University of Groningen

Studies on Anthriscus sylvestris L. (Hoffm.)Hendrawati, Oktavia

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2011

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Hendrawati, O. (2011). Studies on Anthriscus sylvestris L. (Hoffm.): metabolic engineering of combinatorialbiosynthesis of podophyllotoxin. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 16-08-2021

StudieS on Anthriscus sylvestris L. (Hoffm.)

metabolic engineering of combinatorial biosynthesis of podophyllotoxin

Oktavia Hendrawati

This PhD project was carried out in the Department of Pharmaceutical Biology, Department of Molecular Biology of Plants and Department of Plant Physiology according to the requirements of the Graduate School of Science (Faculty of Mathematics and Natural Sciences, University of Groningen). This work was financially supported by Ubbo Emmius Fund of the University of Groningen, the Netherlands

ISBN: 978-94-6182-027-3

Cover design: Hester Nijhoff, www.hesternijhoff.nl

Layout and printing: Off Page, www.offpage.nl

This thesis is also available in electronic format at: http://dissertations.ub.rug.nl/

Copyright © 2011 by O. Hendrawati. All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without prior permission of the author.

RIJKSUNIVERSITEIT GRONINGEN

StudieS on Anthriscus sylvestris L. (Hoffm.)

metabolic engineering of combinatorial biosynthesis of podophyllotoxin

Proefschrift

ter verkrijging van het doctoraat in deWiskunde en Natuurwetenschappenaan de Rijksuniversiteit Groningen

op gezag van deRector Magnificus, dr. E. Sterken,in het openbaar te verdedigen op

vrijdag 28 oktober 2011om 14.30 uur

door

oktavia Hendrawatigeboren op 29 oktober 1977

te Surabaya, Indonesië

Promotores: Prof. dr. O. Kayser Prof. dr. W.J. Quax

Copromotor: Dr. H.J. Woerdenbag

Beoordelingscommissie: Prof. dr. R. Verpoorte Prof. dr. J.T.M. Elzenga Prof. dr. E.M.J. Verpoorte

Didedikasikan untuk papa, mama, Januar Hendrawati, Yuniarto dan Aprilarto Hendronoto Tjandra

The soul filled by love neither tires others nor grows tired

Taize

Paranimfen: Melanie Foeh Christina Avanti

ContentS

Scope of the thesis 9

Chapter 1 Metabolic engineering strategies for the optimization of medicinal and aromatic plants: realities and expectations 11

Chapter 2 Seasonal variations in the deoxypodophyllotoxin content and yield of Anthriscus sylvestris L. (Hoffm.) grown in the field and under controlled conditions 45

Chapter 3 Identification of lignans and related compounds in Anthriscus sylvestris by LC-ESI-MS/MS and LC-SPE-NMR 63

Chapter 4 In vitro regeneration of wild chervil (Anthriscus sylvestris L.) 79

Chapter 5 Agrobacterium mediated transformation of Anthriscus sylvestris with human cytochrome P450 3A4 followed by regeneration 93

Summary 111Samenvatting 113Ringkasan 116Acknowledgments 119

9

SCope of tHe tHeSiS

Podophyllotoxin is a lignan, which is used as a precursor of the anticancer drugs etoposide and teniposide. Both drugs are important in the treatment of lung and testicular cancers, Ewing’s sarcoma, lymphoma, glioblastoma and non-lymphomatic leukaemia. Three different methods are known for the production of podophyllotoxin: isolation from Podophyllum and Linum species, production by plant cell cultures, and organic synthesis.

To date, podophyllotoxin is commercially only obtained by extraction of the rhizome of Podophyllum species. The availability of Podophyllum species, however, is limited and the continuous demand of podophyllotoxin jeopardizes the natural sources. Podophyllum species are already listed on the endangered species list in India. Using plant cell cultures or organic synthesis for production purposes is not feasible, as the yields are limited and the price cannot compete with the collection of plants from wild habitats.

In the near future, the availability of podophyllotoxin will become a major bottleneck in supplying pharmaceutical needs. An alternative source of podophyllotoxin may be obtained from the hydroxylation at the C-7 position of the closely related lignan deoyxpodophyllotoxin yielding (epi)podophyllotoxin, the diastereoisomer of podophyllotoxin. The Apiaceae species Anthriscus sylvestris L. (Hoffm.) is a wild plant generally occurring in northern Europe containing deoxypodophyllotoxin as its main lignan constituent.

The aim of this thesis is to create and explore an alternative source of (epi) podophyllotoxin by production from deoxypodophyllotoxin. Current engineering strategies for the optimization of medicinal and aromatic plants are discussed in the introductory chapter 1. Deoxypodophyllotoxin is found in all parts of A.  sylvestris at concentrations between 0.03 and 0.15% (dw). To get a clear picture of the deoxypodophyllotoxin content over the developmental stages of the plant, we studied the variations of deoxypodophyllotoxin during different developmental stages. The results are described in chapter 2. We followed one biannual life cycle of A. sylvestris collected from 13 different locations in Europe that were grown in the field and in a climate room. Based on the results, we could indicate the best harvesting time of the plant for deoxypodophyllotoxin isolation and for choosing a suitable candidate plant for on-going cultivation. In chapter 3, we identified 14 compounds in the root extract of A. sylvestris using combined method of LC-ESI-MS/MS and LC-SPE-NMR. Based on these results, we propose a hypothetical biosynthetic pathway of 6-hydroxy-aryltetralin lignans in A. sylvestris. A. sylvestris allows an interesting approach for genetic modification to directly produce (epi)podophyllotoxin in the plant. Human cytochrome P450 3A4 was cloned in A. sylvestris via Agrobacterium tumefaciens. For transformation purposes we first needed to establish a regeneration protocol for A. sylvestris. Different media and combinations of auxins and cytokinins were used for regeneration and the outcome is discussed in chapter 4. The transformation of A. sylvestris with human cytochrome P450 3A4 was achieved and discussed in chapter 5. The findings and the interpretation of the results of our study are highlighted in the summary.

ChapterMetabolic engineering strategies for the optimization of medicinal and aromatic

plants: realities and expectations

Oktavia Hendrawati Herman J. Woerdenbag

Jacques Hille Oliver Kayser

Journal of Medicinal and Spice Plants 2010; 15(3): 111-126

1

12

Meta

bolic

eng

ineerin

g stra

tegies

1AbStrACt

In recent years, strategies and techniques for the production of natural compounds (plant derived fine chemicals) and/ or the breeding of medicinal and aromatic plants has expanded. Efficient production of high value natural products with medicinal and cosmetic purpose (e.g. essential oils, paclitaxel, artemisinin, and vincristine) is the main target. Metabolic engineering and pathway optimization with the aim to reduce costs and increase productivity are the main focus of academia and industry. Until now, only a limited number of plant cell cultures and isolated enzymes gave sufficiently high production for commercial purposes. Therefore research strategies have been shifted to metabolic engineering. Engineering of microorganisms has proven to be a valuable tool and the concept has been transferred to plant science opening new promising perspectives. Engineering has been conducted for crop plants, but the application of this technique to medicinal plants has not yet been explored well so far. Nowadays the cloning and expression of multiple genes and genomic integration are of high interest. This allows the reconstitution of biosynthetic pathways in heterologous organisms, being either plants or microorganisms. Combining basic science and applied engineering in this research area has been named combinatorial biosynthesis and later as synthetic biology. Synthetic biology comprises a large number of subareas, including enzymology, protein assembly and interactions, metabolomics, gene regulation, signal transduction and computational biology and is considered as an important future approach for biotechnological plant optimization. Synthetic biology has exciting perspectives for the exploitation of medicinal and aromatic plants, in order to increase the level of desired natural products, to gain insight in metabolic pathways even for new similar chemicals, to improve nutritional and health promoting effects of foods (nutraceuticals), and to reduce the amount of undesired side products with potential toxic or allergic activities.

Keywords: synthetic biology, metabolic engineering, plant breeding, plant genetics, medicinal and aromatic plants

oktavia Hendrawati1, Herman J. Woerdenbag2, Jacques Hille3, oliver Kayser1,4

1Department of Pharmaceutical Biology, University of Groningen, Antonius Deusinglaan 1, 9713 AV, Groningen, The Netherlands; 2Department of Pharmaceutical Technology and Biopharmacy, University of Groningen Antonius Deusinglaan 1, 9713 AV, Groningen, The Netherlands; 3Department of Molecular Biology of Plants, University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands; 4Technical University Dortmund, Technical Biochemistry, Emil-Figge-Strasse 66, 44227 Dortmund, Germany

Journal of Medicinal and Spice Plants 2010; 15(3): 111-126

13

Meta

bolic

eng

ineerin

g stra

tegies

11. introduCtion

Plants are a rich source of bioactive compounds. Compounds of plant origin are used as drugs and precursors of semisynthetic drugs, and may provide valuable leads for novel drug design. Furthermore, plant extracts are still being used to prevent, and to treat a number of diseases although the mechanism of action is frequently unknown. Finally, there is a global demand for “greener” manufacturing processes that are economically attractive to be available in a timely manner.1

Worldwide more than 50,000 plant species are used for medicinal purposes.2 The World Health Organization (WHO) estimated that more than 80% of the population in the world in less developed countries depends primarily on herbal medicine for basic healthcare needs.3 The current herbal drug market has reached a level of US $62 billion, which is forecast to grow to US $5 trillion in 2050.4 The world market for herbal medicines shows an annual growth of 5-15%.5 In the UK, more than 25% of the population uses herbal medicines on a regular basis.3

In the past 30 years, more than 25% of the new drug entities approved was based on a molecule of plant origin and about one third of the approximately 980 new pharmaceuticals originated from or were inspired by natural products.6, 7 About 50% of the top-selling chemicals are derived from the knowledge on plant secondary metabolism.7 About 40% of the pharmaceuticals in the US and Europe use plants as raw source material.8

Besides plants and plant extracts, pure compounds derived from plants play an important role in contemporary pharmacy and medicine. Typical plant compounds (table 1 and figure 1), commonly used drugs, are terpenoids, alkaloids, polyketides, phenylpropanoids, and flavonoids. For examples, morphine and codeine from Papaver somniferum L., artemisinin from Artemisia annua L., paclitaxel from Taxus brevifolia Nutt, genistein from Glycine max L. (Merr.), scopolamine from Dubosia species, camptothecin from Camptotheca acuminata Decne, and podophyllotoxin from Podophyllum species.

Despite the use of and demand for plant-derived compounds, their availability is a major bottleneck in supplying the pharmaceutical needs. Most of these compounds are secondary metabolites, which are present only in low amounts from natural sources. Most medicinal plants are not cultivated, but are collected from the wild and some of them are slowly growing. Because of intensive collection from the wild the current extinction rate of medicinally used plants are estimated to be 100 to 1000 times higher than for other plants. As many as 15,000 out of 50,000 to 70,000 medicinal plant species are now threatened with extinction.9 Currently between 4,000 and 10,000 medicinal plants are on the endangered species list and this number is expected to increase in the future.10

There are a number of limitations to obtain plant-derived compounds. They may be restricted to one species or genus and might be formed only during a particular stage of growth or development or under specific seasonal, stress or nutrient availability conditions.11, 12 Chemists have also been challenged to synthesize plant-derived

14

Meta

bolic

eng

ineerin

g stra

tegies

1

Tabl

e 1.

Ove

rvie

w o

f the

pro

duct

ion

of th

e pl

ant-

deri

ved

and

med

icin

ally

rele

vant

com

poun

ds w

ith th

eir m

etab

olic

eng

inee

ring

stra

tegi

es

Com

poun

dA

ctiv

ity /

func

tion

Plan

ts so

urce

Dem

and

(ton

s/ y

ear)

Cur

rent

pro

duct

ion

Pric

e

$US/

kgR

easo

n fo

r com

bina

tori

al

bios

ynth

esis

Ref

.

Dih

ydro

arte

mis

inic

ac

id (T

erpe

noid

)A

nti-

mal

aria

lAr

tem

isia

annu

a (0

.01

– 0.

86%

in

aeri

al p

arts

)

120

E. co

li1,

000

Avai

labi

lity

in n

atur

e is

lim

ited

Che

mic

al sy

nthe

sis –

eco

nom

ical

ly

not f

easi

ble

13

Pacl

itaxe

l(T

erpe

noid

)A

ntitu

mor

Taxu

s spe

cies

0.3

Plan

t cel

l cul

ture

ES

CA

gen

etic

s (U

SA);

Phyt

on

Cat

alyt

ic (U

SA/

Ger

man

y); N

ippo

n O

il (J

apan

); Sa

mya

ng G

enex

(K

orea

)

28,0

00Lo

w p

rodu

ctio

n in

nat

ure

Slow

ly g

row

ing

tree

The

bark

is n

on re

new

able

–

harv

estin

g th

e ba

rk re

sults

in th

e de

ath

of th

e tr

eeC

hem

ical

synt

hesi

s – e

cono

mic

ally

no

t fea

sibl

e

14-1

6

Vani

llin

(Phe

nylp

ropa

noid

)Fl

avor

Vani

lla p

lani

folia

> 10

,000

chem

ical

synt

hesi

s 12

(s

ynth

etic

)30

-120

(v

anill

a po

ds)

Gre

en m

anuf

actu

ring

17

Gen

istei

n(F

lavo

noid

)Fi

ber

Gos

sypi

um h

irsut

um36

,000

Isol

atio

n fr

om

plan

tsU

nkno

wn?

Cot

tons

eed

free

from

gos

sypo

l for

pr

otei

n so

urce

of h

uman

die

t18

Podo

phyl

loto

xin

(Lig

nan)

Ant

itum

orPo

doph

yllu

m

spec

ies (

4.3%

d.w

.)Pl

ant c

ell c

ultu

re

(Nip

pon

oil –

Ja

pan)

1,49

0En

dang

ered

spec

ies

Che

mic

al sy

nthe

sis –

eco

nom

ical

ly

not f

easi

ble

14, 1

9

15

Meta

bolic

eng

ineerin

g stra

tegies

1

Tabl

e 1.

Ove

rvie

w o

f the

pro

duct

ion

of th

e pl

ant-

deri

ved

and

med

icin

ally

rele

vant

com

poun

ds w

ith th

eir m

etab

olic

eng

inee

ring

stra

tegi

es

Com

poun

dA

ctiv

ity /

func

tion

Plan

ts so

urce

Dem

and

(ton

s/ y

ear)

Cur

rent

pro

duct

ion

Pric

e

$US/

kgR

easo

n fo

r com

bina

tori

al

bios

ynth

esis

Ref

.

Scop

olam

ine

(Alk

aloi

d)

Ant

icho

-lin

ergi

cD

uboi

sia s

peci

es

(1.2

– 2

.4%

in

leav

es)

Tran

sgen

ic h

airy

ro

ot c

ultu

re o

f H

yosc

yam

us

nige

r (H

nH6H

&

NtP

MT)

– 4

11.2

m

g/L

Dub

oisia

spp.

Su

mito

mo

Che

mic

al

Indu

stri

es (J

apan

)

Unk

now

n?M

ultip

le c

hira

l cen

tres

– c

hem

ical

sy

nthe

sis –

eco

nom

ical

ly n

ot

feas

ible

14, 2

0, 2

1

Mor

phin

e(A

lkal

oid)

Ana

lges

icPa

pave

r (20

% d

.w.

in la

tex;

1.2

3-2.

45

% d

.w. w

hole

pla

nt

extr

acts

)

27,8

Isol

atio

n fr

om

plan

tsU

nkno

wn?

Mul

tiple

chi

ral c

entr

es –

che

mic

al

synt

hesi

s – e

cono

mic

ally

not

fe

asib

le

16, 2

2

Vin

crist

in(A

lkal

oid)

Ant

itum

orCa

thar

anth

us ro

seus

(0

.000

3 %

d.w

. in

who

le p

lant

)

0.3

Isol

atio

n of

pla

nts

1,40

0M

ultip

le c

hira

l cen

tres

– c

hem

ical

sy

nthe

sis –

eco

nom

ical

ly n

ot

feas

ible

Hig

her y

ield

23

16

Meta

bolic

eng

ineerin

g stra

tegies

1compounds via organic chemistry. This is often hampered by the chemical complexity, specific stereochemistry and the economic feasibility.

Metabolic engineering may offer prospects to overcome the lack of availability of such compounds, through the advancement of the molecular biology techniques, including cloning, recombinant DNA and knowledge of the plant biosynthetic pathways. In this review, we discuss the major strategies in plant metabolic engineering and their principle approaches and its prospects and limitations for the production of drugs and fine chemicals. Case studies are used as illustrations.

Figure 1. Chemical structures of important plant derived compounds.

17

Meta

bolic

eng

ineerin

g stra

tegies

12. tHe pLAnt AS A SourCe for nAturAL produCtS

2.1. plant Cell CulturesIn principle, whole plants, plant organs and even single cells can be used for the production of natural products. Plant cell culturing was initiated in the 1930s24 and would offer alternatives to improve the production of the secondary metabolites, since natural harvest is sometimes bulky and not feasible from an economic point of view. The main advantages of plant cell culturing are easy up-scaling, simple purification schemes due to product secretion, environmental friendliness, and amenability to strict control in regards to meet FDA manufacturing standards.25 Plant cell cultures are also not subject to changes in environmental conditions, thus the production of the desired compounds could take place at any location and season.26

In 1959, plant cells were first cultured in a 10 L glass or steel bioreactor27 and later, in 1977, a first larger scale of a 20 L stirred tank bioreactor of N. tabacum cells was reported.28 Today, undifferentiated plant cell suspension cultures can easily be scaled up for commercial production purposes, but the productivity is often hampered by the fact that the compounds of interest are not produced in the undifferentiated cells. Currently 14 plant cell cultures have been commercialized for secondary metabolites production for pharmaceutical, food and cosmetic purposes.14, 29 Examples are scopolamine from Dubosia spp (Sumitomo Chemical Industries, Japan), ginsenosides from Panax ginseng (Nitti Denko, Japan), and paclitaxel from Taxus spp. (Phyton Biotech, USA and Samyang Genex, Korea).14

One success story of a plant cell culture produced drug is paclitaxel (Taxol® Bristol Myers Squibb). While the plant itself, Taxus brefolia, only produces paclitaxel at approximately 0.01% of the dry weight of the bark,30 the plant cell suspension culture has been shown to produce stably in the range of 140-295 mg/L, reaching 295 mg/L at a maximum under two-stage culture with the elicitation of methyl jasmonate and high density conditions.15

Main constraints of using plant cell cultures for the production of secondary metabolites include slow growth of plant cells in comparison to microorganisms, no accumulation of desired metabolites in undifferentiated cultures, compartmentalization of the production of secondary metabolites, low and variable yields, and the decrease of metabolite accumulation as the cell line ages.25, 31-33 Differentiated cells produce the same product as the plant itself, but in large-scale production aiming at an economically attractive way the yield remains a bottleneck, especially for slowly growing plants. A variety of approaches such as the growth of differentiated cells (roots and shoot cultures) and the induction of pathways by elicitors had limited success so far.12 The plant production of secondary metabolites is controlled in a tissue-specific manner, thus the dedifferentiation results in loss of production capacity12 and undifferentiated cell cultures which are genetically unstable, often lose, partially or totally, their ability to produce secondary products.34, 35

For example, artimisinin, a potent antimalarial drug, was not found in cell suspension cultures of Artemisia annua, while considerable amounts were detected in

18

Meta

bolic

eng

ineerin

g stra

tegies

1shoot cultures.36, 37 Deoxypodophylltoxin, the main lignan in Anthriscus sylvestris was also detected in trace amounts in callus and cell suspension cultures.38

2.2. transgenic plantsIn 1907, Smith & Towsend reported the cause of crown gall disease of paris daisy (Chysanthemum frustescens) by Bacterium tumefaciens. Later on the bacterium was classified as Phytomonas tumefaciens and finally as Agrobacterium tumefaciens, a gram-negative soil dwelling bacterium.39 In the end of the 1970s, it was reported that the T-DNA of this microorganism was covalently integrated into the plant nuclear genome in tobacco teratoma cell lines.40, 41 This led to many studies up to date. Since 1994, transgenic technology is being used and commercialized to produce new crop products with herbicide tolerance, insect resistance, virus resistance, and improved post-harvest quality.42 Transgenic approaches can be applied to target a rate-limiting step through manipulation of the expression of individual structural genes.5

There are two transformation approaches commonly used to produce recombinant pharmaceuticals in plants. First to subject plants to Agrobacterium-mediated transformation, particle bombardment, electroporation, and second, to infect plants with recombinant viruses that express transgenes during their replication in the host.43-46

Genetic transformation of medicinal plants is usually carried out using Agrobacterium rhizogenes to obtain hairy root cultures, or using Agrobacterium tumefaciens to produce transformed cells that can be maintained in cell cultures or can be regenerated as whole plants.2

In principle, the wounded plant tissues caused by insect or mechanical damage, produce phenolic compounds, which attract Agrobacterium by chemotaxis to infect the plant cell on the wounded site and allow the transfer of T-DNA from Agrobacterium into the plant nuclear chromosome. The T-DNA contains genes that encode enzymes directing the plant cells to produce peculiar amino acids called opines, and express genes to direct the plant cells to produce plant hormones such as cytokinin and auxin. Opines are used as primary sources of carbon and nitrogen by the cohabiting bacteria, and cytokinins and auxin promote cell division and tumor formation, providing a steadily increasing supply of nutrients for the bacteria.47

The infection from A. rhizogenes in the wounded site will cause a number of small roots protrude as fine hairs grow and proliferate rapidly, causing hairy root.48 T-DNA carries the rol and aux genes. The rol genes are responsible for the phenotype of hairy roots and the aux genes are involved in root induction by directing auxin synthesis.49

The major drawbacks of this approach are the unstable gene expression, instability of cell lines that often lose their capacity to produce target molecules over time, and high cost of bioreactors.2 For example, the alkaloid accumulation in transgenic Catharanthus roseus cell cultures resumed quickly to the level of nontransgenic ones.50

The capacity to regenerate whole plants from single cells without changing the genetic features of the cells and the gene transfer mechanism via Agrobacterium tumefaciens facilitate efforts to engineer secondary metabolic pathway.2, 51 The constraint is the

19

Meta

bolic

eng

ineerin

g stra

tegies

1subsequent regeneration of transgenic plants, which remains problematic and time consuming. Unwanted somaclonal variation may be introduced through the tissue culture regeneration system in some cases.52

Plants have been and are used as a host to produce genuine and recombinant proteins and enzymes of industrial and pharmacological value.53 More than 200 novel antibody-based potential products are in clinical trials worldwide, and the market demand will constrain the capabilities of existing production systems.54 One would expect that the biopharmaceuticals from transgenic plants are safer and less expensive than those from animal-based sources, which have the potential for contamination with human pathogens.43

Enhanced productivity of valuable secondary plant metabolites can also be achieved via hairy root cultures.26 Hairy root culture can be obtained from transformed root cultures using Agrobacterium rhizogenes, a gram-negative soil dwelling bacteria. The term “hairy root” was introduced in 190055 and the first transformation of higher plants using A. rhizogenes was achieved in 1973.56

Hairy root cultures are genetically stable, capable of unlimited growth without additional hormones and have an increased capacity for secondary metabolites formation and accumulation.25 Genetically transformed root cultures have been shown to produce levels of secondary metabolites comparable to that of an intact plant. It has further been shown that hairy root cultures can accumulate secondary metabolites that normally occur only in aerial part of the plant. An example is artemisinin in Artemisia annua.57 The transformation of Artemisia annua using A. rhizogenes carrying the cDNA encoding FDS (farnesyl diphosphate synthase) under the 35S CaMV promoter yielded 4-fold higher artemisinin accumulation compared to untreated control plants.58 The transformation of Atropa belladonna with H6H (hyoscyamine 6b-hydroxylase) from Hyoscyamus niger under the control of the 35S CaMV promoter in a binary plasmid via A. rhizogenes mediated transformation resulted in an accumulation of scopolamine up to 5-fold higher compare to untreated control plants.59 Other examples of hairy root cultures producing secondary metabolites can be found in Srivasta and Srivasta.26

Another advantage is that transformed roots are able to regenerate whole viable plants, maintain their genetic regeneration, and in addition produce secondary metabolites that are not present in the parent plant.60 Furthermore, they show fast auxin-independent growth and are suitable for adaptation to bioreactor system.26 In addition to production of secondary metabolites, hairy roots are also used to produce human therapeutic proteins, vaccines and diagnostic monoclonal antibodies. For example, hairy root cultures of potato carrying pBSHER containing the hepatitis B surface antigen (HBsAg)’s gene expressed higher level of HBsAg compared to control cultures.59

Despite discussed potentials, challenges during large-scale cultivation like unusual rheological properties of hairy root cultures have to be addressed. Non-optimal fermentation made it necessary to investigate novel approaches to apply hairy root cultures to fermentor26 and process design.25

20

Meta

bolic

eng

ineerin

g stra

tegies

1In conclusion, the use of hairy roots as factories for the production of novel plant-

based bioactive compounds, vaccines, antibodies and other therapeutic protein offers good prospects for feasibility of commercial production.

3. optimizing bioCHemiCAL pAtHWAyS

3.1. Strategies and goals of metabolic engineeringMetabolic engineering is generally defined as the redirection of one or more enzymatic reactions to produce new compounds in an organism, to improve the production of existing compounds, or to mediate the degradation of compounds.61 Metabolic engineering of plants offers interesting perspectives to improve the productivity of the plant as a cell factory. This approach may create new opportunities in agriculture, environmental applications, production of chemicals, and medicines.12, 52 The main goal of metabolic engineering in general is to produce the desired natural products in a sustainable and economically attractive way.62 Several more specific goals of metabolic engineering are listed in table 2.

The production level of a compound of interest which is present in trace amounts can be enhanced by increasing the flux of precursors, by blocking a competitive (parallel) pathway using the same precursor or intermediate compound, by introducing new routes of metabolism, by overcoming rate limiting steps, by reducing flux through enhancing competing pathways, by over-expressing regulatory genes or transcription factors that induce the pathway, and by inhibiting or limiting catabolism of the molecule or increasing the number of specialized cells producing the compound.2, 63

Some of the scientific challenges comprise a better understanding of the partly known secondary metabolite biosynthetic pathways on a genetic level, the generation of heterologous organisms with desirable biosynthetic characteristics, and optimised tools for pathway manipulation like vectors, synthetic genes and regulating elements.63 Moreover, biosynthetic pathways are often species-specific. Features like cell compartmentalization, tissue differentiation, and multi-enzyme complexes, will make the outcomes unpredictable.

3.2. metabolic pathways of interestA comprehensive understanding of different metabolic pathways and their genetic control is essential for the application of a genomics approach to the improvement of medicinal plants.5 In this review we confine to the metabolic pathway of medicinal compounds among other pathways of interest (figure 2).

Most secondary metabolites are derived from the shikimate, terpenoid and polyketide pathways. The shikimate pathway is the major source of phenylpropanoids and aromatic compounds,65, 66 such as flavonoids, coumarins, isoquinoline and indole alkaloids, lignans, lignins, and anthocyanins.

The terpenoid pathway leads to more than one-third of all known secondary metabolites, including mono-, sesqui-, di-, tri- and tetraterpenes. It is also the source

21

Meta

bolic

eng

ineerin

g stra

tegies

1

of the C5-building block (isoprene) in many skeletons from other biosynthetic origin, such as anthraquinones, naphtoquinones, cannabinoids, furanocoumarins, and terpenoid indole alkaloids.67

The polyketide pathway is a rich source of bioactive molecules such as anthranoids. It is attractive as a model for metabolic engineering studies because the complex structure results from simple C2 units combined in different ways and the modular construction of enzymatic catalysts allows control of enzyme structure.63

The complexity of a metabolic pathway and some strategies to engineer the pathway are illustrated in figure 3. It is supposed that a basic skeleton S (substrate) is present in which three functional groups can be introduced. If highly specific enzymes would catalyze all steps, 12 enzymes (E1 until E12) could be involved in the formation of three different products with one functional group, three different products with two functional groups, and one final product with all three functional groups. If the specificity of the substrate is broad, it is likely that three different enzymes will be adequate. Heterologous genes (E13) can also be introduced into plant metabolic pathway which could catalyze all three functions of the substrate (S1) into product “S1,2,3.”

3.3. Synthetic biology Synthetic biology is a rapidly growing multidisciplinary field among biologists, chemists, physicists, engineers and mathematicians.69 It is defined as the design and construction or engineering driven building of new or artificial biological components or increasingly complex biological entities, such as enzymes, genetic circuits, and cells, or the redesign of existing biological systems for novel applications.70, 71 The goals are

Table 2. Important goals of metabolic engineering in general.

Some important goals of metabolic engineering

Physiological understanding

Novel compounds To yield a novel compound in a plant and other precursors by introducing the appropriate heterologous genes.2

To give a new trait (colour, taste, smell) to food, flowers or ornamental plants.12

To increase To improve production of a desired compound or enzyme in a cell culture and also in the plant itself To achieve production in a related plant species or even in microorganism To improve agronomic traits, such as resistance of a plant to various stresses, pests, diseases and to increase the seed yield of a crop plant through the expression of certain metabolites.12, 64

To decrease To decrease levels of noxious or antinutritional factors in food and feed crops.2

Regulatory understandingTo improve our understanding of pathways regulation and flux when some of the intermediate pathways increase in abundance beyond their usual concentration range.64

22

Meta

bolic

eng

ineerin

g stra

tegies

1

Figure 2. Simplified biosynthetic pathways of primary and secondary metabolism in plants (adapted from Kumar et al.5).

13

of medicinal plants.5 In this review we confine to the metabolic pathway of

medicinalcompoundsamongotherpathwaysofinterest(figure2).

Figure 2.Simplifiedbiosyntheticpathwaysofprimaryandsecondarymetabolismin

plants(adaptedfrom Kumaret al.5).

Most secondary metabolites are derived from the shikimate, terpenoid and

polyketide pathways. The shikimate pathway is the major source of

phenylpropanoids and aromatic compounds,65, 66 such as flavonoids, coumarins,

isoquinolineandindolealkaloids,lignans,lignins,andanthocyanins.

The terpenoid pathway leads to more than one-third of all known secondary

metabolites, including mono-, sesqui-, di-, tri- and tetraterpenes. It is also the

source of the C5-building block (isoprene) in many skeletons from other

biosynthetic origin, such as anthraquinones, naphtoquinones, cannabinoids,

to build complex systems into specific hosts,71 to engineer synthetic organisms,71, 72 to improve understanding of biological systems, and to produce bio-orthogonal systems with new functions.69

The distinguishing element that differentiates synthetic biology from traditional metabolic engineering is the focus on the design and technological construction of core components (part of enzymes, genetic circuits, metabolic pathways, etc), which can be modelled, understood and tuned to meet specific criteria. The assembly of these components into integrated systems which enables a systematic forward-engineering of (parts of) biological systems for improved and novel applications is a second key issue of synthetic biology.70, 71

Synthetic biology is categorized into two broad classes. One uses unnatural molecules to reproduce emergent behaviours from natural biology, with the goal of creating artificial life. The other seeks interchangeable parts from natural biology to assemble into systems that function unnaturally.73

The knowledge about these tools and methods may enable synthetic biologists to design, fabricate, integrate, test and construct artificial biological systems that came from the insights discovered by experimental biologists and their holistic perspectives.72

23

Meta

bolic

eng

ineerin

g stra

tegies

1

Although synthetic biology offers promising applications for novel compounds and novel approaches, the success so far is rather limited since it is quite a young science. It is further hindered by the fact that the production processes of the most effective biological components (promoters, gene, plasmids, etc) have been patented. Royalty payments will increase the costs, which make it economically no longer attractive.70 Another drawback is that living systems are highly complex. Currently biologists lack information about how integration of living systems works.72 The success of synthetic biology depends on its capacity to surpass traditional engineering. It should blend the best features of natural systems with artificial designs that are extensible, comprehensible, user-friendly, and implement stated specifications to fulfill user goals.72

4. metAboLiC engineering StrAtegieS And teCHniqueS in mediCinAL pLAnt bioteCHnoLogy

The major metabolic engineering strategies and techniques applied in medicinal plant biotechnology are discussed in detail in this paragraph (see also figure 1). They include up or down regulating of pathways, redirecting common precursors, targeting metabolites

14

furanocoumarins,andterpenoidindolealkaloids.67

Figure 3. Schematic biosynthetic network, S: basic skeleton to which functional

group 1, 2 and 3 are added. E1-13: enzymes that catalyze biosynthetic steps

(adaptedfromVerpoorteet al.68)

The polyketide pathway is a rich source of bioactive molecules such as

anthranoids. It isattractive asamodel formetabolicengineeringstudiesbecause

thecomplexstructureresultsfromsimpleC2unitscombinedindifferentwaysand

the modular construction of enzymatic catalysts allows control of enzyme

structure.63

The complexity of a metabolic pathway and some strategies to engineer the

pathwayareillustratedinfigure3.ItissupposedthatabasicskeletonS(substrate)

is present in which three functional groups can be introduced. If highly specific

Figure 3. Schematic biosynthetic network, S: basic skeleton to which functional group 1, 2 and 3 are added. E1-13: enzymes that catalyze biosynthetic steps (adapted from Verpoorte et al.68).

24

Meta

bolic

eng

ineerin

g stra

tegies

1to specific cellular compartments, and creating of storage of overproducing secondary metabolites. Examples of techniques used and their application are given in table 3.

Table 3. Techniques used in metabolic engineering and their applications

Techniques Genes/Enzymes

Plant species Target compounds/goals

Ref.

Single transgene (biotransformation)

codeinone reductase (COR)

Papaver somniferum

Increase of morphine 82

CYP80B3 Papaver somniferum

Increase of morphine alkaloid

83

Taxadiene synthase Solanum lycopersicum

Production of taxadiene 84

Polycistronic vectors (artificial chromosomes)

Cholera toxin b-subunit (CTB gene)

Nicotiana tabacum

Production of cholera toxin b-subunit

85, 86

Transcription factor ORCA Catharanthus roseus

Increase of terpenoid indole alkaloid

79

Sense/antisense suppression

antisense CYP80B3 Papaver somniferum

Decrease of benzylisoquinoline alkaloids up to 84% of total alkaloid

83

Virus Inducing Gene Silencing (VIGS)

Phytoene desaturase (PapsPDS)

Papaver somniferum

Reduction of transcript level of endogenous PapsPDS and photobleach phenotype; and assessing gene function

87

T6ODM (thebaine 6-O-demethylase) and CODM (codeine O-demethylase)

Papaver somniferum

Increase of thebaine as a precursor of important pharmaceuticals (oxycodone, naltrexone, naloxone, buprenorphine)

88

RNAi berberine bridge enzyme (BBE)

Eschscholzia californica

Accumulation of (s)-reticuline

89

salutaridinol 7-O-acetyltransferase (SalAT)

Papaver somniferum

Accumulation of salutaridinol

90

d-cadiene synthase Gossypium hirsutum

Reduction of gossypol 18

4.1. upregulating of pathways (overexpression)Transcription factors (in multienzyme pathways) are regulatory proteins that can be used to regulate multiple steps or even to modulate an entire pathway in order to produce a significant yield of a desired product through sequence-specific DNA binding and protein-

25

Meta

bolic

eng

ineerin

g stra

tegies

1protein interactions.64, 74 They can act as activators or repressors of gene expression, which mediate respectively an increase or a decrease in the accumulation of messenger RNA.75 They are also able to regulate steps for which the enzymes are unknown.74

Using this approach, it is often necessary to increase precursor availability and to understand the coordination of multiple branches or sections of the metabolic pathway. The use of transcription factors requires integrated information from genomics, transcriptomics, proteomics, and metabolomics.25

Several transcription factors have now been identified. A relevant example is transcription factor MYB12, a flavonol-specific regulator of the phenylpropanoid biosynthesis in developing seedlings. Total flavanol content of the seed was increased when MYB12 was expressed in developing Arabidopsis thaliana seedlings. The expression of the genes encoding the four flavonoid biosynthetic enzymes was upregulated, increasing the flux through the flavanone pathway.76 Three transcription factors: ORCA1, ORCA2, and ORCA3 (octadecanoid-responsive Catharanthus AP2-domain) have been identified in the medicinal plant Catharanthus roseus and are involved in biosynthesis of terpenoid indole alkaloids. They belong to the AP2/ERF transcription factor family. The overexpression of ORCA3 in C. roseus cultured cells increased the expression of the terpenoid indole alkaloid biosynthesis genes TDC (tryptophan decarboxylase), STR, CPR (cytochrome P450 reductase), D4H (desacetoxyvindoline 4-hydroxylase)77 and SLS (secologanin synthase).78 Moreover, ORCA3 also regulated two genes encoding enzymes (ASa : a subunit of anthranilate synthase and DXS: D-1-deoxyxylulose 5-phosphate synthase) in the primary metabolism leading to terpenoid indole alkaloid precursor formation.79

Next to well-studied pathways, such as the phenylpropanoid biosynthesis, and the well-characterized MYB transcription factor family, finding a transcription factor that acts on specific pathway genes is very challenging.64 Transcription factors are difficult to identify in non-model species.74 An alternative is to design synthetic transcription factors, which target one or more genes of choice.80 As example, the expression of the ZFP-TF (zinc finger protein transcription factor) increased the activity of native g-methyltransferase (VTE4) and the a-tocopherol content in Arabidopsis thaliana seeds.81 Transcription factors, natural or synthetic, are used only if the pathway is endogenous to the plant.64

4.2. redirecting common precursors Many branching points are found in a biosynthetic pathway where enzymes compete for a common precursor. Increasing and redirecting the precursor pool towards the biosynthesis of the target compounds can theoretically increase their production. This can be achieved by blocking the competitive pathway or by inducing overexpression of genes in the precursor pathway.67

GGPP (geranylgeranyl diphosphate) is normally utilized for making carotenoids in tomato. By overexpressing the gene that encodes taxadiene synthase in the tomato that lacks the ability to utilize GGPP for carotenoid production allowed GGPP to be re-routed for the production of taxadiene. The production was 660-20,000 times higher than in Arabidopsis thaliana.84 By overexpressing genes in precursor pathways in both

26

Meta

bolic

eng

ineerin

g stra

tegies

1peppermint91 and lavender92 an increase in the monoterpene fraction of the essential oils was found.

However, due to tight regulation of metabolite accumulation, this approach may also have a limited impact on target products. The increase of intermediate precursor resulted in a limited accumulation of alkaloid target product in C. roseus.93 In this case, the direct overexpression of related genes in the alkaloid pathway was shown to be more effective in increasing the alkaloid accumulation in C. roseus.94 The effect seems to be temporary. It might be due to a result of the same factors, which induce variability in nontransgenic plants.25

4.3. targeting metabolites to specific plant cell compartmentsTargeting gene expression to a specific cellular compartment or organelle that contains the precursors could increase the level of the target compounds. Plants are able to express the transgene with organelle targeting signals from the nuclear DNA and the resulting recombinant proteins will be targeted to the appropriate organelles. Specific amino acid sequences required for targeting of proteins to particular organelles and for retention of proteins in organelles have been identified.52 Thus targeting the enzymes to the compartment of substrate seems feasible. However, the products formed may cause toxicity problems in another compartment than usual.67

Using this approach, overexpression of a target gene either in plastids or the cytosol allows transport of a sufficient pool of common precursors into the right direction and led to an more than 1,000-fold increase in concentration of the sesquiterpenes patchouli alcohol and amorpha-4,10-diene and a 10-30-fold increase of the monoterpene limonene in transgenic tobacco plants compared to untreated control plants.95

4.4. Creation of storage of overproduced secondary metabolitesA plant may have the capacity to produce secondary metabolites but sometimes it lacks a proper subcellular compartment to store them.96 Modifications to metabolic storage of products or secondary metabolic pathways have been generally more successful than manipulations of primary and intermediary metabolism.97, 98 The genes controlling the formation of subcellular compartments have been isolated and characterized in plants.99

The expression of the Or gene encoding a DnaJ cysteine-rich domain-containing protein led to the formation of large membranous chromoplasts in cauliflower curd cells.99 The expression of the same gene in transgenic potato under the control of a potato granule-bound starch synthase promoter increased the total carotenoid up to 6-fold compared to the original, non transgenic plants.99

4.5. downregulating of pathways (silencing)The production of a certain compound can be reduced by decreasing the flux towards that product by reducing the level of enzyme in the pathway, increasing catabolism and increasing flux into competitive pathways.2, 68

A particular step in the pathway that leads to undesirable compounds can be blocked by suppressing genes that upregulate the pathway or by increasing their catabolism.2

27

Meta

bolic

eng

ineerin

g stra

tegies

1Antisense, co-supression and RNA interference (RNAi) methods are used to block, to reduce or to eliminate levels of undesirable compounds. This so called silencing can be targeted to specific plant tissues and organs with minimal interference of the normal plant life cycle, by using tissue or organ-specific RNAi vectors. Mutants with the RNAi effect have been shown to be stable for at least 20 generations.100

5. CHALLengeS in pLAntS metAboLiC engineering

5.1. unexplored regulation of secondary metabolism The lack of complete understanding of the regulation of secondary metabolism, especially in the complex alkaloid biosynthesis, hinders the determination of an effective metabolic engineering strategy to achieve a specific production phenotype. The complexities comprise pathway compartmentalization, the existence of sometimes multiple biosynthetic pathways and the regulatory control mechanisms.101 Up to date, only four biosynthetic routes of alkaloid subclasses have been partially characterized in particular, the benzylisoquinoline, monoterpenoid indole, purine and tropane alkaloids.101 This could be ascribed to limited genome/cDNA sequence information of medicinal plants.5

5.2. pathways are often species-specific A number of genes encoding enzymes, that control key steps of secondary metabolic pathways, has been cloned from a number of a medicinal plant species using classical and modern genomics approaches.5 However this represents a small fraction of a total of about 1,000 plant genes known to function in secondary metabolism.102 The progress in isolating genes involved in secondary metabolism is limited due to species specificity, the difficulty in producing large numbers of mutants, their intermediate precursor availability, their analysis, and to the instability of secondary metabolites caused by environmental factors.103 The major bottleneck for secondary metabolism will remain per definition species specific. Only early parts of pathways are common to most plants, for example in the flavonoid and terpenoid biosynthetic pathways, thus homology between genes can be used for strategies to clone genes from other plants.103 The genes encoding enzymes involved in the more specific ‘decoration’ of the basic skeletons only can be studied at the level of the very producing plant.104

5.3. Cell compartmentalization and tissue differentiation Plant cells have a complicated intercellular organization with metabolite flow between compartments highly regulated and orchestrated depending on the biosynthetic needs of the plants.105 They have numerous organelles of which some are not found in mammalian or yeast cells.52 The highly compartmentalized nature of enzymes, substrate precursors, and metabolic intermediates also contributes to the complexity of secondary metabolites production, which is regulated at a different level.2

28

Meta

bolic

eng

ineerin

g stra

tegies

1Plants also have numerous specialized and differentiated organs in which

physiological processes and gene expression may differ substantially. Next to organelles, the compartmentalization of secondary metabolite pathways also occurs at the subcellular level.106 Furthermore, temporal and developmental processes can profoundly influence whether and when a transgene is active. Thus the issues of compartmentalization complicate the targeting gene strategy. Moreover if the engineered plants are going to be propagated as crops, environmental effects may add to the level of variability and unpredictability which is not encountered in a fermentor based system.52

There is increasing evidence that intra- and intercellular translocation of enzymes are one of the key elements in secondary metabolite production. Localization of enzymes to diverse cellular compartments showed the importance of protein targeting in the assembly of the alkaloid pathway.2 Alkaloids are generally stored in specific types of compartments due to their cytotoxicity and probable role in plant defence responses. The subcellular compartmentalization of alkaloid pathway enzymes is extremely diverse and complex as the cell type-specific localization of the gene transcripts, enzymes, and metabolites.107

Other examples are phenylpropanoid derivatives. Their biosynthesis occurs in the cytoplasm, but the precursors are derived from metabolism in other organelles, including the chloroplasts and mitochondria.105

5.4. unpredicted or unexpected outcomeThe use of metabolic engineering approaches in medicinal plant species to improve the yield of pharmaceutical products has been and is a challenge. There are several limitations such as gene silencing, unpredictable results due to complex network genes, and no increase in concentration of desirable metabolites up to the level of commercialization.5 Techniques used to introduce new genes into plants also do not allow a prediction about the site of integration and the level of gene expression, even when a strong promoter is used.67 It is often hard to accurately guess what are the actual biological roles of certain enzymes solely based on bioinformatics, due to promiscuity towards substrates and the relatively ease to change substrate or product specificity by introducing minor changes in sequence of the enzymes.108

Single-enzyme perturbations of alkaloid pathways resulted in unexpected metabolic consequences suggesting the existence of key rate-limiting steps, potential multi-enzyme complexes, or unsuspected compartmentalization.107 Over-expression of COR1 (codeinone reductase), the final enzyme in morphine biosynthesis, increased the morphine and codeine contents in transgenic poppy.82 However, thebaine, an upstream metabolite in the 23 branch pathways, was also unexpectedly significantly increased. The knock down of COR1 with RNAi technology would expect to suppress 23 upstream biosynthetic steps and the accumulation of codeinone and morphinone, the immediate precursor of COR. The amount of morphinan alkaloids decreased, while the biosynthesis of (S)-reticuline, an early upstream metabolite in the pathway, was increased instead of the target compounds codeinone and morphinone.109

29

Meta

bolic

eng

ineerin

g stra

tegies

1The complexity and redundancy of many biosynthetic pathways coupled to incomplete knowledge of their regulation could lead to an unpredictable outcome from a targeted metabolic engineering strategy.25

Selected case studies using different approaches and strategies in metabolic engineering are discussed in the next paragraph.

6. metAboLiC engineering AppLiCAtionS in mediCinAL pLAnt bioteCHnoLogy

6.1. Case study: podophyllotoxin production in Anthriscus sylvestris Anthriscus sylvestris (L.) Hoffm. (Apiaceae) is a common wild plant in northwest Europe that accumulates considerable amounts of lignans. Deoxypodophyllotoxin, an aryltetralin-lignan as the main attractive constituent which is much more abundant in the plant kingdom than podophyllotoxin, can be used as a precursor for the production of podophyllotoxin. Podophyllotoxin is used as a precursor for the semi-synthesis anticancer drugs: Etoposide phosphate and Teniposide.110 To date, podophyllotoxin is obtained by isolation from Podophyllum species. In the future, the availability of podophyllotoxin from this source is likely to become a major bottleneck. Podophyllum species are on the endangered species list, proving that the increasing demand of podophyllotoxin is a serious threat for the plant.111 An alternative source of podophyllotoxin may be obtained by (biotechnological) hydroxylation of deoxypodophyllotoxin at the C7 position (see figure 1). Human cytochrome P450 3A4 in E.coli DH5a selectively hydroxylates deoxypodophyllotoxin at the C7 position yielding podophyllotoxin.112 Studies to transform A. sylvestris with this cytochrome are in progress.

6.2. Case study: Scopolamine biosynthesis in nicotiana tabacumScopolamine and hyoscyamine are tropane alkaloids. They form an important class of plant derived anticholinergic compounds occurring in several genera of the Solanaceae like Hysoscyamus, Atropa, Duboisia, Scopolia, and Datura.16, 107 Scopolamine has a higher commercial market value than hyoscyamine but has a lower yield from plants than hyoscyamine.16 The world demand of scopolamine is estimated to be about 10 times higher than hyoscyamine and its racemic form atropine. The main source of raw material worldwide are Duboisia leaves containing 2-4% of total alkaloids, with more than 60% scopolamine and 30% hyoscyamine.20 Up to 6% of scopolamine has been achieved by conventional cultivation of selected varieties in Australia, Equador, and Brazil, producing 1 t/ha of plant material for industrial alkaloid extraction.20

The heterologous expression of PMT (putrescine N-methyltranferase) from Nicotiana tabacum in Scopolia parviflora yielded a 8-fold increase in scopolamine and a 4.2-fold increase in hyoscyamine production.113 A similar effect has been achieved in Hyoscyamus muticus and Datura metel.114 Surprisingly, this PMT expression has no effect on alkaloid production when it is expressed in other tropane alkaloid producing hairy root cultures

30

Meta

bolic

eng

ineerin

g stra

tegies

1of Hyoscyamine niger, Atropa belladonna, and Duboisiana hybrid.21, 115, 116 It was suggested that PMT expression in the roots was insufficient to boost the tropane alkaloid synthesis of these plants. Overexpression seems to be species-related due to a different, specific post-translational regulation of the endogenous enzyme in respect to the foreign one.116

The constitutive expression of H6H (hyoscyamine 6b-hydroxylase) from Hyoscyamus niger in Atropa belladonna, a plant that normally accumulates hyoscyamine, converts hyoscyamine into scopolamine up to 1.2% dry weight.117 The alkaloid composition of aerial parts of mature plants changed from over 90% hyoscyamine in controls and wild type plants, to almost exclusively scopolamine in transgenics.117 In transgenic hairy root of Atropa belladonna, up to a 5-fold scopolamine increase was observed.118 In Hyoscyamus muticus hairy root expressing the H6H gene, up to a 100-fold increase of scopolamine was found, while the hyoscyamine content remained unaltered.119

Transgenic tobacco plants expressing constitutively H6H were fed with hyoscyamine and 6b-hydroxyhyoscyamine. These precursors were converted into scopolamine in the leaves of the plants.117 The hairy root cultures of Nicotina tabacum, which do not produce hyoscyamine, were used to express the H6H gene from Hyoscymanus niger. The cultures successfully converted added hyoscyamine into scopolamine. They showed efficient uptake of hyoscyamine (average of 95%) from the culture medium and a higher rate of bioconversion of hyoscyamine into scopolamine (10-45%). Up to 85% of the total scopolamine was released into the culture medium.120 This was in contrast to the normal metabolic behaviour of tropane alkaloid-producing hairy roots in which the scopolamine remained accumulated in the root tissues.62 Feeding exogenous hyoscyamine to cell suspension cultures, which were obtained from the hairy root, showed considerable capacity to convert hysocyamine into scopolamine and the product was secreted into the culture medium.121 The scaling up of the transgenic cells grown in a 5 L turbine stirred tank reactor in a batch mode yielded scopolamine up to a 1.6-fold higher than the small-scale cultures. Almost 18% of the hyoscyamine added to the medium was transformed into scopolamine, which showed 65% of increase with respect to the same alkaloid obtained by bioconversion in shakes flasks.20

The constitutive co-expression of genes encoding the rate-limiting upstream enzyme PMT (putrescine N-methyltransferase) and the downstream enzyme H6H of scopolamine biosynthesis yielded only a modest increase in alkaloid accumulation when it was expressed alone, but exhibited a synergistic effect on alkaloid levels when expressed together.21 It resulted in the highest production of scopolamine in hairy root culture reported of 411 mg/L. It is a 10-fold increase over control cultures and a 2-3-fold increase over cultures, which expressed only H6H.21

6.3. Case study: genistein production in transgenic arabidopsis, tobacco,lettuce, corn, petunia, and tomato

Genistein is a common precursor of the isoflavonoids biosynthesis especially occurring in the subfamily Papilionoideae of the Fabaceae.122 Isoflanoids are interesting because of their pharmaceutical and nutraceutical activity that attract considerable interests in the prospect of introducing them into vegetables, grains, and fruits for dietary disease

31

Meta

bolic

eng

ineerin

g stra

tegies

1prevention.123 Genistein production in nonlegume plants has been performed but yet with unsatisfactory yields. It might be due to the competitive use of naringenin between isoflavon syntase (IFS) and the endogenous flavonoid pathway.123-125

Soy products are the major dietary sources of isoflavonoids (genistein) for human. The IFS isolated from soybean has been introduced into Arabidopsis thaliana, corn (Zea mays) and tobacco (Nicotiana tabacum).125 There was no accumulation of free genistein in Arabidopsis, but genistein was glycosylated to glucose-rhamnose-genistein and rhamnose-genistein.123

The overexpression of soybean IFS in tobacco, petunia (Petunia hybrida Vilm) and lettuce (Lactuca sativa L.) resulted in genistein accumulation in transgenic plants.126 Another approach was the introduction of a heterogenous phenylalanine ammonia-lyase (PAL) and IFS into genetically manipulated plants. This increased the genistein content in tobacco petals (1.80-fold) and lettuce leaves (1.5-fold).126 The overexpression of IFS soybean in tomato (Solanum lycopersicum L) resulted in the presence of genistein 7-O-glucoside as the major isoflavone metabolite in the transgenic plants.127

6.4. Case study: expression of spearmint limonene synthase in lavender Essential oil quantity and quality can be regulated by metabolic engineering.91 In principle, it is possible to engineer the biosynthesis of monoterpenes in order to increase or to modify the essential oil profiles in the target plant.128 For example, the expression of a sense of the 1-deoxy-D-xylulose-5-phosphate (DXP) reductoisomerase cDNA, and with an antisense of menthofuran synthase cDNA under the control of CaMV 35S promoter, resulted in up to 50% more essential oil in Mentha × piperita L. without changing the composition of the monoterpenes as compared to the wild-type.91 Meanwhile the expression of DXP synthase in Lavandula latifolia increased the essential oil up to a 3.5-fold in leaves and up to a 7-fold in flowers as compared to control, without obvious deleterious effects on plant development and fitness.92

Spike lavender (Lavandula laftifolia Med.) is an aromatic shrub that is cultivated worldwide for oil production, which has limonene as a minor constituent (0.5-2%). Overexpression of the LS gene from spearmint (Mentha spicata) in spike lavender under the regulation of the CaMV35S constitutive promoter showed more than 450% increase of limonene content in developing leaves as compared to the control.128

The expression of lemon basil (Ocimum basilicum L. cv. Sweet Dani) geraniol synthase under the regulation of the tomato polygalacturonase promoter in tomato increased the monoterpenes (limonene) and sesquiterpenes content at the expense of reduced lycopene accumulation.129, 130

6.5. Case study: Artemisinin biosynthesis in Artemisia annuaIn the early 1980s, efforts started to establish Artemisia annua L. cultures producing artemisinin.131 A range of variable but always low artimisinin levels were found in callus, shoot and root cultures but no artemisinin in cell suspension cultures, suggesting that some degree of differentiation is required for the production.132 Transformation of Artemisia annua with Agrobacterium rhizogenes resulted in hairy root cultures that

32

Meta

bolic

eng

ineerin

g stra

tegies

1produced artemisinin and currently many efforts are directed toward optimizing production in hairy root cultures.133

Several key genes involved in the biosynthesis of artemisinin have been introduced in A. annua. Approaches in genetic engineering have been focused on the overexpression of cloned key enzymes involved in the biosynthesis, such as farnesyl diphosphate synthase (FPS)134 and amorpha-4,11-diene synthase (AMS).135

Genetic transformation and regeneration of A. annua has been established to introduce genes of interest via A. tumefaciens.136 A. annua expressing FPS from Gossypium arboreum accumulated higher level of artemisinin compared to A. annua expressing the FPS from A. annua. A. annua expressing FPS accumulated up to 10.08 mg/g d.w. artemisinin.136-138

Hairy root cultures of A. annua were established by transforming it via A. rhizogenes carrying the farnesyl diphosphate synthase (FDS) gene. The artemisinin content in the transgenic plants, which were regenerated from the hairy root cultures, was significantly higher than in the control plant.95

Despite all the genetic engineering attempts, the mean of production of artemisinin is still mainly from the plant itself. Recently, the FDA has approved Coartem® (Novartis) as the first artemisinin-based combination treatment (ACT) for malaria in the USA.139 Novartis has stimulated the cultivation of Artemisia annua of more than 1,000 hectares in Kenya, Tanzania and Uganda. In addition, it has also cultivated in China, in total it reaches up to 10,000 hectares.140

6.6. Case study: morphine and thebaine biosynthesis in Papaver somniferumPapaver somniferum remains to be the sole source of morphine. The commercial chemical synthesis of morphine, codeine and other benzylisoquinoline alkaloid is not economically feasible due to the complexity of the molecule and multiple chiral centres.16

Reticuline is an essential precursor leading to the biosynthesis of benzylisoquinoline alkaloids like codeine, berberine and morphine. One of the strategies to increase the flux into morphinan alkaloid is by blocking the BBE (berberine bridge enzymes). Blocking the BBE will increase the (S)-reticuline concentration. The expression of an antisense-BBE construct in transgenic opium poppy plants indeed showed increased flux into the morphinan and the tetrahydrobenzylisoquinoline branch pathway.141

Another strategy is to increase the precursor pool leading to the formation of (S)-reticuline, which is (S)-N-methylcoclaurine 3’-hydrolase. The overexpression of cytochrome P450 monooxygenase (S)-N-methylcoclaurine 3’-hydrolase (CYP80B3) resulted in an up to 450% increase of total morphinan alkaloids.83 The suppression of this gene by an antisense construct led to a reduced total alkaloid content in the transgenic poppy.83

The existence of multi-enzyme complexes has been proposed for flavonoid142-144 and polyamine metabolism.145 The occurrence of multi-enzyme complex as seems also to exist in the morphine biosynthesis,107 therefore the metabolic engineering strategies

33

Meta

bolic

eng

ineerin

g stra

tegies

1have to be developed carefully. However, not all enzymes of the morphinan branch are necessarily involved in such a macromolecular complex.146

A further approach is to increase salutaridinol, a precursor of thebaine, by overexpressing salutaridinol 7-O-acetyltransferase (SaIAT) and salutaridine reductase (SaIR). RNAi-silenced SalAT in opium poppy plants showed an accumulation of salutaridine instead of salutaridinol, which is normally not abundant in the plants.90 Salutaridine may be channelled to thebaine through an enzyme complex that includes salutaridine reductase (SaIR) and SaIAT. Recent results showed that there is an interaction between SaIR and SaIAT.146 Morphine, codeine, and thebaine levels were increased in both saIAT over-expressing and saIAT RNAi plants.90

The codeinone reductase (COR) converts codeinone into codeine. Theoretically, the morphine production can be increased by blocking this enzyme using RNAi technique. On the contrary, there was an accumulation of (S)-reticuline instead of morphine, codeine, oripavine and thebain.109 The reasons are unknown, but there were some speculations. It was suggested that there was a feedback mechanism preventing intermediates from the general benzylisoquinoline synthesis entering the morphine-specific branch.2 The impairment of a required metabolic channel composed of morphinan branch pathway enzymes resulted in accumulation of alkaloid intermediates produced by enzymes that were not part of the same complex.107 The COR could be part of a multi-enzyme complex, which cannot function if one of the enzymes is removed.16 It might be that the side effect of silencing COR was the suppression of 1,2-dehydroreticuline reductase.107 The potential homology between the two reductases could lead to cosilencing.107 The COR seems to be an important target for metabolic engineering. The overexpression of COR in Papaver somniferum yielded a 15 % increase of benzylisoquinoline alkaloids as compared to the high-yielding control genotypes and a 30% increased as compared to the non-transgenic control.82

Recent example is the engineering of thebaine in opium. Thebaine is the natural precursor used in the synthesis of several pharmaceuticals such as oxycodone, naltrexone, naloxone and buprenorphine. Virus inducing gene silencing (VIGS) of T6ODM (thebaine 6-O-demethylase) and CODM (codeine O-demethylase) in opium resulted an efficient blockage of the metabolism at thebaine and codeine, respectively, which increased the thebaine production in transgenic opium up to 95% as compared to control.88

7. CroSSing borderS – HeteroLogouS produCtion of pLAnt CompoundS in miCroorgAniSmS

7.1. Artemisinic acid One of the success stories of using the synthetic biology approach is related to artemisinic acid. It is a naturally occurring precursor of artemisinin, used as an antimalarial drug. Malaria causes nearly a million deaths each year, mostly of children below 5 years. The World Health Organization (WHO) estimated 247 million malaria cases among

34

Meta

bolic

eng

ineerin

g stra

tegies

13.3 billion people at risk in 2006.147 This leads to a demand to supply artimisinin in an economically attractive and environmental friendly way. The relatively low yield (0.01-0.6%) of artemisinin from Artemisia annua is unable to supply the world demand.136 The total chemical synthesis of artemisinin is difficult and costly.148 However, the semi-synthesis of artemisinin or any derivatives from microbial sourced artemisinic acid and its immediate precursor gives an alternative for availability and economic feasibility.149 Using the synthetic biology approach with the use of appropriate promoters and an expression vector resulted in the production of artemisinic acid up to 300 mg/L in the yeast Saccharomyces cerevisiae. 70, 150

7.2. StilbenesStilbenes are polyketides, produced by plants. Resveratrol, as a representative of stilbenes and known as constituent of red wine, which has possible interesting biological activities as anti-cancer agent,151 inhibitor of inflammation, tumor promotion, angiogenesis and metastasis, and regulation of cell cycle progression.152

The biosynthetic pathway and the enzymes have been characterized and metabolic engineering has been achieved in plants, microbes and animals.153 E. coli cells carrying PAL (phenylalanine ammonia-lyase), 4CL (4-coumarate:CoA ligase), STS (stilbene synthase), and ACC (acetyl CoA carboxylase) produced 40 mg/L resveratrol (1 h) from tyrosine The PAL is from the yeast Rhodotorula rubra, 4CL is from actinomycete S. coelicolor A3 (138), and STS is from Arachis hypogaea.154, 155 Resveratrol yield were > 100 mg/L in E.coli expressing 4CL and STS.156

Stilbenes are rapidly absorbed and metabolized when given orally. The modification of the resveratrol scaffold by hydroxylation and methylation enhanced its bioactivities. The recombinant E. coli carrying PAL, 4CL, STS, ACC and OsPMT (pinosylvin methyltransferase in rice) with the addition of tyrosine resulted in the production of 18 mg/L pinostilbene and 6 mg/L pterostilbene. Addition of phenylalanine resulted in production of pinosylvin monomethyl ether and pinosylvin dimethyl ether almost in the same yield of 27 mg/L.157

7.3. CurcuminoidsCurcumin, bisdemethoxycurcumin, and dicinnamoylmethane are called curcuminoids.158 Curcumin is the active ingredient of turmeric (Curcuma longa), which has a surprisingly wide range of beneficial claims, not clinically proven yet. Its use is related to traditional medicine as anti-inflammatory, antioxidant, anti-HIV, chemopreventive, and chemotherapeutic agent. These actions are partly supported by preclinical pharmacology.159, 160

Horinouchi et al., 2009 discovered a type III polyketides synthase (PKS) in Oryza sativa (rice) that can synthesize curcuminods via p-courmaroyl-CoA.157 This PKS, named CUS (curcuminoid synthase), as part of an artificial biosynthetic pathway for production of curcuminoids in E. coli.161 The E. coli expressing PAL, 4CL, CUS, and ACC with the additional supply of 1mM each of the phenylpropanoid acid (p-coumaric acid, cinnamic acid or ferulic acid) yielded about 100 mg/L of curcumin

35

Meta

bolic

eng

ineerin

g stra

tegies

1or dicinnamoylmethane, and bisdemethoxycurcumin, respectively.157 E. coli carrying 4CL, ACC and CUS with the addition of ferulic acid isolated from 1 g of rice bran pitch yielded 60 mg curcumin. Rice bran pitch is a dark and viscous oil, which is a waste from the production of rice edible oil from rice bran. Rice bran pitch (1 g) contains about 22 mg ferulic acid.158

7.4. flavonoids For the first time the complete flavonoid pathway from a plant has been successfully transferred into a microorganism.157 Genes from various organisms were assembled in E. coli on a single pET plasmid for the production of flavanones. They are PAL from the yeast Rhodotorula rubra, 4CL or ScCCL from the actinomycete S. coelicolor A3(2), CHS from the plant Glycyrrhiza echinata, and CHI from the plant Pueraria lobata.162,

163 The construction proved to be optimal using isopropyl b-D-thiogalactopyranoside (IPTG)-inducible T7 promoter and a synthetic ribosome-binding sequence in front of each of the four genes in a recA- host.163 The yield of pinocembrin from 3 mM phenylalanine exogenously added, and naringenin from 3 mM tyrosine were both 60 mg/L.154 Flavanone-3b-hydroxylase (F3H), flavonol synthase (FLS) and flavone synthase (FNS) were introduced into E. coli to modify flavanones into flavonols (kaempferol and galangin). The expression of the genes led to the production of kaempferol (15.1 mg/L) from 3 mM tyrosine and galangin (1.1 mg/L) from 3 mM phenylalanine.164 Cloning of a FNS gene from Petroselinum crispum into pACYC in the E.coli host led to production of flavones: apigenin (13 mg/L) from tyrosine and chrysin (9.4 mg/L) from phenylalanine.164

7.5. Vanillin An example of ‘white biotechnology’ is the production of vanillin as one of the most important aromatic flavour compounds used in foods, beverages, perfumes and pharmaceuticals. The production scale is more than 10,000 tons per year by chemical synthesis.17 The increasing demand of customers for natural flavours has shifted the interest of the flavour industry to produce vanillin from natural sources by biotransformation instead of organic synthesis.17 The aim of the biotransformation of vanillin is to avoid toxic and mutagenic solvents such as phenol and dimethyl sulfate and to avoid corrosive compounds such as hydrogen peroxide which are used for the organic synthesis.165

Many different possibilities have been investigated for the biotechnological production of vanillin using different kinds of bacteria and fungi and different precursors.17 The transformed E. coli BL21(DE3) cells carrying isoeugenol monooxygenase gene of Pseudomonas putida IE27 produced up to 28.3 g/L of vanillin from 230 mM isoeugenol, with a molar conversion yield of 81% at 20oC after 6 h.166 The growing knowledge regarding enzymes involved in biosynthetic pathways as well as the identification and characterization of the corresponding genes offers new opportunities for metabolic engineering and for the construction of genetically engineered production strains.17

36

Meta

bolic

eng

ineerin

g stra

tegies

18. ConCLuSion And future proSpeCtiVe

Plants definitely play an essential role in modern pharmacy and medicine. Efforts to obtain the desired natural compounds to be used as drugs in an efficient way, are going on and included various approaches.

Metabolic engineering has been applied to both plants and plant cell cultures. Plant cell cultures have been shown to be feasible for industrial production only limitedly, as shown for paclitaxel. Understanding secondary metabolism within cells and cell cultures is essential to use them as a means to supply natural products. The characteristics and metabolic capacities of plant cell/tissue and microbial systems are inherently different; therefore they can serve as complementary unit operations in order to solve the long-standing problem of robust secondary metabolites production.101

The lack of complete information about genomes of most medicinal plants is still an immense challenge for applying the appropriate metabolic engineering strategy. Up to date, only few plant genomes (e.g., Oryza sativa, Zea mays, and Arabidopsis thaliana), but none of the medicinal plant, have been fully sequenced. The challenges to unravel the unknown biosynthetic pathways, the encoding genes and the transcription factors are still there. However with the progressing sequencing techniques, it will likely be feasible to fully sequence medicinal plants in a shorter time. Yet the only main constraint will be the funding.

Conventional breeding of medicinal plants is another way to enhance the concentration of the desired compounds. Breeding and genetic engineering are essential to go hand in hand and are necessary to ensure the availability of the desired compounds.

Until now, it seems that there is limited success in engineering medicinal plants in which the product could be commercialized based on economic feasibility. However, genetic engineering strategies have been applied into crop plants such as rice, maize, soybean, and cotton with great significant success. The genetically modified crop plants with bt (Bacillus thuringiensis) toxin for pest resistance have been grown commercially in approximately 42 million hectares worldwide.167 In addition, Bt transgenic rice varieties are in field tests and are close to approval for commercialization.168

The advance of technology holds great promise for the future of plant metabolic engineering. Genomics approaches may lead to the identification of regulatory genes and proteomic approaches may explain why the expression levels of some biosynthetic genes do not correlate with the metabolites profile.107

Finding alternative ways to produce originally plant-derived compounds are still continuing. Microorganisms such as endophytes may serve as an alternative host for production of bioactive substances as reviewed by Staniek et al.169 The success of transferring the biosynthetic pathway from plants into microorganisms or other hosts for the production of artemisinic acid and flavonoids showed that it is feasible to engineer the entire pathway into microorganisms.

The latest promising approach is through synthetic biology for optimizing the biotechnological production of the plant-derived compounds. However, well-

37

Meta

bolic

eng

ineerin

g stra

tegies

1characterized biological components such as the knowledge of the biosynthetic pathways, the genes involved, the promoters, and the precursors are essential to build the system. An integrated approach of synthetic biology and metabolic engineering will be necessary in the near future. For successful engineering, to enhance and optimize the production of the desired metabolites, crossing borders of different disciplines will be needed.

referenCeS

1. Ranjan P. Engineering complex pheno-types in industrial strains. Biotechnol Prog 2008;24:38-47.

2. Gómez-Galera S, Pelacho A, Gené A, et al. The genetic manipulation of medicinal and aromatic plants. Plant Cell Rep 2007;26:1689-1715.

3. Vines G. Herbal harvests with a future: towards sustainable sources for medicinal plants. plantlife international, 2004.

4. Joshi K, Chavan P, Warude D, et al. Molecular markers in herbal drug tech-nology. Curr Sci 2004;87:159 - 165.

5. Kumar J, Gupta, P. Molecular approach-es for improvement of medicinal and aromatic plants. Plant Biotechnol Rep 2008; 2: 11966-11970.

6. Newman DJ, Cragg GM. Natural products as sources of new drugs over the last 25 years. J Nat Prod 2007;70:461-477.

7. Terryn N, Van Montagu M, Inze D, et al. Functional genomic approaches to study and engineer secondary metabolites in plant cell cultures. The Netherlands: Springer, 2006; 291-300.

8. Ganapathy S. Bioreactor technology: a novel industrial tool for high-tech pro-duction of bioactive molecules and bi-opharmaceuticals from plant roots. Bio-technol J 2006;1:1419-1427.

9. Brouwer C, Bruce W, Maddock S, et al. Suppression of transgene silencing by matrix attachment regions in maize: A dual role for the maize 5? ADH1 matrix attachment region. Plant Cell 2002;14:2251.

10. Edward R. No remedy in insight for herbal ransack. New Scientist 2004;181:10-11

11. Kala CP. Medicinal plants of the high altitude cold desert in India: diversity, distribution and traditional uses. Int J Biodiv Sci Manag 2006;2:1-14.

12. Verpoorte R, Contin A, Memelink J. Bio-technology for the production of plant secondary metabolites. Phytochem Rev 2002;1:13-25.

13. Abdin MZ, Israr M, Rehman RU, et al. Artemisinin, a novel antimalarial drug: biochemical and molecular approaches for enhanced production. Planta Med 2003;69:289-299.

14. Eibl R, Eibl D. Bioreactors for plant cell and tissue cultures. In: Oksman-Calden-tey KM, Barz W, ed. Plant biotechnology and transgenic plants. New York: Marcel Dekker, 2002:162-200.

15. Tabata H. Production of paclitaxel and the related taxanes by cell suspension cultures of Taxus Species. Curr Drug Targets 2006;7:453-461.

16. McCoy E, O’Connor SE. Natural products from plant cell cultures. In: Petersen F, Amstutz R, ed. Natural compounds as drugs, volume 1. Basel: Birkhäuser, 2008:329-370.

17. Priefert H, Rabenhorst J, Steinbüchel A. Biotechnological production of vanillin. Appl Microbiol Biotechnol 2001;56:296-314.

18. Sunilkumar G, Campbell LM, Puckhaber L, et al. Engineering cottonseed for use in human nutrition by tissue-specific reduction of toxic gossypol. Proc Nat Acad Sci 2006;103:18054-18059.

19. Arroo RRJ, Alfermann AW, Medarde M, et al. Plant cell factories as a source for anti-cancer lignans. Phytochem Rev 2002;1:27-35.

20. Palazon J, Navarro-Ocana A, Hernandez-Vazquez L, et al. Application of metabolic engineering to the production of scopo-lamine. Molecules 2008;13:1722-42.

21. Zhang L, Ding R, Chai Y, et al. Engineer-ing tropane biosynthetic pathway in Hy-oscyamus niger hairy root cultures. Proc Nat Acad Sci 2004;101:6786-6791.

38

Meta

bolic

eng

ineerin

g stra

tegies

122. United Nations. Technical reports

estimated world requirements for 2005. Statistics for 2003. New York: United Nations Publication INCB Narcotic Drugs, 2004.

23. Kuboyama T, Yokoshima S, Tokuyama H, et al. Stereocontrolled total synthesis of (+)-vincristine. Proc Nat Acad Sci 2004;101:11966-11970.

24. White PR. Potentially unlimited growth of excised tomato root tips in a liquid medium. Plant Physiol 1934;9:585-600.

25. Kolewe ME, Gaurav V, Roberts SC. Pharmaceutically active natural product synthesis and supply via plant cell culture technology. Mol Pharmaceut 2008;5:243-256.

26. Srivastava S, Srivastava AK. Hairy root culture for mass-production of high-value secondary metabolites. Crit Rev Biotechnol 2007;27:29-43.

27. Tulecke W, Nickell LG. Production of large amounts of plant tissue by submerged culture. Science 1959;130:863-864.

28. Noguchi M, Matsumoto T, Hirata Y, et al. Improvement of growth rates of plant cell cultures. In: Barz W, Reinhard E, Zenk MH, ed. Plant tissue culture and its biotechnological application. Berlin: Springer Verlag, 1977: 85-94.

29. Frense D. Taxanes: perspectives for biotechnological production. Appl Microbiol Biotechnol 2007;73:1233-1240.

30. Witherup KM, Look SA, Stasko MW, et al. Taxus spp. needles contain amounts of taxol comparable to the bark of Taxus brevifolia: analysis and isolation. J Nat Prod 1990;53:1249-1255.

31. Deus-Neumann B, Zenk MH. Instability of indole alkaloid production in Catha-ranthus roseus cell suspension cultures. Planta Med 1984;50:427-431.

32. De Jesus-Gonzalez L, Weathers PJ. Tetraploid Artemisia annua hairy roots produce more artemisinin than diploids. Plant Cell Rep 2003;21:809-813.

33. Qu J, Zhang W, Yu X, et al. Instabil-ity of anthocyanin accumulation in Vitis vinifera L. var. Gamay Fréaux suspen-sion cultures. Biotechnol Bioprocess Eng 2005;10:155-161.

34. Rokem JS, Goldberg I. Secondary metab-olites from plant cell suspension cultures: methods for yield improvement. Adv Biotechnol Proc 1985; 4: 241-274.

35. Charlwood BV, Charlwood KA. Terpenoid production in plant cell culture. In: Harborne J B, Tomás-Barberán FA, ed. Ecological chemistry and biochemistry of plant terpenoids. Oxford: Clarendon Press, 1991: 95-132.

36. Liu C, Zhao Y, Wang Y. Artemisinin: current state and perspectives for bio-technological production of an antima-larial drug. Appl Microbiol Biotechnol 2006;72:11-20.

37. Woerdenbag HJ, Lüers JFJ, Uden W, et al. Production of the new antimalarial drug artemisinin in shoot cultures of Artemisia annua L. Plant Cell Tiss Org Cult 1993;32:247-257.

38. Koulman A, Kubbinga ME, Batterman S, et al. A phytochemical study of lignans in whole plants and cell suspension cultures of Anthriscus sylvestris. Planta Med 2003;69:733-738.

39. Conn HJ. Validity of the genus Alcali-genes. J Bacteriol 1942;44:353-60.

40. Zambryski P, Holsters M, Kruger K, et al. Tumor DNA structure in plant cells transformed by A. tumefaciens. Science 1980;209:1385-1391.

41. Yadav NS, Postle K, Saiki RK, et al. T-DNA of a crown gall teratoma is cova-lently joined to host plant DNA. Nature 1980;287:458-461.

42. Sun SSM. Transgenics for new plant products, applications to tropical crops. In: Moore PH, Ming R, ed. Genomics of tropical crop plants. Berlin: Springer, 2008:63-81.

43. Giddings G, Allison G, Brooks D, et al. Transgenic plants as factories for biopharmaceuticals. Nat Biotechnol 2000;18:1151-1155.

44. Dalsgaard K, Uttenthal A, Jones TD, et al. Plant-derived vaccine protects target animals against a viral disease. Nat Bio-technol 1997;15:248-252.

45. Beachy RN, Fitchen JH, Hein MB. Use of plant viruses for delivery of vaccine epitopes. In: Collins GB, Shepherd RJ, ed. New York: Annals of the New York Academy of Sciences, 1996:43-50.

46. Mushegian AR, Shepherd RJ. Genetic elements of plant viruses as tools for genetic engineering. Microbiol Rev 1995;59:548-578.

47. Hughes AM. Plant molecular genetics. Dorset: Addison Wesley Longmann Limited, 1999.

39

Meta

bolic

eng

ineerin

g stra

tegies

148. Balandrin MF, Klocke JA, Wurtele ES, et

al. Natural plant chemicals: sources of in-dustrial and medicinal materials. Science 1985;228:1154-1160.

49. Guillon S, Trémouillaux-Guiller J, Pati PK, et al. Harnessing the potential of hairy roots: dawn of a new era. Trends Biotechnol 2006;24:403-409.

50. Whitmer S, Canel C, van der Heijden R, et al. Long-term instability of alkaloid production by stably transformed cell lines of Catharanthus roseus. Plant Cell Tiss Org Cult 2003;74:73-80.

51. Birch RG. Plant transformation: problems and strategies for practical ap-plication. Ann Rev Plant Physiol Plant Mol Biol 1997;48:297-326.

52. Lessard PA, Kulaveerasingam H, York GM, et al. Manipulating gene expression for the metabolic engineering of plants. Metab Eng 2002;4:67-79.

53. Teli NP, Timko MP. Recent developments in the use of transgenic plants for the production of human therapeutics and biopharmaceuticals. Plant Cell Tiss Org Cult 2004;79:125-145.

54. Stoger E, Sack M, Fischer R, et al. Plan-tibodies: applications, advantages and bottlenecks. Curr Opin Biotechnol 2002;13:161-166.

55. Steward FC, Rolfs FM, Hall FH. A fruit disease survey of western New York in 1900. New York Agric Exp Sta Bact 1900;191:291-331.

56. Ackermann C. Pflanzen aus Agrobacte-rium rhizogenes-tumoren an Nicotiana tabacum. Plant Sci Lett 1977;8:23-30.

57. Wallaart TE, Pras N, Quax WJ. Isolation and identification of dihydroartemisinic acid hydroperoxide from Artemisia annua: a novel biosynthetic precursor of artemisinin. J Nat Prod 1999;62:1160-1162.

58. Chen D, Liu C, Ye H, et al. Ri-mediated transformation of Artemisia annua with a recombinant farnesyl diphos-phate synthase gene for artemisinin production. Plant Cell Tiss Org Cult 1999;57:157-162.

59. Sunil Kumar GB, Ganapathi TR, Srinivas L, et al. Expression of hepatitis B surface antigen in potato hairy roots. Plant Sci 2006;170:918-925.

60. Veerasham C. Medicinal plant biotech-nology. New Delhi: CBS Pub, 2004.

61. DellaPenna D. Plant metabolic engineer-ing. Plant Physiol 2001;125:160-163.

62. Cusido RM, Palazón J, Piñol MT, et al. Datura metel: In vitro produc-tion of tropane alkaloids. Planta Med 1999;65:144-148.

63. Koffas M, Roberge C, Lee K, et al. Metabolic engineering. Ann Rev Biomed Eng 1999;1:535-557.

64. Kinney AJ. Metabolic engineering in plants for human health and nutrition. Curr Opin Biotechnol 2006;17:130-138.

65. Bentley R, Haslam E. The shikimate pathway - a metabolic tree with many branches. Crit Rev Biochem Mol Biol 1990;25:307-384.

66. Herrmann KM, Weaver LM. The shikimate pathway. Ann Rev Plant Physiol Plant Mol Biol 1999;50:473-503.

67. Verpoorte R. Plant secondary metabo-lites. In: Verpoorte R, Alfermann AW, ed. Metabolic engineering of plant secondary metabolism. Dodrecht: Kluwer academic publisher, 2000: 1-29.

68. Verpoorte R. General strategies. In: Verpoorte R, Alfermann AW, ed. Metabolic engineering of plant secondary metabolism. Dodrecht: Kluwer academic publisher, 2000: 31-50.

69. Channon K, Bromley EHC, Woolfson DN. Synthetic biology through biomo-lecular design and engineering. Curr Opin Struct Biol 2008;18:491-498.

70. Keasling JD. Synthetic biology for synthetic chemistry. ACS Chem Biol 2008;3:64-76.

71. Heinemann M, Panke S. Synthetic biology - putting engineering into biology. Bioinformatics 2006;22:2790-2799.

72. Andrianantoandro E, Basu S, Karig DK, et al. Synthetic biology: new engineering rules for an emerging discipline. Mol Syst Biol 2006;2: 1-14.

73. Benner SA, Sismour AM. Synthetic biology. Nat Rev Genet 2005;6:533-543.

74. Broun P. Transcription factors as tools for metabolic engineering in plants. Curr Opin Plant Biol 2004;7:202-209.

75. Latchman DS. Eucaryotic transcription factors. 4th ed. San Diego: Academic Press, 2003.

76. Mehrtens F, Kranz H, Bednarek P, et al. The Arabidopsis transcription factor MYB12 is a flavonol-specific regulator of phenylpropanoid biosynthesis. Plant Physiol 2005;138:1083-1096.

40

Meta

bolic

eng

ineerin

g stra

tegies

177. van der Fits L, Memelink J. ORCA3, a

jasmonate-responsive transcriptional regulator of plant primary and secondary metabolism. Science 2000;289:295-297.

78. Gantet P, Memelink J. Transcription factors: tools to engineer the produc-tion of pharmacologically active plant metabolites. Trends Pharmacol Sci 2002;23:563-569.

79. Memelink J, Gantet P. Transcription factors involved in terpenoid indole alkaloid biosynthesis in Catharanthus roseus. Phytochem Rev 2007;6:353-362.

80. Moore M, Ullman C. Recent develop-ments in the engineering of zinc finger proteins. Brief Funct Genomic Proteomic 2003;1:342-355.

81. Van Eenennaam AL, Li G, Venkatramesh M, et al. Elevation of seed [alpha]-to-copherol levels using plant-based tran-scription factors targeted to an endog-enous locus. Metab Eng 2004;6:101-108.

82. Larkin PJ, Miller JAC, Allen RS, et al. In-creasing morphinan alkaloid production by over-expressing codeinone reductase in transgenic Papaver somniferum. Plant Biotechnol J 2007;5:26-37.

83. Frick S, Kramell R, Kutchan TM. Metabolic engineering with a morphine biosynthetic P450 in opium poppy surpasses breeding. Metab Eng 2007;9:169-176.

84. Kovacs K, Zhang L, Linforth R, et al. Redirection of carotenoid metabolism for the efficient production of taxadiene [taxa-4(5),11(12)-diene] in transgenic tomato fruit. Trans Res 2007;16:121-126.

85. Daniell H. Pharmaceutical proteins, human therapeutics, human serum albumin, insulin, native cholera toxin B subunit in transgenic plastids. WO 01/72959, 2001.

86. Daniell H, Lee SB, Panchal T, et al. Ex-pression of the native cholera toxin B subunit gene and assembly as functional oligomers in transgenic tobacco chloro-plasts. J Mol Biol 2001;311:1001-1009.

87. Lena CH, Sinéad D, Gemma M, et al. Vi-rus-induced gene silencing is an effective tool for assaying gene function in the basal eudicot species Papaver somnifer-um (opium poppy). Plant J 2005;44:334-341.

88. Hagel JM, Facchini PJ. Dioxygenases catalyze the O-demethylation steps of morphine biosynthesis in opium poppy. Nat Chem Biol 2010;6:273-275.

89. Fujii N, Inui T, Iwasa K, et al. Knockdown of berberine bridge enzyme by RNAi ac-cumulates (S)-reticuline and activates a silent pathway in cultured California poppy cells. Transgen Res 2007;16:363-375.

90. Allen RS, Miller JAC, Chitty JA, et al. Metabolic engineering of morphinan alkaloids by over-expression and RNAi suppression of salutaridinol 7-O-acetyl-transferase in opium poppy. Plant Bio-technol J 2008;6:22-30.

91. Mahmoud SS, Croteau RB. Metabolic engineering of essential oil yield and composition in mint by altering expres-sion of deoxyxylulose phosphate reduc-toisomerase and menthofuran synthase. Proc Nat Acad Sci 2001;98:8915-8920.

92. Munoz-Bertomeu J, Arrillaga I, Ros R, et al. Up-regulation of 1-deoxy-D-xylulose-5-phosphate synthase enhances produc-tion of essential oils in transgenic spike lavender. Plant Physiol 2006;142:890-900.

93. Erik HH, Seung-Beom H, Susan IG, et al. Expression of a feedback-resistant anthranilate synthase in Catharanthus roseus hairy roots provides evidence for tight regulation of terpenoid indole alkaloid levels. Biotechnol Bioeng 2004;86:718-727.

94. Canel C, Lopes-Cardoso MI, Whitmer S, et al. Effects of over-expression of stric-tosidine synthase and tryptophan decar-boxylase on alkaloid production by cell cultures of Catharanthus roseus. Planta 1998;205:414-419.

95. Wu S, Schalk M, Clark A, et al. Redirec-tion of cytosolic or plastidic isoprenoid precursors elevates terpene production in plants. Nat Biotechnol 2006;24:1441-1447.

96. Verpoorte R, van der Heijden R, ten Hoopen HJG, et al. Metabolic engi-neering of plant secondary metabolite pathways for the production of fine chemicals. Biotechnol Lett 1999;21:467-479.

97. Kinney AJ. Manipulating flux through plant metabolic pathways. Curr Opin Plant Biol 1998;1:173-178.

98. Stitt M, Sonnewald U. Regulation of me-tabolism in transgenic plants. Ann Rev Plant Physiol Plant Mol Biol 1995;46:341-368.

99. Li L, Van Eck J. Metabolic engineer-ing of carotenoid accumulation by

41

Meta

bolic

eng

ineerin

g stra

tegies

1creating a metabolic sink. Transgen Res 2007;16:581-585.

100. Mansoor S, Amin I, Hussain M, et al. En-gineering novel traits in plants through RNA interference. Trends Plant Sci 2006;11:559-565.

101. Leonard E, Runguphan W, O’Connor S, et al. Opportunities in metabolic engi-neering to facilitate scalable alkaloid pro-duction. Nat Chem Biol 2009;5:292-300.

102. Somerville C, Somerville S. Plant func-tional genomics. Science 1999;285:380-383.

103. Verpoorte R, van der Heijden R, Memelink J. Engineering the plant cell factory for secondary metabolite produc-tion. Transgen Res 2000;9:323-343.

104. Dixon RA. Plant natural products: the molecular genetic basis of biosyn-thetic diversity. Curr Opin Biotechnol 1999;10:192-197.

105. Wu S, Chappell J. Metabolic engineering of natural products in plants; tools of the trade and challenges for the future. Curr Opin Biotechnol 2008;19:145-152.

106. Pasquali G, Porto DD, Fett-Neto AG. Metabolic engineering of cell cultures versus whole plant complexity in pro-duction of bioactive monoterpene indole alkaloids: Recent progress related to old dilemma. J Biosci Bioeng 2006;101:287-296.

107. Ziegler JR, Facchini PJ. Alkaloid biosyn-thesis: metabolism and trafficking. Ann Rev Plant Biol 2008;59:735-769.

108. Lewinsohn E, Gijzen M. Phytochemical diversity: The sounds of silent metabo-lism. Plant Sci 2009;176:161-169.

109. Allen RS, Millgate AG, Chitty JA, et al. RNAi-mediated replacement of morphine with the nonnarcotic alkaloid reticuline in opium poppy. Nat Biotech-nol 2004;22:1559-1566.

110. Ayres DC, Loike JD. Lignans: chemical, biological and clinical properties. Cambridge: Cambridge University Press, 1990.

111. Nayar MP, Sastry APK. Red Data Book of Indian Plants. Calcutta, 1990.

112. Vasilev NP, Julsing MK, Koulman A, et al. Bioconversion of deoxypodophyllo-toxin into epipodophyllotoxin in E. coli using human cytochrome P450 3A4. J Biotechnol 2006;126:383-393.

113. Lee O, Kang Y, Jung H, et al. Enhanced production of tropane alkaloids in

Scopolia parviflora by introducing the PMT (putrescine N-methyltransferase) gene. In Vitro Cell Dev Biol - Plant 2005;41:167-172.

114. Moyano E, Jouhikainen K, Tammela P, et al. Effect of pmt gene overexpression on tropane alkaloid production in trans-formed root cultures of Datura metel and Hyoscyamus muticus. J Exp Bot 2003;54:203-211.

115. Rothe G, Hachiya A, Yamada Y, et al. Alkaloids in plants and root cultures of Atropa belladonna overexpressing pu-trescine N-methyltransferase. J Exp Bot 2003;54:2065-2070.

116. Moyano E, Fornalé S, Palazón J, et al. Alkaloid production in Duboisia hybrid hairy root cultures overexpressing the pmt gene. Phytochem 2002;59:697-702.

117. Yun D, Hashimoto T, Yamada Y. Metabolic engineering of medicinal plants: transgenic Atropa belladonna with an improved alkaloid composition. Proc Nat Acad Sci 1992;89:11799-11803.

118. Hashimoto T, Yun D, Yamada Y. Produc-tion of tropane alkaloids in genetically engineered root cultures. Phytochem 1993;32:713-718.

119. Jouhikainen K, Lindgren L, Jokelainen T, et al. Enhancement of scopolamine pro-duction in Hyoscyamus muticus L. hairy root cultures by genetic engineering. Planta 1999;208:545.

120. Hakkinen ST, Moyano E, Cusido RM, et al. Enhanced secretion of tropane alkaloids in Nicotiana tabacum hairy roots expressing heterologous hyo-scyamine-6b-hydroxylase. J Exp Bot 2005;56:2611-2618.

121. Moyano E, Palazón J, Bonfill M, et al. Biotransformation of hyoscyamine into scopolamine in transgenic tobacco cell cultures. J Plant Physiol 2007;164:521-524.

122. Dixon RA, Ferreira D. Genistein. Phytochem 2002;60:205-211.

123. Liu C, Blount JW, Steele CL, et al. Bottle-necks for metabolic engineering of iso-flavone glycoconjugates in Arabidopsis. Proc Nat Acad Sci 2002;99:14578-14583.

124. Winkel-Shirley B. Flavonoid biosynthe-sis. A colorful model for genetics, bio-chemistry, cell biology, and biotechnol-ogy. Plant Physiol 2001;126:485-493.

125. Yu O, Jung W, Shi J, et al. Production of the isoflavones genistein and daidzein in

42

Meta

bolic

eng

ineerin

g stra

tegies

1non-legume dicot and monocot tissues. Plant Physiol 2000;124:781-794.

126. Liu R, Hu Y, Li J, et al. Production of soybean isoflavone genistein in non-legume plants via genetically modified secondary metabolism pathway. Metab Eng 2007;9:1-7.

127. Shih C, Chen Y, Wang M, et al. Accumu-lation of isoflavone genistein in trans-genic tomato plants overexpressing a soybean isoflavone synthase gene. J Agric Food Chem 2008;56:5655-5661.

128. Muñoz-Bertomeu J, Ros R, Arrillaga I, et al. Expression of spearmint limonene synthase in transgenic spike lavender results in an altered monoterpene com-position in developing leaves. Metab Eng 2008;10:166-177.

129. Davidovich-Rikanati R, Sitrit Y, Tadmor Y, et al. Enrichment of tomato flavor by diversion of the early plastidial terpenoid pathway. Nat Biotechnol 2007;25:899-901.

130. Davidovich-Rikanati R, Lewinsohn E, Bar E, et al. Overexpression of the lemon basil a-zingiberene synthase gene increases both mono- and sesquit-erpene contents in tomato fruit. Plant J 2008;56:228-238.

131. He XZM, Li G, Liang Z. Callus induction and regeneration of plantlets from Artemisia annua and changes in qinghaosu contents. Acta Bot Sin 1983;25:87–90.

132. Krishna S, Uhlemann AC, Haynes RK. Artemisinins: mechanisms of action and potential for resistance. Drug Res Updates 2004;7:233-244.

133. Weathers PJ, De Jesus-Gonzalez L, Kim YJ, et al. Alteration of biomass and artemisinin production in Artemisia annua hairy roots by media steriliza-tion method and sugars. Plant Cell Rep 2004;23:414-418.

134. Matsushita Y, Kang W, Charlwood BV. Cloning and analysis of a cDNA encoding farnesyl diphosphate synthase from Artemisia annua. Gene 1996;172:207-209.

135. Mercke P, Bengtsson M, Bouwmeester HJ, et al. Molecular cloning, expression, and characterization of amorpha-4,11-diene synthase, a key enzyme of artem-isinin biosynthesis in Artemisia annua L. Arch Biochem Biophys 2000;381:173-180.

136. Han J, Wang H, Ye H, et al. High ef-ficiency of genetic transformation and regeneration of Artemisia annua L. via Agrobacterium tumefaciens-mediated procedure. Plant Sci 2005;168:73-80.

137. Chen D, Ye H, Li G. Expression of a chimeric farnesyl diphosphate synthase gene in Artemisia annua L. transgenic plants via Agrobacterium tumefaciens-mediated transformation. Plant Sci 2000;155:179-185.

138. Han J, Liu B, Ye H, et al. Effects of over-expression of the endogenous farnesyl diphosphate synthase on the artemisinin content in Artemisia annua L. J Integr Plant Biol 2006;48:482-487.

139. Novartis. Coartem® receives FDA approval becoming first artemisinin-based combination treatment (ACT) for malaria in the US. 2009: media release.

140. Novartis. Novartis partners with east African botanicals to expand cultivation and extraction of natural ingredient used in anti-malarial Coartem®. 2005: media release.

141. Frick S, Chitty JA, Kramell R, et al. Transformation of opium poppy (Papaver somniferum L.) with antisense berberine bridge enzyme gene (anti-bbe) via somatic embryogenesis results in an altered ratio of alkaloids in latex but not in roots. Transgen Res 2004;13:607-613.

142. Achnine L, Blancaflor EB, Rasmussen S, et al. Colocalization of L-phenylalanine ammonia-lyase and cinnamate 4-hy-droxylase for metabolic channeling in phenylpropanoid biosynthesis. Plant Cell 2004;16:3098-3109.

143. Burbulis IE, Winkel-Shirley B. Interac-tions among enzymes of the Arabidopsis flavonoid biosynthetic pathway. Proc Nat Acad Sci 1999;96:12929-12934.

144. He X, Dixon RA. Genetic manipulation of isoflavone 7-O-methyltransferase enhances biosynthesis of 4’-O-meth-ylated isoflavonoid phytoalexins and disease resistance in alfalfa. Plant Cell 2000;12:1689-1702.

145. Panicot M, Minguet EG, Ferrando A, et al. A polyamine metabolon involving aminopropyl transferase complexes in Arabidopsis. Plant Cell 2002;14:2539-2551.

146. Kempe K, Higashi Y, Frick S, et al. RNAi suppression of the morphine biosyn-thetic gene salAT and evidence of asso-ciation of pathway enzymes. Phytochem 2009;70:579-589.

43

Meta

bolic

eng

ineerin

g stra

tegies

1147. WHO. World Malaria Report 2008.

Geneva, 2008.148. Schmid G, Hofheinz W. Total synthesis

of qinghaosu. J Am Chem Soc 1983;105:624-625.

149. Haynes RKV. Cyclic peroxyacetal lactone, lactol and ether compound. US Patent 5,310,946. 1994.

150. Chang MCY, Eachus RA, Trieu W, et al. Engineering Escherichia coli for produc-tion of functionalized terpenoids using plant P450s. Nat Chem Biol 2007;3:274-277.

151. Oliver Y, Joseph MJ. Nature’s assembly line: biosynthesis of simple phenyl-propanoids and polyketides. Plant J 2008;54:750-762.

152. Kundu JK, Surh YJ. Molecular basis of chemoprevention by resveratrol: NF-kappaB and AP-1 as potential targets. Mut Res 2004;555:65-80.

153. Halls C, Yu O. Potential for metabolic engineering of resveratrol biosynthesis. Trends Biotechnol 2008;26:77-81.

154. Katsuyama Y, Funa N, Miyahisa I, et al. Synthesis of unnatural flavonoids and stilbenes by exploiting the plant biosyn-thetic pathway in Escherichia coli. Chem Biol 2007;14:613-621.

155. Yohei K, Nobutaka F, Sueharu H. Pre-cursor-directed biosynthesis of stilbene methyl ethers in Escherichia coli. Bio-technol J 2007;2:1286-1293.

156. Watts K, Lee P, Schmidt-Dannert C. Biosynthesis of plant-specific stilbene polyketides in metabolically engi-neered Escherichia coli. BMC Biotechnol 2006;6:22.

157. Horinouchi S. Combinatorial biosyn-thesis of plant medicinal polyketides by microorganisms. Curr Opin Chem Biol 2009;13:197-204.

158. Horinouchi S. Combinatorial biosynthesis of non-bacterial and unnatural flavonoids, stilbenoids and curcuminoids by microor-ganisms. J Antibiot 2008;61:709-28.

159. Hatcher H, Planalp R, Cho J, et al. Curcumin: From ancient medicine to current clinical trials. Cell Mol Life Sci 2008;65:1631-1652.

160. Itokawa H, Shi Q, Akiyama T, et al. Recent advances in the investigation of curcuminoids. Chin Med 2008;3:11.

161. Katsuyama Y, Matsuzawa M, Funa N, et al. Production of curcuminoids by Escherichia coli carrying an artifi-cial biosynthesis pathway. Microbiol 2008;154:2620-2628.

162. Hwang EI, Kaneko M, Ohnishi Y, et al. Production of plant-specific flavanones by Escherichia coli containing an artifi-cial gene cluster. Appl Environ Microbiol 2003;69:2699 - 2706.

163. Miyahisa I, Kaneko M, Funa N, et al. Efficient production of (2 S )-flavanones by Escherichia coli containing an artificial biosynthetic gene cluster. Appl Microbiol Biotechnol 2005;68:498-504.

164. Miyahisa I, Funa N, Ohnishi Y, et al. Combinatorial biosynthesis of flavones and flavonols in Escherichia coli. Appl Microbiol Biotechnol 2006;71:53-58.

165. Schafer B. Naturstoffe der chemischen industrie. Munchen: Elsevier, 2007.

166. Yamada M, Okada Y, Yoshida T, et al. Vanillin production using Escherichia coli cells over-expressing isoeugenol monooxygenase of Pseudomonas putida. Biotechnol Lett 2008;30:665-670.

167. Fitt GP. Have Bt crops led to changes in insecticide use patterns and impacted IPM? Integration of insect-resistant genetically modified crops within IPM programs, Progress Biol Contr 2008:303-328.

168. Liu W. Effects of Bt transgenic crops on soil ecosystems: a review of a ten-year research in China. Front Agric China 2009;3:190-198.

169. Staniek A, Woerdenbag HJ, Kayser O. Endophytes: exploiting biodiversity for the improvement of natural product-based drug discovery. J Plant Interact 2008;3:75 - 93.

ChapterSeasonal variations in the

deoxypodophyllotoxin content and yield of Anthriscus sylvestris L. (Hoffm.) grown in the

field and under controlled conditions

Oktavia Hendrawati Herman J. Woerdenbag

Jacques Hille Wim J. Quax Oliver Kayser

Journal of Agricultural and Food Chemistry 2011; 59 (15): 8132-9139

2

46

seaso

na

l Varia

tion

s

2

oktavia Hendrawati1, Herman J. Woerdenbag2, Jacques Hille3, Wim J. quax1, oliver Kayser1,4

1Department of Pharmaceutical Biology, University of Groningen, Groningen, The Netherlands; 2Department of Pharmaceutical Technology and Biopharmacy, University of Groningen, Groningen, the Netherlands; 3Department of Molecular Biology of Plants, University of Groningen, Haren, The Netherlands; 4Technical Biochemistry, Technical University Dortmund, Dortmund, Germany

Journal of Agricultural and Food Chemistry 2011; 59 (15): 8132-9139

AbStrACt

Deoxypodophyllotoxin (DPT) is the main lignan in Anthriscus sylvestris. We carried out two sets of experiments with sixteen plants and seeds, collected from a wide geographical range. The DPT content in roots was significantly lower (p < 0.05) when the plants were cultivated in a non-native environment. For field grown plants the highest DPT content was found in March (2nd year): 0.15 % w/w (dry weight) in roots; 0.03 % w/w in aerial parts. For plants grown in the climate room the highest concentration (0.14 % w/w) was observed in April (2nd year) in the roots and in July (1st year) in the aerial parts (0.05 % w/w). For the isolation of DPT roots are the most suitable part. The best harvest time is March (2nd year) for outdoor plants and April (2nd year) for indoor plants when height content and adequate biomass give the optimal DPT yield.

Keywords: lignan, deoxypodophyllotoxin, podophyllotoxin, Anthriscus sylvestris, environmental factors, plant cultivation; biosynthesis, etoposide, teniposide

47

seaso

na

l Varia

tion

s

2introduCtion

Anthriscus sylvestris (L.) Hoffm. (Apiaceae) is a common wild plant in northwest Europe, in parts of North America, Africa, Asia and New Zealand.1-3 The dried roots are used in Korean and Chinese traditional medicine for the treatment of various diseases, including bronchitis, as an antipyretic, cough remedy, and analgesic herbal drug.4-6 The plant accumulates considerable amounts of lignans, that are held responsible for the biological activity.7

Deoxypodophyllotoxin (DPT) is considered as the plant’s main lignan constituent.7 It has pharmacological properties such as antiproliferative, antitumor, antiviral, anti-inflammatory, anti-platelet aggregation and anti-allergic related disorders including asthma.8-16 The number of studies of DPT mechanism of action against cancer cells17 and other pharmacological properties are growing.4, 18-21 It also has anti-insecticidal properties.22, 23 The closely related podophyllotoxin is used for the semisynthesis of the cytostatic agents, etoposide and teniposide. These drugs are used for the treatment of lung and testicular cancers, Ewing’s sarcoma, lymphoma, glioblastoma and non-lymphomatic leukaemia.24

In several plant species, DPT serves as a biosynthetic precursor for podophyllotoxin (figure 1).25, 26 Feeding experiments with cultures of undifferentiated plant cells (Linum album) or fungi (Penicillium F-0543 and Aspergillus niger) have shown that DPT can be converted into podophyllotoxin.25, 27

To date, podophyllotoxin is obtained by extraction of the rhizome of Podophyllum species. The availability of Podophyllum species is limited28 and the increasing demands of podophyllotoxin jeopardizes the natural sources. Podophyllum species are listed on the endangered species list in India.29, 30 In the near future, the availability of podophyllotoxin will be a major bottleneck in supplying pharmaceutical needs.

Preliminary studies focusing on the question which factors may influence the production of DPT in A. sylvestris suggested that environmental factors highly determine the lignan profile and content.31, 32 It has also been reported that DPT content

Figure 1. Chemical structure of deoxypodohyllotoxin and podophyllotoxin.

48

seaso

na

l Varia

tion

s

2differs at least twofold between the low and high altitude both in aerial (0.13 % d.w. in 900m, 0.33 % d.w. in 1200m) and root part (0.38 % d.w in 900m and 0.78% d.w. in 1200 m) of A. sylvestris.21 Different populations of A. sylvestris yielded different lignan patterns, and that there was no clear genetic factor determining the lignan profiles when seeds from different locations were grown under identical conditions.31, 32 A major drawback of that study was the limited number of plants. Here we conducted a biannual experiment with a higher number of the collected plants and seeds, from a wider geographical range. The plants from the previous studies we included as well (table 2). All plants were grown under identical field conditions and in a climate room. DPT was the main important compound, therefore was used as the biosynthetic marker. The field resembles the natural weather conditions whereas the climate room resembles the controlled cultivation.

The first objective of this study is to determine the DPT content of native and non-native A. sylvestris grown in the field and in a climate room, in order to get insight into the production rate of DPT in A. sylvestris. Second, we aimed to study the influence of growing conditions, both outdoor and indoor, on the DPT production. Furthermore, it is important to identify rate limiting ecological factors for optimal growing conditions of A. sylvestris for DPT production.

reSuLtS

1. dpt content of native and cultivated Anthriscus plants The DPT content of Anthriscus plants from their native location was compared to that of the plants from the seeds of the native location, grown in the botanical garden De Kruidhof. The native A. sylvestris from Iceland (Mogilsa) contained the highest DPT production in the root part (0.182 %) and the native A. sylvestris from Germany (Bonn) contained the highest DPT production in the aerial (0.025 %). For the cultivated plants, the highest DPT production in the root part was the A, sylvestris from Bonn (0.081%) and Coventry (0.062 %), whereas in the aerial part was the A. sylvestris from Olst (0.018%). The results are shown in figure 2.

The DPT content of all field plants cultivated in the Netherlands in 2009 was used for the correlation analysis. There was a significant negative correlation between the DPT content in the aerial part of the field plants with the rainfall (r = -0.624, p < 0.01), with the duration of sun (r = -0.402, p < 0.01) and with the temperature (r = -0.643, p < 0.01). There was also a significant negative correlation between the DPT content in the root part of the field plants with the duration of sun (r = - 0.611, p < 0.01) and with the temperature (r = - 0.599, p < 0.01). In the climate room plants, there was a significant negative correlation between the DPT content in aerial part with the rainfall (r = -0.430, p < 0.01) and with the temperature (r = -0.190, p < 0.01). There was also a significant negative correlation between the DPT content in the root part with the duration of the sun (r = -0.228, p < 0.01), with the temperature (r = -0.464, p < 0.01), and with the rainfall (r = -0.362, p < 0.01). There was a positive correlation between

49

seaso

na

l Varia

tion

s

2

the DPT content in the aerial and in the root of the field plant (r = 0.315, p < 0.01) and the climate room plant (r = 0.374, p < 0.01). There was a significant positive correlation between the total yield of DPT in the aerial part of the field plants with the duration of sun (r = 0.221, p < 0.01). There was a significant negative correlation between the total yield of the DPT in the root part of the field plants with the duration of sun (r = -0.573, p < 0.01) and temperature (r = -0.507, p < 0.01). Thus, a lower temperature and less sun result in a higher DPT production in A. sylvestris.

2. dpt content of the field plants (outdoor plants)The experiments in the field with A. sylvestris plants grown from seeds of the various origins (uncontrolled condition – outdoor plants) and in the climate room (controlled condition – indoor plants) were carried out in parallel. The DPT content variation over the developmental stages was measured and compared both for the aerial (figure 3a) and the root parts (figure 3b). A. sylvestris is a biannual plant, therefore we started sampling in the second year when they develop biomass and flower. The DPT content in the roots of the field plants was up to 0.15 % and in the aerial parts up to 0.03 %. The highest DPT content in aerial part was found in March for all plants and in the root part was found in April, except A. sylvestris from Enebyberg and Egilstaffr, which were in March. The lowest was between July and September 2009. The highest yield of DPT per plant in the field plants was 22.23 mg in the aerial part in June 2009 and 68.87 mg in the root part in March 2009 (figure 3c and 3d). Data are summarized in table 1.

Figure 2. Deoxypodohyllotoxin content of flowering A. sylvestris collected in their native location and the flowering A. sylvetris cultivated from their native fruits in the same field (botanical garden de Kruidhof, the Netherlands). Note: seeds from Hrisey and Mogilsa did not germinate; therefore no data could be recorded in 2009.

50

seaso

na

l Varia

tion

s

2

Figure 3. A. sylvestris grown in the field. (a) Deoxypodophyllotoxin content in the aerial part. (b) Deoxypodophyllo-toxin content in the root part. (c) Deoxypodophyl-lotoxin yield in the aerial part. (d) Deoxypodo-phyllotoxin yield in the root part. 

September2009.ThehighestyieldofDPTperplantinthefieldplantswas22.23

mg in theaerialpart in June2009and68.87mg in the rootpart inMarch2009

(figure3cand3d).Dataaresummarizedintable1.

a

b

0,00

0,01

0,02

0,03

0,04

0,05

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

0,00

0,05

0,10

0,15

0,20

0,25

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

c

d

Figure 3. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in thefield.(b)DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrowninthefield.(c)AerialmassofA. sylvestrisgrown in the field. (d)RootmassofA. sylvestrisgrown in thefield. All DPT contents and yields significantly differ (p<0.01) comparing the indoor and outdoor plants.

0

5

10

15

20

25

30

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

0

10

20

30

40

50

60

70

80

90

100

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Sverige Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

c

d

Figure 3. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in thefield.(b)DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrowninthefield.(c)AerialmassofA. sylvestrisgrown in the field. (d)RootmassofA. sylvestrisgrown in thefield. All DPT contents and yields significantly differ (p<0.01) comparing the indoor and outdoor plants.

0

5

10

15

20

25

30

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

0

10

20

30

40

50

60

70

80

90

100

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Sverige Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

September2009.ThehighestyieldofDPTperplantinthefieldplantswas22.23

mg in theaerialpart in June2009and68.87mg in the rootpart inMarch2009

(figure3cand3d).Dataaresummarizedintable1.

a

b

0,00

0,01

0,02

0,03

0,04

0,05

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

0,00

0,05

0,10

0,15

0,20

0,25

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

September2009.ThehighestyieldofDPTperplantinthefieldplantswas22.23

mg in theaerialpart in June2009and68.87mg in the rootpart inMarch2009

(figure3cand3d).Dataaresummarizedintable1.

a

b

0,00

0,01

0,02

0,03

0,04

0,05

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

0,00

0,05

0,10

0,15

0,20

0,25

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

c

d

Figure 3. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in thefield.(b)DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrowninthefield.(c)AerialmassofA. sylvestrisgrown in the field. (d)RootmassofA. sylvestrisgrown in thefield. All DPT contents and yields significantly differ (p<0.01) comparing the indoor and outdoor plants.

0

5

10

15

20

25

30

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

0

10

20

30

40

50

60

70

80

90

100

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Sverige Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

September2009.ThehighestyieldofDPTperplantinthefieldplantswas22.23

mg in theaerialpart in June2009and68.87mg in the rootpart inMarch2009

(figure3cand3d).Dataaresummarizedintable1.

a

b

0,00

0,01

0,02

0,03

0,04

0,05

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

0,00

0,05

0,10

0,15

0,20

0,25

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

c

d

Figure 3. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in thefield.(b)DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrowninthefield.(c)AerialmassofA. sylvestrisgrown in the field. (d)RootmassofA. sylvestrisgrown in thefield. All DPT contents and yields significantly differ (p<0.01) comparing the indoor and outdoor plants.

0

5

10

15

20

25

30

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

0

10

20

30

40

50

60

70

80

90

100

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Sverige Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

51

seaso

na

l Varia

tion

s

2

a

b

0,00

0,01

0,02

0,03

0,04

0,05

0,06

0,07

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Con

cent

ratio

n (%

)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

0,00

0,02

0,04

0,06

0,08

0,10

0,12

0,14

0,16

0,18

0,20

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Con

cent

ratio

n (%

)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

c

d

Figure 4. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in theclimateroom(b).DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrownintheclimateroom(c).AerialmassofA. sylvestrisgrownintheclimateroom(d).RootmassofA. sylvestrisgrownintheclimateroom

0

1

2

3

4

5

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Yiel

d (m

g)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

0

2

4

6

8

10

12

14

16

18

20

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Yiel

d (m

g)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

c

d

Figure 4. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in theclimateroom(b).DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrownintheclimateroom(c).AerialmassofA. sylvestrisgrownintheclimateroom(d).RootmassofA. sylvestrisgrownintheclimateroom

0

1

2

3

4

5

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Yiel

d (m

g)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

0

2

4

6

8

10

12

14

16

18

20

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Yiel

d (m

g)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

a

b

0,00

0,01

0,02

0,03

0,04

0,05

0,06

0,07

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Con

cent

ratio

n (%

)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

0,00

0,02

0,04

0,06

0,08

0,10

0,12

0,14

0,16

0,18

0,20

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Con

cent

ratio

n (%

)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

c

d

Figure 4. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in theclimateroom(b).DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrownintheclimateroom(c).AerialmassofA. sylvestrisgrownintheclimateroom(d).RootmassofA. sylvestrisgrownintheclimateroom

0

1

2

3

4

5

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Yiel

d (m

g)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

0

2

4

6

8

10

12

14

16

18

20

Bonn Tartu Coventry Essex Buitenpost Olst Diever Nieuwegien Reykjavik Oudeschans Groningen

Location

Yiel

d (m

g)

June2008 July2008 August2008 March2009 April2009 May2009 June2009 July2009 August2009 Sept2009

Figure 4. A. sylves-tris grown in the climate room. (a) D e ox y p o d o p hy l -lotoxin content in the aerial part. (b) Deoxypodophyllo-toxin content in the root part. (c) Deox-ypodophyllotoxin yield in the aerial part. (d) Deoxypo-dophyllotoxin yield in the root part.

c

d

Figure 3. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in thefield.(b)DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrowninthefield.(c)AerialmassofA. sylvestrisgrown in the field. (d)RootmassofA. sylvestrisgrown in thefield. All DPT contents and yields significantly differ (p<0.01) comparing the indoor and outdoor plants.

0

5

10

15

20

25

30

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

0

10

20

30

40

50

60

70

80

90

100

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Sverige Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

September2009.ThehighestyieldofDPTperplantinthefieldplantswas22.23

mg in theaerialpart in June2009and68.87mg in the rootpart inMarch2009

(figure3cand3d).Dataaresummarizedintable1.

a

b

0,00

0,01

0,02

0,03

0,04

0,05

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

0,00

0,05

0,10

0,15

0,20

0,25

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

September2009.ThehighestyieldofDPTperplantinthefieldplantswas22.23

mg in theaerialpart in June2009and68.87mg in the rootpart inMarch2009

(figure3cand3d).Dataaresummarizedintable1.

a

b

0,00

0,01

0,02

0,03

0,04

0,05

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

0,00

0,05

0,10

0,15

0,20

0,25

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Con

cent

ratio

n (%

)

March2009 April2009 May2009 June2009 July2009 Aug2009

c

d

Figure 3. (a) Deoxypodohyllotoxin content in the aerial part ofA. sylvestris grown in thefield.(b)DeoxypodohyllotoxincontentintherootpartofA. sylvestrisgrowninthefield.(c)AerialmassofA. sylvestrisgrown in the field. (d)RootmassofA. sylvestrisgrown in thefield. All DPT contents and yields significantly differ (p<0.01) comparing the indoor and outdoor plants.

0

5

10

15

20

25

30

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegien Reykjavik Enebyberg Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

0

10

20

30

40

50

60

70

80

90

100

Bonn Tartu Coventry Suffolk Buitenpost Olst Diever Nieuwegein Reykjavik Sverige Oudeschans Groningen Egilsstafr

Location

Yiel

d (m

g)

March2009 April2009 May2009 June2009 July2009 Aug2009

52

seaso

na

l Varia

tion

s

2

3. dpt content of the climate room plants (indoor plants)Samples were taken during two years following the developmental stages of the life cycle of A. sylvestris. The DPT content and yield were measured and compared with outdoor grown plants of the same seed origin (figure 4). The DPT content of the climate room plants was up to 0.144% in the roots and up to 0.051% in the aerial part. The highest yield of DPT per plant in the field plants was 2.81 mg in the aerial part and 12.60 mg in the root part in April 2009 (table 1).

In the first year, the highest DPT content in the roots was found in A. sylvestris from Bonn (0.053%) in July 2008. In the aerial parts the highest content of DPT was found in A. sylvestris from Tartu (0.051%) also in July 2008. In the second year, in general the highest DPT content both in aerial as well as root parts were found in April 2009 for all plants except A. sylvestris from Bonn and Oudechans for aerial part and A. sylvestris from Reykjavik and Oudeschans (figure 4a and 4b). At this time the plants started to grow after the cold dormancy period.

Indoor A. sylvestris did not flower at the same time as the outdoor plants. All A. sylvestris plants from Bonn and Nieuwegein already started to flower by mid-April 2009, whereas only about 50% of plants from the other origins started to flower at this time point. The A. sylvestris plants from Estonia did not flower at all by mid-April 2009. By end of May 2009, all plants have flowered. In the Netherlands, A. sylvestris flowers between April and May.

Table 1. Summary of the highest and lowest DPT content, biomass and the DPT yield of Anthriscus sylvestris (figure 3 and 4)

Field Climate roomAerial Root Aerial Root

Highest DPT content (% w/w, dry weight)

0.033EnebybergMar 2009

0.151Nieuwegein

Mar 2009

0.051Tartu

Jul 2008

0.144Bonn

Apr 2009Lowest DPT content (% w/w, dry weight)

0.001Nieuwegein

Jul 2009

0.008ReykjavikSep 2009

0.008Essex

Jun 2009

0.012BuitenpostSep 2009

Highest mass (dry weight in g)

173.4NieuwegeinMay 2009

108.9Oudeschans

Jun 2009

7.0BuitenpostApr 2009

16.4Olst

Apr 2009Lowest mass (dry weight in g)

1.5Suffolk

Aug 2009

8.2EgilstaffrApr 2009

2.0Bonn

Jun 2008

0.8Tartu

Jun 2008Highest yield (dry weight in mg) per plant

22.23 mgOudeschans

Jun 2009

68.87 mgBonn

Mar 2009

2.81BuitenpostApr 2009

12.60 mgBonn

Apr 2009Lowest yield (dry weight in mg) per plant

0.01 mgCoventryJul 2009

0.73 mgReykjavikSep 2009

0.12 mgEssex

Jun 2009

0.32 mgCoventry

Sep 09

53

seaso

na

l Varia

tion

s

2In the first year, the highest DPT yield in the root part was in August and September

2008 for all plants, but in the aerial parts the highest yield was in July 2008 for all plants. In the second year, the highest DPT yield in the aerial as well as root part was in April 2009, except the root part of A. sylvestris from Groningen (figure 4c and 4d).

There is a significant difference comparing the DPT content of the field and climate room plants (p < 0.01). There is also a significant different comparing the total yield of the field and climate room plants (p < 0.01).

diSCuSSion

Anthriscus sylvestris is a biannual plant meaning that a life cycle takes 2 years. In the first year, the rosettes remain in the vegetative phase and form a basal rosette with a taproot. Vernalization will lead to flowering during the next growing season. After flowering, the seeds are produced and then the life cycle is completed.33 The biomass of the aerial and root of the field plants reached the highest in May in the second year for all plants, but the highest DPT content was in March for field plants, both in aerial and root part. However, only the aerial part of the field plants has the highest yield following the highest biomass in May for all plants except A. sylvestris from Oudeschans. The highest yield of the root part for all field plants was in March when the highest DPT content was reached (table 1).

DPT as the main lignan in A. sylvestris was used as a marker to study its metabolic variation over the developmental stages. This is the first study reporting results of climate room experiments covering the entire life cycle of two years. In general, the DPT content in the roots was higher than in the aerial parts.

native versus cultivatedThe first experiment was dedicated to compare the DPT content of plants of the native location with that of the same plants grown on the non-native location (Buitenpost). There were significant differences (p < 0.05) comparing the DPT content in the root parts of all plants except A. sylvestris from Enebyberg and Coventry (figure 2). The DPT content in the aerial part differed moderately over time. Meanwhile the DPT content in the roots decreased except in A. sylvestris from Coventry, UK. It seems that the A. sylvestris from Coventry were adapted easily to the Dutch soil. This is reflected in a more or less equal DPT content (0.007 % in aerial native compared to 0.005 % in aerial cultivation; 0.064% in root native compared to 0.062 % in root cultivation). The DPT content of the A. sylvestris from Bonn, DE, was moderately higher (significantly different of the DPT concentration in the root part, p < 0.05) compared to others (0.025% in aerial native compare to 0.017 % in aerial cultivation; 0.146% in root native compared to 0.081% in root cultivation). Based on these results, A. sylvestris from Coventry and Bonn were chosen as candidates for further cultivation. The DPT content in roots from the cultivated plants from Groningen was 7.5 fold decreased (0.087 % in roots native compared to 0.012 % in roots cultivation). This high producing plant at its

54

seaso

na

l Varia

tion

s

2native location became a low producing one when cultivated in a foreign environment. Interestingly, the DPT content in the root parts of Anthriscus from the local area of the botanical garden (Buitenpost) also decreased fivefold (0.119% in root native compare to 0.024% in root cultivation). In general, the decrease of DPT content in the cultivated root parts varied with a factor between 7.5 and 1 compared to the native plants (0.087 % in roots native compared to 0.012 % in roots cultivation in Groningen Anthriscus, 0.064% in roots native compared to 0.062% in roots cultivation in Coventry plant). These results clearly show that environmental factors play an important role in the DPT content in Anthriscus.

The statistical analysis showed that there is a significant negative correlation among the DPT content, the rainfall, the temperature and the duration of sun. Both in the aerial and root part, the DPT content is expected to increase with decreasing temperature and duration of sun. The DPT in aerial part is also influenced by rainfall. The DPT in aerial part is also expected to increase with decreasing rainfall. There is also a positive significant correlation between the DPT content in the root and aerial part. It means that the DPT in the aerial part is increase with increasing DPT in the root.34

This correlation could explain the significant differences (p < 0.05) found in DPT content of the root part of all plants (except A. sylvestris from Enebyberg and Coventry) from the Netherlands in 2007 and the Dutch cultivated plants in 2009. These plants originated from Buitenpost, Olst, Groningen, Diever and Nieuwegein. The DPT content of the cultivated plants became reduced up to 7.5 times in the root parts and up to 3 times in the aerial parts. The DPT content in the root would decrease when the duration of sun and the temperature were increased. In May 2007 the temperature was 1.7oC higher and there was 38 h less sun as compared to May 2009. This may have caused the DPT content in the root to decrease as shown in figure 2. Whereas in the aerial part, the DPT content would decrease when there were less sun, rainfall and temperature. In 2007, the weather in the Netherlands was unusual as there was no rain in March 2007, but there was 73 mm more rain in May 2007 compared to May 2009. The more abundant rainfall and higher temperature contributed to the decrease of the DPT in aerial part, however the less duration of sun contributed to the increase of the DPT in the aerial part. This might explain why the DPT content was reduced only up to 3 times in the aerial compared to up to 7.5 times in the root part.

outdoor versus indoorThe second experiment was to grow all plants in the same field, under the same environmental conditions. In parallel, we also grew the plants under controlled conditions in the climate room. The A. sylvestris grown in the field (outdoor) received influences from abiotic and biotic stress, such as temperature, light, drought, wind, and insects, and has more space to grow compared to the plants in the climate room. For experiments of controlled cultivation in a climate room we expect that differences found are mainly to be ascribed to genetic factors, as growing conditions are standardized. Furthermore, it is important to identify rate limiting ecological factors for improved cultivation either outdoor or indoor.

55

seaso

na

l Varia

tion

s

2In the climate room experiment, there was a decrease of DPT production during

the vegetative period in the 1st year (August – February) and also after the plants had flowered (2nd year). This biosynthetic profile was also observed in Podophyllum hexandrum. When the plant gets older, the podophyllotoxin content in the rhizome also decreases.35 The DPT content in A. sylvestris increased again in the spring (March – April 2nd year). Interestingly, the DPT pattern over the entire period in the climate room was in general similar for all plants (figure 4).

The DPT profiles of both the indoor and the outdoor plants were comparable, however there were differences in the mass of the aerial and root parts which influenced the yield and also in the flowering time of the indoor plants. There is a significant difference (p< 0.01) either of the DPT content or the DPT yield between indoor and outdoor plants. The differences found in DPT production between indoor and outdoor plants were comparable to a previous study from our group31 and from others.36 The inter-individual variability of the DPT content in the root parts of the outdoors plants was higher than of the indoor plants (e.g. 0.151 % - 0.007% versus to 0.144% - 0.012 % - see table 1). In contrast, the inter-individual variability of the DPT content in the aerial part of the indoor plants was higher than of the outdoor plants (e.g. 0.051% % - 0.008% versus to 0.033 % - 0.001% - see table 1). Previous research showed that the indoor plants contained higher DPT contents in aerial parts than in the roots.31 We now confirm these results as several indoor plants from Estonia (0.051 % in aerial, 0.045% in root), Buitenpost (0.031 % in aerial, 0.024% in root), and Olst (0.028 % in aerial, 0.023% in root), all sampled in July 2008, had higher a DPT content in the aerial part compared to the roots.

The differences found in the flowering time of the indoor plants points towards variations even though the plants were grown under identical conditions. These variations might point to differences in genotype at different geographical locations. The similarity in the DPT content observed under controlled conditions suggested that these genetic differences, if any do not influence the DPT production. However the significant differences in the DPT content and the yield between outdoor and indoor plants suggests that there might be environmental factor that play a role on the DPT production and yield.

The root mass of the outdoor plants was up to 8 times higher than that of the indoor plants and the aerial mass was up to 30 times higher than that of the indoor plants. This influenced the DPT yield. The DPT yield of the outdoor plants was up to 8 times higher in the aerial part and up to 5 times higher in the root part than that of the indoor plants. This variation in the root and aerial mass of the indoor plants was also observed in the previous study.31

In summary, these results show the DPT content over time for both indoor and outdoor plants. The best time to harvest plants for the DPT isolation is when the plants reach the optimal DPT production together with an adequate biomass. For indoor plants this was in April (2nd year) in both aerial and root part. For outdoor plants this was in March (2nd year) for the root part and in June (2nd year) for the aerial part. The A. sylvestris from Bonn showed constant production of DPT in the field over the season

56

seaso

na

l Varia

tion

s

2and may be a suitable candidate for on-going cultivation and breeding for production purposes and commercial isolation of DPT (figure 3). Another candidate is A. sylvestris from Coventry which showed a similar DPT content when cultivated elsewhere (figure 2), but further work to optimise growing conditions have to be carried out.

mAteriALS And metHodS

ChemicalsDeoxypodophyllotoxin (DPT) was a gift from Dr. M. Angeles Castro (Salamanca University, Salamanca, Spain) and was synthesized chemically. The identity and purity of DPT were checked by using HPLC (100 %). Acetonitrile and methanol were HPLC grade from Biosolve (Valkenswaard, the Netherlands). Dichloromethane is from Fisher Scientific (Landsmeer, the Netherlands).

plant materialThe fruits (ripe mericarps, hereafter called seeds) and plants of Anthriscus sylvestris L. were obtained from different sources (table 2). Vouchers specimens are deposited in our department coded Asylv2007-1 until Asylv2007-16. The identity of the seeds was verified and samples were stored in the reference collection of the Groningen Institute of Archaeology. A. sylvestris plants were collected at the flowering stage between May and June 2007, from the locations in the Netherlands and other European countries (table 2), and used for analysis. From the same plants and origin, seeds were collected. Plants were grown from these seeds (identical condition) in the Botanical Garden De Kruidhof (Buitenpost, the Netherlands, 53o 15’ 52” N, 06o 07’ 41” E) and sampled in 2009. The roots were separated immediately from the aerial parts. The plant material was dried at room temperature prior to grinding and extraction.

plant cultivation in the field (outdoor)The seeds were sown in December 2007 on soil in trays for cold stratification at botanical garden De Kruidhof and germinated in February 2008. These seedlings were transferred to the field in May 2008. The field samples were collected in the second year (2009) on a monthly basis, from March until August. The data (table 3) of the weather conditions in the Netherlands during this experiment in 2007 - 2009 were taken from KNMI (the Netherlands Meteorology Institute) website (www.knmi.nl).

plant cultivation in the climate room (indoor)The seeds were sown in December 2007 on soil at the botanical garden De Kruidhof and germinated in February 2008. The seedlings were transferred into individual pots and placed under day-night regime (16 hours light and 8 hours dark) at 20-22oC in May 2008. In the first year (2008), samples were collected in June, July and August. During the subsequent winter period (September 2008 – February 2009), the plants were placed in an open green house to induce a cold dormancy period. In March of the

57

seaso

na

l Varia

tion

s

2Table 2. Origin of the collected Anthriscus plants and seeds

No Origin Country Location amsla Collector1 Nieuwegeinb The

Netherlands52o 00’ 50” N 05o 06’ 68” E 1 Oktavia Hendrawati

2 Olstb The Netherlands

52o 32’ 97” N 06o 10’ 21” E 3 Oktavia Hendrawati

3 Dieverb The Netherlands

52o 85’ 95” N 06o 32’ 91” E 9 Oktavia Hendrawati

4 Groningenb The Netherlands

53o 11’ 34” N 06o 37’ 04” E 1 Oktavia Hendrawati

5 Buitenpost The Netherlands

53o 15’ 52” N 06o 07’ 41” E 1 Jan W. Zwart

6 Oudeschans The Netherlands

53° 08’19” N 07° 08’ 34” E 1 Rene J.T. Cappers

7 Bonn Germany 50o 45’ 04” N 07o 06’ 39” E 60 Frank Klingenstein8 Tartu Estonia 58o 36’ 60” N 26o 69’ 25” E 79 Christina Birnbaum9 Coventry United

Kingdom52° 24’ 03” N 01° 30’ 40” W 76 Dalila Martins

10 Essex United Kingdom

na na 38 Kew Garden, accession No: 0039420

11 Suffolk United Kingdom

na na na Kew Garden, accession No: 0053121

12 Enebyberg Sweden 59° 26’ 06” N 18° 01’48” E 20 Renata Alnsatter13 Reykjavik Iceland 66° 21’ 25” N 21° 71’ 29” W 15 Sigurdur H. Magnusson14 Egillstafr Iceland 64° 59’ 57” N 14° 17’ 22” W 34 Sigvardur Eirnarsson15 Hrisey Iceland 66° 01’ 03” N 18° 23’ 69” W 5 Bjarni E. Gudleifsson16 Mogilsa Iceland 66° 21’25” N 21° 71’ 29” W 65 Sigurdur H. Magnusson

aabove mean sea level (m)bsame locations as in the preliminary studies31, 32

second year (2009), they were placed back into the climate room and samplings were continued every month, from March until August.

Sample preparation and extraction method All sampling were from three individual plants (triplicate) and randomly selected. The samples were divided into aerial part and the root part. Subsequently, the biomass (fresh weight) of aerial and root part was weighed for every plant. The biomass data was an average of three individual plants. The samples from field and climate room experiment were freeze-dried and pulverized for analysis. The fresh weight - dry weight ratio was calculated based on an average of six plants. The ratio for the aerial part was 0.1891 and for the root part 0.2840. The yield was calculated based on DPT concentration and total dry weight of the biomass per plant. The extraction method

58

seaso

na

l Varia

tion

s

2

was as described by Koulman et al.31 Shortly, 100 mg dried plant material were weighed in a sovirel tube. A 2.0 ml portion of 80% methanol was added and the mixture was sonicated during 1 hour. Subsequently 4.0 ml of dichloromethane and 4.0 ml H2O were added. The mixture was vortexed and centrifuged at 1,000 g for 5 minutes. The aqueous layer was discarded and 2.0 ml of organic layer were transferred into a 2 mL Eppendorf tube. The organic layer was left in the fume hood until dried. The residue was redissolved in 2.0 ml methanol and filtered over a 0.45 µm HPLC syringe filter (nylon). The samples were submitted to HPLC analysis.

HpLC analysisThe HPLC analysis was as described by Vasilev et al.37 with some modifications. We used a Shimadzu-VP system consisting of a LC-10AT pump, a Kontron 360 auto sampler, a SPD M10A DAD detector, a FCV-10AL low pressure gradient mixer, a SCL-10A system controller, a FIAtron system CH-30 column heater, operated with LC Solution software, version 1.2. The column used was Zorbax Eclipse C18 (150 x 4.6 mm, 5 µm), together with Phenomenex guard cartridge C18 (4 x 3 mm) Phenomenex, Bester, The Netherlands. The detection wavelength was 240 nm. Mobile phase A [H2O: ACN (95:5)] and B [ACN: H2O (95:5)] were both in 0.1% formic acid and 2 mM ammonium formate. The injection volume was 5 µL. The flow rate was 1 mL/min using a gradient program of 30 min consisting of 1 min of B 30%, followed by a linear gradient until 15 min B 90%, at 20 min B 90%, at 25 min B 30% and at 30 min B 30%. Concentrations of metabolites were calculated using a calibration curve established with the reference compound (DPT).

Table 3. Data on rainfall, temperature and duration of sun in the Netherlands in 2008 and 2009 (www.knmi.nl)

Month1 2 3 4 5 6 7 8 9 10 11 12

Rainfall 2007 (mm) 104 68 85 0 138 90 161 42 97 32 58 77Rainfall 2008 (mm) 96 39 92 34 33 40 127 114 100 92 91 24Rainfall 2009 (mm) 54 55 48 20 65 54 107 47 34 90 120 84Temperature 2007 (°C) 7.1 6.0 8.0 13.1 14.1 17.5 17.0 17.1 13.8 10.1 6.9 3.8Temperature 2008 (°C) 6.5 5.1 5.9 8.9 15.7 16.5 18.1 17.4 13.6 10.1 6.9 2.4Temperature 2009 (°C) 0.8 3.3 6.3 12.2 13.9 15.6 18.1 18.5 15 10.7 9.5 2.2Duration of sun 2007(h) 50 57 157 284 198 172 193 199 116 130 62 72Duration of sun 2008 (h) 50 128 117 185 260 233 182 163 147 134 54 85Duration of sun 2009 (h) 101 48 142 227 236 232 233 226 160 123 51 60

59

seaso

na

l Varia

tion

s

2Statistical analysisStatistical calculations were carried out with the SPSS 16.0. Pearson correlation was used for correlation analysis and T-test was used for significant testing. P values < 0.05 were considered to be significant.

ACKnoWLedgementS

We thank M. A. Castro for giving us pure DPT as a reference compound, J.T.M. Elzenga for the climate room facility, J.W. Zwart and H. Drost for the A. sylvestris cultivation in the field at the Botanical garden De Kruidhof. We also thank J.J. Hogendorf, G. Telkamp, B. Farahanikia and E.J.J. Berm for their technical support. Financial support by the Ubbo Emmius fund, University of Groningen, is acknowledged.

referenCeS1. Jeong G, Kwon O, Park B, et al. Lignans

and coumarins from the roots of Anthr-iscus sylvestris and their increase of caspase-3 activity in HL-60 cells. Biol Pharm Bull 2007;30:1340-1343.

2. Magnusson, S.H. NOBANIS – Invasive alien species fact sheet – Anthriscus sylvesris – From: online database of the north European and Baltic network on Invasive alien species – NOBANIS. Available from: www.nobanis.org. 2006 (accessed 06.09.2011)

3. Hulten E, Fries M. Atlas of north European vascular plants north of the tropic of cancer. Konigstein: Koeltz Sci-entific Books, 1986.

4. Yong Y, Shin S, Lee Y, et al. Antitumor activity of deoxypodophyllotoxin isolated from Anthriscus sylvestris: Induction of G2/M cell cycle arrest and caspase-dependent apoptosis. Bioorg Med Chem Lett 2009;19:4367-4371.

5. Wong S, Tsui S, Kwan S, et al. Identifica-tion and characterization of Podophyllum emodi by API-LC/MS/MS. J Mass Spec 2000;35:1246-1251.

6. Koulman A, Kubbinga ME, Batterman S, et al. A phytochemical study of lignans in whole plants and cell suspension cultures of Anthriscus sylvestris. Planta Med 2003;69:733-738.

7. Koulman A, Bos R, Medarde M, et al. A fast and simple GC MS method for lignan profiling in Anthriscus sylvestris and bio-synthetically related plant species. Planta Med 2001;67:858-862.

8. Kozawa M, Baba K, Matsuyama Y, et al. Components of the root of Anthriscus sylvestris Hoffm. II. Insecticidal acitivity. Chem Pharm Bull 1982;30:2885-2888.

9. Charlton JL. Antiviral activity of lignans. J Nat Prod 1998;61:1447-1451.

10. Jin M, Moon TC, Quan Z, et al. The naturally occurring flavolignan, de-oxypodophyllotoxin, inhibits lipopol-ysaccharide-induced iNOS expression through the NF-kappaB activation in RAW264.7 macrophage cells. Biol Pharm Bull 2008;31:1312-1315.

11. Jin M, Lee E, Yang JH, et al. Deoxypo-dophyllotoxin inhibits the expression of intercellular adhesion molecule-1 induced by tumor necrosis factor-alpha; in murine lung epithelial cells. Biol Pharm Bull 2010;33:1-5.

12. Chen J, Chang Y, Teng C. et al. Anti-platelet aggregation alkaloids and lignans from Hernandia nymphaeifolia. Planta Med 2000;66:251-256.

13. Gordaliza M, Castro MA, García-Grávalos MD, et al. Antineoplastic and antiviral activities of podophyllotoxin related lignans. Archiv der Pharmazie 1994;327:175-179.

14. Lee SH, Son MJ, Ju HK, et al. Dual inhi-bition of cyclooxygenases-2 and 5-lipox-ygenase by deoxypodophyllotoxin in mouse bone marrow-derived mast cells. Biol Pharm Bull 2004;27:786-788.

15. Sudo K, Konno K, Shigeta S, et al. Inhibi-tory effects of podophyllotoxin deriva-

60

seaso

na

l Varia

tion

s

2tives on herpes simplex virus replication. Antivir Chem Chemother 1998;9:263-267.

16. Ikeda R, Nagao T, Okabe H, et al. Anti-proliferative constituents in umbelliferae plants. III. Constituents in the root and the ground part of Anthriscus sylvestris Hoffm. Chem Pharm Bull 1998;46:871-874.

17. Shin SY, Yong Y, Kim CG, et al. De-oxypodophyllotoxin induces G2/M cell cycle arrest and apoptosis in HeLa cells. Cancer lett 2010;287:231-239.

18. Choi H, Lee J, Shin H, et al. Deoxypo-dophyllotoxin reduces skin pigmenta-tion of brown guinea pigs. Planta Med 2004;70:378-380.

19. Suh S, Kim J, Jin U, et al. Deoxypodo-phyllotoxin, flavolignan, from Anthriscus sylvestris Hoffm. inhibits migration and MMP-9 via MAPK pathways in TNF-[α]-induced HASMC. Vasc Pharmacol 2009;51:13-20.

20. Lin CX, Lee E, Jin MH, et al. Deoxy-podophyllotoxin (DPT) inhibits eosi-nophil recruitment into the airway and Th2 cytokine expression in an OVA-induced lung inflammation. Planta Med 2006;72:786-791.

21. Dall’Acqua S, Giorgetti M, Cervellati R, et al. Deoxypodophyllotoxin content and antioxidant activity of aerial parts of Anthriscus sylvestris Hoffm. Z Naturfor-sch C 2006;61:658-62.

22. Rong G, Congfen G, Xuan T, et al. Insec-ticidal activity of deoxypodophyllotoxin, isolated from Juniperus sabina L, and related lignans against larvae of Pieris rapae L. Pest Manag Sci 2004;60:1131-1136.

23. Xu H, Wang JJ. Natural products-based insecticidal agents 5. Design, semisyn-thesis and insecticidal activity of novel 4’-substituted benzenesulfonate deriva-tives of 4-deoxypodophyllotoxin against Mythimna separata Walker in vivo. Bioorg Med Chem Lett 2010;20:2500-2502.

24. Ayres DC, Loike JD. Lignans: chemical, biological and clinical properties. Cambridge, UK: Cambridge University Press, 1990.

25. Federolf K, Alfermann AW, Fuss E. Aryltetralin-lignan formation in two different cell suspension cultures of Linum album: deoxypodophyllotoxin

6-hydroxylase, a key enzyme for the formation of 6-methoxypodophyllotox-in. Phytochem 2007;68:1397-1406.

26. Jackson DE, Dewick PM. Aryltetralin lignans from Podophyllum hexandrum and Podophyllum peltatum. Phytochem 1984;23:1147-1152.

27. Kondo K, Ogura M, Midorikawa Y, et al. Conversion of deoxypodophyllotoxin to podophyllotoxin-related compounds by microbes. Agric Biol Chem 1989;53:777-782.

28. Choudhary DK, Kaul BL, Khan S. Culti-vation and conservation of Podophyllum hexandrum—an overview. J Med Aromat Plant Sci 1998;20:1071-1073.

29. Nayar MP, Sastry APK. Red data book of Indian plants. Calcutta, 1990.

30. Giri A, Narasu ML. Production of podophyllotoxin from Podophyllum hexandrum: a potential natural product for clinically useful anticancer drugs. Cy-totechnol 2000;34:17-26.

31. Koulman A, Batterman S, van Putten FMS, et al. Lignan profiles of indoor-cul-tivated Anthriscus sylvestris. Planta Med 2003;69:959-961.

32. Koulman A, Hendrawati O, Batterman S, et al. The seasonal variations of lignan profiles in Anthriscus sylvestris (L.) Hoffm. Planta Med 2007;73:P_112.

33. van Mierlo JEM, van Groenendael JMA. Population dynamic approach to the control of Anthriscus sylvestris (L.) Hoffm. J Appl Ecol 1991:128-139.

34. Field A. Discovering statistics using SPSS. London: SAGE Publications, 2005; 125-129

35. Pandey H, Nandi S, Kumar A, et al. Po-dophyllotoxin content in Podophyllum hexandrum Royle plants of known age of seed origin and grown at a lower altitude. Acta Phys Plant 2007;29:121-126.

36. Robinson AR, Ukrainetz NK, Kang K, et al. Metabolite profiling of Douglas-fir (Pseudotsuga menziesii) field trials reveals strong environmental and weak genetic variation. New Phytol 2007;174:762-773.

37. Vasilev NP, Julsing MK, Koulman A, et al. Bioconversion of deoxypodophyllo-toxin into epipodophyllotoxin in E. coli using human cytochrome P450 3A4. J Biotechnol 2006;126:383-393.

ChapterIdentification of lignans and related compounds in Anthriscus sylvestris

by LC-ESI-MS/MS and LC-SPE-NMR

Oktavia Hendrawati Herman J. Woerdenbag

Paul J.A. Michiels Herald G. Aantjes

Annie van Dam Oliver Kayser

Phytochemistry 2011; in press

3

64

iDen

tiFica

tion

oF lig

na

ns

3

oktavia Hendrawati1, Herman J. Woerdenbag2, paul J.A. michiels3, Herald g. Aantjes3,

Annie van dam4, oliver Kayser1,5

1Department of Pharmaceutical Biology, University of Groningen, Antonius Deusinglaan 1, 9713 AV, Groningen, The Netherlands; 2Department of Pharmaceutical Technology and Biopharmacy, University of Groningen Antonius Deusinglaan 1, 9713 AV, Groningen, The Netherlands; 3Spinnovation Analytical BV, Nijmegen, the Netherlands; 4Mass Spectrometry Core Facility, University of Groningen, Groningen, The Netherlands; 5Technical University Dortmund, Technical Biochemistry, Emil-Figge-Strasse 66, 44227 Dortmund, Germany

Phytochemistry 2011; in press

AbStrACt

The aryltetralin lignan deoxypodophyllotoxin is much more widespread in the plant kingdom than podophyllotoxin. The latter serves as a starting compound for the production of cytostatic drugs like etoposide. A better insight into the occurrence of deoxypodophyllotoxin combined with detailed knowledge of its biosynthestic pathway(s) may help to develop alternative sources for podophyllotoxin. Using HPLC combined with electrospray tandem mass spectrometry and NMR spectroscopy techniques, we found lignans and related structures in roots of Anthriscus sylvestris (L.) Hoffm. (Apiaceae), a common wild plant in temperate regions of the world. Podophyllotoxone, deoxypodophyllotoxin, yatein, anhydropodorhizol, 1-(3’-methoxy-4’,5’-methylenedioxyphenyl)1-ξ-methoxy-2-propene, and 2-butenoic acid, 2-methyl-4-[[(2Z)-2-methyl-1-oxo-2-buten-1-yl]oxy]-, (2E)-3-(7-methoxy-1,3-benzodioxol-5-yl)-2-propen-1-yl ester, (2Z)- were the major compounds. a-Peltatin, podophyllotoxin, b-peltatin, isopicropodophyllone, b-peltatin-a-methylether, (Z)-2-angeloyloxymethyl-2-butenoic acid, anthriscinol methylether, and anthriscrusin were present in lower concentrations. a-Peltatin, b-peltatin, isopicropodophyllone, podophyllotoxone, and b-peltatin-a-methylether have not been previously reported to be present in A. sylvestris. Based on our findings we propose a hypothetical biosynthetic pathway of aryltetralin lignans in A. sylvestris.

Keywords: Anthriscus sylvestris L. (Hoffm.), Apiaceae, cow parsley, wild chervil, LC-ESI-MS/MS, LC-SPE-NMR, aryltetralin lignans, biosynthetic pathway, deoxypodophyllotoxin

65

iDen

tiFica

tion

oF lig

na

ns

3introduCtion

Anthriscus sylvestris L. (Hoffm.) (Apiaceae) is a perennial herb that grows in Europe and in parts of North America, Africa, Asia and New Zealand.1-3 The aryltetralin lignan deoxypodophyllotoxin (6) is an interesting constituent of the plant because it can be used as a precursor for the production of podophyllotoxin (2).4 Podophyllotoxin (2) is important as a semi-synthetic precursor for the anticancer drugs etoposide, teniposide, and etopophos.5

To date, 2 is obtained by isolation from Podophyllum species. In the future, the availability of 2 from this source is likely to become a major bottleneck. Podophyllum species have been listed on the endangered species list in India, proving that the continuous demand of 2 is a serious threat for their existence.6 An alternative source of 2 may be found by the (biotechnological) hydroxylation of 6 at the C7 position.7 Compound 6 is much more widespread in the plant kingdom than 2. 1-4 A better insight into the occurrence of 6 combined with profound knowledge of its biosynthetic pathway(s) may help to develop alternative sources for the desired lignans.

In addition to 6, several other compounds have been reported to occur in A. sylvestris, including monoterpenes, anthricinol, angeloyl butenoid acid, anthriscusin (13), anthricin, isoanthricin and crocactone.8-10 The objective of the present study was to further investigate the occurrence of lignans and related structures in A. sylvestris. A better insight into the phytochemistry of A. sylvestris will increase our understanding of possible biosynthetic pathways for podophyllotoxin and related lignans and contribute to the development of alternative sources for the desired compounds. LC-SPE-NMR and LC-ESI-MS/MS techniques were used for the identification and structure elucidation.

reSuLtS And diSCuSSion

A typical HPLC chromatogram of the aerial and the root parts of A. sylvestris is shown in figure 1. Because the underground parts contained more compounds and at higher concentrations compared to the aerial parts, we used the root extract for the separation, isolation and identification of lignans and related structures, 14 in total.

Compounds 1, 2, 3, and 6-9 were identified based on comparison with pure reference compounds using HPLC, molecular weights and fragment ions by LC-MS/MS. Compound 4, 5, and 10-14 were identified based on molecular weights, fragment ions, 13C NMR and 1H NMR spectra. The chemical structures are shown in figure 2. The molecular weights and the fragment ions are summarized in table 1. The comprehensive and detailed assignment of the multiplicity and coupling constants of 13C NMR and 1H NMR spectra are summarized in table 2.

We are the first to report on the presence of a-peltatin (1), b-peltatin (3), isopicropodophyllone (4), podophyllotoxone (5) and b-peltatin-a-methylether (9) in A. sylvestris.

66

iDen

tiFica

tion

oF lig

na

ns

3

Figure 1. UV chromatograms obtained at 240 nm of aerial parts and root extract of Anthriscus sylvestris.

Figure 1. UV chromatograms obtained at 240 nm of aerial parts and root extract of

Anthriscus sylvestris

Compounds 1, 2, 3, and 6-9 were identified based on comparison with pure

reference compounds using HPLC, molecular weights and fragment ions by LC-

MS/MS. Compound 4, 5, and 10-14 were identified based on molecular weights,

fragment ions, 13C NMR and 1H NMR spectra. The chemical structures are shown

in figure 2. The molecular weights and the fragment ions are summarized in

table 1. The comprehensive and detailed assignment of the multiplicity and

coupling constants of 13C NMR and 1H NMR spectra are summarized in table 2.

Compound 1 (mw: 400), 2 (mw: 414), 3 (mw: 414), and b-peltatin-a-methylether (9, mw: 428) are aryltetralin lignans. The general fragmentation ions were produced by loss of the pendant phenyl substituent [B+H]+ and subsequent loss of CO2 from the lactone ring [B-CO2+H]+.11 The latter fragment is either not observed or very weak. [B+H]+ shows the dehydration product [B-H2O+H]+ with somewhat higher intensity.11

The fragment ions of 1 are 247 [B+H]+, 229 [B+H-H2O]+, and 185 [A]+. The fragment ions of 2 are 313 [A+H]+, 247 [B+H]+, 229 [B+H-H2O]+, and 185 [B+H-H2O-CO2]

+. Compound 3 has the same fragmentation as 1 with additional 203 [B-CO2+H]+. The fragment ions of compond 9 are 261 [B+H]+ and 217 [B-CO2+H]+. The concentration of these peltatins in A. sylvestris were low and hardly detectable by HPLC. However using combined LC-MS/MS, a technique more sensitive than HPLC-UV alone, we could confirm the identity of compounds 1, 2, 3 and 9, also in comparison to the literature.11

Compounds 4 and 5 have the same molecular weight of 412 and the same fragment ion patterns, 245 [B+H]+ and 201 [B-CO2]

+. We further identified the stereochemistry by 1H-NMR and 13C-NMR and compared it to the literature.12, 13 The coupling constant and multiplicity of especially H-7’, H-8’, H-8 and H-9 determine the stereochemistry. The signals of H-7’ of compound 4 and 5 appear as doublets at δ 4.72 (J = 6.3 Hz) and δ 4.84 (J = 4.6 Hz), respectively. These results are not in agreement with picropodophyllone13 which has a singlet for this proton resonance. The multiplicity of H-8’ of compound 4 is multiplet at δ 3.77 whereas this signal in compound 5 appears

67

iDen

tiFica

tion

oF lig

na

ns

3

as a doublet of doublets (dd) at δ 3.77 (J = 15.6, 4.6 Hz). The multiplicity of H-8 of compound 4 is multiplet at δ 3.83 whereas this signal in compound 5 appears as a doublet of doublets (ddd) at δ 3.59 (J = 15.6 and 4.7 Hz). The multiplicity of H-9 of compound 4 is multiplet at δ 3.87 and triplet at δ 4.50 (J= 7.9 Hz) whereas this signal in compound 5 appears as a doublet of doublets (dd) at δ 4.31 (J=10.3 and 8.4 Hz) and triplet at δ 4.49 (J=7.9 Hz). The 13C-NMR spectra also showed characteristic values distinguishing these stereochemistry at C-3, C-7, C-8, C-9, C-8’, and C-9’. These values are similar with the reported data.12, 13

The main lignan in A. sylvestris is deoxypodophyllotoxin (6). The identity of this compound was based on comparison of the retention time of reference compound by HPLC, the molecular weight (398), the fragment ions, and the 13C NMR and 1H NMR spectra. Our data are in agreement with literature.14

Table 1. LC-ESI-MS and LC-ESI-MS/MS data of the compounds identified in root extract of Anthriscus sylvestris.

No Compound MW Molecular Ions of [M+NH4]

+ or [M+H]+ion

Fragment ions

1 a-Peltatin 400 418 247, 229, 1852 Podophyllotoxin 414 432 313, 247, 229, 1853 b-Peltatin 414 432 247, 229, 203, 1854 Isopicropodophyllone 412 430 245, 2015 Podophyllotoxone 412 430 245, 2016 Deoxypodophyllotoxin 398 416 231, 1877 Yatein 400 418 223, 1818 Anhydropodorhizol 398 416 231, 1359 b-Peltatin-a-methylether 428 446 261, 21710 (Z)-2-angeloyloxymethyl-2-butenoic acid 198 199 18111 Anthriscinol methyl ether 222 223 191, 16112 1-(3’-methoxy-4’,5’-

methylenedioxyphenyl)1-ς-methoxy-2-propene

222 229d 191, 161

13 Anthriscrusin 388 406 19114 2-butenoic acid, 2-methyl-4-[[(2Z)-

2-methyl-1-oxo-2-buten-1-yl]oxy]-, (2E)-3-(7-methoxy-1,3-benzodioxol-5-yl)-2-propen-1yl ester, (2Z)-

388 406 191

a Unknown 428b Unknown 416c Unknown 392

d[M+Li]+

68

iDen

tiFica

tion

oF lig

na

ns

3

Figure 2. Chemical structures of compounds in the root extract of Anthriscus sylvestris

Table 1. LC-ESI-MS and LC-ESI-MS/MS data of the compounds identified in root extract

of Anthriscus sylvestris

No Compound MW Molecular Ions of [M+NH4]+ or [M+H]+ion

Fragment ions

1 α-Peltatin 400 418 247, 229, 185

Figure 2. Chemical structures of compounds in the root extract of Anthriscus sylvestris.

We identified compound 10 as (Z)-2-angeloyloxymethyl-2-butenoic acid, an acyloxycarbocyclic acid, based on the molecular weight (198), and 1H-NMR and 13C-NMR spectra. Our proton NMR data (table 2) are in agreement with those published by Kozawa who isolated this compound from A. sylvestris.8

Compound 11 was identified as anthriscinol methyl ether, a phenylpropanoid derivative, based on the molecular weight (222), the 1H-NMR and 13C-NMR spectra. The NMR spectra were confirmed with reported literature.15 Compound 11 has been reported to display weak insecticidal activity.16

The molecular weight of compound 12 is 222. The 13C NMR and 1H NMR data compared with the literature13 confirmed that compound 12 is 1-(3’-methoxy-4’,5’-methylenedioxyphenyl)-1ξ-methoxy-2-propene.15 The isolation and uses of compound 12 in A. sylvestris are for prevention or treatment of cancer or condyloma.17

69

iDen

tiFica

tion

oF lig

na

ns

3The molecular weight of compounds 13 and 14 is 388 and showed similar fragment

ions in LC-MS data. The NMR data and compared with data from the literature9, 15 suggested that compound 13 is O-[(Z)-2-angeloyloxymethyl-2-butenoyl]-3-methoxy-4,5-methylendioxyamyl alcohol (anthriscusin). It is a phenylpropanoid ester derived from anthriscinol and compound 10.16 Based on our intepretation of 13C NMR and 1H NMR data, we propose that compound 14 is 2-butenoic acid, 2-methyl-4-[[(2Z)-2-methyl-1-oxo-2-buten-1-yl]oxy]-, (2E)-3-(7-methoxy-1,3-benzodioxol-5-yl)-2-propen-1-yl ester, (2Z)-. Selective 1D NOE experiments were carried out to identify the NOE long range coupling for the 4’’ CH to the 3’’ CH3 group. The NOE was difficult to identify in the 2D NOESY experiment because it overlapped with the large NOE between 3’’’ and 4’’’. The weak NOE between CH2 and the CH3 could be explained by spin-diffusion because it is very far away through space. The biological study, isolation and uses of compound 14 in A. sylvestris for prevention or treatment of cancer or condyloma were patented.17 However there was no NMR or MS data reported in the patent.

The major compounds in the roots were compounds 5-8, 12 and 14. Compound 1 and 3 were isolated from P. hexandrum for the first time.18 In A. sylvestris, we now report and structure elucidated these compounds for the first time. Trace amounts of compound 2 (< 0.01 µg/mg d.w.) have been reported in A. sylvestris.19 Compounds 10 and 13 have been isolated earlier from the roots of A. sylvestris.8, 9 Compound 6 was isolated for the first time from the roots of A. sylvestris.10 Compound 7, 11, 12 have been isolated from the roots of A. sylvestris.15 Compound 7 and 12 were also isolated from the fruit of A. sylvestris.20

The low concentration of 1 and 3 might be due to the fact that A. sylvestris was collected after having flowered in May 2007. A decrease of these compounds after flowering has also been shown in Podophyllum species.21 The loss of 1 and 3 during growth might due to metabolic turnover.21

The lignans isolated in this study (1-9) may be involved in the late biosynthesis pathway of aryltetralin lignans in A. sylvestris. Federolf et al.22 has recently proposed hypothetical pathway of 2 in Linum album. Combined with our findings, we propose a biosynthetic pathway of 6 and related lignans in A. sylvestris (figure 3).

Matairesinol was incorporated effectively into 1, 2, 3, 4’-demethylpodophyllotoxin, and 4’-demethyldeoxypodophyllotoxin in P. hexandrum root, P. peltatum leaves and Diphylleia cymosa leaves.23 In A. sylvestris, matairesinol may be incorporated into 7, 8, 4’-demethylyatein (putative intermediate) and other compounds24 which are further metabolized into 6, 2, 5, 1, 3, and 9 (figure 3).

Feeding experiments with labelled 6 and 4’-demethyldeoxypodophyllotoxin in P. hexandrum showed that these compounds are the likely precursors of 1 and 3 respectively, by hydroxylation in the aromatic ring.21 The transformation of compound 6 into 2 and 3 has been shown in a feeding experiment in P. hexandrum.21

In P. hexandrum, it has been shown that 2 is derived from 6 by hydroxylation. Further oxidation yielded 5, the later reaction being reversible.18, 25 Compound 5 may be also reduced in vivo into 2.25 In our studies with A. sylvestris, 2 was detected in low amount together with its oxidation products, 5 and its diasteoisomer, 4 were present.

70

iDen

tiFica

tion

oF lig

na

ns

3 Table 2. NMR spectroscopic data (500MHz, d6-DMSO) for metabolites identified in Anthriscus sylvestris root extracts by LC-SPE-NMR.

Position2 4 5 6

δ C δ H (J in Hz)

δ C δ H (J in Hz)

δ C δ H (J in Hz)

δ C δ H (J in Hz)

1 166.9 128.0 127.9 131.0 6.49, dt (15.8)2 127.3 140.0 141.6 124.9 6.25, dt

(15.8, 5.8, 5.8)3 137.1 6.10, qq 108.0 6.86, s 109.4 6.85, s 71.9 4.00, dd

(6.0, 1.8)4 15.2 1.88, dq (7.4, 1.8,

1.6, 1.5)152.7 152.4

5 20.0 1.81, q (1.45, 1.57, 1.62, 1.46)

147.6 147.6

6 104.7 7.31, s 104.4 7.40, s7 195.0 192.48 44.2 3.83, m 42.8 3.59, ddd

(15.6, 4.7)9 68.5 3.87, m, 4.50,

t (7.9) 66.4 4.31, dd

(10.3, 8.4) and 4.49, t (7.9)

1’ 64.7 4.71, s 135.1 133.2 131.32’ 128.5 106.3 6.35, s 107.5 6.37, s 106.7 6.72, d (1.6)3’ 141.1 6.41, q (7.5, 7.2) 152.6 152.4 143.44’ 15.2 2.00, d (7.2) 136.8 136.8 134.55’ 167.2 152.6 152.4 148.86’ 12.67, s 106.3 6.35, s 107.5 6.37, s 99.4 6.77, d (1.1)7’ 42.7 4.72, d (6.3) 43.6 4.84, d (4.6)8’ 43.5 3.77, m 44.8 3.77, dd (15.6, 4.6)9’ 175.4 173.6

OCH2O 102.6 6.13, d (0.8), 6.13, d (0.8)

102.9 6.16, d (0.8), 6.16, d (0.8)

101.0 5.97, s

C-3’, OCH3

55.5 3.63, s 55.6 3.65, s 56.0 3.83, s

C-4’, OCH3

60.1 3.62, s 59.8 3.63, s

C-5’, OCH3

55.5 3.63, s 55.6 3.65, s

C-3, OCH3

57.0 3.27, s

71

iDen

tiFica

tion

oF lig

na

ns

3Position

7 8 12 13δ C δ H

(J in Hz)δ C δ H

(J in Hz)δ C δ H

(J in Hz)δ C δ H

(J in Hz)

1 130.6 83.1 4.58, d (6.6) 133.2 6.57, dt (15.9, < 1) 133.1 6.61, dt (15.9, < 1)

2 129.1 138.7 5.86, ddd (17.1, 10.3,

6.6)

122.0 6.26, dt (15.9, 6.2) 122.1 6.31, dt (15.9, 6.3)

3 119.6 6.51, s 115.3 5.14 ddd (10.3, 1.8,

1.1); 5.25, ddd (17.2, 1.8, 1.3)

64.5 4.77, m (nd) 64.7 4.78, dd (6.2, 1.2)

4 145.9

5 145.9

6 108.2 6.81, s

7 31.8 2.75, dd (15.4, 11.7); 3.03, dd (16.4, 5.1)

8 32.3 2.62, m

9 71.3 3.96, dd (10.6, 8.2) ; 4.42, t (7.6)

1’ 137.1 135.7 131.0 130.8

2’ 108.0 6.30, s 106.1 6.58, d (1.3) 106.9 6.71, d (1.4) 107.1 6.76, d (1.4)

3’ 152.1 143.1 142.9 143.2

4’ 136.3 134.1 134.9 134.9

5’ 152.1 148.4 148.8 148.8

6’ 108.0 6.30, s 100.2 6.51, d (1.3) 99.7 6.75, d (1.3) 99.7 6.80, d (1.4)

7’ 42.8 4.52, d (5.3)

8’ 45.9 2.99, dd, (13.5, 5.2)

OCH2O 99.8 5.94, d (0.9)5.97, d (0.9)

100.0 5.95, s 101.0, CH2

5.96, s 100.6 5.96, s

C-3’, OCH3

55.6 3.64, s 56.0 3.79, s 56.1 3.83, s 56.1 3.84, s

C-4’, OCH3

59.7 3.62, s

C-5’, OCH3

55.6 3.64, s

C-1, OCH3

55.3 3.20, s

1’’ 165.3 166.2

2’’ 127.2 128.4

3’’ 143.1 6.53, qt (7.2, < 1) 19.3 1.93, tq (5.3, 1.6nd)

4’’ 15.6 2.03, dt (7.2, < 1) 138.1 6.18, tq (5.3, 16)

5’’ 64.6 4.78, m (nd) 61.8 5.02, dq (5.3, 1.8)

1’’’ 166.8 167.0

2’’’ 127.2 127.2

3’’’ 137.5 6.07, qq (7.2, 1.5) 137.8 6.14, qq (7.2, 1.5)

4’’’ 15.3 1.85, dq (7.2, 1.5) 15.5 1.92, dq (nd, 1.5)

5’’’ 20.1 1.78 , p (1.5) 20.1 1.85, p (1.5)

nd: not determined due to overlap

72

iDen

tiFica

tion

oF lig

na

ns

3

Compound 4 was first isolated in nature from P. pleianthum.26 Compounds 4 and 5 have been reported to be present in P. pleianthum, P. hexandrum and P. peltatum.27 Since 5 is readily isomerised to 4 by heat, the latter compound may well be an artefact from 5 in Podophyllum extracts.26 Compound 4 is quite likely to be an artefact formed from 5.

Compound 7 has been reported to be an earlier precursor of deoxypodophyllo-toxin,25 while compound 3 is the product of the metabolization of 6 according to the hypothetical biosynthetic pathway of lignans which has been reported previously.18, 22 Broomhead et al.23 proposed 4’-demethylyatein and 4’-demethyldeoxypodophyllotoxin as putative intermediates between matairesinol and 1 in P. hexandrum root, P. peltatum leaf and Diphylleia cymosa leaf. Although 4’-demethylyatein and 4’-demethyldeoxypodophyllotoxin were not detected in the root extract of

present. Compound 4 was first isolated in nature from P. pleianthum.26

Compounds 4 and 5 have been reported to be present in P. pleianthum, P.

hexandrum and P. peltatum.27 Since 5 is readily isomerised to 4 by heat, the latter

compound may well be an artefact from 5 in Podophyllum extracts.26 Compound 4

is quite likely to be an artefact formed from 5.

Figure 3. Proposed biosynthetic pathway of deoxypodophyllotoxin and related lignans in Anthriscus sylvestris (additional compounds: 1-5, 8, and 9, nd = not detectable).

73

iDen

tiFica

tion

oF lig

na

ns

3A.  syvlestris, these compounds may be putative intermediates between matairesinol and 1 (figure 3). However, we could not trace these earlier precursors such as matairesinol and later products of lignan biosynthesis detected in other plants, such as 6-methoxypodophyllotoxin.22

In the late biosynthetic pathway of podophyllotoxin,18, 22 we suggest a role of other (or: additional) compounds related to the formation of podophyllotoxin in A. sylvestris (figure 3). Our results add to what is known about the biosynthetic pathway of lignans in A. sylvestris. Additional experiments to further unravel this biosynthetic pathway of lignans in A. sylvestris are needed.

In summary, the combination of the liquid chromatography, electrospray ionization mass spectrometry (LC-ESI-MS/MS), and LC-SPE-NMR techniques are proven to be powerful tools to elucidate the structures unknown compounds in A. sylvestris extracts. In total, we identified 14 compounds in the root extract of A. syvlestris which may be involved in the late biosynthesis pathway of aryltetralin lignans as shown in figure 3. We identified the 6-methoxy lignans (1 and 3) as new side pathway, which previously was not known.

mAteriAL And metHodS

ChemicalsDeoxypodophyllotoxin was a gift from Dr. M. Angeles Castro (Salamanca University, Salamanca, Spain). Anhydropodorhizol, yatein, a-peltatin, b-peltatin and b-peltatin-a-methylether were kindly provided from Prof. M. Medarde (Salamanca University, Salamanca, Spain). Podophyllotoxin was from Sigma Aldrich (Zwijndrecht, the Netherlands). The identity and purity of these reference compounds were checked using HPLC. Acetonitrile and methanol were HPLC grade from Biosolve, Valkenswaard, the Netherlands. Dichloromethane was from Fisher Scientific, Landsmeer, the Netherlands. Formic acid and ammonium formate were from Sigma Aldrich, Zwijndrecht, the Netherlands. Deuterated DMSO (99.96 %) was from Eurisotop, Saint Aubin Cedex, France.

plant materialAnthriscus sylvestris L. Hoffm. (Apiaceae) plants at the flowering stage were collected in Groningen, the Netherlands (53o 11’ 34” N and 06o 37’ 04” E) in May 2007. Voucher specimen have been deposited in our department, encoded Asylv2007-4. The roots were separated from the aerial parts and the material was dried at room temperature after harvesting and prior to extraction.

Sample preparationThe extraction method was as described by Koulman et al.19 Shortly, 100 mg dried plant material were weighed into a Sovirel tube. A 2.0 mL portion of 80% methanol was added and the mixture was sonicated during 1 hour. Subsequently 4.0 mL of dichloromethane

74

iDen

tiFica

tion

oF lig

na

ns

3and 4.0 mL H2O were added. The mixture was vortexed and centrifuged at 1,000 g for 5 minutes. The aqueous layer was discarded and 2.0 mL of organic layer were transferred into a 2 mL Eppendorf tube. The organic layer was left in the fume hood until dried. The residue was redissolved in 2.0 mL methanol and filtered over a 0.45 µm HPLC syringe filter (nylon). The samples were submitted to HPLC analysis. For LC-MS/MS and LC-SPE-NMR analysis 1 g dried root material was used and the aliquot residue was reconstituted in methanol at a concentration corresponding to 1 g dried plant material.

HpLC-uV analysisThe HPLC analysis was as described by Vasilev et al.7 with some modifications. A Shimadzu-VP system was used, consisting of an LC-10AT pump, a Kontron 360 auto sampler, an SPD M10A DAD detector, an FCV-10AL low pressure gradient mixer, an SCL-10A system controller, an FIAtron system CH-30 column heater, operated with LC Solution software, version 1.2. The column used was Zorbax Eclipse C18 (150 x 4.6 mm, 5 µm), together with Phenomenex guard cartidge C18 (4 x 3 mm) Phenomenex, Bester, The Netherlands. The detection wavelength was 240 nm. Mobile phase A [H2O: ACN (95:5)] and B [ACN: H2O (95:5)], both in 0.1% formic acid and 2 mM ammonium formate. The injection volume was 5 µL with a flow rate of 1 mL/min using a gradient program of 30 min consisting of 1 min of B 30%, followed by a linear gradient until 15 min B 90%, at 20 min B 90%, at 25 min B 30% and at 30 min B 30%.

LC-eSi-mS/mS analysis HPLC was performed using a Shimadzu LC system, consisting of 2 LC-20AD gradient pumps and a SIL-20AC autosampler. The LC system was coupled to an API 3000 triple-quadrupole mass spectrometer (Applied Biosystems/MDS Sciex) via a TurboIonSpray source. Data were collected and analyzed by Analyst 1.4.2 data acquisition software (Applied Biosystems/MDS Sciex). Chromatographic separation was achieved on an Alltima C18 (Grace Davison) narrow-bore guard column (2.1 x 150 mm, 5 µm). The mobile phase and the gradient system were the same as used for the HPLC analysis. The flow rate was 0.2 mL/min. The injection volume was dependent on the expected analyte concentration. The ionization was performed by electrospray in the positive mode, which resulted in the formation of (M+H)+ and/or (M+NH4)

+ adduct ions. The source temperature was set to 450°C. The instrument was operated with an ionspray voltage of 5.2 kV. Nitrogen was used both for curtain gas and nebulizing gas. Full scan mass spectra were acquired at a scan rate of 1 scan/sec with a scan range of 100-1100 amu and a step size of 1 amu. Defined product ion scans were acquired in specified time windows.

LC-Spe isolationHPLC was applied for the isolation of the individual compounds in pure form. Mobile phase A consisted of 0.1% formic acid in H2O, mobile phase B consisted of 0.1% formic acid in acetonitrile. The following method with gradient was applied with a flow-rate of 1 mL/min: 5 min isocratic at 32% B, followed by a linear gradient until 15 min B

75

iDen

tiFica

tion

oF lig

na

ns

386%, quick increase in 0.1 min to 95% B, where kept constant for 5 min and followed by a quick descent in 1 min to 32% B, where kept for 5 min of equilibration prior to the next analysis. The purification was carried out on an Agilent 1200 system (Agilent Technology, SantaClara, USA) equipped with a Prospekt 2 (Spark Holland), using an Agilent column; Zorbax Eclipse XDB-C18, (150 x 4.6 mm, 5 µm). Compounds were trapped on general purpose (GP) cartridges (Spark Holland) and dried for at least 40 min under nitrogen air flow. Elution was performed using 600 mL of acetonitril, the purified fraction was dried under nitrogen air flow and dissolved in deuterated DMSO prior to NMR analyses.

nmr spectroscopyNMR analysis was performed on a Bruker Avance III 500MHz spectrometer equipped with a 5-mm CPTCI cryo-probe (1H-13C/15N/2H + Z-gradients) operating at 303 K. The structure identification of the purified compounds was based on a selection of 1D 1H, 1D selective NOE, 1H-1H-DQF-COSY, 1H-1H-TOCSY, 1H-13C-HSQC, 1H-13C-HMQC, and 1H-1H-NOESY spectra based on complexity or quantity of the sample. The proton and carbon chemical shifts were referenced to the internal reference TMS (proton, d = 0.00 ppm; carbon, d = 0.00 ppm). The data were processed using Topspin 2.1 pl5 and analysed with Mnova 6.2.1.

ACKnoWLedgementS

Financial support by the Ubbo Emmius Fund, University of Groningen, and the Müggenburg Foundation is greatly acknowledged.

referenCeS 1. Hulten E, Fries M. Atlas of north

European vascular plants north of the tropic of cancer. Konigstein: Koeltz Sci-entific Books, 1986.

2. Jeong G, Kwon O, Park B, et al. Lignans and coumarins from the roots of Anthr-iscus sylvestris and their increase of caspase-3 activity in HL-60 cells. Biol Pharm Bull 2007;30:1340-1343.

3. Magnusson, S.H. NOBANIS – Invasive alien species fact sheet – Anthriscus sylvesris – From: online database of the north European and Baltic network on Invasive alien species – NOBANIS. Available from: www.nobanis.org. 2006 (accessed 06.09.2011)

4. Koulman A, Bos R, Medarde M, et al. A fast and simple GC MS method for lignan profiling in Anthriscus sylvestris and bio-

synthetically related plant species. Planta Med 2001;67:858-862.

5. Ayres DC, Loike JD. Lignans: chemical, biological and clinical properties. Cambridge, UK: Cambridge University Press, 1990.

6. Nayar MP, Sastry APK. Botanical survey of India. Calcutta, 1990.

7. Vasilev NP, Julsing MK, Koulman A, et al. Bioconversion of deoxypodophyllo-toxin into epipodophyllotoxin in E. coli using human cytochrome P450 3A4. J Biotechnol 2006;126:383-393.

8. Kozawa M, Morita N, Hata K. Chemical components of the roots of Anthris-cus sylvestris HOFFM. I. Structures of an acyloxycarboxylic acid and a new phenylpropanoidester, anthriscusin. Yakugaku Zasshi 1978;98:1486-1490.

76

iDen

tiFica

tion

oF lig

na

ns

39. Kozawa M, Morita N, Hata K. Structure

of anthriscusin, a new phenylpropa-noid ester from the roots of Anthriscus sylvestris HOFFM. Chem Pharm Bull 1978;26:1337-1338.

10. Kurihara T, Kikuchi M, Suzuki S, et al. Studies on the constituents of Anthris-cus sylvestris HOFFM. I. On the com-ponents of the radix. Yakugaku Zasshi 1978;98:1586-1591.

11. Schmidt TJ, Hemmati S, Fuss E, et al. A combined HPLC-UV and HPLC-MS method for the identification of lignans and its application to the lignans of Linum usitatissimum L. and L. bienne Mill. Phytochem Anal 2006;17:299-311.

12. Ayres DC, Harris JA, Jenkins PN, et al. Lignans and related phenols. Part XII. Application of nuclear magnetic double resonance to the aryltetrahydronaphtha-lene class. J Chem Soc, Perkin Transac-tions 1 1972:1343-1350.

13. Miguel del Corral JM, Gordaliza M, López J-L, et al. Reassignment of the configuration of several keto-cyclolig-nans prepared from podophyllotoxin. Helv Chim Act 1995;78:1793-1796.

14. Pullockaran AJ, Kingston DGI, Lewis NG. Synthesis of stereospecifically deu-terated desoxypodophyllotoxins and 1H-NMR assignment of desoxypodo-phyllotoxin. J Nat Prod 1989;52:1290-1295.

15. Ikeda R, Nagao T, Okabe H, et al. Anti-proliferative constituents in Umbelliferae plants. III. Constituents in the root and the ground part of Anthriscus sylvestris HOFFM. Chem Pharm Bull 1998;46:871-874.

16. Kozawa M, Baba K, Matsuyama Y, et al. Components of the root of Anthriscus syl-vestris HOFFM. II. Insecticidal activity. Chem Pharm Bull 1982;30:2885-2888.

17. Nishimura K, Umehara K, Nakamura M. Cell differentation-inducing 1,3-di-oxobenzene derivatives from Anthriscus

sylvestris for cancer or condyloma. In: Koho JKT, JP 09194369. Japan, 1997.

18. Jackson DE, Dewick PM. Aryltetralin lignans from Podophyllum hexandrum and Podophyllum peltatum. Phytochem 1984;23:1147-1152.

19. Koulman A, Kubbinga ME, Batterman S, et al. A phytochemical study of lignans in whole plants and cell suspension cultures of Anthriscus sylvestris. Planta Med 2003;69:733-738.

20. Ikeda R, Nagao T, Okabe H, et al. Anti-proliferative constituents in Umbellif-erae plants. IV. Constituents in the fruits of Anthriscus sylvestris HOFFM. Chem Pharm Bull 1998;46:875-878.

21. Kamil WM, Dewick PM. Biosynthesis of the lignans α- and β-peltatin. Phytochem 1986;25:2089-2092.

22. Federolf K, Alfermann AW, Fuss E. Aryltetralin-lignan formation in two different cell suspension cultures of Linum album: Deoxypodophyllotoxin 6-hydroxylase, a key enzyme for the formation of 6-methoxypodophyllotox-in. Phytochem 2007;68:1397-1406.

23. Broomhead AJ, Rahman MMA, Dewick PM, et al. Matairesinol as precursor of Podophyllum Lignans. Phytochem 1991;30:1489-1492.

24. Sakakibara N, Suzuki S, Umezawa T, et al. Biosynthesis of yatein in Anthriscus syl-vestris. Org Biomol Chem 2003;1:2474-2485.

25. Kamil WM, Dewick PM. Biosynthetic re-lationship of aryltetralin lactone lignans to dibenzylbutyrolactone lignans. Phytochem 1986;25:2093-2102.

26. Chang FC, Chiang C, Aiyar VN. Isopicro-podophyllone from Podophyllum pleian-thum. Phytochem 1975;14:1440-1440.

27. Jackson DE, Dewick PM. Tumour-inhibitory aryltetralin lignans from Podophyllum pleianthum. Phytochem 1985;24:2407-2409.

ChapterIn vitro regeneration of wild chervil

(Anthriscus sylvestris L.)

Oktavia Hendrawati Jacques Hille

Herman J. Woerdenbag Wim J. Quax Oliver Kayser

In Vitro Cellular & Developmental Biology – Plant 2011: in press

4

80

In v

Itro reg

enera

tion

4

oktavia Hendrawati1, Jacques Hille2, Herman J. Woerdenbag3, Wim J. quax1, oliver Kayser1,4

1Department of Pharmaceutical Biology, University of Groningen, A. Deusinglaan 1, 9713 AV Groningen, The Netherlands; 2Department of Molecular Biology of Plants, University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands; 3Department of Pharmaceutical Technology and Biopharmacy, University of Groningen, A. Deusinglaan 1, 9713 AV Groningen, The Netherlands; 4Department of Technical Biochemistry, Technical University Dortmund, Emil-Figge-Strasse 66, 44227 Dortmund, Germany.

In Vitro Cellular & Developmental Biology – Plant 2011: in press

AbStrACt

Anthriscus sylvestris (L.) Hoffm. (Apiaceae) is a common wild plant that accumulates the lignan deoxypodophyllotoxin. Deoxypodophyllotoxin can be hydroxylated at the C-7 position in recombinant organisms yielding podophyllotoxin, which is used as a semi-synthetic precursor for the anticancer drugs, etoposide phosphate and teniposide. As in vitro regeneration of A. sylvestris has not yet been reported, development of a regeneration protocol for A. sylvestris would be useful as a micropropagation tool and for metabolic engineering of the plant. Calli were induced from hypocotyl explants and transferred to shoot induction medium containing zeatin riboside. Regenerated shoots were obtained within six months and were transferred onto growth regulator-free root induction medium containing 1% sucrose. Regenerated plants transferred to soil and acclimatized in a greenhouse. Plants were transferred to the field with a 100% survival rate. Regenerated plants flowered and were fully fertile. This is the first report of complete regeneration of A. sylvestris via shoot organogenesis from callus.

Keywords: organogenesis, micropropagation, regeneration, tissue culture, deoxypodophyllotoxin

81

In v

Itro reg

enera

tion

4introduCtion

Anthriscus sylvestris (L.) Hoffm. (Apiaceae) is a common wild plant, native to Europe, North America, Africa, Asia and New Zealand1-3 that accumulates the lignan deoxypodophyllotoxin, especially in the roots. The dried root of A. sylvestris is used in Korean and Chinese traditional medicine for the treatment of various diseases.4

Deoxypodophyllotoxin is structurally closely related to the lignan podophyllotoxin, which is obtained from Podophyllum species and is used for the synthesis of the anticancer drugs, etoposide phosphate and teniposide. Podophyllotoxin is obtained from Podophyllum species, but the supply of podophyllotoxin from this source is likely to become a major bottleneck because Podophyllum species have been listed on the endangered species list in India.5 An alternative and more sustainable source of podophyllotoxin may be obtained by the biotechnological hydroxylation of deoxypodophyllotoxin at the C-7 position. This conversion has already been achieved in E. coli DH5α transfected with the human cytochrome P450.6

Although gene transfer via Agrobacterium can facilitate efforts to engineer secondary metabolic pathways,7, 8 this approach has not yet been evaluated for A. sylvestris.

Plant regeneration from tissue cultures of many plant species has been reported, but some groups, families and genera are still regarded as recalcitrant.9 While there are no reports on regeneration of A. sylvestris, regeneration of the following members of the Apiaceae family has been documented: Daucus carota,10 Apium graveolens,11 Thapsia garganica,12 Pimpinella anisum,13 Ammi majus,14 Carum carvi,15 Coriandrum sativum,16 and Dorem ammoniacum.17

Regeneration from callus and suspension cultured cells of many crop plants, both by organogenesis18 and in vitro embryogenesis.19 Using information on the initiation of A. sylvestris calli and suspension cultures,20 cells from callus tissue were used to establish a regeneration protocol for A. sylvestris. Dormancy problems from A. sylvestris seeds presented an additional concern.21

The main objective of our study was to establish a regeneration protocol for A. sylvestris. This regeneration protocol can subsequently be used either as a micropropagation tool to bypass the dormancy in the seeds or for metabolic engineering of A. sylvestris. Different concentrations and combinations of auxins and cytokinins were used to induce shoots from callus tissue and to induce root formation from the regenerated shoots.

reSuLtS

Seed germination Following chlorine gas sterilization (figure 1a), the seeds did not germinate immediately or at the same time prior to cold stratification at 1oC. The sterilized seeds started to germinate after a 3-months cold stratification (figure 1b) with an average germination rate of 39%. In vitro sterile seedlings were successfully obtained for callus induction (figure 1c).

82

In v

Itro reg

enera

tion

4

Table 1. Effects of different concentrations of auxins and cytokinins for callus induction from roots, hypocotyls and leaves explants of A. sylvestris

Growth regulators (mg l-1) Callus induction (%) (x + SD)a

Auxins Cytokinins Explants2,4-D NAA BA Z Hypocotylsb Rootsc Leaves

2.0 1.0 36.1 + 28.3 28.5 + 35.7 1.33 + 1.162.0 1.0 30.5 + 19.7 61.0 + 39.3 nad

2.0 0.2 12.5 + 13.8 75.0 + 43.3 nad

aMean and standard deviationb,cMeans in this column do not significantly differ from each other (p < 0.05) dMean and standard deviation not available

Figure 1. Regeneration of A. sylvestris L. (Hoffm.). a. Ripe seeds, b. Germination of the sterile seeds after 3 months 1oC, c. In vitro young seedling, d. Leaves and hypocotyls explants, e. Callus growth after 2 months induction, f. Shoot regeneration in Z10 media, g. Further developed shoot regeneration, h. In vitro rooting of shoots, i. Young regenerated plantlets, j. Ex vitro acclimatization of young regenerated plantlets, k. Field planting of the regenerated plants, l. Flowering of the regenerated plants, m. Seed pods of regenerated A. sylvestris. Bars equal to 0.5 cm in all figures.

Root formation and acclimation

For complete plantlet development, regenerated shoots were rooted on plant

growth regulator-free Gamborg B5 medium supplemented with 10 g l-1 sucrose.

After 2 months, roots developed (figure 1h), and regenerated plants were obtained

(figure 1i). Twenty regenerated plants were transferred to small plastic pots for

acclimatization in a growth room for 2 weeks before transfer to bigger pots (figure

1j). After 3 months, the regenerated plants were planted in the field (figure 1k) in

June 2010 with a 100% survival rate.

Flowering and seed formation (R1)

After 2-4 months in the field, 55% of the regenerated plants (11 out of 20) flowered

(figure 1l), formed seed pods (figure 1m) and eventually shed seeds. The

phenotype of the regenerated plants was similar to wild type except for the flowers,

Figure 1. Regeneration of A. sylvestris L. (Hoffm.). a. Ripe seeds, b. Germination of the sterile seeds after 3 months 1oC, c. In vitro young seedling, d. Leaves and hypocotyls explants, e. Callus growth after 2 months induction, f. Shoot regeneration in Z10 media, g. Further developed shoot regeneration, h. In vitro rooting of shoots, i. Young regenerated plantlets, j. Ex vitro acclimatization of young regenerated plantlets, k. Field planting of the regenerated plants, l. Flowering of the regenerated plants, m. Seed pods of regenerated A. sylvestris. Bars equal to 0.5 cm in all figures.

Callus inductionCalli were induced from leaf, hypocotyl (figure 1d) and root explants from in vitro-grown seedlings. Several media were evaluated for callus induction in addition to the one reported previously for A. sylvestris20 (table 1). The best callus formation was obtained from roots with 75% induction using Gamborg's B5 medium supplemented

83

In v

Itro reg

enera

tion

4with 2.0 mg l-1 NAA and 0.2 mg l-1 Z (table 1). The best callus induction from hypocotyl explants was 36% using Gamborg's B5 medium supplemented with 2.0 mg l-1 2,4-D and 1.0 mg l-1 BA (table 1). Calli were formed, on average, after 2 months and 1-2 additional months were required to obtain sustained growth (figure 1e) before transferring to shoot induction media. Calli were either green, friable or compact.

Shoot formationIn general, media supplemented with Z yielded better callus quality, and their growth rate was 2-5.6-fold compared to media supplemented with Kin or BA (table 2).

The growth rate of calli in the shoot induction media with BA was 1.9-2.6, and the quality score was 1.3-3.2. Use of media containing Kin gave a similar growth rate between 1.7 and 2.5 but with better callus quality (1.0-3.9). However, none of these media led to shoot induction after 6 months. The quality of calli was improved using media with Z (from 3.0 up to 4.8). The best media (Z15 and Z16), yielding the fastest callus growth (4.7- and 5.6-fold) and the best callus quality (4.8 and 4.7), did not induce shoot formation within 6 months.

Among the media used, only Gamborg's B5 medium supplemented with 2.0 mg l-1 Z and 1.0 mg l-1 IBA was able to induce shoot formation (figure 1f-g). The shoot induction period required up to 6 months, but the number of regenerated shoots was relatively high (> 100) once shoot induction was obtained. Shoots were regenerated from 3 independent calli from hypocotyl explants, taken from different in vitro seedlings. Although shoot-like structures were observed after 4 months, some shoots seemed stressed (purple-whitish color) and died, but some of them survived and eventually regenerated shoots. The elongation of shoots took 2 months.

root formation and acclimationFor complete plantlet development, regenerated shoots were rooted on plant growth regulator-free Gamborg's B5 medium supplemented with 10 g l-1 sucrose. After 2 months, roots developed (figure 1h), and regenerated plants were obtained (figure 1i). Twenty regenerated plants were transferred to small plastic pots for acclimatization in a growth room for 2 weeks before transferring to bigger pots (figure 1j). After 3 months, the regenerated plants were planted in the field (figure 1k) in June 2010 with a 100% survival rate.

flowering and seed formation (r1)After 2-4 months in the field, 55% of the regenerated plants (11 out of 20) flowered (figure 1l), formed seed pods (figure 1m) and eventually shed seeds. The phenotype of the regenerated plants was similar to wild type except for the flowers, which showed a reduced apical dominance (figure 2). The regenerated plants flowered in the branches and at the basal part (figure 2c-f), whereas wild-type plants flower in the apical part (figure 2a-b). Seeds (R1) from the regenerated plants were collected, dried and placed at 1oC for germination. After a 2-3 months period of cold stratification, seeds from 4

84

In v

Itro reg

enera

tion

4Table 2. Effects of different concentrations of auxins and cytokinins for shoot regeneration of A. sylvestris

Growth regulators (mg l-1) Shoot formation

Growth ratea

(x + SD)cQuality of

callib

(x + SD)cAuxins Cytokinins

2,4-D NAA IAA IBA BA Z KinC 2.0 1.0 - 2.92 + 1.13 3.31 + 0.69S3 0.1 2.0 - 1.94 + 1.37 2.20 + 0.45S5 0.1 2.0 - 1.94 + 0.50 2.67 + 1.03S14 0.05 3.0 - 2.01 + 0.67 2.50 + 0.97S15 0.01 0.3 - 1.59 + 0.40 1.33 + 0.58S16 0.1 2.0 - 2.62 + 1.46 3.24 + 1.22S12 0.1 2.0 - 2.56 + 0.81 3.84 + 0.60*

S9 0.2 2.0 - 2.80 + 0.66 4.30 + 0.56*

S10 0.1 2.0 - 2.33 + 0.68 3.70 + 1.08S11 0.1 2.0 - 2.55 + 0.64 3.95 + 0.86*

S17 0.1 2.0 - 2.54 + 0.80 3.88 + 1.08S18 0.2 2.0 - 2.10 + 2.17 3.38 + 1.33S19 0.05 3.0 - 2.40 + 1.22 3.57 + 0.79S20 0.01 0.3 - 1.89 + 0.61 2.25 + 0.46S22 0.1 2.0 - 1.68 + 0.55 1.00 + 0.00S21 0.1 2.0 - 2.39 + 0.80 3.00 + 0.86Z3 0.1 2.0 - 2.71 + 0.99 3.00 + 0.97Z4 1.0 2.0 - 2.60 + 0.71 4.29 + 0.75*

Z5 0.05 2.0 - 2.18 + 0.60 2.42 + 0.84*

Z6 0.1 1.0 - 2.81 + 1.06 3.05 + 1.28Z7 0.5 1.0 - 3.01 + 1.05 4.25 + 0.68*

Z8 0.05 1.0 - 2.30 + 0.72 3.10 + 1.09Z10 1.0 2.0 + 4.39 + 1.11* 4.55 + 0.00*

Z11 0.05 2.0 - 3.03 + 0.88* 4.00 + 0.97*

Z12 0.1 1.0 - 3.13 + 0.57* 4.33 + 0.59*

Z13 0.5 1.0 - 3.67 + 1.82* 4.41 + 1.23*

Z14 0.05 1.0 - 3.16 + 1.10* 4.25 + 0.93*

Z15 1.0 2.0 - 4.72 + 1.46* 4.79 + 0.41*

Z16 1.0 2.0 - 5.58 + 1.15* 4.68 + 0.48*

a Growth rate : the weight of the calli after 4 weeks divided by the initial weight. The initial weight was 2 g per petri dish. b Quality of calli : score of the greenish appearance (5: all five calli were green and proliferating, 3: average, 1: all five calli were brown/dead). cmean and standard deviation. All data were recorded monthly for at least a 6 months period (n = 30). All media were supplemented with 40 g l-1 sucrose and 8 g l-1 agar. An asterisk (*) shows a significant difference (p < 0.05) compared to control (C).

85

In v

Itro reg

enera

tion

4

regenerated plants germinated. Twenty regenerated seedlings (R1) of each regenerated plant (R0) were grown in the greenhouse.

deoxypodophyllotoxin contentThe DPT content of the aerial part of the regenerated plants was between 0.6-1.2 µg/mg dry weight (d.w.). In the root part of the regenerated plant, the DPT content was 0.3-0.7 µg/mg d.w. The DPT content of the aerial part of the wild type was 0.3-0.5 µg/mg d.w.; in the root part 1.4-1.5 µg/mg d.w.4 There was no difference observed on the profile of metabolites of the regenerated plants compared to the wild type.

diSCuSSion

In general, 3 phases can be distinguished in regeneration (organogenesis) in vitro. The first is dedifferentiation in which the tissue becomes competent to respond to the organogenic stimulus. It is generally initiated by culturing on an auxin-rich callus-inducing medium and may involve a period of callus growth. The second is induction in which cells become determined to form either a root or a shoot. The explants are cultured on a shoot-inducing medium or root-inducing medium that contains a specific auxin/cytokinin ratio combination. The third is the realization in which the explants grow to an organ.9, 22

Initially, seeds of Anthriscus sylvestris (figure 1a) were dormant and have underdeveloped embryos. The embryos of the seeds grow to maturity during cold stratification. Germination frequency increased with the length of the cold stratification period. Approximately 30% of the non-sterilized seeds germinated after 8 weeks of cold stratification at 1oC and then warm stratification at 15oC/6oC or 20oC/10oC or 25oC/15oC.21 The germination increased to 77% with the increased duration of the cold stratification (12 weeks). Most A. sylvestris embryos completed growth during a 12-weeks period of cold stratification at 1oC.21

which showed a reduced apical dominance (figure 2). The regenerated plants

flowered in the branches and at the basal part (figure 2c-f), whereas wild-type

plants flower in the apical part (figure 2a-b). Seeds (R1) from the regenerated

plants were collected, dried and placed at 1oC for germination. After a 2-3 months

period of cold stratification, seeds from 4 regenerated plants germinated. Twenty

regenerated seedlings (R1) of each regenerated plant (R0) were grown in the

greenhouse.

Figure 2. Flowering of A. sylvestris. a. Wild type b. Wild type (no flower at the bottom part). c-d. Reduced apical dominance flowering in the branches of the regenerated plants. e-f. flowering in the bottom part of the regenerated plants. Arrows show the differences. Bars equal to 10 cm (a-b) and 1 cm (c-f). Deoxypodophyllotoxin content

The DPT content of the aerial part of the regenerated plants was between 0.6-1.2

µg/mg dry weight (d.w.). In the root part of the regenerated plant, the DPT content

was 0.3-0.7 µg/mg d.w. The DPT content of the aerial part of the wild type was 0.3-

0.5 µg/mg d.w.; in the root part 1.4-1.5 µg/mg d.w.4 There was no difference

observed on the profile of metabolites of the regenerated plants compared to the

wild type.

Discussion

In general, 3 phases can be distinguished in regeneration (organogenesis) in vitro.

The first is dedifferentiation in which the tissue becomes competent to respond to

the organogenic stimulus. It is generally initiated by culturing on an auxin-rich

callus-inducing medium and may involve a period of callus growth. The second is

induction in which cells become determined to form either a root or a shoot. The

explants are cultured on a shoot-inducing medium or root-inducing medium that

Figure 2. Flowering of A. sylvestris. a. Wild type b. Wild type (no flower at the bottom part). c-d. Reduced apical dominance flowering in the branches of the regenerated plants. e-f. Flowering in the bottom part of the regenerated plants. Arrows show the differences. Bars equal to 10 cm (a-b) and 1 cm (c-f).

86

In v

Itro reg

enera

tion

4The surface sterilization treatment with chlorine gas yielded on average 39%

germination, which is approximately half the germination obtained from the non-surface sterilized seeds.21

Depending on the plant species, only a limited number of cells in an explant show the organogenic response.22 Therefore, different explants from hypocotyls, leaves and roots from 1-2 months young seedling were used for callus induction. The root explants showed the best callus induction (75%) using Gamborg's B5 medium supplemented with 0.2 mg l-1 Z and 2.0 mg l-1 NAA. However, use of seedlings to obtain root explants was not practical because 3 months were needed to obtain in vitro seedlings. Hypocotyl explants were preferred because they are renewable, and the 36% callus induction was sufficient for our purpose. The callus medium contained Gamborg's B5 salts and vitamins, supplemented with 2.0 mg l-1 2,4-D and 1.0 mg l-1 BA (table 1). Callus induction from leaf explants was low compared to hypocotyl and root explants. Callus tissue was formed on the cut sections after 2 months, but a further 1-2 months was required to obtain consistent growth of green friable or compact calli.

There are differences in the organogenetic potential between plant families, genera, species and genotypes, and different genotypes of a species may show widely different responses.23 For example, Capsicum species that belong to the Solanaceae have been shown to be recalcitrant to differentiation and plant regeneration under in vitro conditions compared to Nicotiana tabacum, Lycopersicum esculentum and Solanum tuberosum, which belong to the same family.24 Although plant regeneration, somatic embyrogenesis and genetic transformation have been reported for several Apiaceae family members, tissue culture regeneration systems for different genotypes or species may not generally be applicable to all members of the same family.

Among the 28 different media used, only Gamborg's B5 medium supplemented with 2.0 mg l-1 Z and 1.0 mg -1 IBA led to shoot formation after 6 months (table 2). Among 3 different cytokinins, use of Z led to a better callus quality (score 4.8) and growth rate (up to 5.6). We obtained 3 independent regenerants from different calluses from hypocotyls of different in vitro seedlings.

The regenerated plants were planted in the field in the beginning of June 2010, with a 100% survival rate. In the Netherlands, the flowering season is between April and May. Interestingly, 55% of the regenerated plants flowered between August and October in their first year (figure 2).

Daucus carota and A. sylvestris are biennual plants from the Apiaceae family that flower after exposure to a period of cold (vernalization), generally during the second year of growth.21, 25 Carrot roots, transformed by A. rhizogenes, converted from biennial to annual without vernalization.26, 27 A similar phenomenon was observed in endive plants (Cichorium intybus) containing Ri T-DNA.28 Natural genetic transformation by A. rhizogenes might provide genes that allow flowering in the absence of a cold treatment.26 Annualism was correlated with the segregation of a truncated transferred DNA (T-DNA) insertion. Root locus (rolC) was the primary promoter of annualism.26 Expressin of rolC in transgenic plants mainly leads to a modification of the cytokinin

87

In v

Itro reg

enera

tion

4balance and causes developmental, physiological and morphological alterations.29 In A. sylvestris regenerants, the observed annualism may be caused by the effects of tissue culture as reported for Chichorium intybus28 and long-term growth in shoot induction media containing high a cytokinin concentration.

The regenerated plants in the field showed wild-type phenotypes except that they flowered on the secondary branches (figure 2c-d) and at the base of plants (figure 2e-f), as opposed to the apical flowering of the wild-type plants (figure 2a-b). This phenomenon may be caused by somaclonal variation as a result of tissue culture30 and/or cytokinin imbalance from the long-term growth in shoot induction media. R1 seeds were germinated in February 2011 and grown in the field. Recalcitrance and somaclonal variation are the 2 main problems in plant regeneration via tissue culture. They are complex, multifactorial phenomena based on genotype, media and environmental interaction.30

Calli of A. sylvestris did not produce DPT20 and whole plants were required. Therefore, it is essential to achieve the differentiated form to restore the capability of the cell to produce DPT for metabolic engineering purposes. The DPT content of the aerial part of the regenerated plant was comparable with the root part of the wild-type plants. This is beneficial for the isolation of the bioconversion product for the metabolic engineering purpose. The ability of the regenerated plant to produce secondary metabolites, especially DPT, was not altered.

In conclusion, this is the first report on the regeneration of A. sylvestris plants in vitro. This may provide a valuable tool in creating new opportunities to produce metabolically engineered A. sylvestris.31 Further optimization is needed to increase the efficiency and frequency of regeneration and to reduce the total regeneration period.

mAteriAL And metHodS

plant materialThe fruits (ripe mericarps, hereafter called seeds) of Anthriscus sylvestris L. (Hoffm.) were collected from wild habitat in June 2007 in Groningen, the Netherlands (53o 13’ 34” N and 6o 32’ 43” E). The seeds were surface sterilized with chlorine gas for 3.5 h and subsequently cultured on 25 mL 10 g l-1 water agar (Duchefa, Haarlem, the Netherlands) in Petri dishes (94/16, Greiner Bio One, Alphen aan de Rijn, the Netherlands). To overcome dormancy, the seeds were kept in the Petri dishes at 1oC under fluorescent light illumination for for at least 3 months until the seeds germinated. The germinated seedlings were transferred to Gamborg's B5 medium32 supplemented with 8 g l-1 agar in a sterile container, OS 140 box and ODS filter (Duchefa, Haarlem, the Netherlands) and incubated in a climate room at 22oC/20oC under a 16-h light/8-h dark period (day/night) (60 μmol m-2 s-1 fluorescent and incandescent light).

88

In v

Itro reg

enera

tion

4Callus inductionAfter 2 months, root, hypocotyl and leaf explants from young sterile in vitro-obtained seedlings were excised and cultured on Gamborg's B5 media supplemented with either 2.0 mg l-1 2,4-dichlorophenoxyacetic acid (2,4-D) and 1.0 mg l-1 6-benzylaminopurine (BA)20 or 2 mg l-1 α-naphthaleneacetic acid (NAA) and 0.2 or 1.0 mg l-1 zeatin riboside (Z) (table 1). All media contained 40 g l-1 sucrose and 8 g l-1 agar (pH 5.7). The calli induced from various explants were subcultured on the same medium every 4 weeks.

plant regeneration from callus culturesAfter 3 months, calli were cut into 400 mg pieces and transferred to Gamborg's B5 medium containing various concentrations of auxins such as 2,4-D, NAA, IBA (Indole-3-butyric acid), Indole-3-acetic acid (IAA), and cytokinins such as BA, Z, kinetin (Kin) to induce shoot formation (table 2). Each treatment consisted of at least 5 Petri dishes with 5 calli in each dish. The growth rate and the quality of the calli of each treatment were recorded monthly for at least 6 months. The growth rate was defined as the ratio of the weight of the calli after 4 weeks divided by the initial weight. The quality of calli was given a score from 1 to 5 (5: all 5 calli were green and proliferating, 3: average, 1: all 5 calli were brown/dead). The score in table 2 was the average of at least 6 months data for both the growth rate and the quality of the calli.

rooting and acclimatizationRegenerated shoots measuring about 2 cm in length were transferred into rooting medium. The rooting medium contained Gamborg's B5 salts and vitamins supplemented with 10 g l-1 sucrose and no growth regulators. After 2 months, the rooted plantlets were thoroughly washed with tap water to remove the agar and transferred to sterile soil in small pots and covered with plastic to maintain high humidity. Plantlets were placed in a growth chamber with a 14-h light/10-h dark period, photosynthetic photon flux density 100 μmol m−2  s−1 at 22°C. After 2 weeks, the humidity was reduced by removing the plastic, and 20 plants were transferred to bigger pots and kept for another 2-3 months. Eventually, the plants were transferred to the field. The regeneration experiment was conducted with 3 replicates from 3 different independently obtained calli.

plant extraction and HpLC analysisDeoxypodophyllotoxin (DPT) was a gift from Dr. M. Angeles Castro (Salamanca University, Salamanca, Spain). Acetonitrile and methanol were HPLC grade from Biosolve (Valkenswaard, the Netherlands). Dichloromethane was obtained from Fisher Scientific (Landsmeer, the Netherlands). Samples for HPLC analysis were taken from 10 randomly selected individual young regenerated plantlets, which were divided into an aerial part and a root portion. Plant material was dried at room temperature prior to grinding and extraction. The extraction method was as described by Koulman et al.20 and the HPLC analysis was as described by Hendrawati et al.4

89

In v

Itro reg

enera

tion

4Statistical analysisThe Student’s t-test (p < 0.05) was used to determine statistical significance compared to the control (SPSS software, version 16.0, 2008).

ACKnoWLedgementS

The authors would like to express their gratitude to Margriet Ferwerda for her advice concerning the plant tissue culture, J.T.M. Elzenga, Ger Telkamp, and Jacob J. Hogendorf for the climate room and the field facility. Financial support by the Ubbo Emmius fund, University of Groningen, is acknowledged.

referenCeS1. Jeong G, Kwon O, Park B, et al. Lignans

and coumarins from the roots of Anthr-iscus sylvestris and their increase of caspase-3 activity in HL-60 cells. Biol Pharm Bull 2007;30:1340-1343.

2. Magnusson SH. NOBANIS - Invasive alien species fact sheet - Anthriscus syl-vestris – From: online database of the north European and Baltic network on invasive alien species. Available from: www.nobanis.org. 2006 (accessed 06.09.2011).

3. Hulten E, Fries M. Atlas of north European vascular plants north of the tropic of cancer.Koeltz Scientific Books, Konigstein. Germany, 1986.

4. Hendrawati O, Woerdenbag J, Hille J, et al. Seasonal variations in the deoxypodo-phyllotoxin content and yield of Anthris-cus sylvestris L. (Hoffm.) grown in the field and under controlled conditions. J Agri Food Chem 2011;59:8132-8139.

5. Nayar MP, Sastry APK. Botanical Survey of India. Calcutta, 1990.

6. Vasilev NP, Julsing MK, Koulman A, et al. Bioconversion of deoxypodophyllo-toxin into epipodophyllotoxin in E. coli using human cytochrome P450 3A4. J Biotech 2006;126:383-393.

7. Birch RG. Plant transformation: problems and strategies for practical ap-plication. Annu Rev Plant Physiol Plant Mol Biol 1997;48:297-326.

8. Gómez-Galera S, Pelacho A, Gené A, et al. The genetic manipulation of medicinal and aromatic plants. Plant Cell Rep 2007;26:1689-1715.

9. Ochatt SJ, Atif RM, Patat-Ochatt EM, et al. Competence versus recalcitrance for in vitro regeneration. Notulae Bot Hort Agrobot Cluj-Napoca 2010;38:102-108.

10. Steward FC, Mapes MO, Mears K. Growth and organized development of cultured cells. II. Organization in cultures grown from freely suspended cells. Am J Bot 1958;45:705-707.

11. Catlin D, Ochoa O, McCormick S, et al. Celery transformation by Agrobacterium tumefaciens - cytological and genetic analysis of transgenic plants. Plant Cell Rep 1988;7:100-103.

12. Jäger A, Schottländer B, Smitt U, et al. Somatic embryogenesis in cell cultures of Thapsia garganica. Plant Cell Rep 1993;12:517-520.

13. Salem KMSA, Charlwood BV. Accumu-lation of essential oils by Agrobacterium tumefaciens-transformed shoot cultures of Pimpinella anisum. Plant Cell Tiss Org Cult 1995;40:209-215.

14. Purohit M, Pande D, Datta A, et al. Enhanced xanthotoxin content in re-generating cultures of Ammi majus and micropropagation. Planta Med 1995;61:481,482.

15. Krens FA, Keizer LCP, Capel IEM. Trans-genic caraway, Carum carvi L.: a model species for metabolic engineering. Plant Cell Rep 1997;17:39-43.

16. Wang Y, Kumar PP. Heterologous expres-sion of Arabidopsis ERS1causes delayed senescence in coriander. Plant Cell Rep 2004;22:678-683.

17. Irvani N, Solouki M, Omidi M, et al. Callus induction and plant regeneration

90

In v

Itro reg

enera

tion

4in Dorem ammoniacum D., an endan-gered medicinal plant. Plant Cell Tiss Org Cult 2010;100:293-299.

18. Flick CE, Evans DA, Sharp WR. Or-ganogenesis. In: Evans DA, Sharp, WR, Ammirato PV, Yamada Y, ed. Handbook of plant cell culture. Volume 1. New York: Macmillan Publishing Co., 1983:13-81.

19. Brown DCW, Finstad KI, Watson EM. Somatic embryogenesis in herbaceous dicots. In: Thorpe TA, ed. In vitro em-bryogenesis in plants. Volume 20. Dordrecht: Kluwer Academic Publishers, The Netherlands 1995:345-416.

20. Koulman A, Kubbinga ME, Batterman S, et al. A phytochemical study of lignans in whole plants and cell suspension cultures of Anthriscus sylvestris. Planta Med 2003;69:733,738.

21. Baskin CC, Milberg P, Andersson L, et al. Deep complex morphophysiological dormancy in seeds of Anthriscus sylves-tris (Apiaceae). Flora 2000;195:245-251.

22. De Klerk G, Arnholdt-Schmitt B, Lieberei R, et al. Regeneration of roots, shoots and embryos:physiological, bio-chemical and molecular aspects. Biol Planta 1997;39:53-66.

23. George EF. Plant propagation by tissue culture: part 2 - in practice. Basingstoke: Exegetics, England,1996.

24. Ochoa-Alejo N, Ramirez-Malagon R. In vitro chili pepper biotechnology. In Vitro Cell Dev Biol - Plant 2001;37:701-729.

25. Punja ZK, Jayaraj J, Wally O. Carrot. In: Pua E, Davey MR, ed. Transgenic crops IV. Volume 59: Springer Berlin Heidel-berg, 2007:277-294.

26. Limami MA, Sun LY, Douat C, et al. Natural genetic transformation by Agro-bacterium rhizogenes. Annual flowering in two biennials, Belgian endive and carrot. Plant Physiol 1998;118:543-550.

27. Tepfer D. Transformation of several species of higher plants by Agrobacteri-um rhizogenes: sexual transmission of the transformed genotype and phenotype. Cell 1984;37:959-967.

28. Sun LY, Touraud G, Charbonnier C, et al. Modification of phenotype in Belgian endive (Cichorium intybus) through genetic transformation by Agrobac-terium rhizogenes: conversion from biennial to annual flowering. Transgen Res 1991;1:14-22.

29. Estruch JJ, Chriqui D, Grossmann K, et al. The plant oncogene rolC is re-sponsible for the release of cytokinins from glucoside conjugates. EMBO J 1991;10:2889-95.

30. Cassells A, Curry R. Oxidative stress and physiological, epigenetic and genetic variability in plant tissue culture: impli-cations for micropropagators and genetic engineers. Plant Cell Tiss Org Cult 2001;64:145-157.

31. Hendrawati O, Woerdenbag HJ, Hille J, et al. Metabolic engineering strategies for the optimization of medicinal and aromatic plants: realities and expecta-tions. J Med Spice Plants 2010;15:111-126.

32. Gamborg OL, Miller RA, Ojima K. Nutrient requirements of suspension cultures of soybean root cells. Exp Cell Res 1968;50:151-158.

ChapterAgrobacterium mediated transformation of

Anthriscus sylvestris with human cytochrome P450 3A4 followed by regeneration

Oktavia Hendrawati Jacques Hille

Herman J. Woerdenbag Heribert Warzecha

Frank J. Dekker Wim J. Quax Oliver Kayser

(Manuscript in preparation)

5

94

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5

oktavia Hendrawati1, Jacques Hille2, Herman J. Woerdenbag3, Heribert Warzecha4, frank J.

dekker5, Wim J. quax,1, oliver Kayser1,6

1Department of Pharmaceutical Biology, University of Groningen, Antonius Deusinglaan 1, 9713 AV, Groningen, The Netherlands; 2 Department of Molecular Biology of Plants, University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands; 3Department of Pharmaceutical Technology and Biopharmacy, University of Groningen Antonius Deusinglaan 1, 9713 AV, Groningen, The Netherlands; 4Department of Plant Biotechnology and Metabolic Engineering, Technische Universitaet Darmstadt, Schnittspahnstrasse 3-5, 64287 Darmstadt, Germany; 5Department of Pharmaceutical Gene Modulation, University of Groningen, A. Deusinglaan 1, 9713 AV Groningen, The Netherlands; 6Technical University Dortmund, Technical Biochemistry, Emil-Figge-Strasse 66, 44227 Dortmund, Germany

(Manuscript in preparation)

AbStrACt

Metabolic engineering offers interesting perspectives to improve the productivity of the plant as a cell factory. The recently demonstrated conversion of deoxypodophyllotoxin into epipodophyllotoxin upon expression of human P450 3A4 in E. coli DH5a, has prompted us to investigate the activity of P450 3A4 in planta. Because of the endogenous production of deoxypodophyllotoxin in A. sylvestris, transformation of this plant may result in the production of (epi)podophyllotoxin. A. sylvestris callus tissue was successfully transformed with Agrobacterium tumefaciens carrying the human P450 3A4 gene and the cells were subsequently regenerated. The presence of the P450 3A4 gene in the regenerated transgenic A. sylvestris was confirmed by PCR. The transgenic plants were shown to form epipodophyllotoxin in low amounts whereas the wild type (control) plants do not. This opens the route for a novel plant source for the semi-synthesis of the important anticancer drugs etoposide and teniposide.

Keywords: Anthriscus sylvestris, epipodophyllotoxin, podophyllotoxin, deoxypodophyllotoxin, organogenesis, transformation, regeneration, human P450 3A4, CYP3A4

95

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5introduCtion

Plants are a rich source of bioactive compounds. Compounds of plant origin can be used as drugs or as precursors of semisynthetic drugs, and may provide valuable leads for novel drug design. The amounts of secondary metabolites in plants are often low and the availability can often be a major bottleneck in furnishing the pharmaceutical needs. Chemists have in specific instances succeeded to synthesize typical plant compounds via organic synthesis, but often this is hampered by the high chemical complexity, specific stereochemistry and the economic feasibility. Metabolic engineering of plants offers interesting perspectives to improve the productivity of the plant as a cell factory. This approach may create new opportunities in agriculture, environmental applications, production of chemicals, and medicines.1, 2

Podophyllotoxin (PPT), obtained from Podophyllum species, is used for the semisynthesis of the anticancer drugs etoposide and teniposide. PPT and its derivatives are now intensively investigated for their use in anticancer therapy especially in combination treatment.3 Up to now, three different ways are known for the production of podophyllotoxin: direct isolation from Podophyllum species, via plant cell cultures, and by organic synthesis. Using organic synthesis or plant cell cultures is economically unfeasible. To date, PPT is only obtained by isolation from Podophyllum species. In the future, the availability of PPT from this source is likely to become a major bottleneck, as Podophyllum species have been listed on the endangered species list.4 An alternative and more sustainable source of PPT may be obtained by (biotechnological) hydroxylation of deoxypodophyllotoxin (DPT) at the C7 position (figure 1) yielding epipodophyllotoxin (epi-PPT), the diastereoisomer of PPT. This conversion has been achieved already on laboratory scale in E. coli DH5a transfected with the human cytochrome P450 3A4.5 Both epi-PPT and PPT can be used as a semisynthetic precursor for etoposide and teniposide.

In many plants, DPT serves as the precursor of PPT.6, 7 DPT is the main constituent of Anthriscus sylvestris (L.) Hoffm. (Apiaceae). A. sylvestris accumulates up to 1.5 µg/mg dried weight (d.w.) of DPT in the root part8 and trace amounts of PPT (<0.01 µg/mg d.w) in the aerial part.9 As a wild plant and generally occurring weed native to northwest Europe, the availability of A. sylvestris is hardly limited, in contrast to Podophyllum species.

The human P450 3A4 gene has been used to create different recombinant systems such as in the yeasts Schizosaccharomyces pombe10 and Saccharomyces cerevisiae,11 in the bacterium E. coli,5, 12 and in the insect cells Sf21 Spodoptera frugiferda.13 The transformation of human P450 3A4 has been achieved in plant cells, but so far only in Nicotiana tabacum.14, 15 For functional activity of the 3A4 P450 enzyme a P450 reductase has to be expressed from endogenous or exogenously added DNA.

The main objective of this study was to transform the human P450 3A4 gene in A. sylvestris plants and to determine the capacity of this system to in vivo transform deoxypodophyllotoxin into epipodophyllotoxin (see figure 1).

96

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5

reSuLtS

feeding of dpt to Cyp3A4 expressing n. tabacum cellsN. tabacum cells were transformed with the human full length CYP3A4 cDNA as described15 and the results of the feeding of DPT to transformed and to wild type cells are presented in table 1. The transformed cells (CYP3A4) as well as the wild type (control) cells were able to hydroxylate DPT into epi-PPT, the diastereoisomer of PPT.

Figure 1. Scheme of Anthriscus sylvestris transformed with human P450 3A4. The expression of P450 3A4 is expected to result in the hydroxylation of deoxypodophyllotoxin (DPT) into epipodophyllotoxin (epi-PPT) based on the observed activity of E. coli produced 3A4 P450. The formation of its diastereoisomer, podophyllotoxin (PPT), is not expected to be catalysed by 3A4.

Table 1. DPT and Epi-PPT concentrations, means are given + SD (n=3), (µg/mg d.w. in cells and µg/ml in media) in tobacco plant cell suspension cultures of wild type (WT) and CYP3A4 (3A4) at different time points after feeding with DPT.

means + SD*Day 4 Day 7

WT DPT cells (µg/mg d.w.) 0.102 + 0.030 0.103* + 0.0263A4-DPT in cells (µg/mg d.w.) 0.077 + 0.002 0.035*+ 0.006WT DPT media (µg/ml) 1.316 + 0.178 1.302 + 0.1383A4-DPT in media (µg/ml) 0.721 + 0.124 0.598 + 0.144WT epi-PPT cells (µg/mg d.w.) 0.029 + 0.008 0.035 + 0.0343A4-Epi-PPT in cells (µg/mg d.w.) 0.038 + 0.020 0.102 + 0.017WT epi-PPT media (µg/ml) 1.062 + 0.109 2.940* + 0.4553A4-Epi-PPT in media (µg/ml) 1.843 + 0.244 4.748* + 0.270

An asterisk (*) indicates a significant difference between wild type and 3A4 (p<0.05)

97

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5N. tabacum does not contain the DPT. Therefore the epi-PPT formation was due to the addition of DPT. On day 7, the conversion of DPT into epi-PPT in the medium was significantly higher in transformed cells as compared to wild type cells.

In general, the concentration of epi-PPT was higher in the media than in the cells for both wild type and transgenic cells. Although the wild type was able to form epi-PPT, the conversion was 3 times lower than in the transformed cells and 1.5 fold lower in the media of the transformed cells on day 7 compared to the media of the wild type cells. In the transformed cells, the highest epi-PPT production was seen on day 7, in the cells (0.1 µg/mg d.w.) and in the media (4.8 µg/ml). The high concentration of the product in the medium as compared to the intracellular concentration suggests that there is an efficient secretion mechanism for the epi-PPT formed.

plant transformation and regeneration of A. sylvestrisWe initiated callus cells from young sterile in vitro seedlings (figure 2a-b). The calli were obtained by aseptically excising the hypocotyls of in vitro seedlings (figure 2c). The calli were transformed by A. tumefaciens harboring the binary plasmid pGreen029-3A4 (figure 3). YFP was fused to the P450 3A4 as an additional selection marker for visualization besides the kanamycin resistant gene. After the co-cultivation for 2 days, the calli were subcultured on shoot-inducing medium containing kanamycin for transformant selection and carbenicillin for elimination of overgrowth Agrobacterium. The transformed calli will survive on antibiotics selection. Every 4 weeks, the calli were

within 3 months while wild type would die. The non-transformed calli eventually

died and the transformed calli grew well (figure 2d). About a hundred transformed

calli clumps were subcultured to shoot-inducing medium (figure 2e) for up to 1 year

and the regenerated shoots (figure 2f) were subcultured in the root-inducing media

(figure 2g) for 3 months. Eventually the transformed plants were transferred into

the greenhouse (figure 2h) and grown under controlled conditions.8 After two

independent Agrobacterium-mediated transformations on callus cells, five

kanamycin-resistant plantlets from two independent transformations were

analyzed.

Figure 2. Anthriscus sylvestris transformation and regeneration. (a) Seeds of A. sylvestris. (b) Young sterile in vitro seedling. (c) Callus induction by aseptically excised hypocotyls of the young sterile in vitro seedlings. (d) Transformed callus selection on kanamycin and carbenicillin callus medium. (e) Shoot formation of transformed calli grown in shoot-inducing medium. (f) Elongation of shoot formation. (g) Root formation in shoot-inducing medium. (h) Young transformed plantlets grown in the green house. Bars equal to 1 cm (a-d, f, g), 0.5 cm (e) and 10 cm (h).

Figure 2. Anthriscus syl-vestris transformation and regeneration. (a) Seeds of A. sylvestris. (b) Young sterile in vitro seedling. (c) Callus induction by aseptically excised hypocotyls of the young sterile in vitro seedlings. (d) Transformed callus selection on kanamycin and carbenicillin callus medium. (e) Shoot formation of trans-formed calli grown in shoot-inducing medium. (f) Elon-gation of shoot formation. (g) Root formation in shoot-inducing medium. (h) Young transformed plantlets grown in the green house. Bars equal to 1 cm (a-d, f, g), 0.5 cm (e) and 10 cm (h).

98

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5subcultured to fresh medium. The selection on kanamycin required 3 months and the elimination of Agrobacterium required up to 6 months. Kanamycin 100 mg/L was sufficient to kill the non-transformed cells within 3 months while wild type would die. The non-transformed calli eventually died and the transformed calli grew well (figure 2d). About a hundred transformed calli clumps were subcultured to shoot-inducing medium (figure 2e) for up to 1 year and the regenerated shoots (figure 2f) were subcultured in the root-inducing media (figure 2g) for 3 months. Eventually the transformed plants were transferred into the greenhouse (figure 2h) and grown under controlled conditions.8 After two independent Agrobacterium-mediated transformations on callus cells, five kanamycin-resistant plantlets from two independent transformations were analyzed.

Figure 3. Schematic map of plasmid construct used for Anthriscus sylvestris transformation; LB: left border, D 35 S: double CaMV 35S promoter, YFP: yellow fluorescence protein fused with human P450 3A4 gene, NOS: nos polyadenylation signal (black), NOS: nos promoter (orange), nptII: kanamycin resistance gene.

Figure 4. (a) PCR-amplified human P450 3A4 gene of 561 bp. (b) PCR-amplified kanamycin gene of 687 bp. (c) PCR-amplified YFP gene of 739 bp from two wild type plants (WT) and five transformed plants (Lane 1 – 5). M= molecular weight marker.

We tested that the transformed plants were free from Agrobacterium by growing

the calli and the transformed plants in LB medium at 28oC for 3 days. No grow of

Agrobacterium was observed. The presence of the P450 3A4, YFP and kanamycin

gene in the transformed plants was confirmed by PCR (figure 4). The gene

fragment could be amplified from the aerial part of the transformed Anthriscus,

whereas no band appeared from the wild type aerial part of Anthriscus (negative

control), confirming that the transformed plants could be considered as transgenic

plants. The YFP expression of transformed plants was low and the differences

compared to the wild type were not clear (not shown).

Epi-PPT content in the transformed plants

DPT, PPT or epi-PPT contents in the transformed and wild type plants are

summarized in table 2. Figure 5 shows a typical HPLC chromatogram of A.

Figure 3. Schematic map of plasmid construct used for Anthriscus sylvestris transformation; LB: left border, D 35 S: double CaMV 35S promoter, YFP: yellow fluorescence protein fused with human P450 3A4 gene, NOS: nos polyadenylation signal (black), NOS: nos promoter (orange), nptII: kanamycin resistance gene.

Figure 4. (a) PCR-amplified human P450 3A4 gene of 561 bp. (b) PCR-amplified kanamycin gene of 687 bp. (c) PCR-amplified YFP gene of 739 bp from two wild type plants (WT) and five transformed plants (Lane 1 – 5). M= molecular weight marker.

We tested that the transformed plants were free from Agrobacterium by growing the calli and the transformed plants in LB medium at 28oC for 3 days. No grow of Agrobacterium was observed. The presence of the P450 3A4, YFP and kanamycin gene in the transformed plants was confirmed by PCR (figure 4). The gene fragment could be amplified from the aerial part of the transformed Anthriscus, whereas no band appeared from the wild type aerial part of Anthriscus (negative control), confirming that the transformed plants could be considered as transgenic plants. The YFP expression of transformed plants was low and the differences compared to the wild type were not clear (not shown).

99

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5epi-ppt content in the transformed plants DPT, PPT or epi-PPT contents in the transformed and wild type plants are summarized in table 2. Figure 5 shows a typical HPLC chromatogram of A. sylvestris extracts. The enhanced formation of epi-PPT in transgenic A. sylvestris is clearly seen at a retention time of 7.65 min. The transformed plants were able to form epi-PPT and PPT. Wild type Anthriscus contained trace amounts of PPT in concert with earlier reports9 but no epi-PPT was formed. The PPT content in the transformed plants was slightly higher than in the wild type (areal part). No epi-PPT could be detected in the wild Anthriscus plant (figure 5). No obvious changes could so far be observed in the phenotype of the in vitro transformed regenerated plants.

sylvestris extracts. The enhanced formation of epi-PPT in transgenic A. sylvestris is

clearly seen at a retention time of 7.65 min. The transformed plants were able to

form epi-PPT and PPT. Wild type Anthriscus contained trace amounts of PPT in

concert with earlier reports9 but no epi-PPT was formed. The PPT content in the

transformed plants was slightly higher than in the wild type (areal part). No epi-PPT

could be detected in the wild Anthriscus plant (figure 5). No obvious changes could

so far be observed in the phenotype of the in vitro transformed regenerated plants.

Table 2. DPT, PPT, epi-PPT contents in regenerated wild type, regenerated transformed and wild type plants.

Range concentration (µg/mg d.w.) Transformed plant Wild type (control) plant Aerial Root Aerial Root DPT 0.105 – 0.230 * 0.3 – 0.5 1.4 – 1.5 PPT 0.006 – 0.133∆ * 0.006 – 0.011∆ <0.01 – 0.019∆ Epi-PPT 0.0006 – 0.036∆ * - -

*not determined ∆the concentration was calculated from highly concentrated extracts

Figure 5. HPLC chromatograms of transgenic and wild type Anthriscus. Epi-PPT (epipodophyllotoxin), PPT (podophyllotoxin), DPT (deoxypodophyllotoxin).

Figure 5. HPLC chromatograms of transgenic and wild type Anthriscus. Epi-PPT (epipodophyllotoxin), PPT (podophyllotoxin), DPT (deoxypodophyllotoxin).

Table 2. DPT, PPT, epi-PPT contents in regenerated wild type, regenerated transformed and wild type plants.

Range concentration (mg/mg d.w.)Transformed plant Wild type (control) plant

Aerial Root Aerial RootDPT 0.105 – 0.230 * 0.3 – 0.5 1.4 – 1.5PPT 0.006 – 0.133D * 0.006 – 0.011D <0.01 – 0.019D

Epi-PPT 0.0006 – 0.036D * - -

*not determined Dthe concentration was calculated from highly concentrated extracts

100

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5diSCuSSion

The production of bioactive compounds through in vitro transformed callus or cell cultures has been carried out in several plant species, especially in medicinal plants.16 The expression of human P450 genes in transgenic plants may lead to the formation of new secondary compounds.12

It has been reported that P450 3A4 can be expressed in plant cells (N. tabacum) in an active form.15 Feeding experiments with loratadine to N. tabacum suspension culture cells expressing P450 3A4 led to the formation of desloratadine, an active metabolite of loratadine that is also used as a drug compound. Another feeding experiment with indole in N. tabacum suspension culture cells expressing human cytochrome P450 2A6 and 3A4 led to the production of indican, a metabolite that is usually exclusively present in indigoferous dye plants,14 but not in N. tabacum.

Based on the examples mentioned, it is illustrated that the human cytochrome P450 3A4 in N. tabacum may transform exogenously provided DPT into epi-PPT. In our studies, we showed that N. tabacum cell suspension cultures expressing the human P450 3A4 were able to hydroxylate DPT to epi-PPT with the highest concentration of epi-PPT in the cells and medium on day 7. Also the epi-PPT formation in medium was significantly higher on day 7 in transformed cells compared to wild type cells. Possibly, epi-PPT was excreted into the medium because it is toxic for the cells. Interestingly, the wild type tobacco cell suspension cultures were also capable of converting DPT into epi-PPT, although at a much lower level.

We used the Agrobacterium transformation technique to introduce the human P450 3A4 gene in the callus cells of A. sylvestris. Since callus of A. sylvestris lost the capacity of producing DPT because they are in undifferentiated form,9 it is crucial to induce differentiation in order to restore the capability of the cell to produce DPT. We have developed a regeneration protocol of A. sylvestris17 that indeed after regeneration does restore the ability of the cells to produce DPT. There was no obvious somaclonal variation with respect to DPT production.

The transformed Anthriscus plants were able to form clearly detectable amounts of epi-PPT and small amounts of PPT as well (figure 5). Wild type Anthriscus has been reported to contain trace amounts of PPT (<0.01 µg/mg d.w.).9 We found PPT (0.006-0.011 µg/mg d.w.) in wild type Anthriscus which is in line with this previous finding. The PPT content in the transformed plants was slightly higher than in wild type, however, the involvement of P450 3A4 in this conversion is unlikely. It was reported that DPT can be converted into PPT in Linum album cell cultures,18 but the enzyme responsible for the hydroxylation has never been identified.

The transformation efficiency of A. sylvestris was low (about 1%) compared to N. tabacum (about 30%). Five transgenic plants from two independent transformations were obtained. The shoot regeneration of A. sylvestris required up to a year in the transformed cells instead of 6 months in the wild type plant.17 This might due to carbenicillin which is extensively used for eliminating Agrobacterium. Carbenicillin has a broad spectrum of activity against bacteria and only low toxicity to eukaryotes.19

101

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5Inhibition of callus growth and shoot regeneration has been observed in several plants such as Daucus carota,20 Malus domestica,21 Populus species,22 and Lycopersicon esculentum.23 Carbenicilin was removed from the medium once there was no Agrobacterium growth observed. Rooting took about 3 months which is similar to the wild type plant. Kanamycin showed no inhibition either in the transformed calli growth or the shoot and root formation.

We are the first to show the heterologous expression of the human cytochrome P450 3A4 in A. sylvestris. The construct under the control of a 35S promoter should provide constitutive expression in all parts of A. sylvestris plant. Fusion of a cytochrome P450 and a fluorescence protein has been reported in mammalian cells24 and plants.25 However, the YFP expression was low in our transformed plants (data not shown). This may explain why the amounts of epi-PPT formed in transformed plants are limited. Reverse transcriptase PCR will be carried out to confirm the transcription of the P450, YFP and kanamycin gene in the transformed plants and further work to screen high expression level of P450 3A4 is still in progress. In addition, the progeny will be screened on kanamycin and the segregation analysis will be done. For more effective biotransformation and high level of epi-PPT and / or PPT production, it is necessary to optimize the constructs used such as the promoter, the P450 3A4 gene, and the targeting signal peptide to the compartment where the DPT is produced. To the best of our knowledge, there is no report regarding where in the cell the DPT is produced. Recent papers suggest that the shikimate pathway occurs in the plastids because the shikimate enzymes are generally synthesized as precursors containing a plastid transit peptide that directs them to the plastids.26, 27 Thus, DPT may be synthesized in the plastids. The 3A4 expression construct used in the Anthriscus study was devoid of any targeting signal as the original human signal sequenced was not copied in the PCR fragment. It should be noted that the construct used for expression in N. tabacum did contain an additional endoplasmic reticulum targeting signal.

mAteriAL And metHodS

ChemicalsEpipodophyllotoxin (epi-PPT) was a gift from Dr. M. Angeles Castro (Salamanca University, Salamanca, Spain). The identity and purity of the compound was checked using HPLC. Acetonitrile and methanol were HPLC grade from Biosolve (Valkenswaard, the Netherlands). Dichloromethane was from Fisher Scientific (Landsmeer, the Netherlands). Formic acid, ammonium formate, podophyllotoxin, deuterated CDCl3 (99.96 %) were from Sigma Aldrich (Zwijndrecht, the Netherlands). DPT was synthesized chemically from PPT as described below. PPT (98% purity by HPLC) was obtained from Sino Future Pharmaceutical Company (Xi’an, China).

102

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5Synthesis of deoxypodophyllotoxin from podophyllotoxin Glacial acetic acid (100 mL) was added to PPT (5.0 g, 12 mmol) and Pd/C 10% (0.25 g, 2.4 mmol). The resulting suspension was reacted under hydrogen (H2) atmosphere (3.0 bar) at 50oC for 16 h in a Parr apparatus. The suspension was diluted with ethyl acetate (200 mL) and filtered over Celite. The solvents were evaporated under reduced pressure and the residue was purified using column chromatography using ethyl acetate: hexane (1:1, v/v). The pure product was obtained as white powder in quantitative yield (85 %). 1H-NMR spectra are in agreement with the reported literature.28, 29 13C NMR spectra are in agreement with the reported literature.29 MS ESI-MS m/z [M+H]+ C22H23O7 calc 399.4, found 399.2, [M+NH4]

+ C22H26O7N calc 416.4, found 416.1. [M+Na]+ C22H22O7Na calc 422.4, found 422.1.

plant material and callus inductionFruits (ripe mericarps, hereafter called seeds) of Anthriscus sylvestris L. (Hoffm.) were collected from wild habitat in June 2007 in Groningen, the Netherlands (53o 13’ 34” N and 6o 32’ 43” E). Voucher specimen of the plants used have been deposited in our department, encoded Asylv2007-4. In vitro plants were grown and the calli were initiated as described recently.17

plasmid constructionThe cDNA encoding human P450 3A4 was a gift from F.P. Guengerich (Vanderbilt University School of Medicine, Nashville, USA). The plasmid construct used for Anthriscus transformation is shown in figure 3. Human cDNA devoid of the endoplasmic reticulum retention signal was amplified by PCR to create an overlapping sequence with YFP (pMON999, Monsanto, Missouri, USA), using a set of primers, a forward primer 5’-GCTTAGGAGGTCCATGGCTCTGTTATTAGCAGTTTTTCT-3’ and a reverse primer 5’ CCTTGCTCACCATTCAGGCTCCACTTACGGTGCCATCCCTTGACT-3’. YFP gene was amplified by PCR to introduce an overlapping sequence with cDNA and an XmaI restriction site at the 3’-end, using a set of primers, a forward primer 5’-GGATGGCACCGTAAGTGGAGCCT- GAATGGTGAGCAAGGGCGAGGAGC-3’ and a reverse primer 5’-GGCCGCCGGGCGATCTAGTAACATAGATG-3’. The second round of PCR was performed to create cDNA::YFP::NOS with a set of primers, a forward primer 5’-GCTTAGGAGGTCCATGGCTCTGTTATTAGCAGTTTTTCT-3’, and a reverse primer 5’-GGCCGCCGGGCGATCTAGTAACATAGATG-3’. All PCR was performed using phusion polymerase (Biolabs, Leiden, the Netherlands) and confirmed by sequencing. Thus, P450 3A4 and YFP were fused in frame. We cloned this sequence to plasmid pJIT145, which contained a double 35S promoter (a gift from Mark Smedley, John Innes Centre, UK). The plasmid was cut with KpnI and XmaI and cloned to pGreen029.30 The binary pGreen029 contains the nptII gene for kanamycin selection of transformed plant cells. The resulting plasmid (pGreen029-3A4) was cloned and transformed into Escherichia coli DH5a and mobilized into Agrobacterium tumefaciens EHA 105 (a gift from P.J.J. Hooykaas, Leiden University) using the electroporation method (pulse cells at 1.8 KV in 1 cm cuvette, 600 W and 10 µF capacitance).

103

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5Callus transformation and regenerationAgrobacterium strain EHA 105 harboring the pGreen029-3A4 was grown overnight in LB medium supplemented with rifampicin (50 mg l-1) and kanamycin (100 mg l-1) in a rotary shaker (225 rpm, 28oC). The Agrobacterium culture was centrifuged at 3000 g for 5 min and the pellet resuspended in Gamborg's B5 medium supplemented with 40 g l-1 sucrose. Transformation was achieved by infecting A. sylvestris callus cells with Agrobacterium. Calli were dried on sterile filter paper, transferred to callus medium and incubated for 2 days at 24oC. Incubated calli were subcultured to callus medium supplemented with kanamycin (100 mg l-1) and carbenicillin (250 mg l-1). Shoots were regenerated on shoot-inducing medium as described,17 supplemented with kanamycin (100 mg l-1) and carbenicillin (250 mg l-1). Kanamycin was added to callus, shoot and root-inducing media. Carbenicillin was used until no Agrobacterium growth was observed (up to 6 months). Regenerated shoots measuring about 2 cm in length were transferred into root-inducing medium.17 The rooted plantlets were thoroughly washed with tap water to remove the agar and transferred to sterile soil in small pots and covered with plastic to maintain high humidity. They were placed in a climate chamber with 14-h light/10-h dark regimen, photosynthetic photon flux density 100 μmol m−2 s−1 at 22°C. After 2 weeks, the humidity was reduced to the climate room level by removing the plastic cover and 5 plants were transferred to bigger pots and kept in the green house. The efficiency of the transformation was about 1%.

nucleic acid isolation and detectionPlasmid isolation was performed using Nucleospin Plasmid Isolation kit (Macherey-Nagel). Total plant DNA was isolated using a DNA purification kit (NucleoSpin® Plant II, Macherey-Nagel). The presence of P450 3A4 in transformed plants was confirmed by PCR using a primer pair, a forward primer 5’-AACAGCCTGTGCTGGCTATC-3’ and a reverse primer 5’-GTGTATCTTCGAGGCGACTT-3’ giving a product of 561 bp of an internal fragment of the P450 3A4 cDNA. The primer pair for YFP gene is 5’-GGAATTCCATATGGTGAGCAAGGGCGAGGAGCT-3 and 5’-CGGGATCCTTACTTG- TACAGCTCGTCCATGC-3’ giving a product of 739 bp. The primer pair for kanamycin gene is 5’-GGAGCGGCGATACCGTAAAG-3’ and 5’-CGGCTATGACTGGGCACAAC-3’ giving a product of 687 bp. Plant genomic DNA was used as a template. The PCR condition was as follows: initial denaturation at 98oC for 5 min followed by 35 cycle of amplification at 98oC for 30 sec, annealing at 65oC for 30 sec, and extension at 72oC for 30 sec. The final extension step was at 72oC for 10 min.

feeding experiments with nicotiana tabacum suspension culturesNicotiana tabacum cv. Petit Havana suspension cultures were cultivated in a volume of 120 mL in 300 mL Erlenmeyer flasks, and subcultured on a weekly basis by transferring 60 mL of old culture to 60 mL of fresh LS medium.14 The P450 3A4 cDNA was cloned into the binary plasmid pPSI and cell suspension cultures of N. tabacum cv. Petit Havana were transformed as described.14 The cells transformed with human cytochrome P450 3A4 (CYP3A4) as well as the wild type (control) cells were used for the feeding

104

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5experiment.15 One day after subculture, one set of flasks of CYP3A4 and wild type were supplemented with 2.0 mg DPT in 100 µl DMSO and further cultivated until 7 days. A set of three flasks was used for each feeding experiment. The cells and the supernatant (media) were harvested on day 4, and 7 for analysis. Cells were separated with a nylon mesh in a Büchner-funnel and frozen in liquid nitrogen. Both media and cells were freeze-dried prior to extraction.

extraction, HpLC, LC-mS-mS and nmr analysisThe extraction method of the plant material and the media was as described by Koulman et al.9 Shortly, 100 mg dried cells were weighed in a sovirel tube and all the supernatant were used for extraction. A 2.0 ml portion of 80% methanol was added and the mixture was sonicated during 1 hour. Subsequently 4.0 ml of dichloromethane and 4.0 ml H2O were added. The mixture was vortexed and centrifuged at 1,000 g for 5 minutes. The aqueous layer was discarded and 2.0 ml of organic layer were transferred into a 2 mL Eppendorf tube. The organic layer was left in the fume hood until dried. The residue was redissolved in 2.0 ml methanol and filtered over a 0.45 µm HPLC syringe filter (nylon). The samples were submitted to HPLC analysis. DPT, PPT and epi-PPT were analyzed by HPLC, LC-MS/MS and NMR analysis as described.31

yfp visualizationCalli of the transformed and wild type plants (control) were used to check the YFP expression. Visualization of the YFP expression was performed by using a Zeiss Axioskop fluorescence microscope (Zeiss Netherlands B.V. Weesp, The Netherlands).

Statistical treatment of data All data presented in this study are the mean + SD, of at least three independent experiments. Comparative statistical analyses of groups were performed using the student’s t-test. All statistical tests were considered significant at p values < 0.05.

ACKnoWLedgement

The authors would like to express their gratitude to Margriet Ferwerda for her advice concerning the plant tissue culture, Paul D. Dijkwel and Marcel J.G. Sturre for their advice concerning the construct, J.T.M. Elzenga, Ger Telkamp, and Jacob J. Hogendorf for the climate room and the field facilities. Financial support by the Ubbo Emmius fund, University of Groningen, is acknowledged.

105

Ag

robA

cterIu

m tra

nsFo

rMa

tion

5referenCeS

1. Verpoorte R, Contin A, Memelink J. Bio-technology for the production of plant secondary metabolites. Phytochem Rev 2002;1:13-25.

2. Lessard PA, Kulaveerasingam H, York GM, et al. Manipulating gene expression for the metabolic engineering of plants. Metab Eng 2002;4:67-79.

3. Cragg GM, Grothaus PG, Newman DJ. Impact of natural products on develop-ing new anti-cancer agents Chem Rev 2009;109:3012-3043.

4. Nayar MP, Sastry APK. Red data book of Indian plants. Calcutta, 1990.

5. Vasilev NP, Julsing MK, Koulman A, et al. Bioconversion of deoxypodophyllo-toxin into epipodophyllotoxin in E. coli using human cytochrome P450 3A4. J Biotechnol 2006;126:383-393.

6. Federolf K, Alfermann AW, Fuss E. Aryltetralin-lignan formation in two different cell suspension cultures of Linum album: Deoxypodophyllotoxin 6-hydroxylase, a key enzyme for the formation of 6-methoxypodophyllotox-in. Phytochem 2007;68:1397-1406.

7. Jackson DE, Dewick PM. Aryltetralin lignans from Podophyllum hexandrum and Podophyllum peltatum. Phytochem 1984;23:1147-1152.

8. Hendrawati O, Woerdenbag J, Hille J, et al. Seasonal variations in the deoxypodo-phyllotoxin content and yield of Anthr-iscus sylvestris L. (Hoffm.) grown in the field and under controlled conditions. J Agri Food Chem 2011;59:8132-8139.

9. Koulman A, Kubbinga ME, Batterman S, et al. A phytochemical study of lignans in whole plants and cell suspension cultures of Anthriscus sylvestris. Planta Med 2003;69:733-738.

10. Renaud JP, Cullin C, Pompon D, et al. Expression of human liver cytochrome P450 IIIA4 in yeast. Eur J Biochem 1990;194:889-896.

11. Hayashi K, Sakaki T, Kominami S, et al. Coexpression of genetically engineered fused enzyme between yeast NADPH-P450 reductase and human cytochrome P450 3A4 and human cytochrome b5 in yeast. Arch Biochem Biophys 2000;381:164-170.

12. Gillam EMJ, Guengerich FP. Exploiting the versatility of human cytochrome P450 enzymes: the promise of blue

roses from biotechnology. IUBMB Life 2001;52:271-277.

13. Zhang Z, Li Y, Shou M, et al. Influence of different recombinant systems on the cooperativity exhibited by cytochrome P4503A4. Xenobiotica 2004;34:473-486.

14. Warzecha H, Frank A, Peer M, et al. Formation of the indigo precursor indican in genetically engineered tobacco plants and cell cultures. Plant Biotechnol J 2007;5:185-191.

15. Warzecha H, Ferme D, Peer M, et al. Bioconversion of the antihistaminc drug loratadine by tobacco cell suspension cultures expressing human cytochrome P450 3A4. J Biosci Bioeng 2010;109:288-290.

16. Hendrawati O, Woerdenbag HJ, Hille J. et al. Metabolic engineering strategies for the optimization of medicinal and aromatic plants: realities and expecta-tions. J Med Spice Plants 2010;15:111-126.

17. Hendrawati O, Hille J, Woerdenbag HJ, et al. In vitro regeneration of wild chervil (Anthriscus sylvestris L.). In vitro Cell & Dev Biol - Plant 2011: in press.

18. Seidel V, Windhövel J, Eaton G, et al. Biosynthesis of podophyllotoxin in Linum album cell cultures. Planta 2002;215:1031-1039-1039.

19. Ling HQ, Kriseleit D, Ganal MW. Effect of ticarcillin/potassium clavulanate on callus growth and shoot regeneration in Agrobacterium-mediated transforma-tion of tomato (Lycopersicon esculentum Mill.). Plant Cell Rep 1998;17:843-847.

20. Okkels FT, Pedersen MG. The toxicity to plant tissue and to Agrobacterium tu-mefaciens of some antibiotics. Acta Hort (ISHS) 1988;225:199-208.

21. Modgil M, Sharma R. Effect of antibiot-ics on regeneration and elimination of bacteria during gene transfer in apple. Acta Hort (ISHS) 2009;839:353-360.

22. Bosela M. Effects of β-lactam antibiotics, auxins, and cytokinins on shoot regener-ation from callus cultures of two hybrid aspens, Populus tremuloides x P. tremula and P. xcanescens x P. gradidentata. Plant Cell Tiss Org Cult 2009;98:249-261.

23. Mamidala P, Nanna RS. Influence of antibiotics on regeneration efficiency in tomato. Plant Omics 2009;2:135-140.

106

Ag

robA

cterIu

m tra

nsFo

rMa

tion

524. Steffens S, Frank S, Fischer U, et al.

Enhanced green fluorescent protein fusion proteins of herpes simplex virus type 1 thymidine kinase and cytochrome P450 4B1: applications for prodrug-acti-vating gene therapy. Cancer Gene Ther 2000;7:806-12.

25. Höfer R, Briesen I, Beck M, et al. The Arabidopsis cytochrome P450 CYP86A1 encodes a fatty acid ω-hydroxylase involved in suberin monomer biosynthe-sis. J Exp Bot 2008;59:2347-2360.

26. Tzin V, Galili G. New insights into the shikimate and aromatic amino acids bio-synthesis pathways in plants. Mol Plant 2010.

27. Mustafa NR, Verpoorte R. Chorismate derived C6C1 compounds in plants. Planta 2005;222:1-5.

28. Brewer CF, Loike JD, Horwitz SB, et al. Conformational analysis of podophyllo-toxin and its congeners. Structure-activ-ity relationship in microtubule assembly. J Med Chem 1979;22:215-221.

29. Bogucki DE, Charlton JL. An asymmetric synthesis of (-)-deoxypodophyllotoxin. J Org Chem 1995;60:588-593.

30. Hellens RP, Edwards EA, Leyland NR, et al. pGreen: a versatile and flexible binary Ti vector for Agrobacterium-mediated plant transformation. Plant Mol Biol 2000;42:819-832-832.

31. Hendrawati O, Woerdenbag HJ, Michiels PJA, et al. Identification of lignans and related compounds in Anthriscus syl-vestris by LC-ESI-MS/MS and LC-SPE-NMR. Phytochem 2011: in press.

Summary Samenvatting

Ringkasan Acknowledgement &

111

suM

Ma

ry

&SummAry

In recent years, strategies and techniques for the production of natural compounds (plant derived fine chemicals) and / or the breeding of medicinal and aromatic plants have expanded. The aspiration for efficient production of high value natural products with medicinal purpose (e.g. paclitaxel, artemisinin, and vincristine) has triggered novel approaches. Metabolic engineering and pathway optimization with the aim to reduce costs and increase productivity have become a main focus of academia and industry. Until now, only a limited number of plant cell cultures and isolated enzymes have yielded sufficiently high productivity to allow commercial exploitation. Engineering of microorganisms has proven to be a valuable tool and the concept of metabolic engineering has been transferred to plant science opening promising perspectives. Although pathway engineering has been used for crop plants, but the application of this technique to medicinal plants has not yet been explored well. An overview of the current production of plant-derived and medicinally relevant compounds with their metabolic engineering strategies is presented in chapter 1. In addition, common strategies and techniques in medicinal plant biotechnology and challenges in plants metabolic engineering with several relevant case studies are also discussed in chapter 1.

In this thesis, we focus on the lignan (epi)podophyllotoxin. We aimed to engineer Anthriscus sylvestris (Apiaceae), which contains deoxypodophyllotoxin (DPT) as its main lignin constituent, to produce (epi)podophyllotoxin. A. sylvestris is a wild plant to northern Europe and in other temperate regions of the world. To get a clear picture of the deoxypodophyllotoxin content over the developmental stages of the plant, we studied the variations of deoxypodophyllotoxin during different developmental stages (chapter 2). Two sets of experiments were carried out with sixteen collected plants and seeds, originating from a wide geographical range in Europe. Firstly, we compared the DPT content in various organs of the natively grown plants with that of the same plants now grown under identical conditions in the field. Secondly, we compared the DPT content between controlled climate room and field grown plants during one life cycle. The DPT concentration in both aerial and root parts dropped when the plants were cultivated in their non-native environment. The DPT contents of the climate room and the field plants were comparable, but the biomass and flowering time differed considerably. Environmental factors appear to be an important determinant for the DPT production in A. sylvestris. For field grown plants the highest DPT content was found in March (2nd year) for both root and aerial part (0.15 % w/w dried weight (d.w.) roots; 0.03 % w/w d.w. aerial material). For plants grown in the climate room the highest concentration was observed in April (2nd year) in the root and in July (1st year) in the aerial part (0.14 % w/w d.w. roots; 0.05 % w/w d.w. aerial material). The root part is suggested to be used for the optimal isolation of DPT and the best time to harvest is in April (2nd year) for both indoor and outdoor plants when the plants reach the optimal DPT production and an adequate biomass.

A better insight into the occurrence of DPT combined with profound knowledge of its biosynthestic pathway(s) may help to develop alternative sources for podophyllotoxin

112

suM

Ma

ry

&(chapter 3). Using HPLC combined with electrospray tandem mass spectrometry and NMR spectroscopy techniques, we identified 14 lignans and related structures in roots of A. sylvestris. Podophyllotoxone, deoxypodophyllotoxin, yatein, anhydropodorhizol, 1-(3’-Methoxy-4’,5’-methylenedioxyphenyl)1-ξ-methoxy-2-propene, and 2-butenoic acid, 2-methyl-4-[[(2Z)-2-methyl-1-oxo-2-buten-1-yl]oxy]-, (2E)-3-(7-methoxy-1, 3-benzodioxol-5-yl)-2-propen-1-yl ester, (2Z)- were the major compounds. α-Peltatin, podophyllotoxin, β-peltatin, isopicropodophyllone, β-peltatin-a-methylether, (Z)-2-angeloyloxymethyl-2-butenoic acid, anthriscinol methylether, and anthriscrusin were present in lower concentrations. α-Peltatin, β-peltatin, isopicropodophyllone, podophyllotoxone, and β-peltatin-a-methylether have not been previously reported to be present in A. sylvestris. Based on our findings we propose a hypothetical biosynthetic pathway of aryltetralin lignans in A. sylvestris (figure 3 in chapter 3).

A. sylvestris allows an interesting approach for genetic modification to directly produce (epi)podophyllotoxin in the plant, because of the high content of DPT and its close biosynthetic relationship with podophyllotoxin. Human cytochrome P450 3A4 in E.coli DH5α selectively hydroxylates DPT at the C7 position yielding epipodophyllotoxin, the diastereoisomer of podophyllotoxin. Human cytochrome P450 3A4 was transferred in A. sylvestris via Agrobacterium tumefaciens. For transformation purposes, a regeneration protocol of A. sylvestris was established first (chapter 4). Plants were regenerated from callus material. Calli were induced from hypocotyl explants in Gamborg's B5 medium supplemented with 40 g l-1 sucrose, 2 mg l-1 2,4-dichlorophenoxyacetic acid and 1 mg l-1 6-benzylaminopurin. Subsequently, calli were transferred to shoot induction medium, Gamborg's B5 containing 40 g l-1 sucrose, 2 mg l-1 zeatin riboside and 1.0 mg l-1 indole-3-butyric acid. Regenerated shoots were obtained within 6 months and transferred into root induction medium, Gamborg's B5 supplemented with 10 g l-1 sucrose for 2 months. Regenerated plants were acclimatized in soil and grown in the greenhouse. Eventually they were transferred to the field with 100% survival. Regenerated plants flowered and shed fertile seeds. This protocol facilitates regeneration of genetically transformed plant cells with Agrobacterium.

Feeding DPT to Nicotiana tabacum cell suspension cultures carrying the human P450 3A4 gene led to the bioconversion of DPT to epipodophyllotoxin (chapter 5). The A. sylvestris callus tissue was transformed with Agrobacterium tumefaciens carrying the human P450 3A4 gene and the cells were regenerated. The transformed regenerated plants now were able to form epipodophyllotoxin in low amounts. These results emphasize a future prospect of engineering A. sylvestris for (epi)podophyllotoxin production. Further optimization of the plasmid construction such as the promoter, P450 3A4 gene, and targeting signal peptide are needed to increase the level of expression of human P450 3A4 gene which should lead to a higher level of (epi)podophyllotoxin production. This engineering strategy may be applied to other plants producing DPT.

113

saM

enVa

tting

&SAmenVAtting

De afgelopen jaren is het arsenaal aan methoden en strategieën dat beschikbaar is voor de productie van plantenstoffen en voor het kweken van medicinale en aromatische planten uitgebreid. De wereldwijde ambitie om natuurstoffen voor medische toepassing op efficiënte wijze te produceren (bijvoorbeeld paclitaxel, artemisinine en vincristine) heeft de ontwikkeling van nieuwe technieken gestimuleerd. ‘Metabolic engineering’ en ‘pathway optimization’, met als doel kostenbesparing en productiviteitsverhoging te bewerkstelligen, zijn belangrijke benaderingen die worden toegepast door universiteiten en industrieën. Tot nu toe is met slechts een beperkt aantal plantencelcultures en geïsoleerde enzymsystemen een productieproces ontwikkeld dat rendabel genoeg is voor commerciële toepassing. ‘Engineering’ van micro-organismen is een waardevolle aanpak gebleken en wordt nu ook toegepast in planten, hetgeen de nodige perspectieven biedt. ‘Pathway engineering’ is tot nu toe vooral gebruikt voor planten die als voedingsmiddel dienen; de toepassing ervan voor medicinale planten is nog niet goed onderzocht. In hoofdstuk 1 geven wij een overzicht van de huidige productiemogelijkheden van (medisch relevante) plantenstoffen met behulp van ‘metabolic engineering’. Verder bespreken we in dit hoofdstuk meer algemene benaderingswijzen en technieken in de plantenbiotechnologie, en uitdagingen die het terrein van de ‘metabolic engineering’ biedt. Ter illustratie beschrijven we een aantal case-studies uit de literatuur.

In dit proefschrift richten we ons op het lignaan (epi)podofyllotoxine. We hebben geprobeerd om de plant Anthriscus sylvestris (fluitenkruid), behorend tot de familie van de schermbloemigen (Apiaceae), te onderwerpen aan ‘engineering’-technieken, met als doel het belangrijkste lignaan in de plant, deoxypodofyllotoxine (DPT), om te laten zetten in het chemisch nauw verwante (epi)podofyllotoxine. A. sylvestris komt in het wild voor in Noord-Europa en in andere gematigde gebieden over de hele wereld. Om een helder beeld te krijgen van de variatie aan deoxypodofyllotoxine gedurende de groei van de plant, hebben we het gehalte ervan gemeten in verschillende ontwikkelingsstadia (hoofdstuk 2).

Twee sets aan experimenten zijn uitgevoerd met zestien verschillende planten en zaden, afkomstig uit een breed verspreidingsgebied binnen Europa. Eerst vergeleken we het DPT-gehalte in verschillende organen van planten die geoogst waren op de plaats waar ze van nature voorkomen met dat van dezelfde planten die waren gekweekt onder gelijke condities in het veld. Daarna vergeleken we het DPT-gehalte in planten die onder gecontroleerde omstandigheden in kassen waren opgegroeid met dat van planten in het veld, gedurende een hele levenscyclus (twee jaar). Het DPT-gehalte nam af in zowel bovengrondse delen als in wortels van planten die waren gekweekt in een andere omgeving dan waar ze van nature groeiden. De DPT-gehaltes in de planten uit de kassen en die van het veld waren vergelijkbaar, maar de hoeveelheid gevormde biomassa en de bloeitijd verschilden aanzienlijk. Omgevingsfactoren bleken bepalend voor de DPT-productie in A. sylvestris. Voor planten die waren opgegroeid op het veld vonden we het hoogste DPT-gehalte in maart van het tweede jaar, zowel in

114

saM

enVa

tting

&de wortels als in de bovengrondse delen (0,15% w/w op basis van drooggewicht (dw) in de wortels en 0,03% w/w dw in bovengrondse delen). In kasplanten vonden we het hoogste DPT-gehalte in april van het tweede jaar in de wortels (0,14% w/w dw) en in juli van het eerste jaar in de bovengrondse delen (0,05 % w/w dw). De beste tijd om wortelmateriaal te oogsten voor de isolatie van DPT is in april van het tweede jaar van de levenscyclus van d e plant. Productie en biomassa zijn dan optimaal.

Goed inzicht in het voorkomen van DPT in combinatie met kennis van de biosyntheseroute(s) voor deze verbinding kan nuttig zijn om alternatieve bronnen voor podofyllotoxine te vinden (hoofdstuk 3). Door gebruik te maken van HPLC in combinatie met electrospray tandem massaspectrometrie en NMR-spectroscopie, konden we 14 lignanen en gerelateerde verbindingen in de wortels van A. sylvestris identificeren. Podofyllotoxon, deoxypodofyllotoxine, yateine, anhydropodorhizol, 1-(3’-methoxy-4’,5’-methyleendioxyfenyl)1-ξ-methoxy-2-propeen, en 2-buteenzuur, 2-methyl-4-[[(2Z)-2-methyl-1-oxo-2-buteen-1-yl]oxy]-,(2E)-3-(7-methoxy-1, 3-benzodioxol-5-yl)-2-propeen-1-yl ester, (2Z)- waren hoofdcomponenten. α-Peltatine, podofyllotoxine, β-peltatine, isopicropodofyllon, β-peltatine-a-methylether, (Z)-2-angeloyloxymethyl-2-buteenzuur, anthriscinolmethylether, en anthriscrusine waren in geringere hoeveelheden aanwezig. Het voorkomen van α-peltatine, β-peltatine, isopicropodofyllon, podofyllotoxon, en β-peltatine-a-methylether in A. sylvestris was niet eerder gerapporteerd. Gebaseerd op deze resultaten presenteren we een hypothetische biosyntheseroute voor aryltetralinelignanen in A. sylvestris (figuur 3 in hoofdstuk 3).

A. sylvestris is een interessante kandidaat voor genetische modificatie, om te komen tot een directe productie van (epi)podofyllotoxine in de plant, vanwege het hoge gehalte aan DPT en de sterke overeenkomst met podofyllotoxine in biosynthese en chemische structuur. Humaan cytochroom P450 3A4 in E.coli DH5α kan DPT op de C7-positie selectief hydroxyleren, waardoor epipodofyllotoxine, de diastereoisomeer van podofyllotoxine, ontstaat. Humaan cytochroom P450 3A4 is ingebracht in A. sylvestris met behulp van Agrobacterium tumefaciens. Voor transformatiedoeleinden hebben we een regeneratieprotocol voor A. sylvestris opgezet (hoofdstuk 4). Planten werden geregenereerd uit callusmateriaal. Callus werd geïnduceerd uit hypocotyl-explantaten in Gamborg's B5 medium met 40 g l-1 sucrose, 2 mg l-1 2,4-dichloorfenoxyazijnzuur en 1 mg l-1 6-benzylaminopurine. Vervolgens werden de calli overgebracht in een scheut-inducerend medium, Gamborg's B5 met 40 g l-1 sucrose, 2 mg l-1 zeatineriboside en 1 mg l-1 indool-3-boterzuur. Na zes maanden kregen we geregenereerde scheuten die werden overgebracht in een medium dat wortelvorming induceert, Gamborg's B5 met 10 g l-1 sucrose voor een periode van twee maanden. Gevormde regeneranten werden geacclimatiseerd in potgrond en groeiden verder in een kas. Uiteindelijk werden zij overgebracht naar het veld. Er was 100% overleving. Geregenereerde planten bloeiden en vormden vruchtbaar zaad. Dit protocol wordt gebruikt voor de regeneratie van met Agrobacterium genetisch getransformeerde plantencellen.

Het toevoegen (‘feeding’) van DPT aan celsuspensiecultures van Nicotiana tabacum waarin het humane P450 3A4 gen was ingebouwd, resulteerde in de bioconversie

115

saM

enVa

tting

&van DPT in epipodofyllotoxine (hoofdstuk 5). Callusweefsel van A. sylvestris werd getransformeerd met Agrobacterium tumefaciens voorzien van het humane P450 3A4 gen en de cellen werden geregenereerd. De getransformeerde regeneranten bleken in staat om kleine hoeveelheden epipodofyllotoxine te maken. Deze resultaten tonen aan dat het in de toekomst mogelijk kan zijn om via ‘engineering’ van A. sylvestris (epi)podofyllotoxine te produceren. Om deze benadering succesvol te maken, moet het plasmideconstruct verder worden geoptimaliseerd, via de promoter, het P450 3A4 gen, en een signaalpeptide. Dit is nodig om een hoger expressieniveau van het gen te krijgen, hetgeen zou moeten leiden tot een hogere productie van (epi)podofyllotoxine. Deze strategie van ‘engineering’, ten slotte, kan ook worden toegepast bij andere planten die DPT maken.

116

ring

ka

san

&ringKASAn

Beberapa tahun terakhir ini, strategi dan teknik untuk produksi senyawa alami (bahan kimia yang berasal dari tanaman) dan / atau pengembangbiakan tanaman medisinal dan tanaman penghasil senyawa aromatik berkembang sangat pesat. Strategi baru untuk memproduksi secara efisien senyawa alami yang digunakan untuk obat (misalnya paclitaxel, artemisinin, dan vincristine) telah melahirkan banyak pendekatan baru. Fokus utama riset di bidang akademia dan industri adalah untuk menurunkan biaya produksi dan meningkatkan produktivitas. Sampai saat ini, jumlah kultur sel tanaman dan enzim dengan produktivitas yang tinggi yang bisa dikomersialkan jumlahnya masih sangat terbatas. Rekayasa mikroorganisme telah terbukti menjadi sarana yang penting, lebih lanjut lagi penerapan konsep metabolic engineering di tanaman telah mampu membuka peluang baru yang menjanjikan. Metode engineering pathway banyak digunakan untuk pembiakan tanaman pangan, namun penerapan teknik tersebut untuk tanaman obat masih belum optimal. Pembahasan mengenai produksi terkini dari senyawa obat yang berasal dari tanaman dengan teknik metabolic engineering dibahas di bab 1. Selain itu, beberapa strategi dan teknik umum bioteknologi tanaman medisinal dan permasalahannya dilengkapi dengan beberapa contoh kasus studi juga didiskusikan lebih lanjut di bab 1.

Fokus tesis ini adalah senyawa lignan (epi)podopyllotoxin. Tujuan penelitian ini adalah untuk merekayasa tanaman Anthriscus sylvestris yang secara alami mengandung senyawa utama deoxypodophyllotoxin (DPT) agar dapat digunakan untuk memproduksi senyawa (epi)podophyllotoxin. A. sylvestris (Apiaceae) merupakan tanaman liar di Eropa utara dan beberapa tempat beriklim sedang di dunia. Untuk mendapatkan gambaran tentang kandungan DPT, dilakukan perbandingan tentang variasi kandungan DPT di berbagai tahap perkembangan tanaman di lahan (bab 2). Penelitian ini dibagi dalam dua tahap eksperimen dengan 16 origin tanaman dan biji dari berbagai daerah di Eropa. Pertama, kandungan DPT tanaman yang tumbuh di lingkungan alaminya (native) dibandingkan dengan tanaman dengan daerah asal yang sama yang dikultivasi di lahan (non-native). Kedua, kandungan DPT antara tanaman yang dikultivasi di ruangan dengan iklim terkontrol dibandingkan dengan tanaman yang dikultivasi di lahan selama satu siklus. Hasil eksperimen menunjukkan bahwa kandungan DPT pada bagian aerial dan pada akar tanaman native menurun bila tanaman dikultivasi di lingkungan yang non-native, sedangkan kandungan DPT tanaman di ruangan dengan iklim terkontrol dan di lahan sebanding, namun terdapat perbedaan pada biomasa dan waktu mulai berbunga. Dapat disimpulkan bahwa faktor lingkungan tampaknya menjadi faktor penting untuk produksi senyawa DPT di A. sylvestris. Untuk tanaman di lahan, kandungan DPT tertinggi untuk bagian akar dan aerial adalah pada bulan Maret di tahun kedua (0,15% w/w dibagian akar; 0,03 % w/w di bagian aerial). Tanaman yang di ruangan dengan iklim terkontrol, kandungan DPT tertinggi diperoleh pada tahun kedua di bulan April untuk bagian akar (0,14% w/w) dan di tahun pertama bulan Juli untuk bagian aerial (0,05% w/w). Untuk mengoptimalkan isolasi DPT, disarankan untuk mengambil bagian akar. Waktu terbaik untuk panen

117

ring

ka

san

&adalah pada saat tanaman mencapai produksi DPT secara optimal dan menghasilkan biomasa yang cukup yaitu di tahun kedua pada bulan April baik untuk tanaman di ruangan dengan iklim terkontrol maupun tanaman di lahan.

Pemahaman mengenai terbentuknya DPT dikombinasikan dengan pemahaman mengenai jalur biosintesis yang dijabarkan pada bab 3 dapat digunakan untuk mengembangkan strategi baru dalam memproduksi senyawa (epi)podophyllotoxin secara efisien. Sebanyak 14 senyawa lignans dan senyawa lainnya telah berhasil diisolasi dari bagian akar A. sylvestris dan diidentifikasi menggunakan HPLC yang dikombinasikan dengan ESI-MS (Electrospray Ionization – Mass Spectrometry) dan NMR (Nuclear Magnetic Resonance). Podophyllotoxone, deoxypodophyllotoxin, yatein, anhydropodorhizol, 1-(3’-methoxy-4’,5’-methylenedioxyphenyl)1-ξ-methoxy-2-propene dan 2-butenoic acid, 2-methyl-4-[[(2Z)-2-methyl-1-oxo-2-buten-1-yl]oxy]-, (2E)-3-(7-methoxy-1,3-benzodioxol-5-yl)-2-propen-1-yl ester, (2Z)- adalah merupakan senyawa utama. Sedangkan, α-Peltatin, podophyllotoxin, β-peltatin, isopicropodophyllone, β-peltatin-a-methylether, (Z)-2-angeloyloxymethyl-2-butenoic acid, anthriscinol methylether and anthriscrusin merupakan senyawa minor. Senyawa α-Peltatin, β-peltatin, isopicropodophyllone, podophyllotoxone, dan β-peltatin-a-methylether pada A. sylvestris belum pernah dilaporkan sebelumnya. Berdasarkan penemuan ini, kami mengusulkan hipotesis jalur biosintetik lignan aryltetralin di A. sylvestris (gambar 3 bab 3).

A. sylvestris diduga dapat dimodifikasi secara genetik untuk memproduksi langsung senyawa epi(podophyllotoxin) di tanaman karena kandungan DPT yang tinggi dan kedekatan jalur biosintetis dengan podophyllotoxin. Gen P450 3A4 yang diisolasi dari hati manusia dapat menghidroksilasi DPT pada posisi C-7 menghasilkan epipodophyllotoin (diastereoisomer dari podophyllotoxin) di bakteri E. coli DH5α. Gen P450 3A4 dapat ditransfer ke dalam A. sylvestris melalui Agrobacterium. Protokol regenerasi tanaman A. sylvetris untuk proses transformasi ini perlu dikembangkan terlebih dahulu (bab 4). Tanaman diregenerasi dari kalus sel. Kalus sel diinduksi dari eksplan hipokotil di media Gamborg's B5 ditambah dengan 40 g/l sukrosa, 2 mg/l 2,4-dichlorophenoxyacetic acid dan 1 mg/l 6-benzylaminopurin. Kemudian, kalus sel dipindahkan ke media induksi untuk kultur pucuk (tunas), media Gamborg's B5 ditambah dengan 40 g/l sukrosa, 2 mg/l zeatin riboside dan 1,0 mg/l indole-3-butyric acid. Kultur pucuk diperoleh dalam waktu enam bulan dan kemudian ditransfer ke media induksi akar selama dua bulan, yaitu media Gamborg's B5 ditambah dengan 10 g/l sukrosa. Tanaman baru yang terbentuk dikultivasi di rumah kaca dan kemudian diaklimatisasi di lahan. Setelah beberapa bulan, tanaman baru dipindahkan ke lahan dengan tingkat kesuksesan 100%. Tanaman hasil kultur jaringan berbunga dan menghasilkan biji yang subur. Protokol regenerasi tanaman ini akan memfasilitasi regenerasi sel tanaman yang telah ditransformasi dengan Agrobacterium.

DPT yang ditambahkan (feeding) ke kultur sel suspensi Nicotiana tabacum yang telah ditransformasi dengan gen P450 3A4 dari hati manusia menghasilkan biokonversi dari DPT menjadi epipodophyllotoxin (bab 5). Jaringan kalus A. sylvestris ditransformasi dengan Agrobacterium tumefaciens yang mengandung gen P450

118

ring

ka

san

&3A4 dan kemudian jaringan kalus tersebut diregenerasikan menjadi tanaman utuh yang mengandung gen P450 3A4. Tanaman hasil transformasi dan regenerasi dapat menghasilkan epipodophyllotoxin dengan konsentrasi yang rendah. Hasil penemuan ini membuka prospek baru untuk memproduksi senyawa epipodophyllotoxin melalui rekayasa genetika tanaman A. sylvetris. Konstruksi plasmid, seperti promoter, gen P450 3A4 dan penggunaan target signal peptida perlu dioptimasi lebih lanjut agar dapat meningkatkan produksi senyawa (epi)podophyllotoxin. Strategi rekayasa genetika ini dapat juga diaplikasikan untuk jenis tanaman lain yang juga memproduksi DPT.

119

ac

kn

ow

leDg

eMen

ts

&ACKnoWLedgmentS

Finally this is it. Despites all challenges and struggling, I have finally completed my studies.

First of all, I would like to acknowledge Prof. dr. Wim J. Quax who welcomed me to his department. I thank you very much for your continued support throughout my studies, your critical views on my manuscripts and to let me join the IAPB 2010 conference. This conference was very significant to my subject and it was the highlight of my PhD studies. I would also like to thank Prof. dr. Oliver Kayser for this PhD project.

My deep gratitude goes to Prof. dr. Jacques Hille for his supervision, guidance, and patience in assisting me in the field of plant tissue cultures and the transformation work. I also thank you for your critical views on my manuscripts. I did enjoy the coffee breaks and our chat in the coffee room. I am grateful to Dr. Herman J. Woerdenbag for being my co-promotor because of his help organizing the paper works, the public defense ceremony, and translating the summary in Dutch language. You were always thorough with your feedback on my manuscripts. I do appreciate your structured writing, level of detail and critical views on my manuscripts.

I would also like to express my thanks to Prof. dr. J.T.M. Elzenga for his support and hospitality in plant physiology department and also for being one of the reading committee members. My special thanks go to the other members of the reading committee, Prof. dr. Robert Verpoorte and Prof. dr. E.M.J. Verpoorte for their time and remarks on my thesis.

I am grateful for Janita Zwinderman and Yvonne Vergnes, for their kind assistance on the administrative works during my PhD.

I thank Dr. Paul. J. Dijkwel, Marcel J. G. Sturre and Margriet Ferwerda for their intellectual and technical supports in the beginning of my research in the Department of Molecular Biology of Plant. Paul, I thank you for your help and supervision in the beginning of my research at MBP. Marcel, I thank you for your help that was available any time and your advice concerning plant works. Margriet, I thank you for your help and patience for ‘making my finger green.’ My special thanks go to the entire lab fellows at the MBP, Reza, Jos, Tsanko, Aukje, Bert, Ijaz, Sujeeth, Kamran, Eelco and Jan, I enjoyed a lot the free lab trips, baby visits, BBQ at Jacques’ house, cakes for any occasion and the ultimate Christmas dinner. I would also like to thank Jacob, Ger, and Martin for their help in providing the green house and climate room stuffs.

I thank people who collected the Anthriscus plants and sent it to me: Frank Klingenstein, Christina Birnbaum, Dalila Martins, Sigurdur H. Magnusson, Rene J.T. Cappers, Renata Alnsatter, Sigvardur Eirnarsson, and Bjarni E. Gudleifsson. Without your help, there would not be chapter 2. My special thanks go to Dalila for collecting the plants and the seeds. I also thank you for our friendship. To my students, Lisette and Behnaz, I thank you for doing some of the extraction works. My deep thanks go to Dr. Wouter Hinrichs for all the help with the freeze-drying when our machine was broken.

120

ac

kn

ow

leDg

eMen

ts

&I thank our collaborators, Prof. dr. Heribert Warzecha for the plasmids and the

Nicotiana works, Paul J.A. Michiels and Herald G. Aantjes for the NMR works, and Dr. Frank. J. Dekker for the DPT synthesis.

I am grateful to Jan W. Zwart, Hein Drost, Kees and Katja who helped me for the field experiment in De Kruidhof botanical garden in Buitenpost. Hein, hartelijk bedankt voor je advies en je hulp bij het groein van de planten. Ik begrijp je wel, maar Ik moet ‘heel veel’ denken als ik met je wil praten in het Nederlands. Het is goed om mijn nederlands te verbeteren.

I am thankful to Annie van Dam for her patience and her help on measuring my always ‘urgent samples.’

I thank all my colleagues in the Department of Pharmaceutical biology. Many people have come and gone, nevertheless, it was nice to meet you all. I always enjoyed ‘the cake moment.’ I thank the plant group people, Anna-Margareta, Remco, Torsten, Agata, Magda, Elena, Gany, Insanu, Nizar, Eko and the other groups as well, Gerrit, Robbert, Mariana, Anja, Carlos, Mark, Edzard, Jandre, Marianne, Vinod, Hans, Yufeng, Bert-Jan, Ellen, Gudrun, Ilse, Pol, Luis, Harshwardhan, Mehran, Jan Ytzen, Mariette, Aart, Dan, Ingrid, Anna, and Evelina. Special thanks go to Mariana Wahjudi for your support, tips and tricks in cloning and for our friendship. The highlight of the activities for me was the ski trip to Austria in 2009. Many thanks go to Gudrun who organized it.

I am grateful to all the technicians: Sieb, Rita, Ronald & Pieter. Rita, I thank you for your help during my accident and moving Anthriscus plants to Buitenpost. Ronald, I thank you for your support and help in ordering the chemicals.

I am very grateful to Christina Avanti and Melanie Foeh for being my paranymphs. I would like also to thank Melanie’s family, Peter and Michael for your friendship and support.

My HOST mates, Adri & Kelly, Byeong Ui & Su Jung, Anieljah, Roeland, Gineke, Ben, Corien, Ingrid, Rio, John, Willy, Anieljah, Sarah, Giulia, Jie Lin, Hong You, Nancy, Christian, Geert, Aga, Marije, Michelle, Trix, Martijn, Dirk-Jan, Elisabeth, Magda, Eline, Linda, Erik, Johnson, Lydia, Paulina, Olubunmi, Ivana, Paulina, Louise, Danica, Amy and many others, my deepest thanks go to you for your support, friendships, colorful, and enjoyable moments that we have shared together. I would like to thank Elissa & Nico, Petur & Maartje, Ajay & Kanika, for the BBQ, cooking, baking and eating activities together. Maartje, Kanika & Elissa, girls I did enjoy our shopping trips, as I was always the one who ended up with countless shopping bags. I also enjoyed watching movies together with you. My special gratitude goes to Hester Nijhoff, who designed the cover thesis.

My fellow PhD friends, Dr. Johana Nayoan, you are the first who got the degree. I thank you for your help in editing the Indonesian summary. My special thanks go to Tita Alissa Listyowardojo who made me register to the gym and now I’m kind of addicted to it and have become healthier, also for your valuable inputs on my CV and for all the talks. Special thanks go to Yunia Sribudiani for friendship, companionship, and for being a nice travel buddy, to Laura & Agnes for the sharing and friendships.

121

ac

kn

ow

leDg

eMen

ts

&My special thanks go to my fitness buddies, Maria, Alijca, Wictor, Ridwan, Justina,

Saras, Tita, Klara, Sita, Shinta, Lorenza, Agata for joining the classes and sometimes for having dinner together. Maria, I enjoyed talking to you and joining the body step or body jam classes together. Alijca & Wictor, I did enjoy the sailing very much, thanks a lot. Justina, I thank you for the meal invitations. To Ridwan, I thank you for the gym exercise together and sometimes dinner afterwards. Mbak Saras, Klara, Sita, Tita and Shinta, I thank you all for the gym companionship.

My sincere thanks go to the Indonesian community in Groningen (Shanti, Mira, Neng, Ari, Mas Yayok, Mbak Lia, teh Nisa, Kang Nandang, Adit, Pipit, Wangsa and many others) and the patrimonium group (Stephanie, Wisnu, Yuli, Tita, Dani, Meta, Laura, Daniel, Astrid, Bang Mangarai, Jerome, Kenzie, Agnes, Rosalina W., Mbak Roga, Mbak Ika, Mbak Erna and others). Special thanks go to Rosalina Wisastra (Ocha) who also helped me editing the Indonesian summary.

My very special thanks go to my parents, sister and brothers for all their supports and love.

Finally, my appreciation goes to all people who cannot be mentioned one by one for their contribution to my research projects and my personal life in Groningen. Thank you for everything and God bless you all.

“I believe in the sun, even when it is not shining.

I believe in love, even when I don’t feel it.

I believe in God, even when there is silence.” *

* Graffiti found in 1945 on the wall of a basement in Koln, Germany, where a Jewish believer is thought too have been hiding from the gestapo.