1 A proposed mechanism for TKS.
(A) A starter molecule is attached to the TKS active site Cys157 resulting in the loss of CoA. (B) Acetyl-CoA is
activated following the decarboxylation of malonyl-CoA, (C) and attacks the starter’s C1 carbonyl (D) releasing the
starter from Cys157 and resulting in the formation of diketide. (E) This process is repeated by using the extended
polyketide as a starter compound, resulting in the formation of a triketide. Once the triketide is reattached to the
enzyme, the intermediate may either (F) spontaneously cyclize following the abstraction of hydrogen and a C5O
C1 attack, (G) releasing the lactone from Cys157, and (H) forming PDAL by tautomerization; or (I) be extended by
another two carbons forming a tetraketide, that may either be: (J) reattached to the enzyme, and form a lactone (F);
or (K) diffuse to the environment. A portion of this figure was adapted from Austin and Noel (2003).
197
7.3 The Biosynthesis of Olivetol and OA
If there does exist a diffusible, labile tetraketo-CoA ester, the mechanism for olivetol and OA
biosynthesis would still require the displacement of the CoA moiety. Olivetol biosynthesis itself
requires the hydrolysis of the CoA thioester, as well as a decarboxylation and, although we have
seen that the decarboxylation of cannabinoid acids occurs at higher temperatures and over
extended storage times, the in vitro decarboxylation of any polyketide in solution would drive
further chemical modification, such as cyclization. The biosynthesis of OA and olivetol are
substantially different in that the biosynthesis of olivetol is preceded by the hydrolysis of CoA,
whereas the biosynthesis of OA requires the retention of the CoA ester, which is hydrolyzed only
after the cyclization reaction is complete.
7.3.1 The Biosynthesis of Olivetol
The biosynthesis of olivetol requires a free tetraketoacid in solution, and is preceded by the
hydrolysis of the CoA ester prior to decarboxylation and cyclization of the tetraketoacid (Figure
7.2). Following hydrolysis, the loss of the acidic hydrogen from the carboxyl group subsequently
results in the decarboxylation of the tetraketoacid resulting in the loss of C1. The
decarboxylation results in the production of a short-lived enolate. The enolate is acts as a strong
nucleophile allowing a C2 C7 nucleophilic attack to proceed, resulting in the cyclization of
the intermediate to a phenolic alcohol. The phenolic alcohol is dehydrated, which results in the
biosynthesis of the tautomeric form of olivetol. Note that the hydrolysis of the CoA-tetraketoacid
leads to its spontaneous decarboxylation, which ultimately drives the cyclization to the final
product.
198
Figure 7.2 A proposed reaction mechanism for the synthesis of olivetol.
In this scheme, the R-substitutent is representing a pentyl chain, but may also occur in nature as a
shorter (propyl) or longer (heptyl) alkyl chain.
199
7.3.2 The Biosynthesis of OA
I proposed that the biosynthesis of OA requires a tetraketo-CoA ester, which makes the
tetraketide intermediate resilient to decarboxylation (Figure 7.3). Since CoA acts as a protecting
group which is itself susceptible to hydrolysis in aqueous solutions, the biosynthesis of OA also
requires OAC to anchor and retain the CoA ester during OA cyclization. The biosynthesis of OA
proceeds once the C2 α-hydrogen is abstracted by a base, resulting in the formation of an enolate
at the C1 position. This nucleophile does not hydrolyze CoA due to its position within OAC.
This enolate drives the C2 C7 cyclization, resulting in the formation of an alkoxide. The
alkoxide abstracts hydrogen (presumably from the acid/base that is initially involved) to form an
alcohol. Another α-hydrogen is abstracted from the C2 position, leading to the formation of the
C3 enolate, which is followed by the protonation of the alcohol (again from the acid/base that is
initially involved). The intermediate is then isomerized and dehydrated, resulting in the synthesis
of a CoA-tethered OA tautomer. The ester is hydrolyzed, liberating free CoA and OA (Figure
7.3).
This model for OA biosynthesis shows why OAC is required for proper cyclization, since
without this enzyme, tetraketo-CoA esters would simply hydrolyze and lead to the spontaneous
cyclization of olivetol. Classical aldol-type mechanisms for OAC may be excluded based on the
observations that alanine substitution of lysine residues only resulted in decreased OAC activity
and that OAC does not require a metal. Notably, OAC mutants harboring alanine substitutions at
histidine residues positioned at 5, 75, and 78, which are all located near the proposed active site,
result in the abolishment of OAC activity. Given that optimum OAC activity occurs at a pH of
6.4 (Chapter 4), and that the average pKa of hisitidine is 6.5, it seems likely that these histidine
residues play an important part during the acid/base reactions required to drive OA biosynthesis.
200
OAC
Figure 7.3 A proposed reaction mechanism for the biosynthesis of OA.
The biosynthesis of OA is preceded by the attachement of a tetraketo-CoA ester to OAC.
201
7.4 Cannabinoid Biosynthesis
A revised biosynthetic pathway for all cannabinoids produced in cannabis is here proposed.
Although the discovery of oxidoreductases involved in the cyclization of CBGA have
contributed to the delineation of later cannabinoid biosynthesis, including the biosynthesis of
THCA by THCA synthase (Taura et al., 1995), and for CBDA by CBDA synthase (Taura et al.,
1996), the earlier step involving OA biosynthesis has eluded description due to the lack of a PKS
capable of independently producing OA. The recruitment of OAC during this preliminary step
allows OA to be successfully biosynthesized, and together with TKS, the inclusion of OAC
clarifies the cannabinoid biosynthetic pathway (Figure 7.4). Given its activity and abundance in
trichome cDNA libraries, OAC is a key enzyme in the cannabinoid biosynthetic pathway.
7.5 Metabolic Engineering of Cannabinoid Biosynthesis
The engineering of secondary plant metabolites is an important field that has provided
economically important chemicals to be produced in higher yields that are more financially
worthwhile (Verpoorte et al., 2007). The discovery of OAC may play a role in the future
metabolic engineering of cannabis to either augment or abolish cannabinoid content. These
achievements would result in cannabis strains that have either increased medicinal value or,
conversely, being devoid of cannabinoids, in an agricultural benefit due to the illegality of
psychotropic cannabis. Whether it is used to increase or decrease cannabinoid content, OAC will
undoubtedly play a role in the future engineering of cannabis. A practical application of OAC
during in vitro syntheses will also benefit the future development of synthetic cannabinoid
biosynthesis. The in vitro biosynthesis of cannabinoids has been hampered in part by the lack of
OA production; the inclusion of OAC could ameliorate this problem by providing a suitable
substrate for GOT to produce CBGA.
202
Figure 7.4 A proposed biosynthesis of cannabinoids.
The biosynthesis of cannabinoids is preceded by the biosynthesis of OA by TKS and OAC.
203
7.6 DABB Cyclases May Play a Large Role in Plants
Given their abundance and conservation, OAC-like DABB proteins are possibly involved in a
variety of metabolic processes in plants other than cannabis. Although the occurrence of OA is
rare, other stilbene carboxylic acids in plants have been detected and are known to occur, such as
hydrangeic acid and lunularic acid in Hydrangea macrophylla (Zhang et al., 2009). Although one
study has attributed STCS activity as being responsible for the production of stilbene
carboxylates in Hydrangea, the lack of resolution and detection suggest another factor, or more
specifically another protein with a similar function as OAC, is required for the biosynthesis of
stilbene carboxylates (Eckermann et al., 2003). Similarly, the in vitro biosynthesis of derailed
pyrones instead of anticipated end products, such as in the case of anthrone biosynthesis in Aloe
arborescens, may require an OAC-like cyclase to fold large polyketide into their anticipated
form (Mizuuchi et al., 2009). Furthermore, the associations of OAC-like proteins with plant
stress and defense responses have been noted with regards to PtSP1 (Park et al., 2007; Zhu et al.,
2008). Given the defense-inducing effects of jasmonic acid, a plant hormone associated with
stress which is known to increase trichome production in plants, it is plausible that OAC-like
proteins may be involved in the mediation of unfavorable conditions, such as salt and heat stress,
in plants (Traw and Bergelson, 2003; van Schie et al., 2007). It is possible that OAC-like
proteins are involved in the production of bioactive stilbene carboxylic acids in response to plant
stresses in an attempt to produce antibiotic/antifungal/antiherbivore compounds in a more global
context (Park et al., 2007).
204
7.7 Conclusion
This study set out to determine why OA was not being produced by a type III PKS implicitly
involved in cannabinoid biosynthesis. The hypothesis that a putative cyclase was involved led to
the discovery of OAC, a 12 kDa DABB protein belonging to the family of SRABB plant proteins
of unknown function. The biochemical characterization of TKS and OAC has shown that,
although TKS is capable of producing HTAL, PDAL, and olivetol, the inclusion of OAC during
in vitro assays is required for the successful biosynthesis OA. Furthermore, TKS and OAC
activities are physically independent. It is proposed that TKS generates a linear polyketide that is
subsequently cyclized either spontaneously to a pyrone or olivetol as a result of enzyme product
release, or the hydrolytic release of CoA, respectively. The fate of CoA-linked tetraketide
intermediates is influenced in the presence of OAC, which is able to cyclize the intermediate to
its proper resorcylic acid without a concurrent decarboxylation. In addition to proposing the
mechanistic polyketide origins of OA and olivetol, this work also suggests a plausible
mechanism for biosynthesis of lactones, alternatively known as derailment products, by type III
PKS. The discovery of OAC also attributes the first known function to any protein belonging to
the SRABB protein family. Lastly, this discovery resolves the enigmatic step of OA biosynthesis
which is required during the biosynthesis of cannabinoids.
7.8 Future Directions
The discovery of OAC and its involvement in polyketide biosynthesis will undoubtedly play an
important role in the identification and characterization of other OAC-like proteins in the future.
The possibililty that other plant ‘type II’ PKCs are involved in type III PKS product cyclization
is itself a novel finding that should generate a fresh perspective regarding plant natural product
biosynthesis. It will be interesting to see if other plant OAC-like proteins are involved in
205
polyketide cyclizations. Regardless, OAC itself is an interesting protein, being the first plant
polyketide cyclase that has ever been identified. Much more work will be required to clarify its
mechanisms, which will necessitate more structure/function studies that will hopefully target and
substitute other amino acids putatively involved in OAC activity. Also, the structural
characterization of OAC should be conducted, either by NMR or by X-ray crystallography, to
determine with highest certainty the positions and orientation of OAC’s structural elements.
It will also be interesting to see if OAC is capable of participating in cyclization reactions
not related to cannabinoid biosynthesis. For example, the effect that OAC has when it is assayed
with other interesting plant PKSs, such as Aloe arborescens octaketide synthase, which
accumulates SEK4 and SEK4b instead of chrysophanol anthrone, would be very interesting to
see (Figure 2.7). Similarly, the role that OAC could have during other cyclization reactions that
do not occur in plants, but in fungi and streptomycetes, will also be interesting to perform, since
this could generate a plethora of new compounds, which could be of extreme medical benefit as
potentially new antibiotics or anti-tumor compounds. It will be interesting to see how the role of
OAC plays out in the future.
206
8. REFERENCES
Abe,I., Utsumi,Y., Oguro,S., Morita,H., Sano,Y., and Noguchi,H. (2005). A plant type III
polyketide synthase that produces pentaketide chromone. J Am Chem Soc 127, 1362-1363.
Abel,E.L. (1980). Marihuana, the first twelve thousand years. (Michigan: Plenum Press).
Adams,M.A. and Jia,Z. (2005). Structural and biochemical evidence for an enzymatic quinone
redox cycle in Escherichia coli: identification of a novel quinol monooxygenase. J Biol Chem
280, 8358-8363.
Adams,R. (1940). Marihuana. Science 92, 115-119.
Adams,R., Hunt,M., and Clark,J.H. (1940a). Structure of cannabidiol, a product isolated from the
marihuana extract of Minnesota Wild Hemp. J Am Chem Soc 62, 196-200.
Adams,R., Pease,D.C., Cain,C.K., and Clark,J.H. (1940b). Structure of cannabidiol. VI.
Isomerization of cannabidiol to tetrahydrocannabinol, a physiologically active product.
Conversion of cannabidiol to cannabinol. J Am Chem Soc 62, 2405-2408.
Ainsworth,C. (2000). Boys and girls come out to play: the molecular biology of dioecious plants.
Ann Bot London 86, 211-221.
Altschul,S.F., Madden,T.L., Schaffer,A.A., Zhang,J., Zhang,Z., Miller,W., and Lipman,D.J.
(1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.
Nucleic Acids Res 25, 3389-3402.
Amar,MB. (2006). Cannabinoids in medicine: A review of their therapeutic potential. J
Ethnopharmacol 105, 1-25.
Ames,B.D., Korman,T.P., Zhang,W., Smith,P., Vu,T., Tang,Y., and Tsai,S.C. (2008). Crystal
structure and functional analysis of tetracenomycin ARO/CYC: implications for cyclization
specificity of aromatic polyketides. Proc Natl Acad Sci USA 105, 5349-5354.
Ames,B.D., Lee,M.Y., Moody,C., Zhang,W., Tang,Y., and Tsai,S.C. (2011). Structural and
biochemical characterization of ZhuI aromatase/cyclase from the R1128 polyketide pathway.
Biochemistry 50, 8392-8406.
Armaleo,D., Sun,X., and Culberson,C. (2011). Insights from the first putative biosynthetic gene
cluster for a lichen depside and depsidone. Mycologia 103, 741-754.
Arnold,K., Bordoli,L., Kopp,J., and Schwede,T. (2006). The SWISS-MODEL workspace: a
web-based environment for protein structure homology modelling. Bioinformatics 22, 195-201.
207
Austin,M.B., Bowman,M.E., Ferrer,J.L., Schröder,J., and Noel,J.P. (2004). An aldol switch
discovered in stilbene synthases mediates cyclization specificity of type III polyketide synthases.
Chem Biol 11, 1179-1194.
Austin,M.B. and Noel,J.P. (2003). The chalcone synthase superfamily of type III polyketide
synthases. Nat Prod Rep 20, 79-110.
Ausubel,F.M., Brent,R., Kingston,R.E., Moore,D.D., Seidman,J.G., Smith,J.A., and Struhl,K.
(1994). Current protocols in molecular biology. (New York: Greene Publishing Associates and
Wiley-Interscience).
Baerson,S.R., Schröder,J., Cook,D., Rimando,A.M., Pan,Z., Dayan,F.E., Noonan,B.P., and
Duke,S.O. (2010). Alkylresorcinol biosynthesis in plants: new insights from an ancient enzyme
family? Plant Signal Behav 5, 1286-1289.
Bateman,A., Coin,L., Durbin,R., Finn,R.D., Hollich,V., Griffiths-Jones,S., Khanna,A.,
Marshall,M., Moxon,S., Sonnhammer,E.L., Studholme,D.J., Yeats,C., and Eddy,S.R. (2004).
The Pfam protein families database. Nucleic Acids Res 32, D138-D141.
Benson,D.A., Karsch-Mizrachi,I., Lipman,D.J., Ostell,J., Rapp,B.A., and Wheeler,D.L. (2000).
GenBank. Nucleic Acids Res 28, 15-18.
Bentley,R. and Zwitkowits,P.M. (1967). Biosynthesis of tropolones in Penicillium stipitatum.
VII. The formation of polyketide lactones and other nontropolone compounds as a result of
ethionine inhibition. J Am Chem Soc 89, 676-680.
Bingman,C.A., Johnson,K.A., Peterson,F.C., Frederick,R.O., Zhao,Q., Thao,S., Fox,B.G.,
Volkman,B.F., Jeon,W.B., Smith,D.W., Newman,C.S., Ulrich,E.L., Hegeman,A., Sussman,M.R.,
Markley,J.L., and Phillips,G.N., Jr. (2004). Crystal structure of the protein from gene At3g17210
of Arabidopsis thaliana. Proteins 57, 218-220.
Breithaupt,C., Kurzbauer,R., Lilie,H., Schaller,A., Strassner,J., Huber,R., Macheroux,P., and
Clausen,T. (2006). Crystal structure of 12-oxophytodienoate reductase 3 from tomato: selfinhibition by dimerization. Proc Natl Acad Sci USA 103, 14337-14342.
Brown,I., Cascio,M.G., Rotondo,D., Pertwee,R.G., Heys,S.D., and Wahle,K.W. (2012).
Cannabinoids and omega-3/6 endocannabinoids as cell death and anticancer modulators. Prog
Lipid Res 52, 80-109.
Bruce,N.C., Cain,R.B., Pieper,D.H., and Engesser,K.H. (1989). Purification and characterization
of 4-methylmuconolactone methyl-isomerase, a novel enzyme of the modified 3-oxoadipate
pathway in nocardioform actinomycetes. Biochem J 262, 303-312.
Cahn,R.S. (1932). Cannabis indica resin.Part III.The constitution of cannabinol. J Chem Soc
1342-1353.
Callaway,J.C. (2004). Hempseed as a nutritional resource: An overview. Euphytica 140, 65-72.
208
Camara,B., Bielecki,P., Kaminski,F., dos,S., V, Plumeier,I., Nikodem,P., and Pieper,D.H.
(2007). A gene cluster involved in degradation of substituted salicylates via ortho cleavage in
Pseudomonas sp. strain MT1 encodes enzymes specifically adapted for transformation of 4methylcatechol and 3-methylmuconate. J Bacteriol 189, 1664-1674.
Castoe,T.A., Stephens,T., Noonan,B.P., and Calestani,C. (2007). A novel group of type I
polyketide synthases (PKS) in animals and the complex phylogenomics of PKSs. Gene 392, 4758.
Childs,K.L., Hamilton,J.P., Zhu,W., Ly,E., Cheung,F., Wu,H., Rabinowicz,P.D., Town,C.D.,
Buell,C.R., and Chan,A.P. (2007). The TIGR Plant Transcript Assemblies database. Nucleic
Acids Res 35, D846-D851.
Chung,J.Y., Fujii,I., Harada,S., Sankawa,U., and Ebizuka,Y. (2002). Expression, purification,
and characterization of AknX anthrone oxygenase, which is involved in aklavinone biosynthesis
in Streptomyces galilaeus. J Bacteriol 184, 6115-6122.
Conte,M.G., Gaillard,S., Lanau,N., Rouard,M., and Perin,C. (2008). GreenPhylDB: a database
for plant comparative genomics. Nucleic Acids Res 36, D991-D998.
Crosby,J., Treadwell,T., Hammerton,M., Vasilakis,K., Crump,M.P., Williams,D.S., and Mann,S.
(2012). Stabilization and enhanced reactivity of actinorhodin polyketide synthase minimal
complex in polymer-nucleotide coacervate droplets. Chem Commun (Camb) 48, 11832-11834.
Czichi,U. and Kindl,H. (1977). Phenylalanine ammonia lyase and cinnamic acid hydroxylases as
assembled consecutive enzymes on microsomal membranes of cucumber cotyledons:
cooperation and subcellular distribution. Planta 134, 133-143.
Dayanandan,P. and Kaufman,P.B. (1976). Trichomes of Cannabis sativa L. (Cannabaceae). Am
J Bot 63, 578-591.
de Hoffmann,E. (1996). Tandem mass spectrometry: a primer. J Mass Sprectrom 31, 129-137.
de Jesus,A.E., Hull,W.E., Steyn,P.S., van Heerden,F.R., and Vleggaar,R. (1982). Biosynthesis of
viridicatumtoxin, a mycotoxin from Penicilium expansum. J Chem Soc Chem Comm 0, 902-904.
Debnam,P.M., Shearer,G., Blackwood,L., and Kohl,D.H. (1997). Evidence for channeling of
intermediates in the oxidative pentose phosphate pathway by soybean and pea nodule extracts,
yeast extracts, and purified yeast enzymes. Eur J Biochem 246, 283-290.
Devane,W.A., Dysarz,F.A., III, Johnson,M.R., Melvin,L.S., and Howlett,A.C. (1988).
Determination and characterization of a cannabinoid receptor in rat brain. Mol Pharmacol 34,
605-613.
Devane,W.A., Hanus,L., Breuer,A., Pertwee,R.G., Stevenson,L.A., Griffin,G., Gibson,D.,
Mandelbaum,A., Etinger,A., and Mechoulam,R. (1992). Isolation and structure of a brain
constituent that binds to the cannabinoid receptor. Science 258, 1946-1949.
209
Dgany,O., Gonzalez,A., Sofer,O., Wang,W., Zolotnitsky,G., Wolf,A., Shoham,Y., Altman,A.,
Wolf,S.G., Shoseyov,O., and Almog,O. (2004). The structural basis of the thermostability of
SP1, a novel plant (Populus tremula) boiling stable protein. J Biol Chem 279, 51516-51523.
Dobritsa,A.A., Lei,Z., Nishikawa,S.i., Urbanczyk-Wochniak,E., Huhman,D.V., Preuss,D., and
Sumner,L.W. (2010). LAP5 and LAP6 encode anther-specific proteins with similarity to
chalcone synthase essential for pollen exine development in Arabidopsis. Plant Physiol 153, 937955.
Earley,K.W., Haag,J.R., Pontes,O., Opper,K., Juehne,T., Song,K., and Pikaard,C.S. (2006).
Gateway-compatible vectors for plant functional genomics and proteomics. Plant J 45, 616-629.
Eckermann,C., Schröder,G., Eckermann,S., Strack,D., Schmidt,J., Schneider,B., and Schröder,J.
(2003). Stilbenecarboxylate biosynthesis: a new function in the family of chalcone synthaserelated proteins. Phytochemistry 62, 271-286.
Eckermann,S., Schröder,G., Schmidt,J., Strack,D., Edrada,R.A., Helariutta,Y., Elomaa,P.,
Kotilainen,M., Kilpelainen,I., Proksch,P., Teeri,T.H., and Schröder,J. (1998). New pathway to
polyketides in plants. Nature 396, 387-390.
Elix,J.A. (1974). Synthesis of para-olivetol depsides. Aust J Chem 27, 1767-1779.
Elix,J.A. (1994). New depsides from the lichen Lecanora planaica. Aust J Chem 47, 1199-1203.
Fellermeier,M., Eisenreich,W., Bacher,A., and Zenk,M.H. (2001). Biosynthesis of cannabinoids.
Incorporation experiments with (13)C-labeled glucoses. Eur J Biochem 268, 1596-1604.
Fellermeier,M. and Zenk,M.H. (1998). Prenylation of olivetolate by a hemp transferase yields
cannabigerolic acid, the precursor of tetrahydrocannabinol. FEBS Lett 427, 283-285.
Felsenstein,J. (1985). Confidence limits on phylogenies: An approach using the bootstrap.
Evolution 39, 783-791.
Fesner,W.D., Schneider,A., Held,H., Sinerius,G., Walter,C., Hixon,M., Schloss,J.V. (1996). The
Mechanism of Class II, Metal-Dependent Aldolases. Angewandte Chemie International Edition
in English 35, 2219-2221
Fetterman,P.S., Keith,E.S., Waller,C.W., Guerrero,O., Doorenbos,N.J., and Quimby,M.W.
(1971). Mississippi-grown Cannabis sativa L preliminary observation on chemical definition of
phenotype and variations in tetrahydrocannabinol content versus age, sex, and plant part. J
Pharm Sci 60, 1246-1249.
Fleming,M.P. and Clarke,R.C. (1998). Physical evidence for the antiquity of Cannabis sativa L.
Journal of Industrial Hemp 5, 80-92.
Flores-Sanchez,I.J., Linthorst,H.J., and Verpoorte,R. (2010). In silicio expression analysis of
PKS genes isolated from Cannabis sativa L. Genet Mol Biol 33, 703-713.
210
Flores-Sanchez,I.J. and Verpoorte,R. (2008a). PKS activities and biosynthesis of cannabinoids
and flavonoids in Cannabis sativa L plants. Plant Cell Physiol 49, 1767-1782.
Flores-Sanchez,I.J. and Verpoorte,R. (2008b). Secondary metabolism in Cannabis. Phytochem
Rev. 7, 615-639.
Fridman,E. and Pichersky,E. (2005). Metabolomics, genomics, proteomics, and the identification
of enzymes and their substrates and products. Curr Opin Plant Biol 8, 242-248.
Funa,N., Awakawa,T., and Horinouchi,S. (2007). Pentaketide resorcylic acid synthesis by type
III polyketide synthase from Neurospora crassa. J Biol Chem 282, 14476-14481.
Gagne,S.J., Stout,J.M., Liu,E., Boubakir,Z., Clark, S.M., and Page,J.E. (2012) Identification of
olivetolic acid cyclase from Cannabis sativa reveals a unique catalytic route to plant polyketides.
Proc Natl Acad Sci USA 109, 12811-12816
Gang,D.R., Wang,J., Dudareva,N., Nam,K.H., Simon,J.E., Lewinsohn,E., and Pichersky,E.
(2001). An investigation of the storage and biosynthesis of phenylpropenes in sweet basil. Plant
Physiol 125, 539-555.
Gaoni,Y. and Mechoulam,R. (1964). Isolation, structure, and partial synthesis of an active
constituent of hashish. J Am Chem Soc 86, 1646-1647.
Gaoni,Y. and Mechoulam,R. (1971). The isolation and structure of delta-1-tetrahydrocannabinol
and other neutral cannabinoids from hashish. J Am Chem Soc 93, 217-224.
Gensheimer,M. and Mushegian,A. (2004). Chalcone isomerase family and fold: no longer unique
to plants. Protein Sci 13, 540-544.
Gershenzon,J., McCaskill,D., Rajaonarivony,J.I., Mihaliak,C., Karp,F., and Croteau,R. (1992).
Isolation of secretory cells from plant glandular trichomes and their use in biosynthetic studies of
monoterpenes and other gland products. Anal Biochem 200, 130-138.
Gijsen,H.J., Qiao,L., Fitz,W., and Wong,C.H. (1996). Recent advances in the chemoenzymatic
synthesis of carbohydrates and carbohydrate mimetics. Chem Rev 96, 443-474.
Glish,G.L. and Vachet,R.W. (2003). The basics of mass spectrometry in the twenty-first century.
Nat Rev Drug Discov 2, 140-150.
Goblirsch,B.R., Streit,B.R., Dubois,J.L., and Wilmot,C.M. (2010). Structural features promoting
dioxygen production by Dechloromonas aromatica chlorite dismutase. J Biol Inorg Chem 15,
879-888.
Gomori,G. (1955). Preparation of buffers for use in enzyme studies. Methods Enzymol 1, 138146.
211
Goodstein,D.M., Shu,S., Howson,R., Neupane,R., Hayes,R.D., Fazo,J., Mitros,T., Dirks,W.,
Hellsten,U., Putnam,N., and Rokhsar,D.S. (2012). Phytozome: a comparative platform for green
plant genomics. Nucleic Acids Res 40, D1178-D1186.
Gorham,J. (1995). The biochemistry of the stilbenoids. (London: Chapman and Hall).
Griesebach H, Fritsch H. (1975). Biosynthesis of cyanidin in cell cultures of Haplopappus
gracilis. Phytochemistry 14, 2437-2442.
Guilford,H., Scott,A.I., Skingle,D., and Yalpani,M. (1968). The synthesis of tetra-acetic acid
lactone and a model for the biosynthesis of 6-methylsalicyclic acid. Chem Commun (London)
1127-1128.
Hadfield,A.T., Limpkin,C., Teartasin,W., Simpson,T.J., Crosby,J., and Crump,M.P. (2004). The
crystal structure of the ActIII actinorhodin polyketide reductase: proposed mechanism for ACP
and polyketide binding. Structure 12, 1865-1875.
Hammond,C.T. and Mahlberg,P.G. (1973). Morphology of glandular hairs of Cannabis sativa
from scanning electron microscopy. Am J Bot 60, 524-528.
Hammond,C.T. and Mahlberg,P.G. (1978). Ultrastructural development of capitate glandular
hairs of Cannabis Sativa L. (Cannabaceae). Am J Bot 65, 140-151.
Harris,T.M. and Carney,R.L. (1966). Biogenetically modeled synthesis of -resorcylic acids. J
Am Chem Soc 88, 2053-2054.
Harris,T.M. and Carney,R.L. (1967). Synthesis of 3,5,7-triketo acids and esters and their
cyclizations to resorcinol and phloroglucinol derivatives. Models of biosynthesis of phenolic
compounds. J Am Chem Soc 89, 6734-6741.
Harvey,D.J. (1976). Characterization of the butyl homologues of delta1-tetrahydrocannabinol,
cannabinol and cannabidiol in samples of cannabis by combined gas chromatography and mass
spectrometry. J Pharm Pharmacol 28, 280-285.
Hassani,I. and Razzoul,H. (2005). Olivetol: constituent of lichen Evernia Prunastri Ach.
Physical and Chemical News 26, 98-103.
Hathway,D.E. and Seakin,J.W. (1959). Hydroxystilbenes of Eucalyptus wandoo. Biochem J 72,
369-374.
Hill,A.M. and Staunton,J. (2010). 1.10 - Type I modular PKS. In Comprehensive Natural
Products II, M.Editors-in-Chief:Â Â Lew and B.Hung-Wen, eds. (Oxford: Elsevier), pp. 385452.
Horper,W. and Marner,F.J. (1995). Phenols and quinones from leaves of Primula obconica. Nat
Prod Lett 6, 163-170.
212
Horper,W. and Marner,F.J. (1996). Biosynthesis of primin and micodnidin and its derivatives.
Phytochemistry 41, 451-456.
Hrazdina,G., Wagner,G.J., Siegelman,H.W. (1978). Subcellular localization of enzymes of
anthocyanin biosynthesis in protoplasts. Phytochemistry 17, 53-56.
Iijima,Y., Davidovich-Rikanati,R., Fridman,E., Gang,D.R., Bar,E., Lewinsohn,E., and
Pichersky,E. (2004). The biochemical and molecular basis for the divergent patterns in the
biosynthesis of terpenes and phenylpropenes in the peltate glands of three cultivars of basil. Plant
Physiol 136, 3724-3736.
Intrieri,M.C. and Buiatti,M. (2001). The horizontal transfer of Agrobacterium rhizogenes genes
and the evolution of the genus Nicotiana. Mol Phylogenet Evol 20, 100-110.
Isbell,H. (1973). Research on Cannabis (marijuana). UN Bulletin on Narcotics 25, 37-48.
Jenke-Kodama,H., Sandmann,A., Muller,R., and Dittmann,E. (2005). Evolutionary implications
of bacterial polyketide synthases. Mol Biol Evol 22, 2027-2039.
Jez,J.M., Austin,M.B., Ferrer,J., Bowman,M.E., Schröder,J., and Noel,J.P. (2000). Structural
control of polyketide formation in plant-specific polyketide synthases. Chem Biol 7, 919-930.
Jez,J.M., Ferrer,J.L., Bowman,M.E., Austin,M.B., Schröder,J., Dixon,R.A., and Noel,J.P. (2001).
Structure and mechanism of chalcone synthase-like polyketide synthases. J Ind Microbiol
Biotechnol 27, 393-398.
Jiang,H.E., Li,X., Zhao,Y.X., Ferguson,D.K., Hueber,F., Bera,S., Wang,Y.F., Zhao,L.C.,
Liu,C.J., and Li,C.S. (2006). A new insight into Cannabis sativa (Cannabaceae) utilization from
2500-year-old Yanghai Tombs, Xinjiang, China. J Ethnopharmacol 108, 414-422.
Jin,W. and Zjawiony,J.K. (2006). 5-alkylresorcinols from Merulius incarnatus. J Nat Prod 69,
704-706.
Jirschitzka,J., Mattern,D.J., Gershenzon,J., and D'Auria,J.C. (2013). Learning from nature: new
approaches to the metabolic engineering of plant defense pathways. Curr Opin Biotechnol 24,
320-328.
Jones,D.T. (1999). Protein secondary structure prediction based on position-specific scoring
matrices. J Mol Biol 292, 195-202.
Jones,D.T., Taylor,W.R., and Thornton,J.M. (1992). The rapid generation of mutation data
matrices from protein sequences. Comput Appl Biosci 8, 275-282.
Jones, M. Isolation of plasmid DNA from yeast using the QIAprep® Spin Miniprep Kit. 2000.
(PR04.doc Oct-01), QIAGEN.
213
Jones,T.H., Brunner,S.R., Edwards,A.A., Davidson,D.W., and Snelling,R.R. (2005). 6alkylsalicylic acids and 6-alkylresorcylic acids from ants in the genus Crematogaster from
Brunei. J Chem Ecol 31, 407-417.
Karppinen,K. and Hohtola,A. (2008). Molecular cloning and tissue-specific expression of two
cDNAs encoding polyketide synthases from Hypericum perforatum. J Plant Physiol 165, 10791086.
Keatinge-Clay,A.T. (2012). The structures of type I polyketide synthases. Nat Prod Rep 29,
1050-1073.
Keene,C.K. and Wagner,G.J. (1985). Direct demonstration of duvatrienediol biosynthesis in
glandular heads of tobacco trichomes. Plant Physiol 79, 1026-1032.
Kendrew,S.G., Hopwood,D.A., and Marsh,E.N. (1997). Identification of a monooxygenase from
Streptomyces coelicolor A3(2) involved in biosynthesis of actinorhodin: purification and
characterization of the recombinant enzyme. J Bacteriol 179, 4305-4310.
Kim,E.S. and Mahlberg,P.G. (1997). Plastid development in disc cells of glandular trichomes of
Cannabis (Cannabaceae). Mol Cells 7, 352-359.
Kim,M.I., Kwon,S.J., and Dordick,J.S. (2009). In vitro precursor-directed synthesis of polyketide
analogues with coenzyme a regeneration for the development of antiangiogenic agents. Org Lett
11, 3806-3809.
Kimura,M. and Okamoto,K. (1970). Distribution of tetrahydrocannabinolic acid in fresh wild
cannabis. Experientia 26, 819-820.
Kozubek,A. and Tyman,J.H. (1999). Resorcinolic lipids, the natural non-isoprenoid phenolic
amphiphiles and their biological activity. Chem Rev 99, 1-26.
Lee,W.C., Reniere,M.L., Skaar,E.P., and Murphy,M.E. (2008). Ruffling of metalloporphyrins
bound to IsdG and IsdI, two heme-degrading enzymes in Staphylococcus aureus. J Biol Chem
283, 30957-30963.
Loewe,S. (1945). Marihuana activity of cannabinol. Science 14, 615-616.
Ma,S.M., Zhan,J., Xie,X., Watanabe,K., Tang,Y., and Zhang,W. (2007). Redirecting the
cyclization steps of fungal polyketide synthase. J Am Chem Soc 130, 38-39.
Mandolino,G., Carboni,A., Bagatta,M., Moliterni,V.M., and Ranalli,P. (2002). Occurrence and
frequency of putatively Y chromosome linked DNA markers in Cannabis sativa L. Euphytica
126, 211-218.
Marchler-Bauer,A., Lu,S., Anderson,J.B., Chitsaz,F., Derbyshire,M.K., Weese-Scott,C.,
Fong,J.H., Geer,L.Y., Geer,R.C., Gonzales,N.R., Gwadz,M., Hurwitz,D.I., Jackson,J.D., Ke,Z.,
Lanczycki,C.J., Lu,F., Marchler,G.H., Mullokandov,M., Omelchenko,M.V., Robertson,C.L.,
214
Song,J.S., Thanki,N., Yamashita,R.A., Zhang,D., Zhang,N., Zheng,C., and Bryant,S.H. (2011).
CDD: a conserved domain database for the functional annotation of proteins. Nucleic Acids Res
39, D225-D229.
Marks,M.D., Tian,L., Wenger,J.P., Omburo,S.N., Soto-Fuentes,W., He,J., Gang,D.R.,
Weiblen,G.D., and Dixon,R.A. (2009). Identification of candidate genes affecting 9tetrahydrocannabinol biosynthesis in Cannabis sativa. J Exp Bot 60, 3715-3726.
Martinez,M. (2011). Plant protein-coding gene families: emerging bioinformatics approaches.
Trends Plant Sci 16, 558-567.
Matousek,J., Vrba,L., Novak,P., Patzak,J., De,K.J., Skopek,J., Heyerick,A., Roldan-Ruiz,I., and
De,K.D. (2005). Cloning and molecular analysis of the regulatory factor HlMyb1 in hop
(Humulus lupulus L.) and the potential of hop to produce bioactive prenylated flavonoids. J
Agric Food Chem 53, 4793-4798.
Matsuzawa,M., Katsuyama,Y., Funa,N., and Horinouchi,S. (2010). Alkylresorcylic acid
synthesis by type III polyketide synthases from rice Oryza sativa. Phytochemistry 71, 10591067.
Mechoulam,R. (1988). Alkaloids in Cannabis sativa L. In The Alkaloids: Chemistry and
Pharmacology, A.Brossi, ed. (San Diego: Academic Press), pp. 77-93.
Mechoulam,R., Ben-Shabat,S., Hanus,L., Ligumsky,M., Kaminski,N.E., Schatz,A.R.,
Gopher,A., Almog,S., Martin,B.R., and Compton,D.R. (1995). Identification of an endogenous
2-monoglyceride, present in canine gut, that binds to cannabinoid receptors. Biochem Pharmacol
50, 83-90.
Mechoulam,R. and Hanus,L. (2000). A historical overview of chemical research on
cannabinoids. Chem Phys Lipids 108, 1-13.
Merkus,F.W. (1971). Cannabivarin and tetrahydrocannabivarin, two new constituents of hashish.
Nature 232, 579-580.
Mizuuchi,Y., Shi,S.P., Wanibuchi,K., Kojima,A., Morita,H., Noguchi,H., and Abe,I. (2009).
Novel type III polyketide synthases from Aloe arborescens. FEBS lett 276, 2391-2401.
Moliterni,V., Cattivelli,L., Ranalli,P., and Mandolino,G. (2004). The sexual differentiation of
Cannabis sativa L.: A morphological and molecular study. Euphytica 140, 95-106.
Morimoto,S., Komatsu,K., Taura,F., and Shoyama,Y. (1998). Purification and characterization
of cannabichromenic acid synthase from Cannabis sativa. Phytochemistry 49, 1525-1529.
Morimoto,S., Komatsu,K., Taura,F., and Shoyama,Y. (1997). Enzymological evidence for
cannabichromenic acid biosynthesis. J Nat Prod 60, 854-857.
215
Muggia,L. and Grube,M. (2010). Type III polyketide synthases in lichen mycobionts. Fungal
Biol 114, 379-385.
Mulder,N.J., Apweiler,R., Attwood,T.K., Bairoch,A., Bateman,A., Binns,D., Bradley,P.,
Bork,P., Bucher,P., Cerutti,L., Copley,R., Courcelle,E., Das,U., Durbin,R., Fleischmann,W.,
Gough,J., Haft,D., Harte,N., Hulo,N., Kahn,D., Kanapin,A., Krestyaninova,M., Lonsdale,D.,
Lopez,R., Letunic,I., Madera,M., Maslen,J., McDowall,J., Mitchell,A., Nikolskaya,A.N.,
Orchard,S., Pagni,M., Ponting,C.P., Quevillon,E., Selengut,J., Sigrist,C.J., Silventoinen,V.,
Studholme,D.J., Vaughan,R., and Wu,C.H. (2005). InterPro, progress and status in 2005. Nucleic
Acids Res 33, D201-D205.
Munro,S., Thomas,K.L., and Abu-Shaar,M. (1993). Molecular characterization of a peripheral
receptor for cannabinoids. Nature 365, 61-65.
Munson,A.E., Harris,L.S., Friedman,M.A., Dewey,W.L., and Carchman,R.A. (1975).
Antineoplastic activity of cannabinoids. J Natl Cancer Inst 55, 597-602.
Murzin,A.G., Brenner,S.E., Hubbard,T., and Chothia,C. (1995). SCOP: a structural classification
of proteins database for the investigation of sequences and structures. J Mol Biol 247, 536-540.
Nagel,J., Culley,L.K., Lu,Y., Liu,E., Matthews,P.D., Stevens,J.F., and Page,J.E. (2008). EST
analysis of hop glandular trichomes identifies an O-methyltransferase that catalyzes the
biosynthesis of xanthohumol. Plant Cell 20, 186-200.
O'Shaughnessy,W.B. (1843). On the preparations of the Indian hemp, or gunjah. Prov Med J
Retrosp Med Sci 5, 363-369.
Orengo,C.A. and Thornton,J.M. (1993). Alpha plus beta folds revisited: some favoured motifs.
Structure 1, 105-120.
Paniego,N.B., Zuurbier,K.W., Fung,S.Y., van der,H.R., Scheffer,J.J., and Verpoorte,R. (1999).
Phlorisovalerophenone synthase, a novel polyketide synthase from hop (Humulus lupulus L.)
cones. Eur J Biochem 262, 612-616.
Park,S.C., Lee,J.R., Shin,S.O., Park,Y., Lee,S.Y., and Hahm,K.S. (2007). Characterization of a
heat-stable protein with antimicrobial activity from Arabidopsis thaliana. Biochem Biophys Res
Commun 362, 562-567.
Pelletier,M.K., Murrell,J.R., and Shirley,B.W. (1997). Characterization of flavonol synthase and
leucoanthocyanidin dioxygenase genes in Arabidopsis. Further evidence for differential
regulation of "early" and "late" genes. Plant Physiol 113, 1437-1445.
Pertwee,R.G., Howlett,A.C., Abood,M.E., Alexander,S.P., Di,M., V, Elphick,M.R.,
Greasley,P.J., Hansen,H.S., Kunos,G., Mackie,K., Mechoulam,R., and Ross,R.A. (2010).
International Union of Basic and Clinical Pharmacology. LXXIX. Cannabinoid receptors and
their ligands: beyond CB(1) and CB(2). Pharmacol Rev 62, 588-631.
216
Petkovic,H., Thamchaipenet,A., Zhou,L.H., Hranueli,D., Raspor,P., Waterman,P.G., and
Hunter,I.S. (1999). Disruption of an aromatase/cyclase from the oxytetracycline gene cluster of
Streptomyces rimosus results in production of novel polyketides with shorter chain lengths. J
Biol Chem 274, 32829-32834.
Pieper,D.H., Stadler-Fritzsche,K., Knackmuss,H.J., Engesser,K.H., Bruce,N.C., and Cain,R.B.
(1990). Purification and characterization of 4-methylmuconolactone methylisomerase, a novel
enzyme of the modified 3-oxoadipate pathway in the gram-negative bacterium Alcaligenes
eutrophus JMP 134. Biochem J 271, 529-534.
Potter, D. J. P. The propagation, characterisation and optimisation of Cannabis sativa L. as a
phytopharmaceutical. 2009. King's College London.
Prabhu,V., Chatson,K.B., Abrams,G.D., and King,J. (1996). 13C nuclear magnetic resonance
detection of interactions of serine hydroxymethyltransferase with C1-tetrahydrofolate synthase
and glycine decarboxylase complex activities in Arabidopsis. Plant Physiol 112, 207-216.
Radauer,C., Lackner,P., and Breiteneder,H. (2008). The Bet v 1 fold: an ancient, versatile
scaffold for binding of large, hydrophobic ligands. BMC Evol Biol 8, 286.
Raharjo,T.J., Chang,W.T., Choi,Y.H., Peltenburg-Looman,A.M.G., and Verpoorte,R. (2004a).
Olivetol as product of a polyketide synthase in Cannabis sativa L. Plant Sci 166, 381-385.
Raharjo,T.J., Chang,W.T., Verberne,M.C., Peltenburg-Looman,A.M., Linthorst,H.J., and
Verpoorte,R. (2004b). Cloning and over-expression of a cDNA encoding a polyketide synthase
from Cannabis sativa. Plant Physiol Biochem 42, 291-297.
Raty,K., Kantola,J., Hautala,A., Hakala,J., Ylihonko,K., and Mantsala,P. (2002). Cloning and
characterization of Streptomyces galilaeus aclacinomycins polyketide synthase (PKS) cluster.
Gene 293, 115-122.
Russo,E.B. (2011). Taming THC: potential cannabis synergy and phytocannabinoid-terpenoid
entourage effects. Br J Pharmacol 163, 1344-1364.
Russo,E.B., Jiang,H.E., Li,X., Sutton,A., Carboni,A., del,B.F., Mandolino,G., Potter,D.J.,
Zhao,Y.X., Bera,S., Zhang,Y.B., Lu,E.G., Ferguson,D.K., Hueber,F., Zhao,L.C., Liu,C.J.,
Wang,Y.F., and Li,C.S. (2008). Phytochemical and genetic analyses of ancient cannabis from
Central Asia. J Exp Bot 59, 4171-4182.
Saito,K., Yokoyama,H., Noji,M., and Murakoshi,I. (1995). Molecular cloning and
characterization of a plant serine acetyltransferase playing a regulatory role in cysteine
biosynthesis from watermelon. J Biol Chem 270, 16321-16326.
Sallaud,C., Rontein,D., Onillon,S., Jabes,F., Duffe,P., Giacalone,C., Thoraval,S., Escoffier,C.,
Herbette,G., Leonhardt,N., Causse,M., and Tissier,A. (2009). A novel pathway for sesquiterpene
biosynthesis from Z,Z-farnesyl pyrophosphate in the wild tomato Solanum habrochaites. Plant
Cell 21, 301-317.
217
Schilmiller,A.L., Last,R.L., and Pichersky,E. (2008). Harnessing plant trichome biochemistry for
the production of useful compounds. Plant J 54, 702-711.
Schneider,G. (2005). Enzymes in the biosynthesis of aromatic polyketide antibiotics. Curr Opin
Struct Biol 15, 629-636.
Schröder,G., Brown,J.W., and Schröder,J. (1988). Molecular analysis of resveratrol synthase.
cDNA, genomic clones and relationship with chalcone synthase. Eur J Biochem 172, 161-169.
Schröder, J. Relationship tree of selected plant type III PKS. 2010. http://www.biologie.unifreiburg.de, http://www.biologie.uni-freiburg.de.
Schröder, J. One, two, and three condensations. 2009. http://www.biologie.unifreiburg.de/data/bio2/schroeder/Homepage_English.html, http://www.biologie.uni-freiburg.de.
Schultes,R.E., Klein,W.M., Plowman,T., and Lockwood,T.E. (1974). Cannabis: an example of
taxonomic neglect. Botanical Museum Leaflets. [Harvard University]. Cambridge, MA 23 , 337367.
Sciara,G., Kendrew,S.G., Miele,A.E., Marsh,N.G., Federici,L., Malatesta,F., Schimperna,G.,
Savino,C., and Vallone,B. (2003). The structure of ActVA-Orf6, a novel type of monooxygenase
involved in actinorhodin biosynthesis. EMBO J 22, 205-215.
Seigler,D.S. (1998). Plant secondary metabolism. (Boston: Kluwer Academic).
Shen,B. (2003). Polyketide biosynthesis beyond the type I, II and III polyketide synthase
paradigms. Curr Opin Chem Biol 7, 285-295.
Shen,B. and Hutchinson,C.R. (1996). Deciphering the mechanism for the assembly of aromatic
polyketides by a bacterial polyketide synthase. Proc Natl Acad Sci USA 93, 6600-6604.
Shen,B. and Hutchinson,C.R. (1993b). Tetracenomycin F2 cyclase: intramolecular aldol
condensation in the biosynthesis of tetracenomycin C in Streptomyces glaucescens. Biochemistry
32, 11149-11154.
Shen,B. and Hutchinson,C.R. (1993a). Tetracenomycin F1 monooxygenase: oxidation of a
naphthacenone to a naphthacenequinone in the biosynthesis of tetracenomycin C in Streptomyces
glaucescens. Biochemistry 32, 6656-6663.
Shoyama,Y., Yagi,M., Nishioka,I., and Yamauchi,T. (1975). Biosynthesis of cannabinoid acids.
Phytochemistry 14, 2189-2192.
Sievers,F., Wilm,A., Dineen,D., Gibson,T.J., Karplus,K., Li,W., Lopez,R., McWilliam,H.,
Remmert,M., Soding,J., Thompson,J., and Higgins,D. (2011). Fast, scalable generation of highquality protein multiple sequence alignments using Clustal Omega. Mol Syst Biol 7.
218
Silverman,R.B. (2002). Aldol and Claisen reactions and retroreactions. In The Organic
Chemistry of Enzyme-Catalyzed Reactions, 2, ed. (Waltham. Mass, Academic Press), pp. 453478.
Simonsen,J.L. and Todd,A.R. (1942). Cannabis indica. Part X. The essential oil from Egyptian
hashish. J Chem Soc 1 , 188-191.
Small,E. and Beckstead,H.D. (1973). Common cannabinoid phenotypes in 350 stocks of
Cannabis. Lloydia 36, 144-165.
Soltis,D.E., Smith,S.A., Cellinese,N., Wurdack,K.J., Tank,D.C., Brockington,S.F., RefulioRodriguez,N.F., Walker,J.B., Moore,M.J., Carlsward,B.S., Bell,C.D., Latvis,M., Crawley,S.,
Black,C., Diouf,D., Xi,Z., Rushworth,C.A., Gitzendanner,M.A., Sytsma,K.J., Qiu,Y.L.,
Hilu,K.W., Davis,C.C., Sanderson,M.J., Beaman,R.S., Olmstead,R.G., Judd,W.S.,
Donoghue,M.J., and Soltis,P.S. (2011). Angiosperm phylogeny: 17 genes, 640 taxa. Am J Bot
98, 704-730.
Song,B.H., Wang,X.Q., Li,F.Z., and Hong,D.Y. (2001). Further evidence for paraphyly of the
Celtidaceae from the chloroplast gene mat K. Plant Syst Evol 228, 107-115.
Sparkes,I.A., Runions,J., Kearns,A., and Hawes,C. (2006). Rapid, transient expression of
fluorescent fusion proteins in tobacco plants and generation of stably transformed plants. Nat
Protoc 1, 2019-2025.
Stafford,H.A. (1974). The metabolism of aromatic compounds. Annu Rev Plant Physiol 25, 459486.
Staunton,J. and Weissman,K.J. (2001). Polyketide biosynthesis: a millennium review. Nat Prod
Rep 18, 380-416.
Stearn,W.T. (1970). The Cannabis plant: botanical characteristics. In The Botany and Chemistry
of Cannabis., C.R.B.Joyce and S.H.Curry, eds. (London: J. & A. Churchill), pp. 1-10.
Stout,J.M., Boubakir,Z., Ambrose,S.J., Purves,R.W., and Page,J.E. (2012). The hexanoyl-CoA
precursor for cannabinoid biosynthesis is formed by an acyl-activating enzyme in Cannabis
sativa trichomes. Plant J 71, 353-365.
Sugano,Y., Muramatsu,R., Ichiyanagi,A., Sato,T., and Shoda,M. (2007). DyP, a unique dyedecolorizing peroxidase, represents a novel heme peroxidase family: ASP171 replaces the distal
histidine of classical peroxidases. J Biol Chem 282, 36652-36658.
Sultana,A., Kallio,P., Jansson,A., Wang,J.S., Niemi,J., Mantsala,P., and Schneider,G. (2004).
Structure of the polyketide cyclase SnoaL reveals a novel mechanism for enzymatic aldol
condensation. EMBO J 23, 1911-1921.
219
Suss,K.H., Arkona,C., Manteuffel,R., and Adler,K. (1993). Calvin cycle multienzyme complexes
are bound to chloroplast thylakoid membranes of higher plants in situ. Proc Natl Acad Sci USA
90, 5514-5518.
Sytsma,K.J., Morawetz,J., Pires,J.C., Nepokroeff,M., Conti,E., Zjhra,M., Hall,J.C., and
Chase,M.W. (2002). Urticalean rosids: circumscription, rosid ancestry, and phylogenetics based
on rbcL, trnL-F, and ndhF sequences. Am J Bot 89, 1531-1546.
Taguchi,C., Taura,F., Tamada,T., Shoyama,Y., Shoyama,Y., Tanaka,H., Kuroki,R., and
Morimoto,S. (2008). Crystallization and preliminary X-ray diffraction studies of polyketide
synthase-1 (PKS-1) from Cannabis sativa. Acta Crystallogr Sect F Struct Biol Cryst Commun
64, 217-220.
Takaoka,M. (1939). Resveratrol, a new phenolic compound, from Veratrum grandiflorum.
Nippon Kagaku Kaishi 60, 1090-1100.
Tamura,K., Peterson,D., Peterson,N., Stecher,G., Nei,M., and Kumar,S. (2011). MEGA5:
molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and
maximum parsimony methods. Mol Biol Evol 28, 2731-2739.
Taura,F., Morimoto,S., and Shoyama,Y. (1995). First direct evidence for the mechanism of
delta-1-tetrahydrocannabinolic acid biosynthesis. J Am Chem Soc 117, 9766-9767.
Taura,F., Morimoto,S., and Shoyama,Y. (1996). Purification and characterization of
cannabidiolic-acid synthase from Cannabis sativa L.. Biochemical analysis of a novel enzyme
that catalyzes the oxidocyclization of cannabigerolic acid to cannabidiolic acid. J Biol Chem
271, 17411-17416.
Taura,F., Sirikantaramas,S., Shoyama,Y., Shoyama,Y., and Morimoto,S. (2007).
Phytocannabinoids in Cannabis sativa: recent studies on biosynthetic enzymes. Chem Biodivers
4, 1649-1663.
Taura,F., Tanaka,S., Taguchi,C., Fukamizu,T., Tanaka,H., Shoyama,Y., and Morimoto,S.
(2009). Characterization of olivetol synthase, a polyketide synthase putatively involved in
cannabinoid biosynthetic pathway. FEBS Lett 583, 2061-2066.
Thomas,R. (2001). A biosynthetic classification of fungal and streptomycete fused-ring aromatic
polyketides. ChemBioChem 2, 612-627.
Thomas,R. and Williams,D.J. (1983). Oxytetracycline biosynthesis: mode of incorporation of [113C]- and [1,2-13C2]-acetate. J Chem Soc Chem Comm 128-130.
Thompson,T.B., Katayama,K., Watanabe,K., Hutchinson,C.R., and Rayment,I. (2004). Structural
and functional analysis of tetracenomycin F2 cyclase from Streptomyces glaucescens. A type II
polyketide cyclase. J Biol Chem 279, 37956-37963.
220
Toffoli,F., Avico,U., and Signoretti-Ciranni,E. (1968). Methods of distinguishing biologically
active cannabis and fibre cannabis. Bull Narc 1, 55-59.
Tomato Genome Consortium (2012). The tomato genome sequence provides insights into fleshy
fruit evolution. Nature 485, 635-641.
Touwn,M. (1981). The religious and medical uses of Cannabis in China, India and Tibet. J
Psychoactive Drugs 13, 23-24.
Traw,M.B. and Bergelson,J. (2003). Interactive effects of jasmonic acid, salicylic acid, and
gibberellin on induction of trichomes in Arabidopsis. Plant Physiol 133, 1367-1375.
Turner,C.E., Elsohly,M.A., and Boeren,E.G. (1980). Constituents of Cannabis sativa L. XVII. A
review of the natural constituents. J Nat Prod 43, 169-234.
Ugwu,S.O. and Apte,S.P. (2004). The effect of buffers on protein conformational stability.
Pharm Technol 28, 86.
Van Bakel H., Stout,J.M., Cote,A.G., Tallon,C.M., Sharpe,A.G., Hughes,T.R., and Page,J.E.
(2011). The draft genome and transcriptome of Cannabis sativa. Genome Biol 12, R102.
van Schie,C.C., Haring,M.A., and Schuurink,R.C. (2007). Tomato linalool synthase is induced in
trichomes by jasmonic acid. Plant Mol Biol 64, 251-263.
Verpoorte,R., Alfermann,A.W., and Johnson,T.S. (2007). Applications of Plant Metabolic
Engineering. Springer).
Vreet,T.B., Breimer,D., VanGinneken,C.A.M., and VanRossum,J.M. (1972). Identification in
hashish of tetrahydrocannabinol, cannabidiol and cannabinol analogues with a methyl side-chain.
J Pharm Sci 24, 7-12.
Wagner,G.J. (1991). Secreting glandular trichomes: more than just hairs. Plant Physiol 96, 675679.
Wagner,G.J., Hrazdina,G. (1984). Endoplasmic Reticulum as a Site of Phenylpropanoid and
Flavonoid Metabolism in Hippeastrum. Plant Physiol 74, 901-906.
Wakimoto,T., Morita,H., and Abe,I. (2012). Engineering of plant type III polyketide synthases.
Methods Enzymol 515, 337-358.
Walter,S. and Buchner,J. (2002). Molecular chaperones--cellular machines for protein folding.
Angew Chem Int Ed Engl 41, 1098-1113.
Ward,J.J., Sodhi,J.S., Guffin,L.J., Buxton,B.F., and Jones,D.T. (2004). Prediction and functional
analysis of native disorder in proteins from the three kingdoms of life. J Mol Biol 337, 635-645.
221
Weigel,D. and Glazebrook,J. (2006). Transformation of agrobacterium using electroporation.
CSH Protoc 2006.
Weissman,K.J. (2009). Introduction to polyketide biosynthesis. In Methods Enzymol, David
A.Hopwood, ed. Academic Press), pp. 3-16.
Winkel-Shirley,B. (1999). Evidence for enzyme complexes in the phenylpropanoid and
flavonoid pathways. Physiol Plant 107 , 142-149.
Wollner,H.J., Matchett,J.R., Levine,J., and Loewe,S. (1942). Isolation of a physiologically active
tetrahydrocannabinol from Cannabis sativa resin. J Am Chem Soc 64, 26-29.
Wood,B.T., Spivey,N.W.T., and Easterfield,T.H. (1896). Charas. The resin of Indian hemp. J
Chem Soc Trans 69, 539-546.
Wu,R., Skaar,E.P., Zhang,R., Joachimiak,G., Gornicki,P., Schneewind,O., and Joachimiak,A.
(2005). Staphylococcus aureus IsdG and IsdI, heme-degrading enzymes with structural similarity
to monooxygenases. J Biol Chem 280, 2840-2846.
Yamashita,Y., Matsunami,K., Otsuka,H., Shinzato,T., and Takeda,Y. (2008). Grevillosides A-F:
glucosides of 5-alkylresorcinol derivatives from leaves of Grevillea robusta. Phytochemistry 69,
2749-2752.
Yamashita,Y., Matsunami,K., Otsuka,H., Shinzato,T., and Takeda,Y. (2010). 5-Alkylresorcinol
glucosides from the leaves of Grevillea robusta Allan Cunningham. J Nat Med 64, 474-477.
Yamauchi,T., Shoyama,Y., Aramaki,H., Azuma,T., and Nishioka,I. (1967).
Tetrahydrocannabinolic acid, a genuine substance of tetrahydrocannabinol. Chem Pharm Bull
(Tokyo) 15, 1075-1076.
Yao, R. Biosynthesis of cuticular alkylresorcinols in selected grass species Brachypodium
distachyon and Secale cereale. 2011. The University of British Columbia.
Ylihonko,K., Tuikkanen,J., Jussila,S., Cong,L., and Mantsala,P. (1996). A gene cluster involved
in nogalamycin biosynthesis from Streptomyces nogalater: sequence analysis and
complementation of early-block mutations in the anthracycline pathway. Mol Gen Genet 251,
113-120.
Zdobnov,E. and Apweiler,R. (2001). InterProScan--an integration platform for the signaturerecognition methods in InterPro. Bioinformatics 17, 847-848.
Zhan,J. (2009). Biosynthesis of bacterial aromatic polyketides. Curr Top Med Chem 9, 1958610.
Zhang,H., Matsuda,H., Yamashita,C., Nakamura,S., and Yoshikawa,M. (2009). Hydrangeic acid
from the processed leaves of Hydrangea macrophylla var. thunbergii as a new type of antidiabetic compound. Eur J Pharmacol 606, 255-261.
222
Zhang,W. and Tang,Y. (2009). In vitro analysis of type II polyketide synthase. Methods
Enzymol 459, 367-393.
Zhang,W., Wilke,B.I., Zhan,J., Watanabe,K., Boddy,C.N., and Tang,Y. (2007). A new
mechanism for benzopyrone formation in aromatic polyketide biosynthesis. J Am Chem Soc
129, 9304-9305.
Zhou,H., Li,Y., and Tang,Y. (2010). Cyclization of aromatic polyketides from bacteria and
fungi. Nat Prod Rep 27, 839-868.
Zhu,B., Xiong,A.S., Peng,R.H., Xu,J., Zhou,J., Xu,J.T., Jin,X.F., Zhang,Y., Hou,X.L., and
Yao,Q.H. (2008). Heat stress protection in Aspen sp1 transgenic Arabidopsis thaliana. BMB
Rep 41, 382-387.
Zuardi,A.W. (2006). History of cannabis as a medicine: a review. Rev Bras Psiquiatr 28, 153157.
APPENDIX
APPENDIX A
Determining the concentration of PDAL, HTAL, OA, and olivetol.
223
5.0E+05
6.E+05
peak area at 270 nm
peak area at 270 nm
7.E+05
y = 1088.1x
R² = 0.9993
5.E+05
4.E+05
3.E+05
2.E+05
1.E+05
0.E+00
0
200
[PDAL] pmol
y = 834.47x
R² = 0.9982
3.0E+05
2.0E+05
1.0E+05
0.0E+00
400
0
1.2E+06
100
200
300
[HTAL] pmol
400
500
2.0E+05
y = 2027.1x
R² = 0.9989
1.0E+06
8.0E+05
peak area at 270 nm
peak area at 270 nm
4.0E+05
6.0E+05
4.0E+05
2.0E+05
0.0E+00
y = 307.17x
R² = 0.9988
1.5E+05
1.0E+05
5.0E+04
0.0E+00
0
100
200
300
[OA] pmol
400
500
0
100
200
300
400
[olivetol] pmol
500
Figure A1 Standard curves used for product determination. Error bars are the range in duplicate
samples. Equations are the linear regression lines with correlation coefficient expressed as R2.
224
APPENDIX B
Table B1 Injection and bleed times for the generation of antibodies targetting TKS
Description
Pre-immune and 1° injection
2° Injection
1° bleed
3° Injection
2° bleed
4° Injection
3° bleed / Final 80 days
5° injection (facultative)
4° bleed
6° injection (facultative)
5° bleed
7° injection (facultative)
6° bleed / Final 164 day
8° injection (facultative)
7° bleed
9° injection (facultative)
8° bleed
10° injection (facultative)
9° bleed / Final 248 days
Date
1-Jun-2009
22-Jun-2009
2-Jul-2009
13-Jul-2009
23-Jul-2009
10-Aug-2009
20-Aug-2009
7-Sep-2009
17-Sep-2009
5-Oct-2009
15-Oct-2009
2-Nov-2009
12-Nov-2009
30-Nov-2009
10-Dec-2009
28-Dec-2009
7-Jan-2010
25-Jan-2010
4-Feb-2010
225
Day #
0
21
31
43
52
70
80
98
108
126
136
154
164
182
192
210
220
238
248
APPENDIX C
Q-TOF MS/MS Analysis of an unknown protein.
The identification of an unknown 60 kDa protein was also purified from both recombinant and trichome
extracts, and sent to the PBI mass lab for Q-TOF MS/MS analysis. The unknown protein(s) was identified
as an A. thaliana chaperonin homolog (annotated NP_175945) (Section 3.2.4.2).
226
APPENDIX D
The identification of TKS was also confirmed by Q-TOF MS/MS analysis (Section 3.2.4.2).
227
APPENDIX E
A total of 93 OAC-like sequences (including OAC) were aligned using ClustalW provided in
Mega5 (Tamura et al., 2011) with parameters: Pairwise Alignment: Gap Opening Penalty 10;
Gap Extension Penalty 0.1; Multiple Alignment: Gap Opening Penalty 10; Gap Extension
Penalty 0.2; Protein Weight Matrix Gonnet; Residue-specific Penalties ON; Hydrophobic
Penalties ON; Gap Separation Distance 4; End Gap Separation OFF; Use Negative Matrix OFF;
Delay Divergent Cutoff (%) 30%.
Phylogenetic Relationship Analysis used the following parameters: Phylogeny Reconstruction;
Statistical Method: Maximum Likelihood; Test of Phylogeny: Bootstrap method; Number of
Bootstrap Replications: 1000; Substitution Model: Amino acid; Model/Method: Jones-TaylorThornton (JTT) + Gamma model; Rates among Sites: Uniform rates; Gaps/Missing Data
Treatment: Complete deletion; ML Heuristic Method: Nearest-Neighbor-Interchange (NNI);
Initial Tree for ML: Make initial tree automatically.
228
APPENDIX F
Sequences used during the phylogenetic analysis of OAC.
>A. coerulea 027_00390
MGDLKHMVLVKFKEDVVVEDILKGMEKLASEVDSIKSFEWGQDLESEEMLRQGFTHVFTMTFG
SSEDFAACQTHPSHVEFSTTFSAAIEKILLFHFPAVLVKAPAA
>A. coerulea 029_00097
MEEAKGVVKHVLMAKFKDEITPDRIDLLIKGYANLVPLIPPMKAFHWGKDVSIENLHQGFTHVF
ESTFESLEGVAEYISHPAHVEFATEFLTALDKVIVIDYKPTIANH
>A. lyrata 941765
MATSGFKHLVVVKFKEDAKVDEILEGLENLVSQIDTVKSFEWGEDKESHEMLRQGFTHAFSMTF
ENKDGYVAFTSHPLHVEFSAAFTAVIDKIVLLDFPVAAVKSSVVVTP
>A. thaliana AT3G17210
MAEAKGAVVKHVLLAKFKDGVSPETIEELIKGYANLVNLIEPMKAFHWGEDVSIENLHQGYTHI
FESTFETKEAVAEYIAHPAHVEFATIFLGSLDKALVIDYKPTSVSL
>A. thaliana AT5G22580
MATSGFKHLVVVKFKEDTKVDEILKGLENLVSQIDTVKSFEWGEDKESHDMLRQGFTHAFSMTF
ENKDGYVAFTSHPLHVEFSAAFTAVIDKIVLLDFPVAAVKSSVVATP
>A. thaliana TA34447_3702
GSKMEEAKGPVKHVLLASFKDGVSPEKIEELIKGYANLVNLIEPMKAFHWGKDVSIENLHQGYT
HIFESTFESKEAVAEYIAHPAHVEFATIFLGSLDKVLVIDYKPTSVSL
>B. distachyon 4g04380
MAGSGVVKHILLARFKDEVTPERLDQLIRGYAGLVAVVPSMKAFHWGTDVSIENMHQGFTHVF
ESTFESTEGVKEYIEHPNHVDFATEFLALAEKVIIIDYKPTAAT
>B. distachyon 4g42625
MGEFKHLCLVRFKEGVVVDNIIQELSKLATELDTVKYFGWGKDVLEQEALTQGFTHVFSMSFAT
AGDLAACMAHEKHAAFAATFMAALDKVVVMDFPFVTVKPAPEPEPAT
>B. rapa 021232
MEEAKGPVKHVLLAKFKDDVTPEKIDELIKGYANLVNLIEPMKAFHWGKDVSIENLHQGFTHIF
ESTFDSKEAVAEYVAHPIHVEFANMFLGSLDKVLVIDYKPTSV
229
>B. rapa Bra002422
MATSGFKHLVVVKFKEDAKVDEILKGLENLVSQIDSVKSFEWGEDNESHEMLRQGFTHAFSMTF
ENKDAYVSFTGHPLHVEFSAAFTAVIDKIVVMDFTVAAVKSPVVVAP
>B. rapa Bra020198
MAASEFKHLVVVKFKEDAKADEILKGSENLVSQIDSVKSFEWGEDKESHEMLRQGFTHAFSMTF
ENKDAYVSFTKHPLHVKFSAAFTAVIEKIVILDFTVAAVKSPVVVPP
>C. annuum TA6827_4072
NFASLINTATEQSRRRVGEMEGTKGGVVKHILLAKFKDEISPDKIDQLIKQYANLVNLIEPMKAS
HWGKDVSTENLQQGFTHVFESTFDSTEGVAEYVSHPVHVEFANLMLPQLEKVLVIDYKPEKVRP
LN
>C. clementina 025762m
MKAETKGRDMEEAKGVVKHVLLAKFKEGTAQDQIDQLIKDYANLVNLIEPMKSFQWGKDVSIE
NLHQGFTHIFESTFESTEGVAEYVAHPAHVEYANLFLANLEKVLVIDYKPTTVRV
>C. clementina 025899m
MGEGEEAAMGEFKHLVIVKFKEGVVVEDIVKGMKKLVSEIDAVKSFEWGQDVEGQEMLRQGF
THAFLMTFNKKEDYTTFASHPSHVEFSATFSAAIEKIVLLDFPTVLGKAPAA
>C. papaya 146.67
MEEAKGVVKHVLVAKFKEETLPAQIDELIKGYANLVNLIEPMKSFHWGKDVSIENLHQGFTHVF
ESAFESTEGIAEYIDHPAHVEFANLFLPVLEKFVVIDYKPTTVSF
>C. rubella 10014967m
MEEANKGPVKHILLAKFKDGVSPEKIEELIKGYANLVNLIEPMKAFHWGEDVSIENLHQGYTHIF
ESTFESKEAVAEYVAHPVHVEFATMFLGSLDKVLVIDYKPTSVSL
>C. rubella v10002301m
MATSGFKHLVVVKFKEDTKVDEILKGLENLVSQIDTVKSFEWGEDKESHEMLRQGFTHAFSMNF
ENKDGYVAFTSHPLHVEFSAAFTAVIDKIVVLDFPVAAVKSSVAV
>C. sativa OAC
MAVKHLIVLKFKDEITEAQKEEFFKTYVNLVNIIPAMKDVYWGKDVTQKNKEEGYTHIVEVTFE
SVETIQDYIIHPAHVGFGDVYRSFWEKLLIFDYTPRK
>C. sativa PK00183
230
MEKANKGVVKHLVIMKFKDGISDDQIEQMNKEYANLLNLVPSMKTLQLGKVVEMSPGNYKHG
NGGYTHIFESTFESMEGVAEYTFHPAHLHLGHLYSHTFDKVLVFDYIIPITTIS
>C. sativa PK17532
MGEANKGVVKHVFILKMKEGLSNDQIEQMNKDYANLVNLVPSMKALQWGKLEVNNKLGNGG
YTHIFESTFESMEGVAEYADHPAHLHLRNLYFHTLDKFLVFDYKPTIVLPNSSY
>C. sativa PK18164
MEEAKGVVKHVLLAKFKEGTSDDQIQQLIKGYANLLNLIPSMKSFHWGKDVSFENLHQGFTHIF
ESTFENTEGVAEYVAHPAHVEFANVFLSNLDKVVVFDYKPTTVLLP
>C. sativa PK28464
MEEAKGVVKHIVLLKFKDGITDDEIQQFINAYANIVNKIPVMKDFNWGKDVSCVNLHGGFTHIL
ASTFESIEDVAKYVHHPVHVEFGNLYHHNLEKFLIFDYKPTIFLP
>C. sativus 199990
MGEFKHLVVAKFKEGLNVDEIVAQVEKMVSDIDSVKSFEWGHDVEGQDMLTQGFTHVFSMTF
DDKEAITSFLTHPKHLEFCPTFSAAIDKIVVLDFPSILVKAAVPPPSPQPPPALEETPEAAVPPPSPQP
PPALEETPEAALPPVLTSTPLPTPATAAAATPASASA
>C. sativus 252360
MGDSKGVVKHILLAKFKNGISEEQIDQLIKGYANLVNLIEPMKAFQCGKDVSIENLHQGFTHVFE
STFESTQGIAEYISHPAHVEFANKFLSSLDKVIVIDYKPTLICN
>C. sativus 385330
MGEEFKHLVLVKFKEGAAVEEILKGMENLVSEVEAVKSFEWGQEMEGPEMLTQGFTHAFLMTF
ENKEAHTAFVGHPKHVEFSATFSAAIDKIVVLDFPAIRAKPPLPA
>C. sinensis 1g033664m
MGEGEEAAMGEFKHLVIVKFKEGVVVEDIVKGMKKLVSEIDAVKSFEWGQDVEGQEMLRQGF
THAFLMTFNKKEDYTTFASHPSHVEFSATFSAAIEKIVLLDFPTVLGKAPAA
>C. sinensis 1g033887m
MEEAKGVVKHVLLAKFKEGTAQDQIDQLIKDYANLVNLIEPMKSFQWGKNVSIENLHQGFTHIF
ESTFESTEGVAEYVAHPAHVEYANLFLANLEKVLVIDYKPTTVRV
>E. grandis B02216
MEEAKGLVRRVHLVKFKDGTPLARIEEILKNYANLVNLIEPLKSWHMGKNVSIIYLDQGFTDVFE
VTFESAEGLAEYADHPIHTEFSKQTRPYLDKVIAITYEPTV-CT
231
>E. grandis B02218
MEEAKGLAKRVLLVKFKDGTPPARLEEILKNYANLVNLVEPLKSWRMGKNVSFMNLDQGFTH
VFELTFESAKGLAEYADHPAHTEFSSQCRPYLDKFVVINYEPTVFCT
>E. grandis B03989
MEEASRPKEKTMEEAKGFVRRVLLFKFKAETAPDHIEQLIKDHSNLVNLVESLKSWHMGKDVS
VKYSNEGFTHVIELTFESTEGIAAYMGHPAHLEMHERIWPHLEKIVVIDYKPTLCN
>G. arboreum BM360640
ALREVRLMEEAKGVVKHILLAKFKDEIPPDKIEELIKGYANLVNLIHPMKAFHWGKDVSIENLHQ
GFTHVFESTFEST
>G. hirsutum TA22147_3635
LALREVRLMEEAKGVVKHILLAKFKDEIPPEKIEELIKGYANLVNLIHPMKAFHWGKDVSIENLH
QGFTHVFESTFESTEGIAEYIAHPAHVEFANLFLPSLDKVIVFDYKPTVVRC
>G. max 04g08730
MEEAKGVVKHVLLAKFKDDVTPERIEELIKDYANLVNLIPPMKSFHWGKDVSAENLHQGFTHVF
ESTFESPEGVAEYVAHPAHVEYANLFLSCLEKVIVIDYKPTVVKL
>G. max 06g08830
MEEAKGVVKHVFLAKFKDDVTPERIEELIKDYANLVNLIPPMKSFHWGKDVSAENMHQGFTHV
FESTFESTEGIAEYTAHPAHVEYATLFLSYLDKVIVIDYKPTIVKL
>G. max 09g23640
MGTFNHYVIVKLKDGVAVDELIQGLEKMVSGIDHVKSFEWGKDIESHDMLRQGFTHDFLMAFN
RKEEFNAFQTHLTHLEFTRVFSPAIEKIVVLDFPSNLVKAPA
>G. max 13g11380
MGTFNHYVIVKFNDGVAVDELIQGLEKMVSGIDHVKSFERGKDIESHDMLRQGFTHVFLMAFN
GKEEFNAFQTHVNDLEFTGLFSPAIEKIVVLDFPSNLMKAPA
>G. raimondii TA16529_29730
LALREVRLMEEAKGVVKHILLAKFKDEIPPEKIEELIKGYANLVNLIHPMKAFHWGKDVSIENLH
QGFTHVFESTFESTEGIAEYIAHPAHVEFANLFLPSLDKVIVFDYKPTVVRC
>G. soja TA2928_3848
EEEMEEAKGVVKHVLLAKFKDDVTPERIEELIKDYANLVNLIPPMKSFHWGKDVSAENLHQGFT
HVFESTFESPEGVAEYVAHPAHVEYANLFLSCLEKVIVIDYKPTVVKL
232
>H. lupus CL734Contig1
MEEAKGVVKHVLLAKFKEGIPDDQIEQLIKGYANLVNLIPPMKSFHWGKDVSFENLHQGFTHIFE
STFETTEGVAEYVAHPAHVEFANVFLSVLDKVVVFDYKPTTVLP
>L. usitatissimum 10037823
MEEGKGEVKHVLLAKFKPEISGEEIEKLIKDYANLVTLIEPMKAFHWGTDVSIENMHQGFTHIFV
STFESTEGIAEYVAHPVHVEFANLFLAALDKVLVFDYKPTSVK
>L. usitatissimum Lus10009407
MAAPEFKHLAIIKEGVVVDEMIKGMEKLVSEISLIKSLEWGQDLEGHEMLTQGFTHAFTLTFNSK
EDYIAYDNHPTHLEYAAAFSASIEKFVVLNYPTVLLSQT
>L. usitatissimum Lus10020555
MAAPEFKHLAIIKFKEGVVVDEMIKGMEKLVSQISLIKSLEWGQDLEGHEMLTQGFTHAFTLTFN
SKEDYIAYENHPTHLEYAAGFSASIEKFVVLNYPTVLLKQT
>M. acuminata 6T20510
MAVKHVLLAKFKEEVSPEAVDDLIKGYANLVSLITPMKAFHWGRDVSIENLHQGFTHVFESTFD
GVEGIAEYIAHPAHVEFANKFLPALEKVIVIDYAPTAVN
>M. acuminata 9T09110
MGSIRHLVLAKFKDEAAVEELLQGLQKLVSEIDVVKSFEWGEDVLKDERLGQGFTHAFLLTFGS
AEDLATYIKHPSHVAFGKAFRAAIDKILAIDFPVVTNEISA
>M. domestica MDP0000213440
MEEAKGVVKHVVLAKFKDGIAENKIEEIIKGFANLVNLIEPMKSFNWGKDVSSENLHQGFTHIFE
STFESTEGVAEYVAHPAHVDFANLFLSHLEKVIVIDYKPTAVSV
>M. domestica MDP0000528167
MEEAKGGELKNILMAKFKEGTSESQIEQLIESFANLVNLVEPMKSFHWGKELSIEKEEEGYIHVFE
TTFESVEGMAEYAVHPAHHDFAKLFLPNLEKMCSIGYRLVRRRDGTGQDGTGRDGTEQRFRVM
FGTSKMEQNGTGLND
>M. esculenta 019718m
MEEAKGVVKHVLLAKFKEGISSDEIEKLIKGYANLVNLIEPMKAFHWGTDVSIENLHQGFTHVFE
STFESTEGVAEYISHPAHVDFANLFLPALEKVVVIDYKPTTVRL
>M. esculenta 019824m
233
MGFEHLVIVRFKEGAVVADIMKGLEKLVSDVDLVKTFAWGEDLESPEMLTQGFTHAISMTFDK
KQDYTAFQTHPSHVEFSATFSAAIEKIVVLCFSSVQVKPAIA
>M. esculenta 029781m
MAEAKGVAKHIVLVKFKEEVSAEEIEQLIKGYANLVNLIPPMKSFRWGIDAGIENLTQGYTHAFE
STFETTQGIAEYVAHPAHIEFSNQLAPALHQVLTINYQSNTVHL
>M. esculenta 030990m
MGGYKHLVIFRFKAGAEVEEIMKGLEKLVSETDLVKSFEWGRDLDNPEILTQEFTHAISMTFDKK
QDYDAFQNLPSHIDFSLTFSTAVQKIVVLCFPSVQVKPAVA
>M. guttatus mgv1a016735m
MATEFKHLVLVKFKNDVVVGDVMKGLEKLVLEMDTVKSFVWGEDIESHEMLRQGFTHAFLMT
FGSKEDFTAFTSHPTHVDFSSTFSTAVDKAVLLDFPSVTVKPPPAAA
>M. guttatus mgv1a016769m
MEEGEVKHIVLAKFKESVSEEEIQESIKQYANLVNLVPSMKAFSWGKDLSKENMQQGFTHVFES
TFESTQGVAEYVSHPSHVEYATLLLSQLDNVIVVDYKPTPVRL
>M. guttatus mgv1a016804m
MEGEVKHILLAKFKQGITEDQIEEYIKQYANLVNLVPSMKSFSWGKDVSIENLNEGYTHIFESTFE
STQGIADYISHPDHVQYANTLLPQLEKVLVVDYKPTKVHL
>M. guttatus mgv1a024900m
GKHEVKHMLLAKLKQGISEDEIEGYIKQYANLVNLVPSMKAFAWGKDVSIENLNDGFTHVFEST
FASTQGIAEYLSDPDHVAFANTLLPQLDKVLVVDYMPTKVKL
>M. truncatula 019717m
MEEAKGVVKHIFLAKFKEEISSDQIEKLIKGYANLVNLIEPMKAFHWGTDVSNENLHQGFTHVFE
STFESTEGVAEYVSHPAHVEFANLFLAAAEKVIVIDYKPTTVRL
>M. truncatula 019718m
MEEAKGVVKHVLLAKFKEGISSDEIEKLIKGYANLVNLIEPMKAFHWGTDVSIENLHQGFTHVFE
STFESTEGVAEYISHPAHVDFANLFLPALEKVVVIDYKPTTVRL
>M. truncatula 019720m
MEEAKGVVKHIFLAKFKEEISSDQIEKLIKGYANLVNLIEPMKAFHWGTDVSNENLHQGFTHVFE
STFESTEGVAEYVSHPAHVEFANLFLAAAEKVIVIDYKPTTVRL
>M. truncatula 019721m
234
MEEAKGVVKHIFLAKFKEEISSDQIEKLIKGYANLVNLIEPMKAFHWGTDVSNENLHQGFTHVFE
STFESTEGVAEYVSHPAHVEFANLFLAAAEKVIVIDYKPTTVRL
>N. benthamiana CK296232
FGMEGGKVKHILLAKFKDGIPADQIDQLIKQYANLVNLIEPMKAFHWGENVSIENFHQGFTHVFE
STFDSTEGIAEYIDHPAHVEYANTLLPQLEKVLVIDYKPEKVGP
>N. benthamiana CK296233
FGMEGGKVKHILLAKFKDGIPADQIDQLIKQYANLVNLIEPMKAFHWGENVSIENFHQGFTHVFE
STFDSTEGIAEYIDHPAHVEYANTLLPQLEKVLVIDYKPEKVGP
>N. benthamiana EH366633
FGMEGGKVKHILLAKFKDGIPADQIDQLIKQYANLVNLIEPMKAFHWGENVSIQNFHQGFTHVF
ESTFDSTEGIAEYIDHPAHVEYANTLLPQLEKVLVIDYKPEKVGP
>N. tabacum EB446961
FGMDAAKGGGVVKHILLAEFKDGIPADQIDQLIKQYANLVNLIEPMKAFHWGENVSIENFHQGF
THVFESTFDSTEGIAEYIDHPTHVEYANILLPQLEKVLVIDYKPEKVGP
>O. sativa Os01g33160
MAAETPAAGRSGVLKHIVLARFKEEVTPERLDHLIRGFGGLVNLVPSMKAFNWGTDVSIENMHH
GFTHVFECTFESTEGVKEYIEHPAHLEFAKEILLAMEKTLIIDYMPTAVNNS
>O. sativa Os11g05290
MGEVKHLCLVKFKEEVLVDDILQGMTKLVSEMDMVKSFEWGKDVLNQEMLTQGFTHVFSLTF
ASSEDLTTYMSHERHQEFAGTFMAAIDKVVVVDFPVVASKPAPPAAPAAAA
>P. patens Pp1s189_106V6
MGEVKHLCLVKFKEGVVVEDVLKGMTDLVAGMDMVKSFEWGQDVLNQEMLTQGFTHVFSLT
FAFADDLATYMGHDRHAAFAATFMAALDKVVVIDFPVVVAKPPPPA
>P. patens Pp1s85_83V6
MSGVVKHILLLKFKDSISPDEQTALIQKYAALPNSISAMKGFEWGTDISVENMHQGFTHAFISTFD
SPQGRDEYLVHPVHDSFAKELLGAVDNALVFDFNPTVALKANS
>P. persica ppa013300m
MGSAVDLGNNNMGEFKHLVIVKFKEDVVVDDILKGLENLVAEIDAVKSFEWGQDLESQELLRQ
GFTHVFLMTFDKKDDYAVFQSHPKHQEFSAIFSTVIEKLVLLDFPPTLVKTPPKAAPPEAPPEAPP
A
235
>P. persica ppa013682m
MEEAKGVVKHVLLAKFKEGISENQIEELTKGYAKLVDLIEPMKSFHWGKDVSIENLHQGFTHIFE
STFESMEGVTEYVAHPAHVDFANLFLSHVEKVIVFDYKPTMVRV
>P. trichocarpa 0010s04670
MPRCSMRLVQWKKNVKKASSLVFKKPDSLSVGRNLEETKTESKRERERERDTVLMEEAK-GVVKHVLLAKFKEGIPSDEIEKLIKGYANLVNLIEPMKAFEWGTDVSIENMHQGFTHIFESTCESKE
GIAEYIGHPAHVDFGGLFLPALEKVIVFDYKPTVARF
>P. trichocarpa 0010s04700
MSKRRPAWFSSRNLEETKTESKRERDTVVMEEAKGVVKHVLLAKFKEGIPSDEIEKLIKGYANL
VNLIEPMKAFEWGTDVSIENMHQGFTHIFESTFESKEGIAEYIGHPAHVDFGGLFLPALEKVIVFD
YKPTVARF
>P. trichocarpa 0010s16070
MEQVKGVVKHIVLAKFKEGVTEEEIEKHIKDYANLLNHIEHMKSFHWGTDVSIENLHQGFTHIFE
ITFETLDGRSAYVAHPAHVDFGTAFLTILEKIVVVDFVPTLVKL
>P. trichocarpa 0010s16080
MEGGKGVVKHIVLAKFKEGTTEEEIEKHIKNYADLLNHIEHMKSFEWGTDVSIENLHQGFTHIFE
ATFETLEGRSAYVAHPAHVKFGTALFPTLEKVIVFDYVPKA
>P. vulgaris 091002521m
MEEAKGVVKHIVLAKFKDDITAEKIEELIKGYANLVNLVPPMKSFHWGTDVSAENLHQGFTHVF
ESTFESTEGVAEYVAHPAHVEFANLFISSTEKVIVIDYKPTI
>P. vulgaris 091014673m
MRACCIGLCIYTWLLGTTFIHKSCVVAMGDFNHYVVVKFKDGVAVEELIQGLEKMVSGIHEVKS
FEWGKDIESHDMLRQGFTHVFLMTFNGKEEFNTFRTHPNHIEFSGVFSPSIEKIVVLDFPSKLVKA
PA
>P. vulgaris 091024159m
MLAMGEFKHFVIVKFKEGVAVDELTKGMEKLVSEIDAVKAFEWGRDIESLDVLRQGFTHAFLM
TFNKKEDFAAFQSHPNHVEFSTTFSAAIENIVLLDFPSTLVKAPA
>R. communis 29889.m003356
MAEEVKEVHHLVMAKFKEDITPEKIEELIEGYANLVNLVPPMKAFHWGTDVSIENLNEGFTHIFH
ATIESREGLAEYVNHPAHIEFGNLMTPAMEKVIVVDFSPSLVQLQRE
236
>R. communis 29988.t000006
MAGFKHLVIVKFKEDAAVEEIVKGMEKLVSESDLVKSFAWGQDSEGPEMLTQGFTHAFSMTFD
KKEDYTAFQTHPNHVEYSATFSAAIEKIVVLFFPYVQVKATASA
>R. communis 30170.m013937
MEEAKGVVKHVLLAKFKDEIPTHEIDKLIKGYANLVNLIPPMKAFHWGSDVSIENLHQGFTHVFE
SSFESTEDIAEYISHPAHVEFANLFLRAVDKVIVIDYKPTIVHL
>S. bicolor 03g027060.1
MASGGVVKHILLARFKEDVTQERLDELIRGYAALVAAVPSMKAFHWGTDVSIENMHQGFTHVF
ESTFESTEGIKEYIEHPKHVEFANVFLPVLDKVIVIDYKPTSVN
>S. italic 003420m
MAAGGGSGVVKHIVLARFKEEVTPERLDQLIQGYAALVDAVPSMKSFHWGTNVSIENTHEGFT
HVFESTFESTEGIKEYVKHPAHVEYANEFLSVSEKILIIDYKPTSVN
>S. italica Si011410m
MAEVKHLCLVKFKEGVVVEDILKGMADLAAQMDMVKSFEWGKDVLNQEMLTQGFTHVFSLT
FASADDLTAYMAHEKHAAFAATFMAALEKVVVIDFPVVIAKPPPQA
>S. italica Si027006m
MAPRSIALALRSSLQSHSECGMGEVKHLCLVKFKEGVVVEDVLKGMADLAAQMDMVKSFEWG
QDVLNQEMLTQGFTHVFSLTFASADDLTAYVSHEKHAAFAATFMAALEKVVVIDFPVAIAKPPP
QA
>S. lycopersicum 05g053310
MEGGKGGVVKHILLAKFKDGIPPEQIDQLIKQYANLVNLVEPMKAFQWGKDVSIENLHQGFTHV
FESTFDSLEGVAEYIAHPVHVEYANTLLPQLEKFLIVDYKPQ
>S. lycopersicum 11g066950
MANEFKHLVLVKFKEDVVVEDILKELEKLVQEMDIVKSFVWGKDVESHEMLRQGFTHAIIMTF
NSKEDYQTFANHPNHVGFSATFATVIDKAVLLDFTAISGKAT
>S. moellendorffii 98687
MAQENPKSGLKHVLLARFKPEISEEEVASLIAGYEALPQSIEAMKGFEWGTDVSVENLHQDFTH
VFTSTFESPEGRDAYLVHPAHTAYANKLLPALDKVIVLDFHPKVVCHRS
>S. tuberosum TA38774_4113
237
SRKKSEKIEMEGGKGGVVKHILLAKFKDGIPPEQIDQLIKQYANLVNLVEPMKAFHWGKDVSIE
NLHQGFTHVFESTFDSLEGVAEYVAHPVHVEFANTMLPQLEKVLIIDYKPQ
>T. cacao Tc05_g028060
MVLIKRKHVGPSQNQTQIKPSKFKDETTAEKIEELIKGYANLVNLIEPMKAFQWGKDVSIENLHQ
GFTHVFESTFESTEGIAEYVAHPVHVQFANLFLGHLEKVLVIDYKPTIARC
>T. halophilav10015069m
MATSGFKHLVVVKFKEDAKIDEILKGLENLVSQIDSVKSFEWGEDKESHEMLRQGFTHAFSMTF
ENKDAYVSFTSHPLHVEFSTAFTAAIDKIVLLDFTVAAVKSPVEASP
>V. vinifera GSVIVP00011995001
MEEAKGLVKHVLLAKFKDSTPPDQIEELIKGYANLVNLVPPMKAFHWGKDVSIENMHQGFTHV
FESTFESVEGIAEYVSHPAHVEFANLFLPHLEKVIVFDYKPITVHL
>V. vinifera GSVIVP00019670001
MCGQLKRVEEGKGVVKHVLLAKFKDSTPPDQIEELIKSYANLVSLIPPMKAFHWGKDVSIENMH
QGFTHVFESTFESVEGMAEYVSHPAHVEAANRFLPHLEKVIVLDYKPTAVHL
>V. vinifera GSVIVP00033194001
MIFYKEIASFFYIYNREEGKGVVKHVLLAKFKDSTPPDQIEELIKSYTNLVSLIPPMKAFHWGKDV
SIENMHQGFTHVFESTFESVEGMAEYVSHPAHVEAANRFLPHLEKVIVLDYKPTAVHL
>Z. mays GRMZM2G050730
MGEVKHLCLAKFKDGVVVDDVLKGMTDLAAGMDTVKSFEWGQDVLNQEMLTQGFTHVFSLT
FASAADLAAYMAHDSHTAFAATFMAAIDKVLVVDFPVVVAKPPPPA
>Z. mays GRMZM2G351990
MPTASSRRTFSEAVSTDDEEAARGGKVGRPSPPRCSSRRGTDVSIENMHQGFTHVFESTFESTEGI
KEYIEHPAHVEFASVFLPVLEKVLIIDYKPTSAN
238
APPENDIX G
The Relationship of Type III PKS by J. Schröder (2010). Available at http://www.biologie.unifreiburg.de/data/bio2/schroeder/Plant_TypeIII_Tree.pdf
239
240
Steve Gagne