Kinetic Analysis of the Catabolism of Tabersonine in Catharanthus Roseus Leaf Protein Extracts

Kinetic Analysis of the Catabolism of Tabersonine in Catharanthus roseus Leaf Protein Extracts

A THESIS SUBMITTED TO THE FACULTY OF THE UNIVERSITY OF MINNESOTA BY

Matthew Dale Evans

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE CHEMICAL ENGINEERING

Dr. Guy W. Sander

August 2018

I would like to express gratitude toward my adviser Dr. Guy Sander for teaching me the principles of metabolic engineering, and for providing great guidance and support throughout the entire process of this work.

I also would like to thank Dr. Michael Rother and Dr. Victor Lai for serving on my thesis committee.

Thanks also goes out to the faculty of the Chemical Engineering Department for teaching me the principles of chemical engineering in your courses.

I would also like to thank my fellow Chemical Engineering graduate students for their support and conversations, both on- and off-campus.

And finally, I would like to thank my family and friends for their continued support throughout my graduate studies.

Funding for this project was provided by the Chancellor’s Faculty Small Grant and

Teaching & Research assistantships through the University of Minnesota Duluth.

Abstract

Vinblastine and vincristine are valuable chemotherapy compounds produced by

Catharanthus roseus. However, they are produced in small quantities in the plant because of their cytotoxicity and spatial separation of the metabolic pathways of the two monomeric molecules, vindoline and catharanthine, that condense to form them. There have been extensive attempts to metabolically engineer greater production of these molecules, but the specialized cell types needed for the terminal steps of vindoline biosynthesis from tabersonine and the high regulation of the pathway in the plant have made use of a homologous host difficult. Recently, all metabolic steps of the pathway have been elucidated, but there remains difficulty in culturing a heterologous host to express all metabolic steps. There is almost no knowledge of the catabolism of these alkaloids in the plant, and this knowledge could be used in future metabolic engineering attempts. Leaf protein extracts were incubated with tabersonine to attempt to find possible catabolic products, and initial rate experiments were used to calculate kinetic parameters of a possible catabolic enzyme. Results showed no formation of previously known or unknown metabolites. After performing kinetic studies, evidence was found of

-1 " an allosteric enzyme acting on tabersonine with a vm of 5.4 ± 1.9 μM min , a Km of

16,000 ± 14,000 μM, and a Hill coefficient of 2.0 ± 0.2.

Table of Contents

Abstract ...... ii List of Tables...... iv List of Figures ...... v Background ...... 1 Introduction ...... 1 Terpenoid Indole Alkaloid Enzymatic Pathway ...... 2 Metabolic Engineering for Increased Alkaloid Production ...... 3 Tabersonine Metabolism ...... 6 Enzyme Kinetics ...... 7 Materials and Methods ...... 8 Chemicals...... 8 Protein Extraction ...... 8 Search for Unknown Metabolites ...... 8 Enzymatic Reactions ...... 8 Ethyl Acetate Extraction of Alkaloids ...... 9 Kinetic Studies ...... 9 Initial Rate Experiments ...... 9 Acetone Protein Precipitation Protocol for Alkaloid Extraction ...... 9 HPLC-UV Analysis ...... 10 Theory ...... 10 Determination of Experimental Reaction Rates ...... 10 Initial Kinetic Model to Determine Time Scale ...... 11 Results and Discussion ...... 12 Search for Unknown Metabolites ...... 12 Initial Kinetic Model to Determine Time Scale ...... 12 Initial Enzymatic Reactions ...... 13 Kinetic Studies ...... 13 Error in Ethyl Acetate Extraction ...... 13 Kinetic Studies ...... 14 Kinetic Model ...... 16 Conclusion ...... 18 Future Works...... 19 Illustrations ...... 20 References ...... 28 Appendices ...... 31 Appendix 1: Calculating Rate Constants from Initial Rate Experiments ...... 31 Appendix 2: Polymath ODE Solver Input ...... 36

iii

List of Tables

Table 1: Kinetic constants used in model ...... 21

List of Figures

Figure 1: Terpenoid Indole Alkaloid Metabolic Pathway ...... 20

Figure 2: Metabolic pathway from tabersonine to vindoline...... 21

Figure 3: Results of metabolic model ...... 22

Figure 4: 254 nm chromatogram of extracted reaction medium...... 22

Figure 5: 329 nm chromatogram of extracted reaction medium...... 23

Figure 6: Tabersonine consumption in samples incubated at 30°C for 2 hours with a starting tabersonine concentration of 30 μM and: A: 300 μM of NADPH and s-adenosyl-L-methionine; B: 300 μM s- adenosyl-L-methionine; and C: 300 μM of NADPH ...... 23

Figure 7: Specific activity of samples without (control) and with clotrimazole...... 24

Figure 8: Results of testing both extraction protocols extraction efficiency ...... 24

Figure 9: Kinetic data from non-enzymatic consumption of tabersonine using the enzymatic reaction protocol with boiled enzyme extract...... 25

Figure 10: Lineweaver-Burke plot from the initial rate experiments ...... 25

Figure 11: Trends associated with a Lineweaver-Burk plot for standard Michaelis-Menten enzymes and positive and negative cooperativity allosteric enzymes ...... 26

Figure 12: Modified Eadie-Hofstee plot from the initial rate experiments...... 26

Figure 13: Modified Eadie-Hofstee plot showing Trends associated with standard Michaelis-Menten enzymes and positive and negative cooperativity allosteric enzymes ...... 27

Figure 14: Allosteric kinetic model shown with experimental data points ...... 27

Figure 15: Modelled Langmuir-Hinshelwood plot of a Michaelis-Menten enzyme ...... 31

Figure 16: Modelled Lineweaver-Burke plot of a Michaelis-Menten enzyme ...... 32

Figure 17: Modelled Eadie-Hofstee plot of a Michaelis-Menten enzyme ...... 33

Figure 18: Modelled Langmuir-Hinshelwood plot of a Michaelis-Menten enzyme, a positively cooperative allosteric enzyme, and a negatively cooperative allosteric enzyme ...... 34

Figure 19: Modelled Hill plot of a Michaelis-Menten enzyme, a positively cooperative allosteric enzyme, and a negatively cooperative allosteric enzyme ...... 35

Background

Introduction

Catharanthus roseus produces a wide range of pharmaceutically active alkaloids in its secondary metabolic system. These molecules are enzymatically synthesized in the plant as a defense against folivore and fungal attacks. Different metabolites cause different symptoms in the attacking organism: for example, strictosidine, the first alkaloid in the pathway, causes protein reticulation, while vincristine and vinblastine, closely related dimers in the pathway, disrupt the formation of microtubules during mitosis [1].

This led to those metabolites being the first biologically produced drugs used in cancer treatments [1–3].

Vincristine and vinblastine can currently be produced in two ways: extraction of

the dimer directly from the plant, and semi-synthetic production. There have been several

semi-synthetic methods suggested [4]; however, they require using catharanthine and

vindoline from the plants, which are also found in low levels. Fully synthetic production

techniques are difficult because of the high complexity of the molecules and

stereochemical constraints.

Vincristine and vinblastine are found in very low levels in the plant (0.0003% to

0.01% dry weight) [5] , and it can take 1000 kg dry leaves to produce one gram of product

[6]. Low levels of the terpenoid indole alkaloid (TIA) dimers can be attributed to the

spatial separation of the monomers during TIA biosynthesis [7]. Upstream metabolites in

the pathway are synthesized in internal phloem-associated parenchyma (IPAP) cells before being transported to leaf epidermal cells, where tabersonine and catharanthine are

1 synthesized. Catharanthine is then exuded to the wax on the leaf surface, where it exhibits antifungal properties [7]. However, the terminal stages of vindoline biosynthesis from tabersonine occur in specialized mesophyll and laticifer and idioblast cells [2,8–10]. This allows the separation of the two monomer TIAs and formation of the dimer and its anti- mitotic properties until the plant undergoes folivore attack. For example, when caterpillars were fed a diet of C. roseus leaves and their gut was extracted to test for alkaloid levels, it was found the gut extracts had a higher ratio of dimer to monomer than the leaves while the composition of other TIAs was similar [1].

Terpenoid Indole Alkaloid Enzymatic Pathway

The overall terpenoid indole alkaloid pathway is shown in Figure 1. The first metabolite in the terpenoid indole alkaloid metabolic pathway is strictosidine.

Strictosidine is a result of the enzymatic coupling of secologanin and tryptamine by strictosidine synthase. Tryptophan is decarboxylated to tryptamine by tryptophan decarboxylase in epidermal cells in the indole biosynthetic pathway [2]. Secologanin biosynthesis begins in the methyl-erythritol pathway (MEP), which occurs in specialized internal phloem-associated parenchyma (IPAP) cells. There geraniol is converted to 10- hydroxygeraniol by geraniol-10-hydroxylase before it or another intermediate is transported to the leaf epidermis, and loganin is converted to secologanin by secologanin synthase.

Strictosidine is transformed by strictosidine β-glucosidase, after which it undergoes spontaneous reactions to form 4,21-dehydrogeissochizine which is enzymatically transformed to 19E-geissoschizine by geissoschizine synthase [11]. 19E- geissoschizine then undergoes 6 more enzymatic steps before being converted to either

2 tabersonine or catharanthine by tabersonine synthase or catharanthine synthase [12,13].

While still in the leaf epidermal cells, tabersonine is converted to 16-methoxytabersonine by tabersonine-16-hydroxylase and tabersonine-16-O-methyltransferase [9,14]. It is then transported to the mesophyll cells where the enzymes tabersonine-3-oxygenase and tabersonine-3-reductase transform it to 3-hydroxy-16-methoxy-2,3-dihydrotabersonine

[15], followed by 3-hydroxy-16-methoxy-2,3-dihydrotabersonine-N-methyltransferase

(NMT) transforming it to desacetoxyvindoline [16]. The last two steps toward vindoline biosynthesis occur by the action of desacetoxyvindoline-4-hydroxylase and deacetylvindoline-O-acetyltransferase in specialized laticifer or idioblast cells [2,17,18].

In root cells, where lochnericine and hörhammericine accumulate, tabersonine is converted to lochnericine by tabersonine-6,7-epoxidase, and then further converted by hydroxylation and O-acetylation to hörhammericine [19]. Tabersonine can also be

transformed into minovincinine (unclear reaction), and acetylated by minovincinine-O-

acetyltransferase (MAT), which can also act on hörhammericine [20].

Metabolic Engineering for Increased Alkaloid Production

Many different metabolic engineering techniques have been used to attempt to

increase production of TIAs, but none have been particularly successful due to the

complexity of the metabolic system of C. roseus. Because of the toxicity of some of these alkaloids, the metabolic system is heavily regulated in the plant and not completely understood, especially in the catabolism of TIAs [21,22]. However, past work has given us a better understanding of the pathway. This has included techniques such as gene overexpression of feedback-insensitive anthranilate synthase within C. roseus hairy root cultures [23]. Anthranilate synthase exhibits feedback inhibition from tryptamine and

3 tryptophan. This feedback-insensitive overexpression increased levels of tryptophan and tryptamine, but levels of tabersonine, lochnericine, and hörhammericine were decreased, showing the need for a more complete understanding of the regulation of the pathway.

Later studies found that overexpression of 1-deoxy-D-xylose synthase with either geraniol-10-hydroxylase or anthranilate synthase did increase alkaloid levels [24].

Additionally, a study overexpressing the transcriptional regulator ORCA3 along with geraniol-10-hydroxylase in whole plants significantly increased accumulation of strictosidine, vindoline, catharanthine, and ajmalicine [25]. However, metabolites in other pathways of the plant were also affected, and there was no further accumulation of dimeric TIAs. These findings indicate a need to manipulate multiple genes to achieve an increase in TIA levels.

Other work has included elicitation of ORCA3 by jasmonic acid. Using C. roseus

cell suspension cultures elicited with jasmonic acid and a genetic overexpression of

ORCA3, Peebles et al (2009) [26] found jasmonic acid increased alkaloid levels and expression of TIA pathway genes, even without overexpression of ORCA3. Jasmonic acid was also found to increase levels of ajmalicine in C. roseus cell suspension cultures

[27,28] and vindoline in C. roseus shoot cultures [28].

Additionally, there have been attempts to create heterologous hosts to carry out the biosynthesis of TIAs. In one recent study, virus-induced gene silencing (VIGS) was used to silence candidate genes for the elicitation of the final two uncharacterized enzymes in vindoline biosynthesis from tabersonine. After those genes were discovered, the pathway from tabersonine to vindoline was assembled in Saccharomyces cerevisiae cells which produced vindoline at 0.24 mg gdwt -1 12h -1 when fed with tabersonine [15].

In comparison, vindoline content in whole leaves is 2 mg gdwt -1 after 4 weeks of growth

[15]. S. cerevisiae was also able to produce strictosidine through the addition of exogenous genes and some gene deletions [29].

More recently, the remaining metabolic genes leading from strictosidine to catharanthine, tabersonine, and ajmalicine were elicited and assembled in S. cerevisiae , allowing for the possibility TIA production completely from heterologous hosts. [12,13]

Heterologous hosts also can excrete vindoline and other metabolites into the medium, making separation and purification steps easier to perform.

Difficulties remain in the production of TIAs excluding the gaps left in knowledge of the regulation of the pathway. Extraction of alkaloids from whole plants is difficult and expensive, both in growing the plants and purifying the alkaloids. The specialized cell types in which vindoline is produced also make it difficult to use undifferentiated cell cultures. Suspended C. roseus cell cultures have been produced that

accumulate catharanthine and tabersonine, but they lack the necessary enzymes to

transform tabersonine to vindoline [20]. Transgenic C. roseus cell suspension cultures have also shown genetic instability and decreased production over time [30]. C. roseus hairy root cultures show more genetic stability [31,32], but differentiated root cells metabolize tabersonine away from vindoline. The first two steps of vindoline biosynthesis have been overexpressed in C. roseus hairy root cultures which accumulated

16-hydroxytabersonine and 16-methoxytabersonine. However, two new metabolites were detected, and there were transcriptional changes detected in TIA genes [33]. There is also little known of possible catabolic pathways of TIAs in C. roseus [21,22], and expression of these catabolic genes would likely decrease the economic viability of large scale C.

5 roseus hairy root cultures [34]. Heterologous hosts are an interesting possibility, but as of now would require strictosidine feeding because of low yields in adapting the endogenous metabolism toward strictosidine [29]. Even if higher rates were achieved, it may disrupt other metabolic function in the cell, and slow the growth of cultures or make them unviable. Additionally, it may be difficult to clone and culture a single heterologous host that expresses all the needed genes to achieve viable production without precursor feeding because of the 30+ enzymatic steps needed to complete the pathway [3,34,35].

Tabersonine Metabolism

Thus far, most investigations of tabersonine metabolism have focused on characterizing the kinetics of a single specific enzyme in the pathway or locating the specific cell type or location where that enzyme or its genes are expressed. Most kinetic investigations have involved using a radiolabeled substrate or cofactor of known concentration with a protein extract for the reaction. After extracting the metabolites, they would be separated by chromatography and turnover of the substrate quantified by scintillation counting of the specific product [9,14,17,18]. However, in the characterization of 3, Liscombe et al searched for candidate genes that performed an unknown SAM-dependent methyltransferase and was expressed in the leaves, which had already been established of the enzyme. Three candidate genes were found, produced in a recombinant E. coli strain, and tested for activity with 16-methoxy-2,3-dihydro-3- hydroxytabersonine. Only one candidate gene produced an enzyme that acted upon the substrate, and the kinetics were characterized [5]. The characterization of the third step in the pathway involving tabersonine-3-oxygenase and tabersonine-3-reductase was performed in much the same way [15]. In these previous studies, enzymatic reactions

6 rates were calculated using rate of formation of product, or from medium containing only proteins from the heterologous hosts containing the gene for that specific step.

Enzyme Kinetics

All enzymes characterized thus far in the pathway exhibit single substrate

Michaelis-Menten enzyme kinetics, or have followed that pattern when co-substrates such as S-adenosyl-L-methionine (SAM) have been in excess [5,9,14,15,17,18].

Tabersonine 16-hydroxylase has shown substrate inhibition at tabersonine concentrations above 30 μM [14], while steps involving SAM have shown product inhibition towards S- adenosyl-L-homocysteine, which is regenerated to SAM through another metabolic process. [5,9] The terminal enzyme, deacetylvindoline-O-acetyltransferase, has also shown product inhibition from its co-substrate’s product, CoA (from acetyl-CoA). [17]

The dearth of knowledge of catabolism of alkaloids in C. roseus makes it an essential part of research to increase production of valuable molecules. While a heterologous host may be able to eliminate catabolism because only metabolic genes are added, as yet there has not been a line created that includes all metabolic steps or eliminates the use of expensive precursor feeding. Conversely, homologous hosts exhibit a complex regulatory system, and it is not known what role catabolism plays in this. In this work, the catabolism of alkaloids in C. roseus is investigated using desalted whole

leaf protein extracts.

Materials and Methods

Chemicals

Tabersonine (BOC Sciences, 2mM) and clotrimazole (Acros Organics, 25mM) were dissolved in pure ethanol. Nicotinamide adenine dinucleotide phosphate (NADPH)

(Enzo Life Sciences, 67mM) and S-adenosyl-L-methionine (Matrix Scientific, 2mM)

were dissolved in DI water.

Protein Extraction

Two grams of fresh leaf tissue harvested from mature C. roseus vt. Vinca Cora was homogenized in a mortar and pestle with 3.5 ml of extraction buffer (100 mM tris-

HCl, pH 7.6, 13 μM β-mercaptoethanol) at 4°C. The slurry was centrifuged at 5,000 g for

5 minutes at 4°C, and the supernatant drawn off. An additional 2 ml of extraction buffer

was added to the remaining solids, the tube was vortexed for 1 minute, and centrifuged at

5,000 g for 5 minutes at 4°C. The supernatant was drawn off, mixed with the first

extraction, and filtered through a 0.45 μm syringe filter. The filtrate was then desalted on

a PD-10 G25 desalting column (GE Healthcare) using the spin protocol. Protein

concentrations were determined using the Bradford protein assay protocol.

Search for Unknown Metabolites

Enzymatic Reactions

All reactions were performed in a 1-ml reaction volume. Reactions were incubated at 30°C for 120 minutes. Reactions contained 1.5-2 mg of protein, and the following concentrations for cofactors: 300 μM NADPH and S-adenosyl-L-methionine

8 and 30 μM tabersonine. For reactions that did not contain a cofactor, volumes of the solvent for that stock solution were added to keep ethanol concentrations and total volume of reaction mixture equal across reactions. Reactions were initiated with the addition of NADPH. Reactions were terminated by the addition of 4 ml of ethyl acetate.

Ethyl Acetate Extraction of Alkaloids

After the reaction was terminated by the addition of 4 ml of ethyl acetate, the samples were vortexed for 1 minute, and centrifuged at 10,000 RPM. The organic phase was drawn off and added to a clean glass vial, and an additional 4 ml of ethyl acetate was added to the aqueous phase, extracted in the same manner, and the organic phase was added to the same vial. The samples were dried under nitrogen and dissolved in 0.5 ml methanol and filtered through a 0.2 μm syringe filter for HPLC-UV analysis.

Kinetic Studies

Initial Rate Experiments

Initial rate experiments were run in a 1 ml reaction volume. Reactions contained

1.2 mg of protein, 10 μM clotrimazole, and the following concentrations for cofactors:

300 μM NADPH and S-adenosyl-L-methionine and 10-200 μM tabersonine. Reactions

were initiated with the addition of Tabersonine. Reactions were terminated by the

addition of 4 ml of -20°C acetone after 20 minutes and run in triplicate for each

tabersonine concentration.

Acetone Protein Precipitation Protocol for Alkaloid Extraction

After the reaction was terminated by the addition of 4 ml of cold (-20°C) acetone,

the samples were incubated at -20°C for 1 hour and filtered with a 0.45 μm syringe filter

9 to remove the precipitated protein. The mixture was dried under nitrogen until ~2 ml of solution remained, frozen at -86°C, and freeze-dried under vacuum. The samples were then dissolved in 0.5 ml methanol and filtered through a 0.2 μm syringe filter for HPLC-

UV analysis.

HPLC-UV Analysis

The HPLC method used is a modification of a method described by Peebles et al.

(2005) [23]. 20 microliters of the alkaloid extract were injected into a Phenomenex Luna

C18(2) HPLC column (250 mm x 4.6 mm). The HPLC system was a Waters system consisting of a 1525 Binary Pump, a 717plus Autosampler, and a 2996 Photodiode

Detector. For the first 5 minutes, the mobile phase was a 30:70 mixture of acetonitrile:100mM ammonium acetate (pH 7.3) at a flow rate of 1 mL/min. Over the next 10 minutes, the mobile phase was linearly ramped to a 64:36 mixture, which was maintained for 15 minutes. Over the next 5 minutes, the flow rate was linearly increased to 1.4 mL/min. For the next 5 minutes, the mobile phase ratio was increased to 80:20, and maintained for 15 minutes. Flow was then returned to starting conditions and column equilibrated. Standards used were strictosidine, tetrohydroalstonine, serpentine, vindoline, catharanthine, ajmalicine, hörhammericine, lochnericine, tabersonine, vincristine, and vinblastine.

Theory

Determination of Experimental Reaction Rates

Ci - Ci-Cb - Cf v = (1) ∆t

Equation 1 shows how experimental reaction rates were calculated. v is the rate of

reaction, Ci is the beginning concentration of tabersonine, Cb is the concentration of tabersonine after incubation with boiled protein extracts, Cf is the concentration of tabersonine after incubation with protein extracts, and Δ t is the time of incubation. The

consumption of tabersonine by nonenzymatic reactions, likely oxidation reaction and

calculated by substracting the initial tabersonine concentration by concentration after

incubation with boiled enzymes, is used to ensure kinetic calculations only include

enzymatic rates of reaction. Enzymatic activity was calculated by converting rate to units

of katals (mol/sec) and dividing by amount of protein used in that reaction.

Initial Kinetic Model to Determine Time Scale

To determine a probable time scale to identify downstream metabolic products of

tabersonine in a protein extract solution, a mathematical model was created using previously experimentally deduced kinetic parameters and Polymath ODE Solver. Since protein extract from the leaves of the plant were used, only reactions going towards vindoline from tabersonine were included in the model, shown in Figure 2 with all co- substrates and inhibitors. Briefly, it was assumed that all co-substrates for each enzymatic step were in excess, and so the biochemical reaction rates were only a function of the tabersonine metabolite and followed single substrate Michaelis-Menten kinetics. The differential equations for each step are shown in Equation 2:

dCS vm,S-1CS-1 vm,S CS = rS-1 - rS = - (2) dt Ci Ci Km,S-1 1+ +CS-1 Km,S 1+ +CS KI,S-1 KI,S where CS is the concentration of the metabolite, rS-1 is the rate of reaction generating metabolite CS, rS is the rate of reaction consuming metabolite CS, and CS-1 is the

concentration of the metabolite in the step before CS. The rate equations are single

11 substrate Michaelis-Menten equations with competitive inhibition, where vm and K m refer to the max velocity and Michaelis-Menten constant for steps S and S-1, and KI and Ci are the inhibition constant and concentration of inhibitor for that enzyme. If there is no inhibition on a step, the concentration of that inhibitor would be zero, and so the inhibition term would drop out of the equation. The first step in the pathway also exhibits substrate inhibition for tabersonine concentrations greater than 30 μM [14], and so the initial concentration for tabersonine was set to that value. The kinetic constants used are shown in Table 1

Results and Discussion

Search for Unknown Metabolites

Initial Kinetic Model to Determine Time Scale

The results of the initial kinetic model are shown in Figure 3. As expected, there is negligible accumulation of intermediate metabolites in the pathway. This is because tabersonine is one of the metabolic nodes in the pathway, where the first step away from tabersonine is likely the rate limiting step. The concentration of SAH, an inhibitor to and product of a required substrate in two steps in the pathway, increases at twice the rate of tabersonine. In a biological system, SAH is regenerated into SAM in a separate biochemical process, and so would not accumulate to those levels. However, in this model, it did not make a large impact on the rates of reaction for the steps in which it acts as an inhibitor. Using the results of the model, two hours was chosen as the reaction time for the initial sets of enzymatic studies.

Initial Enzymatic Reactions

Initial enzymatic reactions were run at 30°C with a starting concentration of tabersonine of 30 μM and incubated for two hours. Figure 4 and Figure 5 show the chromatograms extracted at 254 nm and 329 nm, respectively. No accumulation of any alkaloid, either previously known or unknown, was observed. The vindoline present in the samples is from native production in the leaves and does not significantly change throughout the two hours. Tabersonine was also nearly completely consumed in the two- hour timeframe, and there was no difference in tabersonine consumption between samples containing either NADPH and SAM or both substrates, shown in Figure 6.

Additional experiments were run with clotrimazole, an inhibitor of tabersonine 16- hydroxylase, with a concentration of 200 μM clotrimazole, which is 4 times the IC 50 [14].

In both cases, as shown in Figure 7, the enzymatic activity observed, 3230 ± 160 and

3100 ±150 pkat g protein -1 for samples with and without clotrimazole, respectively, was an order of magnitude greater than observed by St-Pierre et al for tabersonine 16-

hydroxylase, which was 340 pkat g protein -1. It should be noted that the difference in enzymatic activity between samples is similar to the activity of tabersonine 16- hydroxylase observed by St-Pierre et al under similar conditions. These data appear to support the presence of another enzyme acting upon tabersonine in the leaves of C. roseus.

Kinetic Studies

Error in Ethyl Acetate Extraction

Kinetic calculations were attempted to determine the kinetic parameters of the enzyme. However, when a higher initial concentration of tabersonine was used, it was

13 found that the results exhibited a high amount of experimental error. This was likely caused by an increased amount of ethanol in the reaction medium from the tabersonine stock solution. The increased ethanol concentration could have affected the solute partitioning coefficient, or the alkaloids may have been in a solid, gel-like phase observed to form between the aqueous and organic phases during extractions. Tabersonine has a low water solubility, and so ethanol was used as the solvent for the stock solution. To achieve a 200 μM initial concentration, 11% of the volume added to the sample mixtures was ethanol. As shown in Figure 8, when tabersonine was added to protein extract and immediately extracted using the ethyl acetate method, the ethyl acetate method produced extraction efficiencies of 22 ± 13% and 43 ± 5% for 50 and 200 micromolar samples, respectively. A protein precipitation method to achieve separation was developed, and showed a consistent extraction efficiency of 87 ± 3% and 87 ± 5% for 50 and 200 micromolar samples, respectively, when tested in the same manner. To carry out subsequent kinetic experiments, the acetone protein precipitation method was used, and the concentration obtained by HPLC-UV was divided by 87% to normalize the boiled enzyme and final concentrations in respect to the initial added tabersonine concentration.

Kinetic Studies

Initial rate experiments were performed with tabersonine to determine the kinetic parameters of this possible new enzyme. Experiments were performed with clotrimazole

to further decrease the activity of tabersonine 16-hydroxylase. First, the rate of non-

enzymatic degradation was found using boiled enzyme extracts, shown in Figure 9.

The rate of non-enzymatic consumption of tabersonine appears to be linear in this concentration range. This data was then used to eliminate non-enzymatic consumption for calculating the rate of enzymatic consumption of tabersonine at each initial concentration.

To analyze the kinetics of tabersonine consumption, the data was plotted on a

Lineweaver-Burke plot, shown in Figure 10. For a Michaelis-Menten enzyme-catalyzed reaction, this plot would be expected to be linear. The exponential increase of inverse velocity with increasing inverse substrate concentration is a trend associated with positively cooperative allosteric enzymes shown in Figure 11 [36,37]. Additionally, other methods of linearizing Michaelis-Menten enzymatic rates to calculate rate constants also yielded trends associated with cooperative allosteric enzymes, shown in Figure 12 and

Figure 13. These trends support the presence of a positively cooperative allosteric enzyme in the leaves of C. roseus.

n vm [Tab ] v= " n (3) Km + [Tab ]

Equation 3 shows the rate equation for allosteric enzymes, where v is velocity of

" the reaction, vm is the max velocity for the enzyme, Km is the Michaelis-Menten constant

for allosteric enzymes, [Tab ] is the concentration of tabersonine, and n is the Hill coefficient with a value greater than one corresponding to positive cooperativity. To calculate the kinetic parameters of allosteric enzymes, there is a graphical method to linearize the data, as in Michaelis-Menten enzymes. However, to perform this analysis it is necessary to know one of the kinetic parameters ( " , n, or vm). Since it is uncertain

that vm was achieved with the initial concentrations used, a model was used to determine

kinetic parameters.

Kinetic Model

vm Tab v = (4) K + Tab m Equation 4 shows the rate equation for a simple Michaelis-Menten enzyme kinetics, where v is velocity of the reaction, vm is the max velocity for the enzyme, Km is the Michaelis-Menten constant, and [Tab ] is the concentration of tabersonine. The allosteric rate equation is shown in Equation 3. The step size for the model was 1 μM tabersonine and the model was fit to the experimental data in Microsoft Excel using the least sum of squared residuals method and the solver add-in. The results of the model are shown in Figure 14.

As expected from the results of the previous graphing methods, the allosteric model fit closer with the data, with more reasonable results. To fit a Michaelis-Menten

10 -1 12 model, vm and Km values went to the order of 10 μM min and 10 μM respectively to

yield a linear trend within reasonable tabersonine concentrations. This was caused by the

sigmoidal shape of the velocity vs. tabersonine concentration data, while a Michaelis-

Menten enzyme exhibits a hyperbolic curve shape. The allosteric model produced values

-1 of 5.4 ± 1.9 μM min for vm, 16,000 ± 14,000 μM for " , and a Hill coefficient of 2.0 ±

0.2. The determined value of vm was an order of magnitude larger than the value found by

St-Pierre et al [14] for tabersonine 16-hydroxylase and the two terminal steps towards vindoline [17,18], and smaller than the vm for the middle enzymatic steps [5,9]. This

shows a reasonable value for vm was determined. It should also be noted the value of " does not hold the same physical significance for allosteric enzymes, as its value is very dependent on the value of the Hill coefficient and so will be much larger for positively cooperative allosteric enzymes.

These calculated values are also plausible when examining how a catabolic enzyme would fit into the regulation of tabersonine metabolism in the leaves of C. roseus. Looking at the known steps between tabersonine and vindoline, the first step has the highest Km and one of the lower vm values. Those two values together are what makes

it the rate limiting step; even though the vm is low by itself, it also takes a higher concentration of tabersonine to reach its max velocity. With the middle steps, a higher max velocity and slightly smaller Km lead to rapid consumption of substrate by those

enzymes. The terminal two steps have max velocities in the same order of magnitude as

the first step, but much lower Km values lead to those rates approaching the max rate

much quicker. This leads to almost no accumulation of intermediates between

tabersonine and vindoline. Substrate inhibition of tabersonine 16-hydroxylase, the first

step in the pathway, at concentrations above 30 μM [14] also fits with the calculated parameters, as just above this concentration of tabersonine the rate of reaction for the

catabolic enzyme is highly effected by small changes in tabersonine concentration. This

would lead to large increases in reaction rate from relatively small increases in

tabersonine concentration.

It is plausible that this allosteric enzyme catalyzes the degradation tabersonine in

C. roseus. Experimental evidence in Figure 3 and 4 shows there was no significant accumulation of any other metabolites with the consumption of tabersonine. Enzymatic activity was also much higher than that seen in previous works studying the metabolism of tabersonine toward vindoline in leaf protein extracts [14]. Allosteric enzymes are commonly affected by the endpoint of the pathway of the substrate they act upon [37], and the presence of endogenous vindoline in the desalted protein extract could have a

17 positive homotropic effect on the enzyme [36]. Tabersonine itself would also have a homotropic effect on the enzyme, and likely any other intermediate metabolite because of their close conformational relationship. It would also fit that tabersonine is the substrate of this degradation enzyme because of its differing metabolism in various locations in the plant. The same enzyme could act as regulatory enzyme in the whole plant, with other tabersonine metabolites having positively cooperative interactions with the enzyme.

Allosteric kinetics in the first step of a degradation pathway away from tabersonine would also help in the regulation of the alkaloid levels in the plant. If concentrations of tabersonine and tabersonine metabolites were low, the rate of the degradation pathway would not be greatly affected by increasing tabersonine concentration. However, when concentrations of tabersonine metabolites reached static levels in the plant, an increasing tabersonine concentration would greatly increase the rate of the degradation of tabersonine to decrease the flux going toward vindoline in the leaves or other metabolites in the roots of the plant, keeping those alkaloid concentrations at stable levels.

Conclusion

Evidence was found to support the presence of an allosteric enzyme acting upon tabersonine in C. roseus. A model was fit to the data to calculate the kinetic parameters from the Hill equation, although there was a large amount in error for the binding

" constant (Km), which is due to the mathematical nature of the equation. As there was no accumulation of any similar alkaloids observed in any HPLC chromatogram, it is likely this enzyme catalyzes the first step in a degradation pathway out of the TIA pathway.

Continued pursuit of eliciting the catabolic pathways of TIAs is pertinent to homologous production of vinblastine and vincristine in C. roseus.

Future Works

Future works on this topic include designing a procedure to get a higher initial concentration of tabersonine so that the max velocity of the reaction would be reached.

This could involve protocols that add the ethanolic solution to the reaction tube, drying the ethanol off, and then adding the other components of the reaction medium to initiate the reaction. A concern with such a protocol would be the rate at which tabersonine dissolves back into solution, as that would influence the rate of the enzyme. The exact solubility of tabersonine in water and ethanol are unknown, and a more concentrated stock solution could also be attempted.

Other future works on this topic could include performing the same experiments on root protein extract to identify if the enzyme or a similar one is expressed there.

Methods used to explore the cell type and subcellular location of enzymes in the vindoline pathway could also be used, including carborundum abrasion [9], laser-capture microdissection [2], and separation of soluble and microsomal proteins [14].

Radiolabeled tabersonine could also be used to confirm the presence of a degradation pathway. Such experiments would also have to factor in non-enzymatic consumption of tabersonine.

Illustrations

Figure 1: Figure 1a shows the terpenoid indole alkaloid pathway going toward vinblastine and vincristine. The steps after tabersonine shown occur in the leaves. Figure 1b shows the metabolic pathway of tabersonine that occurs in the roots of the plant. Solid line arrows are known biochemical steps, dotted arrows are unknown steps, and double arrows are multiple known steps. Abbreviations; T16H: tabersonine 16-hydroxylase; 16OMT: tabersonine 16-O-methyltransferase; T3O: tabersonine 3-oxygenase; T3R: tabersonine 3-reductase; NMT: 3-hydroxy-16-methoxy-2,3-dihydrotabersonine N-methyltransferase; D4H: deacetoxyvindoline 4-hydroxylase; DAT: deacetylvindoline O-acetyltransferase; PRX 1: peroxidase 1 or 3’,4’-anydrovinblastine synthase; T6,7E: tabersonine 6,7-epoxidase; 19H: tabersonine 19-hydroxylase; MAT: minovincinine 19-hydroxy-O-acetyltransferase

Figure 2: Metabolic pathway from tabersonine to vindoline shown with all co-substrates, products, and inhibitors.

Enzyme Inhibitor Substrate vm / ( μM / min) Km / μM Ki / μM Literature

T16H -- Tab 0.24 11 -- [14]

16OMT -- 16-OH 34.80 2.6 -- [9]

16OMT SAH ------0.6 [9]

T3O &T3R -- 16-MEO ------[15]

NMT -- HMDHT 8.04 8.8 -- [5]

NMT SAH ------7.3 [5]

D4H -- DOX 0.175 0.03 -- [18]

DAT -- Dace 0.361 1.3 -- [17]

Table 1: Kinetic constants used instyle="baselin model. Kinetic studies have not been performed on T3O and T3R to determine kinetic constants, and so it was assumed that this step is not rate limiting and 16- methoxytabersonine was immediately transformed to 3-hydroxy-16-methoxy-2,3-dihydrotabersonine in the model. Enzyme abbreviations are the same as in Figure 1. Abbreviations; Tab: Tabersonine; 16-OH: Tabersonine 16-hydroxytabersonine; SAH: s-adenosyl-L homocysteine; 16-MEO: 16-methoxytabersonine; HMDHT: 3-hydroxy-16-methoxy-2,3-dihydrotabersonine; DOX: desacetoxyvindoline; Dace: deacetylvindoline.

20 Concentration Concentration / μM

0 0 50 100 150 200 250 300 350 400 Time / Minute

16-hydroxytabersonine Tabersonine SAH Deacetylvindoline Vindoline

Figure 3: Results of metabolic model. SAH is s-adenosyl-L-homocysteine, the product of a co-substrate in two steps in the pathway.

0.20

0.18 Vindoline 0.16

0.14

0.12

0.10 A U

0.08

0.06

0.04

0.02

0.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00 65.00 70.00 Minutes

Figure 4: 254 nm chromatogram of extracted reaction medium. The top chromatogram is the extract of reaction medium after 2 hours of incubation at 30°C with NADPH and s-adenosyl-L-methionine, and the bottom chromatogram is the extract immediately after addition of tabersonine.

0.018

0.016 Vindoline Tabersonine 0.014

0.012

0.010

AU 0.008

0.006

0.004

0.002

0.000

0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00 55.00 60.00 65.00 70.00 Minutes Figure 5: 329 nm chromatogram of extracted reaction medium. The top chromatogram is the extract of reaction medium after 2 hours of incubation at 30°C with NADPH and s-adenosyl-L-methionine, and the bottom chromatogram is the extract immediately after addition of tabersonine.

10 Tabersonine Consumed Tabersonine Consumed / μM 5

0 ABC

Figure 6: Tabersonine consumption in samples incubated at 30°C for 2 hours with a starting tabersonine concentration of 30 μM and: A: 300 μM of NADPH and s-adenosyl-L-methionine; B: 300 μM s-adenosyl-L- methionine; and C: 300 μM of NADPH. Error bars represent the standard deviation of three samples.

3500

3000

2500

2000

1500

1000

Specific Specific Activity / pkat / g protein 500

0 Control Clotrimazole

Figure 7: Specific activity of samples without (control) and with clotrimazole. Activities were 3230 ± 160 and 3100 ± 150 pkat g protein -1 for samples without and with clotrimazole, respectively. Specific activity for tabersonine 16-hydroxylase was 140 pkat g protein-1 as reported by St-Pierre et al. Error bars represent the standard deviation of samples run in triplicate.

100% 90% 80% 70% 60% 50% 40% 30%

Extraction Effeciency 20% 10% 0% 50 μM 200 μM Ethyl Acetate Extraction Acetone Precipitation Extraction

Figure 8: Results of testing both extraction protocols extraction efficiency. Extraction efficiency is defined as alkaloid extract concentration over initial concentration in test sample. Error bars represent the standard deviation of samples run in triplicate.

2.5

2 y = 0.0083x R² = 0.9895

1.5

1 Rate Rate of Nonenzymatic Consumpution Consumpution / (μM/min) 0.5

0 0 50 100 150 200 250 [Tabersonine] / μM

Figure 9: Kinetic data from non-enzymatic consumption of tabersonine using the enzymatic reaction protocol with boiled enzyme extract. Error calculated by propagating standard deviation of final concentrations of samples run in triplicate through change in concentration over time for boiled samples.

μM) 40

/ / (min/ 30 v 1/

0 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 1 / [Tabersonine] / ( μM-1)

Figure 10: Lineweaver-Burke plot from the initial rate experiments. Error calculated by propagating the inverse of error in rate calculations.

v 1 / / 1

1 / [S]

Michaelis-Menten Positive Cooperitivity Negative Cooperitivity

Figure 11: Trends associated with a Lineweaver-Burk plot for standard Michaelis-Menten enzymes and positive and negative cooperativity allosteric enzymes. Figure adopted from [36]

0.03

0.025

0.02

0.015

0.01 / / / [S](min-1)

v 0.005

0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 v / (μM / min)

Figure 12: Modified Eadie-Hofstee plot from the initial rate experiments. x-axis error calculated by propagating standard deviation of final whole protein and boiled protein samples run in triplicate through Equation 1. Y-axis error is error in v divided by beginning substrate concentration for that sample set.

/ / [S] v

No Cooperitivity Positive Cooperitivity Negative Cooperitivity

Figure 13: Modified Eadie-Hofstee plot showing Trends associated with standard Michaelis-Menten enzymes and positive and negative cooperativity allosteric enzymes. Figure adopted from [36]

v / (μM/min) v 2

0 0 50 100 150 200 250 [Tabersonine] / μM

Figure 14: Allosteric kinetic model shown with experimental data points. Error bars are the standard deviation from calculations at that concentration, and dashed lines are the error from the model. The solid line is the best fit from the data. Error was calculated by fitting the model to plus the standard deviation from each point, and fitting to minus the standard deviation.

References

[1] Bernonville, T. D. De; Carqueijeiro, I.; Lanoue, A.; Lafontaine, F.; Bel, P. S.; Liesecke, F.; Musset, K.; Oudin, A.; Glévarec, G.; Pichon, O.; et al. Folivory Elicits a Strong Defense Reaction in Catharanthus Roseus : Metabolomic and Transcriptomic Analyses Reveal Distinct Local and Systemic Responses. Nat. Publ. Gr. 2017 , No. July 2016, 1–14 [2] Murata, J.; Luca, V. D. Localization of Tabersonine 16-Hydroxylase and 16-OH Tabersonine-16- O- Methyltransferase to Leaf Epidermal Cells Defines Them as a Major Site of Precursor Biosynthesis in the Vindoline Pathway in Catharanthus Roseus. Plant J. 2005 , 44 (4), 581–594 [3] Sottomayor, M.; Barcel, A. R. Peroxidase from Catharanthus Roseus (L.) G. Don and the Biosynthesis of Alpha-3’,4’-Anhydrovinblastine: A Specific Role for a Multifunctional Enzyme. Protoplasma 2003 , 222 (1–2), 97–105 [4] Duangteraprecha, S.; Hirata, K.; Morihara, E.; Nakae, M.; Katayama, H.; Honda, M.; Miyamoto, K. Non-Enzymic Coupling of Vindoline and Catharanthine to Synthesize 3’,4’- Anhydrovinblastine under Irradiation with near-Ultraviolet Light. J. Ferment. Bioeng. 1997 , 83 (3), 227–232 [5] Liscombe, D. K.; Usera, A. R.; O’Connor, S. E. Homolog of Tocopherol C Methyltransferases Catalyzes N Methylation in Anticancer Alkaloid Biosynthesis. Proc. Natl. Acad. Sci. U. S. A. 2010 , 107 (44), 18793–18798 [6] Sottomayor, M.; Lopes Cardoso, I.; Pereira, L. G.; Ros, B. Peroxidase and the Biosynthesis of Terpenoid Indole Alkaloids in the Medicinal Plant Catharanthus Roseus (L.) G. Don. Phytochem. Rev. 2004 , 3, 159–171 [7] Roepke, J.; Salim, V.; Wu, M.; Thamm, A. M. K.; Murata, J.; Ploss, K.; Boland, W.; De Luca, V. Vinca Drug Components Accumulate Exclusively in Leaf Exudates of Madagascar Periwinkle. Proc. Natl. Acad. Sci. 2010 , 107 (34), 15287–15292 [8] St-Pierre, B.; Laflamme, P.; Alarco, A. M.; De Luca, V. The Terminal O-Acetyltransferase Involved in Vindoline Biosynthesis Defines a New Class of Proteins Responsible for Coenzyme A- Dependent Acyl Transfer. Plant J. 1998 , 14 (6), 703–713 [9] Levac, D.; Murata, J.; Kim, W. S.; De Luca, V. Application of Carborundum Abrasion for Investigating the Leaf Epidermis: Molecular Cloning of Catharanthus Roseus 16- Hydroxytabersonine-16-O- Methyltransferase. Plant J. 2008 , 53 (2), 225–236 [10] Vazquez-Flota, F.; De Carolis, E.; Alarco, A. M.; De Luca, V. Molecular Cloning and Characterization of Desacetoxyvindoline-4-Hydroxylase, a 2-Oxoglutarate Dependent- Dioxygenase Involved in the Biosynthesis of Vindoline in Catharanthus Roseus (L.) G. Don. Plant Mol. Biol. 1997, 34 (6), 935–948 [11] Qu, Y.; Thamm, A. M. K.; Czerwinski, M.; Masada, S.; Kim, K. H.; Jones, G.; Liang, P.; De Luca, V. Geissoschizine Synthase Controls Flux in the Formation of Monoterpenoid Indole Alkaloids in a Catharanthus Roseus Mutant. Planta 2018 , 247 (3), 625–634 [12] Qu, Y.; Easson, M. E. A. M.; Simionescu, R.; Hajicek, J.; Thamm, A. M. K.; Salim, V.; De Luca, V. Solution of the Multistep Pathway for Assembly of Corynanthean, Strychnos, Iboga, and Aspidosperma Monoterpenoid Indole Alkaloids from 19 E -Geissoschizine. Proc. Natl. Acad. Sci. 2018 , 115 (12), 3180–3185 [13] Caputi, L.; Franke, J.; Farrow, S. C.; Chung, K.; Payne, R. M. E.; Nguyen, T.-D.; Dang, T.-T. T.; Carqueijeiro, I. S. T.; Koudounas, K.; Bernonville, T. D. de; et al. Missing Enzymes in the Biosynthesis of the Anticancer Drug Vinblastine in Madagascar Periwinkle. Science (80-. ). 2018 , 4100

[14] St-Pierre, B.; De Luca, V. A Cytochrome P-450 Monooxygenase Catalyzes the First Step in the Conversion of Tabersonine to Vindoline in Catharanthus Roseus. Plant Physiol. 1995 , 109 (1), 131–139 [15] Qu, Y.; Easson, M. L. A. E.; Froese, J.; Simionescu, R.; Hudlicky, T.; De Luca, V. Completion of the Seven-Step Pathway from Tabersonine to the Anticancer Drug Precursor Vindoline and Its Assembly in Yeast. Proc. Natl. Acad. Sci. U. S. A. 2015 , 112 (19), 6224–6229 [16] Giddings, L. A.; Liscombe, D. K.; Hamilton, J. P.; Childs, K. L.; DellaPenna, D.; Buell, C. R.; O’Connor, S. E. A Stereoselective Hydroxylation Step of Alkaloid Biosynthesis by a Unique Cytochrome P450 in Catharanthus Roseus. J. Biol. Chem. 2011 , 286 (19), 16751–16757 [17] Power, R.; Kurz, W. G. W.; De Luca, V. Purification and Characterization of Acetylcoenzyme A: Deacetylvindoline 4-O-Acetyltransferase from Catharanthus Roseus. Arch. Biochem. Biophys. 1990 , 279 (2), 370–376 [18] De Carolis, E.; De Luca, V. A Novel 2-Oxoglutarate-Dependent Dioxygenase Involved in Vindoline Biosynthesis: Characterization, Purification and Kinetic Properties. Plant Cell. Tissue Organ Cult. 1994 , 38 (2–3), 281–287 [19] Rodriguez, S.; Compagnon, V.; Crouch, N. P.; St-Pierre, B.; De Luca, V. Jasmonate-Induced Epoxidation of Tabersonine by a Cytochrome P-450 in Hairy Root Cultures of Catharanthus Roseus. Phytochemistry 2003 , 64 (2), 401–409 [20] Laflamme, P.; St-Pierre, B.; De Luca V. Molecular and Biochemical Analysis of a Madagascar Periwinkle Root-Specific Minovincinine-19-Hydroxy-O-Acetyltransferase. Plant Physiol. 2001 , 125 (1), 189–198 [21] Verma, P.; Mathur, A. K.; Khan, S. A.; Verma, N.; Sharma, A. Transgenic Studies for Modulating Terpenoid Indole Alkaloids Pathway in Catharanthus Roseus: Present Status and Future Options. Phytochem. Rev. 2017 , 16 (1), 19–54 [22] Sander, G. W. Quantitative Analysis of Metabolic Pathways in Catharanthus Roseus Hairy Roots Metabolically Engineered for Terpenoid Indole Alkaloid Overproduction, 2009 [23] Peebles, C. A. M.; Hong, S. B.; Gibson, S. I.; Shanks, J. V.; San, K. Y. Transient Effects of Overexpressing Anthranilate Synthase α and β Subunits in Catharanthus Roseus Hairy Roots. Biotechnol. Prog. 2005 , 21 (5), 1572–1576 [24] Peebles, C. A. M.; Sander, G. W.; Hughes, E. H.; Peacock, R.; Shanks, J. V.; San, K. Y. The Expression of 1-Deoxy-d-Xylulose Synthase and Geraniol-10-Hydroxylase or Anthranilate Synthase Increases Terpenoid Indole Alkaloid Accumulation in Catharanthus Roseus Hairy Roots. Metab. Eng. 2011 , 13 (2), 234–240 [25] Pan, Q.; Wang, Q.; Yuan, F.; Xing, S.; Zhao, J.; Choi, Y. H.; Verpoorte, R.; Tian, Y.; Wang, G.; Tang, K. Overexpression of ORCA3 and G10H in Catharanthus Roseus Plants Regulated Alkaloid Biosynthesis and Metabolism Revealed by NMR-Metabolomics. PLoS One 2012 , 7 (8), 1–14 [26] Peebles, C. A. M.; Hughes, E. H.; Shanks, J. V.; San, K. Y. Transcriptional Response of the Terpenoid Indole Alkaloid Pathway to the Overexpression of ORCA3 along with Jasmonic Acid Elicitation of Catharanthus Roseus Hairy Roots over Time. Metab. Eng. 2009 , 11 (2), 76–86 [27] El-Sayed, M.; Verpoorte, R. Effect of Phytohormones on Growth and Alkaloid Accumulation by a Catharanthus Roseus Cell Suspension Cultures Fed with Alkaloid Precursors Tryptamine and Loganin. Plant Cell. Tissue Organ Cult. 2002 , 68 (3), 265–270 [28] Vázquez-Flota, F.; Hernández-Domínguez, E.; De Lourdes Miranda-Ham, M.; Monforte-González, M. A Differential Response to Chemical Elicitors in Catharanthus Roseus in Vitro Cultures. Biotechnol. Lett. 2009 , 31 (4), 591–595

[29] Brown, S.; Clastre, M.; Courdavault, V.; O’Connor, S. E. De Novo Production of the Plant-Derived Alkaloid Strictosidine in Yeast. Proc. Natl. Acad. Sci. 2015 , 112 (11), 3205–3210 [30] Whitmer, S.; Canel, C.; Van Der Heijden, R.; Verpoorte, R. Long-Term Instability of Alkaloid Production by Stably Transformed Cell Lines of Catharanthus Roseus. Plant Cell. Tissue Organ Cult. 2003 , 74 (1), 73–80 [31] Sun, J.; Ma, L.; San, K. Y.; Peebles, C. A. M. Still Stable after 11 Years: A Catharanthus Roseus Hairy Root Line Maintains Inducible Expression of Anthranilate Synthase. Biotechnol. Prog. 2017 , 33 (1), 66–69 [32] Peebles, C. A. M.; Sander, G. W.; Li, M.; Shanks, J. V.; San, K. Y. Five Year Maintenance of the Inducible Expression of Anthranilate Synthase in Catharanthus Roseus Hairy Roots. Biotechnol. Bioeng. 2009 , 102 (5), 1521–1525 [33] Sun, J.; Zhao, L.; Shao, Z.; Shanks, J.; Peebles, C. A. M. Expression of Tabersonine 16- Hydroxylase and 16-Hydroxytabersonine-O-Methyltransferase in Catharanthus Roseus Hairy Roots. Biotechnol. Bioeng. 2018 , 115 (3), 673–683 [34] Zhao, J.; Verpoorte, R. Manipulating Indole Alkaloid Production by Catharanthus Roseus Cell Cultures in Bioreactors: From Biochemical Processing to Metabolic Engineering. Phytochem. Rev. 2007 , 6 (2–3), 435–457 [35] Zhu, X.; Zeng, X.; Sun, C.; Chen, S. Biosynthetic Pathway of Terpenoid Indole Alkaloids in Catharanthus Roseus. Front. Med. 2014 , 8 (3), 285–293 [36] Hammes, G. G.; Wu, C.-W. Kinetics of Allosteric Enzymes. Annu. Rev. Biophys. Bioeng. 1974 , 3, 1–33 [37] Traut, T. Allosteric Regulatory Enzymes ; 2008

Appendices

Appendix 1: Calculating Rate Constants from Initial Rate Experiments

Initial rate experiments are commonly used to determine enzymatic rate constants.

They involve using a known starting concentration of substrate and measuring the change of concentration over a specified time. These results can be graphed in a Langmuir-

Hinshelwood plot, shown in Figure 15.

5 4.5 4 3.5 3 2.5 2 1.5

/ (μM/min) 1 v v 0.5 0 0 50 100 150 200 [S] / μM

Figure 15: A modelled Langmuir-Hinshelwood plot of a Michaelis-Menten enzyme with a Km of 10 μM and -1 a vm of 5 μM min

However, it is difficult to determine kinetic parameters from this graph, as the vm is the asymptote of the hyperbolic curve, and the Km is the substrate concentration where the

rate is half of vm. Because of this, several different methods of linearizing Michaelis-

Menten enzyme rates have been developed. Here, the two discussed in this thesis will be described.

The first linearization method is a Lineweaver-Burke plot, or a reciprocal plot.

The linearized equation is shown in Equation 5.

1 K 1 1 = m + (5) v vm [S] vm

1 1 In this method, vs is plotted, shown in Figure 16. v [S]

0.6

0.5

0.4 v

1/ 0.3

0.2

0.1

0 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 1/[S]

Figure 16: Modelled Lineweaver-Burke plot of a Michaelis-Menten enzyme with a Km of 10 μM and a vm of 5 μM min -1

1 K 1 In this linearization method, the y-intercept is , the slope is m, and the x-intercept is . vm νm Km

Another method is the Eadie-Hofstee plot. The linearized equation is shown in

Equation 6.

v v = v - K (6) m m [S]

v In this method, v vs. is plotted, shown in Figure 17. [S]

v 3

0 0 0.1 0.2 0.3 0.4 0.5 0.6 v/[S]

Figure 17: Modelled Eadie-Hofstee plot of a Michaelis-Menten enzyme with a Km of 10 μM and a vm of 5 μM min -1

vm In this linearization method, the y-intercept is vm, the x-intercept is , and the slope is m K

-Km.

For allosteric enzymes, a Langmuir-Hinshelwood plot will have a sigmoidal shape instead of the hyperbolic shape seen in Michaelis-Menten enzymes. The shape of the sigmoid will depend on the value of n, with values higher than one corresponding positively cooperative enzymes, lower than 1 negatively cooperative enzymes, and values equal to 1 being a Michaelis-Menten enzyme. A plot showing all three is shown in Figure

18. For illustrative purposes, an “effective Km” ( KE) is used to show the effect the n value has on the changes in velocity over changes in substrate concentration. The KE is defined in Equation 7.

K = K1/n (7) E m

5 4.5 4 3.5 3

v 2.5 2 1.5 1 0.5 0 0 5 10 15 20 25 [S] n = 1 n = 2 n = 0.7

Figure 18: A modelled Langmuir-Hinshelwood plot of a Michaelis-Menten enzyme (n = 1), a positively cooperative allosteric enzyme (n = 2), and a negatively cooperative allosteric enzyme (n = 0.7) with a KE -1 of 10 μM and a vm of 5 μM min .

For Michaelis-Menten enzymes, Km is the substrate concentration where v = 0.5v m.

Defining a KE in the way it has been allows that value to also be the substrate concentration where v = 0.5v m for positively and negatively allosteric enzymes, which

makes graphically comparing these enzymes easier. For negatively cooperative allosteric

enzymes, changes in substrate concentration close to the KE has a smaller effect on

reaction rate than for a Michaelis-Menten enzyme, while the reverse is true for positively

cooperative allosteric enzymes. For the linearization of allosteric enzyme initial rate

experiment data, a hill plot is used, which requires the vm to be known. The linearized equation for this plot is shown in Equation 8.

v ln =nln [S]-lnKm (8) vm-v v ln vs. ln[S] is then plotted, shown in Figure 19. vm-v

2 )) -v m 0

v/(v 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 ln( -2

-4

-6 ln([S])

n = 1 n = 2 n = 0.7

Figure 19: A modelled Hill plot of a Michaelis-Menten enzyme (n = 1), a positively cooperative allosteric enzyme (n = 2), and a negatively cooperative allosteric enzyme (n = 0.7) with a KE of 10 μM and a vm of 5 μM min -1.

In this plot, the y-intercept is -ln( Km) and the slope is n. It should be noted that the y- intercept relates to the actual Km of the enzyme, not the KE discussed earlier.

Appendix 2: Polymath ODE Solver Input

Note: for the third step, kinetic constants have not been characterized, so the rate equation used is to prevent the metabolite consumed in that reaction, 16-methoxy-2,3- dihydro-3-hydroxytabersonine, from accumulating.