Phenolic Metabolism in Higher Plants I. Catechol Biogenesis in Gaultheria Ii. the Biogenesis of Rosmarinic Acid in Mentha Iii. D

PHENOLIC METABOLISM IN HIGHER PLANTS

I. CATECHOL BIOGENESIS IN GAULTHERIA II. THE BIOGENESIS OF ROSMARINIC ACID IN MENTHA III. DEGRADATION OF AROMATIC COMPOUNDS BY STERILE PLANT TISSUES

BRIAN EDWARD ELLIS

B.Sc, University of New Brunswick, 19&5

A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of Botany

We accept this thesis as conforming to the

required standard

THE UNIVERSITY OF BRITISH COLUMBIA

November, I969 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 tha permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.

Department of

The University of British Columbia Vancouver 8, Canada

Date -i-

Abstract

I. Previous studies on biogenesis of simple phenols

ir plants have been restricted, to hydrocuinone.

Ariong the other simple phenols, catechol is of par•

ticular interest because of its potential role as

a ring-cleavage substrate. Tracer studies on the

biogenesis of catechol in Gaultherla leaf discs

showed that it was formed from salicylic acid by

oxidative decarboxylation. Salicylate decarboxyl-

ating activity could be detected in buffered ex•

tracts of very young leaves,

II. Among the numerous caffeic acid esters presently

known in plants, only 3-0-'caffeoylquinic acid (chlor-

os;enic acid) has been stud!.ed in detail. Rosmarin-

ic acid (alpha-O-caffeoyl-3»^-dihydroxyphenyllactlc

acid) has been reported to occur in a number of

plants but nothing was known of its biosynthesis or

metabolic role. Tracer studies demonstrated that

^n Mentha the caffeic acid moiety was formed from,

phenylalanine via cihnamic and para-coumarlc acids.

In contrast, the .structurally similar 3»^-dih.yd.roxy-

phenyllactic acid moiety was formed from tyrosine

and J>,k-dihydroxyphenylalanine. There was no evid•

ence of the participation of a para-coumaroyl ester

intermediate. Time-course studies and use of -ii-

labelled rosmarinic acid showed that endogenous rosmarinic acid was turning over slowly. The caffeoyl moiety, however, does not appear to be contributing to the formation of insoluble polymers, as has been suggested for chlorogenic acid in other plants.

III. Bacteria and fungi readily degrade aromatic com• pounds to carbon dioxide. Despite the large quan• tities of aromatic compounds formed in plants, little attention has been paid to the ability of plant tissues to degrade aromatic rings. No re• ported studies have used completely sterile plants and techniques. This has left open the possibility that the microflora associated with the plant might be carrying out the observed reactions.

The ability of sterile plant tissue cultures to de• grade aromatic ring-l^C compounds to carbon dioxide was studied. It was established that a number of tissues (Ruta, Triticum, Phaseolus, Melilotus) have the ability to cleave the aromatic ring of phenylal• anine. Melilotus tissue could also degrade cinnamic acid-ring-

14C suggesting that a dihydroxy phenolic acid may be the ring-cleavage substrate. Neither Ruta nor Melilotus tissues were able to degrade ben• zoic acid or salicylic acid-ring--*-^C. Tryptophan- benzene ring- C was shown to be degraded to carbon dioxide by both Ruta and Melilotus. In summary, the ability of plants to cleave the benzene ring of aromatic compounds when free of micro-organisms was thus established. -iv-

TABLE OF CONTENTS

Page

ABSTRACT i

TABLE OF CONTENTS iv

COMMON AND SYSTEMATIC NAMES OF COMPOUNDS vii

LIST OF TABLES viii

LIST OF FIGURES ix

ACKNOWLEDGEMENT x

PREFACE xi

I. CATECHOL BIOGENESIS IN GAULTHERIA

INTRODUCTION 1

MATERIALS AND METHODS

Plant material k Radioactive compounds ^ Isolation of catechol beta-D-glucoside and catechol 5 Salicylate hydroxylase assay 6

RESULTS AND DISCUSSION Populus 8 Gaultheria . 9 -v- Page

LITERATURE CITED 16

II. THE BIOGENESIS OF ROSMARINIC ACID IN MENTHA

INTRODUCTION 19

MATERIALS AND METHODS Plant material 26 Labelled compounds and their administration 26 Isolation of rosmarinic acid and caffeic acid 2? Specific activity determination 29

RESULTS AND DISCUSSION Isolation of rosmarinic acid 30 Biosynthetic studies 31 Time-course studies 35

LITERATURE CITED 4-3

III. DEGRADATION OF AROMATIC COMPOUNDS BY STERILE PLANT TISSUES

INTRODUCTION ^7

MATERIALS AND METHODS Plant tissue cultures 5? Radioactive compounds 58 Detection of ring-cleavage 59

Microbial contamination 6l -vi- Page

RESULTS AND DISCUSSION Aseptically grown plants 62 Plant tissue cultures 62 Tryptophan 63 Phenylalanine 63 Tyrosine 68

LITERATURE CITED 75

APPENDIX 80 -vii-

Common and systematic names of compounds

Common name Systematic name anthranilic acid 2-aminobenzoic acid caffeic acid 3»^-dihydroxycinnamic acid catechol 1,2-dihydroxybenzene chlorogenic acid 3-0-caffeoylquinic acid o-coumaric acid 2-hydroxycinnamic acid p_-coumaric acid 4-hydroxycinnamic acid

DOPA 3,4—dihydroxyphenylalanine dopamine 3,4-dihydroxyphenylethylamine

DOPL 3,^-dihydroxyphenyllactic acid esculetin 6,7-dihydroxycoumarin ferulic acid 3-methoxy-i)--hydroxycinnamic acid gentisic acid 2,5-dihydroxybenzoic acid homogentisic acid 2,5-dihydroxyphenylacetic acid hydroquinone 1,k-dihydroxybenzene phloroglucinol 1»3» 5-trihydroxybenzene protocatechuic acid 3,^-dihydroxybenzoic acid o-pyrocatechuic acid 2,3-dihydroxybenzoic acid rosmarinic acid alpha-0-caff eoy1-3,^-di hydroxyphenyllactic acid salicylic acid 2-hydroxybenzoic acid -viii-

List of tables

Table Page

Catechol Biogenesis in Gaultheria

I. The Incorporation of Radioactive Precursors into Catechol in Leaf Discs of Gaultheria adenothrix after 24 hr 7777777777 11

II. Decarboxylation of Salicylic Acid-7-l^C by Crude Extract of G. adenothrix Leaves 14

The Biogenesis of Rosmarinic Acid in Mentha

I. Esterified Forms of Caffeic Acid in Plants 22

II. Incorporation of Phenylpropanoid Compounds into Rosmarinic Acid in Mentha arvense L 33

III. The Turnover of Labelled Rosmarinic Acid Readministered to Mentha 41

Degradation of Aromatic Compounds by Sterile Plant Tissues

I. Oxidation of DL-Tryptophan-benzene ring_u-l4c to 1^C02 by Ruta and Melilotus over seven days 64

II. Metabolism of DL-Phenylalanine-l^C and Cinnamic-l4c by Ruta and Melilotus over seven days 65

III. Oxidation of Benzoic acid and Salicylic acid ring-U-14c by Ruta and Melilotus over seven days ...... 67

IV. Degradation of Tyrosine-1*^ and DOPA-^C by Ruta and Melilotus over seven days. -ix-

List of figures

Figure Page

Catechol Biogenesis in Gaultheria 1.. Proposed pathways for the formation of catechol in G-aultheria 12

The Biogenesis of Rosmarinic Acid in Mentha 1. Hydroxycinnamic acids in plant phenolic metabolism 20 2. Proposed phenylpropanoid metabolism leading to lignins 23 3. Rosmarinic acid 25 k. Proposed pathway for the biogenesis of rosmarinic acid in Mentha 36 5. Changes in the specific activity of rosmarinic acid in Mentha after a two hour exposure to 1^'C02 • • • 38 6. Changes in the specific activity of rosmarinic acid and the total activity of insoluble esters in Mentha after a pulse-feeding of phenylalanine-^-^C for (a) 2 hr and (b) 0.5 hr 39

Degradation of Aromatic Compounds by Sterile Plant Tissues 1. Feeding flask. CO2 regeneration system ... 60 -X-

Acknowledgement

To Dr. G.H.N. Towers, under whose supervision this work was carried out, goes my deepest appreciation for his guidance and constant encouragement. He has been generous with his time as well as his research facilities.

The faculty, staff and students of the Department of Botany have been helpful in many ways. I am particularly grateful to Dr. B.A. Bohm for helpful discussions and ad• vice in the course of this work, and to Dr. C.O. Person for the frequent use of his sterile transfer facilities. The financial assistance of the National Research Council of Canada and of the Department of Botany is gratefully acknowledged.

Finally, I am< very grateful for the encouragement and assistance of my wife, Margaret, both in the lab and in the preparation of this manuscript. -xi- Preface

Phenolic compounds occur universally in the plant king• dom. They include a vast variety of structures ranging from phenol itself to the highly polymerized lignins, and new compounds are being discovered each year as more plant species are examined chemically. More recently the focus of interest has shifted from the structural identification of these compounds to their biogenesis and metabolic role in plant tissues. The link between glycolytic metabolism and the aromatic compounds has been demonstrated and the general pattern of metabolic relationships has been re• vealed in some groups of phenolics such as the flavonoids, coumarlns, cinnamic and benzoic acids and the aromatic amino acids. Even the complexities of lignin formation are being gradually unravelled.

Other groups of compounds, however, have received less attention and many questions have remained unanswered in the metabolism of the simple phenols, the glycosides and the phenolic esters. Little serious attention has been paid to the role and extent of aromatic ring cleavage in plant tissues. As "secondary metabolites", phenolics have generally been considered to be isolated from oxidative metabolic pathways and yet considerable carbon and energy is tied up in these compounds for which no role has yet been demonstrated in many cases. Presumably it would be an advantage to the plant to be able to oxidize the arom• atic compounds as microorganisms do with such ease.

Work was begun on some of these questions and eventually -xii- three problems proved to be particularly interesting and were studied more or less simultaneously. This limited the extent of the area of study in each topic but the in• ability to carry out experiments on peripheral points of interest has been offset to some extent by the valuable opportunity to become familiar with the state of knowledge and the techniques in a number of areas in phenolic bio• chemistry. Catechol Biogenesis in Gaultheria -1-

INTRODUCTION

The simple phenols (hydroxy and methoxy-benzene com• pounds) seem relatively limited in their distribution in 1 2 the plant kingdom ' . Hydroquinone, 1,4-dihydroxybenzene, occurs in a number of Ericaceae"*", and in totally unrelated 3 genera such as Pyrus and Bergenia . Catechol, 1,2-dl- hydroxybenzene, has been reported in Allium^, in Psoro- 1 6 7 6 8 9 spermum , in Populus ' and Salix ' and in Gaultheria . The other simple phenols are even more limited in their reported distribution.

Catechol is of particular interest in phenolic meta• bolism because of its frequent occurrence as a shortlived 10 intermediate in ring-cleavage metabolism in microorganisms In order to oxidize the aromatic ring, these organisms usually require the presence of orthohydroxyl groups on the compound and this often results in higher molecular weight aromatic compounds being hydroxylated and degraded to catechol. It does not normally accumulate, however, "but is oxidized to aliphatic acids and further metabol- lzed Catechol occurs most prominently in higher plants in 6,7 6,8 Gaultheria and the Salicaceae. In Populus and Salix it is demonstrably present in small concentrations in the unhydrolyzed extracts of leaves and has been isolated from

Populus bark"*""1". There are conflicting reports of its 8 occurrence in a bound form in the plant but to date there is no report of the isolation of a bound form of catechol -2- from the Salicaceae and catechol monoglycosides (the most likely form of such a compound) are absent from chromato- 12 grams of Populus leaf extracts . In contrast, catechol occurs in three Gaultheria spp. in substantial quantities as the beta-D-glucoside with little or no free catechol 9 detectable in the extracts . When this study was undertaken the only work on simple phenol biogenesis in higher plants had been carried out on hydroquinone. Its beta-D-glucoside, arbutin, occurs in Pyrus where Grisdale and Towers determined that the hydro• quinone moiety was formed from phenylpropanoid precur-

13 sors . This was later extended by Zenk, who showed that para-hydroxybenzoic acid was oxidatively decarboxylated to In• form hydroquinone in Bergenia The analogous reaction could be expected to result in conversion of salicylic acid to catechol, but a number of other interesting possibilities exist. These include non-oxidative decarboxylation of 2,3-dihydroxybenzoic 15 , 16 acid or 3.4-dihydroxybenzoic acid , and a complex con• version of anthranilic acid to catechol either via 2,3- 1? 18 dihydroxybenzoic acid or ortho-amino phenol . The latter reaction has been demonstrated in vitro with an 18, enzyme system from Tecoma stans (Bignoneaceae) but no 19 catechol could be detected in extracts of Tecoma and it was not shown to be a normal metabolite of the plant.

The other reactions have only been demonstrated in micro• organisms, which have also been found to form catechol by 20 hudroxylation of phenol and by oxidative decarboxylation of salicylic acid . The latter reaction has been studied 21 in vitro with a partially purified enzyme system In view of the scarcity of information regarding sample phenol formation iin plants, and the potential role of cat• echol as a ring-cleavag:e substrate, the biogenesis of cat• echol in higher plant tissues was considered a worthwhile s tudy. MATERIALS AND METHODS

Plant material. Populus leaves were obtained from trees growing on or near the campus of the University of British Columbia, Vancouver. Gaultheria adenothrlx (Miq.) Mich, was grown in flats in the departmental greenhouse under daylight supplemented with sixteen hour-day fluorescent lighting. Young leaves were used when approximately two-thirds mature size and still light green in colour. G. ovatifolia Gray was collected near Alta Lake, British Columbia, in May.

Radioactive compounds. 14 i4 Benzoic acid-rlng-1- C, tryptophan-benzene ring-U- C Inl• and phenylalanine-U- C were purchased from the Radio- 3 chemical Centre, Amersham, England, tryptophan-GL- H from 14 ICN, California, salicylic acid-7- C from New England Nuclear, Boston, Massachusetts and salicylic acid-ring- 14 U- C from Mallinckrodt Nuclear Corporation, Orlando, Florida. Tritiation of the ring proton positions of 2,3- dihydroxybenzoic acid and 4-hydroxybenzoic acid was carried out by refluxing 1 mM of the compound in 10 ml of 3 3 CF3COO H (prepared from H2O and trifluoroacetic anhyd• ride) with 10 mg of Pd on charcoal for forty-eight hours. Labile tritium was removed by successive solution and evaporation and the compound twice crystallized from water. The 2,3-d.ihydroxybenzoic acid and the 4-hydroxy-

benzoic acid had specific activities of 266 p.C/mM and.752 p-C/mM, respectively. All the acids were administered as their ammonium salts in aqueous solution. Ten 1.2 cm discs cut from washed leaves were infiltrated with the radioactive solution (5 ml) using water aspirator vacuum. The discs were then floated on the residual solution in small Petri dishes under constant illumination (9000 lux) at 20° for twenty- four hours.

Isolation of catechol beta-D-glucoside and catechol. The rinsed discs were extracted with hot 95$ ethanol until colourless. The solvent was evaporated and the res• idue extracted with hot water followed by filtration through Celite. The filtrate was concentrated to a few milliliters and run on to a small column of Avicel (micro- crystalline cellulose) which was then eluted with 2% formic acid. The fractions containing catechol glucoside (ab• sorbing strongly at 278 nm) were taken to dryness and purified by preparative thin-layer chromatography on 1 mm Avicel layers, using as solvents n-butanol:acetic acid:water (4:1; 2. 2), n-butanol:pyridine:water (75'. 15'. 10) and ethyl acetate:formic acidswater (75:10:10). The catechol gluco• side was located by spraying a strip of the plate with o diazotized para-nitro aniline reagent and NaOH . (catechol glucoside-magenta; catechol-blue). The glucoside was then -6-

dissolved in 2 ml water and incubated with emulsin (beta-

glucosidase) for twenty-four hours. The hydrolysate was

banded on Avicel plates and developed in benzeneJacetic

acid:water (10:7:3-organic phase). The catechol band was

eluted and the concentration measured spectrophotometri-

cally at 276 nm. The radioactivity was determined by

liquid scintillation counting. Further chromatography did

not alter the specific activity.

Salicylate hydroxylase assay.

2-5 gm of young G. adenothrix leaves were washed

successively with distilled water, 0.05$ Tween 80, 70%

aqueous ethanol and distilled water (twice). They were

then blotted dry and ground with twice their weight of

Polyclar AT (polyvinylpyrrolidone) and 1-2 gm of acid- washed sea sand, wetting the mixture with sufficient 0.05

M phosphate buffer (pH 7) to make a thick slurry. The

slurry was squeezed through five thicknesses of cheese•

cloth and centrifuged at 10,000 g for ten minutes. The

clear supernatant (app. 20 ml) was used as enzyme sol• ution.

The incubation mixture consisted of 9 ml of the super•

natant to which were added 15 NADH2» 2.5 ^uM FAD, 5 ^ Ik NADPH2 and 1.3 /iM salicylic acid-7- C (3/*C) for a final

volume of 10 ml. The incubation was carried out at room

temperature in a 125 ml side-arm flask equipped for a

flow-through of air. A slow stream of air was passed through the flask during incubation (2 hours), the re• action mixture acidified with 2 ml cone. HG1 and a rapid airstream (app. 100 ml/min.) passed through the flask for one-half hour. The C0£ was trapped in 10 ml 2-phenylethylamine, which was subsequently counted by liquid scin^ tillation counting. The control was treated identically except that the supernatant was held at 100° for 20 min• utes prior to being added to the incubation mixture. -8-

RESULTS AND DISCUSSION

Populus. The presence of catechol in unhydrolyzed extracts of the leaves of local species of Populus could be demon• strated chromatographically, although the levels were low. There was no sign of catechol monoglycosides in the ex• tracts though, and no glycoside fraction could be shown to consistently yield catechol upon hydrolysis. Tomaszew- ski has found a similar situation in Populus wllsonii, P. 22 tremula and two Salix spp . Vazquez et al found free catechol in Salix viminalis but they maintain that a bound g form is also present . Pearl and Darling isolated cate- 11 7 chol from extracts of P. balsamifera bark and leaves . In the extensive studies of the phenolic glycosides of the 23 7 Salicaceae by Thieme ^ and Pearl and Darling , they have never reported a catechol glycoside or an indication of one from hydrolysis products.

If, however, Populus leaf discs are exposed to a dilute solution of catechol they form a compound chromatograph- 12 ically identical with catechol-beta-D-glucoside It would appear, then, that the catechol found in the Salicaceae is to a large extent present in the free form and must be compartmentalized in some way within the tissues. Efforts to demonstrate incorporation of label from aromatic precursors into the catechol in P. tricho- carpa leaves of all ages were unsuccessful. Similar diff- -9- iculties have been encountered by Zenk . The reason may be that synthesis is taking place elsewhere in the tree followed by translocation to the leaves, or that some un• suspected biosynthetic pathway is involved.

Gaultheria.

Catechol is the most prominent phenolic compound in hydrolyzed extracts of Gaultheria adenothrix (Miq.) Mich.

In the original report of the occurrence of catechol-beta-

D-glucoside in Gaultheria spp. it was suggested that it might arise from either 2,3-dihydroxybenzoic or 3,4- dihydroxybenzoic acid, both of which occur in G. adenoth- 9 rix . On the basis of the biosynthetic studies on hydro- 14 quinone , however, the decarboxylation of salicylic acid seemed a more likely possibility despite the apparent ab- o sence of this acid from the extracts . As can be seen from Table I, salicylic acid was incor• porated into catechol with less dilution of specific ac• tivity than any other compound tried. Both labelled ben• zoic acid and phenylalanine served as precursors to cate• chol as well. Phenylalanine is readily converted to cinnamic acid in plants, and the latter can undergo beta- oxidation to yield benzoic acid"^. Benzoic acid has been shown to be a good precursor to salicylic acid in higher 25 plants . The pathway involved may therefore be phenyl• alanine—¥ cinnamic acid—> benzoic acid—> salicylic acid —> catechol. The large variations in incorporation of -10-

benzoic acid (five replicates) may be an indication that the conversion cinnamic acid—> ortho-coumaric acid > 26 salicylic acid plays a dominant role at times (Figure 1.). The age of the leaf tissue determined the extent of conversion of salicylic acid to catechol, young leaves possessing activity at least ten times that of mature leaves. All reported work was carried out on leaves less than 2 weeks old. A similar effect of age has been noted previously in in vivo studies on hydroxylation of benzoic 25 13 acids and formation of hydroquinone . In one experiment 14 salicylic acid-ring-U -C was administered to leaf discs of G. ovatlfolia, one of the other two species of Gaulther• ia reported to contain catechol. Labelling of the cate• chol was again observed (Table I.) the greater dilution probably being due to the maturity of the leaves obtained. The possibility that decarboxylation of 2,3-dihydroxy• benzoic acid or 3,4-dihydroxybenzoic acid is involved in catechol biosynthesis was examined. The former compound (three replicates) and 4-hydroxybenzoic acid (three rep- 30 licates), a likely precursor of the latter were fed to leaf discs; in neither case was any label detected in catechol. Salicylic acid is converted in part to 2,3- dihydroxybenzoic acid, however, since after feeding 20 ^iG 3 of salicylic GL- H (sp. act. 238 mC/mM) for 24 hours, 9.3 x 10"^ dpm were found in catechol (sp. act. 475 yuC/mM) and 1.34 x 10^ dpm were found in 2,3-dihydroxybenzoic acid. -11-

Table I. The Incorporation of Radioactive Precursors into Catechol in Leaf Discs of Gaultheria adenothrix after 24 hr.

Specific Activity Activity activity Compound fed fed taken up of Dilution (jxC) (/AC) catechol (/iC/mM)

i4 2 1.53 2.06 2.4xl05 Phenylalanine-U- C, 495 ^C/uM 1 4 2 1.80 3.48 730 Benzoic acid-ring-1- C, 2.54 /iC//iM 14 2 1.94 0.12 2100 Benzoic acid-ring-1- C, 2.54 ;uC//iM Salicylic acid-ring- 2 1.68 5.88 161 UL- C, 0.95 JuC/uM Salicylic acid-ring- 2* 1.26 0.24 3890 UL-1^, 0.95 pC/jM Salicylic acid-ring- 2** 1.96 5.90 161 UL-1^, 0.95 }xQ/jm 1.83 0.15 6330 2,3-Dihydroxybenzoic 0.114 0.10 nil acid-3H, 0.27 pC/pM

4-Hydroxybenzoic acid-1 1 0.80 nil 3H, 0.75 p.C/p&

Tryptophan-benz ene- 4 2.99 nil ring-UL-^C, 52 uC/uM • Tryptophan-GL-3H, 50 21 0.14 1.93xl07 2.7 C/mM

G. ovatifolia leaves ** Leaves less than 2 weeks old

*** Leaves more than 6 weeks old -12-

Pigure 1. Proposed pathways for the formation of catechol in Gaultheria

COOH

CINNAMIC ACID PHENYLALANINE

COOH

0-COUMARIC ACID BENZOIC ACID COOH OH SALICYLIC ACID \ COOI-I OH COOH OH OH

HO" H GENTISIC ACID CATECHOL 0-PYROCATECHUIC ACID -13-

Tryptophan degradation by Aspergillus niger leads to 17 catechol ' but there was no clear evidence for this in Gaultheria. The small amount of activity in catechol after feeding 50 yuC of tryptophan-GL--% may have been re• cycled through the shikimate pathway from loss of the tryptophan side-chain. The possibility of a small amount of degradation to the aromatic nucleus, however, cannot be excluded. Unlike catechol, salicylic acid is quite common in Gaultheria, appearing in some species in relatively large o quantities as a glycoside of methyl salicylate . The sub• group, Amblyandra, however, would seem to have evolved a system for converting this salicylic acid to catechol which is then stored as its beta-glucoside. It is inter• esting to note that an acid which is apparently absent from extracts of the plant is actually actively involved in its phenolic metabolism and hence does not accumulate in the tissues. The formation of a number of methoxyl- 27 ated phenols in higher plants has recently been reported In each case the corresponding benzoic acid was oxidatively decarboxylated. This result confirms the pattern of simple phenol formation in higher plants. An attempt was made to detect salicylate decarboxyl- ating activity in an in vitro system. By using the super• natant from buffered extracts of the very youngest leaves 14 14 and either ring- C or carboxyl- C salicylic acid as sub• strate, the reaction could be detected. Using the more -In•

sensitive CO2 trapping assay, approximately 0.25$ of the label in the supplied substrate could be detected in the CO? trap after 2 hours (Table II.).

Table II. 14 Decarboxylation of Salicylic Acid-7- C by Crude

Extract of G. adenothrix Leaves

total dpm in trap % of substrate fed

supernatant 17,600 0.26

control 580

This is comparable to the rate of decarboxylation re• ported for the semi-purified salicylate hydroxylase iso- 21 lated by Katagiri et al from a soil pseudomonad . Fur• ther studies of the actual cofactor requirements of the Gaultheria system have not been possible due to the lack of adequate supplies of young tissue. The use of Pyrus or Bergenia tissues as sources of enzyme might be an easier approach to this enzyme or group of enzymes in plants. Katagiri et al did not examine 4-hydroxybenzoic acid as a substrate for their enzyme. They did report, however, that while 2,3-dihydroxybenzoic acid was a sub- 21 strate, anihranilic and benzoic acids were not Chen studied decarboxylation of benzoic acids by ace- 28 14 tone powder extracts of higher plant tissues . The CO2 was recovered from incubation mixtures containing car- -15-

14 boxyl- C-benzoic acids but the expected phenolic prod• ucts (assuming a non-oxidative decarboxylation) could not be detected. On the basis of the work now published the reaction products in Chen's work could reasonably be ex• pected to be the hydroxylated forms instead (e.g. hydroxy- hydroquinone from 3.4-dihydroxybenzoic acid). Two simple phenols which may prove to be exceptions to this pattern of formation are toluhydroquinone (2,5- dihydroxytoluene) and'phloroglucinol (l,3t5-trihydroxy- 29 benzene). The former occurs in quantity in Pyrola spp. but has been proposed as a general intermediate in plasto- 30 quinone biosynthesis^ . The suggested pathway would mean that tyrosine gives rise to toluhydroquinone via homogentisic acid (2,5-dihydroxyphenylacetic acid). This has 24 recently been shown to be the case in the Ericaceae . It is also possible that where phloroglucinol accumulates it may be the product of a polyacetate cyclization, either 31 directly or, indirectly, via flavonoid breakdown . -16-

LITERATURE CITED

1. Karrer, W. 1958. Konstltution und Vorkommen der organischen Pflanzenstoffe. Blrkhauser-Verlag, Basel and Stuttgart.

2. Pridham, J.B. 19&5• Low molecular weight phenols in higher plants, Ann. Rev. Plant Physiology 16:13. 3. Friedrich, H. 1958. Untersuchungen uber die phenolischen Inhalt stoffe von Pyrus communis, Die Pharmazie 13:153.

4. Friedrich, H. 1961. Physiology and biochemistry of glycoside synthesis, Planta med. (Stuttgart) 9:425.

5. Link, K.P. and Walker, J.C. 1933. The isolation of catechol from pigmented onion scales and its significance in relation to disease resistance in onions. J. Biol. Chem. 100;379.

6. Tomaszewskl, M. i960. Occurrence of p_-hydroxybenzoic acid and some other phenols in vascular plants, Bull. Acad. Polon. Sci. 8j6l.

7. Pearl, I.A. and Darling, S.F. 1968. Studies on the leaves of the family Salicaceae XI. The hot water extractives of the leaves of Populus balsamifera, Phytochem. 7:1845. 8. Vazquez, A., Mendez, J., Gesto, M.D.V., Seoane, E. and Vieitez, E. 1968. Growth substances isolated from woody cuttings of Salix viminalis L. and Ficus carica L., Phytochem. 7:161.

9. Towers, G.H.N., Tse, Aida and Maass, W.S.G. 1966. Phenolic acids and phenolic glycosides of Gaultheria species, Phytochem. 5:677.

10. Towers, G.H.N. 1964. Metabolism of phenolics in higher plants and microorganisms, p. 249, Biochem• istry of Phenolic Compounds, ed. J.B. Harborne, Academic Press, London and New York.

11. Pearl, I.A. and Darling, S.F. 1968. Studies on the barks of the family Salicaceae XIX. Continued studies on the hot water extractives of Populus balsamifera bark, Phytochem. 7:1851.

12. Ellis, B.E. 1968. Unpublished observations. -17-

13. Grisdale, S.K. and Towers, G.H.N. I960. Biosyn• thesis of arbutin from some phenylpropanoid compounds in Pyrus communism Nature 188:1130.

14. Zenk, M.H. 1964. Einbau von p_-hydroxybenzoesaure in die Hydrochinonkomponente des Arbutins in Bergenia crassifolla, Z. Naturforsch, 196s856.

15. Subba Rao, P.V., Moore, K. and Towers, G.H.N. 1967. o-Pyrocatechuic acid carboxy-lyase from Aspergillus niger, Arch. Biochem. Biophys. 122:466.

16. Cain, R.B., Bilton, R.F". and Darrah, J.A. 1968. The metabolism of aromatic compounds by micro• organisms. Metabolic pathways in the fungi, Biochem. J. 108:797.

17. Subba Rao, P.V., Moore, K. and Towers, G.H.N. 1967. The conversion of tryptophan to 2,3-dihydrox- benzoic acid and catechol by Aspergillus niger, Biochem. Biophys. Res. Commun. 28:1008.

18. Nair, P.M. and Vaidyanathan, C.S. 1966. Conversion of isophenoxazine to catechol in Tecoma stans, Arch. Biochem. Biophys. 115:515*

19. Subba Rao, P.V. Personal communication.

20. Evans, W.C. 1947. Oxidation of phenol and benzoic acid by some soil bacteria, Biochem. J. 41:373. 21. Katagiri, M., Yamamoto, S., and Hayaishi, 0. 1962. Flavin adenine dinucleotide requirement for the enzymic hydroxylation and decarboxylation of salicylic acid, J. Biol. Chem. 237sPC 2413. 22. Tomaszewski, M. Personal communication.

23. Thieme, H. and Richter, R. I966. Isolierung eines neuen Phenolglykosids aus Populus tremula L., Die Pharmazie 211251.

24. Zenk, M.H. Personal communication.

25. el-Basyounl, S.Z., Chen, D., Ibrahim, R.K., Neish, A.C. and Towers, G.H.N. 1964. The biosynthesis of hydroxybenzoic acids in higher plants, Phyto- chem. 3:485.

26. Volmer, K.0. and Grisebach, H. 1966. Zur Biosynthese der Benzoesauren in Gaultheria procumbens III, Z. Naturforsch. 216:435. -18-

27. Bolkart, K.H. and Zenk, M.H. 1968. Zur Biosynthese methoxylierter Phenole in hoheren Pflanzen, Z. fur Pflanzenphysiologie 59'^39 28. Chen, D.C.T. 1963. M.Sc. Thesis, Dalhousie Univ., Halifax, N.S. 29. Inouye, H. 1956. Uber die Bestandteile der Pirol- aceae-Pflanzen. VI. Uber die Bestandteile von Pirola incarnata Fisch. Pharmaceutical Bull. Japan 4:281.

30. Whistance, G.R. and Threlfall, D.R. 1968. Bio• synthesis of Phytoquinones. Biosynthetic or• igins of the nuclei and satellite methyl groups of plastoquinone, tocopherols, and tocopherol- quinones in maize shoots, bean shoots and ivy leaves, Biochem. J. 109:577. 31. Patschke, V.L., Hess, D. and Grisebach, H. 1964. Uber den Abbau von 4,2',4*,6'-Tetrahydroxy- chalkon-2-glucosid und 4,2',4'-Trihydroxychalkon- 4-glucosid in Rotkohlkeimlingen und Petunien, Z. Naturforsch. 19b:1114. The Biogenesis of Rosmarinic Acid in Mentha -19-

INTRODUCTION

The hydroxylated and methoxylated derivatives of cinn• amic acid form the most important and most studied pool of phenolic intermediates in plant tissues'''. From the cinn• amic acids arise the coumarins, flavonoids and anthocyanin pigments, the lignins, the benzoic acids and hence most simple phenols (Figure 1.). Plants form these aromatic compounds almost exclusively from the amino acids phenyl- 2 alanine and tyrosine, especially the former . The ability to deaminate phenylalanine to cinnamic acid seems to be universal in higher plants but the corresponding activity toward tyrosine to form para-coumaric acid (4-hydroxycinn• amic) is generally low outside the Gramineae and effect- 3 ively absent in many groups of plants . Tyrosine is hydroxylated in some plants to form 3,4- 4 dihydroxyphenylalanine (DOPA) . A few plants have yielded acetone powders which will deaminate DOPA to caffeic 3»5 acid ' (3,4-dihydroxycinnamic) but whether this reaction is of any significance in vivo is unknown. The reported distribution of DOPA in the plant kingdom is limited and so far it is known to be a precursor to some alkaloids 8 and to the betacyanin pigments . Phenyllactic and phenylpyruvic acids are metabolically active in plants and have been demonstrated to give 9 rise to cinnamic acid products such as lignin . There has been no report, however, of a direct dehydration of -20-

Figure 1. Hydroxycinnamic acids in plant phenolic metabolism

£22

shikimate pathway

simple phenols -21- phenylactic acid to cinnamic acid with an enzyme system from plants. With the discovery of phenylalanine ammonia- lyase it has generally been assumed that phenylalanine, phenyllactic and phenylpyruvic acids form a readily in• terchanging pool'*""1' from which phenylalanine provides the main, or sole, entrance to the cinnamic acids. The same situation "occurs with respect to tyrosine, p_-hydroxyphenylpyruvic and p_-hydroxyphenyllactic acids.

In plants lacking appreciable tyrosine ammonia-lyase activity, the distribution of label from radioactive phen• ylalanine and tyrosine is strikingly different within the non-nitrogenous metabolites. Label from phenylalanine is normally incorporated predominantly into the phenolic compounds while very little label from tyrosine appears 12 in these compounds . The label from tyrosine is spread throughout the organic acids and carbohydrates, indic• ating extensive degradation of the phenylpropanoid struc- 13 -ture, possibly including aromatic ring cleavage . Al• though an enzyme system isolated from spinach will hy- 14 droxylate phenylalanine to form tyrosine , there has been no indication from in vivo tracer studies that this takes place to any extent in the intact plant1''. The fates of these two amino acids are, then, largely distinct except when tyrosine or DOPA ammonia-lyase activity is present. The cinnamic acids ordinarily do not occur free in higher plant tissues in more than trace quantities. In- -22-

16 stead, they are found as glycosides , or esterified with the hydroxyls of glucose, glucosides and a variety of 17 aliphatic acids (Table !.)•

Table I.

Esterified Forms of Caffeic Acid in Plants.

Ester Distribution

20 caffeoylquinic acid widespread

(mono-and dicaffeoyl forms) 21,22 shikimic acid prob. widespread 23 24 tartaric acid Vitis , Chlcorium (mono-and dicaffeoyl forms)

malic acid Phaseolus 25 17 glucose widespread 26 glucosamine Nicotiana 27,28 glucose as glucosides widespread 3,4-dihydroxyphenyl- many Labiateae, lactic acid Boraginaceae and others '

"e thanol-i ns oluble' widespread?^0 compounds

They are also found esterified with some "ethanol- insoluble" component(s) of the cell which has not been completely characterized yet but is believed to be a poly- peptide or protein . -22a-

Caffeic acid is the most widely accumulated hydroxycinnamic acid in the plant kingdom; in one survey it was

identified in 80$ of the 122 species examined1^. Caffeic acid glycosides have been reported much less frequently than caffeoyl esters, chlorogenic acid (3-0-caffeol- 20 quinic acid) being particularly widespread , The pos• itional isomers of mono- and dicaffeoyl quinic acid are not easily distinguished chromatographically. They have only recently been unequivocably chemically character• ized^1'-^2 and it is likely that they will be found to be widespread as well2<^. Caffeic acid is an established intermediate in the pathway leading from phenylalanine to the coniferyl and sinapyl alcohols which are the proposed immediate precur• sors to the lignins1. The extent to which the pools of esterified caffeic acid participate ln this pathway, how• ever, is still not clear. Tracer studies on phenyl- 12 propanoid metabolism in Salvia established the sequence shown in Figure 2. The results suggested, however, that .only a small fraction of the pool of caffeic acid was turning over rapidly, possibly the free caffeic acid as opposed to the bulk of the esterified compound. Studies in wheat shoots, on the other hand, have shown that the "ethanol-insoluble" cinnamic esters became labelled much more rapidly than the soluble pools and it was suggested that these insoluble compounds were the 18 30 most likely precursors to lignin . A study of the -23-

Figure 2. Proposed phenylpropanoid metabolism leading to lignins. cinnamic acid I p-coumaric acid caffeic acid i ferulic acid > (coniferyl alcohol) ^ J^) lignins sinapic acid > (sinapyl alcohol)

response of phenolic metabolism in wheat to rust infection confirmed the earlier results in healthy wheat shoots. Susceptible-reacting plants, however, differed in showing more activity in the soluble esters than the insoluble 33 pool when fed labelled phenylalanine^ . More recently, work on the biosynthesis of scopoletin in tobacco tissue cultures using an elegant double-labelling technique has demonstrated that in those cultures the trace quantities of free cinnamic acids are the metabolically most active forms, the soluble esters accumulating label more slowly and the insoluble esters showing little 21 labelling from phenylalanine or cinnamic acid . There were very low levels of insoluble esters, however . The biosynthetic studies of the soluble caffeoyl esters have been restricted to chlorogenic acid. Use of tracers 33 3^ 22 3^ 38 in a number of plants (Nicotiana ' ,Solanum 39»40 Xanthium ' has established that both direct esterification of caffeic acid with quinic acid, or the form- -24-

ation of 3-0-p_-coumaroyl quinic acid with subsequent hydroxylation of this ester can be involved, Recent studies on chlorogenic acid metabolism in 39 40 Xanthiunr ^* have shown that in this plant the caffeoyl esters are not metabolic end-products but are turning over, with a half-life of about fourteen hours for 3-0- caff eoyl quinic acid. This ester appeared to be serving as a precursor to 3t5-di-0-caffeoyl quinic acid, and it was suggested that the latter, and perhaps both, compounds were serving as substrates for polymerization reactions which would produce lignins. The "alcohol-insoluble" es• ters contained comparatively little label from phenylpropanoid precursors but the use of aqueous methanol in the extraction procedure-^ makes it difficult to compare this result with the earlier work on wheat where only 80% 30 ethanol was used for extraction of the "soluble" esters-' . If caffeic acid esters are really playing an active metabolic role in plants it becomes of interest to examine the other less well-known esters. Rosmarinic acid is the caffeoyl ester of 3,4-dihydroxyphenyllactic acid (DOPL) esterified through the alpha hydroxyl, and is particularly interesting because nothing is known concerning the source of the DOPL moiety (Figure 3.). This ester was originally isolated from Rosmarinus 4l officinalis . It occurs together with chlorogenic acid in Sanicula spp. where the distribution of the two esters has been studied quantitatively within the plant and -25-

42 within the genus . Its distribution in the plant king• dom has been examined chromatographically in higher 29 43 plants 'and in ferns . These reports show rosmarinic acid to be widespread within a limited number of plant families, notably the Labiateae and Boraginaceae, and at present it is the second most common caffeoyl ester re• ported in plants. This study was undertaken to determine the biogenesis of this caffeic acid ester and, if possible, to see if it plays a role similar to that of chlorogenic acid in Xanthlum.

Figure 3. Rosmarinic acid.

OH -26-

MATERIALS AND METHODS

Plant material. Rosmarinus officinalis was grown on the campus of the University of British Columbia. Mentha arvense was normally grown in flats in the de• partmental greenhouses under daylight supplemented by sixteen-hour-day fluorescent lighting. M. piperita and M. arvense used in feeding experiments with labelled ros• marinic acid were growing outdoors in June and July.

Shoots were chosen which carried 5-7 pairs of fully ex• panded leaves but which had no flower heads. The freshly cut shoots had their ends recut under water before use.

Labelled compounds and their administration. 14 14 L-Phenylalanine-U- C, DL-phenylalanine-2- C, 14 3 L-tyrosine-3- C, L-phenylalanine-G- H and L-tyrosine- 3 G- H were purchased from the Radiochemical Centre, Amer- 14 sham, England. Cinnamic acid 2- C was purchased from 14 14 ICN, California. DL-DOPA-2- C, malonic acid-2- C and 14 barium carbonate- C were purchased from New England 14 Nuclear, Boston. The radiochemical purity of the C- amino acids was checked by paper chromatography and strip- 14 14 scanning. p_-coumaric acid-2- C and caffeic acid-2- C were prepared by the condensation of the appropriate benzalde- 14 hyde with malonic acid-2- C in pyridine with a trace of 44 piperidine . The products were purified by crystalliz- -27-

ation from water, and in the case of caffeic acid, by TLC and lead acetate precipitation as well. The acids were administered as their ammonium salts in

1 aqueous solution (0.5-1.0 ml) to 5-10 gm of mint shoots under constant illumination (9000 lux) at 22°. The sol• ution was normally completely taken up in one-half to two hours and was followed with distilled water.

Isolation of rosmarinic acid and caffeic acid. Rosmarinic acid was isolated on a large scale using the classical lead acetate precipitation technique employed in 41 its original isolation The ester could be isolated from 5-10 gm of mint shoots by use of the above technique on a much reduced scale. After the appropriate incubation period with a labelled precursor, the shoots were homogenized in boiling 95% eth• anol and extracted until colourless. The ethanol extract was taken to dryness, taken up in a small volume of hot water (15 ml) and filtered through Celite. The filter cake was washed with hot water and the combined filtrates were treated with 20% (w/v) aqueous neutral lead acetate until no further precipitate appeared. The precipitate was separated by centrifugation and washed twice with dis• tilled water by resuspension and centrifugation. The re- suspended precipitate was then decomposed by bubbling

H2S through the suspension. Decomposition was completed by heating on the steam bath and the precipitate of lead -28- sulphide was removed by filtration while still hot. The resulting solution of "catechols" or polyphenols was ex• tracted three times with a small volume of ether. The ether extract was taken to dryness, taken up in a minimal volume of hot water and filtered through decolourizing charcoal on a small sintered glass funnel. The filtrate was stored at 4° and crystallization of the rosmarinic acid initiated by seeding with a minute particle of the pure compound. The ester was recrystallized from water and was sufficiently pure after one recrystallization. The yield of purified rosmarinic acid from one batch of shoots varied from 2 to 10 mg.

The distribution of label between the two moieties of the ester was determined by hydrolysis and chromatographic isolation of the caffeic acid by TLC on Avicel plates.

Most hydrolyses were carried out in 2N NaOH under N2 for 22 2 hr. Later, it was found that pectinase hydrolyzed the ester overnight with better yields of products. The hydrolysis products were extracted into ether after acid• ification and the extract chromatographed on Avicel plates using as solvent chloroform shaken with 0.2 volume of water methanol:formic acid (71:125:4). Multiple dev• elopment separated caffeic acid from residual rosmarinic acid, DOPL and esculetin (the coumarin formed by light- 45 activated cyclization of cls-caffeic acid ). The caffeic acid band, easily detected by its bright blue fluorescence under 366 nm UV light, was eluted with 95$ ethanol. -29

Specific activity determination. The concentration of rosmarinic acid or caffeic acid in 95$ ethanol solution was determined spectrophotometrically at 331 nm and 325 nm respectively. An aliquot of the same solution was used for liquid scintillation counting. Blanks for compounds isolated by TLC were pre• pared by elution of bands at the same R^ on blank plates developed in the appropriate solvent. -30-

RESULTS AND DISCUSSION

Isolation of rosmarinic acid. Attempts to obtain a sample of pure rosmarinic acid from Mentha by preparative paper chromatography were un• successful because of the inevitable oxidation of the ortho-di hydroxyl structures during manipulation. The ester appeared to be immobile on polyamide and charcoal columns using the usual solvents. By using the lead acetate precipitation technique applied in the original isolation of rosmarinic acid from R. officinalis a white crystalline solid was obtained in 0.1$ yield (fresh wt.) from R. officinalis (reported 0.01- 41 0.02$) and in 0.2$ yield from Mentha arvense. The iso• lated products were compared with samples of authentic rosmarinic acid (courtesy of Dr. M.L. Scarpati and Dr. K. Hiller) and were shown to have the same melting point, ultra violet and infrared spectra and chromatographic be• havior. The melting point of 204° reported by Scarpati 41 and Oriente must be incorrect since the sample supplied by Scarpati, a sample supplied by Hiller and both of the present products all melted at 172-174°. Hydrolysis of the isolated ester by NaOH or pectinase yielded caffeic acid (identified by m.p., U.V. spect• rum and chromatographic behavior) and an uncrystallizable

28 compound with ^Xmax ^ ^ (EtOH) and colour reactions of an unconjugated catechol nucleus. This compound, 3,4- -31- dihydroxyphenyllactic acid (DOPL), was very water soluble and unstable to high pH. A reference sample for chroma• tographic purposes was isolated from hydrolyzed rosmarinic acid by column chromatography (Avicel, 2% formic acid) of the ether-extracted hydrolysate.

Biosynthetic studies.

While on the basis of current knowledge of cinnamic acid biochemistry, the caffeic acid moiety was expected to arise from phenylalanine via cinnamic acid, the poss• ible routes to form DOPL were more numerous. From phenyl• alanine a route could be envisioned passing through phenyllactic acid to 3,4-dihydroxyphenyllactic acid, or the corresponding phenylpyruvic acid derivatives. An unlikely but intriguing possibility was the hydration of the double bond of caffeic acid or some other cinnamic acid. More likely were the possibilities arising from tyrosine. The hydroxylation of 4-hydroxyphenyllactic or 4-hydroxyphenylpyruvic acids, followed by reduction in the latter case, would yield DOPL. Finally, the conversion of tyrosine to

DOPA followed by transamination and reduction of the keto group would form the simplest and most easily demonstrated pathway. Any tracer studies involving phenyllactic or phenylpyruvic acids, or the hydroxy analogues, would be made difficult by their interconversions with phenylalan- , 15 ine and tyrosine, respectively In a preliminary experiment, both labelled phenylalanine and labelled tyrosine fed to Mentha arvense for 24 hr. were incorporated into rosmarinic acid. Radioautography of the chromatographed hydrolysis products indicated that phenylalanine was labelling the caffeoyl moiety while tyrosine was labelling the DOPL. This was confirmed in subsequent feedings for 8 hr. (Table II.). The problems inherent in accurately measuring caffeic acid concentra• tions spectrophotometrically (the readily interconvertible

46 x cis and trans isomers have different £ values ) make it difficult to decide if the slight interconversion of phen• ylalanine and tyrosine implied by the label distribution is real. 14 14 Administered C-cinnamic acid and C-p_-coumaric acid were incorporated more efficiently than phenylalanine and labelled only the caffeoyl moiety, as expected. On the 14 other hand, C-DOPA proved to be as good a precursor to rosmarinic acid as tyrosine (Table II.), indicating that the pathway did not directly involve 4-hydroxyphenyllactic or '4-hydroxyphenylpyruvic acids. Direct esterification of caffeic acid with aliphatic 17 alcohols is one known mode of caffeic ester formation Studies on chlorogenic acid biosynthesis have suggested that an alternate route to these esters is via the _- 22 coumaryl quinic ester , and this may be the major route

35-37 in some cases . With DOPA as an intermediate in the formation of DOPL, the involvement of p_-coumaroyl-DOPL is a possibility. Earlier esters are unlikely since the -33-

Table II

Incorporation of Phenylpropanoid Compounds

into Rosmarinic Acid in Mentha arvense L.

Rosmarinic Acid Activity specific $ label in Precursor fed activity caffeic fed (jiC) (;AC/mM) Dilution moiety

14 1.88 1200 96.8$ Phenylalanine-2- C 2

2.5 mC/mM 14 3.67 1850 3.7$ Tyrosine-3- C '6.7 mC/mM 14 3.00 407 102$ Cinnamic acid-2- C 2

1.15 mC/mM

4-OH Cinnamic acid 0.62 126 99$

-2-^C

65.8 ;aC/mM

li( Caffeic acid-2- 'C 1 0.51 236 98$

121 /iC/mM

DOPA-2-^C 3.32 1285

4.07 mC/mM -34- ring hydroxylation evidently precedes the transamination and reduction to yield an esterifiable alpha-hydroxyl. Careful examination of the phenolics extracted from Mentha failed to show any sign of such an ester, which could be expected to show typical p_-coumaric acid fluor• escence (blue when viewed under 366 nm U.V. light with ammonia) and a strong colour reaction between the DOPL 47 14 hydroxyls and diazotized p_-nitroaniline . When C shikimic acid, phenylalanine and tyrosine were fed to Mentha for three hours and the compounds with catechol hydroxylation were isolated and chromatographed for auto• radiography, the results showed that all three precursors labelled rosmarinic acid well but no other phenolics con• tained appreciable label. The feeding time may have been sufficiently long, however, to allow the label to pass through a small pool of p_-coumaroyl ester. One of the main criteria, used in studies on chlorogenic acid formation, for the involvement of a p_-coumaroyl precursor has been the relative inefficiency of adminis• tered caffeic acid as a precursor to the ester. The results have not been consistent, however, and have varied from caffeic acid being incorporated with less dilution than ^coumaric acid to its being incorporated with much greater dilution and possibly even degradation and 35 resynthesis . The picture is complicated by the fre• quent destruction of part of the administered caffeic acid by polyphenol oxidases as it passes through cell -35- membranes or comes in contact with injured tissue such as leaf discs, tuber discs or cut shoots. 14 C-caffeic acid administered to Mentha was incorpor• ated specifically into the caffeoyl moiety of rosmarinic acid with an average dilution of about 240 x (four repli• cates), about twice that of p_-coumaric acid. However, considering the losses encountered in feeding such a labile compound, the observed incorporation suggests that direct esterification of caffeic acid with DOPL is the main route of biosynthesis of rosmarinic acid in Mentha (Figure 4.).

Time-course studies. Recent results showing that chlorogenic acid in Xanthium leaves and SPlanum tubers is actively turning 39.40 overJ/' prompted an examination of the turnover of ros• marinic acid in Mentha. In a closed illuminated system, 50 shoots of M. arvense were exposed for two hours to 14 continuously circulating air containing 150 )xC of CO2 14 (released from Ba CO3 by addition of dilute acid). The specific activity of the ester and of its caffeoyl moiety were checked over 40 hours (Figure 5.). In experiment 1, the label reached the rosmarinic acid more slowly than in experiment 2, which may reflect differences in age or physiological condition of the Mentha at that time. The short decline in specific activity levelled off and little change was seen over 40 hours. The initial peak may have -36-

Figure 4. Proposed pathway for the biogenesis of rosmarinic acid in Mentha. SHIKIMIC ACID PREPHENIC ACID

NH 2 NH 2 COOH

COOH

DOPL

HO-

HO HO V OH Rosmarinic acid

OH -37- been produced by the pulse of label passing through the pools of free precursors such as monosaccharides and ar• omatic amino acids. The label initially incorporated into other compounds would reach these precursors more slowly through metabolic turnover and help to maintain the level of activity in the ester. On the other hand the uptake 14 of COg among the 40-50 shoots in the chamber was un• doubtedly not completely uniform and some specific activ• ity variation from this is unavoidable. The general pict•

ure, then, is one of rapid labelling from C02 followed by relative stability of the amount of label in this pool. The distribution of label between caffeic acid and DOPL varied from about 70% in the caffeoyl moiety in the early part of the experiments, to 52-60$ in the caffeoyl moiety after 40 hr., indicating that the cinnamoyl com• pounds become labelled somewhat faster than the DOPL- DOPA pool. The relatively large dilution of DOPA as a labelled precursor (Table II) suggests that the limiting step may be transamination and/or reduction. 14 Using phenylalanine-U- C as a precursor, there is a definite indication of turnover in the rosmarinic acid pool (Figure 6>.), but it is still not very rapid. The incorporation of label from phenylalanine into the "ethanol-insoluble" ester pool in Mentha was demonstrated (Figure 6b.) with at least as much label in this pool as in the estimated total pool of rosmarinic acid over 2 hr. The relationship between these pools is of great interest -38-

Experiment 1

0 4 8 12 16 20 24 28 32 36 40 44 48 52 Hours Experiment'2 Figure 5. Changes in the specific activity of rosmarinic acid in Mentha after a two hour exposure to 14 CO- specific activity rosmarinic acid •- specific activity caffeoyl moiety A- Hours Hours

Figure 6. Changes in the specific activity of rosmarinic acid and the total activity of insoluble esters in Mentha after a pulse-feeding of phenylalanine-l1^ for (a) 2 hr and (b) 0.5 hr. Specific activity rosmarinic acid Total activity insoluble esters -40-

but was not pursued in this study. Finally, in an attempt to see if rosmarinic acid was being used as a substrate for the formation of insoluble polymers such as lignin, labelled rosmarinic acid was pre• pared biosynthetically. The comparatively low specific 14 activities obtained from C-precursors made it necessary 3 to resort to H-labelling. 3 3 Phenylalanine-G- H (250 ;uC: 0.2 mg) and tyrosine-G- H (250 JAC: 0.2 mg) were administered to separate lots of M. piperita for 2 hr. and the rosmarinic acid isolated and purified. In each case the rosmarinic acid was specif• ically labelled in one moiety or another. Each sample was readministered to Mentha for two feeding periods, (i) the length of time required to take up the solution of rosmar• inic acid, and (ii) 24 hours later (Table III). 3 The label from rosmarinic acid (caffeic- H) remained entirely in the soluble compounds over 24 hours. The specific activity of the isolated rosmarinic acid showed a small decrease, suggesting some turnover in the caffeoyl moiety. The label from rosmarinic acid (DOPL- H) in the soluble compounds, however, showed a marked decrease over 24 hours, as did the specific activity of the isolated ester. An attempt was made to locate the insoluble label with little success. An increase in the labelling of the 30 "insoluble" ester pool^ was evident but was apparently 48 tractoo esmal ofl labeto laccoun couldt bfoe r detectethe decreasd in eth notee Klasod aboven ligni. Nno -41-

Table III.

The Turnover of Labelled Rosmarinic

Acid Readministered to Mentha

Rosmarinic % fed acid amount activity time activity specific Compound fed fed fed ethanol activity fed (mg) (dpm) (hr) soluble (dpm/mM)

RA ~ (caffeic-^H)* 1.5(0.33V 6.8xlo| 1 88 1.5(0.29) 6.8x105 24 83 2.5(0.45) 1.1x10° 2 85 3.28x10^ 2.5(0.52) 1.1x10° 24 82 2.86x10°

RA o (D0PL-JH)** 1.5(0.36) 9.5x10-5 1 82 1.5(0.33) 9.5x105 24 53 2.5(0.42) 1.6x10° 2 75 1.28x10? 2.5(0.45 1.6x10° 24 55 8.30x10°

Rosmarinic acid (caffeic moiety with greater than 96$

of the 3H), sp. act. 73.5yiC/mK

•* Rosmarinic acid (D0PL moiety with greater than 98$ of

the 3H), sp. act. 102 ^uC/mM

t figures in parentheses (mg fed/gm fresh wt. mint) -42-

(or soluble hydrolysate) prepared from the 24 hour rosmar- 3 inic acid (DOPL- H) sample. Since the shoots used were young actively growing material containing lignin (app• roximately \% fresh wt.) it is reasonable to assume that lignin was actually being synthesized during the feeding period. It is possible that the lability of DOPL itself, or of its tritiated protons, at high pH produces large losses in procedures such as "insoluble ester" hydrolysis and thus yields low results. There is, however, no rea• son to believe from these preliminary experiments that either moiety of rosmarinic acid is contributing in a major way to lignin formation as suggested for chloro- 39 genie acid in other plants The interesting questions involving further metabolism of rosmarinic acid and of its DOPL moiety will require much higher specific activities in the ester and the use 14 3 of C to avoid the well-known difficulties with the H labelT9. -JO-

LITERATURE CITED

1. Harborne, J.B., ed. 1964. Biochemistry of Phenolic Compounds. Academic Press, London and New York. 2. Neish, A.C. 1961. Formation of m- and p_-coumaric acids by enzymatic deamination of the correspond• ing isomers of tyrosine, Phytochem. lil. 3. Rosa, N. 1966. Ph.D thesis, Dalhousie Univ., Halifax, N.S. 4. Kovacs, P. and Jindra, A. 1964. Biosynthesis of'al• kaloids. On the transformation of tyrosine to 3,4-dihydroxyphenylalanine in Papaver somnlferum L. plants, Experientia 21:18.

5. MacLeod, N.J. and Pridham, J.B. 1963. Deamination of beta-(3,4-dihydroxyphenyl)-L-alanine by plants, Biochem. J. 88:45p.

6. Karrer, W. 1958. Konstitution und Vorkommen der organischen Pflanzenstoffe. Birkhauser-Verlag, Basel and Stuttgart.

7. Leete, E. 1966. Alkaloid biogenesis, Chap. 17, Bio• genesis of natural compounds (2nd ed.), P. Bern- feld, ed., Pergamon Press, Oxford.

8. Horhammer, L., Wagner, H., and Fritzsche, W. 1964. Zur Biosynthese der Betacyane I, Biochemische Zeitschrift 339:398. 9. Wright, D., Brown, S.A. and Neish, A.C. 1958. Studies of lignin biosynthesis using isotopic carbon. VI. Formation of the side-chain of the phenyl-propane monomer, Can. J. Biochem. Physiol. 36:1087.

10. Koukol, J. and Conn, E.E. I963. The metabolism of aromatic compounds in higher plants. IV. Pur• ification and properties of the phenylalanine deaminase of Hordeum vulgare, J. Biol. Chem. 236:2692.

11. Gamborg, O.L. and Wetter, L.R. I963. An aromatic am• ino acid transaminase from mung bean, Can. J. Bio• chem. Physiol. 41:1733. 12. McCalla, D.R. and Neish, A.C. 1959. Metabolism of phenylpropanoid compounds in Salvia. II. Biosyn• thesis of phenolic cinnamic acids, Can. J. Bio• chem. Physiol. 37:537. _44-

13. Ibrahim, R.K., Lawson, S.G. and Towers, G.H.N. 1961. Formation of labelled sugars from L-tyrosine-l^c in some higher plants, Can. J. Biochem. Physiol. 39:873. 14. Nair, P.M. and Vining, L.C. 19^5• Phenylalanine hy• droxylase from spinach leaves, Phytochem. 4:401.

15. McCalla, D.R. and Neish, A.C. 1959. Metabolism of phenylpropanoid compounds in Salvia. I Bio• synthesis of phenylalanine and tyrosine, Can. J. Biochem. Physiol. 37:531. 16. Runeckles, V.C. and Woolrich, K. 19&3* Tobacco poly• phenols. I. The biosynthesis of O-glucosides and 0-glucose esters of hydroxycinnamic acids, Phytochem. 2:1.

17. Harborne, J.B. and Corner, J.J. I96I. Plant poly• phenols. 4. Hydroxycinnamic acid-sugar derivat• ives, Biochem. J. 81:242.

18. El-Basyouni, S.Z. and Neish, A.C. 1966. Occurrence of metabolically-active bound forms of cinnamic acid and its phenolic derivatives in acetone powders of wheat and barley plants, Phytochem. £:683.

19. Tomaszewski, M. i960. Occurrence of p_-hydroxybenzoic acid and some other phenols in vascular plants, Bull. Acad. Polon. Sci. 8:6l.

20. Sondheimer, E. 1958. On the distribution of caffeic acid and the chlorogenic acid isomers in plants, Arch. Biochem. Biophys. 74:131.

21. Fritig, B. 1968. Ph.D. Thesis, Univ. of Strasbourg, Strasbourg, France.

22. Levy, C.C. and Zucker, M. i960. Cinnamyl and p_- coumaroyl esters as intermediates in the bio• synthesis of chlorogenic acid, J. Biol. Chem. 235:2418.

23. Ribereau-Gayon, P. 1965. Identification d'esters des acides cinnamiques et de l'acide tartarique

dans les limbes et les baies de V. vlnifera. Comptes rendus Acad. Sci. Paris. 260:341.

24. Scarpati, M.L. and Oriente, G. 1958. Chicoric acid (dicaffeyltartaric acid): its isolation from chicory (Chicorium intybus) and synthesis Tetrahedron. 4:43. -45-

25. Scarpati, M.L. and Oriente, G. I960. Isolation from kidney beans (Phaseolus vulgaris) of phaselic acid; its constitution and synthesis. Gazz. chim. ital. 90:212.

26. Bergmann, L., Thies, W. and Erdelsky, K. I965. Das Vorkommen von Glucosaminestern der Hydroxyzim- tsauren in Gew ebekulturen von Nicotiana tabacum. Z. Naturforsch. 206:1297.

27. Harborne, J.B. 1967. Comparative biochemistry of the flavonoids. Academic Press, London and New York. 28. Birkhofer, L., Kaiser, C. and Thomas, V. 1968. Acteosid und Neoacteosid: Zuckerester aus Syringa vulgaris (L.). Z. Naturforsch. 23b:1051. 29. Harborne, J.B. 1966. Caffeic acid ester distrib• ution in higher plants, Z. Naturforsch. 21b: 604. 30. El-Basyouni, S.Z., Neish, A.C. and Towers, G.H.N. 1964. The phenolic acids in wheat. III. In• soluble derivatives of phenolic cinnamic acids as natural intermediates in lignin biosynthesis, Phytochem. 3:627.

31. Scarpati, M.L. and Esposito, P. I963. Neochlorogenic acid and "Band 510" structure, Tet. Lett. #18:1147. 32. Scarpati, M.L. and Guiso, M. 1964. Structure of the three dicaffeoyl-quinic acids of coffee (iso- chlorogenic acid), Tet. Lett. #19:2851.

33. Rohringer, R. and Samborski, D.J. 1967. Aromatic compounds in the host-parasite interaction, Ann. Rev. Phytopath. 5:77. 34. Fritig, B. Personal communication.

35* Runeckles, V.C. 1963. Tobacco polyphenols. II. On the biosynthesis of chlorogenic acid, Can. J. Biochem. Physiol. 41:2249.

36. Steck, W. 1968. Metabolism of cinnamic acid in plants: chlorogenic acid formation, Phytochem. 7:1711.

37. Kojina, M., Minamikawa, T., Hyodo, H. and Uritanl, I. 1969. Incorporation of some possible radioactive intermediates into chlorogenic acid in sliced sweet potato tissue, Plant and Cell Physiol. 10:471. -46-

38. Gamborg, O.L. 1967. Aromatic metabolism in plants. V. The biosynthesis of chlorogenic acid and lignin in potato cell cultures, Can. J. Biochem. 45:1451.

39. Taylor, A.O. and Zucker, M. 1966. Turnover and met• abolism of chlorogenic acid in Xanthium leaves and potato tubers, Plant Physiol^ 41:135Q»

40. Taylor, A.O. I968. The distribution and turnover rate of soluble and insoluble caffeoyl esters in Xanthium, Phytochem. 7s63.

41. Scarpati, M.L. and Oriente, G. 1958. Isolament e constituzione dell' acido rosmarinico (del rosmarinus off.), La Ricerca Scientifica 28:2329.

42. Hiller, K. and Kothe, N. I967. Chlorogen- und Rosmarinsaure-Vorkommen und quantitative Verteilung in Pflanzen der Saniculoideae, Phar- mazie 22:220.

43. Bohm, B.A. 1968. Phenolic compounds in ferns. III. An examination of some ferns for caffeic acid derivatives, Phytochem. 7:1825. 44. Austin, D.J. and Meyers, M.B. 1965. The formation of 7-oxygenated coumarins in hydrangea and lav• ender, Phytochem. 4:245.

45. Kagan, J. I966. The photochemical conversion of caffeic acid to esculetin. A model for the synthesis of coumarins in vivo, J. Am. Chem. Soc. 88:2617.

46. Kahnt, G. 1966. Uber das Gleichgewicht zwischen den Stereoisomeren einiger Zimtsaurederivate in Abhangigkeit von der molaren Konzentration und ihre quantitative spektrophotometrische Messung bei Pflanzenanalysen, Biologisches Zentralblatt. 85:545.

47. Bray, H.G., Thorpe, W.V. and White, K. 1950. The fate of certain organic acids and amides in the rabbit. 10. The application of paperchromatography to metabolic studies of hydroxyacids and amides, Bio• chem. J. 46:271.

^8. Modern methods of Plant Analysis. II. p. 205, 1955, Springer Verlag.

49. Wang, C.H. and Willis, D.L. I965. Radiotracer meth• odology in biological science. Prentice-Hall Inc., New Jersey. Degradation of Aromatic Compounds by Sterile Plant Tissu -47- INTRODUCTION

Aromatic compounds are ubiquitous in higher plants, principally as phenolics. Aside from the usual accumu• lation of soluble phenolic compounds, most plants convert an appreciable part of their carbon pool into the lignin polymers which provide a matrix for secondary cell walls'*'. A second pool of aromatic compounds is the free and protein-bound aromatic amino acids, phenylalanine, tryo- sine and tryptophan. Some plants also accumulate complex nitrogenous aromatics, notably the alkaloids formed from 2 the aromatic amino acids . A great deal of work has been devoted to revealing the pathways of formation of the aromatic amino acids and the 3 4 many classes of phenolics including the alkaloids . All can be said to be formed from the pools of glycolytic metabolites, notably the monosaccharides, pyruvate and acetate, through such well-known routes as the shikimate pathway, acetate (polyketide) condensation and the 3 mevalonate-isoprenoid pathway . Many of the accumulated products have structures which differ greatly from those of the first aromatic Intermediates, so it is not easy to draw a distinction between pathways of metabolism and cat- abolism in this group of compounds. In plants these first intermediates are phenylpyruvic acid, para-hydroxyphenylpyruvic acid and (enroute to tryptophan) anthranilic acid. Aside from its role as an intermediate in tryptophan -48- biosynthesis, the only information on anthranilic acid metabolism in plants is the presence of an enzyme in Tec• oma stans which can convert this compound to catechol, in 5 vitro . In microorganisms the degradation of tryptophan can give rise to anthranilic acid and hence catechol , to 7 8 3-hydroxyanthranilic acid or to kynurenate derivatives . In all three cases, the final steps involve aromatic ring fission and complete oxidation of the fission products to

CC>2. In plants, however, interest in the metabolic fate of tryptophan has been largely restricted to the formation 9 of indoleacetic acid and the numerous alkaloids which are 4 generally regarded as inert metabolic end-products . There is considerable evidence for a wide-spread ability to oxidize IAA10, a reaction which has been suggested as one means of controlling the profound physiological eff• ects of that auxin. The products of this oxidation have not been well characterized, however, and there is little reason to believe that the reaction or reactions involve aromatic ring fission. The further metabolism of phenylpyruvate and 4-hydroxy- phenylpyruvate in plants is probably dominated by their transamination to form phenylalanine and tyrosine respec• tively11. Both of these compounds undergo extensive mod• ifications in their further metabolism, including ring- and side-chain hydroxylations, side-chain cyclization, deamin- 3 ation and beta-oxidation . Microorganisms readily degrade both phenylalanine and -49-

tyrosine, usually via the corresponding keto compound or 12 cinnamate . The side-chain may be degraded to any ex• tent from loss of one carbon to loss of all three, and most microorganisms have the ability to hydroxylate the ring. After introduction of a hydroxyl, either ortho to another hydroxyl (or para to another hydroxyl and ortho to a side-chain), microorganisms can readily oxidize the ring with atmospheric oxygen to produce aliphatic 12 13 acids ' . Few microorganisms seem to require polyphen• ols compounds for any structural or physiological role. These reactions, therefore, must be assumed to be directed toward scavenging the carbon and energy tied up in the considerable quantities of aromatic compounds in the en• vironment, notably in dead plant material. In plants, phenylalanine and tyrosine form the aromatic 4 nuclei of extensive series of alkaloids . Of more inter• est with respect to degradative metabolism, however, are the conversions of phenylalanine, and to some extent tyro-. 3 sine, to numerous plant phenolics . The principal route is through the formation of cinnamic and 4-hydroxycinnamic acid which are both extensively modified by ring hydroxyl- 14 15 16 ation * and methoxylation , beta-oxidation of the 17 side-chain and decarboxylation of the resulting benzoic acids'*"0,'"^ . The latter reactions in particular could be considered degradative, as three of the nine carbon atoms in the original precursor have been made available for further metabolism. The products of phenylpyruvate- -50- phenylalanine metabolism include 2,3- and 3,4-dihydroxy- 20 21 benzoic acid , 3,4-dihydroxycinnamic acid , 2,3-dihydroxy- 22 19 phenylacetic acid and catechol . These compounds bear ortho dihydroxyls and are readily ring-cleaved by micro• organisms. Some or all of them could possibly serve as ring-cleavage substrates in plant cells. Other potential substrates are 2,5-di-hydroxybenzoic and 3,^,5-trihydroxy- benzoic acids. Possible ring-cleavage substrates known to be formed from tyrosine include 3,4-dihydroxyphenylalanine 23 24 (DOPA) , 3,4-dihydroxyphenyllactic acid (DOPL) , dopa- 4 22 mine and 3,b- and 2,5-d.ihydroxyphenylacetic acid

More complex structures such as flavonoids and alka• loids can also bear ortho-dihydroxyls and their degrad• ative metabolism could potentially include ring-fission.

While degradation and reutilization of phenolics in plants is often assumed to occur to some extent, relative• ly little experimental work has been carried out on either general degradation of phenolics or on aromatic ring cleavage in plants. No work has been published on ring- fission by plant tissues growing in the absence of micro• organisms. This condition is essential because of the facility with which bacteria and fungi hydroxylate and degrade aromatic structures, and the generally low levels of such activity one may expect to find in plant tissues. Essentially all published studies on phenolic degradation have used the technique of administering the labelled com• pound in aqueous solution to leaf discs, cut shoots or -51-

25 26 leaf cuttings ' . Some studies have apparently ignored any possibility of microbial metabolism either within the 27 plant tissues or on the surface of the submerged tissue, since no precautions are described to minimize such a 26,28 possibility

The metabolism of tyrosine in plants has attracted the most attention. In contrast to phenylalanine, which pre• dominantly forms phenolics, tyrosine is apparently de• graded quite readily. Label from uniformly labelled ty• rosine quickly appears in monosaccharides, malate and com• pounds derived from these precursors. This phenomenon has 25 29 been examined in a number of studies (Nicotiana , Pyrus , 28 30 Eucalyptus ) and commented on in others . In only one case was any evidence of ring-fission presented, when in two of three cases more than 33$ of the label from tyrosine- 14 U- C was recovered in non-aromatic compounds after 7-23 29 hours of metabolism in the light . This is an extra• ordinary demonstration of aromatic degradation considering the apparent stability of other aromatic compounds (in• cluding 4-hydroxyphenylpropanoids) and the fact that tyro• sine is a protein amino acid. In other studies, the label appearing in non-aromatics could easily have arisen by side-chain degradation without invoking ring-cleavage.

In no case could the participation of microbial metabol• ism be excluded, although the magnitude and specificity of the degradation observed makes it likely that it was, at least in part, a plant tissue phenomenon. Unfortunately, -52- there have been no reported studies on the metabolic route or routes involved. Recent work on the biosynthesis of the betanidin nu• cleus of the betacyanin pigments in the Centrospermae has provided evidence of the ability of plants to oxidize the aromatic ring of DOPA, as previously postulated on the 30 basis of the structure of betanin . It remains to be seen if the non-aromatic ring-fission product is utilized in reactions other than betanidin formation. There has been no report of studies on ring-fission of DOPL or dopa• mine . It has been suggested that 2,3- and 3,4-dihydroxyphenylacetic acids undergo ring-fission in Astilbe and 22 Sinapis . Using uniformly-labelled compounds in a five day feeding experiment, about 5% of the administered label appeared in sucrose. Considering the apparent lack of controls over microbial contamination during this extended period, this result is suggestive, but hardly convincing evidence of ring-fission. The loss of all or part of the labelled side-chain could produce the same result. Sim- 14 ilarly, the results of feeding alpha- C-hydroxycinnamic 14 acids and phenylalanine-U- C to tobacco leaf discs over forty hours show that the side-chains of these compounds 25 are readily degraded and it is reasonable to suspect plantbeta-oxidatios n of the cinnamates, a well-known reaction in 17 The case of 2,5-dihydroxyphenylacetic acid (homo- -53- gentisic acid) may be different. It is a well-documented substrate for ring-cleavage in the degradation of phenylalanine and tyrosine in animals and microorganisms , In plants it has recently been demonstrated to be an intermediate in the formation of plastoquinone from tyrosine^ .

There is little reason to suspect that plastoquinone it• self is a substrate for appreciable ring-fission but it has been suggested that the homogentisic acid intermediate may be an immediate substrate for ring-cleavage2^. Such a reaction could explain the rapid degradation of tyrosine in plants but this interesting possibility has yet to be examined. Preparation of ring-labelled tyrosine and homo- gentisate and the use of sterile plants will be necessary to obtain an unequivocal answer.

Studies on the degradation of flavonoids in plants have been restricted to the catechins and chalcones. C- 14 catechins prepared biosynthetically from COg were read- 14 ministered to tea shoots. The observed release of COg amounted to ca. 10$ of the fed label over the first 25-40 26 hours and a remarkable 80$ after 50-75 hours . If this is any indication of the rate of turnover of the large pool of catechins in tea shoots, they are metabolically 14

extremely active. Even if only the early phase of C02 release is attributable to plant metabolism, the Cg-Cj-Cg structures are obviously being degraded very rapidly to 28 areportet leasdt to anbe dmetabolicall Cg-C-^ componentsy activ. e Catechinin othesr worhavke abees weln l -54-

With no report of the distribution of -^C within the cat• echins, and the very long feeding period, no conclusive case can be made for ring-fission in this experiment.

An interesting report that a specifically labelled

chalcone, 4,2',4',6'-tetrahydroxychalcone-2'-glucosidef was being split by Petunia to form phloroglucinol and 32 para-coumaric acid^ has never been pursued, A-ring- labelled 5,7,4'-trihydroxyflavanone was found to give rise 14 to CO2, presumably after formation of phloroglucinolJ->.

Both aspects of this work need to be confirmed using diff• erent flavonoids and sterile tissues. A number of micro• organisms can break flavonoids down to phloroglucinol and a corresponding phenylpropionate derivative-^, or phloro• glucinol carboxylic acid, carbon monoxide and a CgC-^ 35 acid . The aromatic products can then be further met• abolized, although the mode of cleavage of phloroglucinol has not been clearly demonstrated in any organism. 14

Uniformly-labelled phloroglucinol- C incubated aseptically with nine species of marine algae was partly de• graded to "^CC^ in each case^. The efficiency of this pathway was increased by prior incubation of the algae with cold phloroglucinol. An attempt to detect ring-clea- 14 14 vage of C-catechol and phenylalanine-ring- C by asept- 37 ically cultured tomato seedlings was unsuccessful .

This suggests that catechol is not a widespread substrate for ring cleavage in plants, whereas it is in microorgan• isms . -55-

14 The production of CCv, from toluene and benzene ring- 14 38

C by avocado-" cannot be definitely attributed to plant

metabolism. Exposure of intact, unsterilized, tree-grown

fruit to vapours of the aromatic substrate (100 JJLC) gave 14

rise to only minute amounts of COg over four hours.

Much larger amounts of label were found in compounds other

than benzene and toluene, but they were not identified. 14 14 Benzene- C has been shown to give rise to COg (2-3$ over 72 hours) when fed through the sterilized roots 39

of tea plants, or directly into the stemy/. Considerable quantities of the label appeared in non-aromatic compounds

throughout the plant. Similarly, leaf homogenates (Thea, 40

Vitis) converted benzene to phenol and muconic acid . The oxide in microorganisms involves phenol, catechol and cis, pathway known for the conversion of benzene to carbon di• 41 sen cis-muconic acid ; the same general route has been proposed .40 for the observed degradation of benzene in plants

Since benzene is not a normal metabolite of these plants

the enzymes involved may normally act on other substrates

but possess low substrate specificity. On the other hand,

the sterilization procedures (0.01$ mercuric bichloride or bromine water) may have been inadequate, leaving a micro• bial flora quite capable of carrying out such degradations.

While reference is made to the use of sterilely-grown 39 plants in other cases, no details are given^ . One of the best controlled studies to date involved the 14 administration of phenylalanine-ring-UL- C to cut shoots -56-

of a number of plants and measurement of evolved "^COg^. Antibiotics were used in the feeding solution and attempts were made to reduce the likelihood of microbial metabolism being involved. The results showed that Triticum, Hordeum and Picea could respire as much as 7.3$ of fed -^C. No attempt was made to determine the pathway being followed to 14 ring-fission. The release of CO^ seemed to vary with the state of development of the shoots in the one species ex• amined in detail (Hordeum). While gross contamination was ruled out, and the likeli• hood of the plant tissues themselves carrying out the ring- fission was high, the validity of the results of any ex• periments such as these using non-aseptic tissues and tech• niques will remain in doubt. Two approaches which could eliminate this doubt are the use of aseptically grown plants or tissue cultures. With the recent availability of a number of aromatic ring- labelled compounds, a preliminary study of their possible ring-cleavage and oxidation to COg was undertaken using both approaches. -57-

MATERIALS AND METHODS

Plant tissue cultures. The two tissues used in the present work were derived from Ruta graveolens and Melilotus alba. The cultures were initiated by Dr. O.L. Gamborg, Prairie Regional Laboratory, NRC, Saskatoon and inocula were sent to the University of British Columbia. They were maintained on the standard media^ (Appendix A), B5 (Ruta) and B5C2 (Melilotus), both as callus cultures and as liquid suspension cultures. Callus cultures were grown on 30 ml of the appropriate medium plus 0.6$ agar in 50 ml cotton-stoppered Erlenmeyer flasks. Grown at 23° under continuous fluorescent light• ing (9000 lux) they required subculturing every 2-4 weeks. Liquid suspension cultures were grown in 250 ml cotton- stoppered Erlenmeyer flasks with 50 ml of medium. They were established by inoculation with one or two small pieces of callus tissue. Stocks of liquid suspension cul• tures were maintained on reciprocating and gyrotary shakers but all experiments were run on the latter, which produced better growth. Subculturing was necessary at 1-3 week in• tervals .

The two tissues are very different in morphology. Ruta tissue is uniformly green and in callus often forms a mass of minute shoots. In suspension, the lumps of inocula in• crease in size, producing extensions of tissue in all dir• ections. These usually do not spontaneously break up into -58- smaller pieces. Experiments with Ruta typically ended when the mass of tissue clogged the rotary flow in the feeding flask. The Melilotus, in contrast, was not a green tissue and grew in liquid suspension as single cells and small clumps of cells.

Radioactive compounds. DL-Phenylalanine-ring-U--*-^C, DL-phenylalanine-ring-1- 14 1 4 I LL T LL C, D-glucose-U- C, DL-DOPA-2- C and DL-tyrosine-2-X4,C were purchased from New England Nuclear, Boston, Mass. 14 Cinnamic acid-2- C was purchased from ICN, California. 14 14 14 DL-tyrosine-1- C, DL-tyrosine-3- C, L-tyrosine-U- C, 14 DL-tryptophan-benzene ring-U- C and benzoic acid-ring-U- 14 C were purchased from the Radiochemical Centre, Amersham, 14 England. Salicylic acid-ring-U- C was purchased from Mallinckrodt Nuclear, Orlando, Fla. 14 Cinnamic acid-ring-1- C was prepared from DL-phenyl- 14 alanine-ring-1- C by use of phenylalanine ammonia-lyase. The enzyme was prepared from Ustilago hordel E^ and purif• ied to the end of the first ammonium sulfate fraction• ation^. By using no cold phenylalanine, a large excess of enzyme and overnight incubation at 35°, more than 90$ of the L-isomer was converted to cinnamic acid. The prod• uct was purified by ether extraction from the incubation mixture and TLC of the ether extract. All compounds were made up in aqueous solution and -59- sterilized by Millipore filtration.

Detection of ring-cleavage. A feeding flask consisted of a cotton-stoppered 250 ml Erlenmeyer flask with a center well (Figure 1.). To begin an experiment, a heavy inoculum of tissue and an aliquot of a radioactive compound were added to the 50 ml of med• ium. After 24 hours' incubation on the gyrotary shaker (100 rpm, 22-23°, continuous low light), the center well was charged with 2 ml 5N KOH and a sterile filter paper

wick (4x5 cm, Whatman #50). In early experiments the C02 trapping was carried on continuously, changing the KOH and wick at 24 hr. intervals. This was found to allow app• reciable losses of COg through the stopper, however. Later experiments (most of those reported) were run by trapping COg for 24 hr, sealing the flask with a tight foil cap over the stopper. This was followed by 24 hr in• cubation without trapping. There was no indication that the 24 hr flask sealing depressed growth of the cultures. In the course of an experiment (7-14 days) the Ruta tissue increased in mass about Jx, while the mass of Melilotus increased from 5 to lOx.

The KOH and wick aseptically removed from a feeding flask were placed in a 125 ml side-arm flask which was then connected to the apparatus shown in Figure 1. The COg was regenerated from the carbonate by addition of 2 ml of 50$ lactic acid. The system was then flushed with air (ca./ -60-

Figure 1.

Feeding flask

CO2 Regeneration system -61-

60 ml/min) for 20 min and the CO2 trapped in 10 ml of 46 2-phenylethylamine . After 20 min, 2 ml of water were added to the 2-phenylethylamine to aid in dissolving car• bonate deposits and 2 ml of the final 12 ml volume were counted in duplicate by liquid scintillation counting (to 14

a minimum of 1000 cpm.). Tests with Na2 CO3 in the medium of the feeding flask and 16 hr trapping after acidif• ication showed that the recovery of activity in the com• plete procedure was 70-80$. At the end of a feeding experiment, the residual radio• activity in the medium was determined, after removal of the tissue by filtration.

Microbial contamination. Contamination of feeding experiments, stock cultures or radiochemical solutions was checked by plating the tissue or solution on two media-yeast extract-glucose and malt . o extract (Appendix A). The plates were incubated at 33 o for two days and then 23 for three days. -62-

RESULTS AND DISCUSSION

Aseptically grown plants. A number of attempts were made to grow sunflower seed• lings aseptically and examine their ability to oxidize the ring of aromatic compounds. Using various growth systems and feeding plants phenylalanine-ring-U- C either through the roots or through cut stems, radioactive COg was usually detected. Close examination of the system at the end of each run frequently showed some microbial contamination, leaving the entire series of results in doubt. This approach was, therefore, abandoned in favour of plant tiss• ue cultures.

Plant tissue cultures. Plant tissues grown aseptically in appropriate media can be induced to grow without differentiation. This is frequently accompanied by an altered metabolism, as demon• strated by their requirement for more or less complex nutrient media ' and growth factors .

A two week study with Dr. O.L. Gamborg at the Prairie Regional Laboratory (NRC), Saskatoon made it possible to 48 learn his tissue culture techniques and to attempt some preliminary studies on suspension cultures which he pro• vided. In two experiments, cultures of Triticum, soybean and mung bean could be shown to yield small amounts of 14 14 COg from phenylalanine-ring-U- C. The procedures used -63-

in this work were improved upon and the studies were con•

tinued at UBC using Ruta graveolens and Melilotus alba.

Tryptophan.

DL-Tryptophan-benzene ring-U-^C administered to both 14 species was metabolized to a small extent to CO2 (Table I.). The uptake of this amino acid was consistently poorer than that of other compounds used, but this does not app• ear to be the reason for the one low value for Ruta. The cultures in use at that time were growing more slowly than usual and there may have been different demands on the met• abolic pool of tryptophan, resulting in less degradation of the compound. 14 Because of the low levels of released CGv,, no attempt was made to trace the route of degradation. Phenylalanine.

DL-Phenylalanine-ring-l-x^C was degraded to ^CCv, by both tissues (Table II.). This confirms the ability of plants to oxidize the ring of phenylalanine as suggested 43 by earlier work in a non-sterile system. J If it is assumed that only the L-isomer is actively metabolized and the D-

isomer is stored as an N-acetyl or N-malonyl derivative^9, the percent recovery of activity in CO^ can be doubled. There are a number of routes through which phenylalanine could be degraded. Deamination to cinnamic acid could lead 21 to ortho-dihydroxy compounds such as caffeic acid . Fur- -64-

Table I.

14 14 Oxidation of DL-Tryptophan-benzene ring-U- C to CO by Ruta and Melilotus over seven days.

activity fed % activity % activity

Tissue (JULC) taken up recovered in C02*

Ruta 1 67 0.22

1 70 0.26

4 89 0.03 Melilotus 2 51.5 0.23

* as percent of activity taken up. -65-

Table II.

Metabolism of DL-Phenylalanine-1^ and Cinnamic-1^

by Ruta and Melilotus over seven days t

% activity Compound fed activity % activity recovered fed Tissue (/AC) taken up in C02*

DL-phenylalanine-

ring-l-XMC

3.44 mC/mM Melilotus 2.8 86 0.14 7.5 79 0.14

Ruta 2.5 94.9 0.31 4.2 95.9 0.17 5.0 95.3 0.12 cinnamic acid- 14 ring-1- C 3.44 mC/mM Melilotus 1.2 84 0.04 3.5 71.1 0.39

Ruta 2.1 89.9 N.D.** DL-phenylalanine- 14 2-C 2.5 mC/mM Ruta 2 88 1.1

cinnamic acid- 14 2-C

50 mC/mM Melilotus 1 73.5 3.8***

Ruta 2 88.2 0.71

* as percent of activity taken up. ** not detectable, less than 0.02$. *** incubation for four days. -66-

ther degradation of the cinnamyl side-chain by beta- 17 oxidation would yield acetate and benzoic or hydroxy- 20 benzoic acids . On the other hand, oxidative deamin• ation or transamination of phenylalanine would yield phenylpyruvic acid'*"''". The latter could give rise to 2-hydroxy- and 2,3-dihydroxyphenylacetic acids as has 22 been observed in Astilbe . There is also a possibility of conversion of 2-hydroxyphenylacetic acid to homogentisic acid (2,5-dihydroxyphenylacetic acid). Although this last reaction has not been observed in plants, in an anal- gous reaction salicylic acid is hydroxylated to yield gentisic acid^ .

Cinnamic acid-ring-1- C was degraded to C02 by Melilotus quite well in one experiment and only in traces in a second experiment. The reason for this variation is not known. This tissue definitely has the ability to de• grade cinnamoyl compounds, both ring- and side-chain lab- 14

elled. The greatly increased yield of C02 when using 14 cinnamic acid-2- C is probably indicative of beta-oxid• ation of the side-chain. The same tissue, however, is apparently unable to degrade the benzoic or salicylic acids (Table III.). This suggests that the ring-fission substrate is a C^C^ compound, such as caffeic acid, but more work is required to confirm this. The one attempt to detect ring cleavage of cinnamic acid by Ruta was unsuccessful. While this requires con• firmation, it may mean that phenylalanine is being de- -67-

Table III.

Oxidation of Benzoic acid and Salicylic acid ring-U-^C by Ruta and Melilotus over seven days.

activity % activity Compound fed % activity recovered

fed Tissue {jxC) taken up in C02*

Benzoic acid- 14 ring-U- C 45 mC/mM Ruta 5 96.7 N.D.* *

Ruta 5 96.. 5 N.D.

Melilotus 3 65.9 N.D.

Salicylic acid- 14 ring-U- C 5 82.1 N.D. 0.95 mC/mM Ruta 3 88.2 N.D. Melilotus

* as percent of activity taken up.

»* not detectable, less than 0.02$. -68- graded through the phenylpyruvic acid-hydroxyphenylacetic acids route in Ruta. In support of this, the release of 14 14 COg from cinnamic acid-2- C was strikingly lower in Ruta than in Melilotus. No degradation of the ring of benzoic or salicylic acids could be detected (Table III.). The cultures of Ruta in use have an active cinnamate metabolism since they accumulate coumarins and rutin^x. There is no sign of an accumulation of cinnamic or benzoic acids, however, when hydrolyzed extracts of the tissue are chromatographed. The same is true of the Melilotus cul• tures, which also lack the coumarins and rutin. It would be interesting to extend this work to tissues which accum• ulate the pools of hydroxycinnamic and hydroxybenzoic acids more typical of higher plants.

Tyrosine. Because of the great interest in tyrosine degradation in plants, an attempt was made to measure ring-cleavage of 14 this amino acid. Ring-labelled tyrosine- C is unavailable 14 at present, but comparison of the extent of COg release 14 from cultures fed tyrosine-1,2,3-(side-chain) C and cul- 14 tures fed tyrosine-U- C (same total and specific activity) might be expected to demonstrate ring-cleavage indirectly. The inevitable experimental variation makes traces of ring-cleavage undetectable, but earlier work in non-sterile systems suggested that ring-cleavage of tyrosine in plants 29 may be quite extensive . -69-

Using Ruta, a 1:3 ratio of activity from tyrosine-U- 14 14 C and tyrosine-1,2,3- C was not exceeded, despite the remarkably high rate of degradation of the side-chain

(Table IV.). Similarly, the slight increase in this ratio using Melilotus may not be significant. The rate of side- chain degradation with Melilotus was lower but still con• siderable. Ring-cleavage of tyrosine does not appear to be occurring to any great extent in these tissues. Phenyl• alanine is not being converted to tyrosine very readily, 14 as is shown by the comparatively low rate of CO^ rel- 14 ease from phenylalanine-2- C (Table II.). To obtain some indication of the extent of degradation 14 of the side-chain of tyrosine, tyrosine-2- C and tyro- 14 sine-3- C were compared as substrates. As can be seen from Table IV., the degradation of the side-chain extends

to all three carbons. The second carbon is lost most readily, while carbon three is lost to a smaller extent

(approx. 60% of carbon two). In experiments run beyond 14 seven days the ratio of C0£ recovery from tyrosine-3- 14 14 C vs. tyrosine-2- C steadily increases. This suggests 14 that the release of C from both positions might occur in

the initial reaction, but the subsequent metabolic fate of the two carbons differs. On the other hand, the two carbons may be lost in a stepwise fashion in distinct reactions. Tyrosine is known to be a precursor to the plasto- 32 quinones . This route involves 2,5-dihydroxyphenyl- -70-

Table IV.

14 14 Degradation of Tyrosine- C and DOPA- C by Ruta and

Melilotus over seven days.

activity % activity Compound fed % activity recovered fed Tissue (yuC) taken up in COg*

DL-tyrosine- 14 1,2.3- C 5.14 mC/mM Ruta 2.16 n.m.#* 19.69 L-tyrosine- 14 U- C

5.14 mC/mM Ruta 2.0 n.m. 6.65

DL-tyrosine- 14

1,2,3- C 5.14 mC/mM Ruta 1 93.9 19.20 L-tyrosine- U-^C 5.14 mC/mM Ruta 1 92.3 4.35

DL-tyrosine- l^^-1^ Melilotus 1 81 7.07

L-tyrosine-

U-^C Melilotus 1 83 2.58 -71-

Table IV. (cont'd.)

activity % activity Compound fed % activity recovered

fed Tissue (p.c) taken up in C02*

DL-tyrosine- 14 2-C 50 mC/mM Ruta 2 n.m. 19.50

DL-DOPA-

2-^c

4.1 mC/mM Ruta 2 n.m. 0.50

DL-tyrosine-

2-^C 50 mC/mM Ruta 1 n.m. 21.00 DL-tyrosine-

3-^c

6.85 mC/mM Ruta 1 n.m. 10.30

DL-tyrosine-

2-^C

0.90 mC/mM Ruta 1 93.9 17.35 DL-tyrosine-

3-l4C 0.90 mC/mM Ruta 1 94.8 12.30

* as percent of activity taken up.

** n.m. not measured: activity in C02 is % activity fed. -72- acetic acid (homogentisic acid) and is probably common to all photosynthetic higher plants. Tyrosine also gives rise to 3»k-dihydroxyphenylacetic acid in Astilbe but the extent of this reaction in the plant kingdom is not known. In a number of plants, tyrosine is hydroxylated 23 to form 3>4-dihydroxyphenylalanine (DOPA) . iii 14 Ruta was incubated with D0PA-2- C and the C02 14 release compared with that from tyrosine-2- C (Table IV.) Although some of the DOPA may be lost by oxidative poly• merization within the cells, it seems unlikely that ty• rosine is being degraded via its hydroxy derivative in this case. The possibility of 2,5-dihydroxyphenylacetic acid (DPA) involvement in the tyrosine side-chain degradation was checked in two ways. (i) 25 yuM of DPA was added to one of two cultures of 14 Melilotus metabolizing tyrosine-2- C. Over the subse- 14

quent 24 hr incubation, a 30% decrease in the C02 out• put by the treated culture (relative to the control) was observed. The same amount of DPA added to one of two 14 cultures metabolizing glucose-U- C had no effect on 14

C02 release. (ii) A heavy inoculum of Melilotus tissue was incubated 14 for 24 hr with 2 ^uC DL-tyrosine-2- C and 25 jm DPA. The tissue was then extracted with 95$ ethanol (10 dpm eth• anol soluble). This extract was taken up in hot water, filtered, hydrolyzed with acid and extracted with ether -73-

for 48 hours. The ether extract was chromatographed on Avicel TLC plates in the following solvent systems:

(1) benzene:acetic acid:water (10:7:3 org. phase) 2x dev. (2) methyl ethyl ketone:acetone:formic acid:water (80:4:2:12) (3) 2% formic acid The DPA band after this purification procedure still con• tained at least 20,000 dpm, despite considerable losses of the compound by oxidation. These experiments indicate that tyrosine is forming 2,5-dihydroxyphenylacetic acid in a plant tissue that is presently non-photosynthetic and presumably has no require• ment for plastoquinone synthesis. It is tempting to sugg• est that the formation of DPA normally serves as a route to the plastoquinones as well as for the degradation of tyrosine. The control over the extent of tyrosine con• version to DPA may be chloroplast- and light-dependent, as 25 29 earlier studies on tyrosine degradation have suggested The Ruta used in the present studies is a green tissue

with structurally normal chloroplasts^2, and it shows a notably greater tendency to degrade tyrosine than does Melilotus (Table IV.). While these experiments are preliminary in nature, the ability of plant tissues to carry out aromatic ring- fission seems established. Since both phenylalanine- ring-l-^C and tryptophan-benzene ring-U-^^C are prepared by specific chemical syntheses, there is no reason to sus- -74- pect substantial labelling of these compounds in non- aromatic positions. The possibility exists that a vola• tile aromatic acid is being synthesized from phenylalan• ine and tryptophan in these tissue cultures. The low vapor pressures of the known aromatic acids, however, and the two step CO2 isolation procedure make it very unlikely that the activity recovered in the 2-phenylethylamine is due to compounds other than CO2. In at least one case, the aromatic ring of cinnamic acid can also be degraded, an important point in plant phenolic metabolism. Whether or not the low percentage of ring-cleavage observed in the sterile cultures is typical of the true level in intact plants can only be ascertained by use of sterilely cultured plants and sterile procedures. It is possible that a re• duced emphasis on "secondary metabolism" in the cultures reduces the observed ring-fission. Definitive studies on tyrosine degradation await the availability of ring-^C- tyrosine. -75-

LITERATURE CITED

1. Brown, S.A. 1964. Lignin and tannin biosynthesis, Biochemistry of Phenolic Compounds, p. 361, J.B. Harborne, ed., Academic Press, London and New York.

2. The Alkaloids, Chemistry and Physiology. 1964.' Manske, R.H.F., ed., Academic Press, London and New York.

3. Neish, A.C. 1964. Major pathways of biosynthesis of phenols, Biochemistry of Phenolic Compounds, p. 295, J.B. Harborne, ed., Academic Press, London and New York.

4. Leete, E. 1967. Alkaloid biogenesis, Biogenesis of Natural Compounds (2nd Ed.). P. Bernfeld, ed., P« 953, Pergamon Press, Oxford. 5. Nair, P.M. and Vaidyanathan, C.S. 1966. Conversion of isophenoxazine to catechol in Tecoma stans, Arch. Biochem. Biophys. 115;515« 6. Subba Rao, P.V., Moore, K. and Towers, G.H.N. I967. The conversion of tryptophan to 2,3-dihydroxy• benzoic acid and catechol by Aspergillus niger, Biochem. Biophys. Res. Comm. 28:1008~

7. Jakoby, W.B. and Bonner, D.M. 1953. Kynureninase from Neurospora: purification and properties, J. Biol. Chem. 205:699. 8. Taniuchi, H. and Hayaishi, 0. 1963. Studies on the metabolism of kynurenic acid. III. Enzymatic formation of 7,8-dihydroxykynurenic acid from kynurenic acid. J. Biol. Chem. 238:283.

9. Moore, T.C. and Shaner, C.A. I967. Biosynthesis of indoleacetic acid from tryptophan-l4c in cell- free extracts of pea shoot tips, Plant Physiol. 42:1787. 10. Galston, A.W., Bonner, J. and Baker, R.S. 1953. Flavoprotein and peroxidase as components of the indoleacetic acid oxidase system of peas, Arch. Biochem. Biophys. 42:456.

11. Gamborg, O.L. and Wetter, L.R. 19^3. An aromatic amino acid transaminase from mung bean, Can. J. Biochem. Physiol. 4l:1733. -76-

12. Towers, G.H.N, and Subba Rao, P.V. 1969. Degrad• ative metabolism of phenylalanine, tyrosine and DOPA, in press 9th Annual Symposium of Phyto- chemical Society of North America, Banff, Alta.

13. Cain, R.B. Bilton, R.F. and Darrah, J.A. I968. The metabolism of aromatic compounds by microorgan• isms. Metabolic pathways in the fungi, Biochem. J. 108:797. 14. Nair, P.M. and Vining, L.C. 1965. Phenylalanine hydroxylase from spinach leaves, Phytochem. 4:401.

15. Vaughan, P.F.T. and Butt, V.S. 1969. The hydroxyl• ation of p_-coumaric acid by an enzyme from leaves of spinach beet (Beta vulgaris L.), Biochem. J. 113:109. 16. Byerrum, R.U., Flokstra, H.H., Dewey, L.J. and Ball, CD. 195^. Incorporation of formate and the methyl group of methionine into methoxyl groups of lignin, J. Biol. Chem. 210:633. 17. Vollmer, K.O., Reisener, H.J. and Grisebach, H. 1965. The formation of acetic acid from p_- hydroxycinnamic acid during its degradation to 2-hydroxybenzoic acid in wheat shoots, Biochem. Biophys. Res. Comm. 21:221.

18. Zenk, M.H. 1964. Einbau von p_-hydroxybenzoesaure in die Hydrochinonkomponente des Arbutins in Bergenia crassifolia, Z. Naturforsch. 196:856.

19. Ellis, B.E. and Towers, G.H.N. 1969. The biogenesis of catechol in Gaultheria, Phytochem. 8:l4l5. 20. Vollmer, K.0. and Grisebach, H. 1966. Zur biosynthese der Benzoesauren in Gaultheria procumbens. III. Z. Naturforsch. 216:435.

21. McCalla, D.R. and Neish, A.C. 1959. Metabolism of phenylpropanoid compounds in Salvia. II. Bio• synthesis of phenolic cinnamic acids, Can. J. Biochem. Physiol. 37:537. 22. Kindl, H. 1969. Biosynthesis and metabolism of hydroxyphenylacetic acids in higher plants, European J. Biochem. 7:340. -77-

23. Kovacs, P. and Jindra, A. 1964. Biosynthesis of alkaloids. On the transformation of tyrosine to 3,'4-dihydroxyphenylalanine in Pa paver somnif- erum L. plants, Experientia 21:18. 24. Ellis, B.E. and Towers, G.H.N. 1969. The biogenesis of rosmarinic acid in Mentha, in prep. 25. Runeckles, V.C. I963. Formation of sugars from phenylpropanoid compounds in tobacco leaf discs, Can. J. Botany 41:823.

26. Zaprometov, M.N. 1959. On the ability of higher plants to cleave the benzene ring. Deep oxid• ation of l^C-catechins in tea, Doklady Academii Nauk SSSR 125:1359. 27. Henry, E.W., Valdovinos, J.G. and Jensen, T.E. 1968. Invasion of plant tissue by bacteria under in vitro conditions, Plant. Physiol. 43:1730.

28. Hillis, W.E. and Isoi, K. 1965. The biosynthesis of polyphenols in Eucalyptus species, Phytochem. 4:905. 29. Ibrahim, R.K., Lawson, S.G. and Towers, G.H.N. 1961. Formation of labelled sugars from L-tyrosine- l^C in some higher plants, Can. J. Biochem. Physiol. 39:873. 30. Dougall, D.K. and Shimbayashi, K. i960. Factors affecting growth of tobacco callus tissue and its incorporation of tyrosine, Plant Physiol. 25:396.

31. Miller, H.E., Rosier, H., Wohlpart, A., Wyler, H., Wilcox, M.E., Frohofer, H., Mabry, T.J. and Dreiding, A.S. 1968. Biogenese der Betalaine. Biotransformation von DOPA und Tyrosin in den Betalaminsaureteil des Betanins, Helv. Chim. Acta ^51:1470.

32. Whistance, G.R. and Threlfall, D.R. 1968. Biosyn• thesis of phytoquinones. Biosynthetic origins of the nuclei and satellite methyl groups of plastoquinone, tocopherols, and tocopherol- quinones in maize shoots, bean shoots, and ivy leaves, Biochem. J. 109:577. -78-

33. Patschke, V.L., Hess, D. and Grisebach, H. 1964. Uber den A'b'bau von 4, 2',4*, 6'-Tetrahydroxy- chalkon-2-glucosid und 4,2',4'-Trihydroxy- chalkon-4-glucosid in Rotkohlkeimlingen und Petunien, Z. Naturforsch. 19°:1114. 34. Patschke, U.L., Barz, W. and Grisebach, H. 1964. Uber den Einbau von 5.7.4'-Trihydroxyflavanon- 2,6,8,10-l4c in Cyanidin und die Isoflavone Biochanin-A und Formononetin, Z. Naturforsch. 19bs1110.

35. Jayasankar, N.P., Bandoni, R.J. and Towers, G.H.N. 1969. Fungal degradation of phloridzin, Phyto• chem. 8;379.

36. Westlake, D.W.S. and Spencer, J.F.T. 1966. The utilization of flavonoid compounds by yeasts and yeast-like fungi, Can. J. Microbiol. 12;165• 37. Craigie, J.S., McLachlan, J. and Towers, G.H.N. 1965. A note on the fission of an aromatic ring by algae, Can. J, Botany 43:1589. 38. Garay, A.S. and Towers, G.H.N. 1965. On the ability of tomato seedlings to cleave the benzene ring, unpublished results. 39. Jansen, E.F. and Olson, A.C. 1969.Metabolism of carbon-l4-labeled benzene and toluene in avocado fruit, Plant Physiol. 44:766. 40. Durmishidze, S.V. and Ugrekhelidze, D. Sh. 1969. Splitting of phenol by the tea plant, Doklady Akademii Nauk SSSR 184:228. 41. Durmishidze, S.V., Ugrekhelidze, D. Sh., Dzhikia, A.N. and Tsevelidze, D. Sh. 1969. Intermediate products of the fermentative oxidation of benzene and phenol, Doklady Akademii Nauk SSSR 184:466. 42. Evans, W.C., Smith, B.S.W., Linstead, R.P. and Elvidge, J.A. 1951- Chemistry of the oxidative metabolism of certain aromatic compounds by micro-organisms, Nature 168:772.

43. Rosa, N. 1966. Ph.D. thesis, Dalhousie University, Halifax, N.S. 44. Gamborg, O.L., Miller, R.A. and Ojima, K. 1968. Nutrient requirements of suspension cultures of soybean root cells, Exp. Cell Research 50:151. -79-

45. Towers, G.H.N, and Subba Rao, P.V. I969. Phenyl• alanine ammonia-lyase (Ustilago hordei) in press Methods in Enzymology.

46. Wang, C.H. and Willis, D.L. 1965. Radiotracer methodology in biological science, p. 168, Prentice-Hall, Inc^., New Jersey.

47. Staba, E.J. 1969. Plant tissue culture as a tech• nique for the phytochemist. Recent Advances in Phytochemistry vol. II. Seikel, M.K. and Runeckles, V.C. ed, Appleton-Century-Crofts, New York.

48. Gamborg, O.L. and Eveleigh, D.E. 1968. Culture methods and detection of glucanases in suspen• sion cultures of wheat and barley, Can. J. Biochem. 46:417. 49. Rosa, N. and Neish, A.C, 1968. Formation and occurrence of N-malonylphenylalanine and rel• ated compounds in plants, Can, J. Biochem. 46:797. 50. el-Basyouni, S.Z., Chen, D., Ibrahim, R.K., Neish, A.C. and Towers, G.H.N. 1964. The biosynthesis of hydroxybenzoic acids in higher plants, Phytochem. 3:485.

51. Gamborg, O.L. Personal communication.

52. McBride, D.L. I969. Unpublished observations. -80- Appendix "A"

(i) Tissue culture media B5 medium*

per liter per liter glass-distilled glass-distilled Compound water Compound water

150 mg Sucrose 20 gm NaH2P0k.H20 KNO^ 2500 mg 2,4-D*** 1 mg

(NHj^gSO^ 13k mg Vitamin solution 10 ml

MgS0k.?H20 250 mg Micronutrients

CaCl2.2H20 150 mg solution 1 ml

Fe(EDTA)** 28 mg PH 5.5

KI 0.75 mg

Add 10$ excess water to compensate for losses in auto•

claving and incubation.

* B5C2 medium consists of B5 medium plus 2 gm/liter N-Z. Amine Type A casein hydrolysate, Sheffield Chem• ical, Norwich, N.Y. (B5C2 medium thus contains 102 mg phenylalanine /.L, 62 mg tyrosine /L and 28 mg try• ptophan /.L. )

Sequestrene 330 Fe, Geigi Agricultural Chemicals, Ardsley, N.Y.

*** 2,4-dichlorophenoxyacetic acid dissolved in 95$ ethanol (1 mg/ml).

£ Stock solution. Dissolved in 100 ml distilled water:

1000 mg MnS04-H20, 300 mg H3BO3, 300 mg ZiiSOk. 7H2O, 25

mg Na2Mo0k.2H20, 25 mg CuSOk, 25 mg CQC12.6H20 stored frozen.

t Stock solution. Dissolved in 100 ml distilled water: 10 mg nicotinic acid, 100 mg thiamine HC1, 10 mg pyr• idoxine HC1 and 1000 mg myo-inositol stored frozen in plastic bottle. -81-

(ii) Plating media for contamination monitoring. (a) Difco malt extract agar 4.5$

(b) Difco yeast extract 1$ Glucose 1$ agar 1.5$