UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Aspects of physiology and trichome chemistry in the medicinal plant Tanacetum parthenium (L.) Schultz-Bip. Usher, Kevin Bernard 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-715493.pdf [ 10.81MB ]
Metadata
JSON: 831-1.0090789.json
JSON-LD: 831-1.0090789-ld.json
RDF/XML (Pretty): 831-1.0090789-rdf.xml
RDF/JSON: 831-1.0090789-rdf.json
Turtle: 831-1.0090789-turtle.txt
N-Triples: 831-1.0090789-rdf-ntriples.txt
Original Record: 831-1.0090789-source.json
Full Text
831-1.0090789-fulltext.txt
Citation
831-1.0090789.ris

Full Text

Aspects of Physiology and Trichome Chemistry in the Medicinal Plant Tanacetum parthenium (L.) Schultz-Bip. by Kevin Bernard Usher B . S c , Okanagan University Co l lege , 1994 A T H E S I S S U B M I T T E D IN P A R T I A L F U L L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E S T U D I E S D E P A R T M E N T O F B O T A N Y We accept this thesis as conforming to the required standard G.H.N. Towers, Supervisor (Botany, University of British Columbia) .E.P. Taylor, Co^ipen/ i sor (Botany, University of British Columbia) P.A. Bowen, Committee Member (Pacific Agriculture Research Center, Agriculture and Agri-Food Canada) A.D./vKala^s, Commit$e4v1ember (Botany, University of British Columbia) T H E UNIVERSITY O F BRITISH COLUMBIA September 2001 © Kevin Bernard Usher, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date S" Oct , Zoo( DE-6 (2/88) 11 ABSTRACT This study investigated aspects of physiology and terpenoid chemistry in feverfew, a medicinal plant used for migraine therapy. T h e sesqui terpene lactone parthenolide accumulates in feverfew shoots and is thought to contribute to feverfew's antimigraine activity. The first part of this study examined the effects of nitrogen application and irrigation on shoot yield and shoot parthenolide concentrat ion. Reduced shoot yield was observed under treatments of low nitrogen application and irrigation frequency. Leaf parthenolide concentration increased in plants grown with high nitrogen application rates and decreased with high irrigation rates. In the second part of this study, shoot yield, parthenolide concentration and tr ichome distribution were examined in response to developmental changes . Days longer than approximately 12 hours induced f lowering. Fever few grown under days shorter than 12 hours for extended periods remained in a vegetative stage and their leaves accumulated parthenolide in concentrat ions up to 10x that of flowering plants. Yie ld was lower in vegetative plants grown under short days but leaf to stem ratio was high. Glandu lar t r ichomes are the site of parthenolide biosynthesis and storage. Leaf parthenolide concentrat ion is related to glandular tr ichome densi t ies on leaf sur faces. Young leaves of vegetative plants have high tr ichome densi t ies while young leaves of f lowering plants have low tr ichome densit ies. Tr ichome densit ies dec reased with leaf expans ion and as density dec reased , parthenolide concentrat ion dec reased . The third part of this study investigated the two terpenoid biosynthetic pathways involved in parthenolide biosynthesis. Exper iments using 1 4 C and 1 3 C labeled substrates revealed that both the mevalonate ( M E V ) and methylerythritol phosphate ( M E P ) pathways contribute isoprene subunits to parthenolide biosynthesis. Two of the Ill three isoprene subunits in parthenolide were enr iched after feeding shoots 1- 1 3C-mevalonate and 2 - 1 3 C-acetate, an enrichment pattern consistent with the M E V pathway. Parthenol ide's enrichment pattern after feeding 2 - 1 3 C-pyruvate and 1 - 1 3 C-G lucose was consistent with isoprene contributions from both pathways. After feeding 2- 1 3 C-pyruvate however, there was a higher proportion of 1 3 C-enr i chment from the M E P pathway than from the M E V pathway. TABLE OF CONTENTS Abstract ii Tab le of Contents iv List of Tab les vii List of Figures viii Acknowledgements x Chapter 1: Genera l Introduction 1.1. Plant Natural Products 1 1.2. Terpenoids 4 1.3. Sesqui terpene lactones 10 1.4. Feverfew: Historical use and modern medic ine 13 1.5. Object ives 17 1.6. References 18 Chapter 2: Effects of irrigation and nitrogen application on feverfew shoot yield and parthenolide concentration 2.1. Introduction 26 2.2. Materials and Methods 2.2.1. Genera l methods 30 2.2.2. Field irrigation trial 33 2.2.3. Field irrigation and nitrogen appl icat ion trial 34 2.2.4. Greenhouse irrigation and nitrogen appl icat ion trial 36 2.3. Resul ts 2.3.1. Field irrigation trial: effects of irrigation f requency on parthenolide concentrat ion and plant growth ... 38 2.3.2. Field irrigation and nitrogen appl icat ion trial 39 2.3.3. Greenhouse irrigation and nitrogen appl icat ion trial 42 2.4. D iscuss ion 44 2.5. References 52 Chapter 3: Development and Regenerat ion 3.1. Introduction 55 3.2. Materials and Methods 3.2.1. Genera l methods 58 3.2.2. Field growing medium and fertigation trial 60 3.2.3. Greenhouse growing medium and fertigation trial 62 3.3. Resul ts 3.3.1. Field growing medium and fertigation trial 63 3.3.2. Greenhouse growing medium and fertigation trial 69 3.4. D iscuss ion 72 3.5. References 82 Chapter 4: Developmenta l effects on glandular t r ichomes and leaf chemistry 4 .1 . Introduction 85 4.2. Materials and Methods 87 4 .3 . Resul ts 4 .3 .1 . Parthenol ide variability during leaf development and flowering 89 4.3.2. Fever few glandular tr ichome development , density, and concentration 94 4.4. D iscuss ion 102 4.5. Re fe rences 108 Chapter 5: Biosynthet ic studies using 1 4 C and 1 3 C incorporation into Parthenol ide 5.1. Introduction 110 5.2. Materials and Methods 113 5.2.1. 1 4 C feeding experiments 114 5.2.2. 1 3 C feeding experiments 116 5.2.3. Extraction methods, parthenolide isolation, and N M R analys is 117 5.3. Resul ts 5.3.1. 1 4 C labeling of parthenolide 119 5.3.2. 1 3 C enriched parthenolide 120 5.4. D iscuss ion 126 5.5. Conc lus ion 136 5.6. References 137 vi Chapter 6: Genera l D iscuss ion 6.1. Overv iew 139 6.2. Future Resea rch Directions 143 6.3. References .' 146 Append ix 148 LIST OF TABLES vii Tab le 2.1. Parthenol ide concentration of feverfew leaves 43 days and 87 days after transplanting. 38 Tab le 2.2. Average whole plant and organ dry weights of field grown feverfew. 39 Tab le 2.3. Leaf water status in field grown feverfew measured at 3 am (night), 12 pm (mid-day), and 6 pm (evening). 40 Tab le 2.4. Dry leaf parthenolide content measured over a 3 month period in the field. 41 Tab le 2.5. Leaf and flower parthenolide concentrat ions and total parthenolide content per plant of s tems, leaves and f lowers. 42 Tab le 2.6. Leaf parthenolide concentration in feverfew leaves grown in the greenhouse under irrigation and nitrogen treatments. 43 Tab le 3.1. Field experiment treatments and abbreviat ions. 61 Tab le 3.2. Treatments in the greenhouse trial and abbreviat ions. 63 Tab le 3.3. Dry weight and dry to fresh weight ratios of feverfew plants grown in the field and greenhouse. 64 Tab le 3.4. Leaf water potential and osmot ic potential of g reenhouse and field grown plants measured at 3 p.m. (light) and 4 a.m. (dark). 66 Tab le 3.5. Ave rage leaf parthenolide concentration in g reenhouse and field-grown plants. 67 Tab le 3.6. Parthenol ide content at harvest in leaf, s tem, and f lower t issues, based on subsample analys is. 68 Tab le 4 .1 . Parthenol ide concentration in leaves of different ages from vegetative and reproductive shoots. 92 Tab le 4.2. Tr ichome density and parthenolide concentrat ion of leaves measured from the apex to the base o f the s tem in vegetat ive and reproductive plants. 100 Tab le 5.1. Incorporation of 1 4 C labeled substrates into parthenolide. 119 Tab le 5.2. 1 3 C - N M R ass ignments and enrichment of parthenolide after feeding 1 3 C-en r i ched substrates. 124 viii LIST OF FIGURES Figure 1.1. Mevalonate and methylerythritol phosphate routes to isoprene Biosynthes is . 5 Figure 1.2. The diversity of terpenoid biosynthesis showing examples of the compounds derived from this pathway 8 Figure 1.3. Major skeletal types of sesqui terpene lactones showing the common pathway through the germacranol ides. 9 Figure 1.4. Scann ing electron micrograph of non-glandular and glandular t r ichomes on the abaxial leaf surface of Tanacetum parthenium. 12 Figure 4 .1 . A glandular tr ichome derived from an epidermal cell with a subcut icular extracellular space where secretory cel ls secrete non-polar compounds for storage. 86 Figure 4.2. Greenhouse-grown feverfew leaf parthenolide concentration during development from vegetative to reproductive growth. 89 Figure 4 .3 . Leaf parthenolide concentration of greenhouse-grown feverfew over t ime. 91 Figure 4.4. H P L C chromatograms of leaf and flower extracts from feverfew shoots in the vegetative and reproductive s tages. 93 Figure 4 .5 . H P L C chromatograms of a composi te disk f lower tr ichome extract and a receptacle tr ichome extract. 94 Figure 4.6. Scann ing electron micrographs of feverfew leaf surface showing glandular and non-glandular t r ichomes. 96 Figure 4.7. Scann ing electron micrographs of feverfew f lower glandular t r ichomes on the floret petals of the inf lorescence. 97 Figure 4.8. Increased visibility of t r ichomes on sl ide preparations of leaf epidermal peals after drying. 99 Figure 4.9. Tr ichomes before and after treatment with dichloromethane. 101 Figure 5.1. The methylerythritol phosphate pathway and the mevalonate pathway to terpenoid biosynthesis. 111 Figure 5.2. 1 H - N M R spectra of parthenolide. 121 Figure 5.3. Carbon numbering of parthenolide and the predicted conformation of the three isoprene units. 121 Figure 5.4. 1 3 C N M R of parthenolide isolated from feverfew shoots fed with 2 - 1 2 C-meva lono lac tone or 2 - 1 3 C-meva lono lac tone . 122 Figure 5.5. 1 3 C enrichment patterns in parthenolide after feeding enr iched substrates. 125 Figure 5.6. G l u c o s e catabol ism. 127 Figure 5.7. 1 - 1 3 C-D-g lucose feeding experiment. Observed and predicted patterns of 1 3 C enrichment. Figure 5.8. 2 - 1 3 C-ace ta te feeding experiment. Observed and predicted patterns of 1 3 C enrichment. Figure 5.9. 2 - 1 3 C-meva lono lac tone feeding experiment. Observed and predicted patterns of 1 3 C enrichment. Figure 5.10. 2 - 1 3 C-pyruva te feeding experiment. Observed and predicted patterns of 1 3 C enrichment. ACKNOWLEDGEMENTS I extend my gratitude to Professor G . H. Neil Towers for his gu idance and support throughout this thesis. The unrestricted f reedom to explore plant physiology and phytochemistry in the lab and f ield, and the liberty to explore where I felt the research led, was invaluable to me. I give my sincerest thanks to Dr. Pat A . Bowen for giving me an early start in my career as a scientist, for support and encouragement throughout my thesis and the invaluable gu idance in crop physiology. Dr. Bowen was instrumental to the success of this thesis by providing the resources for field and greenhouse exper iments. I a lso thank my co-superv isor Pro fessor lain E. P. Taylor and committee member Pro fessor Anthony D. M. G l a s s for their advice and support throughout my thesis and particularly in the preparation and writing of this dissertat ion. Many thanks to the staff at the University of British Co lumb ia , Department of Chemist ry N M R laboratory and at the Faculty of Pharmaceut ica l Sc i ences L C - M S laboratory. I thank the many members of Dr. Towers research group for their involvement with my research and for the many wonderful memor ies. In particular I want to thank Zyta Abramowsk i the cornerstone of Dr. Towers laboratory, Dr. Jon P a g e , Eduardo Jove l , J i Y a n g , F iona Cochrane , Andres Lopez , and Keith Pa rdee . M y gratitude to Heidi Remp le and Brenda Frey for the numerous hours maintaining my field and greenhouse crops in A g a s s i , B . C . Th is research was supported by Natural Sc i ences and Engineer ing Reserch Counc i l of C a n a d a in the form of operating and equipment grants to Pro fessor G . H. N. Towers , and by Agriculture and Agr i -Food C a n a d a at the Paci f ic Agr i -Food Research Centre who provided the laboratory and horticultural facilit ies under the supervis ion of Dr. P. A . Bowen . Finally, I thank my wife Kathy Usher , for her unending support and encouragement to pursue my interests and dreams. 1 Chapter 1 Introduction to Phvtochemistry and Feverfew 1.1. PLANT NATURAL PRODUCTS It is est imated that 8 0 % of the known natural products are of plant origin (Rob inson, 1980; Swa in , 1974). W e are becoming increasingly aware that secondary compounds have important survival functions within plants including roles as plant defensive mechan isms against herbivory and infection, al lelopathic agents and protection from damage caused by U V radiation, oxidat ion, and free radicals. They a lso serve as photoreceptors and provide communicat ion between plants and other organisms through chemical or visual stimuli (Harborne, 1993). Biological ly active secondary metaboli tes interact with enzymes or other chemica ls to elicit responses that result in physiological changes at the cellular level. Ove r the past century, as people d iscovered these properties, interest in plant natural products has increased. Humans have exploited the abundance of secondary compounds as cosmet ics , medic ines, recreational drugs, pigments, perfumes, and pest ic ides (Buchanin et al., 1980). Consumpt ion of natural products can protect against free radical damage , oxidation, and many d i seases such as cancer . Recent ly, attention has focused on food plants due to f indings that certain grains, fruits, vegetables, and nuts contain important non-nutritional secondary compounds that contribute to human health. Foods with beneficial pharmacological properties are cal led nutraceuticals. The full s igni f icance of health benefits provided by nutraceuticals is not yet known. Phytochemica ls can also be 2 detrimental to human health. For example, the toxic g lycoalkalo ids a - s o l a n i n e and « -chaconine are present in potato tuber skins. W h e n the tubers are exposed to light they turn green and the glycoalkaloid concentration increases. So lan ine is an establ ished acetylchol ine esterase inhibitor, a key component of the nervous sys tem, and signs of neurological impairment have been recorded after ingestion of the toxin (Dalvi and Bowie , 1983). Thus , the more that is known about natural products in plants, the better we can develop safe and effective pharmaceut icals, nutraceuticals, and foods. Establ ishing the function of secondary metaboli tes in plants is often difficult. T h e s e chemica ls may have more than one function and c losely related compounds may have completely different functions. S o m e investigators have suggested that plant secondary compounds are waste products of metabol ism, storage compounds, or a means to keep primary pathways open in t imes of reduced growth (Swain, 1977). The large number and diversity of plant natural products cannot be d ismissed this easily. O n e of the underlying rationalizations for the alternative chemica l defense theory of secondary metaboli tes is that plants are sess i le and must be capable of defending themselves (Harborne, 1993). The rich diversity of secondary metabol i tes in the plant k ingdom, and the ability of plants to continue generating novel compounds , may have contributed greatly to their evolutionary s u c c e s s (Rausher , 1992). Chemica l defenses can exist throughout the plant in key locations such as specia l ized cel ls and compartments, or at specif ic s tages in the life cyc le. Biosynthesis of defensive compounds is regulated through either constitutive or inducible mechan isms or may be controlled temporally, changing the chemical complement over t ime (Gershenzon & Cro teau, 1990). Plants a lso protect themselves through physical barriers such as thick or waxy epidermis, hairs, thorns or sp ines that may be impregnated with chemica ls that are unpalatable or toxic. S o m e plants use combinat ions of these defense mechan isms 3 (Denno & McC lu re , 1983). For example, glandular hairs are spec ia l ized extrusions on the epidermis containing toxic or repellent chemica ls . The glandular t r ichomes rupture, releasing their contents on contact with insects or herbivores. After the attack is f inished the chemica ls may remain on the surface of the plant providing protection from infection or herbivore feeding. The tr ichomes thus provide both physical and chemical obstac les to phytopathic organisms. The secondary chemistry of a plant spec ies is often consistent among individuals. This is the foundation of chemotaxonomy, which uses spec ies-spec i f ic chemica ls in the identification of plants. However, natural product formation is inherently var iable and this may confound attempts to determine chemotaxonomic relat ionships (Harborne, 1993; Ca tes , 1987). It s e e m s that chemica l variation ensures reproductive s u c c e s s through natural select ion. By keeping the chemical mixture variable, plants increase probability of survival and evolutionary s u c c e s s . The factors influencing variation are genetically determined and but concentrat ions may be modulated by cl imate, predation, infection, competit ion, edaph ic factors, nutrition, or other environmental stimuli (Denno & McClu re , 1983). The three most prevalent c lasses of plant secondary compounds are the alkaloids, phenol ics, and terpenoids. The alkaloids are nitrogen containing, basic compounds , many of which are derived from the shikimate pathway. Phenol ics are derived from a combinat ion of the acetate and shikimate pathways and include f lavonoids (Mann, 1986). Terpenoids are synthesized two ways in plants, the first is through the c lass ica l mevalonate pathway via acetate, the second is through the methylerythritol phosphate ( M E P ) pathway which uses pyruvate and g lyceraldehyde-3-phosphate as substrates (Dewick, 1999; Rohmer , 1999; E isenre ich e r a / . , 1998). The 4 terpenoids are of interest due to their broad range of biological activities and potential usefu lness as medic ines or agrochemicals . 1.2. T E R P E N O I D S Terpenoids are ubiquitous and some are perceived as necessary for the existence of life (Rohmer, 1999). G lasby (1982) compi led a list of more than 10,000 structures known at that time. They form the largest group of natural products, now numbering between 15,000 and 22,000 character ized compounds (Rohmer et al., 1996; Ge rshenzon and Cro teau, 1991). In contrast, there are over 10,000 alkaloids character ized (Southon and Buck ingham, 1989) and approximately 8,000 phenol ics including the f lavonoids (Harborne, 1988). S ince ancient t imes terpenoids have been used in oi ls, perfumes, soaps , drugs, and pigments. The lower terpenoids, mono- and sesqui terpenes, are characteristically volatile and often have an odor (Gershenzon & Cro teau, 1990). For example, camphor is an aromatic monoterpene traded s ince the 1 1 t h century for its f ragrance (Banthorpe, 1991). Terpenoids are general ly unsaturated, lipophilic, and predominantly cycl ic compounds that may be highly oxygenated and contain var ious functional groups. These compounds have been studied s ince the 19 t h century when Wal lach put forward the isoprene rule of terpenoid biosynthesis (Fowler et al., 1999). The rule says that terpenoids are made up of repeating five carbon isoprene units joined together head to tail. Isopentenyl pyrophosphate (IPP) is the isoprene unit from which all terpenes are synthes ized. A s with all rules there are except ions and in this c a s e tail-to-tail and head-to-head condensat ion of isoprene units does occur, but are rare. T h e reactions of isoprene biosynthesis are shown in Figure 1.1. The details of the mevalonate pathway M E P p a t h w a y o OH + M e v a l o n a t e p a t h w a y -SCoA + Pyruvate G-3-P H 3 C acetyl-CoASH O H 3 C -SCoA C02 HcC C o A S H O SCoA + OH H M G S H 3 C SCoA H20 oASH 3 OH O • NADPH/H + NADP+ HOOC 2 NADPH/2H+ SCoA 2 NADP+ *d C o A S H ^ r H M G R H 3 C . OH 3 OH H O O C . > ^ s OH 3R-mevalonic acid (MVA) 2 ATP 2 A D P < 3 OH IPP C H 2 HOOC OPP 5-PPMVA O P P DMAPP F i g u r e 1.1. Two routes to isoprene biosynthesis. The mevalonate pathway and the methylerythritol phosphate pathway converge at isopentenyl pyrophosphate. H M G S = hydroxymethylglutaryl synthase, H M G R = hydroxymethylglutaryl reductase, 5 - P P M V A = 5-pyrophosphomevalonate, IPP = isopentenylpyrophosphate, D M A P P = dimethylal ly lpyrophosphate, G - 3 - P = g lycera ldehydes-3-phosphate, D X P = 1-deoxyxylulose-5-P, M E P = 2-C-methyl-D-erythri tol-4-P. 6 were worked out in the 1950's when it was d iscovered that in flax, rat liver and yeast preparat ions, ace ty l -CoA was the precursor to mevalonic ac id (MVA) and IPP (Fisher, 1999). Demonstrat ion of radiolabeled mevalonic acid incorporation into various plant terpenoids reinforced the evolving principle that the mevalonate pathway was the route to all terpenoids. S ince the elucidation of the M V A pathway in the 1950's there have been reports contradicting the universal role of M V A for terpenoid biosynthesis. Labeled M V A and acetate were not incorporated, or were incorporated in low levels, into monoterpenes, di terpenes, carotenoids and phytol, but were readily incorporated into sterols, tr iterpenes and somet imes into sesqui terpenes (Sagner et al., 1998; Lichtenthaler et al., 1996). Rhomer d iscovered that 1 3 C enr iched acetate incorporation into bacterial terpenoids resulted in unexpected patterns of enrichment. Subsequent investigations led him to the discovery of an alternative pathway to terpenoid biosynthesis cal led the methylerythritol phosphate ( M E P ) pathway (Rohmer et al., 1993; F lesch & Rohmer , 1989; F lesch & Rohmer, 1988). The precursors for this novel pathway are pyruvate and glyceraldehyde-3-phosphate (Figure 1.1). In eukaryotes the mevalonate ( M E V ) pathway functions in the cytosol while the M E P pathway functions in chloroplasts. (Arigoni et al., 1999; Paseshn ichenko , 1998; Disch et al., 1998; Putra et al., 1998; Rohmer et al., 1996). It was suspected that compartmental izat ion was playing a role in terpenoid distribution at the cellular level, but until the elucidation of the M E P pathway the plastidic contribution to terpenoid synthesis could not be determined (Disch e ra / . , 1998; Nabeta etai, 1998; Lichtenthaler et al., 1997a, 1997b). The regulation of all terpenoids begins with regulating isoprenoid biosynthesis. Hydroxymethylglutaryl reductase ( H M G R ) is the rate limiting step and regulation point in the M E V pathway controll ing the production of IPP (Bach et al., 1990). Al l s teps of the 7 M E P pathway have not been determined, including the rate-limiting step. Methylerythritol phosphate is the first committed metabolite in this pathway so the rate-limiting step may be at the formation of, or downstream from, this compound. The mevalonate and M E P pathways converge at IPP after which all the reactions likely proceed via enzymes of the same or similar structure. However, there has not been an investigation into the possibil ity that the enzymes downst ream from IPP may be different in the plastidic and the cytosol ic pathways. After IPP is synthes ized, it reacts with dimethylallyl pyrophosphate (DMAPP) forming a monoterpene. Sequent ia l condensat ion of IPP units forms the higher terpenoids (Figure 1.2) (Towers & Stafford, 1990). The cyc lase enzymes control the branch point of terpenoids into the subc lasses such as mono- or sesqui terpenes. Once cycl ization occurs , the molecule is committed and is usual ly not prenylated to form a higher c lass of terpenoid. For example, sesqui terpene lactones are composed of three isoprene units to form farnesyl pyrophosphate. Farnesyl cyc lase converts farnesyl into a ten membered ring after which it is committed to sesqui terpene biosynthesis (Figure 1.3) 8 M i x e d B i o s y n t h e s i s (c„) -Prenylatiorn— Po ly isoprenoids T i Plasto-, ubi-quinones Prenylation •<— DMAPP/IPP T Flavonoids, alkaloids ipp nucleicacid bases coumarins, proteins benzoquinones i -Prenylation <* GPP/NPP Benzo-, naptha-quinones ipp alkaloids, dlavonoids cannabinoids P u r e B i o s y n t h e s i s -> (C„) Gutta rubber Polyprenols (C5) Isoprene Monoterpenes (^10) iridoids Pseudoalkaloids -Prenylation Porphyrins IPP Esterification Porphyrins IPP DMAPP = dimethylallyl pyrophosphate IPP = isopenteny pyrophosphate GPP = geranyl pyrophosphate NPP = neryl pyrophosphate FPP = farnesy pyrophosphate GGPP = geranylgeranyl pyrophosphate GFPP = geranylfarnesyl pyrophosphate FPP FPP-GGPP _^ ( C . Sesquiterpene lactones ^ 1 5 ' Abscisic acid Sterols, brassins (C30) saponins, sardenolides pseudoalkaloids -> (C2o) Gibberellins 2 0 pseudoalkaloids GGPP-(C40) Carotenoids t GFPP (C25) Sesterterpenes F i g u r e 1.2. Terpenoid biosynthesis showing the diversity of this chemica l c lass . A few examples of the compounds derived from this pathway are given. Prenylat ion refers to the addition of the respect ive C5,Cio,Ci5, etc.. units to another non-terpenoid compound. Adapted from Towers and Stafford (1990). 9 Eremophilanolide F i g u r e 1.3. Major skeletal types of sesqui terpene lactones showing the common pathway through the germacranol ides. 10 1.3. S E S Q U I T E R P E N E L A C T O N E S The largest and most diverse c lass of terpenoids is the sesqui terpene lactones (STLs) . In 1979 the number of identified naturally occurr ing S T L s was 950, by 1987 it had tripled to 3200 (Fischer, 1990) and presently may exceed 5000. There are over 200 skeletal types (Fowler et al., 1999) but the majority of known compounds fall into 9 major skeletal c l asses . The majority of sesqui terpene lactones are formed by an initial cycl ization of farnesyl pyrophosphate to form a germacrene and then germacranol ides, which are modified to form the other skeletal types (figure 1.3). The 12,6-lactonization is represented in figure 1.3. The 12,8-lactonization is a lso common and occurs in all these skeletal types (Fischer et al., 1979). The obvious control point for S T L biosynthesis is at the branch point where farnesyl cycl izat ion forms germacrene (figure 1.3). The biosynthesis of S T L s has not been studied in enough detail to know how the pathways are regulated. The majority of S T L s have been isolated from the As te raceae where they are prevalent. There are many other plant famil ies containing S T L s including Acan thaceae , Amaran thaceae, Burseraceae, B o m b a c a c e a e , Cor iar iaceae, l l l ic iaceae, Magno l iaceae, Men ispermaceae , Lamiaceae , Lauraceae , Po lygonaceae , and Win teraceae. S T L s have also been found in the gymnosperms in the family C u p r e s s a c e a e as well as fungi, liverworts, and marine organisms (Fischer, 1990). Interest in S T L s has grown for two reasons: they are useful in chemotaxonomy, and they have a range of biological activities (Bruneton, 1995; F ischer et al., 1979). Many S T L s have been shown to be antiviral, antibacterial, antifungal, cytotoxic, and nematocidal as well as having other pharmacological activities (Bos et al., 1998; Beekman etal., 1998; Maruta et al., 1995; Woynarowsk i & K o n o p a , 1981; Hoffmann et al., 1977). They a lso have potential as herbic ides and insect ic ides due to their al lelopathic and antifeedant properties (Macias et al., 1999; Mac ias et al., 1996; Mac ias 11 et al., 1993; Mac ias et al., 1992). The S T L s have a bitter taste so they contribute to f lavour of foods. The bitter taste in the ch icons of chicory roots is due to S T L s (Peters et al., 1997). The S T L s have met with limited application as medic ines due to their inherent toxicity. The best known medical application of an S T L has been the development of the antimalarial drug, artemisinin, from Artemisia annua. This chemica l , in its purified form, is currently undergoing clinical trials. Artemisinin has very few known side effects unlike other antimalarials such as chloroquine and quinine. Plasmodium falciparum and Plasmodium vivax , the parasi tes responsib le for malar ia, have acquired resistance to many pharmaceut ical drugs in current use (Bruneton, 1995; K layman , 1985). From a chemical ecology perspect ive we know that organisms can adapt rapidly and develop resistance to chemica ls . Th is is apparent in the antibiotic resistance developed by bacteria, which has scientists around the world searching for new antibiotics. In malaria treatment, P. falciparum has not become resistant to artemisinin. Its use however, has been as a traditional plant preparation and not a pure drug. It will be interesting to watch this drug as its use in pure form increases in western medic ine to see if P . falciparum deve lops resistance to artemisinin. O n e important functional group in S T L s is the a-methylene-y- lactone which confers biological activity to these compounds. The methylene group can bind to sulphydryls such as those in cysteine residues of proteins to causes loss of function (Heptinstall et al., 1988; Heptinstall et al., 1987). Sesqu i te rpene lactones can cause allergic eczematous contact dermatitis and can cross sensi t ize a person to other S T L s , a direct effect of the lactone methylene (Spettoli et al., 1998; Lamminpaa et al., 1996; Burry, 1980; Rodr iguez et al., 1977; Schu l z et al., 1975). Other functional groups a lso give S T L s increased biological activity such as a 2,3-double bond, epox ides, peroxides, 12 and cyclopentenone groups (Yuuya et al., 1999; Beekman et al., 1998; Goren et al., 1996; Woerdenbag, 1986; Elissalde etai, 1983). STLs are not free in plant cells but are sequestered and cannot cause toxic damage. STLs are thought to be defensive compounds and appear to be located in strategic areas of the plant, such as laticifers, canals, and trichomes or exuded to external surfaces of the plant. The position and form of these structures can be important in the effectiveness of defensive mechanisms. They are generally near or on structures likely to be attacked by plant eating organisms. The forms of trichomes for example include glandular hairs, trigger hairs, root hairs, and scales that provide physical barriers and sometimes toxic chemical barriers. Trichomes on leaf surfaces can be divided into three categories, simple (unbranched), complex (branched), and glandular (Behnke, 1984). Glandular trichomes are easily accessible and easily removed or manipulated for biosynthetic studies (Tellez et al., 1999; Gershenzon et al., 1992; Gershenzon et al., 1987; Croteau & Johnson, 1984). A scanning electron micrograph of glandular and simple trichomes on the leaf surface of T. parthenium (Figure 1.4) shows the abundance of these structures. Figure 1.4. A) Simple and B) glandular trichomes on the abaxial leaf surface of A B Tanacetum parthenium. 13 1.4. FEVERFEW: Historical use and Modern medicine Peop le have been using plants as medic ines s ince antiquity. In his "Archidoxa" of the "Arcanum" written early in the 1 6 t h century, Pa race l sus referred to the need to d iscover the active component of a remedy or the "secret" of a treatment (Di Stephano, 1951). The search for active constituents in plants began in the 1780s with Schee le ' s work on organic ac ids (referred to by Sneader , 1985). In the early part of the 19th-century investigation into several wel l -known medicinal plants led to the discovery of a number of biologically active alkaloids. Morphine, atropine, papaver ine, and codeine were a few of the alkaloids d iscovered and subsequent ly became cornerstones of modern medic ine and remain among the most important pharmaceut ical compounds isolated from plants (Foye et al., 1995; Sneader , 1985). The sesqui terpene lactones have not been of the s a m e medicinal value as the alkaloids, but they also have not been studied as intensively. Many traditional herbal drugs are from the As te raceae and contain S T L s (Spring e r a / . , 1999; O 'Hara e r a / . , 1998; Heinr ich et al., 1998; Bruneton, 1995). Fever few (T. parthenium) is a plant in the As te raceae containing sesqui terpene lactones which have been shown to be effective for migraine prophylaxis in clinical trials (Murphy e r a / . , 1988). Fever few is a spec ies native to Western Europe and has been used medicinally for centuries. It now occurs in many p laces around the world due to its popularity as a garden plant and herbal remedy and its propensity to become weedy. It's classif ication has been changed several t imes starting with a move from the genus Rudbeckia to Matricaria ass igned by L innaeus, then to Leucanthemum parthenium (L.) Gren and Godron , to Pyrethrum parthenium (L.) S m . , then Chrysanthemum parthenium (L.) Bernh. , and finally, as it is known today, to Tanacetum parthenium (L.) Schul tz Bip. (Tutin et al., 1976). It is known as mutterkraut in Germany , feddygen fenyw in W a l e s , 14 San ta Mar ia in Latin Amer i ca as well as federfoy, noseb leed , midsummer daisy, featherfew, flirtwort, and bachelor 's buttons. One of the earl iest recorded uses of feverfew is found in Culpepper 's Complete Herbal written in 1649, where it was recorded as a cure for agues and headache. Fever few has been cal led the aspirin of the 1 8 t h century (Berry, 1984). In England there was a resurgence of interest in feverfew when in the early 1970's newspapers reported feverfew as a cure for migraine and arthritis. Thousands of people started taking this plant even though nothing was known about its pharmacology. Engl ish scientists began researching feverfew and it quickly became one of the best-studied medicinal plants on the market. It was found that feverfew leaves taken daily reduce both intensity and f requency of migraine attacks (Pattrick et al., 1989; Murphy et al., 1988; Johnson etal., 1985). The active compounds are bel ieved to be S T L s and the one in highest concentrat ion is parthenolide. Parthenol ide has been found at concentrat ions of up to 2 % of dry weight in aerial organs (Heptinstall et al., 1992b; Awang et al., 1991). Thus the focus of pharmacological research on feverfew has been on parthenolide (Knight, 1995; A b a d et al., 1995). However, differing b ioassay results have been reported for pure parthenolide, different plant extracts, and fresh vs . dry plant preparat ions (Barsby et al., 1993; Groenewegen & Heptinstall, 1990; R o s s et al., 1999; Bejar, 1996; Barsby et al., 1992) . The problem in identifying the pharmacological agent(s) of feverfew may lie in the approach. Many scientists look for a single compound responsible for the pharmacological activity. However, it is quite possib le that the activities of this and other herbal drugs lie within the mixture of compounds present in the plant. More than 25 S T L s belonging to three c lasses have been isolated from feverfew leaves (Maries, 2000), as well as biologically active f lavonoids (Wil l iams et al., 1999; Smith & Burford, 1993) and monoterpenes (Knight, 1995). Parthenol ide, however, was investigated 15 because it is easy to isolate, is avai lable in large quantit ies in feverfew, and has a range of biological activities. In vivo, feverfew inhibits platelet aggregation (Losche et al., 1987) and the A D P -or adrenal ine- induced re lease of serotonin (Bejar, 1996; Mar ies et al., 1992). It inhibits the degranulat ion of granulocytes (Hayes & Foreman, 1987; E l issa lde et al., 1983), inhibits the re lease of enzymes involved in the inflammatory process (Jain & Kulkarni, 1999; Hehner er al., 1999; Makheja & Bai ley, 1981), exhibits a protective effect on vascu lar endothelial cells (Voyno-Yasenetskaya et al., 1988) and blocks voltage-dependent potassium channels (Barsby et al., 1993). Many of these effects are related to current knowledge about the physiology of migraine but direct ev idence of specif ic chemica ls in feverfew responsible for migraine prophylaxis has not been establ ished and may not be until migraine physiology is understood. The quality of many plant drugs is a s s e s s e d by the content of specif ic chemicals . There are two approaches to rating quality. The first is to use chemotaxonomy which is the use of chemica l traits in the identification of a plant spec ies . The second is to determine the presence and quantity of the compounds responsib le for the attributed physiological response or medicinal properties. However, these two methods are often confused and chemotaxonomic markers become promoted as medicinal compounds. B e c a u s e the use of chemical profiles for spec ies identification does not assure medicinal quality of a plant, the best method for quality control is a combinat ion of these two approaches (Bruneton, 1995). In feverfew, parthenolide is used as a taxonomic marker and is thought to be active against migraine. In this c a s e the chemotaxonomic marker and an active compound are the same. Most over-the-counter (OTC) herbal drugs have not undergone rigorous clinical testing to determine which chemica ls are the bioactive principles and whether the remedies are effective or dangerous. In addition, 16 secondary compounds in plants vary within spec ies which makes it difficult to evaluate the medicinal quality of an unpurified product. There are reports of substantial chemical variation in commerc ia l preparations of O T C plant drugs, including feverfew (Heptinstall et al., 1992a; Groenewegen & Heptinstall, 1986). This has raised concern over both the eff icacy and safety of unproven and untested herbal remedies. In order to reduce variability of the chemica ls of interest and to retain a consistent reliable product it's important to understand the cause of these chemica l variations. The use of parthenolide as a marker compound to authenticate feverfew products is common and it is suggested by Health C a n a d a that feverfew products contain a minimum of 0 .2% parthenolide (Bruneton, 1995). However , without the use of another chemical marker or additional taxonomic information, the authentication of feverfew using parthenolide may be misleading because there are a number of plants which produce parthenolide (Hendricks & Bos , 1990; F ischer e r a / . , 1979; Hoffmann et al., 1977; Wiedhopf et al., 1973) which could potentially be mistaken for, or used as an adulterant in commerc ia l feverfew preparations. O n e of these plants, Tanacetum vulgare (common tansy) contains parthenolide and has been used as an adulterant in feverfew (Heptinstall et al., 1992a)(Smith, 1994; Mit ich, 1992; Hendr icks & B o s , 1990). T. vulgare is a common weed spec ies and contains the monoterpene thujone which is a potent neurotoxin caus ing epileptic and tetanic type seizure (Bruneton, 1995). Not only do we need to ensure that authentic plant spec ies are being used in the correct dosage , but tests for possible contaminating plant material (weeds or adulterants), such as common tansy in feverfew crops, are also necessary . The premise of this thesis is that by determining the c a u s e s of parthenolide variation, the location of synthesis and parthenolide accumulat ion within feverfew, and the biosynthetic origins of parthenolide, we can better understand sesqui terpene 17 metabol ism in plants. This knowledge may be used to grow higher quality crops by better understanding how S T L levels vary in plants. Th is information is important if regulations are to be establ ished for achieving safe herbal drugs. Finally, elucidation of the M E P pathway in sesqui terpene biosynthesis must be establ ished before successfu l manipulation of the biosynthetic pathways through molecular techniques, in t issue culture and through horticultural methods can be utilized to their fullest potential. 1.5. O B J E C T I V E S I am interested in the fifteen carbon terpenoids cal led sesqui terpenes and how development, cl imate, edaphic factors, water and nitrogen affect variability in these compounds . My thesis is that these external factors affect plant growth, in addition to variability, local izat ion, and biosynthesis of the sesqui terpene, parthenolide, in the medicinal plant Tanacetum parthenium (L.) Schu l tz Bip. and that biosynthesis occurs through two independent pathways. The research in this thesis examined characterist ics of tr ichome chemistry and plant growth in the medicinal plant Tanacetum parthenium (feverfew). There were three areas of investigation; (1) The influence of water, nitrogen, photoperiod, growing media, and regeneration after harvesting on chemical variation and plant growth in both greenhouse and field culture, (2) Parthenol ide localization and chromatographic profiles of glandular tr ichome contents in leaves and f lowers throughout plant development, and (3) The isoprene biosynthetic route to parthenolide, a c losely related compound to germacrene and the first committed step in S T L biosynthesis. 18 1.6 R E F E R E N C E S A b a d , M .J . , Bermejo, P. & Vil lar, A . (1995) A n approach to the genus Tanacetum (Composi tae) - a phytochemical and pharmacological review. Phytotherapy Research, 9:79-92. Ar igoni , D., E isenre ich , W. , Latzel , C , Sagner , S . , Radykewicz , T., Zenk, M.H. and Bacher , A . (1999).Dimethylallyl pyrophosphate is not the committed precursor of isopentenyl pyrophosphate during terpenoid biosynthesis from 1-Deoxyxylulose in higher plants. Proceedings ofthe National Academy of Sciences ofthe United States of America, 96:1309-1314. A w a n g , D .V .C . , Dawson , B.A., Kindack, D .G. , Crompton, C . W . and Heptinstall, S . (1991) Parthenol ide content of feverfew (Tanacetum parthenium) a s s e s s e d by H P L C and 1 H - N M R spectroscopy. Journal of Natural Products, 54:1516-1521. B a c h , J . , Weber , T. and Motel , A . (1990) S o m e properties of enzymes involved in the biosynthesis of 3-hydroxy-3-methylglutaryl-CoA in plants. In: Recent advances in phytochemistry; Biochemistry o f t he mevalonic acid pathway to terpenoids, pp. 1-82. P lenum P r e s s , New York. , Eds . Stafford and Towers , G . H . N . Banthorpe, D.V. (1991) Classi f icat ion of terpenoids and general procedures for their character izat ion. In: Methods in plant biochemistry, vol . 7., Terpenoids. S a n Diego, C a . , A c a d e m i c Press . , Eds . Char lwood, B. V . and Banthorpe D. V . Barsby, R.W. , Knight, D.W. and M c F a d z e a n , I. (1993) A chloroform extract o f t he herb feverfew blocks vol tage-dependent potassium currents recorded from single smooth musc le cel ls. Journal of Pharmacy and Pharmacology, 45:641-645. Barsby, R.W. , S a l a n , U., Knight, D.W. and Hoult, J . R . (1992) Fever few extracts and parthenolide irreversibly inhibit vascu lar responses of the rabbit aorta. Journal of Pharmacy and Pharmacology, 44:737-740. B e e k m a n , A . C . , Wie renga , P.K. , Woerdenbag , H.J . , V a n Uden , W . , P ras , N., Konings, A .W.T . , E l -Fera ly , F .S . , Ga la l , A . M . and Wikst rom, H.V. (1998) Artemisinin-derived sesqui terpene lactones as potential antitumour compounds : cytotoxic action against bone marrow and tumour cel ls. Planta Medica, 64:615-619. Behnke , D.H. (1984) Plant t r ichomes - structure and ultrastructure: Genera l terminology, taxonomic appl icat ions, and aspects of t r ichome-bacter ia interaction in the leaf tips of dioscorea. In: Biology and chemistry of plant t r ichomes. pp. 1-21 . P lenum P r e s s , New York., Eds . E. Rodr iguez, P. L. H e a l e y a n d I. Mehta. Bejar, E. (1996) Parthenol ide inhibits the contractile responses of rat s tomach fundus to fenfluramine and dextroamphetamine but not serotonin. Journal of Ethnopharmacology, 50:1-12. Berry, M. (1984) Feverfew. Pharmaceutical Journal., 232 , 611-614. 19 B o s , R., Hendr iks, H., Scheffer, J . J . C . and Woerdenbag , H .J . (1998) Cytotoxic potential of valer ian constituents and valerian tinctures. Phytomedicine, 5:219-225. Bruneton, J . (1995) Pharmacognosy , phytochemistry, medic inal plants, pp. 500-510. Lavo iser Publ ish ing, New York. E d . Bruneton, J . Buchan in , R.A. , Otey, F .H . and Bagby, M. (1980) Botanochemica ls . In: Recent advances in phytochemistry. vol. 14., The resource potential in phytochemistry. pp. 1-22. P lenum Press , New York. Burry, J . N . (1980) Compos i tae dermatitis in South Austra l ia : contact dermatitis from Chrysanthemum parthenium. Contact Dermatitis, 6, 445 . Ca tes , R . G . (1987) Influence of biological rhythms, t issue development, and physiological state of plants and insects on their interactions. In: Insects-Plants: Proceed ings of the 6th international sympos ium on plant-insect relationships. E d s . Dordrecht, W. , Labeyr ie, V . , Fabres , D., and Lacha ise , D. Cro teau , R. and Johnson , M.A. (1984) Biosynthesis of terpenoids in glandular t r ichomes. In: Biology and chemistry of plant t r ichomes. pp. 133-185. P lenum P r e s s , New York. E. Rodr iguez, P. L. Hea ley and Mehta , I. eds . Denno, R .F . and McC lu re , M .S . (1983) Variabil i ty in plants and herbivores in natural and managed sys tems, pp. 717. A c a d e m i c P r e s s , New York. Denno, R. F. and McC lu re , M. S . eds. Dewick, P . M . (1999) The biosynthesis of C - 5 - C - 2 5 terpenoid compounds . Natural Product Reports, 16:97-130. Di S tephano, V . (1951) Australian journal of medical herbalism., 6:5. D isch , A . , Schwender , J . , Muller, C , Lichtenthaler, H.K. and Rohmer , M. (1998) Distribution of the mevalonate and glyceraldehyde phosphate/pyruvate pathways for isoprenoid biosynthesis in unicellular a lgae and the cyanobacter ium Synechocys t is P C C 6714. Biochemical Journal, 333:381-388. E isenre ich , W. , Schwarz , M., Car tayrade, A . , Ar igoni , D., Zenk, M .H . and Bacher , A . (1998) The deoxyxylu lose phosphate pathway of terpenoid biosynthesis in plants and microorganisms. Chemistry and Biology, 5 :R221 -R233 . E l issa lde, M.H. , Ivie, G .W. , Rowe, L.D. and El issa lde, G . S . (1983) Considerat ions of the structure of sesqui terpene lactones on biological activity: inf luence of the a lpha-methylene-gamma-lactone moiety on mast cell degranulat ion. American Journal of Veterinary Research, 44:1894-1897. F ischer, N.H. (1990) Sesqui terpene lactones: b iogenesis and biomimetic transformations. In: recent advances in phytochemistry, Vo l . 24. pp. 161-201., P lenum P r e s s , New York. Towers , G . H. N. and Stafford, H. A . eds . 20 Fischer , N.H. , Olivier, E . J . and Fischer, H.D. (1979) The b iogenesis and chemistry of sesqui terpene lactones. In: Progress in the chemistry of organic natural products. W i e n , Spr inger-Ver lag. , Eds . Herz, W. , Gr i sebach , H., and Kirby, G . W . Fisher, J . E., Rogers , M. J . , Halasy, J . M. , Luckman, S . P. , Hughes , D. E. , Masarach ia , P. J . , Weso lowsk i , G . , Russe l l , R. G . , R o d a n , G . A . , and R e s z k a , A . A . (1999) Alendronate mechan ism of act ion: geranylgeraniol , an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and k inase activation in vitro. Proceedings of the National Academy of Science, 96:133-138. F lesch , G . and Rohmer , M. (1988) Prokaryotic hopanoids: the biosynthesis of the bacter iohopane skeleton. Formation of isoprenic units from two distinct acetate pools and a novel type of carbon/carbon l inkage between a triterpene and D-r ibose. European Journal of Biochemistry, 175:405-411. F lesch , G . and Rohmer , M. (1989) Prokaryotic triterpenoids. A novel hopanoid from the ethanol-producing bacterium Zymomonas mobilis. Biochemical Journal, 262:673-675. Fowler, D .J . , Hamil ton, J . T . G . , Humphrey, A . J . and O 'hagan, D. (1999) Plant terpene biosynthesis. The biosynthesis of linalyl acetate in Mentha citrata. Tetrahedron Letters, 40 , 3803-3806. Foye , W . O . , Lemke, T.L . and Wi l l iams, D.A. Pr inciples of medicinal chemistry. (1995) Wi l l iams and Wi lk ins, Balt imore., Eds . Foye , W . O. , Lemke , T. L., and Wi l l iams, D. A . G e r s h e n z o n , J . and Croteau, R. (1990) Regulat ion of monoterpene biosynthesis. In: Recent advances in phytochemistry., vol . 24., Biochemistry o f t he mevalonic acid pathway to terpenoid biosynthesis, pp. 99-160, P lenum P r e s s , New York. Tower, G . H. N. and Stafford, H. A . (Eds.) G e r s h e n z o n , J . and Croteau, R. (1991) Terpenoids. In: Herbivores, Their Interactions With Secondary Plant Metabol i tes, Vo lume 1. pp. 165-219. A c a d e m i c P ress , S a n Diego. Rosentha l , G . A . and Berenbaum M.R. (Eds.) . G e r s h e n z o n , J . , Duffy, M.A., Karp, F. & Cro teau, R. (1987) Mechan ized techniques for the select ive isolation of enzymes from plant epidermal g lands. Analytical Biochemistry, 163:159-164. G e r s h e n z o n , J . , Mccask i l l , D., Rajaonarivony, J. I .M., Mihal iak, C . , Karp, F. and Croteau, R. (1992) Isolation of secretory-cel ls from plant glandular t r ichomes and their use in biosynthetic-studies of monoterpenes and other gland products. Analytical Biochemistry, 200:130-138. G lasby , J . S . (1982) Encyc loped ia o f t h e alkaloids. John Wi ley, Chichester . G lasby , J . S . (Ed.) 21 G o r e n , N., Woerdenbag , H.J . and Bozok johansson, C . (1996) Cytotoxic and antibacterial activities of sesqui terpene lactones isolated from Tanacetum praeteritum subsp . praeteritum. Planta Medica, 62:419-422. G roenewegen , W . A . and Heptinstall, S . (1986) Amounts of feverfew in commercia l preparat ions of the herb. Lancet, 1:44-45. G roenewegen , W . A . and Heptinstall, S . (1990) A compar ison of the effects of an extract of feverfew and parthenolide, a component of feverfew, on human platelet activity in-vitro. Journal of Pharmacy and Pharmacology, 42:553-7. Harborne, J . B . (1988) The f lavonoids: Advances in research s ince 1980. C h a p m a n and Hal l , London. Harborne, J . B . (1993) Ecologica l biochemistry. A c a d e m i c P r e s s , S a n Diego. Hayes , N.A. and Foreman, J . C . (1987) The activity of compounds extracted from feverfew on histamine re lease from rat mast cel ls. Journal of Pharmacy and Pharmacology, 39:466-470. Hehner, S . P . , Hofmann, T . G . , Droge, W . and Schmi tz , M.L. (1999) The anti inf lammatory sesqui terpene lactone parthenol ide inhibits N F - kappa B by targeting the I kappa B kinase complex. Journal of Immunology, 163:5617-5623. Heinr ich, M. , Rob les , M. , West , J . E . , Demontel lano, B . R . O . and Rodr iguez, E. (1998) Ethnopharmacology of mexican asteraceae (Composi tae) . Annual Review of Pharmacology & Toxicology 38:539-565. Hendr icks, H. and Bos , R. (1990) The presence of parthenolide in dutch tansy. Planta Medica, 56:540. Heptinstal l , S . , A w a n g , D.V., Dawson , B.A., K indack, D., Knight, D.W. and May, J . (1992) Parthenol ide content and bioactivity of feverfew (Tanacetum parthenium (L.) Schultz-bip.) . Estimation of commerc ia l and authenticated feverfew products. Journal of Pharmacy and Pharmacology, 44:391-395. Heptinstal l , S . , G roenewegen , W.A . , Spangenberg , P. and L o e s c h e , W . (1987) Extracts of feverfew may inhibit platelet behaviour via neutral ization of sulphydryl groups. Journal of Pharmacy and Pharmacology, 39:459-465. Hoffmann, J . J . , Torrance, S . J . , Widehopf, R . M . and Co le , J . R . (1977) Cytotoxic agents from Michel ia champaca and Ta lauma ovata: parthenolide and costunol ide. Journal of Pharmaceutical Science, 66:883-884. Ja in , N.K. and Kulkarni , S .K . (1999) Ant inocicept ive and anti-inflammatory effects of Tanace tum parthenium L. extract in mice and rats. Journal of Ethnopharmacology, 68:251-259. Johnson , E . S . , K a d a m , N.P. , Hy lands, D .M. and Hylands, P . J . (1985) Eff icacy of feverfew as prophylactic treatment of migraine. British Medical Journal (Clinical Research Edition), 291:569-573. 22 K layman , D.L. (1985) Q inghaosu (artemisinin): an antimalarial drug from Ch ina . Science, 228:1049-1055. Knight, D.W. (1995) Feverfew: chemistry and biological activity. Natural Product Reports, 12:271-276. Lamminpaa , A . , Est lander, T., Jo lanki , R. and Kanerva L. (1996) Occupat ional allergic contact dermatitis caused by decorat ive plants. Contact Dermatitis., 34:330-335. Lichtenthaler, H.K., Rohmer , M. and Schwender , J . (1997a) Two independent b iochemical pathways for isopentenyl d iphosphate and isoprenoid biosynthesis in higher plants. Physiologia Plantarum, 101:643-652. Lichtenthaler, H.K., Schwender , J . , D isch , A . and Rohmer , M. (1997b) Biosynthesis of isoprenoids in higher plant chloroplasts proceeds via a mevalonate- independent pathway. FEBS Letters, 400:271-274. Lichtenthaler, H.K., Schwender , J . , S e e m a n n , M. and Rohmer , M. (1996) Biosynthesis of carotenoids and other isoprenoids via a non-mevalonate pathway in Scendesmus obliquus. Plant Physiology, 111:588. Losche , W. , Mazurov, A . V . , Heptinstall, S . , G roenewegen , W . A . , Rep in , V . S . and Til l, U. (1987) A n extract of feverfew inhibits interactions of human platelets with col lagen substrates. Thrombosis Research, 48:511-518. Mac ias , F.A., Gal indo, J . C . G . and Massanet , G . M . (1992) Potential al lelopathic activity of several sesqui terpene lactone models. Phytochemistry 31, 1969-1977. Mac ias , F.A., Ol iva, R . M . , Vare la , R . M . , Torres, A . and Molini l lo, J . M . G . (1999) A l le lochemica ls from sunflower leaves cv. peredovick. Phytochemistry, 52:613-621 . Mac ias , F.A., Torres, A . , Molinil lo, J . M . G . , Va re la , R . M . and Caste l lano, D. (1996) Potential al lelopathic sesqui terpene lactones from sunflower leaves. Phytochemistry, 43:1205-1215. Mac ias , F.A., Va re la , R . M . , Torres, A . and Molinil lo, J . M . G . (1993) Potential al lelopathic guaianol ides from cultivar sunf lower leaves, var Sh -222 . Phytochemistry, 34:669-674. Makhe ja , A . N . and Bai ley, J . M . (1981) The active principle in feverfew. Lancet, 2:1054. M a n n , J . Secondary metabol ism. (1986) Oxford University P r e s s , N e w york. Mar ies , R . J . (2000) Persona l communicat ion. Mar ies , R . J . , Kaminsk i , J . , A rnason , J.T. , P a z o s - S a n o u , L., Heptinstal l , S . , F ischer, N.H. , Crompton, C .W. , Kindack, D .G . and A w a n g , D.V. (1992) A b ioassay for inhibition of serotonin re lease from bovine platelets. Journal of Natural Products, 55:1044-1056. 23 Maruta, Y . , Fukush i , Y . , Ohkawa , K., Nakanish i , Y . , Tahara , S . and Mizutani , J . (1995) Antimicrobial stress compounds from Hypochoeris radicata. Phytochemistry, 38:1169-1173. Mit ich, L.W. (1992) Tansy . W e e d Technology 6:242-244. Murphy, J . J . , Heptinstal l , S . and Mitchell , J . R . (1988) Randomized double-bl ind placebo-control led trial of feverfew in migraine prevention. Lancet, 2:189-192. Nabeta , K., Sai toh, T., Adach i , K. and Komuro, K. (1998) Biosynthesis of phytyl s ide-chain of chlorophyll a : Apparent reutilization of carbon dioxide evolved during acetate assimilat ion in biosynthesis of chloroplast idic isoprenoid. Chemical Communications, 6:671-672. O 'Hara , M. , Kiefer, D., Farrel l , K. and Kemper , K. (1998) A review of 12 commonly used medicinal herbs. Archives of Family Medicine, 7:523-536. Pachter , H .M. (1951) Magic into Science. Henry S c h u m a n Inc., New york, Pachter, H .M. (Ed.) Paseshn ichenko , V . A . (1998) A new alternative non-mevalonate pathway for isoprenoid biosynthesis in eubacter ia and plants. Biochemistry 63:139-148. Pattrick, M. , Heptinstal l , S . and Doherty, M. (1989) Fever few in rheumatoid arthritis: A double blind, p lacebo controlled study. Annals of Rheumatoid Diseases, 48:547-549. Peters, A . M . , H a a g s m a , N. and Vanamerongen , A . (1997) A pilot study on the effects of cultivation condit ions of chicory (Cichoruium intybus L.) roots on the levels of sesqui terpene lactones in ch icons. Food Research and Technology, 205:143-147. Putra, S . R . , Lo is , L .M. , C a m p o s , N., Boronat, A . and Rohmer , M. (1998) Incorporation of [2,3-C-13(2)]- and [2,4-C-13(2)]-D-1-Deoxyxylulose into ubiquinone of Escherichia coli via the mevalonate- independent pathway for isoprenoid biosynthesis. Tetrahedron Letters, 39:23-26. Rausher , M.D. (1992) Natural select ion and the evolution of plant insect interaction. In: Insect chemica l ecology, pp. 20-88 C h a p m a n and Hal l , N e w York. Roi tberg, B. D. and Isman, M. B. (Eds.) Rob inson , T. Organic constituents of higher plants. (1980) N. Amherst , Mass . , Cordus P r e s s . Rodr iguez, E. , Epste in , W . L . and Mitchell, J . C . (1977) The role of sesqui terpene lactones in contact hypersensitivity to some North and South Amer ican spec ies of feverfew (Parffren/ i /m-Compositae). Contact Dermatitis, 3:155-162. Rohmer , M. (1999) The discovery of a mevalonate- independent pathway for isoprenoid biosynthesis in bacteria, a lgae and higher plants. Natural Product Reports, 16:565-574. 24 Rohmer , M. , Knani , M. , S imonin , P. , Sutter, B. and S a h m , H. (1993) Isoprenoid biosynthesis in bacteria: a novel pathway for the early s teps leading to isopentenyl d iphosphate. Biochemical Journal, 295:517-524. Rohmer , M. , S e e m a n n , M., Horbach, S . , Br ingermeyer, S . and S a h m , H. (1996) Glycera ldehyde 3-phosphate and pyruvate as precursors of isoprenic units in an alternative non-mevalonate pathway for terpenoid biosynthesis. Journal of the American Chemical Society, 118:2564-2566. R o s s , J . J . , A rnason , J.T. and Birnboim, H.C. (1999) Low concentrat ions o f t he feverfew component parthenolide inhibit in vitro growth of tumor l ines in a cytostatic fashion. Planta Medica, 65:126-129. Sagner , S . , Latzel , C , E isenre ich, W. , Bacher , A. and Zenk, M .H . (1998) Differential incorporation of 1-deoxy-D-xylulose into monoterpenes and carotenoids in higher plants. Chemical Communications, 2 :221-222. Schu lz , K .H . , Hausen , B .M. , Wallhofer, L. and Schmidt-Loff ler, P. (1975) Chrysan themum allergy. Pt. II: Experimental studies on the causat ive agents. Archive of Dermatologia Forschricht, 251:235-244. Smi th, R . M . and Burford, M.D. (1993) Compar ison of f lavanoids in feverfew varieties and related spec ies by principal components-analys is . Chemometrics and Intelligent Laboratory Systems, 19:133. Smi th, R . M . and Burford, M.D. (1994) G L C of supercrit ical fluid extracts of essent ial oils from the medicinal herbs, feverfew, tansy, and german chamomi le . Journal of Chromatographic Sc ience . 32:265-269. Sneader , W . (1985) Drug discovery: The evolution of modern medic ines. John Wi ley & S o n s , Chichester . Southon, I.W. and J . Buck ingham (1989). Dictionary of alkaloids. C h a p m a n and Hal l , London. Spettol i , E. , Si lvani , S . , Lucente, P. , Guer ra , L. and V incenz i , C . (1998) Contact dermatit is caused by sesqui terpene lactones. American Journal of Contact Dermatitis, 9:49-50. Spr ing, O. , Hei l , N. and E l iasson , U. (1999) Chemosys temat i c studies on the genus Scalesia (Asteraceae). Biochemical Systematics and Ecology, 27:277-288. Swa in , T. (1974) B iochemica l evolution in plants. Comparative Biochemistry 29:595-622. Swa in , T. (1977) Secondary compounds as protective agents. Annual Review of Plant Physiology, 28 :479-501. Tel lez, M.R., C a n e l , C , R imando, A.M. and Duke, S . O . (1999) Differential accumulat ion of isoprenoids in glanded and g landless Artemisia annua L. Phytochemistry, 52:1035-1040. 25 Towers , G . H . N , and Stafford, H.A. (1990) Biochemistry of the mevalonic acid pathway to terpenoids. In: Recent advances in Phytochemistry, vol . 24., P lenum Press , N e w York, Towers , G . H. N. and Stafford H. A . (Eds.) Tutin, T . G . , Heywood, V . H . , Burges, N.A., Moore , D .M. , Valent ine, D.H. , Wal ters, S . M . and W e b , D.A. (1976) Flora Europaea . Cambr idge University P ress , Cambr idge . Voyno -Yasene tskaya , T.A. , Loesche , W. , Groenewegen , W . A . , Heptinstal l , S . , Rep in , V . S . and Til l , U. (1988) Effects of an extract of feverfew on endothelial cell integrity and on c A M P in rabbit perfused aorta. Journal of Pharmacy and Pharmacology, 40:501-502. Wiedhopf, R . M . , Y o u n g , M. , Bianchi , E. and Co le , J . R . (1973) Tumor inhibitory agent from Magnolia grandiflora (Magnol iaceae). I. Parthenol ide. Journal of Pharmaceutical Science, 62:345. Wi l l iams, C .A . , Harborne, J . B . , Geiger , H. and Hoult, J . R . (1999) The f lavonoids of Tanacetum parthenium and T. vulgare and their anti- inflammatory properties. Phytochemistry, 51:417-23. Woerdenbag , H.J . (1986) Eupatorium cannabinum L A review emphas iz ing the sesqui terpene lactones and their biological activity. Pharmaceutical Weekly 8:245-251. Woynarowsk i , J . M . and Konopa , J . (1981) Inhibition of D N A biosynthesis in H e L a cells by cytotoxic and antitumor sesqui terpene lactones. Molecular Pharmacology, 19:97-102. Y u u y a , S . , Hagiwara, H., Suzuk i , T., Ando , M. , Y a m a d a , A . , S u d a , K., Ka taoka , T. and Naga i , K. (1999) Guaiano l ides as immunomodulators. Synthes is and biological activities of dehydrocostus lactone, mokko lactone, eremanthin, and their derivatives. Journal of Natural Products, 62:22-30. 26 Chapter 2 Effects of irrigation frequency and nitrogen application on feverfew shoot yield and parthenolide concentration 2.1 . I N T R O D U C T I O N Concentrat ions of terpenoids may vary within a plant spec ies . C h a n g e s in secondary metabol ism in response to environmental factors appear to be spec ies -specif ic traits (Gershenzon , 1984). Abiot ic st ress, herbivory, and infection induced responses are well known and often affect mono- and sesqui terpene concentrat ions. There are no simple explanat ions or rules to account for the variability of essent ial oil and terpenoids in response to environment. For example , increased nitrogen resulted in higher essent ia l oil content in sage (Salvia officinalis) (Rohricht et al., 1996) but a lower content in sweet basil (Ocimum basilicum) (Adler et al., 1989). Essent ia l oil content increased under moderate water stress in basil (Simon et al. 1992) but dec reased in Cymbopogon sp. (Vivek et al., 1998) and peppermint (Char les et al., 1990). Terpenoid profiles and concentrat ions of essent ial oils in Artemisia annua and Oreganum vulgare varied depending on the geographic location and climate in which they were grown (Kokkini et al., 1994; Wallaart et al., 2000). Sesqui terpene lactone content, particularly parthenolide, is highly variable in feverfew (Fontanel et al., 1990; Hendr iks et al., 1997; Heptinstall et al., 1992; A w a n g et al., 1991) but the cause of this variation has not been investigated. In an effort to elucidate some factors influencing parthenolide biosynthesis, experiments presented in this chapter were des igned to examine water and nitrogen effects on parthenolide content in feverfew. 27 In the field, light intensity, photoperiod, temperature, and rainfall are determined by season and cl imate, but it is possib le to manipulate some of these environmental parameters. Plasticulture is an intensive agriculture sys tem which increases air and soil temperatures in the field by covering both the soil and crop with vented plastics, which may result in increased crop yield and quality (Brown and Channel l -Butcher, 1999; Ricotta and Mas iunas , 1991). Trickle or drip irrigation is often used with plasticulture and is an efficient delivery system for both nutrients and water. Fertigation is the application of soluble fertilizer with irrigation water to del iver nutrients directly to the root zone (Barua et al., 2000). Combin ing these sys tems provides a variety of advantages to farmers and scientists al ike. B lack or dark green plastic mulch used to cover the ground leads to soil warming, el iminates weeds , and reduces labour costs (Ricotta and Mas iunas , 1991; Bonanno, 1996). Heating the soil early in the season and covering the crop with miniature tunnels of c lear plastic, advances the growing season and results in earlier maturation, higher yields and quality ( Dubois, 1978; Ga lambos i and Szeben i -Ga lambos i , 1992). Fertigation enab les accurate manipulation of nutrient and water delivery and for these resources plasticulture with fertigation in the field is similar to greenhouse hydroponic sys tems but maintains the advantages of field edaphic factors. G reenhouses are used for growing high value crops and have been indispensable for research. C rops can achieve high levels of production, quality, and c leanl iness when cultured in greenhouses. T h e s e benefits come from artificial lighting, cl imate control, and soi l - less nutrient delivery sys tems. Greenhouse hydroponics provides a variety of benefits including easy manipulation of the nutrient supply, p H , and salt concentrat ion, and uniform application of nutrient solution to the plants. One primary advantage of g reenhouses in colder cl imates is the ability to grow plants 28 throughout the winter. W h e n the cl imate is right, however, the field has a distinct advantage over greenhouses. In the Fraser Val ley, British Co lumb ia , common cloudy condit ions reduce irradiance levels. W h e n there is a further reduction of 3 0 % of incident light reflected and absorbed by greenhouses , the light which reaches the leaves is significantly lower in greenhouse than in field crops. G reenhouse growers often use sterile and soi l - less medium devoid of soil microflora. Soi ls most often contain beneficial microorganisms, provide unlimited space for root growth, and general ly have a high capacity for nutrient and water retention (Vince-Prue and Cockshu l l , 1981; Downs and Hel lmers, 1975). Plant growth depends on temperature, CO2, light, nutrition and water availability. Nitrogen forms a large part of plant element composit ion and is the fourth most abundant element in plants after carbon, oxygen and hydrogen (Sal isbury and R o s s , 1991). General ly , plant growth increases with nitrogen (N) supply but excess ive N supply can inhibit growth. Many spec ies alter their terpenoid composi t ion in response to N availability but the complement of lower terpenoids in plants is species-speci f ic . Nitrogen-deprived Heterotheca subaxillaris has been found to have greater monoterpene and sesqui terpene content than nitrogen-rich plants (Mihaliak and L incoln, 1989). The terpene content of Abies grandis did not respond to N fertilizer (Muzika, 1993). Sesqui terpenes in basil (Ocimum basilicum) increased with increasing nitrogen but dec reased with excess ive nitrogen supply (Youssef et al., 1998). The terpene-rich essent ia l oil content of geranium (Pelargonium graveolens) was greatest at moderate N fertilizer application levels (100 kg/ha) but decl ined with either excess or low N supply (Rao et ai, 1990). S T L s vary among spec ies and among individuals within a spec ies in response to nitrogen and water supply. For example, S T L accumulat ion in chicory (Chichorium intybus L.) roots is dependent on the cultivar, 29 growing location and N supply (Peters et al., 1997). In general there appears to be an opt imum nitrogen level for max imum essent ial oil and S T L content. The variation observed in secondary metabol ism, and particularly terpenoids, among genotypes and when grown in different environments and under different cultivation methods may be a concern because many of these compounds are toxic and they are often consumed in foods and medic ine. High levels of some terpenoids may be detrimental to health. Water is necessary in plants as a medium for transporting nutrients and organic compounds , it is the reaction medium in plant cells and t issues, it is involved in holding the plant upright and turgid, it enables growth through expans ive osmot ic forces of the vacuo le , and it is a source of molecular oxygen and hydrogen in chemica l reactions. Reduced water supply to plants can result in a variety of responses . The stress caused by reduced water supply leads to decreased synthesis of cytokinin, and increased synthesis of absc is ic ac id . This may result in dec reased growth, stomatal conductance, and photosynthetic rate (Kramer and Boyer, 1995; Le tchamo, 1996). The effects of water stress on secondary chemistry are not well understood and may vary between spec ies . There s e e m to be no consistent effects of water stress on terpenoid content in plants. Water stress in mint has been shown to increase cycl ic monoterpenes in the essent ial oi l, whereas the essent ial oil in well-watered plants contain predominantly acycl ic monoterpenes (Gershenzon , 1984). The sesqui terpene and diterpene content of a desert sunf lower (Helianthus ciliaris) increased, by double, under water stress (Gershenzon , 1984), whereas essent ia l oil in lemon grass (Cymbopogon flexuosis) was not affected by mild water stress (Singh, 1999), and in marigold (Tagetes pagula) there was a reduction in the sesqui terpene containing volatile oil (Razin and Omer , 1994). Plant growth is reduced by limitations or e x c e s s e s of resources such as nutrients, light, and water. Thus from an economic perspect ive, the potential benefits 30 of a stress- induced increase in synthesis of valuable phytochemicals are worth investigating. If b iomass is reduced substantial ly while commercia l ly important secondary compounds increase, then the economics of process ing and marketing high versus low potency materials must a lso be cons idered. Reported in this chapter are the results of three exper iments that explored the effects of irrigation f requency and nitrogen fertilizer on parthenolide content and growth of feverfew. The first was a field experiment to determine the effects of irrigation frequency. The second experiment, a lso in the field, examined the interactions of nitrogen fertilizer supply (3 levels) and irrigation f requency (2 levels). The third experiment was conducted in the greenhouse, and also determined the interactions of two levels of irrigation and two levels of nitrogen fertilizer in a hydroponic sys tem. 2.2. M A T E R I A L S A N D M E T H O D S 2.2.1. G e n e r a l m e t h o d s Propagation of field transplants Feverfew was grown from seed suppl ied by 5-B Produce (Langley, B.C.) , a commerc ia l grower in the Fraser Val ley. The seeds were sown into flats containing a peat and perlite mixture, in rows spaced 5 cm apart and covered with 3 m m of perlite. Germinat ion was apparent one week after sowing and the seedl ings were grown for 30 days before transplanting into trays with 100 ml cel ls. T h e repotted seedl ings were grown in the greenhouse, fertilized weekly for 4 weeks with commerc ia l 20-20-20 (N :P :K) fertilizer at 2 g/l (50 ml/plant), then were grown for another 4 weeks without fertilization, at which they reached an average height of 10 cm. Immediately after transplanting, the crop was watered. 31 Field culture methods Fever few field trials were conducted in 1996 and 1997 in A g a s s i z , B . C . The soil is a Monroe silt loam with 5.5 to 5.8% organic matter, c lassi f ied as a Eutric Eluviated Brunisol . In the year before each planting liquid dairy manure was appl ied in the spring and a rye cover crop was seeded in mid summer and incorporated by ploughing the following Apri l . In south-coastal British Co lumbia , approximately 1000mm of rainfall perculates through the soil in the fall and winter leaching all nitrate N from the soil rooting zone . The field was prepared each year for transplanting by cultivation fol lowed by the formation of raised beds that were 30 cm high, 1.1 m wide and 1.8 m apart from center to center. A drip irrigation line was laid down the centre of the beds to deliver water and d issolved fertilizer. The drip emitters were spaced 12 cm apart and an in-line pressure regulator maintained pressure at 10 - 12 psi . The raised bed and the irrigation line were covered with 2 mm, U V resistant dark green polyethylene mulch (IRT-76; A E P industries, South Hackensack , N.J.) appl ied with a tractor-drawn applicator. The plants were spaced 45 cm apart within the row and 45 cm between staggered rows. For transplanting, a smal l slit was cut into the plastic and the seedl ings were planted through the slit and into soi l . A mini-tunnel sys tem was erected over the new crop to create a greenhouse effect. The mini-tunnels consisted of 2mm clear polyethylene stretched over 0.8 m tall wire hoops spaced 2 m apart. The tunnels were ventilated with 8 cm holes along the tunnel tops. The tunnels were removed in mid-July. Ai r and soil temperature was monitored hourly using thermistor sensors attached to a data logger (Campbel l Scientif ic). A i r temperature w a s measured at the canopy level and 10 cm below the soil surface. W h e n the temperature was above 35 °C in the tunnels, the s ides were lifted 30 cm at every second hoop to increase 32 ventilation and reduce the temperature. Soi l matric potential was measured at a depth of 21 to 24 cm midway between two plants in a row using a tensiometer capped with a rubber septum and measured using a tensimeter (Soil Measurement Sys tems , Tucson , Ariz.) pressure sensor . Soi l matric potential was measured every three or four days or more often when warm, dry condit ions warranted more frequent monitoring. Subsampling and harvesting for yield measurement The plants in each plot or experimental unit (EU) were harvested at the same time. A n experimental unit is equal to one replicate of one experimental treatment. The plants were cut 12 cm above the ground. One or more plants from each E U were taken as subsamples for organ partitioning. The remaining plants were weighed fresh. The subsamp les were partitioned into leaves, s tems, and f lowers for measurement of dry to fresh weight ratios and number of f lowers. Subsamp les were dried at 50 °C for three days to measure t issue dry weights. Parthenolide extraction and quantification A s specif ied in the respective sect ions, fresh or dry feverfew leaves, s tems, and f lowers were extracted separately with dichloromethane for 30 seconds . Both fresh and dry leaves were not ground and only the leaf sur faces were extracted. This extraction method was used to selectively extract non-polar compounds on the leaf sur faceand in glandular t r ichomes. After the dichloromethane extraction the solvent was vacuum filtered through Whatman No. 1 filter paper. The leaves were extracted two addit ional t imes (30 seconds each) with 50 ml aliquots of d ichloromethane and vacuum filtered. The three extracts were combined into a round bottom flask and evaporated to dryness under vacuum at 30 °C in a rotary evaporator (rotovap). The extracts were resuspended in 2 x 10 ml methanol , agitated in a sonicat ing water bath, 33 transferred quantitatively to a 25 ml g lass vial and sea led . Approximately 1 ml of the extract was used for H P L C analys is. The H P L C system used for ana lyses was a Waters 6 0 0 E controller, 790 photodiode array detector, 770 autosampler, operated with Mi l lennium software. The column w a s Waters C 1 8 reverse phase, 150 m m x 3.9 mm. The mobi le phase was isocratic wateracetoni t r i le (55:45) for ten minutes at a flow rate of 1 ml/minute. A ten point standard curve was prepared in duplicate from a parthenolide standard (97% pure) purchased from Sigma-Aldr ich. The regression line passed through zero and R 2 was 0.998. Quantitation was based on peak area and retention t ime. Statistical analysis Treatment effects on all plant response var iables were ana lyzed using analysis of var iance. 2.2.2. F ie ld irrigation trial Experimental design Transplant ing was on 26 May, 1996. Two irrigation treatments were randomized in each of four blocks for a total of eight E U s . S ix raised beds consisted of two outer beds as guard rows and four inner beds, each one designated a block. The beds were oriented north/south. E a c h E U was half the length of a bed with two rows, each with six plants. The two plants at the row ends were guard plants. Guard plants were used to reduce the effects of growing on the edges of the plots and were not used for measurements . Therefore, each E U had a total of eight plants for measurements . Fertilizer and irrigation schedule Ferti l izer w a s we ighed, d issolved in warm water to make 20 L of concentrated solution. The pH of the concentrate was adjusted to 6 before it w a s injected into the 34 irrigation line over 45 minutes. Fertil izer appl icat ion rate was based on bed surface area and calculated in ki lograms per hectare (kg-ha" 1). The crop w a s fertigated every two weeks according to the fertigation schedule in Append ix 1. There were two irrigation treatments that started one week after transplanting. Wate r was applied according to tensiometer measurements at a threshold of - 2 0 k P a for the high irrigation f requency and at - 8 0 k P a for the low irrigation f requency treatment or with the scheduled fertigation. The low and high irrigation rates were appl ied for three months. Leaf sampling for extraction The fourth fully expanded leaf from the shoot apex was samp led from each of two shoots per plant every three weeks . At each sampl ing, the sixteen leaves from each E U (two from each plant) were pooled into one sample for chemica l analys is. The pooled samp les were dr ied, we ighed, extracted with dichloromethane and analysed by H P L C , as descr ibed in Sect ion 2.2.1. The crop was harvested 23 August , 1997. Shoot organs were partitioned into stem leaf and flower and dry weights measured. 2.2.3. Field irrigation and nitrogen application trial Experimental design Transplant ing was on 16 May, 1996. The treatments consisted of a factorial combinat ion of three rates of nitrogen fertilizer and two irrigation rates, appl ied in a randomized block design with four blocks. The nitrogen fertilizer rates were per year totals of 0, 50, and 100 kg-ha" 1 appl ied in split appl icat ions every two weeks . The nitrogen source was ammonium nitrate. High and low frequency irrigation treatments were appl ied as in 2.2.2 according to tensiometer measurements taken every two or three days . There were eight raised beds orientated north/south. The two outside 35 beds contained guard rows. A block consisted of 1.5 beds. E a c h bed split into four E U s . The treatments were applied at random to E U s within a block. Each E U contained 12 plants in two rows of six. The two plants at the end of the rows in each E U were guard plants. The guard rows and guard plants were not included in data col lect ion, leaving 8 plants for data collection in each E U . There were four replicates and treatments were randomized within a replicate block. Fertilizer and irrigation schedule Ferti l izers to be applied (Appendix 1) were d issolved in warm water to make 20 L of concentrate. After adjusting the pH to 6, the concentrated fertilizer solution was injected into the irrigation line. Al l nutrients, with the except ion of ammonium nitrate, were appl ied at the same rate. Ferti l izer application rate was based on the bed surface area. The bimonthly fertigations were 45 minutes long. Irrigation was appl ied when the soil matric potential was below - 20 k P a for high f requency irrigation rate, and below - 8 0 k P a for the low f requency rate. Irrigation was appl ied for 70 minutes. Leaf water potential and leaf osmotic potential Leaf water potential and leaf osmot ic potential measurements were recorded three t imes over 20 hours, at 3 a m , 12 pm, and 6 pm. The measurements were taken on a c loudless day/night, July 25 , 70 days after transplanting. Wate r potential measurements were taken with a portable pressure bomb. Leaves with petioles, approximately 12 c m long, were cut and immediately p laced into the pressure bomb with petioles extruded. S ix leaves from each E U were measured . The pressure was recorded (bars) when sap emerged from the cut petiole. The s a m e leaves were frozen in preparation for osmot ic potential measurements . The leaf samp les were thawed and their liquid contents pressed into a vial for measurement . The osmot ic potential of the 36 liquid was determined using a freeze-point depress ion osmometer (Advanced Instruments). Leaf sampling for chemical analysis Leaves were sampled every three weeks . The leaf samp les were the fourth leaves from the ap ices of two shoots per plant. The 16 leaves sampled per E U were pooled together for ana lyses . Each leaf was cut in half and the petiole removed. One -half of each leaf was weighed fresh and then dried and reweighed to obtain dry to fresh weight ratio. The second half of the leaf was weighed and extracted fresh. The dry weight for the second leaf half was then est imated from its fresh weight and the dry to fresh weight ratio of the first leaf half. The extraction method and H P L C analysis is descr ibed in Sect ion 2.2.1. 2.2.4. Greenhouse irrigation and nitrogen application trial Propagation Feverfew was grown as descr ibed in Sect ion 2.2.1 with the following except ions. The seedl ings were transferred from the cell trays to 8 x 8 x 8 cm rockwool cubes. On 21 January, 1997 the plants in rockwool cubes were transferred onto sawdust-f i l led, white polyethylene bags (pillow bags). The pillow bags were approximately 1 meter long and 18 cm in diameter. The bags were laid horizontally and holes cut on the top surface for plant roots to penetrate, and the bottoms were slit for drainage. Greenhouse culture system The sawdust pillow bags were suppl ied nutrient solution via trickle irrigation. A pump submersed in a 500 L nutrient tank pumped dilute solution through half-inch polyethylene tubing to a smal l diameter polyethylene "spaghett i" tubing. A plastic spike attached to the end of the spaghetti tubing held it at the base of the plant. The flow 37 del ivered 35 ml/min/per drip line. The flow through fertigation sys tem al lowed excess nutrient solution out through slits at the bottom of the bag. Experimental design This experiment had four treatments, which were a factorial combinat ion of two levels of nitrogen and two irrigation rates, as follows (concentrat ions refer to ammonium nitrate in the nutrient solution): 1. 170 ml of 0.91 m M N solution per day 2. 170 ml of 0.45 m M N solution per day 3. 170 ml of 0.91 m M N solution plus 70 ml water per day 4. 170 ml of 0.45 m M N solution plus 70 ml water per day E a c h treatment was ass igned at random to six E U s in six b locks. E a c h pillow bag was one E U and had two plants spaced 30 cm apart. In all there were 48 plants in 24 E U s in six blocks. Guard plants were placed around the perimeter of the experiment and were not included in data collection. Pi l low bags were spaced 60 cm between rows and plants spaced approximately 45 cm within rows. Fertilizer and irrigation schedule The formula of the nutrient solution is outlined in the Appendix . E a c h N solution was held in a 500 L tank. The E C was adjusted by adding N a C l to the lower N concentrat ion tank to match that of the higher N tank which was 2 mSv . The pH was adjusted to 6.0 in both tanks using sulphuric ac id . Al l nutrients other than ammonium nitrate were the s a m e concentration in both tanks. Nutrient delivery was controlled with an A R G U S environmental control sys tem. The nutrient solution was appl ied four t imes per day to deliver 170 ml/plant/day. Nutrients were del ivered at the s a m e time for the s a m e duration in all treatments. There was a secondary irrigation system for watering the high irrigation rate which del ivered an additional 70 ml of water/plant/day. 38 Sampling methods Leaf samp les were col lected every three weeks and included two leaves from each plant, both the fourth leaf from the apex of two shoots. The four leaves sampled from each E U were pooled into one sample for chemica l analys is . The leaves were cut in half and ana lysed fresh and dry on a dry weight basis according to the procedure descr ibed in Sect ions 2.2.1 and 2.2.3. 2.3. RESULTS 2.3.1. Field irrigation trial: effects of irrigation frequency on parthenolide concentration and plant growth Irrigation rate had a smal l but significant affect on leaf parthenolide concentrat ion and shoot yield. Parthenol ide content was similar between treatments at 43 days, but at 87 days it was higher in leaves of plants grown at the lower irrigation rate. Under the higher irrigation frequency, leaves maintained a constant parthenolide concentrat ion between the two t imes. Shoot dry weight yield was 8% higher in response to the higher irrigation rate after 87 days. Table 2.1. Parthenol ide concentration of feverfew leaves 43 days and 87 days after transplanting. Units for parthenolide concentrat ion are mg parthenolide / gram dry leaf and plant dry weights are in grams. Leaf parthenolide concentration (mg/g) Plant dry weight (g) Irrigation f requency 43 days 87 days 87 days Low 5.10 5.39 236.6 High 4.49 4.46 258.0 Signi f icance N S * * * Signi f icance (p < 0.05) N S No Signi f icance (p > 0.05) 39 2.3.2. Field irrigation and nitrogen application trial B iomass was affected by irrigation rate but not by nitrogen treatments (Table 2.2). There was a 9% reduction in shoot dry weight ( leaves, s tems and flowers combined) in response to the low compared to high irrigation rate (p=0.017). In a mature plant, s tems contributed more to total dry weight than did f lowers and leaves combined (Table 2.2). Leaf dry weight was 14% lower (p = 0.049) under the lower f requency irrigation treatment. S tem dry weight was 1 0 % lower (p=0.012) under the lower f requency irrigation rate. Dry to fresh weight ratios of leaves, s tems, and flowers were not affected by irrigation treatments. Table 2.2. Average whole plant and organ dry weights. S a m p l e s were harvested 87 days after transplanting. Dry weiqht yield (q/plant) Nitroqen rate (N)(n=8) stem leaf f lower total 0 Kg /ha 130.4 34.89 71.65 236.9 50 Kg/ha 133.4 37.52 72.10 243.0 100 Kg/ha 136.6 39.89 71.33 247.8 signi f icance N S N S N S N S Irriqation rate (I) (n=12) low 127.0 34.57 70.09 231.6 high 139.9 40.30 73.29 253.6 signi f icance ** * N S * * signif icance p<0.05 ** signif icance p<0.01 N S no signif icance p>0.05 Leaf water potential, osmot ic potential, and turgor were highest at 3 a.m. and decreased through the day at 12 pm and 6 pm (Table 2.3). Plants under the lower f requency irrigation treatment had lower water potentials, osmot ic potentials, and turgor pressure at all three measurement t imes than did plants under high frequency 40 irrigation, specif ical ly in the late afternoon and at night. The calculated turgor was approximately zero under low frequency irrigation at 6 pm. The N treatments affected leaf water status. The 50 kg/ha N treatment had higher water potential and turgor than both the 0 and 100 kg/ha treatments at 12 pm. At 6 pm the water potential was lower in the 50 kg/ha N treatment resulting in a turgor pressure of 0 M P a . T a b l e 2.3. Leaf water status in field grown feverfew measured at 3 am (night), 12 pm (mid-day), and 6 pm (evening). Leaf water potential was measured in the apoplast and osmot ic potential was measured from the expressed cell contents. Turgor pressure was calculated by subtracting osmot ic from water potential. Al l va lues reported in M P a . Nitrogen rate 0 Kg /ha 50 Kg/ha 100 Kg /ha signi f icance Irrigation rate Low High signi f icance Water potential Osmot ic potential Turgor pressure 3 am 12 pm 6 pm 3 am 12 pm 6 pm 3 a m 12 pm 6 pm -0.453 -1.19 -1.36 -1.15 -1.33 -1.44 0.696 0.139 0.077 -1.18 -1.26 -1.41 0.733 -1.20 -1.31 -1.49 0.750 N S N S N S N S -0.447 -1.04 -1.44 -0.447 -1.16 -1.41 N S N S -0.496 -1.15 -1.55 -1.19 -0.402 -1.11 -1.25 -1.16 N S ** N S 0.224 0 0.148 0.085 N S -1.31 -1.49 0.697 -1.29 -1.41 0.755 N S 0.161 0 0.180 0.160 N S * signi f icance p<0.05 ** signi f icance p<0.01 N S no signi f icance p>0.05 Leaf parthenolide concentration increased during early development and then decreased as the crop matured to the reproductive stage (Table 2.4). After the feverfew crop was harvested on 24 July (87 days after transplanting) it was regenerated and leaf parthenolide concentration increased. Leaf parthenolide varied over time but was consistent between treatments. There were significant treatment 41 effects due to irrigation frequency and nitrogen rate in samples taken 69 days after transplanting just prior to harvest. The low irrigation frequency and 50 Kg/ha nitrogen resulted in the highest leaf parthenolide concentration. The regenerated growth showed significant nitrogen treatment effects after 28 days of regeneration. Applying no N resulted in a lower parthenolide concentration than did the other nitrogen treatments. After 53 days of regeneration lower frequency irrigation resulted in lower parthenolide content. T a b l e 2.4. Dry leaf parthenolide content measured over a 3 month period in the field. Sampling dates are indicated and days after transplanting are bracketed. Units are mg parthenolide/gram dry leaf. Leaf parthenolide concentration (mg parthenolide/q dry leaf) First crop Regenerated crop Nitrogen rate 14-Jun 26-Jun 09-Jul 24-Jul 21-Aug 15-Seo (32) (44) (58) (73) (101) (126) 0 Kg/ha 3.96 5.50 2.84 2.71 2.82 4.49 50 Kg/ha 4.25 6.10 3.09 3.40 3.72 4.51 100 Kg/ha 4.34 5.46 2.81 2.45 3.74 4.39 Significance NS NS NS * * NS Irrigation frequency Low 3.93 5.52 3.01 3.16 3.32 3.99 High 4.43 5.85 2.82 2.55 3.53 4.93 Significance NS NS NS * NS * * significance p<0.05 ** significance p<0.01 NS no significance p>0.05 The parthenolide concentration in dried flowers was more than twice that in dried leaf lamina (Table 2.5). Stems contained very little parthenolide. Approximately 80% of the parthenolide in a flowering plant was contained in the flowers. Leaves contained about 18% followed by stems at 2%. The field treatments affected both the flower and 42 leaf parthenolide content. Leaf parthenolide content was 2 0 % higher under the lower irrigation rate than the higher rate. Among the three nitrogen rates, 50 Kg/ha N applied resulted in the highest leaf parthenolide content. A s appl ied N w a s increased under the lower irrigation rate, f lower parthenolide content dec reased , but under the higher irrigation rate, increasing N application resulted in increased f lower parthenolide content (Table 2.5). T a b l e 2.5. Leaf and flower parthenolide concentrat ions and total parthenolide content per plant of s tems, leaves and f lowers. S a m p l e s were harvested from mature flowering plants grown in the f ield. mg Parthenol ide per gram mg Parthenol ide per plant Leaf F lower Total 92.59 594.0 706.1 139.9 603.0 762.9 99.40 597.6 717.5 N S 117.8 604.4 741.3 103.4 591.9 716.3 N S * signi f icance p<0.05 N S no signif icance p>0.05 2.3.3. G r e e n h o u s e i r r iga t ion a n d n i t rogen app l i ca t i on t r ia l Irrigation and N fertilization effects on leaf parthenolide deve loped two months after treatment initiation. The mean leaf parthenolide concentrat ion was 8.4 mg/g dry weight at transplanting. A s the crop matured, young flowering tops were removed to repress flowering and promote crop homogeneity for 54 days . During this time, the mean leaf parthenolide concentration increased to 37 mg/g dry weight and was unaffected by treatments (Table 2.6). During the next 16 days treatment effects began Nitrogen rate Leaf Flower S tem 0 Kg /ha 2.71 8.36 19.57 50 Kg/ha 3.40 8.39 20.02 100 Kg /ha 2.45 8.40 20.49 Signi f icance * N S N S Irrigation frequency Low 3.16 8.63 19.05 High 2.55 8.13 21.00 Signi f icance * N S N S 43 developing and were maintained as the crop matured to flowering. At 70 days, leaves on plants fed with 1mM N had 40 mg parthenolide /g dry weight while those on plants fed 0.5 mM N had less than 30 mg parthenolide/g dry weight after 70 days. Subsequently, flowering was allowed to proceed, and parthenolide concentration decreased to 1.8 mg/g under low N and 13 mg/g under high N fertilization. The N treatment effect on leaf parthenolide was significant after two months and was maintained through the transition from vegetative to reproductive growth. Irrigation did not have significant effects on parthenolide content and there was no interaction between irrigation and N rate that effected parthenolide content. The higher irrigation treatment was continued for an additional 40 days after the lower irrigation treatment was stopped. T a b l e 2.6. Leaf parthenolide concentration in feverfew leaves grown in the greenhouse under irrigation and nitrogen treatments. Days from treatment initiation are in brackets. Flowering tops were picked off plants 31 days, 40 days, and 54 days after treatments were applied. Parthenolide concentration is mg parthenolide/g dry leaf Days after transplanting for sample collection 29 50 71 98 Nitrogen rate 0.45 mM 8.44 18.7 35.9 27.9 1.77 0.91 mM 8.44 21.7 37.3 39.6 12.6 significance NS NS NS ** ** Irrigation freguencv Low 8.44 21.8 37.2 34.5 High 8.44 18.5 36.0 33.4 5.80 significance NS NS NS NS NS * significance p<0.05 NS no significance p>0.05 44 2.4. D I S C U S S I O N The exper iments in this chapter showed that low irrigation f requency in the field had a smal l but significant effect on both plant yield and leaf parthenolide content. In the two field trials, yield was 8% (Table 2.1) to 9% (Table 2.2) lower three months after a low irrigation rate was initiated. Dry matter partitioning among organs of mature feverfew in Tab le 2.2 revealed that most of the dry weight (55%) was in s tems fol lowed by f lowers (29%) and leaves (16%). The average leaf dry weight per plant under the low irrigation rate was 14% less (p = 0.049), and stem dry weight was 1 0 % less (p=0.012) than plants irrigated at the higher f requency. Treatment effects in the 1997 field irrigation trial were probably reduced due to the high rainfall that summer. Water stress can reduce plant growth and yield but has variable affects on terpenoid chemistry (Hanson and Ne lson , 1980; G e r s h e n z o n , 1984). Under water st ress, nutrient acquisit ion and mobility is reduced, primary metabol ism slows down, and specia l ized stress responses may occur such as increased absc is ic acid synthesis, and increases in the synthesis of osmoregulators like proline, and betaine (Losche, 1996). Reduced stomatal conductance and metabol ism are the main causes of reduced b iomass underwate r stress (Losche, 1996). Leaf water potential is a measure of water status of a plant and is regulated by stomatal conductance which in turn determines rate of CO2 uptake. Low C 0 2 uptake can reduce carbon assimilat ion and b iomass. In the field study, feverfew responded to low irrigation f requency with reduced b iomass in leaves and stems. Feverfew under low irrigation f requency had higher water potentials and osmot ic potentials, specif ical ly in the late afternoon (6 pm) when transpiration was likely at a peak and during the night (3 am) when the stomata and the plant xylem equilibrate with the soi l . A s indicated by the lower plant dry weight, stomatal conductance was probably lower in response to the lower irrigation frequency 45 and reduced C 0 2 uptake and carbon assimilat ion. Low turgor pressure was calculated for low f requency irrigation treatments which a lso lead to reduced growth and wilting. Reduced CO2 uptake and low turgor are both possib le precursors for the yield reduction that was observed in stem and leaf yields under low irrigation treatments. Unlike s tems and leaves, f lower dry weights were not significantly affected by irrigation rate. The flowering process is a primary sink for carbon and often does not show the s a m e response to reduced water availability as leaves and s tems which exhibit reduced growth and development. Nitrogen is a component of all plant enzymes and many metaboli tes and thus is an important plant nutrient which is often a limiting factor in field crops. Feverfew is a weedy spec ies commonly found growing on roadsides and poor soi ls. The nutrient requirements of feverfew are probably low s ince it s e e m s able to thrive in unfertilized waste lands where it survives well and reproduces under low N condit ions. In the field, nitrogen application did not significantly affect feverfew dry weight yields and there were no visible s igns of nitrogen def ic iency such as leaf yel lowing or reduced growth when no N was appl ied. A plausible explanation is that residual nitrogen in the field was sufficient for feverfew growth and therefore addit ional nitrogen did not greatly contribute to yield. Nitrogen affected both water potential (Table 2.3) and turgor measured at 12pm. W h e n soil water is limited, plants can adjust the leaf osmot ic potential to maintain turgor at low water potentials. Plants utilize osmoregulators like proline, and betaine which contain nitrogen (Losche, 1996). Nitrogen availability may influence synthesis of osmoregulators in feverfew. There was an effect of N fertilizer rate on water potential at mid-day (12 pm) which resulted in a higher water potential for plants fertilized with 50 kg/ha N compared to 0 and 100 kg/ha N treatments. A n explanat ion for this result may 46 be that 50 Kg /ha N is sufficient for growth while maintaining the salinity of the soil lower than at 100 kg/ha N. Higher salinity in the soil caused by e x c e s s nitrogen could dec rease soil water potential reducing the plants ability for water uptake. Therefore, soil with higher nitrogen content and higher salinity may result in lower leaf water potentials in response to the soil osmot ic potential. This in turn would result in lower turgor which is the result obtained for 100 Kg/ha compared to 50 kg/ha N treatments. Tugor pressure at 12 pm under the 0 Kg/ha treatment was the lowest of all treatments and was primarily due to the low leaf water potential. Fever few under 0 kg/ha N treatment may not have adequate N and therefore must keep stomates open to fix carbon for root growth, and to maintain the transpirational st ream for nitrogen acquisi t ion. Interestingly the treatments with significant dif ferences in water relations are the s a m e as those with differences in parthenolide concentrat ions. The role of secondary chemistry in plants under water stress remains, for the most part, unknown. St ress induced increases of secondary compounds may provide antioxidant potential to combat increased oxidation due to water st ress (Losche, 1996). St ress induced susceptibil i ty to pathogens may elicit increased chemica l defences. Pr imary metabol ism may be affected while secondary metabol ism remains unchanged resulting in higher relative concentrat ions of the secondary compounds . Finally, secondary metabol ism may decrease in response to water st ress as a conservat ion mechan ism to minimize waste of resources (Gershenzon , 1984; Gershenzon and Cro teau, 1991). There have been many studies of the effects of water stress on essent ial oil content in a number of spec ies showing var ious and somet imes contradictory results, but few that specif ical ly examine sesqui terpene lactones. In the 1996 and 1997 field trials, feverfew's parthenolide content 73 days after planting (DAP) (Table 2.4) and 87 D A P (Table 2.1) respectively were higher under low irrigation rates 47 while shoot b iomass dec reased . The average leaf parthenolide concentrat ion was 17% (table 2.1) to 2 0 % (figure 2.5) higher under low irrigation treatments compared to high irrigation treatments after three months. T h e s e treatment effects took more than two months to develop. In addit ion, only the plants in full f lower showed this treatment effect. At harvest, leaf parthenolide concentration in regenerated vegetative plants was higher under the high irrigation f requency compared with the low frequency. Nitrogen moves in the soil with water. Under low irrigation treatments N may build up in the soil because it is less prone to leaching compared with higher irrigation treatments. This can result in higher N availability under low irrigation compared to high irrigation treatments. If the plant takes up more N under low irrigation f requency a higher rate of enzyme production may result al lowing greater S T L production. Another explanation for higher parthenolide concentration in low irrigation treatments was briefly mentioned above where parthenolide and S T L synthesis may not have been directly affected by water or nitrogen availability but was an indirect result of reduced b iomass caused by these factors. For example if biosynthesis remained constant but leaf s ize was reduced due to limited resources then leaf S T L concentrat ion may be higher. The response of feverfew to mild water stress was similar to that found with Mentha piperita in which essent ia l oils (including sesqui terpenes) increased by 3 7 % and was accompan ied by decreased b iomass (Char les et al., 1990). In field grown Ambrosia maritime and Achillea millefolium, dec reased growth in response to reduced irrigation was correlated with decreased leaf and flower sesqui terpene lactone concentration (Sidky, E l -Mergawi , 1997; El -Kholy, 1984; S imon er al., 1992). In these types of investigations however, problems arise when differentiating between changes in b iomass and secondary chemistry s ince it is often unclear if biosynthetic rate of the 48 secondary compound or plant b iomass are the main effect when they both may change in response to treatments. Irrigation and nitrogen affected both flower and leaf parthenolide. At harvest 72 D A P leaf parthenolide was highest under the 50 kg/ha N treatment regardless of irrigation rate (Table 2.5). Simi lar results were obtained by R a o (1990) in a study of essent ial oil y ields of geranium (Pelargonium graveolens) which were 2 4 % higher at moderate nitrogen levels (100 kg/ha) compared with excess or limited nitrogen treatments. In a study of basi l (Ocimum basilicum), sesqu i terpenes increased with increasing nitrogen application but decreased when nitrogen was excess ive (Youssef et al., 1998). The main effects of irrigation and nitrogen on f lower parthenolide concentrat ion were insignificant, but an interaction between the two factors had an effect on parthenolide concentration resulting in oppos ing effects of N when different irrigation rates were appl ied. Nitrogen treatments on feverfew grown in a greenhouse had significant effects on leaf parthenolide. Leaf parthenolide was the highest in response to lower irrigation level and higher nitrogen treatments 71 and 98 D A P , a result similar to the field trials. However, in contrast to the field trials leaf parthenolide substantial ly increased as the crop matured. This increase was enhanced by the removal of young flowering tops to repress f lowering. Treatment effects on leaf parthenolide were significant after two months, and were sustained through the transition from vegetat ive to reproductive growth. In the third month while the plants were in a vegetative stage, the highest leaf parthenolide concentrat ion (maximum 5%) was recorded. Plants grown under high N rate had 3 0 % greater leaf parthenolide (4%) than plants treated with a low N rate (2.8%). This is in contrast to the field trials, in which parthenolide decreased as the crop developed and never attained levels greater than 1%. After f lowering was al lowed 49 to proceed in the greenhouse trial, the crop continued its development to the reproductive stage during which leaf parthenolide substantial ly dec reased . Pref lowering leaf parthenolide was very high in this g reenhouse trial compared to that reported e lsewhere. Leaf parthenolide from vegetative plants was reported to be much lower than in leaves from flowering plants (0.33% versus 1.27%) (Awang, 1991). Fever few leaf parthenolide concentration in flowering plants was reported as high as 2 .77% in apical leaves, but general ly ranges between 0 .3% - 1.5% in mature leaves (Hendriks et al., 1997; Dolman et al., 1992; Brown er al., 1996). The majority of research on feverfew appears to have been performed in g reenhouses , and results of detai led field studies have not been reported. It would be unfair to compare directly the field and greenhouse trials presented in this chapter because they were done at different t imes under different condit ions. However, in the g reenhouse the significant dec rease in parthenolide content after being held in a vegetative state for 3 months and then al lowed to flower suggested there was a large effect correlated with the developmental stage that might a lso explain patterns observed in the field. Exper iments that explore developmental affects on parthenolide and yield are presented in chapter 4. Fever few is marketed with an assurance of minimum parthenolide concentration. Parthenol ide content was proposed as a marker compound to assure feverfew authenticity. It has been incorrectly presented as a quality assurance and somet imes as the primary active principle. Not all the active components have been identified and parthenolide is found in many other plants (see Chapter 1). Farmers are paid by dry weight of the herb regardless of parthenolide concentration and therefore high levels of field production with low overhead costs are important regardless of chemical quality. Y ie ld and parthenolide variability have been d iscussed separately but could be 50 considered together to determine crop performance based on parthenolide concentrat ion of the whole plant. The weight contributions of different organs are important when consider ing parthenolide content. For example , the field crop was harvested two months after transplanting when the plants had approximately 8 0 % mature f lowers. In whole plants, f lowers contributed most parthenolide (80%) per plant fol lowed by leaves (18%) and stems (2%). However, s tems contributed more than 5 5 % total dry weight fol lowed by f lowers (29%) and leaves (16%). Low irrigation rate reduced leaf and stem dry weights, but not f lower dry weight which was the major contributor of plant parthenolide. Therefore, total parthenolide was higher in plants receiving lower irrigation rates and 50 kg/ha N. By growing plants under specif ic condit ions it should be possible to achieve the most favourable organ proportions and chemical composi t ion if yield is not the first priority. Variabil i ty of secondary compounds within a spec ies is well establ ished and appears to be under both genetic and environmental control. For example, S T L accumulat ion in chicory (Chichorium intybus L.) root was dependent on the cultivar, growing location and nitrogen supply (Peters et al., 1997). Parthenol ide content in feverfew has been examined in plants grown in t issue culture (Brown et al., 1993, Brown et al., 1996), grown in different regions and cl imates (Maries et al., 1992), and in commerc ia l preparations (Awang et al., 1991; Heptinstall et al., 1992). Parthenol ide variability has been attributed to many factors such as process ing methods, chemical degradat ion (Smith and Burford, 1992), fillers and adulterants (Awang, 1991), varietal dif ferences (Awang, 1989) and extraction and analytical methods (Brown et al., 1996). As ide from this investigation, there have not been studies on the effect of the growing environment on feverfew. Indirect ev idence that growing condit ions affect parthenolide variability comes from the literature which reports large variation in feverfew 51 parthenolide content in leaves (0 % - 2.8 %), f lowers (0.5% - 2.3%), parthenolide concentrat ions that are very low in Canad ian-grown plants (Awang et al., 1991; Heptinstall and A w a n g , 1998), and even undetectable in Mex i can - and Yugos lav ian-grown feverfew (Maries et al. 1992). If high parthenolide concentrat ion and high plant yields are desirable, these experiments show that we can grow high quality feverfew in one part of C a n a d a in spite of previous c la ims that our cl imate is unsuitable. 52 2.5. REFERENCES Adler P .R. , S imon J . E . , and Wi lcox G . E . (1989) Nitrogen form alters sweet basil growth and essent ia l oil content and composi t ion. Hortsc ience, 24:789-790 Awang D .V .C , Dawson B.A., Kindack D.G. , Crompton C .W. , and Heptinstall S . (1991) Parthenol ide content of feverfew (Tanacetum parthenium) a s s e s s e d by H P L C and 1 H-nmr spectroscopy. Journal of Natural Products, 54:1516-1521 Barua P., Barua H.K., and Borah A . (2000) Plant growth and yield of A s s a m lemon as inf luenced by different drip irrigation levels and plastic mulch. Anna ls of Biology, 16:17-20 Bonanno A . R . (1996) W e e d management in plasticulture. Horttechnology, 6:186-189 Brown A . M . G . , Lowe K .C . , Davey M.R., and Power J . B . (1996) Fever few (Tanacetum parthenium): T i ssue culture and parthenolide synthesis. Plant Sc ience , 116:223-232 Brown A . M . G . , Lowe K . C . , Davey M.R. , Power J . B . , Knight D.W., and Heptinstall S . (1996) Compar ison of extraction procedures for parthenolide in Tanacetum parthenium. Phytochemical Ana lys is , 7:86-91 Brown J . E . , and Channel l -Butcher C . (1999) Effect of three row covers and black plastic mulch on the growth and yield of 'C lemson Sp ine less ' okra. Journal of Vegetab le Crop Product ion, 5:67-71 Char les D.J . , Joly R . J . , and S imon J . E . (1990) Effects of osmot ic stress on the essent ia l oil content and composit ion of peppermint. Phytochemistry, 29:2837-2840 Do lman D.M. , Knight D.W., Sa lan U., and Topl is D. (1992) A quantitative method for the estimation of parthenolide and other sesqui terpene lactones containing a lpha methylenebutyrolactone functions present in feverfew Tanacetum parthenium. Phytochemical Analys is , 3:26-31 El -Kholy S . A . (1984) Soi l moisture and macronutrient effects on yield and quality of Achillea millefolium L. Minufiya Journal of Agricultural R e s e a r c h . 8:331-348 Fontanel D, Bizot S , and Beaufi ls P (1990) Determination of parthenolide content in the chamomi le Tanacetum parthenium L. Schulz-b ip by H P L C . P lantes Medic inales et Phytotherapie, 24:231-237 Ge rshenzon J . (1984) C h a n g e s in the levels of plant secondary metaboli tes under water and nutrient stress. In: Recent advances in phytochemistry vol . 18; Phytochemica l adaptat ions to stress. P lenum P r e s s , New York, pp 273-320 Gershenzon J . , and Croteau R. (1991) Terpenoids, In: Herbivores, their interactions with secondary plant metabolites. Academ ic P r e s s , S a n Diego, pp 165-220 53 Hanson A . D . , and Nelson C E . (1980) Water: Adaptat ion of plants and other organisms. In: The biology of crop production. Academ ic P ress , New York, pp 78-147 Hendriks H., Andersonwi ldeboer Y . , Engels G . , Bos R., and Woerdenbag H.J . (1997) The content of parthenolide and its yield per plant during the growth of Tanacetum parthenium. P lanta Med ica 63:356-359 Heptinstall S . , A w a n g D.V., Dawson B.A., K indack D., Knight D.W., and May J . (1992) Parthenol ide content and bioactivity of feverfew (Tanacetum parthenium (L.) Schul tz-Bip.) . Estimation of commercia l and authenticated feverfew products. Journal of Pharmacy and Pharmaco logy 44:391-5 Heptinstall S . and Awang D .V .C . (1998) Feverfew: a review of its history,'its biological and medicinal properties, and the status of commerc ia l preparations of the herb In: Phytomedic ines of Europe, chemistry and biological activity. Amer ican Chemica l Society, Wash ington, Eds . Lawson , L. and Bauer R. Kokkini S . , Karousou R., and Vokou D. (1994) Pattern of geographic variation of Origanum vulgare t r ichomes and essent ial oil content in G r e e c e . B iochemica l Systemat ics and Eco logy 22:517-528 Letchamo W . G . A . (1996) Transpirat ion, essent ial oil g lands, epicuticular wax and morphology of Thymus vulgaris are inf luenced by light intensity and water supply. Journal of Horticultural Sc ience . 71:123-134 Losche R. (1996) Plant water relations: metabol ic responses to water deficit and surplus. Progress in Botany 57:17-31 Mar ies R . J . , Kaminsk i J . , A rnason J.T. , P a z o s - S a n o u L., Heptinstall S . , F ischer N.H. , Crompton C .W. , Kindack D.G. , and Awang D.V. (1992) A b ioassay for inhibition of serotonin release from bovine platelets. Journal of Natural Products 55:1044-1056 Mihal iak C A . and Lincoln D.E. (1989) C h a n g e s in leaf monoterpene and sesqui terpene metabol ism with nitrate availability and leaf age in Heterotheca subaxillaris. Journal of Chemica l Eco logy 15:1579-1588 Muz ika R . M . (1993) Terpenes and phenol ics in response to nitrogen fertilization: a test of the carbon/nutrient balance hypothesis. Chemoeco logy . 4:3-7 Peters A . M . , H a a g s m a N., and Vanamerongen A . (1997) A pilot study on the effects of cultivation condit ions of chicory (Cichoruium intybus L) roots on the levels of sesqui terpene lactones in ch icons. Food Resea rch and Techno logy 205:143-147 R a o B . R . R . and Kai lash S . , and Bhattacharya A . K . (1990) Effect of nitrogen and method of harvesting on herbage and essent ia l oil y ields of geranium (Pelargonium graveolens L. Herit.). Indian Journal of Agronomy. 35:312-313 54 Raz in A . M . and O m e r E.A. (1994) Effect of water regime on the growth, f lower yield and volatile oil content of marigold {Tagetes patula). Egypt ian Journal of Horticulture. 21:195-202 Ricotta J .A . and Mas iunas J . B . (1991) The effects of black plastic mulch and weed control strategies on herb yield. Hortsc ience. 26:539-541 Rohricht C , Grunert M. and Solf M. (1996) The influence of graduated nitrogen fertilizer application on yield and quality of sage (Salvia officinalis L ) . Zeitschrift fur A rzne i - & Gewurzpf lanzen. 1:117-122 Sal isbury F .B . and R o s s C . W . (1991) Mineral nutrition. In: Plant physiology. Wadswor th Publ ishing , Cal i fornia Sidky M . A . M . and El -Mergawi R.A. (1997) Relat ionship between growth, biosynthesis and accumulat ion of major active constituents of Ambrosia maritima and some exogenous and endogenous factors. Bulletin of Facul ty of Agriculture, University of Cai ro. 48:631 -654 S imon J . E . , Re iss -Bubenhe im D., Joly R . J . and Char les D .J . (1992) Water stress-induced alterations in essent ial oil content and composi t ion of sweet basi l . Journal of Essent ia l Oil Resea rch . 4:71-75 Singh M (1999) Effect of irrigation and nitrogen on herbage, oil yield and water use of lemongrass (Cymbopogon flexuosus) on alf isols. Journal of Agricultural Sc ience . 132:201-206 Smith R . M . and Burford M.D. (1992) Supercri t ical fluid extraction and gas chromatographic determination o f the sesqui terpene lactone parthenolide in the medicinal herb feverfew (Tanacetum parthenium). Journal of Chromatography. 627:255-261 V ince -Prue D. and Cockshu l l K .E . (1981) Photoper iodism and crop production. In: Physio logical p rocesses limiting plant productivity. Butterworths, London, pp 175-197 V ivek P., S h a r m a J . R . , Naqvi A . A . and Sha rma S . (1998) Effect of soi l-moisture status on economic performance and divergence in Cymbopogon spec ies . Journal of Medic inal & Aromat ic Plant Sc iences . 20:388-393 Wal laart T . E . , P ras N., Beekman A . C . , Quax W . J . (2000) S e a s o n a l variation of artemisinin and its biosynthetic precursors in plants of Artemisia annua of different geographical origin: proof for the existence of chemotypes. Planta Med ica 66:57-62 Yousse f A . A . , Talaat I.M. and Omer E.A. (1998) Physio logical response of basil G reen Ruffles (Ocimum basilicum L.) to nitrogen fertilization in different soil types. Egypt ian Journal of Horticulture. 25:253-269 55 Chapter 3 Feverfew production under field and greenhouse conditions 3.1. I N T R O D U C T I O N Whi le feverfew reduces the f requency and intensity of migraine headaches , both migraine physiology and S T L (parthenolide) act ions are complex and not understood. Parthenol ide and other S T L s are considered the primary candidates for the antimigraine activity due to their vasoregulatory properties. There is indirect ev idence that parthenolide has antimigraine activity but until direct ev idence is presented for parthenolide, and we have evaluated the antimigraine action of other feverfew constituents, the medicinal quality of feverfew should not be a s s e s s e d only on parthenolide concentrat ion. O n c e the active compounds are estab l ished, greenhouse production may give the flexibility to manipulate chemical content through cultivation methods. The controlled environment of g reenhouses makes it possib le to grow a wide range of fruits, vegetables, herbs, f lowers and houseplants throughout the year and in areas with harsh cl imates. Greenhouse experiments al low for strict control of nutritional factors and environmental parameters. Al though greenhouse research is important for many of the common medicinal plants, most commerc ia l production is outdoors or plants are harvested wild. Dif ferences between the greenhouse and field may be temperature, light quality, nutrient and water availability, wind, rain, growing media , root restriction, and length of growing s e a s o n , each of which may in turn cause significant physiological and chemical changes in a plant. Most of these factors affect stomatal 56 conductance and leaf water potential, which directly affect y ield, physiology and alter chemical composi t ion. In Chapter 2, I reported that f ield-grown feverfew with low water potential had a lower dry weight yield and higher leaf parthenolide concentrat ion when plants were at full f lower, a result a lso found in Cymbopogon martinii under irrigation treatments where low irrigation resulted in reduced growth and higher essent ia l oil y ields (Shabih et al., 1999). Two causes of low water potential are reduced water availability and high temperature leading to a high vapor pressure deficit. For an herbaceous plant like feverfew in which leaves have commercia l value, water relations are important because they directly affect leaf yield. Therefore, the difference in water relations between the greenhouse and field may be important in determining whether greenhouse production can be used to improve yield and alter chemical concentrat ions and/or composit ion of feverfew. Fever few is commercial ly grown in the field but has potential as a greenhouse crop. The cost of greenhouse operation is high and often restricts greenhouse growers ability to produce high-value crops. To gain a market advantage greenhouse producers select crops with extremely high fresh market value or grow seasona l crops during the out-of-season periods when they can't be grown outdoors. Currently feverfew may not be grown economical ly in the greenhouse because its market value is low, it grows well in the field, and it can be dried and stored for long periods. Fever few is general ly sold as dried crushed shoots, therefore visual quality of the raw plant material does not contribute to value like other quality parameters such as chemical composit ion. Greenhouse production is only viable if high quality, high yield, or fresh product are required. There is ev idence that fresh leaves have different pharmacological activity compared with dried leaves, so the value of feverfew may be very different if sold as a 57 fresh product. In a study testing the effects of feverfew leaf extracts on aortic ring relaxation and contraction, fresh extracts caused relaxation while dried extracts caused contraction (Barsby et al., 1993). In contrast to fresh leaves , the dried leaves did not contain S T L s . More research is required but this may be ev idence toward the preferential use of fresh feverfew leaves. In this Chapter I report that field-grown feverfew had greater dry weight yields than the greenhouse. However, total shoot yield may not be as important as the yield of speci f ic organs such as leaves or f lowers. Leaves have been the traditional source and mixed preparations of leaves, f lowers and s tems may be of lower quality (efficacy) than leaves a lone. O n e way to change quality is by altering the proportion of leaf, stem and flower t issues. A n economica l way for producers to ach ieve this is to grow feverfew in a way that provides the best ratio of shoot organs. Another, more labor intensive way to increase quality is separat ing leaves from flowers and s tems. The removal of s tems would reduce yield but may increase potency. Reduc ing variability in chemical composit ion of medicinal plants is important in the herbal drug industry s ince variability makes it difficult to administer the correct dosage . Water relations, yield, and chemical composit ion are three related attributes in plants that vary between field and greenhouse production. The research presented in this chapter examines the potential for manipulating leaf, s tem, and flower proportions in the g reenhouse and field by manipulating growing condit ions. In this study I used feverfew as a model to investigate the relationship between yield and chemical content (STLs) in plants and how they are influenced by greenhouse and field environments. In this chapter compar isons are made between the greenhouse and field by conducting exper iments in both environments at the s a m e time with comparab le treatments. R e s p o n s e of yield, parthenolide concentrat ion, and water status to media treatments 58 and to production environment (field or greenhouse) were determined. Specif ical ly three primary quest ions were addressed : Does growing med ium affect yield and parthenolide concentrat ion of feverfew leaves, s tems, and f lowers? How do potted plants in the field or greenhouse compare with plants grown in field soi l? What is the effect of crop regeneration on yield, parthenolide concentrat ion, and water status in feverfew grown in the g reenhouse? 3.2. M A T E R I A L S A N D M E T H O D S 3.2.1. G e n e r a l m e t h o d s Propagation of feverfew seedlings Feverfew was grown from seed col lected from a previous feverfew trial at the Paci f ic Agr i -Food Resea rch Center in Agass i z , British Co lumb ia , C a n a d a . The seeds were sown (April 6 t h ) into flats containing a peat and perlite mixture, in rows spaced 5 c m apart and covered with 3 mm of perlite. Germinat ion was one week after sowing and the seedl ings were grown for 30 days in the flats before being transplanted into cell trays with 4 x 4 x 4 cm cel ls filled with a mixture of peat moss , sawdust , and compost in a 1:1:0.5 ratio. Seed l ings for the field experiment were moved outside into cold f rames June 10 t h and transplanted to the field June 14 t h (72 days after seeding). Seedl ings for the greenhouse trial were transplanted June 14 t h (72 days after seeding). On June 3 0 t h (16 days after transplanting) the growing tips were removed from plants in both the greenhouse and field trials. Harvesting and Shoot Regeneration Seventy -seven days after transplanting, the greenhouse and field grown crops were harvested. The shoots were pruned to crowns 10 cm above the base o f the plant. Th is left sufficient leaf material on the crown which promoted quick regenerat ion. The 59 field experiment was terminated after the harvest. The greenhouse experiment was continued by regenerating the plants that were pruned to crowns. New growth was observed one week after pruning. The second crop was harvested 95 days after the first crop was harvested (172 days after transplanting). The total shoot fresh weights were measured for each experimental unit (EU) . Subsampling Shoot subsamples for organ partitioning were se lected randomly from plants in each E U . The subsamp les were separated into leaves, f lowers, and s tems. Their fresh and dry weights were measured and the number of f lowers counted. Sampling Leaf samp les for chemical analysis were harvested from the greenhouse and field grown crops 0, 36, and 77 days after transplanting. Two addit ional samples were taken from the regenerated greenhouse crop 128 and 173 days after transplanting (50 and 95 days after the first harvest). Leaves sampled were the fourth leaf from a shoot apex taken from four shoots per plant. Leaves from plants in each experimental unit (EU) were dried in an open container at 40 °C for 6 hours and then weighed. Water potential and osmotic potential Prior to the crops being harvested at the end of summer and end of autumn, samples for leaf water potential and osmot ic potential measurements were col lected. Leaf water potential was measured at 2 pm and 4 am using a portable pressure bomb. Leaves with petioles that were approximately 10 cm long were cut and immediately p laced into the pressure bomb with petioles extruding. The pressure was recorded when sap extruded from the cut petiole. The same leaves were frozen for osmot ic potential measurements . The osmot ic potential of the expressed sap from the thawed 60 leaves was measured with a f reeze point depress ion osmometer (Advanced Instruments). Surface extraction Leaves were extracted by dipping three t imes with dichloromethane for 30 seconds each time. The extract was vacuum filtered through Wha tman No. 1 filter paper. The filtrate was evaporated to dryness under vacuum at 30 °C using a rotary evaporator. The extract was quantitatively transferred to a vial with 20 ml H P L C grade methanol for chemical analys is. Chemical analysis: parthenolide extraction and quantification Methods for parthenolide extraction and quantification are descr ibed in Chapter 2.2.2. Statistical analysis Statistical analys is was performed using S A S and S Y S T A T software. Analys is of var iance with P<0.05 was used to determine signif icance between treatments. 3.2.2. Field fertigation frequency and growing medium trial Experimental design The experiment was a randomized block design with five treatments and four blocks. The five field treatments are outlined in Tab le 3.1. Treatments were selected to compare field production methods with greenhouse production methods. The plasticulture treatment represented field production with plants directly in the field soi l . In this treatment, there were ten plants per experimental unit (6 experimental plants and 4 guard plants on the row ends) , spaced 45 c m between the rows and 45 cm within the rows. The fertigated plasticulture treatment was the s a m e as previous plasticulture trials (Chapter 2) except that the fertigation system was through individual drip l ines, 1 61 per plant, instead of drip tape under the plastic. The remaining four treatments represented greenhouse hydroponics and consisted of plants grown in 35 cm pots. E a c h pot was spaced 30 cm apart. There were three pots in each E U and each pot contained 2 plants spaced 20 cm apart. There were a total of six raised beds. The two outside beds were guard beds and the inner four beds were the blocks within which the five treatments were arranged randomly. The four b locks each with five treatments resulted in a total of 20 experimental units. Al l experimental plants were sampled for physiological and chemical measurements. T a b l e 3.1. Field experiment treatments and abbreviat ions. Treatment descript ions Abbreviat ions Plast iculture, low fertigation Soi l filled pots set into the ground, low fertigation Soi l filled pots above ground, high fertigation Soi l filled pots above ground, low fertigation Sawdust filled pots above ground, high fertigation Plasticulture, Low Inset Pots , Low Soi l Pots , High Soi l Pots , Low Sawdus t Pots , High Field culture system Seed l ings were transplanted to the field after propagation in the greenhouse (see Sect ion 3.2.1). A fertigated plasticulture system was used in combinat ion with pot-culture. O n e of the treatments utilized plasticulture with plants grown in field soi l . The other treatments used pot-culture. E a c h pot was set either into the ground or on top of the ground on dark green polyethylene covered raised beds. Pots above the ground in the field had a white polyethylene s leeve covering the black pot exterior to minimize high pot temperatures. A clear polyethylene tent was erected over the crop to reduce exposure to wind and rainwater. 62 Fertigation schedule The fertigation system in the field was des igned to mimic the greenhouse nutrient delivery sys tem. A nutrient solution was drawn into the irrigation line via a Mazz i injection sys tem, which draws nutrient solution by a hydraulic vacuum into the irrigation line. There were two rates of nutrient solution (high fertigation and low fertigation). Nutrient solution was del ivered either four (low rate) or eight (high rate) t imes daily, for two minutes each . The flow rate averaged 35 ml/min, which del ivered 280 ml (low) or 560 ml (high) nutrient solution per day per plant. Al l treatments were fertigated with the formulation outl ined in the Appendix . 3.2.3. Greenhouse irrigation frequency and growing medium trial Greenhouse culture system Fever few seedl ings were transplanted 20 cm apart into sawdust or soil treatments in the greenhouse. Sawdust filled pillow bags (cylindrical shaped , polyethylene, 1 m long x 20 cm diameter) were used for one treatment. Pots , 35 c m diameter and 35 cm deep, were used for the other three treatments (two soil and one sawdust). The plants were grown on a bench one-meter high, under natural day-length. Experimental design The greenhouse trial was des igned to match c losely the field trial descr ibed in sect ion 3.2.2. Treatments were arranged randomly in b locks from west to east. Three treatments received a high rate of nutrient feed and one soil treatment received a low rate of nutrient feed (table 3.2). There were ten experimental units per treatment. With the except ion of the pillow bag treatment, the other three treatments matched those in the field. The experiment lasted for 6 months. Th is was 3 months longer than the field trial. Al l plants were used for physiological and chemical measurements . 63 Table 3.2. Treatments in the greenhouse trial and abbreviat ions. Greenhouse treatment descript ions Tab le abbreviat ions Sawdust filled pots, high fertigation Sawdust filled pillow bags, high fertigation Soi l filled pots, high fertigation Soi l filled pots, low fertigation Sawdus t Pots , High Sawdus t Pi l lows, High Soi l Pots , High Soi l Pots , Low Greenhouse nutrient solution and feeding schedule The fertil izers to be applied were d issolved in a tank containing 500 L of water. The electrical conductivity o f t he nutrient solution was adjusted between 1.8 to 1.9 m S v using sodium chloride and the pH was adjusted to 6 with sulphuric ac id. Two pumps were submersed in the tank, each attached to 1.8 c m diameter polyethylene tubing that del ivered nutrient solution to the experiment. Nutrient solution was del ivered from this line to each plant through smal ler drip line (0.15 mm i.d.) at a rate of 35 ml/minute. The pumps and timing of nutrient solution delivery were regulated by an Argus control sys tem. Nutrient solution was del ivered four t imes per day for the low rate treatment or eight t imes per day for the high rate treatments. There was one drip line per plant. E a c h feed delivery was for 2 minutes to provide 280 ml (low rate treatment) or 560 ml (high rate treatment) solution per day. The nutrient feed formula is in the Appendix . 64 3.3. R E S U L T S 3.3.1 F i e l d m e d i a a n d fer t iga t ion tr ial The field trial compared plasticulture with culture in pots containing soil and sawdust media. The plants grown in plasticulture and in pots set into the ground had greater yields than plants grown in pots above-ground (Table 3.3). Plants in the plasticulture treatment had the highest stem proportion and shoot dry weights 2-3 t imes T a b l e 3.3. Dry weight and dry to fresh weight ratios of feverfew plants grown in the field and greenhouse. Plants were partitioned into leaves, s tems and flowers. The field and first greenhouse crops were harvested when the plants were in full flower, 142 days after seed ing and 77 days after transplanting. The second greenhouse crop, regenerated from crowns, was harvested 95 days after harvesting the first crop. The regenerated greenhouse crop was in the vegetative developmental stage when harvested November 30. Dry weight yield (g) Drv:Fresh weight # flowers/ gdry Field Leaf Stem Flower Total Leaf Stem Flower shoot Plasticulture, Low 36.7 70.3 48.1 155 0.156 0.239 0.238 7.39 Inset Pots, Low 16.3 26.6 23.3 66.3 0.165 0.258 0.234 7.70 Soil Pots, High 9.56 17.1 11.4 38.0 0.228 0.295 0.265 7.74 Soil Pots, Low 10.7 12.9 12.3 35.8 0.241 0.337 0.275 7.57 Sawdust Pots, High 9.72 15.5 11.6 36.8 0.208 0.290 0.249 7.07 Significance ** ** ** ** ** ** NS NS 3reenhouse (1st harvest) Sawdust Pots, High 15.5 18.4 33.6 67.5 0.149 0.239 0.311 10.5 Sawdust Pillows, High 18.5 20.6 39.9 78.9 0.158 0.229 0.309 10.9 Soil Pots, High 18.5 21.2 30.1 69.9 0.156 0.251 0.365 9.33 Soil Pots, Low 20.9 18.9 24.9 64.7 0.172 0.290 0.344 8.55 Significance NS NS * * * ** NS NS Breenhouse (2nd harvest) Sawdust Pots, High 10.8 11.6 22.4 0.093 0.0873 Sawdust Pillows, High 22.2 19.0 41.2 0.110 0.0990 Soil Pots, High 16.4 17.5 33.9 0.102 0.0823 Soil Pots, Low 20.4 21.9 42.2 0.111 0.106 Significance * * * * NS * s igni f icance p<0.05 ** signif icance p<0.01 N S no signif icance p>0.05 65 greater than plants in the pot-culture treatments. Soil-f i l led pots set into the ground had greater aerial organ yield than the above-ground treatments. T h e pattern of leaf, s tem, and flower proportions was consistent among treatments; s tems (35%-45%) contributed most of the plant dry weight fol lowed by f lowers (30-35%) and leaves (24-26%). Plant water status was a s s e s s e d using water potentials, osmot ic potentials, and dry to fresh weight ratios. The water and osmot ic potentials were significantly lower in the two above-ground soil-filled pot treatments, and the low rate fertigation treatment had the lower water potentials of the two (Table 3.4). T h e calculated cell turgor was higher at night, particularly in the three above-ground treatments, but during the day turgor was significantly less and was negative in low rate fertigation soil and sawdust pots (Table 3.4). The plasticulture and the inset-pot treatments had the highest water and osmot ic potentials during the day. Turgor at night in these treatments was lower than that of other treatments. The dry to fresh weight ratios (Table 3.3) were significantly lower in the plasticulture and the inset soil-filled pot treatments indicating the higher water content of these plants. The flower dry to fresh weight ratio was the s a m e as the stem dry to fresh weight ratio in the plasticulture treatments, unlike in the other treatments where stem dry to fresh weight ratios were greater than that for flowers and leaves. Al l three of the above-ground soil or sawdust pot-culture treatments had low water potentials, low osmot ic potentials, high dry to fresh weight ratios and the lowest dry weight yields. Overal l , s tems had the highest dry to fresh weight ratio fol lowed by f lowers and leaves. 66 T a b l e 3.4. Leaf water potential and osmot ic potential of g reenhouse and field grown plants measured at 3 p.m. (light) and 4 a.m. (dark). Turgor was calculated by subtracting osmot ic potential from water potential. Potentials and turgor are reported in M P a . Location and Treatments Water potential Osmotic potential Turgor Field Dark Light Dark Light Dark Light Plasticulture, Low -0.288 -1.35 -1.12 -1.27 0.83 0 Inset Pots, Low -0.275 -1.21 -1.04 -1.30 0.77 0.09 Soil Pots, High -0.300 -1.56 -1.23 -1.52 0.93 0 Soil Pots, Low -0.770 -2.05 -1.66 -1.85 0.89 0 Sawdust Pots, High -0.283 -1.59 -1.20 -1.43 0.92 0 Significance * ** ** ** * * Greenhouse (1st harvest) Sawdust Pots, High -0.722 -1.71 -1.57 -1.85 0.850 0.14 Sawdust Pillows,High -0.772 -1.76 -1.55 -1.90 0.775 0.14 Soil Pots, High -0.783 -2.04 -1.56 -1.96 0.777 0 Soil Pots, Low -0.974 -2.17 -1.53 -2.08 0.558 0 Significance ** * NS NS * NS Greenhouse (2nd harvest) Sawdust Pots, High -0.619 -1.55 -1.37 -1.65 0.760 0.10 Sawdust Pillows,High -0.567 -1.59 -1.28 -1.70 0.714 0.11 Soil Pots, High -0.752 -1.83 -1.36 -1.76 0.607 0 Soil Pots, Low -0.862 -1.73 -1.21 -1.68 0.548 0 Significance * NS NS NS NS * s igni f icance p<0.05 ** signif icance p<0.01 N S no signif icance p>0.05 67 T a b l e 3.5. Ave rage leaf parthenolide concentration in g reenhouse and field-grown plants. The first leaf sample was a crop average for both the greenhouse and field taken just prior to transplanting. The first sample was 7.4 mg parthenol ide/gram dry leaf. Both field and greenhouse crops were harvested 73 days after transplanting ( D A P . The greenhouse crop was regenerated. Samp les 123 D A P and 168 D A P were taken from regenerated plants. Units are mg parthenolide/g dry leaf. Locat ion and Treatments T ime of Sampl ing Field First qrowth S e c o n d growth 32 D A P 73 D A P 123 D A P 168 D A P Plasticulture, Low 6.00 1.74 Inset Pots , Low 7.65 2.73 Soi l Pots , High 6.99 2.03 Soi l Pots , Low 6.30 1.89 Sawdus t Pots , High 6.11 3.32 Signi f icance N S * Greenhouse Sawdus t Pots , High 6.09 2.13 24.1 23.7 Sawdus t Pi l lows, High 5.37 2.84 22.3 27.6 Soi l Pots , High 6.66 4.61 24.2 27.2 Soi l Pots , Low 7.71 2.08 20.8 27.7 Signi f icance N S * N S N S * signi f icance p<0.05 ** signif icance p<0.01 N S no signif icance p>0.05 68 T a b l e 3.6. Parthenol ide content (mg/plant) at harvest (73 D A P ) in leaf, s tem, and flower t issues, based on subsample analys is. Units of measurement are mg parthenolide/g dry weight. Locat ion and Treatments Parthenol ide content per plant Field Leaf S tem Flower Total Plasticulture, Low 69.5 10.5 385 460 Inset Pots , Low 43.7 3.99 187 234 Soi l Pots , High 19.7 2.56 91.0 113 Soi l Pots , Low 20.3 1.93 98.0 120 Sawdust Pots , High 30.9 2.32 93.0 126 Signi f icance ** ** ** ** Greenhouse (1 s t harvest) Sawdus t Pots , High 17.7 2.76 269 289 Sawdus t Pi l lows, High 23.9 3.08 319 346 Soi l Pots , High 49.0 3.19 241 293 Soi l Pots , Low 29.7 2.83 200 232 Signi f icance * N S * * Greenhouse ( 2 n d harvest) Sawdust Pots , High 261 1.75 263 Sawdust Pi l lows, High 631 2.85 634 Soi l Pots , High 454 2.63 457 Soi l Pots , Low 592 3.28 595 Signi f icance ** N S ** * signif icance p<0.05 ** signi f icance p<0.01 N S no signif icance p>0.05 Parthenol ide concentrat ion of feverfew leaves was a s s e s s e d from samples taken three t imes during a 73-day period (Table 3.5). The initial sampl ing was done just prior to the application of treatments and parthenolide concentrat ion was 7.4 mg/g dry leaf. A s the summer progressed, plants developed from the vegetative to the flowering stage. Parthenol ide concentration remained nearly unchanged (avg. 6.8 mg/g dry leaf) over the first 36 days. However, when the crop was at the peak of f lowering (73 D A P ) leaf parthenolide concentration decreased significantly (avg. 2.5 mg/g dry leaf). This 69 pattern of dec reased parthenolide concentration as the crop matured was observed in exper iments d i scussed in Chapter 2. Parthenol ide concentrat ion was greatest in plants grown in the sawdust filled pots (3.3 mg/g dry leaf) and in plants grown in the soil filled pots set into the ground (2.8 mg/g dry leaf), both of which had low water and osmotic potentials. Ana lys is of shoot organs and the parthenolide content per plant showed that f lowers produced 7 4 % to 8 4 % o f the total plant parthenolide, leaves contributed 1 5 % to 2 5 % , and stem produced 1.4% to 2 .3% (Table 3.6). High dry weight yields of f lowers and leaves in response to the plasticulture treatment resulted in very high whole-plant parthenolide concentrat ions, even though leaves under this treatment produced the lowest parthenolide content by dry weight. Total average plant parthenolide content ranged from 450 mg per plant under the plasticulture treatment to 125 mg per plant from above-ground, soil-filled pots. Regard less of irrigation f requency or medium, total plant parthenolide in leaves, s tems, and f lowers was very similar from the three above-ground pot treatments. 3.3.2 Greenhouse growing medium and fertigation trial There were four treatments in the greenhouse trial, three of which were also appl ied in the concurrent field trial. A fourth greenhouse treatment was the use of sawdust filled pillow bags, a treatment which had been successfu l ly used in another experiment (Chapter 2) resulting in high leaf parthenolide concentrat ion. Flowers accounted for the most dry matter (37% - 50%), fol lowed by leaves (21% - 31%) and s tems (21% - 27%) (Table 3.3). The sawdust treatments produced more flowers and therefore a 2 5 % higher f lower dry weight than soil treatments. Plants in the greenhouse had 10% - 30 % more f lowers than field plants. 70 Fever few is a long day plant, requiring short periods of darkness for f lowering. In my exper ience, f lower induction has occurred under days longer than approximately 12 hours. Crop regeneration in the greenhouse was under short days which held plants in a vegetative stage for the three months until the experiment was terminated. Leaf and stem yields were similar between the regenerated crop and the first growth (Table 3.3). The yields from the first crop were not significantly affected by treatments but from the regenerated crop, the sawdust-pot treatment had lower stem and leaf yields compared with that in the other three treatments. Dry to fresh weight ratios were measured to a s s e s s plant water content. The dry to fresh weight ratios in f lowers were higher (avg. 0.34) than in the s tems (avg. 0.26) and the leaves (avg. 0.16) of plants grown in the greenhouse (Table 3.3). The field grown plants had lower dry to fresh weight ratios for f lowers (avg. 0.25) than for s tems (avg. 0.29). In the greenhouse crop, dry to fresh weight ratios in leaves and stems were greater in response to the low level fertigation in soil filled pots compared to the other treatments. The dry to fresh ratios in the regenerated crop was 3 5 % lower in leaves, and 7 0 % lower in s tems compared with that of the first crops from the greenhouse and field. Unl ike the flowering plants, the regenerated vegetative plants had dry to fresh weight ratios that were greater in vegetative leaves (avg. 0.105) than in s tems (avg. 0.09)). Leaf water and osmot ic potentials show that plants from the low-level fertigation soil treatment resulted in the lowest mid-day and night water potentials at - 2 . 2 M P a and - 0 . 9 5 M P a respectively (table 3.4). T h e s e results resemble the field trial where the low-fertigation soil-fil led pot treatment had the lowest water potentials at mid-day and at night. The soil treatments had lower water potentials than did the sawdust treatments. Turgor was very low for plants in the soil-filled pots during the day and the low-irrigation 71 treatment resulted in significantly lower turgor at night than did the other treatments. In the regenerated crop, water potential was higher than the first growth but lower than in the field trial. Water potentials in the regenerated crop were the lowest in response to the low water treatment. Overal l , plants in the greenhouse trial had lower water potentials than those of the field trial. In the first g reenhouse crop, mid-day and night leaf water potentials averaged -1.95 M P a and - 0 . 8 3 M P a , respectively, whereas the field trial mid-day and night water potentials averaged - 1 . 6 3 M P a and - 0 . 5 3 M P a , respect ive ly . Leaf parthenolide concentration changed with plant development. Leaf parthenolide concentrat ion in the first g reenhouse crop showed the s a m e relationship with development as in the field trial where leaf parthenolide concentrat ion decreased as the plants matured (Table 3.5). After crop regeneration and growth under short days however, leaf parthenolide concentration was 10 t imes higher in the vegetative plants compared to the flowering plants of the previous crop. Th is s a m e relationship was observed in the work descr ibed in Chapter 2 (Table 2.6) where parthenolide levels became high when plants were held in a vegetative stage in the greenhouse. There were no significant treatment effects on parthenolide concentrat ion in the regenerated feverfew crop. Compar ing dry weights of plants from the field and greenhouse experiments, the regenerated crop had dry leaf weights within the same range as those from both the field trial and the first growth of the greenhouse trial under comparable treatments (Table 3.3). G reenhouse flower yields were two to three t imes higher than field f lower yields from plants from the potted treatments. The high ratio of f lower to stem dry weight in the greenhouse crop did not occur in the field. Fever few grown in plasticulture produced higher yields than it did under all other treatments in the field and 72 greenhouse. Most of the parthenolide from flowering plants was in f lower t issue. For plants in the vegetative stage most of the parthenolide was in leaf t issue (Table 3.6). Total plant yield of regenerated greenhouse plants was lower than the first greenhouse crop yield, and approximately 3 0 % of the yield from the plasticulture treatment. Due to high leaf parthenolide in the regenerated greenhouse plants, the whole-plant parthenolide concentration was greater than for the large f lowering plants of the field plasticulture treatment. Of the four greenhouse treatments used to produce the regenerated crop, the sawdust-f i l led pillow bag and the soil-fil led pot, with low-level fertigation produced the most parthenolide per plant. 3.4. D I S C U S S I O N Historically (>400 years) feverfew has been harvested from outdoor sources such as gardens or from the wild. It is field-grown commercia l ly but much of the research on its phytochemistry has been conducted on plants grown in greenhouses (Heptinstall et al., 1998). Feverfew from both the greenhouse and field has been used in clinical trials, pharmacological studies, and for compar isons of commercia l products (Heptinstall et al., 1998; Awang et al., 1991) but the effects of the two environments on plant chemistry have not been studied until now. Cl in ical trials showing feverfew efficacy in migraine prophylaxis used either greenhouse or f ield-grown plants (Johnson et al., 1985; Murphy et al., 1988) and in s o m e c a s e s the origin was not ment ioned (Pattrick et al., 1989). L ikewise, the developmental stage at which feverfew leaves were harvested is a lso not mentioned. In this study, leaf parthenolide concentration during vegetative growth under short days in the greenhouse was 2-4 t imes higher than had been previously reported by Awang et al. (1991), a result consistent between two greenhouse exper iments (see also Chapter 2). In both the field and greenhouse 73 exper iments presented here, leaf parthenolide concentrat ion dec reased as the crop matured from the vegetative to reproductive stage. Developmenta l s tage appears to be an important determinate in S T L biosynthesis. The exper iments descr ibed in this chapter were des igned to examine differences between greenhouse and field production methods and explore the effect of environment on plant physiology and S T L metabol ism. Exper iments presented in Chapter 2 were conducted in the greenhouse and field but different growing media and at different t imes of the year made direct compar isons inappropriate. The field and greenhouse trials presented in this chapter were des igned with matching treatments, and conducted at the same time of year to eliminate the difference of daylength, growing media , and pot s ize between the two locations. Limited capaci ty for root growth in the confined or compacted pots, large diurnal fluctuation of root temperature of above ground pots, and dif ferences in drainage between treatments are three factors which may have caused yield difference between field plasticulture and potted treatments. In the annual spec ies Abutilon theophrasti and Setaria faberii, limitations in physical space for root growth resulted in reduced vegetative growth and lower reproductive output. Abutilon theophrasti decreased al location to reproductive t issues relative to vegetative t issues whereas Setaria faberii responded with earlier flowering and higher reproductive output in smal ler soil vo lumes (Mcconnaughay and B a z z a z , 1991). Thus the developmental response to root restriction depends on the spec ies but yield usual ly dec reases . For example, root restriction in Gossypium hirsutum caused reduced shoot b iomass and leaf area (Thomas and strain, 1991), and with Salvia splendens reduced growth was positively correlated with container volume (Vaniersel , 1997). Compac ted soil can have a similar effect on root restriction and can affect drainage. Fever few grown in plasticulture had 74 5 0 % - 7 5 % higher dry weight yields than potted plants in both field and greenhouse (Table 3.3) in a season where yields were not optimal. S ince above ground environments were the same among treatments in the f ield, the difference may have been caused by an effect on the roots. Above ground potted treatments had higher shoot dry weight yields in the greenhouse than in the field at the first harvest (Table 3.3). Y ie lds were nearly double in the greenhouse, with f lower weights contributing most to the higher yield. S tems contributed most to dry weight of f ield-grown plants. O n e of the differences between the field and greenhouse was temperature, which fluctuated in the field between 35°C and 10°C while in the greenhouse ranged between 38°C and 18°C. In Chrysanthemum sp. , warmer temperatures result in earlier flowering (Larsen, 1982). In Chamomilla recutita, fresh weight yield increases in response to temperature (Fahlen et al., 1997). Another difference between field and greenhouse cl imates is air movement. Wind alters stomatal conductance, which in turn alters C 0 2 uptake affecting carbon assimilat ion and growth (Brenner et al., 1995; Cordero , 1999). W ind also increases lignification of s tems which is consistent with the higher s tem dry weights in field-grown compared to greenhouse-grown plants. S o m e of the energy required for stem lignification might otherwise be redirected to growth and flowering in the greenhouse. Soil-fi l led pots, particularly under low-level fertigation, resulted in lower leaf water and osmot ic potentials, lower turgor, and higher dry to fresh weight ratios. Plant water content was low in these potted soil treatments compared to sawdust treatments, which may have been caused by a low matric potential in soil compared to that in sawdust. If that was the case , a greater water potential gradient existed for plants grown in soil resulting in lower water content in plant t issues. The low-level fertigation soil treatment 75 resulted in a lower osmot ic potential which may have been a response to the higher gradient. Lower water and osmot ic potentials coincided with lower dry weight yields in the potted soil treatments. Dry to fresh weight ratios revealed that the water content was higher in g reenhouse plants compared to field plants (Table 3.3). General ly , plants with high water potential and low dry to fresh weight ratio had higher dry weight yields. Other studies have found that low water potential usual ly results in high dry to fresh weight ratio and low yield (Flenet et al., 1996; Kimura et al., 1994). In the f ield, f lower dry to fresh weight ratios were lower than stem dry to fresh weight ratios (Table 3.3 and Chapter 2), but the opposi te was found in greenhouse plants. A s with most plants used in traditional medic ine, chemica l quality is difficult or impossib le to a s s e s s because there is little research on active constituents and their physiological roles. Therefore, when we know how speci f ic chemica l concentrat ions vary there is little information on which to a s s e s the pharmacological quality. In an effort to establ ish some regulation for quality control of herbal remedies, Health and Wel fare C a n a d a proposed that feverfew products contain a minimum of 0.2 % parthenolide (Heptinstall and A w a n g , 1998). This guidel ine is widely used in C a n a d a to ensure spec ies authenticity and a standard of quality. Parthenol ide was probably chosen because it has been the focus of pharmacological research and found to be a vasoregulator (Barsby et al., 1991; Groenewegen and Heptinstal l , 1990). The antimigraine activity of feverfew is thought to be mediated by the effect of parthenolide on serotonin (Maries et al., 1992; Biggs et al., 1982; Groenewegen et al., 1992). Parthenol ide is an early metabolite in sesqui terpene lactone (STL) biosynthesis (Fischer et al., 1979) and most other S T L s in feverfew are likely der ived from parthenolide or its immediate precursor. S ince most S T L s have pharmacological activity, parthenolide 76 might be a useful indicator of total S T L biosynthesis. For example , low parthenolide in flowering plants may be a result of metabol ism and hence a higher concentration of downstream S T L s . O n e important quest ion to producers of medicinal plants is: What conditions result in the highest crop quality and yield? Methods of quality assessmen t may include visual appearance , purity of selected organs such as leaves, and chemica l composi t ion. For feverfew, these quality standards have not been set. Currently, feverfew shoots are ground and sold compressed or encapsula ted, so v isual quality or the proportion of leaves, s tems, and flowers is not identifiable by the consumer and thus may not be considered important. High leaf and low f lower/stem composi t ion in commercia l products should be an indicator of higher quality s ince leaves appear to have been selected historically and they have thus become the focus for pharmacological and clinical research. The present study found that it would be eas iest to produce plants with high leaf to stem ratio, no f lowers, and high leaf parthenolide concentration in the greenhouse under short days. Ideally, the chemical compl iment and concentration are the most important measure of quality in a medicinal plant. If parthenolide and other S T L s are the only group of active compounds in feverfew that relieve migraine, flowers rich in S T L s may be an important ingredient in feverfew preparat ions for migraine prophylaxis. Select ion for high quality feverfew crops in terms of S T L concentration in leaf and flower while maintaining high yields is possib le. Th is study shows that organ proportions and parthenolide concentration in feverfew can be manipulated by modifying the environment. This work provides information on how quality can potentially be maximized without genet ic modification or significantly increasing overhead costs. 77 Many commerc ia l preparations use the whole aerial portion of feverfew. It was interesting to look at parthenolide concentration of the whole plant based on its component organs to see the environmental effect on both yield and parthenolide concentrat ions. Plants in the reproductive stage had the majority of parthenolide in the f lowers (74% to 84%), followed by leaves and very little in s tems even though stems represented 2 5 % to 5 0 % of the plant dry weight. Th is is consistent with results of another study in which flower heads were found to contain the highest parthenolide concentration (Brown et al., 1996). A m o n g the three comparab le treatments in the greenhouse and field, shoot parthenolide content was considerably higher in g reenhouse plants than in field plants. The high greenhouse f lower yield contributed most to the higher total parthenolide concentration in g reenhouse plants. Flower parthenolide concentrat ion based on dry weight was consistent ly high among treatments. The leaf parthenolide contribution to the whole plant was a lso higher in the greenhouse. For commerc ia l preparations, the signi f icance of harvesting at the correct developmental stage is important for quality and possibly eff icacy. Commerc ia l preparations from flowering plants may contain high proportions of f lowers, which are not traditionally used for migraine therapy. A feverfew preparation rich in f lowers could contain very high parthenolide levels among other S T L s . Preparat ions containing a low proportion of leaves to f lowers and stems have not been tested for s ide effects or eff icacy, which may be important because f lowers are reported to contain different terpenoids than the leaves (Banthorpe et al., 1990; Dolman et al., 1992). More research is required on the pharmacology of feverfew to identify all the antimigraine compounds before we will know which organs are the safest and most eff icacious to use. 78 Regenerat ion of feverfew in the greenhouse resulted in a 4 0 % lower dry weight yield than the first crop, primarily due to the lack of f lowers. In my exper ience, feverfew is a long day plant requiring periods of darkness shorter than 12 hours to flower. The regenerated plants remained in a vegetative stage due to the short photoperiod. However, dry weight yields of s tems and leaves in the second crop were similar to the first crop. Plants grown in the sawdust pil lows had the highest yields and were the only treatment to produce more leaf t issue than stem t issue. In addit ion, the first crop grown in the greenhouse in sawdust-f i l led pillow bags had lower yield than the regenerated crop. The pillow bag is long and tubular and contains the s a m e volume of sawdust as the pot but has greater surface area. This may result in higher root temperatures and contribute to the higher shoot yield found in plants grown in the pillow bag treatment. Regrowth from the establ ished crown should have resulted in quicker maturation and growth due to the resources avai lable in the roots. The dry to fresh weight ratios of the regenerated crop reveal two interesting features of the vegetative plants. Leaf dry to fresh weight ratios were greater than stem dry to fresh weight ratios, and both were low compared to those of the first crop indicating the second growth had high water content. Th is may have resulted because s tems of a vegetative plant are not as woody as s tems from flowering plants. Stem and leaf dry to fresh weight ratios were highest in response to the low-level fertigation soi l-filled pot, the same affect as was found in the first g reenhouse crop. Sawdust-f i l led pillow bags are used in commercia l g reenhouses throughout the Fraser Val ley, B .C . to grow a variety of vegetables. The results for yield and dry to fresh weight ratios indicate pillow bags are a viable way of producing feverfew and may have potential in g reenhouse production of other herbaceous medicinal plants. Plant regeneration in the greenhouse under short days produced plants with a low proportion of s tem t issue and, 79 in the pillow bag treatment, higher leaf yields. In British Co lumb ia , sawdust is readily avai lable as a forestry byproduct. It provides good aerat ion, adsorbs minerals, has a high water holding capacity while allowing excel lent dra inage, and is inexpensive and biodegradable. T h e s e attributes make sawdust an excel lent media for growing some spec ies of plants including feverfew. In the greenhouse experiment presented in Chapter 2, feverfew grown in a sawdust medium produced a high parthenolide concentrat ion compared to plants grown in the field and compared to other reported results. The high parthenolide concentration could not be attributed to any particular feature in the greenhouse s ince many dif ferences existed between greenhouse and field condit ions. The experiments in Chapter 3 were des igned partly to determine if sawdust medium was promoting high leaf parthenolide concentrat ions. It is now clear that growing feverfew in the sawdust medium did not account for the high parthenolide concentrat ions observed in prior exper iments. Instead, developmental status of the plants is an important determinant in the accumulat ion of parthenolide. The regenerated greenhouse crop was held in a vegetative stage and it produced high leaf parthenolide concentrat ions regardless of sawdust or soil media. Exper iments in this thesis consistently show plants sampled in the vegetative stage had the higher leaf parthenolide than f lowering plants. In the greenhouse and field experiments conducted in the summer , parthenolide concentration was within the range of reported va lues (<1%) early in the s e a s o n when plants were still vegetative. In the regenerated crop, the extended growth period (3 months) of the vegetative stage may cause a build-up of parthenolide to high concentrat ions, a result not observed in the vegetative stage of field trials. It is possib le that the onset of flowering triggers a reduction in parthenolide biosynthesis or an increase in parthenolide catabol ism, resulting in lower concentrat ions as plant development proceeds. 80 The primary contributor to whole plant parthenolide concentrat ion in the regenerated crop was the leaves with only smal l amounts from the stems. The higher levels of total plant parthenolide in regenerated plants compared with flowering plants of f ield-grown and greenhouse-grown crops were surprising because the yield was low in the regenerated crop. The plasticulture treatment was the only one that produced similar amounts of parthenolide, primarily due to high b iomass but low parthenolide concentrat ion. One major difference between regenerated and first crops was that leaves were the primary source of parthenolide in the regenerated plants, whereas in the first g reenhouse crop and field-grown crops the primary source of parthenolide was f lowers. The regenerated greenhouse crop had more than 10x the leaf parthenolide concentrat ion than the flowering crops. The organ source of parthenolide may be important in consider ing product efficacy because there are different chemica ls found in the tr ichomes of each organ (see next chapter). Leaves from vegetative plants contain primarily parthenolide but leaves and f lowers from flowering plants contain complex mixtures of compounds and parthenolide may not be the most abundant. The dif ferences in parthenolide concentrat ions and dry weight yields between the regenerated crop and flowering crops may have been affected by regeneration or by photoperiod. The effects of photoperiod are investigated in Chapter 4. Daylength or photoperiod has a variety of physiological effects in plants. T h e s e effects in the plant are mediated by photoreceptors which sense changes in light duration, intensity, and wavelength. Much work focused on how day length affects f lowering and how light perception differs in spec ies . A s already ment ioned, f lowering in feverfew is induced by long days. In some way the critical photoperiod for f lowering, or the flowering process itself plays a role in modifying the parthenolide concentrat ion of leaves. Normally plants germinate in the spring when days get longer. This results in the plant remaining in a 81 vegetative stage for only a short period, developing quickly toward f lowering. The changing S T L composit ion in feverfew may be part of an adaptat ion strategy against herbivores and pathogens to ensure reproductive s u c c e s s . Pharmacolog ica l research (Groenewegen and Heptinstal l , 1990; Groenewegen et al., 1986; Heptinstall et al., 1987; Losche et al., 1988) and clinical trials (Johnson et al., 1985; Murphy et al., 1988) have been conducted with plants grown either in the greenhouse or field and often location or developmental stage is not mentioned. From the results presented in this chapter, there are significant di f ferences in the effects of g reenhouse and field condit ions on parthenolide concentrat ion, dry weight yield, and organ proportions. Reports of chemical investigations and clinical trials should thus state the developmental stage of the plant and whether it was greenhouse or field grown. 82 3.5. R E F E R E N C E S A w a n g , D .V .C . , Dawson , B.A., Kindack, D .G. , Crompton, C . W . , and Heptinstal l , S . (1991) Parthenol ide content of feverfew (Tanacetum parthenium) a s s e s s e d by H P L C and 1 H - N M R spectroscopy. Journal of Natural Products 54:1516-1521 Banthorpe, D. V . , Brown, G . D., J a n e s , J . F, and Marr, I. M. (1990) Parthenol ide and other volati les in the f lowerheads of Tanacetum parthenium I. Schu l tz bip. Flavour and Fragrance Journal 5(3): 183-186. Barsby, R., S a l a n , U., Knight, D.W., and Hoult, J . R . (1991) Irreversible inhibition of vascu lar reactivity by feverfew Lancet 338:1015 Barsby, R., S a l a n , U., Knight, D.W., and Hoult, J . R . (1993) Fever few and vascular smooth musc le : extracts from fresh and dried plants show oppos ing pharmacological profiles, dependent upon sesqui terpene lactone content. Planta Med ica 59:20-25. Biggs, M .J . , Johnson , E . S . , Pe rsaud , N.P. , and Ratcliffe, D .M. (1982) Platelet aggregat ion in patients using feverfew for migraine. Lancet 2:776 Brenner, A . J . , Jarv is , P . G . , and Vandenbeldt , R . J . (1995) Windbreak-Crop interactions in the sahe l . 2. Growth -Response of millet in shelter. Agricultural and Forest Meteorology 75:235-262 Brown, A . M . G . , Lowe, K . C . , Davey, M.R. , and Power , J . B . (1996) Fever few (Tanacetum parthenium): T i ssue culture and parthenolide synthesis. Plant S c i e n c e 116:223-232 Cordero , R.A. (1999) Ecophys io logy of Cecropia schreberiana sapl ings in two wind regimes in an elfin cloud forest: Growth, gas exchange, architecture and stem b iomechanics Tree Physio logy 19:153-163 Do lman, D .M. , Knight, D.W., Sa lan , U., and Topl is, D. (1992) A quantitative method for the estimation of parthenolide and other sesqui terpene lactones containing alpha methylenebutyrolactone functions present in feverfew Tanacetum parthenium. Phytochemica l Ana lys is 3:26-31 Fah len , A . , Welander , M. , and Wenners ten , R. (1997) Effects of light-temperature regimes on plant growth and essent ial oil yield of selected aromatic plants. Journal o f the Sc ience of Food and Agriculture 73:111-119. F ischer , N .H . , Olivier, E . J . , and F ischer , H.D. (1979) T h e b iogenes is and chemistry of sesqui terpene lactones. In: Progress in the chemistry of organic natural products. Spr inger-Ver lag, W ien . Eds . Herz, W. , Gr i sebach , H., and Kirby, G . W . Flenet, F., Bounio ls , A . , and Sara iva , C . (1996) Sunf lower response to a range of soil water contents. European Journal of Agronomy 5:161-167 83 Groenewegen , W . A . , and Heptinstall, S . (1990) A compar ison of the effects of an extract of feverfew and parthenolide, a component of feverfew, on human platelet activity in-vitro. Journal of Pharmacy and Pharmaco logy 42:553-557 Groenewegen , W . A . , Knight, D. W. , and Heptinstall, S . (1986) Compounds extracted from feverfew that have anti-secretory activity contain an alpha-methylene butyrolactone unit. Journal of Pharmacy and Pharmaco logy 38:709-712. G roenewegen , W . A . , Knight, D.W., and Heptinstall, S . (1992) Progress in the medicinal chemistry of the herb feverfew. Progress in Medic inal Chemist ry 29:217-238 Heptinstal l , S . , and A w a n g , D .V .C . (1998) Feverfew: a review of its history, its biological and medicinal properties, and the status of commerc ia l preparations of the herb. In: Phytomedic ines of Europe, chemistry and biological activity pp. 158-175, Amer ican Chemica l Society, Wash ington. E d s . Lawson , L. and Bauer R. Heptinstal l , S . , G roenewegen , W . A . , Spangenberg , P. , and L o e s c h e , W . (1987) Extracts of feverfew may inhibit platelet behaviour via neutralization of sulphydryl groups. Journal of Pharmacy and Pharmaco logy 39:459-465. Johnson , E . S . , K a d a m , N.P. , Hy lands, D.M. , and Hylands, P . J . (1985) Eff icacy of feverfew as prophylactic treatment of migraine. British Medica l Journal 291:569-573 K imura , M. , Ichimura, M. , and Tomitaka, Y . (1994) Effects of watering on growth, yield, essent ia l oil concentration and evapotranspirat ion of sweet basil (Ocimum basilicum L ) . Journal of Tropical Agriculture. 38:65-72. Larsen , R. (1982) The effect of night temperature on f lower initiation and early differentiation of Chrysanthemum morifolium. Swed ish Journal of Agricultural R e s e a r c h . 12:95-102. Loesche , W . , Groenewegen , W.A . , Krause , S . , Spangenberg , P. , and Heptinstall, S . (1988) Effects of an extract of feverfew (Tanacetum parthenium) on arachidonic acid metabol ism in human blood platelets. B iomed ica et B iochimica Acta 47 :S241-243 Loesche , W. , Miche l , E. , Heptinstal l , S . , Krause , S . , G roenewegen , W.A . , Pesca rmona , G . P . , and Th ie lmann, K. (1988) Inhibition of the behaviour of human polynuclear leukocytes by an extract of Chrysanthemum parthenium. P lanta Med ica 54:381-384 Mar ies , R . J . , Kaminsk i , J . , A rnason , J .T. , P a z o s - S a n o u , L., Heptinstal l , S . , F ischer , N.H. , Crompton, C .W. , Kindack, D .G. , and A w a n g , D.V. (1992) A b ioassay fo r inhibition of serotonin re lease from bovine platelets. Journal of Natural Products 55:1044-1056 McConnaughay , K . D . M . , and B a z z a z , F.A. (1991) Is physical s p a c e a soil resource. Ecology, 72:94-103 84 Murphy, J . J . , Heptinstal l , S . , and Mitchell , J . R . (1988) Randomised double-bl ind placebo-control led trial of feverfew in migraine prevention. Lancet 2:189-192 Pattrick, M. , Heptinstal l , S . , Doherty, M. (1989) Fever few in rheumatoid arthritis: a double blind, p lacebo controlled study. Anna ls of Rheumato id D isease 48:547-549 Thomas , R .B . , Strain, B.R. (1991) Root restriction as a factor in photosynthetic accl imation of cotton seedl ings grown in elevated carbon-dioxide. Plant Phys io logy 96:627-634 Van ie rse l , M. (1997) Root restriction effects on growth and development of salvia (Salvia splendens). Hortscience 32:1186-1190 Chapter 4 85 Developmental effects on glandular trichomes and leaf chemistry 4.1 I N T R O D U C T I O N Var ious environmental cues affect plant metabol ism. Light is an important environmental st imulus that plants respond to physiological ly via photoreceptors, which initiate signal transduction pathways (Sal isbury, 1982). Per iod ic stimulation of some a photoreceptors such as phytochrome affect periodic p rocesses like f lower initiation. Both photoperiod and the developmental changes assoc ia ted with f lowering may affect tr ichome chemistry and density (Circel la et al., 1995; Hendr iks et al., 1996; Voir in et al., 1990). Tr ichomes develop on the aerial organs of most plants from a single protodermal cell (Szymansk i et al., 2000). They are spec ia l ized structures that protrude from the epidermis and can be uni- or multi-cellular, and glandular or non-glandular. O n e or more types of t r ichomes may occur on any one spec ies and can be useful attributes for taxonomic identification. Glandular t r ichomes typically produce, accumulate, and secrete chemica ls that general ly are terpenoids, phenol ics, and resins (Behnke, 1984). The glandular tr ichome in Figure 4.1 is the type present on the epidermis of a feverfew leaf. It has cells that produce compounds that are secreted into a subcut icular space for storage. The contents of the g lands provide the plant with a chemical defense against insects, herbivores, and pathogens (Behnke, 1984). Both glandular and non-glandular tr ichomes provide another defensive function by forming physical barriers to insects. They also create a boundary layer of air that reduces water loss and dampens temperature fluctuations. Tr ichome traits conserved within a spec ies include time of tr ichome initiation, developmental patterns, and spac ing (Szymansk i et 86 al., 2000). Tr ichomes are important for plants as the first line of defense against phytopathic organisms and therefore their involvement in chemica l eco logy may be critical. Sesqu i terpene lactones (STLs) are produced in feverfew tr ichomes but little is known about biosynthesis and regulation within these structures. Biosynthesis of sesqui terpenes and monoterpenes has been investigated in the glandular tr ichomes of Artemisia annua (Tel lez et al., 1999), Helianthus annuus (Spring et al., 1992), and Mentha spec ies (McConkey et al., 2000). In these plants biosynthesis occurred in the underlying secretory cel ls of the subcuticular space . F i g u r e 4 . 1 . A glandular tr ichome derived from an epidermal cell with a subcuticular, extracellular space where secretory cel ls secrete non-polar compounds for storage. Fever few contains a mixture of more than 25 S T L s in the glandular tr ichomes. Parthenol ide concentrat ion is general ly higher than the other S T L s but feverfew chemotypes have been identified in which parthenolide was not found (Heptinstall, A w a n g , 1998). Variat ion in S T L composit ion of feverfew leaves is common. In Chapters 2 and 3, I reported that parthenolide concentrat ion in feverfew is affected by development as the plant matures from vegetative to reproductive growth. The mechan ism for this variation is unknown but changes in tr ichome density, leaf f*' G landular Tr ichome Leaf epidermis 87 metabol ism, developmental stage, or photoperiod may affect changes in parthenolide concentrat ion. Exper iments presented in this chapter explore how tr ichome chemistry changes with development and organ type, age and location. 4.2 M A T E R I A L S A N D M E T H O D S General procedures Fresh plant material for spectroscopy was grown at U B C , Vancouver , B .C . , in a controlled greenhouse environment. Growing and sampl ing methods for plants used to collect data reported in figure 4.2 were the same as those descr ibed in the Materials and Methods sect ion of Chapter 2.2. The growing and sampl ing methods used in the other exper iments, and to obtain the t issues for microscopy, are descr ibed in Chapter 3.2. The standard extraction procedures and H P L C analys is used are descr ibed in Chapter 2.2. Al l solvents used for extraction and H P L C analys is were H P L C - g r a d e (Fisher). Trichome isolation and extraction Flower t r ichomes were removed individually from disk floret petals, composi te inf lorescence receptacles and leaves using a fine 27-gauge syringe needle and a dissect ing microscope at 300 x magnif ication. Exc ised t r ichomes were extracted in methanol and filtered using a 0.45 u.m P T F E syringe filter. The filtrate was dried under nitrogen, and resuspended in 1 ml methanol for H P L C analys is . Microscopy A new method for observing glandular tr ichomes on leaf sur faces was developed for this experiment. Four to eight epidermal peels from a single leaf were mounted on a sl ide with water and covered with a g lass cover slip. The sl ide preparations were placed in an oven at 40 °C for approximately 30 minutes until the epidermal peels were 88 dry before tr ichome densit ies were determined. Tr ichome images (Figures 4.9 and 4.10) were made usine a Ze iss Axioplan 2 imaging light microscope equipped with a D V C digital v ideo imaging camera . Scann ing electron micrographs were obtained with a Cambr idge 250T scanning electron microscope ( S E M ) . Preparat ions of fresh leaves and f lowers for the S E M were mounted on a S E M stud and coated with gold under vacuum in a Nanotech Semprep 2 sputter coater. Trichome density and parthenolide content of leaves Sl ide preparations of the dried epidermal peels were observed under a light microscope at 400 x magnif ication. The field of v iew was cal ibrated using a micrometer. O n e half of each leaf was used for epidermal sl ide preparat ions and tr ichome density measurements while the other half was extracted and ana lyzed for parthenolide concentration using H P L C . Tr ichome density was measured on a minimum of eight and a max imum of fifteen locations on the epidermal peels of each half leaf. Leaves from vegetative and flowering plants were sampled at 10 posit ions from the top to the bottom of the plant. The leaf samples were harvested from 8 plants, 4 flowering and 4 vegetative. Leaf rinsing Six fresh feverfew leaves of about the same age from a flowering plant were cut in half. The first half was dipped three t imes in 300 ml water for one second each and the other was not d ipped. Both halves were dried in an oven at 40 °C, seperately extracted with dichloromethane and parthenolide concentrat ion analyzed by H P L C . The water into which the leaf halves were dipped was extracted with dichloromethane. 89 4.3. RESULTS 4.3.1 Parthenolide variability during leaf development and flowering Feverfew grown under a short photoperiod (< 12 hours) in the greenhouse produced vegetative shoots for approximately 60 days after transplanting (DAT). Leaf parthenolide concentrat ions increased significantly over this two month period (from 7.5 to 36 mg parthenolide/g dry leaf) (Figure 4.2). The plants matured to the flowering stage under longer photoperiods (>12 hours of light / day) and leaf parthenolide concentrat ions decreased (34 to 5.8 mg parthenolide/g dry leaf). 1 29 50 71 98 days after transplanting Figure 4.2. Greenhouse-grown feverfew leaf parthenolide concentrat ion during development from vegetative to reproductive growth. The shaded area is the period when internodes elongated and the plants started f lowering. Error bars are standard error o f the mean. 90 In a second experiment where feverfew was grown under photoperiods >15 hours per day, the onset of reproductive growth occurred approximately 30 DAT. Aga in , the leaf parthenolide concentration decreased with the transition from vegetative to reproductive s tages (from 7.4 to 3 mg parthenolide/g dry leaf respectively) (Figure 4.3). Half of the plants were harvested when flowering was at a peak at 77 DAT. New shoots sprouted from the remaining crown under a photoperiod less than 12 hours. T h e s e new shoots remained in an extended vegetative stage. The other half of the plants were not harvested and were al lowed to continue f lowering. The new shoots that developed from the crown on these flowering plants remained in a vegetative stage while shoots with inf lorescence cont inued to mature but did not produce new flowers. Leaf samp les taken 127 and 172 D A T from the regenerated vegetative plants had significantly higher parthenolide concentrat ions than leaves from flowering plants (26 compared to 3.6 mg parthenolide/g dry leaf respectively) (Figure 4.3). L ikewise, leaves from the new vegetative shoots growing from the flowering plants had higher parthenolide concentrat ions than the leaves from flowering shoots of the same plant. Parthenol ide concentration varied with leaf age (Table 4.1). The growing tip sampled from vegetative plants consisted of the shoot apical meristem and smal l , unopened leaves. Parthenol ide concentrat ions in the growing tips were 13 mg/g dry weight which is similar to the concentration found in f lowers (11 mg/g dry weight) but were not the highest concentration in the plant. In vegetative shoots, parthenolide concentration was highest in leaves at the middle of the shoot whereas on flowering shoots the greatest concentration was in the older (basal) leaves and was less in younger leaves (top of plant). 91 F i g u r e 4 .3 . Leaf parthenolide concentration of greenhouse-grown feverfew over time. On day 77 half the crop was harvested and regenerated and the other half was not harvested. Plants not harvested remained in a flowering stage. 92 T a b l e 4 .1 . Parthenol ide concentration in leaves of different ages from vegetative and reproductive shoots. Developmental stage Vegetat ive** Reproduct ive Leaf posit ion* (mg parthenolide / g dry leaf) growing tip / f lower head 12.6 ± 2 . 0 10.7 ± 0 . 6 7 1&2 18.6 ± 0 . 8 0.80 ± 0 . 1 6 5&6 27.6 ± 3 . 0 2.49 ± 0.90 9&10 33.4 ± 2.2 4.27 ± 0 . 7 3 13&14 24.2 ± 2 . 9 3.64 ± 0.99 19&20 9.4 ± 2 . 0 6.88 ± 1.2 * leaf position from top to bottom of shoot ** mg parthenolide / g dry leaf ± standard error of the mean H P L C chromatograms of leaf surface extracts revealed that vegetative shoots had a high concentration of parthenolide and a peak area greater than the other chromatogram peaks at all leaf ages (Figure 4.4). However, shoots in the reproductive stage had lower parthenolide concentration and lower peak areas than other chromatogram peaks in both young and old leaves. Older leaves of vegetative shoots had a lower proportion of parthenolide, as did younger leaves of the reproductive shoots, relative to the other peaks in the chromatograms. 93 o . i o 0 . 0 8 -0.06-1 0.04-1 0.02 0.00—1 0. 00 3 < tz <D O c ro i— o V) . O < 0.10 0.08_| 0.06-0.04-0.02-0.00-0, 60 0.10 0.08—1 0.06 0.04 0.02 1L: o . o o 10.00 B 10.00 0.00 i 1 r -p. / • x i o ! o o Time (mm.) S 0.14 ~ 0.12^ 0.10-1 e a 0.08^ 0.06^ s 0.04-3 i 0.02^ u 1 0.00^ 0.14. 0.12. 0.10^ 0.08^ 0.06. 0.04-0.02. 0.00. 0.00 i o ! o o 0.00 Time (min.) 10.00 94 Figure 4.4. (Previous page) H P L C chromatograms of leaf and f lower extracts from feverfew shoots in the vegetative and reproductive s tages. Leaves were counted from the shoot apex beginning with the first fully expanded leaf and counting toward the base of the plant. Leaves from shoots in the vegetative stage (A-D), f lowers (E) and leaves from shoots in the reproductive stage (F-H). A ) Growing tip and unopened leaves. B) First and second fully expanded leaves. C) Ninth and tenth leaves. D) Ninteenth and twentieth leaves. E) Mature flowers. F) First and second fully expanded leaves. G ) Ninth and tenth leaves. H) Seventeenth and eighteenth leaves. The chromatograms of whole f lower surface extracts showed a complex mixture of compounds including parthenolide (Figure 4 .4-E) . Parthenol ide was the major peak in H P L C chromatograms from flower receptacle t r ichomes but not in the tr ichomes from disk f lower petals (Figure 4.5). Figure 4.5. H P L C chromatograms of extracts from isolated t r ichomes of composi te disk flowers (A) and the receptacle (B). 4.3.2 Feverfew glandular trichome development, density, and content Two types of t r ichomes were present on feverfew leaves. Glandular, approximately 20 x 12 urn, and ribbon-like approximately 150 to 300 urn long (Figure 4.6). The glandular tr ichomes were multi-cellular capi tate-sessi le g lands and appeared 95 to have two extracellular compartments with the cuticle forming the outer surface. The developing bi- lobed head can be seen at three different s tages in Figure 4.6.a and different developmental phases together with non-glandular ribbon-like tr ichomes are shown in Figure 4.6.b. Leaf veins had mostly non-glandular hairs with a low density of glandular t r ichomes. Figures 4.6 c and 4.6.d show leaves at lower magnification with both types of tr ichomes present and high densit ies of non-glandular t r ichomes on the midrib and veins. Methanol ic extracts of the non-glandular hairs did not contain parthenolide or other compounds detected as H P L C peaks normally found in extracts of the feverfew leaf surface and glandular tr ichomes. Glandular t r ichomes were also present on disk floret petals around the opening of the floret (Figure 4.7). The floret tr ichomes (50u.m x 30 urn) were always glandular, were much larger than leaf tr ichomes (20 pm X 12 urn) and had a shape similar to leaf t r ichomes but with irregular bulges. Both glandular t r ichomes and non-glandular tr ichomes were abundant on the green inf lorescence receptacle. They were a similar s ize and shape as the leaf tr ichomes and were the primary location of parthenolide in the f lowers. 98 S E M was the best method to observe the feverfew tr ichomes although they could be seen with a light microscope or a dissect ing microscope. However, using the latter two methods, glandular tr ichomes on fresh leaf epidermis are not a lways easi ly visible due to their translucent appearance (Figure 4.8a). In addit ion, due to the uneven surface of the fresh epidermis, not all t r ichomes were visible in one plane of focus. W h e n epidermis sl ide preparations were al lowed to dry in the oven or at room temperature, the oil-filled tr ichomes remained intact and could be easi ly dist inguished in one plain of focus (Figure 4.8b). This method for counting tr ichomes was fast and inexpensive when compared with S E M and did not appear to alter tr ichome s ize or density. Leaf glandular tr ichome density was measured using a light microscope and the dried epidermal peel method. There were consistent patterns in the relationships among tr ichome density, parthenolide concentrat ion, and leaf age in vegetative and reproductive shoots (Table 4.2). In both vegetative and reproductive plants, leaf parthenolide concentrat ion increased with increasing tr ichome density. Leaf tr ichome density of vegetative shoots was high on young leaves and dec reased with leaf age. Converse ly the tr ichome density in reproductive shoots was lowest in young leaves and increased with leaf age. Plants in the reproductive stage had lower leaf tr ichome densit ies than plants in the vegetative stage. 100 Tab le 4.2. Trichome density and parthenolide concentration of leaves measured from the apex to the base of the stem in vegetative and reproductive plants. Vegetative Reproductive Leaf position* Trichome density** Parthenolide*** Trichome density** Parthenolide*** 1 21.5 ± 1.6 40.6 ± 1.8 0.113 ± 0 . 0 0 7 0.598 ± 0.086 2 13.3 ± 0 . 5 7 35.6 ± 1.6 0.327 ± 0.006 0.849 + 0.072 3 11.1 ± 0 . 6 1 35.9 ± 0.6 0.653 ± 0 . 0 1 0 . 9 1 0 ± 0 . 1 2 4 12.5 ± 0 . 7 7 36.2 ± 2 . 1 2.39 ± 0 . 0 7 2.69 ± 0.43 5 13.1 ± 0 . 5 2 39.3 ± 1.3 1.94 ± 0 . 0 8 3.12 ± 0 . 3 8 6 10.1 ± 0 . 5 2 26.1 ± 1 . 7 1.66 ± 0 . 0 5 3.31 ± 0 . 4 7 7 6.48 + 0.21 21.7 ± 1.2 2.51 ± 0 . 0 8 6.59 ± 0 . 2 6 8 5.56 ± 0 . 1 9 17.2 ± 0 . 5 8 2.87 ± 0 . 0 6 5.12 ± 0 . 3 4 9 3.89 ± 0 . 2 3 21.5 ± 2 . 8 3 . 1 0 ± 0 . 1 4.57 ± 0 . 2 1 10 6.19 ± 0 . 3 9 27.5 ± 0 . 9 1 1.72 ± 0 . 0 9 3.18 ± 0 . 6 8 * Leaf position 1 was the youngest leaf (top of the shoot) and position 10 was the oldest. ** number of trichomes / mm 2 ± standard error of the mean *** mg parthenolide / g dry leaf ± standard error of the mean 101 Dipping feverfew leaves in water resulted in a 15 % decrease in parthenolide concentration. Prior to surface extraction with water, leaf parthenolide concentration was 3.68 ± 0.17 mg parthenolide / g dry leaf and after extraction was 3.12 ± 0.21 mg parthenolide / g dry leaf. Examination of the trichomes after the water extraction showed that they were not damaged. In contrast, leaf surface extraction with dichloromethane (DCM) ruptured glandular trichomes (Figure 4.10). • -fit . ~ • •s." t-^afc " " <* »Jfc Figure 4.9. Trichomes before (a) and after (b) treatment with dichloromethane. Trichomes are circled with a dotted line. 102 4.4 D I S C U S S I O N Photoper iod controls f lowering through photoreceptors which induce the developmental shift from vegetative to reproductive growth. Al tered secondary metabol ism may result from the cascade of events that occur during the developmental shift. This phenomenon is exhibited in Artemisia annua, Origanum majorana and s o m e Mentha sp. which require a long photoperiod to flower. In Artemisia annua the concentration of the S T L artemisinin was greatest in the vegetative stage and decreased during flowering (Liersch et al., 1986). Lower terpene concentrat ions in O. majorana occured under long photoperiods, a result of the developmental transition from vegetative to reproductive growth (Circel la et al., 1995). L ikewise, essent ial oil content in Mentha sp. (M. arvensis, M. citrata, and M. cardiaca) was highest under short photoperiods and during a prolonged vegetative stage. However, in the Mentha sp. both the chemica l composit ion and concentrat ions in the essent ia l oil changed with photoperiod (Farooqi et al., 1999). Essent ia l oils contain many compounds and their individual concentrat ions can vary without significantly affecting total oil y ields. In another study of feverfew, the total concentration of essent ia l oil in the leaves decreased as the plant matured. In this same plant however, the camphor concentration increased while the crysanthenyl acetate concentrat ion decreased with maturation (Hendriks et al., 1996). The variable mixtures of chemica ls in essent ial oils go unnoticed when only the total essent ial oil content in a plant is measured . Investigation of the individual compounds in essent ial oils al lows a better understanding of how individual chemica ls in the oils are accumulated, stored, and metabol ized in response to environmental and developmental cues . In many feverfew varieties parthenolide is the most abundant S T L and is accumula ted, stored, and probably synthesized in glandular t r ichomes. This 103 accumulat ion is altered in response to long photoperiods and developmental changes. S ince parthenolide is probably an early intermediate in the biosynthesis of the more complex S T L s present in feverfew, the study of parthenolide metabol ism and regulation may provide insights into the role of S T L s and help to explain their dynamics. The high accumulat ions of parthenolide early in development with subsequent dec reases , and the increased abundance of other S T L s later in development suggests this is a point of regulation in S T L metabol ism. Exper iments presented in this chapter show that feverfew grown under short photoperiods remained in a vegetative stage as leaf parthenolide concentrat ion increased by a factor of ten, compris ing 4 % of the leaves by dry weight. Under long photoperiods the plants f lowered and leaf parthenolide content dec reased . The transition from vegetative to reproductive growth was also followed by altered chemica l composit ion in the leaves. The change may have been due to parthenolide metabol ism into other S T L s . It is not c lear whether photoperiod directly affected S T L metabol ism in feverfew or if the altered chemistry was a biproduct of the developmental changes . It is possible that photoperiod has a direct effect on S T L metabol ism. A photoperiodic effect on monoterpene enzyme activity in peppermint (Mentha x piperita) was reported by Voir in et al. (1990). Long photoperiods were required for the hydroxylase enzyme to convert menthone to menthol (Voirin et al. 1990). The concentrat ion of specif ic chemica ls in an essent ia l oil may fluctuate independently if they are regulated by seperate factors. This, occurs in essent ial oils of both feverfew and Mentha sp. due to photoperiodic, abiotic and developmental factors (Ghosh and Chatterjee, 1993; Je l iazkova et al., 1999; Maffei et al., 1986; Hendriks et al., 1996). Many environmental factors may change throughout plant development. Secondary metabol ism will a lmost certainly be inf luenced by these changes. 104 Exper iments in this thesis showed that nitrogen, irrigation, and photoperiod influenced feverfew leaf chemistry. Regard less of these treatments there were consistent parthenolide concentrat ion gradients within the plant, such as parthenolide concentration that increased with leaf age. This suggests that there is a regulated pattern of parthenolide biosynthesis which can be modified by environmental factors and development. In vegetative plants the highest concentrat ions were in leaves at the middle and toward the bottom of plants while in the reproductive stage the highest parthenolide concentrat ion was in leaves at the base of plants which also contained significant concentrat ions of other compounds. The lower leaves of f lowering shoots originated from vegetative shoots which may explain the higher parthenolide concentrat ions. Simi lar developmental patterns were found in Origanum majorana where terpene concentrat ions were higher in older leaves (Circel la et al., 1995). In Mentha spec ies (M. spicata, M. longifolia, M. rubra), menthol and menthone were higher in the third to fifth leaf pairs compared to younger leaf pairs (Fahlen et al., 1997). E n z y m e activit ies, protein levels, and the rate of monoterpene biosynthesis in peppermint (Mentha x piperita) glandular tr ichomes were determined early in leaves 12 to 20 days old which indicated monoterpene biosynthesis was developmental ly regulated at the level of gene expression during early leaf development (McConkey et al., 2000). Terpene biosynthetic enzyme concentrat ions and activities in feverfew may be determined early and then modified with development from vegetative to reproductive growth. The mechan ism or level of control of environmental modulators, whether the target is genetic or enzymat ic, is largely unknown for sesqui terpene biosynthesis. G landu lar t r ichomes are very smal l and the use of a microscope is necessary for measurements . Scann ing electron microscopy ( S E M ) was the best method for 105 visual izat ion but is t ime consuming and expens ive (Duke and P a u l , 1993). Another method involves acid staining to enhance the visibility of t r ichomes under the light microscope (Kelsey and Shaf izadeh, 1980). I developed a quick method which does not require staining or S E M . Us ing epidermal peals mounted on a sl ide with water and then dried, glandular tr ichomes remained intact and were highly visible in one plain of v iew under a light microscope. A compar ison of observat ions made using the S E M and dried epidermis preparations showed tr ichome density and s ize were not affected by the latter procedure. Us ing this new technique I was able to measure the s ize and densit ies of t r ichomes. The relationship between leaf parthenolide concentrat ion, leaf age, and development may be explained by tr ichome densit ies (Brun et al., 1991). On a single mature leaf, t r ichomes are a uniform s ize and shape and contain similar amounts of parthenolide. In mature leaves there were high parthenolide concentrat ions when glandular tr ichome densit ies were high. Both parthenolide concentrat ions and tr ichome densit ies on leaves of vegetative plants were significantly higher than on leaves of f lowering plants. Th is pattern was also found in leaves from Mentha verdae which had higher tr ichome densit ies before flowering and the lowest during f lowering. The Mentha tr ichome densit ies were positively correlated with essent ia l oil concentrat ion (Maffei et al., 1986). In the absence of t r ichomes in g landless Artemisia annua mutants, S T L s were not synthesized (Tellez ei* al., 1999). Certainly t r ichome density affects terpenoid concentrat ion in a variety of plants but the question remains whether tr ichome density is a mechan ism of control for leaf terpenoid concentrat ions. Tr ichome patterns are determined in the initial s tages of leaf emergence, therefore the number of tr ichomes per leaf is determined at leaf emergence (Szymansk i et al., 2000). A s the leaf grows and expands both the s ize and the distance between tr ichomes increases. Tr ichome 106 density dec reased during leaf expansion but tr ichome s ize increased, resulting in a maintenance or increase of essent ial oils per unit a rea . A n y environmental effect on tr ichome density probably occurs before or during leaf emergence . Observat ions from tr ichome measurements presented in this chapter indicate that mature leaves did not have new tr ichomes developing. During feverfew's vegetative stage, older leaves had higher parthenolide concentrat ions than predicted from their tr ichome densit ies. This may be explained if t r ichomes had burst on older leaves exuding their contents onto the leaf surface resulting in lower tr ichome densit ies. In this case the tr ichome contents may have remained on the leaf surface. In some plants, g landular t r ichomes develop a weak area or pore at the apex to facilitate their rupture (Afolayan and Meyer , 1995; A s c e n s a o and Pa i s , 1987). This suggests an inherent mechan ism to distribute tr ichome contents on the leaf surface. Although there were no obvious weak areas or pores on feverfew's glandular t r ichomes, there may be a s ize limit that when exceeded results in rupture. If feverfew tr ichomes naturally rupture on older leaves and exude their contents to the leaf surface, water may be sufficient to rinse the exposed parthenolide away. In an experiment where feverfew leaves were rinsed with water, t r ichomes were not ruptured by the water but there was a 1 5 % decrease in leaf parthenolide concentration after rinsing. The ecological s igni f icance for the re lease of tr ichome contents onto the leaf surface may be in deterring feeding insects or killing pathogens. Glandular t r ichomes on disk florets of the inf lorescence were significantly larger than leaf t r ichomes. If t r ichomes were to rupture at a critical s ize , the flower tr ichomes must have different s ize characterist ics. Individual t r ichomes sampled from petals of disk florets contained low parthenolide concentrat ions. Th is seemed to contradict results from whole f lower extracts that had high parthenolide concentrat ions. Most of 107 the parthenolide in f lowers was in the dense arrangements of glandular tr ichomes on the receptacle and only smal l amounts were present in the disk or ray florets. The disk floret petal t r ichomes were a different s ize , had an irregular shape with a lumpy appearance and their chemistry was different in compar ison to leaf glandular tr ichomes. T h e s e three characterist ics suggest that the disk floret petal tr ichome and leaf t r ichomes are different types and develop differently. The density and chemistry of glandular t r ichomes is variable and the environment affects this variability. It is bel ieved that plants produce glandular t r ichomes and their chemical arsenals as defensive mechan isms. However, environmental factors affecting tr ichomes may indicate that t r ichomes have functions other than defense. For example, Lycopersicon hirsutum had higher tr ichome densit ies when grown under short photoperiods (Weston et al., 1989). In this case , there may be a physiological function for tr ichomes under short days such as cold tolerance in preparation for autumn and winter weather (Behnke, 1984). Only leaves, which develop during or after the environmental effect would exhibit the change in tr ichome density. With the isolation of new g landless mutants it may be possib le to study the role of t r ichomes in relation to the plant's environment. 108 4.5 R E F E R E N C E S Afo layan, A . J . , and Meyer , J . J . M . (1995) Morphology and ultrastructure of secreting and nonsecret ing foliar t r ichomes of Helichrysum aureonitens (Asteraceae). International Journal of Plant Sc iences 156:481-487 A s c e n s a o , L , and P a i s , M . S . S . (1987) Glandular t r ichomes of Artemisia campestris (ssp. maritima): ontogeny and histochemistry of the secretory product. Botanical Gazet te 148:221-227 Behnke , D.H. (1984) Plant tr ichomes - structure and ultrastructure: Genera l terminology, taxonomic appl icat ions, and aspects of t r ichome-bacter ia interaction in the leaf tips of dioscorea. In: Biology and chemistry of plant tr ichomes. New York, P lenum Press . Eds . E. Rodr iguez, P. L. Hea ley and I. Mehta . Brun, N., Co l son , M., Perr in, A . , and Voir in, B. (1991) Chemica l and morphological-studies of the effects of aging on monoterpene composi t ion in Mentha x piperita leaves. Canad ian Journal of Botany 69:2271-2278 Ci rce l la , G . , F ranz , C , Novak, J . , and R e s c h , H. (1995) Influence of day length and leaf insertion on the composit ion of marjoram essent ia l oil. F lavour and Fragrance Journal 10:371-374 Fah len , A . , Welander , M. , and Wenners ten , R. (1997) Effects of light-temperature regimes on plant growth and essent ial oil yield of se lected aromatic plants. Journal of the Sc ience of Food and Agriculture 73:111-119 Farooqi , A . H . A . , S a m g w a n , N.S. , and Sangwan , R . S . (1999) Effect of different photoperiodic regimes on growth, flowering and essent ia l oil in Mentha spec ies . Plant Growth Regulat ion 29:181-187 G h o s h , M.L., and Chatterjee, S .K . (1993) Physio logical and b iochemical indexing of synthesis of essent ia l oil in Mentha spp. grown in India. Ac ta Horticulturae. 331:351-356 Hendr iks, H., B o s , R., and Woerdenbag , H. J . (1996) The essent ia l oil of Tanacetum parthenium (I.) Schultz-bip. Flavour and Fragrance Journal 11:367-371. Heptinstal l , S . , and A w a n g , D .V .C . (1998) Feverfew: a review of its history, its biological and medicinal properties, and the status of commerc ia l preparations of the herb. In: Phytomedic ines of Europe, chemistry and biological activity, pp. 158-175 Amer ican Chemica l Society, Washington. E d s . Lawson , L. and Bauer R. Je l iazkova , E.A. , Zhel jazkov, V . D . , Craker , L .E . , Yankov , B., and Georg ieva , T. (1999) N P K fertilizer and yields of peppermint, Mentha x piperita. Ac ta Horticulturae 502:231-236 Ke lsey , R . G . , and Sha f i zadeh , F. (1980) Glandular t r ichomes and sesqui terpene lactones of Artemisia nova (Asteraceae). B iochemica l Systemat ics and Ecology 8:371-377 109 Liersch, R., So i cke , H., Stehr, C , and Tullner, H.U. (1986) Formation of artemisinin in Artemisia annua during one vegetation period. Planta Med ica 5:387-390 Maffei, M. , Gal l ino, M. , a n d S a c c o , T. (1986) Glandular t r ichomes and essent ial oils of developing leaves in Mentha viridis lavanduliodora. P lanta Med i ca 3:187-193 M c C o n k e y , M .E . , Ge rshenzon , J . , and Cro teau, R .B . (2000) Developmenta l regulation of monoterpene biosynthesis in the glandular t r ichomes of peppermint. Plant Phys io logy 122:215-223 Sal isbury, F .B. (1982) Photoper iodism. Horticultural Rev iews 4:66-105 Spr ing, O. , Rodon , U., and Mac ias , F.A. (1992) Sesqu i te rpenes from noncapitate glandular t r ichomes of Helianthus annuus. Phytochemistry 31:1541-1544 Szymansk i , D.B., L loyd, A . M . , and Marks, M.D. (2000) Progress in the molecular genet ic analys is of tr ichome initiation and morphogenes is in Arabidopsis. Trends in Plant Sc iences 5:214-219 Tel lez, M.R., C a n e l , C , R imando, A . M . , and Duke, S . O . (1999) Differential accumulat ion of isoprenoids in g landed and g land less Artemisia annua L. Phytochemistry 52:1035-1040 Voir in, B., Brun, N., and Bayet, C . (1990) Effects of daylength on the monoterpene composi t ion of leaves of Mentha piperita. Phytochemistry 29:749-755 Wes ton , P.A. , Johnson , D.A., Burton, H.T., and Snyder , J . C . (1989) Tr ichome secretion composi t ion, tr ichome densit ies, and spider mite resistance of ten access ions of Lycopersicon hirsutum. Journal of the Amer ican Society for Horticultural Sc ience 114:492-498 110 Chapter 5 Biosvnthetic studies using 1 4 C and 1 3 C incorporation into parthenolide 5.1. Introduction Terpenoids are present in all living organisms. There are over 25,000 terpenoids reported in the plant k ingdom (Eisenreich et al., 2001). T h e s e compounds are diverse in function. They include carotenoids which serve as light-harvesting and light protecting pigments, sterols which modulate membrane properties, the phytol side chain of the chlorophylls and a profusion of insect attractants and repellents. Important pharmaceut icals and nutraceuticals are synthesized from terpenoids including taxol the cytostatic diterpene from Taxus brevifolia, lycopene and lutein which are oncoprotective agents, and the antimalarial sesqui terpene artemisinin from Artemisia annua. In the 1950s investigations into isoprenoid and sterol biosynthesis resulted in the elucidation of the mevalonic acid (MVA) pathway. Subsequent ly the biogenetic isoprene rule was formulated which states: 1) Terpeno ids are composed of repeating five carbon isoprene units united in a head-to-tail fashion, 2) Isoprene units are derived from acetate in the activated forms of ace ty l -CoA and acetoacety l -CoA, and 3) Mevalon ic acid is an essent ial intermediate. It was thirty years before the discovery of an alternative pathway to isoprenoid biosynthesis in the early 1990s. The methylerythritol phosphate ( M E P ) pathway util izes g lycera ldehyde-3-phosphate and pyruvate to form the basic isoprene unit. The M V A and M E P pathways converge at the formation o f t h e isoprene unit, isopentenyl pyrophosphate (IPP) (figure 5.1). The M E P pathway is a lso referred to as the alternative pathway, the Rohmer pathway, or acronyms derived from the intermediate compound deoxyxy lu lose-5-phosphate such as D O X - P , D O X , and D X P . I l l M E P p a t h w a y o C H 3 OH DXP • NADPH/H + 1^ *- NADP+ H 3 C OH HO OP OH MEP -ATP -ADP H 3 C OPP IPP CH. ,OPP DMAPP CH, M e v a l o n a t e p a t h w a y SCoA SCoA HOOC 2 NADPH/2H+ SCoA 2 NADP CoASH HMGR 3 OH HOOC. >^ / OH 3R-mevalonic acid (MVA) 2 ATP 2 ADP < C . H 3 OH HOOC 5-PPMVA H,C OPP OPP IPP F i g u r e 5.1. The methylerythritol phosphate pathway and the mevalonate pathway to terpenoid biosynthesis. H M G S = hydroxymethylglutaryl synthase, H M G R = hydroxymethylglutaryl reductase, 5 - P P M V A = 5-pyrophosphomevalonate, IPP = isopentenylpyrophosphate, D M A P P = dimethylal ly lpyrophosphate, G - 3 - P = g lycera ldehydes-3-phosphate, D X P = 1-deoxyxylulose-5-P, M E P = 2-C-methy l -D-erythritol-4-P. 112 In plants, the M V A pathway occurs in the cytoplasm and is responsible for synthesis of triterpenes including sterols, and the prenyl s ide chain of ubiquinones (Eisenreich et al., 2001). The M E P pathway is local ized in plastids (Araki et al., 2000). The plastidic terpenoids, which include isoprene, monoterpenes, di terpenes, carotenoids, the prenyl s ide chains of p lastoquinones and chlorophylls are all synthesized through the M E P pathway (Rohmer, 1999). Isopentenyl pyrophosphate ( IPP), farnesyl pyrophosphate ( F P P ) , and geranyl pyrophosphate ( G P P ) are present, and may be synthes ized, in both compartments (Rohmer, 1999). Both the compartmentat ion and utilization of different substrates by the two pathways explains the low incorporation of 1 4 C labeled acetate or mevalonate via the mevalonate pathway into plastidic terpenoids such as phytol and plastoquinone. The low levels of incorporation however, indicate participation of M V A derived terpenoids in the plastids and that transport or diffusion must occur between the two compartments. The M V A pathway is widespread but not ubiquitous. It is found in the A rchaea and certain bacter ia, yeasts , fungi, a lgae, higher plants, s o m e protozoa and animals. Despi te the elusive history of it's d iscovery and descript ion, the M E P pathway is widely distributed in nature. Many bacteria (including pathogens) (Giner and Jaun , 1998), green a lgae, higher plants, and protozoa (including the malaria parasite Plasmodium falciparum) use the M E P pathway. Bacter ia appear to use either the M V A or the M E P pathways but not both. M o s s e s and liverworts, marine diatoms and higher plants use both pathways. S o m e parasit ic micro-organisms appear to lack isoprenoid biosynthetic capabil i t ies and may use terpenoids from their host (Eisenreich et al., 2001). Over 5000 sesqui terpene lactones (STLs) have been identified and a large number of these are biologically active against bacter ia, fungi, and/or v i ruses. Many of the S T L s interact with protein or D N A to elicit physiological responses . In higher plants 113 it is evident that monoterpenes and diterpenes are synthesized primarily by the M E P pathway but few studies have been conducted on the role of the M E P pathway in sesqui terpene biosynthesis. Recent reports show both the M V A and M E P pathways contribute to the biosynthesis of sesqui terpenes (Adam and Zapp , 1998). The purpose of exper iments descr ibed in this chapter was to elucidate the biosynthetic origins of the sesqui terpene lactone parthenolide. This was accompl ished by feeding feverfew shoots 1 4 C and 1 3 C enr iched substrates (glucose, acetate, pyruvate and mevalonate) and quantifying 1 4 C incorporation and the posit ion of 1 3 C enr ichment in parthenolide. 5.2. Materials and Methods The feeding experiments were carried out in a modified fume-hood with full spectrum vita lights programmed for 16 hours light / 8 hours dark. The average temperature was 24 °C with the lights on and 20 °C with the lights off. Solut ions were made with distilled water filtered through a milli-Q (Millipore) filtration sys tem. Al l chemica ls were purchase from S igma unless otherwise stated. In both the 1 3 C and 1 4 C exper iments, labeled feeding solutions were administered until the entire volume was taken up by the shoots after which water or g lucose solution was suppl ied to allow the plants a longer period for label incorporation. The 1 4 C - g l u c o s e experiments were conducted prior to the 1 3 C experiments. The 1 4 C-py ruva te , 1 4 C - a c e t a t e , and 1 4 C-meva lono lac tone exper iments were conducted simultaneously at the same location and under the s a m e condit ions. Al l shoots were harvested fresh and recut underwater before treatments were appl ied. 114 5.2.1. 1 4 C feeding experiments UL-14C-glucose feeding 11.31 (j,Ci UL- 1 4 C- [D] -g lucose (purchased from Cambr idge Isotope Laborator ies, CIL) was added to 100 ml of cold 1% D-glucose for a final activity of 0.1131 u.Ci / ml. Twelve young shoots with 4 to 7 leaves each were placed in four vials, each containing 25 ml U L - 1 4 C - D - g l u c o s e solution. The shoots were grown in an environmentally controlled fume hood under 16 hours light (24 °C) / 8 hours dark (20 °C) for seven days. The U L - 1 4 C - D - g l u c o s e solution was taken up in three days and replenished with cold 1% D-glucose for an additional four days. The control treatment was grown under the s a m e condit ions with the exception of cold 1% D-glucose instead of U L - 1 4 C - D - g l u c o s e as the substrate. 1-14C-glucose feeding O n e hundred and sixty ml of cold 1% D-glucose was mixed with 40 u.Ci of 1 - 1 4 C -D-glucose (purchased from CIL) for a final activity of 0.25 u,Ci/ml. Twelve young shoots with 5 to 8 leaves each were placed in four vials containing the radio-labeled g lucose solution. The shoots were incubated in a fume hood under 16 hours of light (24 °C) / 8 hours dark (20 °C) for nine days. The 1 - 1 4 C-D-g lucose solution w a s taken up in six days and then replenished with cold 1% D-glucose for three days. A control treatment was conducted at the same time using cold 1% D-glucose as a substrate. Al l other condit ions were as descr ibed above. 1-14C-pyruvate feeding Cold 1 m M sodium pyruvate solution (40 ml) was prepared and mixed with 9.33 u.Ci sod ium-1- 1 4 C-pyruva te for an activity of 0.23 u.Ci/ml. Two vials were filled with 20 ml of the feeding solution and three young shoots were placed in each vial. The shoots 115 were incubated in a fume hood under 16 hours of light (24 °C) / 8 hours dark (20 °C) for seven days. The labeled solution was used up within four days and the containers were refilled with distilled water for the remaining three days. The s a m e number of shoots and condit ions were used for the control treatment with the except ion of 1 2 C-pyruva te instead of 1 4 C-py ruva te . 2-14C-acetate feeding A 1 m M Na + -ace ta te solution (40 ml) was prepared and mixed with 10.2 u.Ci N a + 2 - 1 4 C - a c e t a t e for a specif ic activity of 0.25 u.Ci/ml. Two vials were filled with 20ml of the feeding solution. E a c h vial contained three young shoots. The shoots were incubated in a fume hood under 16 hours of light (24 °C) / 8 hours dark (20 °C) for seven days. The labeled solution was taken up within four days and the containers were refilled with distilled water for the remaining four days. The s a m e number of shoots and condit ions were used for the control treatment except 1 2 C - a c e t a t e was used instead of 1 4 C - a c e t a t e . 2-14C-mevalonolactone feeding A 1mM mevalonolactone solution (40 ml) was prepared and mixed with 10.5 uCi 2 - 1 4 C-meva lono lac tone for a specif ic activity of 0.26 u.Ci/ml. Two vials were filled with 20 ml o f the feeding solution. E a c h vial containing three young shoots was incubated in a fume hood under 16 hours of light (24 °C) / 8 hours dark (20 °C) for seven days. The labeled solution was taken up within four days and the containers refilled with distilled water for the remaining four days. The same number of shoots and condit ions were used for the control treatment except 1 2 C-meva lono lac tone was used instead of 1 4 C -mevalonolactone. 116 5.2.2. 1 3 C f e e d i n g e x p e r i m e n t s 1 3 C - E n r i c h e d substrates (pyruvate, acetate, mevalonolactone, and D-glucose) were fed to feverfew cuttings (all 1 3 C experiments were conducted at the s a m e time under the same condit ions) grown in a modified fume hood equipped with full spectrum vita-lights. The cuttings were inserted in vials filled with the enr iched solution for a total of seven days under 16 hours of light at 24 °C and 8 hours dark at 20 °C. The control treatments of unlabeled substrate was applied for three days fol lowed by distilled water for four days. The control treatment was applied to five vials with three shoots in each . V ia ls from all treatments were arranged randomly together under one set of vita-lights. 1- 13C-glucose feeding O n e hundred mg of 9 9 % enriched 1 - 1 3 C-D-g lucose was added to 300 ml of a 1% D-glucose solution. The solution was divided into 10 vials, each containing three young shoots with 4 to 7 leaves. The solution was taken up by the plants in three days and replenished with 1 % g lucose for another four days. 2- 13C-pyruvate, 2-13C-acetate and 2-13C-mevalonolactone feeding Three hundred ml each of 3 m M 2- 1 3 C-pyruva te , 3 m M 2 - 1 3 C-ace ta te and 2.5 m M 2- 1 3 C-meva lono lac tone solutions (99% enriched) were each divided into ten vials (30 vials total). Three shoots with 4-7 leaves each were put into each of the vials. The 1 3 C -enriched solutions were taken up in three days. V ia ls were replenished with water which was taken up for an additional four days. 117 5.2.3. Extraction methods, parthenolide isolation, and NMR analysis After feeding radio-labeled substrates to feverfew shoots, the surface of the fresh shoots were extracted twice using a rapid dichloromethane rinse. The extract was vacuum filtered with Wha tman No. 1 filter paper and dried under vacuum at 30 °C using a rotary evaporator. The dry extract was weighed and resuspended in 1 ml methanol. A n aliquot of 100 pi was put into a 7 ml scintillation vial with 5 ml of scintillation fluid to measure 1 4 C activity using a scintillation counter. Activity in the extract was calculated from this sample . 14C-Parthenolide isolation Aluminum backed si l ica T L C plates with UV254 f luorescence indicator were used for the initial U L - 1 4 C - g l u c o s e experiment. For the rest of the 1 4 C exper iments, g lass backed 0.25 mm preparative si l ica T L C plates with UV254 f luorescence indicator were used . The extracts were loaded in bands at the origin 1cm from the base and placed in the 8 0 % dichloromethane : 2 0 % acetone solvent system in a g lass T L C chamber. This sys tem resolved many bands including parthenolide (Rf=0.81). B a n d s were visual ized under UV254 and with vanill in-sulfuric acid spray reagent. E a c h band on the T L C plate was scraped off and transferred to a 7 ml plastic scintillation vial with 5ml scintillation fluid and activity recorded with the scintillation counter. Purity of the resolved parthenolide band was confirmed with H P L C using methods descr ibed in Chapter 2. 13C-Parthenolide isolation Extracts were loaded in bands on 0.25mm si l ica T L C plates with UV254 indicator. The plates were developed with 8 0 % D C M : 2 0 % acetone. B a n d s were visual ized under UV254 and with vanill in-sulfuric acid spray reagent. Parthenol ide (Rf = 0.81) quenched the UV254 indicator but was not f luorescent. Developing the plate with vanillin sulfuric acid and heating resulted in a dark blue color for the parthenolide band. The 118 parthenolide band was scraped and the parthenolide eluted from si l ica with methanol. The extracted band w a s vacuum filtered with Wha tman No . 1 filter paper and evaporated to dryness using a rotovap under vacuum at 30 °C. T h e dried extract was resuspended, transferred in 4 ml methanol to a vial , and concentrated under nitrogen at 40 °C. Parthenol ide was purified further using preparative H P L C with a Waters N o v a P a k 2.5 X 10 cm (5u.m) C i 8 radial compress ion co lumn. The mobile phase was acetonitri le:water (70:30) at 25 ml/min with detection at 210 nm. Injection volumes were 200 uJ and elution of the parthenolide peak was at 6 minutes. The fraction was dried under vacuum leaving a white residue (parthenolide). Purity of the isolated parthenolide was checked with analytical H P L C . NMR analysis The 1 3 C - N M R spectra of parthenolide were recorded in deuterated chloroform (CDCI3) on a Bruker A M X - 5 0 0 spectrometer. The 1 H - N M R spectrum of parthenolide was recorded in CDCI3 on a Bruker A C - 2 0 0 E spectrometer. The residual chloroform CHCI3 (5 7.24 ppm) signal was used as reference. Enr iched and non-enriched parthenolide samp les were run sequential ly under the s a m e condit ions. Enr ichments were obtained through subtraction of the peak integrations and peak intensities of non-enr iched samp les from enriched samples . 119 5.3. R e s u l t s 5.3.1. 1 4 C labe l i ng o f pa r theno l i de The 1 4 C feeding experiments were carried out to determine feeding parameters and incorporation rates for use with the subsequent 1 3 C exper iments. A range of shoot s izes were tested with water or g lucose. Shoots performed best when they were very young with 4 to 8 leaves and a g lucose solution was better than water for shoot survival. 1 4 C was incorporated into parthenolide after feeding the five labeled substrates to feverfew shoots. The 1 4 C incorporation rates (Table 5.1), reported as relative specif ic activity ( R S A ) , were 5.3 t imes greater when U L - 1 4 C - g l u c o s e was fed to shoots compared with 1 - 1 4 C-g lucose fed shoots. The highest R S A s after feeding the single labeled substrates were in shoots fed with either 1- 1 4 C-pyruvate or 1 - 1 4 C-g lucose . The substrates predicted to move solely through the mevalonic acid pathway, 2 - 1 4 C -mevalonolactone and 2 - 1 4 C-ace ta te , had the lowest R S A s . T a b l e 5.1. Incorporation of 1 4 C labeled substrates into parthenolide. Radio labeled substrate Activity Activity in taken up by surface the shoots extract (nCi) (nCi) Spec i f ic activity InCii 2- 1 4 C-meva lono lac tone U L - 1 4 C - g l u c o s e 1 - 1 4 C-g lucose 1- 1 4 C-py ruva te 2- 1 4 C - a c e t a t e 3430 37560 9325 10153 9455 188 4860 1122 726 689 0.624 1.28 0.387 0.189 0.276 0.0182 0.00341 0.00415 0.00186 0.00292 120 5.3.2. 1 3 C e n r i c h e d pa r theno l i de The 1 H - N M R spectrum of parthenolide (Figure 5.2 a) shows a characterist ic chemical shift of two doublets around 5.7 and 6.3 ppm representing the two protons of the exocycl ic methylene on the lactone ring (protons attached to carbon 13 of parthenolide in Figure 5.3). The 400 M H z 1H-NMR (Figure 5.2 b) shows one of these doublets with the corresponding satellite peaks. The satellite peaks are a result of 1 H -1 3 C spin-spin coupl ing and can be used to quantify the 1 3 C incorporation at that position. In Figure 5.2 b the large doublet (doublet 1) is the 1 H attached to 13- 1 2 C-par thenol ide and the smal l doublet (Figure 5.2 b, doublet 2) is the 1 H attached to 1 3 - 1 3 C -parthenolide. The ratio of integrals for the parent and satellite peaks gave the 1 3 C abundance for carbon position 13 in parthenolide. After feeding the substrates, 1 3 C abundance in carbon 13 of parthenolide calculated from 1H-NMR was 4 % from 1 - 1 3 C -glucose, 2 % from 2 - 1 3 C-ace ta te , 1.3% from 2- 1 3 C-pyruva te , and 1.2% from 2 - 1 3 C -mevalonolactone fed shoots. 121 F i g u r e 5.2. 1 H - N M R spectra of parthenolide. A) Full spectrum at 90 m H z and an expanded region B) at 5.4 to 5.7 ppm of a 400 mhz 1 H - N M R spectrum showing the parent (1) and satellite peaks (2) of the attached proton of carbon 13. F i gu re 5.3. Carbon numbering of parthenolide and the predicted conformation of the three isoprene units. 122 m R S f? iS * :; s s a) 13 11 10 6 | 5 7 9 8 2 12 15/14 cn m n OJ mm ni ni R in iQ ru as R IO ID ID r- —< IO O ^ M B N s ID to ^ n n ni « -b) C-9 (4') C -3 (4) F i g u r e 5.4. 1 3 C N M R of parthenolide isolated from feverfew shoots a) fed with 2 - 1 2 C -mevalonolactone and b) fed with 2 - 1 3 C-meva lono lac tone . 1 3 C enrichment in carbon 9 and carbon 3 is greater in spectrum b than a. 123 The natural abundance of 1 3 C is 1.1%. Patterns of 1 3 C enrichment in excess of the natural abundance were found in parthenolide isolated from shoots from each of the four 1 3 C feeding experiments (Table 5.2). 1 3 C Abundance calculated from 1 H - N M R (Figure 5.2) was consistent with the 1 3 C incorporations calculated from 1 3 C - N M R (Table 5.2 and Figure 5.5). Carbon 13 (see Fig. 5.3 for numbering scheme) of parthenolide was enr iched after feeding the shoots 1 - 1 3 C-g lucose but not significantly enr iched after feeding 2 - 1 3 C-meva lono lac tone or 2 - 1 3 C-pyruva te . Parthenol ide isolated from shoots after feeding 2 - 1 3 C-meva lono lac tone was enr iched at Carbon 3 and Carbon 9, consistent with C-4 and C-4 ' of the isoprene subunits (Figure 5.3) whereas C - 4 " of the third isoprene subunit (equivalent to C-12) was not enr iched. Parthenol ide isolated from plants fed with 2 - 1 3 C-pyruva te was enriched in carbons 4, 10, and 11 corresponding to C - 3 , C -3 ' , and C - 3 " of the isoprene subunits. Ca rbons 6, 2, and 8 had intermediate enr ichments which corresponded to C - 1 , C - 1 ' , and C - 1 " of the isoprene subunits. Many carbons in parthenolide were enriched after the 1 - 1 3 C-g lucose feeding experiment with the except ion of carbons 4, 10, 11, and 12. Three of these non-enr iched carbons (4, 10, and 11) corresponded to C - 3 , C -3 ' , and C - 3 " of the isoprene subunits whereas the fourth non-enriched carbon corresponded to C - 4 " of the isoprene subunit. c TJ 0 a 0 -4—' 03 L _ -*—' co .Q co L _ CD *i 03 CD JD O C CD SZ n CO Q. c CD E .rz o ' i _ c <D O CO CD •4—' CD O CO i O CM CD CO > I 6 CM CD c O -*—» o 03 O c o 03 > CD E 6 00 1— I CM 0 CO o o O CO O t o ra C CD Q . > o '5 W F T c r oc CO w 0 0 C D 0 0 C 0 T - < 0 C D l 0 O r ^ 0 0 l 0 ' < t C T ) 00 CO o C 0 O T - O 0 0 0 0 O 0 0 O O 0 0 0 0 T - O O C M l O T - C M - r - C N C O O } C M C M C ) 0 - « -d ^ d ^ ^ CM CN o I O O O O O T - 0 > T - T - 0 0 0 ( \ | T - r t O N N ( M C O C D N CM CM CM CM T- T- CN 0 0 O CO CM CO O CN O T - T - CN •<fr 00 lO CO CN CN CD CN 00 C O C O r O T - C M c M c o r t 05 CO CO CN CN CN (D 6 s O) O) T - 00 ^" CO CD CN CM CO T t lO CN CO XT LO CM oo ^ in CO lO 00 LO CN CD CD T3 CD CO o T3 _C CD L-03 CO ZJ CD - C o c CD i O CO O) C 5 .8 0 c CD L_ Q . O CO CD <1> co T3 C CO 00 c CD CD CD .O C D) O c — TJ C CO 0 3 o _0 O E 0 o <: CD 0 J3 C 0 E ^ c £ "o o Q- E ^ .E 0 O t o •— B * 125 Pathways Actua l Enr ichment Predic ted Meva lona te Pred ic ted M E P F i g u r e 5.5. 1 3 C enrichment patterns in parthenolide after feeding enr iched substrates. Observed enr ichments from table 5.2 are in the first column and predicted enrichment patterns are in the next two co lumns. Fil led circles are carbons with strong enrichment and open squares indicate carbons with intermediate enrichment. 126 5.4. D i s c u s s i o n The biosynthetic origins of parthenolide were examined using 1 3 C isotopic enrichment. The 1 3 C feeding experiments showed parthenolide biosynthesis in feverfew utilized isoprene units from both the M V A and M E P pathways. Eich inger et al. (1999) and Eisenre ich et al. (1996) conducted feeding studies using 1 3 C - e n r i c h e d substrates including 1 - 1 3 C-g lucose and found different labeling patterns in IPP dependent on the biosynthetic route uti l ized. In these studies the posit ions of enr iched carbons in the isoprene units after feeding 1 - 1 3 C-g lucose were in carbons 1 and 5 when the M E P pathway was utilized and carbons 2, 4 and 5 when the M V A pathway was used . After feeding 1 - 1 3 C-g lucose to feverfew shoots, two isoprene units in parthenolide were enr iched in C - 1 , C - 2 , C-4 and C -5 consistent with contributions from both the M V A and M E P pathways (numbering scheme in Figure 5.3). The third isoprene unit was enriched in C - 1 " , C - 2 " and C - 5 " but not enr iched in C - 4 " . This pattern is difficult to explain but is more consistent with isoprene from the M E P pathway due to the strong enrichments in C - 1 " and C - 5 " . The enrichment in C - 2 " may have resulted from metabol ic turnover of the 1 - 1 3 C-g lucose . Alternatively, the enrichment in C - 4 " may have been masked by the three oxygen atoms in its vicinity which shifts the 1 3 C peak downfield. In combination with the quaternary carbon, this caused reduced signal intensity. A s expected, the C-3 posit ions of the isoprene units were not labeled. The enr ichment in C - 3 , C - 3 ' and C - 3 " is c lose to natural abundance and is in agreement with both biogenetic schemes (Figures 5.6 and 5.7). Acetate is an immediate precursor of the M V A pathway. Three molecules of ace ty l -CoA condense to form IPP via mevalonic ac id . There is no direct pathway for acetate to form G - 3 - P or pyruvate, and therefore it was not expected to move through 127 a) H OH a-D-glucose b) D-fructose-1,6-bisphosphate < r = 0 H O - - H H — - O H H - - O H ^ O P F-1,6-BiP : 0 Glyceraldehyde-3-phosphate - O H - O P H 3 Q O ^ O H DHAP O - ^ O H » — O P V - O H 0 = ^ — O P 1,3-bisphosphoglycerate - O H ) O H 0 = ^ O P 3-13C-Pyruvate Phosphoenolpyruvate 3-phosphoglycerate F igu re 5.6. Glucose catabolism. a) The conversion of D-glucose to D-fructose-1,6-bisphosphate (F-1,6-BiP) and b) the production of glyceraldehyde-3-phosphate (G-3-P) and pyruvate from (F-1,6-BiP). The black circle shows the carbon position as glucose is catabolized. Methvlerythritol phosphate pathway o Pyruvate o O H PO G-3-P • i OH PO^Y^V0 DXP OH C H 3 H O P P O . IPP I C H 9 FPP C 0 2 Mevalonate pathway 128 C o A S H 2 - 1 3 C-ace ty l -SCoA C o A S H O o SCoA 2 - 1 3 C - a c e t y l - S C o A - ^ ^ C o A S H O H O H O O C H M G - S C o A S C o A 3 O H H O O C O P P [ > C 0 2 C H 2 O P P IPP C H 3 | OPP Observed 1 3 C incorporation Predicted MEP pathway Predicted Mevalonate Pathway F i g u r e 5.7. 1 - 1 3 C-D-g lucose feeding experiment. Observed and predicted patterns of 1 3 C enrichment. 129 the M E P pathway. Feed ing feverfew shoots with 2 - 1 3 C-ace ta te resulted in enrichment in posit ions C - 2 , C-4, C -5 , C -2 ' , C-4', C -5 ' , C - 2 " , C-4", and C - 5 " in parthenolide as predicted from the M V A pathway (Figure 5.8). Interestingly the third isoprene equivalent which forms the lactone ring (C-2" , C4" and C-5" ) had lower levels of enrichment than the other two isoprenes. This suggests another source or pool of IPP is utilized for this isoprene unit, possibly derived from the M E P pathway. This is consistent with the 2 - 1 3 C-meva lono lac tone feeding experiment where there was enrichment at C-4 and C-4' in the first two isoprene units but insignificant enrichment in C-4" of the third isoprene compris ing the lactone ring of parthenolide (Figure 5.9). The 2 - 1 3 C-pyruva te feeding experiment also indicated contributions from two separate IPP pools (Figure 5.10). Isotopic enrichment of six carbons (C-1 and C -3 of each isoprene unit) was predicted if pyruvate was catabol ized to acetate and then condensed to form IPP through the M V A pathway whereas only three carbons (C-3 in each isoprene unit) were predicted to be enriched if pyruvate went through the M E P pathway. Enrichment was greater in C - 3 , C - 3 ' , and C - 3 " while lower levels of enr ichment were seen in C - 1 , C - 1 ' and C - 1 " . The higher enr ichments at C -3 relative to C-1 posit ions o f t he isoprene units were consistent with contributions of IPP from the M E P pathway which resulted in higher enr ichments at the C -3 posit ions. Due to the over lap in carbon enrichment from the two pathways, the results from 2- 1 3 C-pyruva te feeding can only confirm a contribution of IPP from the M E P pathway when there is no IPP contribution from the M V A pathway. A problem in identifying the contributions to the M E P pathway from pyruvate or G - 3 - P enriched in carbons 2 or 3 ar ises because carbon 1 of pyruvate is lost as CO2 in the M E P pathway o Predicted Mevalonate Pathway F i g u r e 5.8. 2 - 1 3 C-ace ta te feeding experiment. Observed and predicted patterns of 1 3 C enrichment. C losed circles are carbons with strong enrichment and open boxes are carbons with moderate enrichment 131 Methylerythritol phosphate pathway Mevalonate pathway Predicted Mevalonate Pathw F i g u r e 5.9. 2 - 1 3 C-mevalonolactone feeding experiment. Observed and predicted patterns of 1 3 C enrichment. C losed circles are carbons with strong enrichment. Methylerythritol phosphate pathway Mevalonate pathway Predicted MEP pathway Predicted Mevalonate Pathway F i g u r e 5.10. 2 - 1 3 C - p y r u v a t e feeding experiment. Observed and predicted patterns of 1 3 C enrichment. C l o s e d circles are carbons with strong enr ichment and open boxes are carbons with moderate enrichment. 133 with it's convers ion to acetate. Only the enrichment of carbon 1 of G - 3 - P would be speci f ic for the M E P pathway without seeing contributions from the M V A pathway. 1 3 C Enr iched deoxyxylu lose is an alternative substrate to study the M E P pathway but is not commercia l ly avai lable. Another approach is to use double- labeled substrate which gives the addit ional advantage of studying 1 3 C - 1 3 C interactions. The two independent terpenoid biosynthetic pathways are physical ly separated by compartmentat ion, but there is interaction between the two pathways. There is ev idence that IPP, G P P , and F P P are exchanged between the compartments (Eisenreich et al., 2001). The superposit ion of two different labeling patterns in terpenoid compounds and the chimera nature of some terpenoids suggest a transport mechan ism between the chloroplast and cytoplasm. 1 3 C - D e o x y x y l u l o s e phosphate fed to Catharanthus roseus resulted in significant 1 3 C enrichment in phytol, p-carotene and lutein and low levels of enrichment in the phytosterols (Arigoni, 1997). It has been establ ished that plant sterols are primarily synthesized through the M V A pathway (Eisenreich et al., 2001). Mixed biosynthesis was found in cal lus cultures of the hornwort Anthoceros punctatus. 2 H and 1 3 C labeled mevalonate were incorporated preferentially into an F P P derived portion of chlorophyll a , p-carotene and some di terpenes which are all primarily synthesized through the M E P pathway (Itoh, 2000). Exper iments with chamomi le sesqui terpenes showed mixed biosynthesis. Two isoprenes from the M E P pathway and a third from the M V A pathway condensed to form the sesqui terpene chamazu lene (Adam et al., 1999; A d a m and Zapp , 1998). It is not c lear how the plastidic terpenoid transport mechan isms work, but in chamomi le either M V A derived IPP was transported into the chloroplast, or M E P derived G P P was exported to the cytoplasm, to form F P P . Sesqu i te rpenes were thought to be synthesized primarily in the cytoplasm but there are reports of sesqui terpene synthesis 134 via the M E P pathway in plastids. The leaves of l ima bean (Phaseolus lunatus) incorporated 2 H labeled deoxyxylu lose into volatile monoterpenes and sesqui terpenes in chloroplasts after jasmonic acid treatment or spider mite infestation (Rohmer, 1999). Inducible sesqui terpene synthesis may occur in the plastids because of the proximity to the source of carbon fixation. The two precursors, pyruvate and G - 3 - P , of the M E P pathway are derived from the products of photosynthesis and thus may be induced faster than the M V A pathway. S o m e Streptomyces spec ies synthesize sesqui terpenes via M E P (Eisenr ich et al., 1998), and the sesqui terpenoids from mychorrhyzal barley roots are a lso synthesized through the M E P pathway (Walter et al., 2000). It is possible that parthenolide is b iosynthesized in both the plastids and cytoplasm via the two independent pathways. This is the case for A B A biosynthesis which can be direct via M V A in the cytoplasm or through the catabol ism of carotenoids in plastids (Dewick, 1999). S ince parthenolide may be a defense compound it is possib le that synthesis via the M E P pathway occurs under stress condit ions such as herbivory, infection, or physical damage , but under unstressed condit ions synthesis is through the M V A pathway. This could explain the superposit ion of the 1 3 C labeling patterns found in parthenolide. The compartmental separat ion of the two IPP biosynthetic pathways is not absolute. At least one metabolite can be exchanged between the compartments. The extent of the transport depends on the spec ies as well as on the presence and concentrat ion of exogenous precursors. The movements of terpenoids between compartments appear general ly smal l in intact plants under physiological condit ions. The nature of the metabol ic exchange between the compartments and the regulation of the process remain to be establ ished. The sharing of terpenoids between the two pathways is one of the reasons for the failure in history to recognize the M E P route. 135 The low incorporation rates of mevalonate and acetate in many plant terpenoids were not because of restricted uptake or restricted use of the labeled precursors but likely reflected the smal l contributions of the M V A pathway to the biosynthesis of terpenoids predominantly formed via M E P . Feed ing experiments are tedious and may not give the exact picture, but they do shed light on the main terpenoid biosynthetic route and on possib le exchanges between the two routes. The presence of two different biosynthetic routes leading to the same metabolite (IPP), which functions in two separate compartments, may be a mechan ism for plants to regulate isoprene biosynthesis. Exper iments using molecular techniques may uncover the nature of interactions between the two pathways. Contradict ions between some recent isotopic feeding studies of the M E P and M V A pathways may be explained by some inherent problems with feeding studies. Excess i ve feeding and the subsequent accumulat ion of metabol i tes, which are not naturally accumulated by cel ls such as M V A or D X P , do not represent normal physiological condit ions. High concentrat ions of a substrate may unnaturally induce or increase synthesis of a terpenoid and lead to an over-est imation of the pathway's contribution. Feed ing of an earlier metabolite such as g lucose may result in a more accurate estimation of the contribution of each pathway to isoprenoid biosynthesis. However, feeding earlier precursors such as g lucose may lead to confusing results due to prolonged metabol ism. For example, in phototrophic organ isms the recycling of CO2 may result in confusing labeling patterns. Other factors to cons ider with intact organisms are the differences in uptake and metabol ism dynamics for each substrate. T h e s e affect incorporation into terpenoids and may invalidate direct compar isons between incorporation rates of different substrates. In feverfew for example , labeled compounds must move through the plant from stem to tr ichome and into parthenolide. 136 Labeled pyruvate moves to the cytoplasm, is metabol ized to acetate and used in the M V A pathway but must move into the chloroplast to be used by the M E P pathway. 5.5. C O N C L U S I O N S It is c lear that parthenolide biosynthesis in feverfew uses the mevalonate pathway. It appears that the M E P pathway also contributes to parthenolide biosynthesis. The extent of terpenoid movement between the two pathways seemed substantial . Further experiments are required to obtain conclus ive ev idence showing the extent of M E P contribution to S T L s and parthenolide. Th is may be ach ieved with the use of inhibitors specif ic for either pathway such as mevinol in which inhibits the M V A pathway and fosmidomycin which inhibits the M E P pathway (Fel lermeier et al., 1999). Feed ing substrates in the presence of an inhibitor for one pathway would show the potential involvement of the other pathway. In addit ion, labeled methylerythritol phosphate or deoxyxylu lose phosphate would give a better indication of M E P pathway involvement in parthenolide biosynthesis. The increasing antibiotic drug resistance of human pathogens has created an urgent need for novel therapeutic approaches. The M E P pathway is used by many pathogenic microorganisms and by the protozoa P. falciparum but is not present in humans or animals. Therefore, the M E P pathway is an ideal target for the development of novel antibiotics and antimalarial agents (Rohmer, 1998). In plants the inhibition of the M E P pathway may lead to the development of new herbic ides. Understanding the mechan isms and regulation of both the M V A and M E P pathways will likely benefit the biotechnological production of commercial ly or pharmaceut ical ly interesting isoprenoids by engineer ing increased production or production from non-typical sources . 137 5.6. R E F E R E N C E S A d a m , K., Thie l , R., and Zapp , J . (1999) Incorporation of 1-[1- 1 3 C]deoxy-d-xylulose in chamomi le sesqui terpenes. Arch ives of Biochemistry and Biophys ics 369:127-132 A d a m , K., and Zapp , J . (1998) Biosynthesis of the isoprene units of chamomi le sesqui terpenes. Phytochemistry 48:953-959 Arak i , N., Kusumi , K., Masamoto , K., N iwa, Y . , and Iba, K. (2000) Temperature-sensit ive arabidopsis mutant defective in 1-deoxy-D-xylulose 5-phosphate synthase within the plastid non-mevalonate pathway of isoprenoid biosynthesis. Physio logia Plantarum 108:19-24. Ar igoni , D., Sagner , S . , Latzel , C , E isenre ich, W. , Bacher , A . , and Zenk, M.H . (1997) Terpenoid biosynthesis from 1-deoxy-D-xylulose in higher plants by intramolecular skeletal rearrangement. Proceed ings of the National A c a d e m y of Sc i ence 94:10600-5. Dewick, P . M . (1999) The biosynthesis of C5-C25 terpenoid compounds. Natural Product Repor ts 16:97-130 Eichinger, D., Bacher , A . , Zenk, M. H., and E isenre ich , W . (1999) Ana lys is of metabolic pathways via quantitative prediction of isotope labeling patterns: a retrobiosynthetic C N M R study on the monoterpene loganin. Phytochemistry 51:223-236 E isenre ich , W. , Menhard , B., Hy lands, P. J . , Zenk, M. H., and Bacher , A . (1996) Stud ies on the biosynthesis of taxol: the taxane carbon skeleton is not of mevalonoid origin. Proceed ings o f the National A c a d e m y of Sc ience 93:6431-6 E isenre ich , W . , Schwarz , M. , Car tayrade, A . , Ar igoni , D., Zenk , M.H. , and Bacher , A . (1998) The deoxyxylu lose phosphate pathway of terpenoid biosynthesis in plants and microorganisms. Chemist ry and Biology 5 :R221 -R233 E isenre ich , W. , Rohd ich , F., and Bacher , A . (2001) Deoxyxy lu lose phosphate pathway to terpenoids. Trends in Plant Sc ience 6: 78-84 Fel lermeier M. , Kis K., Sagne r S . , Maier U., Bacher A . , and Zenk M. H. (1999) Cel l- free convers ion of 1-deoxy-d-xylulose 5-phosphate and 2-c-methyl-d-erythritol 4-phosphate into beta-carotene in higher plants and its inhibition by fosmidomycin. Tetrahedron Letters 40:2743-2746 Giner, J .L . and Jaun , B. (1998) Biosynthesis of isoprenoids in Escherichia coli: Retention of the methyl H-atoms of 1-deoxy-D-xylulose. Tetrahedron Letters 39:8021-8022. Itoh, D., Karunagoda, R . P . , Fush ie , T., Katoh, K., and Nabeta , K. (2000) Nonequivalent labeling of the phytyl s ide chain of the chlorophyll a in cal lus of the hornwort Anthoceros punctatus. Journal of Natural Products 63:1090-1093 138 Rohmer , M. (1998) Isoprenoid biosynthesis via the mevalonate- independent route, a novel target for antibacterial drugs? Progress in Drug R e s e a r c h 50:135-54. Rohmer , M. (1999) The discovery of the mevalonate- independent pathway for isoprenoid biosynthesis in the bacteria, a lgae, and higher plants. Natural Product Repor ts 16:565-574 Walter, M.H. , Fester , T., and Strack, D. (2000) Arbuscu lar mycorrhizal fungi induce the non-mevalonate methylerythritol phosphate pathway of isoprenoid biosynthesis correlated with accumulat ion of the 'yellow pigment' and other apocarotenoids. Plant Journal 21:571-8. CHAPTER 6 General discussion 139 6.1. R E S E A R C H O V E R V I E W Phytochemica ls are synthesized from water, nutrients and CO2 using energy from light to drive the reactions. Variat ion in any of these components affects plant biochemistry and physiology including secondary metabol ism. Fluctuation of secondary compounds in response to the supply of resources for growth suggests plasticity in secondary metabol ism. The role of many secondary compounds is for defence. Chemica l concentrat ions and the level of chemical or physiological protection from the environment may thus depend on the availability of resources. For humans, secondary compounds are the primary source of biological activity of most medicinal plants as well as in f lavours and aromas of culinary plants. If we can understand how different environments affect the regulation of secondary metabol ism, we can manipulate medicinal and culinary plants toward higher production of commercia l ly valuable compounds while maintaining high crop yields. W e could improve agriculture and resource management if we had the ability to predict how secondary metabol ism changes with the availability of resources. This thesis examined the relationship between growing condit ions and growth and development, physiology and biochemistry of feverfew shoots. In Chapter 2, feverfew plants were grown with different nitrogen and water application rates in the field and greenhouse. Ana lys is of shoot yields and leaf parthenolide concentrat ions showed that higher rates of nitrogen appl icat ion resulted in higher leaf parthenolide concentrat ions and higher shoot yields. Plants under mild water st ress had lower shoot yields but higher leaf and flower parthenolide concentrat ions than non-st ressed plants. 140 Parthenol ide concentration has been examined in feverfew t issue culture (Brown, 1993, Brown et al., 1996), feverfew grown in different regions and cl imates (Maries et al., 1992), and in feverfew plants and commerc ia l preparat ions (Awang et al., 1991; Heptinstall et al., 1992). T h e s e reports showed high variability in parthenolide concentration when measured in commercia l products and authenticated feverfew plants. The range of parthenolide concentrat ions found in the literature is 0 % to 2.8 % in leaves and between 0.8% and 2 .3% in f lowers by dry weight (Awang et al., 1991; Mar ies et al., 1992). Leaf and f lower parthenolide concentrat ions found in the exper iments descr ibed in Chapter 2 were within the reported range. Variabil i ty in parthenolide concentrat ions resulted from varying water and nitrogen supply. Thus these two resources in the environment affected sesqui terpene metabol ism. Environmental cues are required for normal plant development. As ide from providing energy for photosynthesis, light plays an important role in regulating many physiological p rocesses such as germination, f lowering and s e n e s c e n c e . For example, the transition from vegetative to reproductive growth in many plants is cued by photoperiod. Exper iments in Chapter 3 examined shoot yield and parthenolide concentrat ion during shoot development in vegetative and flowering s tages. Feverfew is a long-day plant that requires a dark period shorter than 12 hours per day to flower. W h e n grown under short days for extended periods it remained in a vegetative stage and accumulated high concentrat ions of parthenolide (over 4 % by dry weight) in leaf glandular t r ichomes. During the transition from vegetative to reproductive growth, leaf parthenolide concentrat ions decreased while the concentrat ion of other more complex S T L s increased. The accumulat ion of parthenolide may be a result of restricted biosynthesis of S T L s downstream from parthenolide synthesis during the vegetative stage. The restriction may be due to an inhibition, or an absence of the required 141 enzymes for convers ion of parthenolide to other S T L s . The biosynthesis of other S T L s during reproductive growth may be triggered by the s a m e mechan ism that triggers f lowering. There are direct appl icat ions of these findings to the commerc ia l cultivation of feverfew. Product ion of leaves with a high concentrat ion of parthenolide and low levels of other S T L s can be ach ieved by growing plants under short days . Alternatively, if the other S T L s or f lowers are pharmacological ly important, feverfew can be induced to f lower resulting in lower leaf parthenolide concentration and higher concentrat ions of other S T L s . Tr ichomes are the site of parthenolide biosynthesis in feverfew. The relationship between tr ichomes, leaf development and parthenolide concentrat ion was examined in Chapter 4. The mechan isms regulating S T L biosynthesis are unknown. Exper iments in Chapters 2 and 3 showed nitrogen and water availability and developmental status of feverfew had effects on leaf parthenolide concentrat ions. It is unknown exactly how these factors affected parthenolide biosynthesis. Parthenol ide concentrat ions may be regulated by manipulation of tr ichome s ize and spatial distribution. I developed a new method for observing tr ichomes using sl ide preparations of dried epidermal peals under the light microscope. Drying the epidermis left the lipophilic t r ichomes intact and clearly visible. Tr ichome density was found to be related to leaf parthenolide concentration and both varied with leaf age and developmental stage. The difference in s ize and density of t r ichomes on leaves of different ages or between leaves of vegetative and reproductive shoots suggests a mechan ism of control al lowing the plant to alter total terpene production by changing the spac ing and s ize of t r ichomes initiated on the leaf epidermis. Microsampl ing techniques made it possib le to perform chemica l analysis of individual t r ichomes from leaf and flower sur faces. G landu lar t r ichomes from flower petals were significantly larger and contained low concentrat ions of parthenolide 142 compared to glandular tr ichomes on the leaves. This result seemed to contradict the consensus of reports showing that f lowers had higher parthenolide concentrat ions than leaves (Banthorpe et al., 1990; Fontanel et al., 1990; A w a n g et al., 1991; Heptinstall et al., 1992). My research showed that parthenolide in feverfew flowers is most concentrated in t r ichomes on the receptacle and least concentrated in t r ichomes on the f lower petals. Receptac le tr ichomes have high densit ies and therefore high parthenolide concentrat ions. The receptacle t r ichomes are of similar shape and s ize as the leaf t r ichomes and individually contain similar amounts of parthenolide. The relationship of tr ichome density and s ize to parthenolide and S T L concentrat ion has appl icat ions in research and agriculture. Tr ichomes can be easi ly isolated and manipulated by scientists to study terpenoid biosynthesis and tr ichome development. Var iet ies of feverfew and other crops with glandular t r ichomes of commercia l importance may be selected for tr ichome density and s ize . Var iet ies could be selected for f lowers with large receptacles or for other organs with high parthenolide concentrat ion. Molecular techniques may be used to manipulate tr ichome density and s ize which could affect chemical concentration in the plant. The biosynthesis of terpenoids is under renewed investigation due to the recent d iscovery of a second terpenoid pathway (Rohmer et al., 1993). In plants, the methylerythritol phosphate ( M E P ) pathway to terpenoid biosynthesis occurs in plastids while the c lassical ly understood mevalonate pathway occurs in the cytosol (Rohmer, 1999). There is little known about the interaction between the two pathways and their contributions to S T L biosynthesis. There appears to be transport of terpenoids across the plastidic membrane but the mechan ism and rate of flux is unknown (Eisenreich et al., 2001). S o m e terpenoids produced in high concentrat ions, like isoprene and monoterpenes, are synthesized exclusively by the M E P pathway. Both pathways can 143 contribute subunits to the s a m e terpenoid while other types of terpenoids appear to be synthesized exclusively through a single pathway. For example , the sesqui terpene chamazu lene from chamomi le is of mixed biosynthesis from the two pathways (Adam and Zapp , 1998; A d a m et al., 1999) while the sterols are synthes ized exclusively through the mevalonate pathway (Lichtenthaler et al., 1997). The pathway to parthenolide biosynthesis was examined in Chapter 5. Like the biosynthesis of chamazu lene from chamomi le , both the mevalonate and M E P pathways contribute to parthenolide synthesis. The mevalonate pathway was clearly demonstrated as a significant contributor to the biosynthesis of parthenolide. The extent of involvement of the M E P pathway however remains unclear and requires further study. 6.2. F U T U R E R E S E A R C H The broad scope of this thesis presents many potential a reas for future research. One of the underlying and unanswered quest ions is what are the control mechan isms linking environmental stimuli and shoot development to sesqui terpene biosynthesis? The nitrogen and water effects on sesqui terpene metabol ism may be a response to altered physiology rather than a result of direct regulation of a terpenoid pathway. L ikewise, shoot development and the transition from vegetat ive to reproductive growth alters physiology, which in turn may affect terpenoid biosynthesis. The possible indirect nature of these effects will make it difficult to answer the quest ion. The best approach may be to determine what the genetic regulatory e lements are in genes involved in the terpenoid biosynthetic pathways and then from these identify which factors cause the changes in terpenoid metabol ism. Perhaps one of the most important but poorly understood p rocesses in plant terpenoid metabol ism is the transport of terpenoids ac ross organel le membranes. 144 Transport proteins likely exist but have not been found to date. The non-polar terpenoids may diffuse across membranes but they often contain phosphate groups that would inhibit diffusion. A pass ive mechan ism is unlikely s ince there are c lasses of compounds such as monoterpenes and triterpenes that utilize just one pathway for synthesis. Therefore it is more likely that there is strict control over the movement of terpenoids between compartments. The regulation of terpenoid pools in different cel lular compartments is currently a primary target of investigation. The M E P pathway is responsible for synthesiz ing many important plastidic compounds for photosynthesis, electron transport, and photoprotection. Local izat ion of the M E P pathway in plastids suggests that photosynthetic p rocesses may be tightly l inked to M E P pathway regulation. Research should focus on the relationship between plastidic p rocesses , such as photosynthesis, and M E P pathway regulation. More research is required to determine the involvement of the M E P pathway in parthenolide biosynthesis. The ideal isotope feeding experiment to would be with enr iched methylerythritol phosphate or 1-deoxyxylulose-5-phosphate. Both of these compounds are substrates for the M E P pathway. Methylerythritol phosphate is a committed precursor in the M E P pathway. Another approach is to use the mevalonate pathway inhibitor mevinolin while feeding labelled pyruvate. Th is will al low only the M E P pathway to contribute to parthenolide biosynthesis. The complimentary experiment using the M E P pathway inhibitor fosmidomycin with label led mevalonate substrates could a lso provide insight into parthenolide biosynthesis. The variability of pharmacological ly active terpenoids in commercia l crops, particularly in p rocessed medicinal herbs used for one compound of interest, raises concerns over safety and efficacy of these products. Manipulat ing terpenoid biosynthesis by cultivation methods is idealistic and may be impractical for many 145 producers in diverse cl imates and geographies. To ensure safety for consumers of medicinal and culinary crops with potentially harmful terpenoids, genet ic and chemical analys is could be performed to authenticate spec ies , determine levels of unsafe compounds , and ensure safe dosage . Determining chemica l or genet ic markers to authenticate a spec ies is relatively s imple and should be pursued by regulating bodies. On the other hand, determining correct dosage for medicinal plants is general ly difficult and understudied. The principle pharmacological activities in most herbal medic ines are unknown. Th is is an area that requires significantly more research. Feverfew is among the most studied medicinal herbs during the past 30 years . Even though its eff icacy in relieving migraine has been demonstrated, compounds other than parthenolide are thought to contribute to feverfew's antimigraine medicinal properties. Thus although feverfew is a well-studied medicinal plant, dosage will be difficult to recommend unless the other antimigraine compounds are identified and their interaction with parthenolide identified. A s with all medic inal plants, systemat ic research should be conducted to develop standard procedures for product authentication, and whenever possib le standardizat ion of dosage , to ensure safe products for consumers . 146 6.3 R E F E R E N C E S A d a m , K .P . , Thie l , R., and Zapp , J . (1999) Incorporation of 1-[1-C-13]deoxy-D-xylulose in chamomi le sesqui terpenes. Arch ives of Biochemistry and B iophys ics 369:127-132. A d a m , K .P . and Zapp , J . (1998) Biosynthesis o f the isoprene units of Chamomi le sesqui terpenes. Phytochemistry 48:953-959. A w a n g , D .V .C . , Dawson , B.A., Kindack, D .G. , Crompton, C . W . , and Heptinstal l , S . (1991) Parthenol ide content of feverfew (Tanacetum parthenium) a s s e s s e d by H P L C and 1 H - n m r spectroscopy. Journal of Natural Products 54:1516-1521. Banthorpe, D. V , Brown, G . D, J a n e s , J . F, and Marr, I. M. (1990) Parthenol ide and other volati les in the f lowerheads of Tanacetum parthenium I. Schu l tz bip. F lavour & Fragrance Journal 5:183-186. Brown, G . D. (1993) Product ion of anti-malarial and anti-migraine drugs in t issue culture of Artemisia annua and Tanacetum parthenium. Ac ta Horticulturae 330:269-276. Brown, A . M . G . , Lowe, K . C . , Davey, M.R., and Power , J . B . (1996) Fever few (Tanacetum parthenium): T i ssue culture and parthenolide synthesis. Plant Sc ience 116:223-232. E isenre ich , W. , Rohd ich , F., and Bacher , A . , (2001) Deoxyxy lu lose pathway to terpenoids. Trends in Plant Sc ience 6:78-84 Fontanel , D., Bizot, S . , and Beauf i ls , P. (1990) H P L C determination o f the parthenolide content of Chamomi le Tanacetum parthenium (L.) Schu lz-b ip . Plantes Medic ina les et Phytotherapie 24:231-237. Heptinstal l , S . , A w a n g , D.V., Dawson , B.A., K indack, D., Knight, D.W., and May, J . (1992) Parthenol ide content and bioactivity of feverfew (Tanacetum parthenium (L.) Schul tz-Bip.) . Estimation of commercia l and authenticated feverfew products. Journal of Pharmacy and Pharmaco logy 44:391-395. Lichtenthaler, H.K., Rohmer , M. , and Schwender , J . (1997) Two independent b iochemical pathways for isopentenyl d iphosphate and isoprenoid biosynthesis in higher plants. Physio logia Plantarum 101:643-652. Mar ies , R . J . , Kaminsk i , J . , A rnason , J.T. , P a z o s - S a n o u , L., Heptinstal l , S . , F ischer , N.H. , Crompton, C .W. , Kindack, D .G. , and A w a n g , D.V. (1992) A b ioassay fo r inhibition of serotonin release from bovine platelets. Journal of Natural Products 55:1044-1056. Rohmer , M. (1999) The discovery of a mevalonate- independent pathway for isoprenoid biosynthesis in bacteria, a lgae and higher plants. Natural Product Reports 16:565-574. 147 Rohmer , M. , Knani , M. , S imonin , P. , Sutter, B., and S a h m , H. (1993) Isoprenoid biosynthesis in bacteria: a novel pathway for the early s teps leading to isopentenyl d iphosphate. B iochemical Journal 295:517-524. Smith R . M . and Burford M.D. (1992) Supercri t ical fluid extraction and gas chromatographic determination of the sesqui terpene lactone parthenolide in the medicinal herb feverfew (Tanacetum parthenium). Journal of Chromatography 627:255-261. 148 APPENDIX 1 Table 1. Field irrigation and nitrogen application trial fertigation schedu le . The 50 kg/ha feed schedule was used for the field irrigation f requency trial. Date Application Rate Fertilizer kq/ha qm/N treatment area 0 50 100 0 50 100 May 22 0-53-34 15 15 15 53 48 53 (transplanted 34-0-0 0 18.4 36.8 0 59 130 May 16) M g S 0 4 20 20 20 71 64 71 Che la tes 1.5 1.5 1.5 5.3 4.8 5.3 June 5 0-53-34 15 15 15 53 48 53 34-0-0 0 18.4 36.8 0 59 130 M g S 0 4 40 40 40 142 128 142 Che la tes 1.5 1.5 1.5 5.3 4.8 5.3 June 19 34-0-0 0 18.4 36.8 0 59 130 Che la tes 1.5 1.5 1.5 5.3 4.8 5.3 July 3 34-0-0 0 18.4 36.8 0 59 130 M g S 0 4 30 30 30 106 97 106 Che la tes 1.5 1.5 1.5 5.3 4.8 5.3 July 17 34-0-0 0 18.4 36.8 0 59 130 Che la tes 1.5 1.5 1.5 5.3 4.8 5.3 July 31 34-0-0 0 18.4 36.8 0 59 130 M g S 0 4 30 30 30 106 97 106 Che la tes 1.5 1.5 1.5 5.3 4.8 5.3 August 14 34-0-0 0 18.4 36.8 0 59 130 Che la tes 1.5 1.5 1.5 5.3 4.8 5.3 August 28 34-0-0 0 18.4 36.8 0 59 130 M g S 0 4 30 30 30 106 97 106 Che la tes 1.5 1.5 1.5 5.3 4.8 5.3 1) May 10, 400 kg/ha of gypsum was appl ied to the beds and raked in. 2) Dissolve nutrients in warm water and adjust pH between 5.5-6.5. Make up to 20 L with water. 3) Fertigate both high and low water treatments, of the s a m e N rate, at the same time. / 149 Table 2. Field and greenhouse irrigation and medium trial nutrient feed formula. Field solution was made in 100 I tank and greenhouse solution was made in 500 I water. Nutrient solut ions were replenished weekly. Al l greenhouse-grown plants received this nutrient feed. This feed formula was used for all g reenhouse exper iments. Nutrient Ferti l izer Concentrat ion weight del ivered to (g) plants (g/l) N H 4 N O 3 (34-0-0) 133 0.266 C a C I 2 196 0.392 KH2PO4 (0-53-34) 100 0.2 KH2SO4 (0-0-50) 278 0.556 micronutrients 7.5 0.015 H2SO4 adjust pH to 6 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0090789/manifest

Comment

Related Items