Biosynthesis and Translocation of Secondary Metabolite Glycosides In

."t c,ÞrûFU$ Qq*i'. {ù. "gÐ"/ ;"$ 21 0cI 1996 a Biosynthesis and translocation of secondary ^"fu¡rr"'%Sry"V glycosides in the grapevin e Vitis viniÍera¿- #

by

Mansour Gholami (M.Sc. Horticulture, Tehran)

Department of Horticulture, Viticulture and Oenologt

l'[/ aite Agriculture Re s earch Institute The University of Adelaide South Australia

Thesis Submitted for the Degree of Doctor of Philosophy

ln The University of Adelaide Facttlty of Agriculture and Natural Resources Science

March,1996 Table of contents

Summary...- I

Declaration.. -l.Y

Acknowledgements.. v

List of tables ...1{l

List of figures. .Yrii

List of publications. ..xu

Abbreviations ..EU

Chapter l. I

Chapter 2. J

Chapter 3 24

Chapter 4 37

Chapter 5 55

Chapter 6 67

Chapter 7 99

Chapter I lt7

References t2t

Appendix. 147

Publications 153

Summary

In this study different experiments were conducted to investigate the site of biosynthesis of flavour compounds in the grapevine. Most of the secondary metabolites, including flavour compounds, are glycosylated and stored in plant tissues as glycosides. The chemical properties of these compounds, especially their water solubility, suggests that glycosides might be possible forms of translocated secondary metabolites in plants. The levels of total glycosides in grapevine fruits and leaves was determined. Whole berries, leaf blades and petioles of three grape cultivars were sampled and extracted throughout the growing period. The glycosidically bound fraction was isolated by retention on Cl8 reverse-phase silica gel, and after hydrolysis with sulfuric acid at 100 'C, the released glycosyl-glucose (G-G) was enzymatically determined. There were large quantities of glycosidically bound secondary metabolites in the extracts of leaf blades but much lower levels of glycosides were detected in petioles and berries. Seasonal changes in the amounts of glycosides, and also varietal differences between cultivars were observed. The glycoside levels increased on a per berry basis but decreased on a fresh weight basis during berry development. Muscat Gordo, as a floral grape, and Sultana, as a neutral grape, had different patterns of total glycoside accumulation during ripening. The effect of leaf genotype on the berry flavou¡ constituent profile and levels was investigated b1, grafting bunches between floral and non-floral grape cultivars. Inflorescences of Muscat Gordo and Shiraz grapevines grown in a glasshouse were grafted onto shoots of each other before flowering, and grown until ripe. Monoterpene glycosides were isolated fronr the fruits, enzvmatically hydrolysed and the released monoterpene aglycones liquid-liquid extracted and analysed by GC-MS. Muscat Gordo fruit from both ungrafted control vines and from fruit grafted onto Shiraz vines yielded fruit with monoterpenes at levels and types typical of this floral grape variety. Control Shiraz fruit (ungrafted) and Shiraz fìuit grafted onto Muscat Gordo vines contained only low levels of monoterpene glycosides. Similar results were obtained in a field experiment with Muscat Gordo and Sultana fruits similarly grafted, The lack of difference between flavour compounds in grafted and non-grafted f¡uit suggests that berries have the ability to synthesise aroma compounds independently from the rest of the vine. Rose coloration of skin was seen to develop late during ripening on the normally white grape berries of cv. Muscat Gordo Blanco. The nature of the pigment was investigated by HPLC analysis of skin extracts of single ber¡ies. The predominant anthocyanin was identified as cyanidin-3-glucoside with minor amounts of delphinidin- and peonidin-3- ii glucosides. This composition resembles the skin composition of coloured, Muscat a petits grains cultivars which it also resembles by the monoterpene composition of the juice. The pigments occurred only in berries with levels of total soluble solids in excess of 24"Brix in the juice and such berries tended to have smaller fresh weight. Berry pigmentation occurred on vines with various root systems. The specific conditions under which pigment developed in Muscat Gordo berries may offer a useful tool in the study of anthocyanin biosynthesis. As grafting bunches of Muscat Gordo onto Shiraz and also grafting of Shiraz onto Muscat Gordo did not effect the anthocyanin profile of the berries, it appears that anthocyanins are another group of secondary metabolites synthesised in the berries independent of the leaves. Photosynthetically fixed assimilates translocate from source leaves to sinks via phloem tissue in vascular bundles. The phloem sap composition was analysed for three grapevine cultivars. Phloem sap was collected from the cut end of grape bunch laterals using an EDTA-facilitated exudation technique. Samples of phloem sap exudates were collected at different berry developmental stages and also continuously over a few days and nights at different periods of berry ripening. Sucrose, amino acids and mineral ions were analysed in these phloem exudates. The sucrose concentration in the phloem exudates collected from different cultivars was found to be high while glucose and fructose were low. In addition to sucrose, substantial levels of amino acids and potassium were also detected in the exudates. Glutamine was the principal amino acid in the phloem exudates of grapevines followed by glutamic and aspartic acids , and alanine. On'rission of EDTA from the buffer solution or girdling of canes either side of the bunch greatly decreased the levels of all these sugars and amino acids in the exudate. These controls, together with the observed composition of the collected material, suggested that the exudates were predominantl;- derived from phloem sap entering the berries in the laterals. The metabolites exuded from the phloem of fruit bunch stem showed seasonal and diurnal variations. The diurnal pattern of sucrose exudation showed high levels of sucrose exuded at night. Total amino acids and some individual amino acids also showed diurnal variation. This is the first time that the diurnal pattern of phloem sap flux in grapevines has been reported. Using a similar EDTA-lacilitated collection technique, the phloem exudates of grapevine leaves were collected from the cut end of petioles of excised leaves. Sucrose, potassium and amino acids were present in the samples collected from leaves. The results indicate that sucrose is the main sugar translocating in grapevine phloem, and that amino acids and potassium are also major constituents of the phloem sap . iii

To investigate the site of biosynthesis of flavour compounds in grape berries, the glycosidically bound fraction in phloem sap collected from fruits and leaves of three cultivars was qualitatively and quantitatively analysed by GC-MS. Total glycoside levels in the collected solution were quantified by determination of G-G. The aglycones identified were mostly monoterpenes, C13-norisoprenoids, and benzene derivatives. There were varietal differences in the GC-MS profile of compounds exuded into the buffer solution. Glycosylated secondary metabolite concentrations were not decreased in exudates from bunches after girdling of the canes, indicating that the compounds being analysed may not have been derived from phloem sap. Analysis of grape bunch peduncle tissue showed that some of most abundant glycosylated compounds in the exudates, such as those of geraniol, were also present in the rachis tissue in high concentrations. It seems that these compounds reside in the tissues surrounding the phloem tissue and, due to their solubility, they may easily leak out and enter the buffer solution used for sap collection. The experimental fîndings in this thesis lead to the interpretation that, wilh regard to secondary metabolite production, grape berries are metabolically independent of leaves and other parts of the plant and all secondary metabolic pathways are a part of the berry's genetic identity. Although dependent on the rest of the vine for supply of precursor compounds and primary metabolites, it appears that the berr), itself is the site of synthesis of secondary metabolites including anthocyanins and flavour compounds. IV

I)eclaration

I hereby declare that this work conducted in the Department of Horticulture, Viticulture and Oenology, within the University of Adelaide and contains no material which has been previously accepted for the award of any other degree or diploma at any University. To the best of my knowledge this thesis contains no material previously published or written by another person except where due references is made in the text.

I confent to this thesis being made available for loan and photocopying, when deposited in the university library.

Mansour Gholami V

Acknowledgement

My sincere appreciation goes to my supervisors, Dr Patrick Williams, Dr Simon

Robinson, and Dr John Jackson for their excellent advice, patience and encouragement throughout this project and during the preparation of the thesis. I would like to express my profound gratitude to Dr Bryan Coombe for his constant guidance, constructive criticism and assistance during the investigation and in preparation of this manuscript. I wish to acknowledge gratefully the support of the University of Bu-Ali Sina,

Hamadan, Islamic Republic of Iran fo¡ awarding a scholarship which enabled me to persue my studies at the University of Adelaide, Australia.

Special thanks are due to Mr Yoji Hayasaka for GC-MS analysis. Dr Wies Cynkar, Ms

Mariola Kwiatkowski, and Mr Ken Pocock of Australian Wine Resea¡ch Institute for their technical support for the chemical analyses.

I would like to thanii the Cooperative Research Centre for Viticulture for their financial

support all those people associated within CRCV Program 3 for their friendship.

Finally, I wish to express my appreciation to my wife and children for their patience and

encouragement throughout my studies. VI

List of tables

Table 2.1 Translocation of photosynthetic products from leavesato immature benies (from Peynaud and Ribéreau-Gayon, 197 I).

Table 4.1. Concentrations of monoterpene compounds identified in ripe grafted and non- grafted berries from the glasshouse experiments. Peak numbers are shown in Fig. 4.4.

Table 5.1 Total polyphenolics and anthocyanins in the skin of Muscat Gordo Blanco in two groups: white (non-pigmented) and pigmented (showing a light rose or pink colour). The fruits were from vines grown in the field.

Table 5.2 Concentrations of monoterpene compounds in whole non-pigmented (White) and rose- or pink-coloured (Colour) Muscat Gordo berries and white juice. Compounds are numbered as in Figure 4.3 of chapter four.

Table 6.1 Sugars, amino acids and potassium measured in phloem sap exudate of three grapevine cultivars during the growing season (p mol / 1.5 mL of buffer solution).

Table 6.2 The seasonal changes in phloem exudate amino acids of Muscat Gordo, Sultana and Shiraz. Values iven are Yo of total amino acids before veraison (B), at veraison (V). and at ripening time (R).

Table 6.3 Comparison of sugars exuded from fruit bunches on the vine with excised bunchstems as in Fig. 6.3, (cv. Muscat Gordo at pre veraison, "Brix : 4'5).

Tabte 6.4 The effect of exclusion of EDTA from the buffer solution used for sap collection on the levels of different metabolites in collected sap samples (cv. Muscat Gordo at veraison, "Brix :6.4).

TTable 6.5 The effect of girdling the phloem tissue above and below the bunch on the levels of different metabolites in collected sap samples (cv. Muscat Gordo at ripening. "Brix : 10.2).

Table 6.6 The effect of girdling and exclusion of EDTA on the phloem sap exudation (cv Muscat Gordo at ripening, 'Brix = 15.8). v'ii

Table 6.7 The effect of exclusion of EDTA on the phloem sap exudation (cv. Sultana at ripening 'Brix: 18.0).

Table 6.8 The effect of exclusion of EDTA on the phloem sap exudation (cv. Sultana, ripe 'Brix = 23.8).

Table 6.9 The levels of sucrose, glucose, fructose, potassium and total amino acids of leaf phloem sap exudated under light and dark, collected at ripening period.

Table 7.1 Glycosyl glucose levels measured in phloem sap exuded into buffer solution from peduncle laterals and leaves, and also in berry homogenates of Muscat Gordo, Sultana and Shiraz at three berry development stages.

Table 7.2 The aglycones determined in hydrolysates of phloem exudate collections and tissue extracts

Appendix Table 1. The seasonal changes in phloem sap amino acids of Muscat Gordo, Sultana and Shiraz. Values given areo/o of total amino acids before veraison (B), at veraison (V), and at ripening time (R).

Appendix Table 2. The differences in amino acid profiles of leaf exudates in light and dark collected at berrv ripening time (% of total).

Appendix Table 3. The seasonal changes in amino acid profiles (Yo of total) of Muscat Gordo peduncle exudates.

Appendix Table 4. The seasonal changes in amino acid profiles (% of total) of Sultana peduncle exudates.

Appendix Table 5. The seasonal changes in amino acid prohles (% of total) of Shiraz

peduncle exudates. viii

List of figures

(2) Figure 2.1 Water and dry matter (1) and type of changes in the amounts of the sugils in grape berries during berry development. (From coombe, 1992).

Figure 2.2 Concept of synthesis and hydrolysis of secondary metabolite glycosides in plants (from Hosel, 1981)'

Figure 2.3 Types of secondary metabolite glycoconjugates that have been identified as flavour precursors in fruits (from Williams 1993).

Figure 3.1 G-G assay steps.

Figure 3.2.1 to 6 Evolution of G-G in leaf blades ( I ), leaf petiole ( 2 ) and fruit ( 3 ) and changes in berry volume ( 4 ), berry density ( 5 ), and fruit soluble solids ( 6 ) in Muscar Gordo with time in weeks after flowering. G-G data in Fig. ( 3 ), are expressed as concentration in macerated whole bunch on a fresh weight basis (left axis) and in mg per separated berry (right axis)' "Brix Vertical bars thruogh mean values represent SE. : inception of increase.

Figure 3.3.1 to 6 Evolutionof G-Ginleaf blades( I ), leaf petiole(2)andfruit(3 ) and changes in berry volume ( 4 ), berry density ( 5 ), and fruit soluble solids ( 6 ) in Sultana with time in weeks after flowering. G-G data in Fig. ( 3 ), are expressed as concentration in whole bunches on a fresh weight basis (left axis) and in mg per separated berry (right axis). Vertical bars thruogh mean values represent SE. : inception of 'Brix increase

Figure 3.4.1 to 6 Evolution of G-G in leaf blades ( I ), leaf petiole (.2 ) and fruit ( 3 ) and changes in ben-v volume ( 4 ), berry density ( 5 ), and fruit soluble solids ( 6 ) in Flame Seedless with time in weeks after flowering. G-G data in Fig. ( 3 ), are expressed as concentration in whole bunches on a fresh weight basis (left axis) and in nig per separated berry (right axis). Vertical bars thruogh mean values represent SE. : indication of 'Brix increase.

Figure 3.5.1 to 3.5.3 Glycosyl glucose (G-G) concentration in different tissues of all three cultivars comparing G-G in leaf blades (Fig.1), in leaf petioles (Fig.2 ) and whole bunches (Fig.3). M, Muscat Gordo; S, Sultana; F, Flame Seedless. ix

Figure 3.6 Glycosyl glucose concentration in berries of Muscat Gordo, Sultana and Flame Seedless plotted against juice 'Brix (data from 1992-93)'

Figure 3.7 Glycosyl glucose concentration in berries of Muscat Gordo and Sultana plotted against juice "Brix (data from 1994'95).

Figure 4.1 Transfer of grape bunches by approach-grafting. Before flowering, the donor vine with bunch (A) was grafted onto the host vine (C). The donor vine was cut off five weeks later (D) and the fruit was allowed to ripen on the host vine (E). Ungrafted vines were grown under the same conditions as a control (B).

Figure 4.2 Glasshouse grown Muscat Gordo grapes bunch grafted onto Shiraz vine.

Figure 4.3 Glasshouse grown Shiraz grapes bunch grafted onto a Muscat Gordo vine.

Figure 4.4Typical GC-MS chromatograms of the aglycons from Shiraz control berries and from fruit of Muscat Gordo control and Muscat Gordo grafted onto Shiraz. The y- axis scale is 'relative ion current' (RIC).

Figure 4.5. Identity of 16 monoterpene compounds from GC-MS chromatograms

Figure 4.6 Histograms of monoterpene aglycones in berries (shown in Fig. 4.2) from the field experiment of Sultana ungrafted (A), Sultana grafted on Muscat Gordo (B), Muscat Gordo ungrafted (C) and Muscat Gordo grafted on Sultana (D), at three ripening stages.

Figure 4.7 Typical GC-MS chromatograms of the aglycons from Sultana control berries and from fruit of Muscat Gordo control and Muscat Gordo grafted onto Sultana. The y-axis scale is 'relative ion current' (RIC).

Figure 4.8 Typical GC-MS chromatograms of the aglycons from Muscat gordo control berries and from fruit of Sultana control and Sultana grafted onto Muscat Gordo. The y-axis scale is 'relative ion current' (2C).

Figure 5.1 Muscat Gordo berries showing pink colour

Figure 5.2 HPLC chromatograms of anthocyanin pigments profile of the extracts for Muscat Gordo berries skin (a) from own-rooted vines, (b), from bunches grafted onto Shiraz vines. chromatogram (c) shows authentic cyanidin-3-glucoside. X

Chromatogram (b) showed a faster elution profile due to different instrument conditions

Figure 5.3 Relationship for 221Muscat Gordo berries of: (a) total soluble solids of juice ('Brix) and absorbance of skin extracts at 520 nm, (b) berry weight and absorbance of skin extracts at520 nm and (c) berry weight and "Brix.

Figure 6.1 Phloem sap collection from a peduncle lateral.

Figure 6.2 Phloem sap collection from detached leaves.

Figure 6.3 Exudate collection from peduncle tissue off the vine. Ped: peduncle; BL: bunch lateral; ET: Eppendorf tube; HB: holding block.

Figure 6.4 Girdling the cane above and below the fruit bunch in order to stop phloem sap flux to the bunch. G: girdled area; Ped. L: peduncle lateral from which sap was collected.

Figure 6.5 The diurnal pattern of sucrose (A), total amino acids (B) and potassium (C) levels in phloem sap exudates collected from pre-veraison grape bunches during the day (vertical bars indicate SE).

Figure 6.6 (A-F) The diurnal patterns and the levels of major amino acids detected in the grapevine phloem sap þre-veraison).

Figure 6.7 The diurnal pattern of continuous sucrose exudation from post-veraison Muscat Gordo bunch laterals during light and dark periods with different starting times (each value represents mean of l0 bunches), "Brix:10.2. A, B, and C represent three samples where the bunch laterals were removed at different times. For girdled sample (D) the canes were girdled the day before sampling. The values are for two hours of each sampling period,

Figure 6.8 The diurnal pattern of continuous sucrose exudation from ripe Muscat Gordo bunch laterals during light and dark periods with different starting times (each value represents mean of l0 bunches), "Brix : 15.6.

Figure 6.9 The diurnal pattern of total amino acids measured in ripe Muscat Gordo grapevine exudates during days and nights (samplings started at different times xt

and each value represents mean of l0 bunches), (first and second series of samples presented in Fig. 6.8).

Figure 6.10 The diurnal pattern of potassium measured in ripe Muscat Gordo grapevine exudates during days and nights (samplings started at different times and each value represents mean of l0 bunches)'

Figure 6.11 (A-F) The diurnal pattern of individual amino acids measured in ripe Muscat Gordo grapevine phloem sap exudates during days and nights (samplings started at different times and each value represents mean of 10 bunches).

Figure 7.f GC-MS chromatograms of the aglycones isolated from phloem exudates collected from ripe fruit bunch laterals of Muscat Gordo (A), Sultana (B), and Shiraz (C) grapes.

Figure 7.2The concentration of secondary metabolite aglycones shown as bars (right axis) in the phloem exudate samples collected over a 36 h period. The sucrose concentration (left axis) in the same samples also shown. x'ii

List of Publications

Published;

Gholami, M., Hayasaka Y., Coombe, 8.G., Jackson, J.F., Robinson, S.P. and Williams, P.J' (1995) Biosynthesis of flavour compounds in Muscat Gordo Blanco grape berries. Australian Journal of Grape and Wine Research, I(L): 19-24.

Gholami, M., and B.G. Coombe, 1995, Occurrence of anthocyanin pigments in berries of the white cultivar Muscat gordo Blanco (Vitis vinifera L.). Australian Journal of Grape and Wine Research l(2):67-70.

Manuscripts in preparation; No. I Title: Distribution of glycosides in the grapevine tissues during berry growth and development (Chapter 3). Authors: M.Gholami, B. Coombe, and P.J. V/illiams

No.2 Title: Grapevine phloem sap analysis; I. Sucrose, amino acids, potassium concentrations, seasonal and diurnal patterns (Chapter 6). Authors: M. Gholami. S. Robinson, and B. Coombe

No.3 Title: Grapevine phloem sap analysis; II. Secondary metabolite flavour compounds (Chapter 7). Authors: M.Gholami. P.J. Williams, Y. Hayasaka, B. Coombe, and S. Robinson

From this work was presented; In Australian Plant Physiology conference (APP) 1994 , oral presentation and abstract was published in the proceeding (abstract no. 162). Posters in; CRCV annual meeting, 1993. CRCV annual meeting, 1994 (two posters). Wine industry technical conference, 1995. xii i

Abbreviations

G-G Glycosyl Glucose HEPES N-[2-Hydroxy Ethyl] Piperazine-N' [2-Ethan Sulfonic acid] EDTA EthyleneDiamineteTraacetic Acid) TSS Total Solubl Solids. P-protein Phloem protein ASP Aspartic acid GLU Glutamic acid ASN Asparagine SER Serine GLN Glutamine HIS Histidine GLY Glycine THR Threonine ALA Alanine ARG Arginine TYR Tyrosine CYS-CYS Cysteine VAL Valine MET Methionine TRP Tryptophan PHE Phenylalanine ILE Isoleucine LEU Leucine LYS Lysine PRO Proline This thesis dedicated to :

My parents who supported me throughout my studies and passed away while I was completing my Ph D and before I could see them again.

My wife and my children for their patience and encouragement during this work. .:rl'fË t-,,,itiixiili i ir.i:,Ê r :" 1 ' ¿i ii;it¡,-iì;lili ',1ì:;lüí:i,,^, , , Chapter 1

Introduction and aims

1.1 Introduction

The importance of flavour compounds in grapes , grapejuice and derived beverages has attracted much research interest. This has led to the identification and classification of many different compounds important to the flavour of grapes (Strauss et al., 1986; V/illiams et al., 1993; Noble, 1994). Terpenoids, particularly monoterpenes, have been extensively studied both for their evolution during berry growth and development, and as constituents of different cultivars (Wilson et al., 1984; Gunata et al., 1985b; Park et al., 1991). It is know appreciated that the majority of flavour compounds occur in grape berries as flavourless glycosides which could readily release the flavour compounds by hydrolysis. Glycosidically conjugated monoterpenes were the earliest recognised as flavour precursor in grape berries. Some research has shown that leaves have an important role in the development and the composition of fruits, and decreasing the leaf area to low levels could result in a decrease in yield and quality (Kliewer, 1970a; Sidahmed and Kliewer, 1980; Bledsoe et al., 1988; Hunter and Visser, 1989; Candolfi-Vasconcelos and Koblet, 1990; Hunter and Visser, 1990 ). Viticultural practices and environmental factors have also been shown to have an influence on grape composition and flavour Qlloble, 1994). Few data are available concerning the presence of glycosides of secondary metabolites in leaves or other parts of the grapevine. Gunata et al., (1986) reported the presence of monoterpene glycosides in grapevine leaves and suggested the possible translocation of these compounds from leaves to berries. There is no information in the literature about the translocation of glycosides synthesised in leaves to the berries. In the absence of reliable evidence about the mobility of glycosylated secondary metabolites in general, and monoterpene glycosides in particular, the mechanism of Their translocation remains an assumption.

1.2 Aims

The general aim of this work was to investigate the relationship between leaves and berries of grapevines in terms of the translocation of secondary metabolite, and in particular monoterpene flavour compounds. The objectives of this project are as follows: 2

l. To investigate the levels of total glycosides in differentparts of the grapevine and their concentration changes at va¡ious growth and physiological stages.

2. To determine the contribution of leaves, as a site of metabolite synthesis and export, to the berry composition.

3. To evaluate the role and importance of the berry itself in terms of utilisation and metabolism of the imported metabolites.

4. To collect and analyse phloem sap translocating in the grapevine phloem tissue from leaves to the berries for primary metabolites (carbohydrates and amino acids)'

5. To analyse grapevine phloem sap, exuded from the leaf petiole and peduncle lateral, for secondary metabolites throughout the growing season. 3

Chapter 2

General literature review

Contents Page

2.1 Berry composition ...... 4 2.1.1 Primary metabolites ...... 4 2.l.l.l CarbohYdrates ... 4 2.1.1.2 Organic acids .-... 6 2.1.1.3 Amino acids ...... 7 2.1.1.4 Minerals 7

2. 1.2 Secondary metabolites and flavour compounds...... 8 2.1.2.1 Monoterpenes ...... 9

2.1.2.2 Norisoprenoids ... -...... 9

2.1.2.3 Sesquiterpenoids ...... - ..10

2.1.2.4 Shikimic acid derivatives...... 10 2.1.2.5 Anthocyanins...... t0 2.1.3 Secondary metabolite glycosides ...... -.-. 11

2.1.3.1 Glyco sy I ation of secondary metabolites t2 2.1.3.2 Glycosides in fruits 13 2.1.3.3 Glycosides as precursors of volatile flavour compounds 13 2.2 Assimilate transport in grapevine (Vitis vinifera L.) ...... ls 2.2.1 Analomical aspects l5 2.2.2 Teclniques of phloem sap collection l6 2.2.3 Leaves as sources of assimilates ....-.....- t7 2.2.4 Relranslocation from reserves ...... l8 2.2.5 Transport of carbohydrates ...... 19 2.2.5.1Sucrose 20 2.2.5.2 Hexoses 20

2.2.6 T ransport o f nitrogen compounds 2l 2.2.7 Transport of organic acids ...... 2l 2.2.8 Transport of minerals 22 2.2.9 Transport of secondary metabolites 23 4

2.1B,erry composition

The grape berry contains a mixture of primary and secondary metabolites' Primary metabolites are essential for plant growth and development and include lipids, carbohydrates, amino acids, proteins and nucleotides (Mann, 1987; Taiz and Zeiger, 1991). Plant secondary metabolites are a large and diverse array of organic compounds that do not appear to have a direct function in growth and development. Unlike primary metabolites, these compounds may have a restricted distribution in the plant kingdom; sometimes a particular secondary plant product is typical for only one plant species or a taxonomically-related group of species (Taiz and Zeiget,l99l).

2.1.L Primary metabolites

2.l.l.l Carbohydrates

Grape juice, which is largely the vacuolar content of the pericarp cells of ripened grape berries, contains large amounts of two hexoses, glucose and fructose, and somewhat lesser amounts of tartaric acid. malic acid. plus a wide variety of other compounds. Potassium is the dominant mineral in the juice. The major constituent of berries is water followed by, in the case of ripe berries. D(+)-glucose and D(-)-fructose which together amount to about 150-250 grams per litre ofjuice. These two carbohydrates constitute a large part of berry dry matter (Peynaud and Ribéreau-Gayon, 1971, Coombe, 1992). Berry sugar level increases at the inception of ripening (veraison) along with softening of the berry, decline in acidity, change in skin colour, and increase in berry volume (Coombe, 1989, 1992). Glucose constitutes 85o/o of the grape berry sugar in the hrst growth period, but after veraison the ratio of glucose to fructose gets approaches to unity, while in very ripe grapes there is generally a slight excess of fructose (Kliewer, 1967a; Hanis et al., 1968; Coombe, 1975) (Fig. 2.1). It is known that sucrose, which is an end product of photosynthesis, is translocated to the berries (Swanson and El-Shishini, 1958; Glad et al., 1992a), but the concentrations of glucose and fructose are approximately 30-fold that of sucrose in ripe berries (Hrazdina et al., 1984; Coombe, 1987b). It seems that sucrose is metabolised following unloading into the berry, first by hydrolysis to its hexose moieties and then utilised in different metabolic pathways (Hardy, 1967,1968). In immature 5 berries, carbohydrates are transformed into malic acid but this reaction decreases as the berry ripens (Peynaud and Ribéreau-Gayon, L97l). After veraison, malate decline tends to be initially rapid while tartaric acid per berry tends to remain constant. In green grapes glucose can be converted to malate, just as the reverse may occur in ripe grapes by gluconeogenesis through PEP-carboxykinase (Ruffner and Kliewer,1975; Ruffner et al., teTs).

:.3 wltcil ::!¡ r!t ,1. l?i ¡l lF ll¡al

1 2

Ok xov xoY

Fig. 2.1 Water and dry matter (l) and type of changes in the amounts of the sugars (2) in grape berries.during berry development. (From Coombe, 1992)

There are some other sugars present in grape juice but in very low concent¡ation compared to glucose and fructose. At the inception of ripening intense physiological activities occur and a considerable accumulation of sugar takes place, for example, the sugar content of a berry increases about seven fold within a week or so. This increase is difficult to explain in terms of increase in photosynthesis rates. Brown (1981) suggested that the onset of rapid sugar accumulation in the grape be.ry results from changes in phloem unloading rather than from concentrative accumulation by the cell. This author also found that the skin of a pre-veraison berry is a glucose accumulator, but post-veraison it becomes a fructose and glucose accumulator. Coombe and Matile (1980) showed that concentrations of fructose, both diffr,¡sible and compartmented, were double that of glucose in green skin but there was no difference in the concent¡ations of the hexoses in ripening berries. 6

2.1.1.2 Organic acids

The main organic acids in the grape berries are L(+)-tartaric acid and L(-)-malic acid; together these acids account for 90Yo of total berr)' acidity with citric' ascorbic and phosphoric acids contribuuting to the other 10% (Winkler et al., 1974). In plant tissues generally, carbohydrates are broken down through the pentose phosphate cycle or through glycolysis which transforms the glucose molecule into pyruvic acid. Pyruvic acid can give rise to malic acid through carboxylation or the operation of the tricarboxylic acid (TCA) cycle. Introduction of raC-acetic acid into the grape berries resulted in the formation of highly labelled glutamic acid formed by amination of cr-oxoglutarate (Peynaud and Ribéreau-Gayon, lgTl). There is no direct relation between the operation of the cycle as a whole and the accumulation of malic acid (Peynaud and Ribéreau-Gayon, l97l). If it is assumed that sucrose is the sole sugar transported into the berry, then hydrolysis of sucrose into glucose and fructose under the effect of invertase could be the key reaction to supply carbon for all metabolic reactions. The conversion of glucose to malate via phosphoglycerate and phosphoenolpyruvate can be demonstrated by supplying immature berries with uniformly labelled glucose thus emphasising the close interconnections between the metabolism of malic acid and of glucose (Peynaud and Ribéreau-Gayon, 1971). In immature berries, carbohydrates are transformed into malic acid but the intensity of this reaction decreases as the berry ripens and in ripe grapes, the reverse process (gluconeogenesis) may take place (Peynaud and Ribéreau-Gayon, l97l)' However, the activity of gluconeogenesis enzymes in converting malic acid to sugar in ripening berry could account for only 5o/o of sugff accumulation if there is no fresh malic acid production after veraison (Ruffner and Hawker, 1977). Accumulation of major organic acids in the berry occurs during the first stage of berr), growth and total acid content of the berries increases significantly until veraison time. Tartrate accumulates in grape berries before veraison and during ripening tartrate content per berr1, remains constant (Hale, 1977; Carroll and Marcy, 1982). The metabolism of tanaric acid differs from that of malic acid and, once formed, it remains relatively stable (Hrazdina et al.. 1984). It is also knownthat immature berries are able to synthesise their own tartaric acid during the early stages of development. At this period, tafaric acid synthesis is as active as malic acid synthesis but it decreases rapidly until at veraison it practicall),ceases (Hrazdina et al., 1984). According to Hardy (1968), the high percentages of roC found in organic acids only a few hours after supplying immature grapes with labelled sugars indicated that most of the malate and the tartrate in the berries originates from glucose and fructose (most probably supplied by hydrolysis of the sucrose translocated via the phloem). The malate content of the berry declines sharply during ripening. The free acids tend to convert to salts (mono and di-basic) during ripening (Iland and Coombe. 1988). The decline in the acid content of berries could be also due to 7 decreased synthesis and increased metabolism. Malate is a major tricarboxylic cycle acid which is metabolised under the effect of malate dehydrogenase enzyme and converted to carbohydrates or recycled in the tricaboxylic acid cycle. It can also be converted to acetyl Co-A and participate in the metabolism of fatty acids or enter into transformations leading to the formation of secondary metabolites (Nii and Coombe, 1983; Possner et al., 1983; Hrazdtnaet al., 1984).

2.1.1.3 Amino acids

Total nitrogen concentration in mature grape berries ranges from 100-200 mg/ 100 ml, but nitrate and protein only account for a small proportion of the total. The protein content of grape juice is reported to be 50-100 mg/L ofjuice (Somers andZiemelis, 1973; Murphey et al., 1989; Pueyo et al., 1993). Towards the end of fruit ripening large amounts of free amino acids accumulate in the berries and at maturity make up 50% to 90o/o of the total N, with arginine and proline being the most prevalent amino acids on a molar basis (Kliewer, 1969, I970b; Wermelinger, 1991). The amino acid composition of ripe grape berries of numerous cultivars of Vitis spp have been studied in past years and is reported to be: glutamic acid I .2 -12.4 mM; arginine from 0.3 - I 1.2 mM; proline 2.3 - 40.0 mM; serine 0.6 - 4.6 mM; threonine 0.6 - 3.5 mM; alanine I.7 - 11.3;1-aminobutyric acid 0.4 - 4.2 m}y'r; aspartic acid 0.2 - 2.3 mM (Rubelakis-Angelakis,l99l; Kliewer, 1969, 1970b) and the total free amino acid concentration ranges from 10.4 to 64.5 mM. Arginine accounted for 5YrTo 43%o of total nitrogen in grape juice. Proline levels in grapes often increase more than arginine late in the season (Kliewer, 1969, 1970b; Bath et al.,

1991 ). In summary, although the concentrations of free amino acids in ripe grape berries may depend on several factors. including cultivar, environmental conditions, soil characteristics, and cultural practices. it seems that in l/itis vinifera cultivars, proline and arginine, and to a lesser extent glutamic acid and alanine, are the predominant amino acids (Kliewer, 1967b, l e6e).

2.1.1.4 Minerals

Constant uptake of anions and cations and distribution of minerals to the various part of the plants including the berries occurs all through the growth period. The berries are relatively low in mineral elements compared with other parts of the plant. During the second growth period there is an increase per berry in cations such as K*, Na+, Ca**, Mg** , Cr**, Mn** and in anions such as phosphate (POa-l¡ (Hrazdina et al., 1984; Boselli et al., 1995). Phosphoric acid (which increases all through the berry growth 8

Boselli et al., 1995). Phosphoric acid (which increases all through the berry growth period) is the third major acid present in grapes. The cation content per berry increases 2- 3 times in the skin and 1.2-l.g times in the pulp during berr),ripening (Hale, I977;lland and Coombe, 1988). On the basis of berry fresh weight' potassium remains almost constant during the beny development period whereas levels of other cations such as calcium, manganese, copper and magnesium decrease during berry development which is probably due to berry expansion (Hrazdina et al., 1984; Possner and Kliewer' 1985; Boselli et al., 1995). Potassium is the most abundant cation in the grape berry accounting for 45Vo-70%o of themineral cation content of the berry (Somers, 1975; Hale, 1977). During the same period heavy metals increase by 50%. There is an increase in anions, particularly phosphate, which is mostly concentrated in seeds. The phosphate content of skin and pulp also increases steadily as the fruit ripens (Peyanud and Ribéreau-Gayon, r97r).

2.1.2 Secondary metabolites and flavour compounds

The chemical structure of thousands of secondary metabolites are known (Lukner 1990), and every month there are reports in phytochemical journals of more novel structures. Plant secondary products can be divided into three groups according to their mode of biosynthesis: terpenes, phenolic and nitrogen containing compounds' Terpenes are synthesised from acetyl CoA via the mevalonic acid pathway. Phenolic compounds are aromatic substances formed via the shikimic acid and/or the polyketide pathways. Nitrogen containing secondary products, such as alkaloids, are biosynthesised primarily from amino acids. One of the defining characteristics of secondary metabolites is their propensity to accumulate in particular organs or tissues of the plant (Haslam, 1975). Many volatile compounds of different metabolic origins have been identified in grapes and related products (Rapp, 1988b). Volatile flavour compounds of grapes can be classified into three groups according to their metabolic origin; l) mevalonic acid derivatives, such as terpenoids, 2) shikimic acid pathway derivatives which includes volatile phenols and 3) fatty acid derived compounds (Williams et al., 1989). Terpenoids are a large. diverse and widespread group of secondary metabolites with structures made up of C5 isoprenoid units. In grapes, monoterpenes (C I o) , norisoprenoids (C7.8. tr. r¡), shikimate-derived compounds, sesquiterpenes (C¡5) and carotenoids (Cae) have been identified (Schreier et al., 1976; Strauss et al., 1986; Versini et al. ,1988; Razungles et al.. 1988; Williams et al., 1989; Sefton et al., 1989, 7993,1994; Banthorpe, 1991 Williams et al., 1992; Stahl-Biskup et al., 1993). Compounds responsible for grape varietal flavour can be classified into two groups ie, l) free volatiles, which are responsble for the perceived flavour of the fruit, and 2) more I abundant glycosidically-bound, flavourless compounds. These non-volatile compounds are formed by coupling of free volatiles to a glucose, or to a disaccharide, the latter comprising a glucose and one other sugar unit. Importantly, these glycosides can contribute significantly to aroma upon hydrolysis (Noble et al., 1988; Williams et al., 1989; Shoseyov et al., 1990; Fong et al.,l99l;Noble, 1994) and thus represent a potential source of flavour compounds.

2.1.2.1 Monoterpenes

o/o The formation of significant quantities of monoterpenes (>0.I fresh weight ) appears to be confined to some 50 families of higher plants and more than 1000 monoterpenes are presently known as natural products of higher plants (Charlwood and Banthorpe,l9ST; 'West, Croteau, 1987; 1990; Charlwood and Charlwood, 1991). Monoterpenes have been identified as important contributors to the aroma of the essential oils of many plant species and they often occur as glycosidically bound components of plant tissues. The composition and flavour role of the monoterpene glycosides of plant-derived foods have been the subject of many studies in the last few years (Williams, 1993). The monoterpene flavour compounds are an extensively studied group of secondary metabolites in grapes (Ribéreau-Gayon et al.,1975; Williams et al., 1980; Williams et al., 1983; Maris, 1983; Rapp et al., 1983; Wilson et al., 1984; Strauss et al., 1986, 1988, Rapp, 1988b). Monoterpene compounds, alone among the many and varied constituents of grapes, show a relation with varietal flavour of Vitis vinifera (Rapp 1988a). In muscat grapes they were found in both volatile and non-volatile forms (Strauss et al., 1986; Park and Noble, 1993). Cordonnier and Bayonove (1974) first suggested the existence of monoterpene glycosides as precursors of free monoterpenes in muscat grapes. Subsequent work showed that monoterpene volatiles in grapes are generated by hydrolysis of glycosidic compounds and of polyhydroxylated monoterpenes (ie polyols). The latter polyols may be glycosylated or free (Strauss et al., 1986). Monoterpene compounds accumulate in the grape berries during berry ripening in both free and glycosylated forms (Wilson et al.. 1984; Williams et al., 1985; Park et al., l99l).

2.1.2.2 Norisoprenoids

Norisoprenoids with l3 carbon atoms (C13) are another important class of flavour and aroma components(Winterhalter and Schreier, 1994; Ohlofl 1978). Carotenoids are assumed to be the precursor of norisoprenoids in grapes (Schreier et al., 1976: Razungles et al., 1988; Sefton et al.. 1989; Kanasawud and Crouzet, 1990; Williams et al., 1991) 10 occurrence of norisoprenoids and carotenoids in grapes (Williams et al., 1982c). However, little is known about the precursors of the norisoprenoid compounds and the reactions by which they are formed (Winterhalter and Schreir, 1938). Norisoprenoids contribute to the flavour and aroma of floral grape varieties along with monoterpenes. In addition to free forms they are also present in non-volatile glycosylated forms (V/illiams et al., I982a;Williams et al., 1989; Winterhalter et al., 1990b) and in some floral varieties such as Riesling levels of norisoprenoid glycosides are higher than monoterpene glycosides (Williams et al., 1987).

2.1.2.3 Sesquiterpenoids

Sesquiterpenes are the largest class of terpenoids, with more than 100 sesquiterpenoid structures known and several thousand compounds of the class isolated and identified (Fraga, l99l). Sesquiterpenes are C¡5 terpenoids and have been identified as important components of aroma in plants (Ramaswami et al., 1986). From the sesquiterpenoids found in nature (Ramaswami et al., 1986), several have been identified as constituent of grapesandgrapejuice(Schreieret a1.,1976;Maris, 19S3). Ohloff (1978b)inareviewon naturally occurring aroma compounds has given many other examples of sesquiterpenoids with fl avour properties.

2.1.2.4 Shikimic acid derivatives

The shikimic acid pathway gives rise to large number of aromatic compounds including many polyphenols. Since volatile phenols were not generally detected in grape juices, the origin of shikimate-derived phenols in wines was suggested to be berry skin, seeds and cluster stem tissues during alcohol fermentation (Dzhakhua et al., 1978). The shikimate- derived volatile phenols were later identified in the hydrolysate of Cl8 reversed-phase isolates of grape juices (Strauss et al., 1987; Williams et al., 1989; Winterhalter et al., l e90).

2.L.2.5 Anthocyanins

Anthocyanins as a flavonoid subclass are secondary metabolites widespread among plants (Hahlbrock and Scheele, 1989). Grape berries of red varieties contain relatively large amounts of anthocyanins and other polyphenols which contribute greatly to their appearance, and quality. Wulf and Nagel (1978) have separated and quantified about 20 different anthocyanin compounds from Vitis vinifera Z. berries. The distribution of anthocyanins in the fruit is complex and varies according to variety. 11

The anthocyanins are fundamentally responsible for all colour differences between one grapes (Ribereau-Gayon, lg82). The level of anthocyanins in the berry skin is of the parameters being used for quality assessment in black grapes. The accumulation of anthocyanin begins at veraison, an event signalling the end of the development period of the berries and the onset of the ripening process which is coincident with or shortly after the increased accumulation of sugars (Coombe, 1973; Pirie and Mullins, Ig77, lg80; Hrazdina et al., 1984; Crippen and Morrison, 1986; Roggero et al', 1986; Darné, 1993).

2.1.3 Secondary metabolite glycosides

Many secondary products occur in plants in the free form (i.e. as aglycones) to only a limited extent or not at all. Usually they are conjugated with a variety of monomeric or oligomeric saccharide moieties. Aliphatic and phenolic hydroxyl groups, carboxyl functions and amino and mercapto groups of aglycones are involved in conjugation reactions. Glycosides of terpenoids, phenols or other alcohols are widely distributed in nature. Flavonoids and anthocyanins have their hydroxyl groups linked to sugars. The N- glycosides are of utmost importance as they are structural units in coenzymes, nucleic acids, nucleotides. etc.(Kurt and Torssell, 1983). From a physiological point of view, conjugation reactions are of importance for a number of reasons (Barz and Koster, 1981): a (i.e. through ) they drasrically alter the chemical (i.e. solubility) or physiological transport cells or membranes and biological activity) properties of secondary metabolites; b ) they may result in a site of accumulation different from that occupied by the aglycone thus resulting in the formation of several interrelated metabolic pools (Ackerman et al., 1989); c ) the conjugated compound may enter a different metabolic pathway to the corresponding aglycone; and d) they can determine whether a compound is a metabolically active species or a metabolically inactive (end) product. Therefore. conjugation reactions are considered an important way to transport and store materials or to detoxiff unwanted materials (Barz and Koster, 1981;Hosel. l98l;Croteau. 1984; Croteau, 1987). Glycosylation has been observed to effectively redistribute monoterpenes throughout skin and juice fractions of grape berries (Wilson et al.. 1986: Park et al., l99l). C¡3 norisoprenoids and shikimiate- derived compounds also have been isolated in glycoconjugated forms from grape juice (Sefton et al., 1989; Williams et al., 1989). 12

2.1.3.1 Glycosylation of secondary metabolites

Transglycosylation from a glycosylating agent to the secondary metabolite aglycone is always catalyzed by transferases utilizing a nucleotide sugar (NDP-sugar) as a cofactor, usually uridine diphosphate-D-glucose (UDPG) (Luckner, 1972; Hosel, l98l; Goodwin and Mercer, 1983). UDPG is a transfer agent which leads to the formation of a ß- glucoside. Although the mechanism of glycosylation is understood, the trigger for this conjugation reaction in plants is not known. Formation and breakdown of glycosides are a consequence of the appropriate enzymes present in plants and can be represented in terms of enzymatic substrate specificities (Hosel, 1981). The glycosylation steps are usually at the end of the particular biosynthetic pathways (Hosel, 1981)'

Nucleotide suga¡

Transferase

R-OH R-O-sugar

Glycosidase

Sugar

Fig. 2.2- Concept of synthesis and hydrolysis of secondary metabolite glycosides in plants (from Hosel. l98l ). 13

There is evidence that glycosylation may be enhanced at some particular physiological stages in plants, e.g. after flowering in peppermint leaves (Croteau and Martinkus, 1979; Croteau, 1984) and after veraison (berrl'softening) in grape berries (Wilson et al., 1984). Recent evidence suggests that glycosylation takes place after extra hydroxyl groups ¿ìre introduced into a monoterpene skeleton in the formation of polyols (Strauss et al., 1988). Contrary to these findings, Ackerman et al. (1989) demonstrated that for two rose species, Rosa damascena and Rosa gallica, the concentration of glycosidically bound compounds was greatest at the early stage of flowering, and decreased as flowers developed; at the same time they observed an increase in the concentration of free volatiles. The mechanism causing these differences in the time and pattern of glycoside accumulation for these plant species is not known. Study of the enzymes involved in the conjugation reactions would help understanding of how glycosides accumulate. It would also produce a possible means by which glycosylation and thus other related pathways may be controlled.

2.1.3.2 Glycosides in fruits

The changes in the chemical properties of glycosylated secondary products make them different from free aglycones in water solubility and chemical reactivity. This may explain why glycosylated compounds, rather than the free aglycones, are accumulated in the plant vacuole and are less reactive towards other cellular components than are the aglycones, Glycosylated compounds are therefore often thought of as excretion products or as physiologically inactive plant storage forms. Hydrolysis of secondary plant glycosides, would release the physiologically active aglycones (Hosel, l98l). The composition of the monoterpene glycosides of plants and their possible biochemical function have been recently reviewed (Salles et al., 1988; Williams, 1993; Stahl-Biskup et al., 1993).

2.1.3.3 Glycosides as precursors of volatile flavour compounds

Many glycosylated secondary metabolites are known to be flavou¡less (Noble et al., 1988; Fong et al., 1991). Nevertheless, their ability to indirectly influence grape juice flavour, ie their precursor properties (see 2.1.2 above), justifies the interest in their analysis, structural elucidation and understanding of their sensory aspects. Apart from this, structural elucidation of glycosidic secondary metabolites may prove useful to plant physiologists and others. The biochemical and physiological properties of many secondary products cannot be understood unless their conjugation moieties are known. Analysis of plant extracts for aglycone alone will not reveal the full complexity and 14 possible conjugate-caused compartmentation of secondary metabolism (Barz and Koster, re81). However, no information about precursors of flavour compounds in fruits was obtained until acid and enzyme hydrolysis studies on Muscat grapes indicated that glycosidic derivatives of monoterpenes may have been involved (Cordonnier and Bayonove, 1974). Since then a growing number of reports have been published concerning flavou¡ precursors in a range of fresh and processed fruits, as well as in related beverages, and in some other consumed plant and leaf products. Most of the precursor compounds that have been isolated have been found to be glycosidic derivatives (V/illiams, 1993 ). The common structural element of many glycoconjugates that have been identified in grapes and many other plant products is a glucopyranosyl unit attached through ß- glucosidic linkage, to an aglycone which, in turn, can be one of several well-known secondary metabolites (Fig. 1.3). Both glycosides and polyols must be considered as precursors of flavour which could be released by hydrolysis of these compounds. Thus, enrymatic and chemical hydrolysis of glycosides and acid hydrolysis of polyols is very important for recovering the non- volatile flavour constituents of fiuits (Williams, 1993).

R

HO- i

OH AlÞbatic rcsiduæ c1, c5, c6, cs R Monoærpencs o-L-arabinoñ:ranosyl

Scsquitcrpcnes

Norisoprcnoids - riramnop yra.o o s y I o-L C¡¡, C¡1

Shikinic acid octaboliæs OH

þD-xylopyranosyl

HOOrrflOH OH -ol o-

þapioturanosyl i

Fig. 1.3 Types of secondary metabolite glycoconjugates that have been identified as flavour precursors in fruits (from Williams, 1993) 15

2.2 Assimilate transport in grapevin e (Vitis viniferø L.)

Transport in higher plants is a broad topic and has been the subject of many studies. In vascular plants long distance transport of water occurs through the transpiration stream via primary and secondary xylem. Inorganic ion transport occurs principally by the same route. Long distance transport of assimilates occurs predominantly through the phloem' The phloem tissue transports carbohydrates, and other substances, to meristematic growing points, developing fruits, storage organs, and other sites of carbohydrate utilisation. It is considered by some researchers that translocation takes place on a source to sink pathway, that the loading of the nutrients into the phloem controls the rate of movement, and that the direction of the movement is controlled by unloading at the sink and consequence effects on concentration gradients. Carbohydrates are known as the major substances translocating in phloem of plants (Kursanov, 1984). During the final growth phase of the grape berry water and solutes, principally the hexoses glucose and fructose, accumulate in large amounts in pericarp tissue- From the beginning of the ripening phase (veraison) until 4 to 8 weeks later when it slows or ceases, the total soluble solids may increase from 5Yoto more than 20%o, sometimes nearly 30o/o; fructose may increase from 5-10 to more than 160 mg g-t (Coombe and Philips, 1982l. Coombe, l9S7). Phloem transport in the sieve tubes ensures the transfer of C and N assimilates from leaves to the different sink tissues, i.e. fruits, growing shoots, and perennial woody parts (Pate, 1980). Information about grapevine phloem sap composition is limited to that published by Swanson and El-Shishiny (1958) and Glad et aL, (1992a) which show that sucrose, which is the principal sugar. is present along with amino acids in the sap. A precise knowledge of phloem sap composition is important for the understanding of the mechanisms of assimilate transport. Analysis of phloem exudate could provide a better picture of the substances present in the mature sieve cells and the flux of sucrose and other metabolites into the berry during ripening.

2.2.1 Ãnatomical aspects

The specialised phloem tissue of vascular plants transports carbohydrates (produced as a result of photosynthesis) and other substances, to meristems, developing fruits, storage organs, and other sites of carbohydrate utilisation. According to Flowers and Yeo, (1992) sieve tubes were discovered in 1837 by Harting, he also description of exudation from the phloem was made in 1860 by the same researcher. Phloem is often difficult to identify in transverse sections of plants but is generally spatially associated with the xylem; together they comprise the vascular tissue of the plant. The phloem tissue is 16 composed of conducting sieve elements, companion cells parenchyma cells, fibers and sclereids. Both of the latter a¡e supporting cells' Sieve tube elements are highly specialised cells through which translocation actually takes place. Mature sieve elements a¡e unique among living plant cells and lack many structures normally found in other living cells such as nucleus, tonoplast, golgi bodies, ribosomes, microfilaments and micro tubules. In addition to the plasma membrane, organelles that are retained include somewhat modified mitochondria, plastids, and smooth endoplasmic reticulum (Taíz andZeigel l99l). Sieve tube members of the dicots are usually rich in phloem proteins called P-proteins which, with callose, appears to function in sealing off damaged sieve elements by plugging up the sieve plate pores (Taiz and Zeiger, 1991). Phloem proteins, which could be visualised by light and electron microscopy, were also found in the phloem exudates from cucurbitaceae (Eschrich and Heyser, 1975; Cronshaw and Sabnis, 1990). Some of the P-proteins could have important roles in the loading of sugars and amino acids into sieve tubes and even in cell division at the meristematic regions connected to the end of sieve elements (Ishiwatari et al., 1995). The end walls of the sieve elements are modified to form sieve plates which are perforated so as to allow the massive rapid flow of nutrient assimilates (Woding, 1978). Each sieve tube is associated with one or more companion cells. The two types of cells arise in pairs from the same parent cell. Numerous intercellular connections, the plasmodesmata, penetrate the walls between sieve tube members and their companion cells, suggesting a close functional relationship and ease of transport between the two cells (Esau, 1965). Structurally, companion cells appear less specialised than sieve elements; they retain a nucleus and have numerous ribosomes and organelles which indicates a high level of metabolic activity (Esau, 1965; Ridge, 1991). The phloem parenchyma serves as 'packing' but it may also be modified to form transfer cells and is then involved in the local transfer of materials (Ridge, 1991). Sometimes companion cells in the small leaf veins, which are the most active sites of solute loading into sieve tubes, become specialised to form transfer cells (fudge, 1991).

2.2.2 Techniques of phloem sap collection

Various techniques have been used in the past to collect phloem sap from many plant species. One of the oldest methods uses the natural exudation property of certain plant species, which produce phloem sap after an incision is made to the tubes. This method is easy to use but is limited to only a small number of species, mostly woody dicotyledons (Zimmermann, 1957,1960; Milburn 1970; Richardson et al., 1982; Chino et al., l99l). Some researchers have used isolated sieve elements for their investigations on phloem transportation of metabolites. Isolating sieve elements from phloem tissue so that they 17 would be separated from other cells without losing their properties is an extremely difficult task, especially when account is taken of the high lability of the differentiating sieve tubes and their strong plasmodesmatal links with companion cells' Successful tapping of sap from phloem has been achived by the use of insects such as aphids and coccids which parasitize the phloem by inserting their stylet tips into the sieve tubes. Upon cutting the aphid away from the stylet which is left penetrating the sieve tubes, sap exudates can be collected. The stylectomy method is unsurpassed in many respects. Using this method it is possible to obtain pure sap, uncontaminated by traces of substances from the cut surface of other tissues. On the other hand there are limitations for this method such as low flow rate and long sampling time as well as that of finding a suitable insect. Another method of identiffing translocated material is the replacement of translocates by labelled substances of the same type. ßCOz is usually supplied to one leaf and va¡ious plant parts are subsequently analysed for radioactive substances (Aronoff, 1955; Swanson and El-Shishiny, 1958;Hale and Weaver, 1962). An alternative technique involves tapping the sap (phloem exudate) from differentiated sieve tubes dissected or punctured for the purpose. Phloem exudate provides a more complete picture of the composition of the substances present in the lumen of the mature sieve elements. The facilitated exudation technique was initiated by King and Zeevaart (1g74) and is based on the use of chelating agents e.g. (EDTA) in the collecting solution in which the cut end of the severed tissue is immersed. Chelators prevent blockage of the sieve-plate pores after wounding by their action in absorbing calcium, which is an essential element in the presence of P-proteins for the synthesis of the plugging material, callose' This technique is valuable in numerous herbaceous and woody species (Pate et a1.,1974l' GIad et al..l992a). It is easy to use and allows the collection of relatively large volumes of phloem sap considered to represent true phloem sap composition (Weibull et al., 1990; Girousse et al.. l99l) and sufficient for biochemical analysis particularly in plants on which insect stylets cannot be used. It is generally accepted that the exudates collected by these techniques are samples of sieve tubes contents, and using these methods, many analytical studies have been carried out to determine the composition of phloem exudates.

2.2.3 Leaves as sources of assimilates

Photosynthesis in the productive leaves of higher plants provides the assimilatary driving force for metabolism of the sink organs and sustains plant growth and development. Phloem transport in sieve tubes enables the transfer of assimilates from source tissues (mature leaves) to sink tissues elsewhere in the plant. Leaves are sites for photosynthetic 18 activity and the main source of carbohydrates. Assimilate production in the leaves depends on both compartmentation and distribution of assimilates between storage and transport compartments (Wardlaw, 1990). Export of photosynthate assimilates from source leaves is therefore dependent on the functioning of many complex metabolic events that control the production of solutes such as sucrose, which translocate in phloem, and delivery of these solutes to the phloem (Stitt and Quick, 1989; Hunter and Visser, 1989, 1988 a;Geiger and Fondy, 1991; Candolfi-Vasconcelos and Koblet, 1990). In some plants there is a relationship between sucrose synthesis and the rate of export from the leaf which suggests that the rate of phloem transport is controlled directly by the rate of sucrose synthesis (Stitt and Quick, 1989; Geiger and Fondy, 1991). In other plants leaf sucrose levels have no correlation with rates of export, suggesting that it may be phloem loading rather than sucrose synthesis which is important in determining export rates (Geiger and Fondy,lggl; Goldschmidt and Huber, 1992). Developing leaves, including those of grapevine, change from assimilate importers to exporters of photosynthetic assimilates at the time they achieve 30-50% of their maximal size (Hale and Weaver,1962: Turgeon and Webb, 1973; Fellows and Dalling, 1974;Twgeon et a1.,1975).

2.2.4 Retranslocation from reserves

At the early growing stage before leaf photosynthesis is sufficient, soluble carbohydrate reserves are mobilised from the woody tissues and translocated towards the growing buds via the xylem stream. Some carbohydrates may originate from the breakdown of the callose located in dormant woody tissues (Glad et al., 1992b; Aloni et al., 1991). The ripening grape berry is a strong sink for dry matter supplied by current photosynthesis or reserves in wood (Coombe, 1989), and mobilisation of sucrose from reserves to fruit can contribute signifìcant amounts of carbohydrates during fruit maturation and ripening (Candolfi-Vasconcelos et al., 1994; Williams and Biscay, l99l). Sucrose translocating in the grapevine could originate from either of carbon fixation or reserve tissue in the leaves or other perennial parts of the vine. Current leaf photosynthesis is the main source of sugar for maturation of the berry but remobilisation from stem reserves occurs under abnormal conditions such as defoliation (Matsui et al., 1979). Conradie (1980) analysing all parts of two-year-old vines showed that the majority of dry mass increase (about four-fold) in the ripening grapes was from photosynthesis with little change in the dry mass of other parts of the vine. Hunter and Visser (1988) showed that redistribution of photosynthates to vegetative organs is likely to take place at berry ripeness and that bunches were mainly supported by basal leaves. However, mobilisation of storage reserves in the wood of stems and roots may also contribute in the dry mass accumulation (Kliewer and Antcliff, 1970). 19

2.2.5 Transport of carbohYdrates

V/ater is quantitatively the most abundant substance transported in the phloem' Dissolved in the water are the translocated solutes, which consist mainly of carbohydrates in almost all of the plants which have been studied so far. The sugar composition of phloem exudates is highly variable from species to species. It has been firmly established plants, that sucrose is the main mobile component transported in the phloem of most with oligosaccharides (e.g. raffinose, stachiose, verbascose), and sug¿Ìr alcohols (e'g. sorbitol, manitol), are translocated in some species (Gamalei, 1985; Madore, l99I; Flowers and yeo, 1992; Van Bel, 1993). In the sieve element these account for at least 85% of all mobile compounds (Kursanov, 1984). Sucrose is the main mobile component, occurring at concentrations of 0.3-0.9 M. Once carbohydrate is transported from, SâY, leaves to fruits, considerable transformation may occur. For example, Table 2.1 shows the different metabolites in the toc}z- leaves and immature berries of grapes after feeding the leaves with labelled

Table 2.1 Translocation of photosynthetic products from leavesa to immature berries (from Peynaud and Ribéreau-Gayon, l97I).

Radioactivity (1 s)

leaves Berries Compound 48 hours 6 hours 24 hours 48 hours Total carbohydrates 265 l0 50 130 Total organic acids I l3 I 8 60 Total amino acids l2 0 I t7 Fructose 50 tr. t7 19 Glucose 154 tr. l4 6l Sucrose 53 tr. t7 38 Malic acid 104 0 6 4t 14 Tartaric acid 1 0 I Citric acid 2 0 tr. I u Leaves were fed 'oCOz during a 48 hour period. Analyses were made of the leaves and adjacent grape berries of the same vine-stock after various times of exposure as shown. tr. : trace. 20

2.2.5.1Sucrose

Ziegler (1956), reported paper chromatographic analysis of sieve tube exudates from some European trees which showed that the only sugar found was sucrose, and that hexoses and their products were absent. Zimmermann (1957), analysed exudates of l6 North American tree species and found sucrose to be predominant in a number of them, with some carrying sugars such as raffinose besides sucrose. In many plants, sucrose is the only sugar present, but other sugars, namely raffinose, stachyose, and verbascose, and occasionally, the sugar alcohols manitol and sorbitol may occur in some species (Zimmermann, 1960). Sucrose is the only sugar detected in the phloem sap of maize (Moss and Rasmussen, 1969, Maillard and Procher l99l), Ricinus cummunis L.(Hall and Baker, lg72), sugil beet (Bet vulgaris), ( Fondy and Geiger,1977); bean, (Giaquinta and Geiger, lg77), Pisume sativum (Barlow and Randolph, 1978), Oryza sativa (Kawabe et al., 1980), Vicia faba, (Delrot, 1981), Triticum aestivum (Hayashi and Chino, 1986), Medicago sativa (Girousse et al., l99l; Chino et al., 1991), and Banksia prionotes (Jeschke and Pate, 1gg5). The concentration of sucrose in the phloem sap of alfalfa is reported to be aboutg}yo of assayed sugars (Girousse et al., 1991). Sucrose is the main sugar produced by photosynthesis and translocated in grapevines. This compound may be derived from recent photosynthesis carbon fixation or be mobilised from reserve tissues in leaf or other perennial parts of the grapevine (Swanson and Elshishiny, i958; Koblet, 1977; Glad et al-,1992a).

2.2.5.2 Hexoses

Besides sucrose. however, glucose and fructose have been reported in amounts that seem to depend largely on the sampling site, experimental procedures, and times. All reports concerning sieve tube exudates, covering a total of about 45 species, stress the complete absence of hexoses even in chromatographically detectable traces and reports on displacement of naturally occurring translocates with 'oCOr-labelled photosynthates supports this (Zimmermann. 1960; Flowers and Yeo, 1992). There are indications that sucrose is the only translocated sugar in the grapevine, and that the hexoses are secondary products, derived from hydrolysis after sucrose has moved out of the actual translocational channels (Swanson and El-Shishini, 1958). It is possible that some sucrose molecules diffuse out of sieve tubes during normal translocation and are hydrolysed in adjacent cells to glucose and f¡uctose which are not normally mobile in phloem (Swanson and El-Shishini, I 958). 21

2.2.6 Transport of nitrogen compounds

Although nitrogen solutes frequently are the major components of dry matter in grapevine xylem exudates (Anderson and Brodbeck, 1989a; 1989b, Glad et al-, 1992b; lg94) they are also important in the phloem, being second only to carbohydrates (Pate, 1980; Glad et al., 1992a; Rubelakis-Angelakis and Kliewer, 1992). Phloem transport involves mostly the same nitrogen forms as xylem transport (Pate, 1980; Bray, 1983), but phloem exudates usually contain 10-20 times the xylem concentration of nitrogenenous solutes Although nitrate concentration may be high in xylem, it is often absent or much lower in phloem (Pate, 1980). Phloem translocation of nitrogen is important to supply these compounds to fruits (Pate, 1983). Nitrogen occurs in phloem largely in the form of amino acids and amides especially glutamate/glutamine and aspartate/asparagine. The amino acids glutamine, asparagine, glutamate, aspartate, alanine, and serine have been reported to be the main nitrogen-containing compounds in phloem sap of plants (Hall and Baker, 1972; Leckstein and Llewellyn, 1975; Pate, 1980; Hoking, 1980; Simpson and Dalling, 1981; Urquhart and Loy, 1981;Fukumorita and Chino, 1982; Fisher and Mac- Nicol, 1986; Chino et al., l99l;Feller and Hoelzer,Iggt; Girousse et al., 1991; Wiener et al.,l99l; Glad ef al.,1992b; Jeschke and Pate, 1995). Glad et al. (1992a), reported the nitrogenous solute composition of grapevine phloem sap (Vitis vinifera L. cv. Pinot noir). During flowering, glutamine was the major form of transported nitrogen in the phloem sap of the grapevine (60%) and proline the second (10%).

2.2.7 Transport of organic acids

Organic acids as photosynthesis assimilates translocate into conducting cells of leaves along with amino acids. but in a lower concentration. Malic acid, tartaric acid, citric acid and keto acids have been detected in phloem sap of some woody plants (Ziegler, 1962, cited in Kursanov. 1984). However, it is not clear whether these organic acids are transported directly from leaves as photosynthetic products or whether they are synthesised in sieve tubes themselves as a result of glycolysis reactions and tricarboxylic acid cycle (Kursanov, 1984). In phloem transport studies, experiments withlaCOz have shown that labelled organic acids are the earliest photosynthetic products emerging in conducting bundles after the leaves have been supplied with laCO2. The relative amount of organic acids among the laC assimilates entering the conducting cells is small, being equal to only 2.5o/o of labelled compounds compared with carbohydrates, (90olo), and amino acids, (7 .3%) (Kurasanov, 1984). 22

Tartaric acid, malic acid and citric acid were detected at levels of 170, 110, and 25 nmol respectively from phloem sap exudate of grape bunches at flowering time (Glad et al., 1992a).

2.2.8 T ransport of minerals

Potassium is the predominant cation in plant phloem exudates normally accounting for more than 60%o of the total cations (Kursanov, 1984). Bukovac and V/ittwet (1957) studied absorption and translocation of a number of foliar-applied radioactive isotopes. Using recovery in untreated plant parts as the control, they found Rb, Na, and K ions were the most readily absorbed and most highly mobile; P, Cl, S,Zn, Cu, Mn, Fe, and Mo ions were intermediate with decreasing mobility in the order given; while Ca, Sr, and Ba ions were absorbed into the leaf but not exported. In ripe grape berries, K+ constitutes 70o/o of the mineral cation content of the berry and 45o/o of this total accumulates in the berry skin (Somers, 1975; Hale, 1977; Iland and Coombe, 1988). Changes in K+ content in fruits is generally associated with phloem flow and in grapes is usually contrasted with Ca2+ accumulation. Therefore, change in xylem flow after veraison results in decreasing calcium accumulation in the berry,. Calcium ions move passively within the xylem via the transpiration stream, and the Ca2+ content in fruits has been used as an indicator of cumulative xylem flow ( Lang and Thorp, 1989; Morrison and Iodi. 1990). Creasy et al. (1993), using Pinot noir grapevines, showed that K+ in the berry increased rapidly after veraison whereas this was not the same with Ca2+ because of berr.v xylem discontinuities at the veraison stage. Calcium is well known to be transported only in the xylem (Marschner, 1983) while phloem sap contains little or no Ca and is not considered to be a medium for Ca movement (Pate, 1975). Additionally, among the cations, potassium has a role in stimulating the unloading of assimilates (Ho,

I 988). During the second growth period of grapevine there is an increase per berry in cations such as K*, Na*, Ca2* and Mg2- and in anions such as POo-1. During fruit maturation and ripening, mobilisation of K* from reserves to fruit increases in significant amounts (Williams and Biscay, 1991). Potassium was the main mineral ion found in the Chardonnay grapevine xylem exudate (7300pM), followed by calcium, phosphate, nitrate, sulphate, magnesium, and chloride (Glad et al., 1992b). Potassium was exuded at levels close to 800 nmol/ sample whereas nitrate was not detected (Glad et al., 1992a). They found that phloem sap was slightly alkaline and quoted the pH value of grapevine phloem

sap exudate as being 7 .7 (Glad et al., 1992a). Boselli et al. (1995) reported that the potassium content of berries of two cultivars inc¡eased to approximately 0.02-0.03 mg berry-r d-r during the entire period of fruit 23 development, but the concentration of calcium and magnesium in the berry reached weight basis' maximum at veraison and then decreased. Their results show that, on a fresh calcium the potassium level was constant during the berry development period whereas and magnesium levels decreased during the whole berry development period'

2.2.9 Transport of secondary metabolites

There is little information regarding the translocation of secondary metabolites in plants. Secondary metabolites such as polyphenols do not seem to be typically involved in phloem transport. Nevertheless, according to some reports small amounts of phenolic substances have been reported in phloem exudates (Kursanov, 1984 and references cited therein). The possibility of transport of large amounts of polyphenols in the phloem seems unlikely because, in differentiated sieve elements, the cytoplasm is not protected against the contents of the central lumen by a tonoplastic membrane but is partially dispersed in it; this creates conditions conducive to protein precipitation (Kursanov'

1 984). Considerable quantities of different secondary metabolites accumulate in grape berries during growth and ripening. The ability of the various tissues to accumulate secondary metabolites is specific to the developmental stage, the tissue, and cell type in the mature grape berry, and it seems that epidermal cells are the most appropriate site for the accumulation of secondary metabolites such as pigments, other non-pigment polyphenols, and flavour compounds. Considine ( I 979) remarks upon the occurrence of polyphenols in most pericarp cells at anthesis; those of the outer epidermis and the mesocarp stained differently. By day 16 after anthesis there were two cell types between the hypodermal layers and the position of the vascular bundles, one containing polyphenols and the other not. By day 26 there was a reduction in polyphenols in all except the dermal system. It seems that, despite the similar function of inner and outer mesocarp, cells of these tissues, they are metabolically dilferentiated in the hrst growth phase. The important question which arises here is whether or not secondary metabolites are actually translocating into grape berries via phloem sap and, if so, is this the mechanism responsible for the accumulation of flavour compounds in the grape berries. 24

Chapter 3

Distribution of glycosides in grapevine tissues

Contents Page

3.1 Introduction ...... 25 3.2 Materials and methods...... 25 3.2.1 P1ants...... 25

3.2.2 F ruit growth...... 25 3.2.3 Plant materials... 26 3.2.4 Tissue extraction .26 3.2.5 Glycosyl-glucose determination 27 3.3 Results ..28

3.4 Discussion...... 35 3.4.1 Leaves...... -...... 35 3.4.2 Bunches and berries..-.... 35 3.5 Conclusion...... 36 25

3.1 Introduction The aim of this study was to investigate the amounts of total glycosides as an indicator of total secondary metabolites in different tissues of grapevines. In particular, studies were made of the changes in concentration of glycosides in the leaf blades and petioles as well as in berries at various growth and physiological stages throughout the growing season for cvs Muscat Gordo, Suløna and Flame Seedless vines.

3.2 Materials and methods

3.2.1 Plants

Mature (approximately 1l year-old) vines of Vitis vinifera L., Muscat Gordo Blanco (syn. Muscat of Alexandria), Sultana (syn. Thompson Seedless) and Flame Seedless grapes were used in this experiment. The vines were grown in the V/aite Agricultural Research Institute experimental vineyard at Glen Osmond, South Australia.

3.2.2 Fruit growth

Inflorescences showin g 50% flower cap fall were selected on the same day (40-60 bunches for each cultivar), labelled, and the date of flowering recorded. For berry growth measurements, 25-30 berries of different size and from different sides of the bunch (4 to 6 berries on each) were labled and measured weekly on the vine for diameter. length. and berry deformability. At the same time every week, 10-15 berries were picked randomly from clusters and 'Brix were used to determine berry dimensions, berry weight and (total soluble solids) in all the three cultivars. Berry density was calculated by dividing berry fresh weight by berry volume calculated from measurement of th¡ee berry axes, length, smaller diameter and larger diameter (Coombe et al. 1987). BerÐ, deformability was measured 'Brix with callipers and juice was measured by hand refractometer. 26

3.2.3 Plant materials

Beginning at three weeks after flowering in the 1992-93 season, triplicate samples of fruits and leaves were taken at three week intervals up to veraison and then a sample at veraison, at early ripening and at ripeness. Mature leaves at the fourth to sixth nodes from the base of the shoot were picked and separated to leaf blades and leaf petioles. V/hole uniform bunches, including berries and bunchstem, were used as a fruit sample in this experiment and are labelled whole bunch to distinguish from 'berries'. A separate sample of berries was taken in 1992-93 for the determination of glycosyl glucose. A series of berry samples was also collected for cvs Muscat Gordo and Sultana in the 1994-95 season from before veraison until ripeness. All plant 'C samples were frozen in liquid nitrogen and stored at'20 until extraction'

3.2.4 Tissue extraction

Extraction was carried out as follows: Samples were immersed in liquid nitrogen and the solid frozen tissue was ground, first in a laboratory blender and then in a coffee grinder to give a ñne powder. For extraction of glycosides, the method used by Gunata et al. (1985a) was adopted after modification when it was found that 0.3 M sodium acetate buffer, pH 5.0 improved the recovery. Ground tissue sample (5g) was 'C. added to acetate buffer (40 mL) at 90 After pH adjustment to 5.0 by addition of NaOH (0.5 M), the solution was held for 5 minutes at 90'C, then cooled and centrifuged at 9000 x g for l5 min. aT2 C in a Sorvall superspeed RC2-B Automatic refrigerated centrifuge. After removal of the supernatant, the solids were re-extracted with 30 ml of hot buffer in the same way. The supernatants were combined for isolation of glycosides. A different procedure was used for ripe fruit. Whole bunch samples, which included the bunchstem, skin and flesh of the ripe berries were ground under liquid nitrogen then homogenized with an Ultra-Turrax Tl25 high speed homogenizer with an F25N dispersing head (Janke & Kunkel GmbH & Co., Germany). This mix was extracted with hot buffer as for other tissue samples. All extracts and juice samples were stored

at. -20'C until used for assay of glycosyl glucose (G-G). Muscat Gordo and Sultana berry samples from the 1995 season were processed differently. In this case a sample of berries was counted, weighed, homogenised and extracted with 50% ethanol as described in Williams et al. (1995) for G-G determination. 27

3.2.5 Glycosyl-glucose determination

The glycoside assay for determination of glycosyl-glucose according to the principles outlined by Williams et al. (1995) was adopted in this work. An outline of the procedure is shown in Fig. 3.1.

Sample Homogenized fruit extract or ground tissue extact or Juice

+ Isolation of glycosides Cl8 Reverse Phase cartridge

G ide fraction

+ Hydrolysis 1.5 M H2SOa, t h, 100'C

Glucose

+ Glucose analysis enzymlc assay

Glycosyl Glucose G-G value

Fig. 3.1 G-G assay steps 28

3.3 Results

In this study glycosidically bound secondary metabolites were isolated and measured for Muscat Gordo Blanco, Sultana and Flame Seedless cultivars. The measurements for each sample were calculated in pmol of glycosyl glucose per gram of tissue fresh weight. The means of measurements for each cultivar during the growth period are plotted against time after flowering in Figures 3 .2.1 to 3 , 3 .3 '1 .to 3, and 3.4.1 to 3. Figures3.2.4to 6,3.3.4. to 6, and 3.4.4. to 6 show changes in berry density, soluble solids and berry volume. G-G accumulation in berries was calculated on a per berry basis based on berry samples separate from the whole bunch samples (Figs.3.2.3 , 3.3.3 and 3.4.3). In some graphs standard enor (SE) of the mean is given by a vertical bar. In Muscat Gordo Blanco the concentration of glycosides in leaf blades and petioles increased gradually in weeks th¡ee to 14 after flowering and then increased sharply from week 14 to 18 (Fig. 3.2.1 and 3.2.2). There was a sharp decrease in the concentration of glycosides in whole bunch samples with the initial acceleration in berry volume, but the second phase of rapid berry growth (see Fig. 3.2.4) had no such effect. In Sultana the concentration of glycosides in leaf blades and petioles increased in a parallel fashion throughout the period of berry development (Fig.3.3.1 and 3.3.2)' Leaf blades had seven-fold higher G-G concentration compared to petioles. The changes in whole bunch glycosides showed a similar trend to that of Muscat Gordo berries, although the concentrations of G-G in Sultana were higher than were found in the Muscat berries in the first six weeks and declined to lower concentration at berry ripeness (compare Fig 3.2.3 with Fig. 3.3.3). In contrast to the patterns seen in Muscat Gordo and Sultana, the rate of increase in G-G concentration of Flame Seedless leaves and petioles showed no increase until the sample of ten weeks after flowering when berries were ripe (Fig. 3.a.1,3.4.2); concentrations then declined in the fourth sample. The G-G concentration of Flame Seedless whole bunches showed a decrease from before to after veraison followed by a small increase thereafter (Fig. 3.4.3). The graphs of G-G concentration in leaf blades, petioles andbunches from Figs.3.2, 3.3 and 3.4 are regrouped in Figs. 3.5 to illustrate the comparative changes in the three cultivars. Concentrations of G-G were much higher in leaf blades compared with petioles and fruit (Figs 3.5.1 to 3). Of the concentrations in leaf blades of the three cultivars (Fig. 3.5.1) levels were greatest in Sultana, intermediate in Muscat Gordo and lowest in Flame Seedless (except for one high value at week l0 when berries were 2OoBrix). G-G concentrations tended to increase in leaf blades during the berry development period in all cultivars, e.g. by 95o/o and 53Yo in Sultana and Muscat Gordo respectively. 29

The changes in the concentrations of G-G in petioles (Fig. 3.5.2) were, in general, simila¡ to those in leaf blades but at a lower level - at about one fifth. The levels in Sultana were not predominant as in leaf blades being lower at the early and late sampling times. Again, there was a single higher value at week 10 in the petioles of Flame Seedless, as seen in leaf blades, and also petioles showed the same tendency to increase over the measurement period, Sultana by 98% and Muscat Gordo by 157%' The pattern of change in the G-G concentration in whole bunches (Fig. 3.5.3) was quite different to that shown by leaf blades and petioles. In all three cultivars the levels declined from their highest level at the first sampling (three weeks after flowering, just after berry setting). The decline was most rapid until veraison (eight weeks after flowering in Muscat Gordo and Sultana and five weeks in Flame Seedless) after which the levels steadied during berry ripening. Concentrations were similar in Muscat Gordo and Flame Seedless but showed greater extremes in Sultana fruit, being highest early and declining to lowest levels at berry ripeness. This pattern of decrease in G-G concentration in whole bunches was shown to be reversed when the G-G was determined on as a per berry basis (Figs. 3.6 and 3.7)' This determination was made on berry samples taken separately from the whole bunch samples but from the same vines. They are graphed in Fig. 3.6 against "Brix of juice (as an alternative indicator of berry development). The graphs show that levels of G-G per berry at the first sampling (5"Brix) started at levels of < 0.02 pmoles, then increased considerably. The increase was greatest in Muscat Gordo eventually reaching a level of 0.74 pmoles G-G per berry. On the other hand the amount in Sultana berries at the second sampling (6'Brix) had risen to 0.2 pmoles but, unexpectedly, showed no further increase thereafter. As a check on these results, a second series of samples of Muscat Gordo and Sultana were taken in the 1994-95 season and are shown in Fig 3.7. Thcsc results showed similar trends to those in Fig. 3.6 except that the values of G-G per berry were somewhat higher than those in 1992-93. t6 50(Xt 30 1 4 t5 0 E ¡ooo Ì E l4 o

Er l3 5 .o* o o E 12 \ t 2000 o q¡ ct 11 (5 1000 10

I 0 0 2 ¡f 6 I l0 12 14 16 18 20 22 24 0 2 4 6 I r0 12 14 16 18 20 Weeks alter f lowerlng Weeks after llowering

4.0 1.2

2 cl 5 , 3.5 o È cD i 3.0 1.t Er caD o õ 2.s t, E o 9(5 2.0 ¡t 1.0 l5

1.5

1.0 0.9 02468 10 12 14 16 18 20 22 24 0 2 1 6 I 10 12 14 't6 Weeks after flowering Weeks after llowering

1.6 1.0 30 + 3 + 6 1.4 ?5 ¿ 0.8 ì !

¿12 ct o 20 0.6 tt cD :\ x O L r< o F E co'" E .É 0.4 =- o 0.8 Or^ (J (, F 0.2 0.6

0.4 0.0 0 0 2 4 6 I 10t2141618202224 o 2 4 6 8 1012141618202224 Weeks alter llowerlng Weeks atter llowerlng

Figs3.2.1 to6EvolutionofG-Ginleafblades(l), leafpetiole(2)andfruit(3)and changesinberryvolume(4),berrydensity(5),andfruitsolublesolids(6)inMuscat Gordo with time in weeks after flowering. G-G data in Fig. ( 3 ), are expressed as concentration in macerated whole bunch on a fresh weight basis (left axis) and in mg per separated berry (right axis). Vertical bars thruogh mean values represent SE.A: inception of "Brix increase. t000 31 2 1 4

t aoo ' E Þtg ts ol Ër* =o E Et¿ o (t I o ct to l0 200

6 0 0 2 4 6 I 10 12 14 16 l8 20 0 2 1 6 I t0 12

Weeks after f lowering Weeks alter f lowering

30 1.2 2 5 ¿ o ì z.s j oE I .'l Et cD È z.o E =can oo (5 1.0 (, 1.s o to

1.0 0.9 0 2 4 6 I l0 12 14 16 18 20 o 2 4 6 I f0 12 14 16 Weeks after f lowering Weeks after f lowering

3.0 0.30 30 + 3 --+-- 6 ¿ o¿5 25 È lj an P ozo j 20 ô x an \ o '" o n,. E ; 1s E 1 ó 1.0 o 10 o 0.10 ó g ,

0.5 0.05 5

ó

0.0 0.00 0 0 2 4 6 8 l0 12 14 16 18 20 o 2 4 6 I t0 12 .t4 16 .tI 20 22 Weeks after flowerlng weeks afler llowerlng

Figs 3.3.1 to 6 Evolution of G-G in leaf blades ( I ), leafpetiole ( 2 ) and fruit ( 3 ) and changes in berry volume ( 4 ), berry density ( 5 ), and fruit soluble solids ( 6 ) in Sultana with time in weeks after flowering. G-G data in Fig. ( 3 ), are expressed as concentration in whole bunches on a fresh weight basis (left axis) and in mg per separated berry (right axis). Vertical bars thruogh mean values represent SE. A: inception of "Brix increase. 20 r200 32

1 4 t000 ¿t6 o I E j E soo

12 o ctt E : 600 o o E 18 à 4oo o o (, c¡ 4 200

0 0 2 4 6 I t0 12 2 4 6 I 10 12 14 Weeks after llowering Weeks after flowering

3.0 1,2 2 5

2.6 o E È o 1.1 22 Et Eì

o =al, E tr 1.8 !o o 1.0 (5 o 1.4 lt

1.0 0.9 0 2 4 6 I 't0 12 14 0 2 4 6 I 10 12 14

Weeks af ter f lowering Weeks after flowering

25 + 0.5 D 1.8 3 --o-- b 20 = 0.4 j

1.4 q) ct) -o :- 15 o uJ _ x E sù 1.0 1 (5 02 ro (5 ó (tI 06 01 5

0 o.2 00 't0 o24 6 I l0 12 1{ 4 6 I 12 14 Weeks after llowering Weeks af ter f lower¡ng

Figs 3.-f.1 to 6 Evolution of G-G in leaf blades ( I ), Ieaf petiole (.2 ) and fruit ( 3 ) andchangesinberryvolume(4),berr,vdensity(5),andfruitsolublesolids(6)in Flame Seedless with tinre in weeks after flowering. G-G data in Fig. ( 3 ), are expressed as concentration in r.vhole bunches on a fresh weight basis (left axis) and in mg per separated berry (right axis). Vertical bars thruogh mean values represent SE.A: indication of 'Brix increase. 33 21 1 ¿20 = j

Ên 16 o E1p o t5

4 0 2 4 6 I 10 12 14 16 18 20 22 24

Weeks atter flowerlng

4 2

I

3 Et o ts a o

0 2 4 6 I 10 12 14 t6 18 20 22 24

Weeks after f lowering

3 3

Muscal Gordo -3 -.È-¡ts Sultana + Flame Seedless õ)

o E

(J (,

0 0 2 4 6 I 10 12 14 16 18 20 22 24

Weeks af ter llowerlng

Figs 3.5.1 to 3.5.3 Glycosyl glucose (G-G) concentration in different tissues of all th¡ee cultivars comparing G-G in leaf blades (Fig.l), in leaf petioles (Fig.2 ) and whole bunches (Fig.3). M, Muscat Gordo; S, Sultana; F, Flame Seedless. 34

1.0

0.8

Muscat Gordo o 0.6 '.c¡

=\o E 0.4 Flame Seedless (, I (5 0.2 Sultana

0.0 0 5 10 15 20 25 30 'Brix

Fig. 3.6 Glycosyl glucose concentration in berries of Muscat Gordo, Sultana and Flame

Seedless plotted against juice 'Brix (data from 1992-93).

1.2

1.0

Muscat Gordo ¡o) 08

o 0.6 E J (, Sultana I 0.4 (5

o.2

0.0 E 'I 0 'I 5 20 25 30 'Brix

Fig. 3.7 Glycosyl glucose concentration in berries of Muscat Gordo and Sultana plotted against juice 'Brix (data from 1994-95). 35

3.4 Discussion

3.4.1 Leaves

Leaf btade. Leaf blades are known to be highly active sites of photosynthesis and carbon assimilation. Tazaki et al. (1993) reported that glycosides are a major fraction of leaf extracts in a wide variety of plants. In this study the results showed that grapevine leaves accumulate large quantities of glycosylated compounds. Although it is also known that grapevine leaves begin to transport assimilates after gaining about 50Y, oftheir final size (Hale and Weaver,1962; Turgeon and Webb, 1973; Flowers and Geiger, I974; Turgeon et al., I975), the results reported in this chapter show that the storage of glycosylated compounds continues after the leaves mature and reach their maximum size (see Fig. 3.2.I,3.3.1 and 3.4.1). This could be due to storage of excess of translocating metabolites or it may be due to physiological changes in the leaves during their senescence development. At the end of the growth period and during fruit ripening, leaves showed different patterns of glycoside evolution depending on the cultivar. Muscat Gordo and Sultana leaves contained highest levels at the late berry ripening stage with the former showing the sharpest increase in concentration (see Fig. 3.2.1 and 3.31). In contrast, glycoside concentration in Flame Seedless leaves decreased at the late ripening stage (Fig. 3.a.1). Because it is an early ripening cultivar the leaves of Flame Seedless may have reached advanced senescence at the last sampling time. The G-G levels of leaf tissue are compa¡ed in Fig. 3.5.1.

Leaf petiole. Leaf petiole is considered as a passage for the assimilates transported from the leaves to other parts of plant. In this experiment leaf petioles for all three cultivars analysed. showed a pattern of concentration of glycosides similar to that of the corresponding leaf blades. however, the concentrations were at levels of about one fifth of those found in the leaf blades. (see Fig. 3.2.1-2,3.3.1-2 and 3.4.1-2). Reasons for the lower levels of glycosides in the petioles could be due to firstly, petiole tissue is not as active photosynthetically as leaves are, and secondly, the petiole tissue, which connects the vascular bundles of leaves to the shoots, is not known as a storage tissue. The G-G levels of petiole tissue are compared in Fig. 3.5.2.

3.4.2 Bunches and berries

Measurement of glycosides in whole bunches cannot be converted directly to per berry because of the presence of peduncle tissues in bunch samples. The data reportted in Chapter 7 suggest that secondary metabolites are abundant in peduncle tissue. This observation is supported by calculation of concentration of G-G per g berry weight 36

(dividing the amount per berry by berry fresh weight- data not shown) with G-G per g fesh weight of whole bunch. Such comparison shows that the berry samples had concentations 18, 19,23,26,34,and32%of bunch samples at successive sampling dates. These calculations do not detract from the following conclusions. Glycoside concentration in the whole bunches of all three cultivars decreased from berry set until veraison (Fig 3.4.3) but glycoside amount per berry increased (Fig. 3.7). This reversal is explained by the large amount of berry growth that occurs during this period of berry development, especially in the large-berried Muscat Gordo for example Muscat Gordo berries at the first sampling averaged 0.47 g fresh weight but 3.41 g at the last sampling. In other words, the decline in G-G concentration in fruit, which differs so markedly from the increase shown in leaf blades and petioles, \ryas not reflected in a decrease in the amount of G-G per berry but to the dilution of an increasing weight of G- G by a proportionally greater increase in water and dry mass in the developing berry. Leaf blades and petioles showed small changes in fresh weight per organ; for example, the range of fresh weights of Muscat Gordo leaf blades and petioles only fluctuated + 9yo and t 4o/o in all samples. These small differences would not have altered the general pattern of increase in G-G shown in the concentration values shown in Figs. 3.5'l and 3.5.2. Thus all three organs - leaf blades, petioles and fruit - showed an accumulation of G-G per organ during the berry development phase of these vines. This differs from report of the high accumulation of secondary metabolites in immature plant organs (Stahl- Biskup et al. 1993). Given the high percentage of secondary metabolites that are glycosylated (Barz and Koster, l98l;Kurt and Torssell, 1983), the measurement of glucose released from isolated glycosides, as done in the G-G assay, appears to be a convenient index of total secondary metabolites in plant tissue'

3.5 Conclusions

During the berry ripening period of a grapevine's yearly cycle of development, the concentrations of glycosyl-glucose (G-G) increase in leaf blades and petioles but decline in fruit. This difference in pattern is shown to be due to the large weight increase that takes place in berries during this period compared with leaves; in fact on a per berry basis, the weight of G-G also increases during its development. Comparing cultivars, the amount of G-G per Muscat Gordo berry about doubles as oBrix increases from 6 to23, while for Sultana there is only a small increase over this period; this reflects their relative flavour status. The glycosyl-glucose assay shows promise as an indicator of the content of total secondary metabolites in plant tissues. 37

Chapter 4

The inVolt'ment of leaves in the accumulation of flavour compounds in Muscat Gordo grape berries

Contents Pase

4.1 Introduction ......

4.2 Materials and methods... -

4.2.1 Plarft materials.. -

Glas s house experiment. Field experiment. A.2.2Methods of growing plants and grafting bunches..

P re-ro oted cuttings...... Grafting bunches from field grown vines onto potted vines 40 Chip-budding onto pre-rooted cutting...... 40 Bench-grafiin4...... 4l 4.2.3 Sample preparation..-....-..... 44

4.2.4 G as chromatography-mass spectrometry (GC-MS)...... -...... 44 4.3 Results 45 4.4 Discussion...... 5l 4.5 Conclusion...... 52 38

4.1 Introduction

Monoterpene compounds contribute significantly to the characteristic flavour of floral grapes and are generally present only at low levels in non-floral varieties. These aroma components, which are cornmon constituents of many fiuits, are present in free form and, more abundantly, as non-volatile glycosides (Williams et a1.,1982a, V/ilson et al., 1984, Gunata et al., 1985a). The sites of synthesis of these flavour compounds in grapevines are unknown. There is limited information concerning the composition of secondary metabolites in grapevine leaves. Wildenradt et al. (1975) reported six-carbon compounds and also terpenes and their derivatives at high levels in grapevine leaves. Gunata et al. (1986) reported the presence of terpenols and aromatic alcohols in free and bound forms in Muscat of Alexandria (syn. Muscat Gordo Blanco) leaves in much higher concentration than in berries. Gunata et al. (1986) also analysed Muscat Gordo leaf blades and petioles for terpenols and reported that, despite having the same profile of these compounds, the changes in levels of terpenols in the leaf blade and petiole were different during the period of fruit development. In the leaf blade the levels of terpenyl glycosides increased progressively during grape maturation, whilst in the stalk both free and bound forms of terpenols decreased. Their results also showed that the leaf blade was higher in free and bound aromatic alcohols than either the petiole or the berry. They raised the question that if the petiole is a passage for metabolites transported from leaves to other parts of the plant, the changes in concentration of terpenols in the petiole would be similarto those in the blade and if the glvcosides are considered a preferential form of transport then the petiole would contain more of the bound forms. They concluded that if these compounds are translocated from the leaves to the berries, this was not reflected in the levels of the compounds in the various plant parts during the season. Bravdo et al. (1989) after feeding Muscat Gordo grapevine leaves with laC-labelled mevalonic acid found that both leaves and berries contained labelled monoterpene glycosides l2 to 72 h after feeding. The authors suggested that monoterpene glycosides are translocated from leaves to berries. Skouroumounis and Winterhalter (1994) found C¡3 norisoprenoids present in Riesling grapevine leaves in concentrations l0-100 times higher in Riesling wine. None of the references suggested anv clear mechanism for translocation of secondary metabolite compounds between leaves and berries. 39

In a textbook on viticulture, Winkler et al. (1974, pp.l70-171) described results of an unpublished experiment which implied that grape berries have the capacity to produce flavour compounds directly. The experiments repofed, through informal tastings only, the retention of floral flavour characteristics in clusters of Muscat berries grafted and grown on vines of a non-floral cultivar, and the non-development of Muscat flavour in neutral fruit grafted and grown on Muscat vines' Since knowledge about translocation of secondary metabolites in Vitis vinifera grapevine cultivars is lacking, the aim of this study was to investigate the involvement of leaves as a source of metabolites and their influence on the composition of the fruits. The development of analytical techniques for the analysis in grapes of glycosidically-bound monoterpenes (Wilson et al., 1984) provides methods for objective investigations into the sites of biosynthesis of flavour compounds in Vitis vinifera L. The experiments described in this chapter used these techniques combined with bunch grafting of fruit to determine if monoterpene flavour compounds are synthesised in grape berries or translocated into them.

4.2 Materials and methods

4.2.1Plant materials

Glasshouse experimenl Cuttings of Muscat Gordo Blanco and Shiraz vines were gro\¡rrt in the glasshouse by the method of Mullins and Rajasekaran (1981) to produce fruit. Two pre-rooted cuttings, one of each cultivar, u,ere planted in a pot close together. One of the rootlings in each pot was allowed to produce a shoot and the other trained to produce an inflorescence. About two weeks before flowering the inflorescence of one vine was grafted onto the shoot of the other by approach-grafting (Fig. a.l ). After five weeks, by which time berries had formed, the bunch was cut off from its parent plant and allowed to develop on the other cultivar's shoot until the berries were ripe (Fig. a.2). Control plants were left ungrafted and grown under the same conditions. Berries were harvested and kept frozen at -20"C until they were used for extraction.

Field experiment In the following season. bunches were grafted using pairs of shoots on mature Muscat Gordo and Sultana vines grown side by side in the same row so that their shoots intermingled. The timing and method of grafting were similar to that used in the Muscat Gordo/Shiraz glasshouse experiment. Six bunches were set up for each of four treatments (Muscat on Sultana. Muscat ungrafted, Sultana on Muscat and Sultana ungrafted). 40

Duplicate bunches were harvested at three ripening stages þre-ripe, early ripe and very ripe).

4.2.2Methods of growing plants and grafting bunches

Pre-rooted cuttings For the glasshouse experiment dormant cuttings were obtained from the field before bud burst. The cuttings \ryere the basal part of canes, about 25 cm in length. All the buds were removed except the two uppermost buds, and cuttings were planted in a heated rooting- bed containing heating cable and coarse sand in a box. The rooting box was in a room with the air temperature held at 4 "C. The cuttings were planted so that the two upper buds remained out of the sand and the lower end was in the wann sand. Tap water was used for watering every two days. After about four weeks, two of the rooted cuttings, one from each cultivar, were planted in a pot of 20 cm diameter filled with a6:3:l mixture of perlite, vermiculite and peat moss and placed in the glasshouse and watered with Hoagland solution. When buds burst the method of Mullins and Rajasekaran (1981) was used retaining only one bud to obtain inflorescences on the donor vine. For the host vine, one of the buds was allowed to grow and the inflorescences were severed when they appeared on the newly growing shoot.

Grafting bunches fromfield grown vines onto potted vines Because of the low rate of fruiting Muscat Gordo cuttings obtained by the method described above another method was developed to graft bunches from field grown vines onto potted vines. In this method, two-year-old potted vines were kept outdoors during winter, and at the time of bud burst, were trained to give one or two shoots, depending on the vine's vigour. Before the clusters on the field grown vines had flowered, these pots were tied under the vines so that their shoots could intermingle with the field vines. The fruit cluster from a freld vine (donor) was grafted onto the potted plant (host) by approach-grafting and after three weeks the donor shoot tip was removed. Five weeks from grafting, the bunch was cut off from the donor plant and the potted plant with grafted bunch was transferred to the glasshouse and grown until ripening (Fig. 4.2 and 4.3). This method was most convenient and was virtually 100% successful.

Chip-b udding onto pre-rooted cutting A chip-budding method was used on pre-rooted cuttings and, after transferring a dormant bud from one cultivar onto a pre-rooted cutting of the other cultivar, the grafted plants were kept in a glasshouse and treated according to the method of Mullins and Rajasekaran (1981) and maintained as described before. Using this method, inflorescences 41

\/ere obtained from the scion after the grafted bud grew (Appendix Fig.l) but they failed to develop into berries after setting.

Bench-grafting Cuttings of different cultiva¡s were grafted using the bench-grafting technique and then the graft region and the basal end of the grafted cuttings was held at 25" C under wet sand for two weeks to produce a callose at the graft area to improve the success of grafting. After a callose was developed grafted cuttings were planted in pots and placed in the glasshouse. This method failed to produce any successfully grafted plants.

à Host vine Donor vine

I

I

Donor vine cut+ off

Ungrafted Grafted I fru¡t ripens lru¡t r¡pens

I Analyse frurt Analyse fruit

Fig. 4.1 Transfer of grape bunches by approach-grafting. Before flowering, the donor vine with bunch (A) was grafted onto the host vine (C). The donor vine was cut off five weeks later (D) and the fruit was allowed to ripen on the host vine (E). Ungrafted vines were grown under the same conditions as a control (B). 42

¡lig.4.2 Glasshouse grown Muscat Gordo grapes bunch grafted onto Shiraz vine 43

f ig. 4.3 Glasshouse grown Shiraz grapes bunch grafted onto a Muscat Gordo vine. 44

4.2.3 Sample preparation

The fruits of the Muscat Gordo grafted onto a Shiraz shoot and Shiraz grafted onto a Muscat Gordo shoot, plus ungrafted Gordo and Shiraz (control), were used for monoterpene analysis. Berries were weighed after the seeds had been removed and were homogenised in redistilled water purified in a Milli-Q apparatus (Waters, Millipore Corp.) using an Ultra-Turrax homogeniser. The homogenate was centrifuged at 9,000 g at 4'C for l5 min and the supernatant decanted and retained. The centrifuged pellet was resuspended in redistilled water, homogenised, centrifuged, decanted and the two lots of supernatants were combined. After the addition of 200 nmol each of n-octyl ß-D-glucopyranoside and n-dodecyl ß-D- glucopyranoside as internal standards (IS) (10 nmol for the ungrafted Shiraz), the supernatant was passed through a C-18 reversed-phase adsorbent cartridge (Sep-Pak, Waters) to adsorb free monoterpenes and their glycosidically bound forms. The C-l8 cartridge was washed three times with l5 mL of the Milli-Q water. Then the sample was eluted twice with 0.5 mL of redistilled ethanol and evaporated to dryness under a stream of nitrogen. The residue was redissolved in l0 mL of an aqueous pH 5 buffer [mixture of 0.1 M of aqueous citric acid solution, 0.2 M of aqueous disodium hydrogen phosphate solution and sodium azide (final conc. :0.02%)l and then incubated at37"Cfor 17 h with 0.5 mL of an e¡¿yme solution made up with I mL of Novoferm 12 ( Novonordisk, Switzerland) in 25 mL of redistilled water. The hydrolysate was diluted with redistilled water to 20 mL and then extracted continuously with redistilled dichloromethane at 55"C for 24 h using a liquid/liquid extractor, The solvent extract was concentrated by distillation through a column packed with Fenske's helices and then subjected to GC-MS analysis.

4.2,4 G as ch ro m ato gra p h-v-- m as s s pe ct ro m etry (G C-MS)

A Finnigan Mat TSQ-70 mass spectrometer was interfaced with a Varian model 3400 gas chromatograph. A DB-1701 capillary column (J&W Scientific) of 30 m x0.25 mm and frlm thickness of 0.25 pm, was used. The injector and transfer line temperatures were 200'and 250'C, respectively. The concentrated extract was injected using splitless mode venting after 0.5 min. The column was held at 50'C for I min, programmed at 5"C/min to 250"C, and held at this temperature for 20 min. The ionisation energy and emission current for electron impact mode vn'ere 70 eV and 0.2 mA. Mass spectra were taken over mlz 35-350 in 0.5 sec. Monoterpenes were identified from their mass spectra and retention times by comparison with data obtained previously in this laboratory (Strauss et al., 1986, 1988; Winterhalte¡ et al.. 1990). 45

peak The concentration of the identified monoterpenes wÍls estimated by comparing the The area of the individual compounds relative to those of internal standards (IS). calculations were based on the two assumptions that the individual monoterpenes had the same behaviour during sample preparation and the same response factor in GC-MS the analysis as those of IS. Analyses of all samples from the first experiment and some of grafted samples from the second set were done in duplicate'

4.3 Results

Figure 4.3 shows typical GC-MS chromatograms of the aglycones from berries of Muscat Gordo bunches grafted onto Shiraz shoots, ungrafted Muscat Gordo fruit and ungrafted Shiraz fruit from the glasshouse experiment. The differences in amounts of the monoterpenes in the samples are evident from these chromatograms. The compounds were identified from their mass spectra and are shown in Fig. 4.4 with their concentrations in Table 4.1. The large differences in monoterpene composition between Muscat Gordo and Shiraz is emphasised by the data for the respective controls. Summation of the data in Table 4.1 showed an approximately 50 times higher level of total monoterpenes in Muscat Gordo compared with Shiraz. Most monoterpenes were either not detected or were present in only trace amounts in Shiraz grapes. Importantly, the results show that the Muscat Gordo berries produced the same range of monoterpene compounds regardless of the type of vine on which they were grafted. In fact, for this experiment the levels of most compounds determined were 2-8 times higher in the grafted fruit, possibly relating to differences in ripeness between the grafted (26.2'Brix) and the control (21.0 'Brix) grapes. Shiraz fruit from the grafted vines, which was available in only a small amount, was essentially unchanged in composition from its control. GC-MS analysis of Shiraz berries grafted onto Muscat Gordo shoots was repeated in 1995 using more berries for extraction. Chromatograms of aglycones from GC-MS analysis of Shiraz berries grafted onto Muscat Gordo and ungrafted Shiraz berries from this experiment are shown in Fig.4.5. The results showed the same pattern as the previous results with detected compounds in higher concentration in the Shiraz fruit (Table 4.1). Results of the field experiment using grafted and ungrafted (control) Muscat Gordo and Sultana fruits are shown in Fig. 4.6. Like Shiraz, the Sultana contained monoterpenes at levels of less than one tenth of Muscat Gordo berries. As in the glasshouse experiment, the Muscat Gordo fruit had similar monoterpene profiles in the grafted and ungrafted berries. Grafting of Sultana onto Muscat Gordo vines had no influence on the monoterpene composition. The concentration of individual compounds varied during ripening of the 46 berries (Fig. 4.6), generally showing a large increase, especially in the more abundant compounds, as total soluble solids rose from 10-15 to 25-28 'Brix. GC-MS chromatograms of these samples are shown in figures 4.7 and 4.8.

Table 4.1. Concentrations of monoterpene compounds identified in ripe grafted and non-grafted berries from the glasshouse experiments. Peak numbers are shown in Fig. 4.4.

Levels (nmol / e fresh weieht)

Peak Muscat Muscat Shiraz Shiraz Shiraz no. Gordo on Gordo on Muscat Control on Muscat Shiraz Control Gordo * Gordo (1995)

I 1.8 0.8 n.d. trace trace 2 7.4 2.6 n.d. n.d. n.d.

J 0.8 0.4 n.d. trace n.d. 4 6.6 2.2 n.d. n.d. n.d.

5 0.4 0.07 n.d. n.d. n.d.

6 0.5 0.1 n.d. n.d. n.d. 7 3.4 1.7 n.d. n.d. n.d.

8 24.8 15.9 n.d. trace 0.4 9 34.5 13.2 trace n.d. 0.2

10 0.9 n.d. n.d. n.d. n.d. ll 4.9 0.8 n.d. n.d. n.d. t2 6.7 t.7 n.d. 0.2 0.1

13 3.6 1.9 n.d. 0.5 2.7 l4 8.8 2.1 n.d. n.d. n.d. l5 0.3 n.d. n.d. n.d. n.d. l6 5.3 1.7 n.d. 0.1 3.4 Total I10.7 45.17 trace 0.8 6.8 "Brix 26.2 2l .0 24.0 22.0 22.2 n.d.: not detected trace: less than 0.05 + Due to the small amount of sample the compounds levels were low in the extract Shiraz controt 47 40

30 12 I

3 20 f3 I 16

10

'r00 I tJîtf: I

5o 9 IS

Muscat Gordo AO control

1o I

) 12 20 7 16 4 14 1 I 1 10 \ I yI

0 5 )UU l00a_, 1100

51 q 12 IS I fS 40 Muscat Gordo grafted

¿ )O

11 > 16 4 2n / 7 11 \ 15

t0 10 \ 6 \ ü {

c'0Ct l(¿î:r. t:00 f ig. 4.4 Typical GC-MS chromatograms of the aglycons from Shiraz control berries and from fruit of Muscat Gordo control and Muscat Gordo grafted onto Shiraz. The y-axis scale is 'relative ion current' (RIC). 48

OH

OH o I ) 3 416

OH OH OH

OH

OH OH OH

7 8 9 r0 ll OH OH OH OH

OH OH OH HO OH

t2 l3 t4 15 t6

Fig. 4.4. Identity of l6 monoterpene compounds from GC-MS ch¡omatograms (see Fig. l): 1: Furan linalool oxide, 2: Linalool, J: Hotrienol, 4: Pyran linalool oxide isomer, 5: cr- Terpineol, ó: Pyran linalool oxide isomer, Z: Nerol, 8: Geraniol,9:2,6-Dimethylocta-3,7- diene-2,6-diol,I0:3,7-Dimethylocta-1,6-diene-3,5-diol ,1122,6-Dimetþlocta-1,7-diene- 3,6-diol, I2:Z-2,6-Dimethylocta-2,7-diene-1,6-diol,I3:E-2,6-Dimethylocta-2,7-diene- 1,6-diol, I4z 3,7-Dimethylocta-2-ene-1,7-diol, 15: p-Menth-l-ene-6,8-diol, 1ó: p-Menth- l-ene- 7,8-diol. s0 I rs 49 {0 Muscat control !0 I

12 20 ? ? 16 4 t{ I 6 l0 11 \ I ! l

0 500 100f-, t.,00

50

a1 Shiraz control

IS

)o 12

3 20 13 t6

l0

r.0Ð l7(tt: 1500

S

S

Shiraz grafted

I

1i

I 9 *¡ I tl

Figure 4.5 Typical GC-MS chromatograms of the aglycons from Muscat Gordo control berries and from fruit of Shiraz control and Shiraz grafted onto Muscat Gordo. The y-axis scale is 'relative ion current' (RIC). 50

20 20 'Brix A B 16 ABCD 16 Pre-ripe 15.0, 14.4, 11.2, 10.4 .9 o El Early ripe 18.0, 17 .8, 17 .8, 17 .8 28.0, 25.0, 26.2,28.o 7tz 12 fl Ripe -c tJ) 0.) ¡8 I oo o4 4 E (õ 9^ o 0 c 820 20 c .9 c D Ero 16 c ()o c 12 (Jo12 tt cq) qe8 8 (¡)

54 4

0 -",Jl 0 JI rJlr* 1 2 3 4 5 6 7 I I 10 11 12 1314 15 16 1 2 3 4 5 6 7 I I 10 11 12 13 14't5 16

Peak number

Figure 4.6. Histograms of monoterpene aglycones in berries (shown in Fig. 4.2) from the field experiment of Sultana ungrafted (A), Sultana grafted on Muscat Gordo (B), Muscat Gordo ungrafted (C) and Muscat Gordo grafted on Sultana (D), at three ripening stages. 51

4.4 Discussion

Grape berries contain a complex mixture of secondary metabolites which contribute to the a¡oma and flavour of the fruit and provide the chemical basis for varietal characteristics of the grapes. Monoterpene compounds are particularly abundant in floral grape varieties such as Muscat Gordo and most of these compounds are present as glycosides (Gunata et al., 1985b, V/illiams et al., 1982a, Wilson et al., 1934). In this present study, the total levels of monoterpene glycosides in Muscat Gordo fruit were approximately 50 times that of the non-floral varieties. Indeed, many of the monoterpene glycosides present in the Muscat Gordo fruit were not detected in berries of the non-floral varieties. In the grafting experiments, the same spectrum of monoterpene glycosides was present in Muscat Gordo fruit, irrespective of whether the fruit developed on its own vine or following bunch grafting onto Shiraz vines (Table 4.I, Fig. a.3) or Sultana vines (Fig. 4.5, 4.6). The amount of each of these monoterpene glycosides present in grafted Muscat Gordo fruit was equal to or even greater than that in ungrafted fruit. The quantitative differences observed, probably relate to differences in ripeness of the fruit, as these compounds are known to continue to accumulate as the fruit ripens (Gunata et al., 1985b, Wilson et al., 1984). Of equal importance, grafting of fruit of the non-floral varieties Sultana and Shiraz onto Muscat Gordo vines did not increase the low level of monoterpene glycosides in these grapes, nor did it induce accumulation of any monoterpene compounds different from those of ungrafted berries (Table 4. l, Fig. 4.3 and Fig. 4.6). These results confirm the earlier deductions made by Winkler (Winkler et al., 1974) and suggest that the ability to accumulate monoterpene glycosides resides within grape berries and is independent of the rest of the vine. As with many secondary compouncls, monoterpene glycosides are also present in grapevine leaves (Gunata et al., 1986; Bravdo et al., 1989). These compounds ma),be synthesised in the leaves and transported to the berries or they may be fbrmed in the berries themselves from precursor molecules imported via the phloem sap. The results indicate that the monoterpene glycosides are actually synthesised in the berries. This would suggest that the enzymes required to synthesise these compounds are present in the berries and that some or all of these enzymes are absent or inactive in fruit of non-floral varieties. It is, however. still possible that the monoterpene glycosides are synthesised elsewhere in the grapevine and imported into the fruit. If this is so, the Muscat Gordo berries must have mechanisms to selectively import and accumulate these compounds while these mechanisms are presumabl¡- not active in grapes of the non-floral varieties. The grafting experiments do indicate that the inability of the non-floral grapes to accumulate monoterpene glycosides does not result from a lack of these compounds or their 52 precgrsors in the phloem sap circulating around the vine, since fruit of Muscat Gordo grafted onto these varieties still produced the normal levels of monoterpene glycosides. The dependence on the rest of plant for synthesis of metabolites in fruit may be less than is currently accepted.

4.5 Conclusion

These experiments reported in this chapter have demonstrated through fruit composition analysis that biosynthesis of monoterpene glycoside flavour compounds in the grape berry is independent of other parts of the vine. Muscat Gordo ungr¡ltcd (control)

2 53

2 11

/ 12 II

10 1 4

Sullane grafted on Musc¡t Gordo rtc aaa

rst

Sulaana ungr¡flcd (control) ¡¡c

lsl

¡l

fig.4.7 Typical GC-MS chromatograms of the aglycons from Sultana control berries and from fruit of Muscat Gordo control and Muscat Gordo grafted onto Sultana. The y-axis scale is 'relative ion current' (RIC). Sullenr ungreftcd (control) 54

16

I

Muscat Gordo greftcd on Sult¡n¡ tlc lsf

2

7

16 4 6

10

Muscat Gordo ungrafted (control)

I

7 2 1¡l I 12 't 1 10

Fig. 4.8 Typical GC-MS chromatograms of the aglycons from Muscat gordo control berries and from fruit of Sultana control and Sultana grafted onto Muscat Gordo. The y- æris scale is 'relative ion current' (RIC). 55

Chapter 5

Occurrence of anthocyanin pigments in berries of the white cultivar Muscat Gordo Blanco

Contents Page

5.1 Introduction ...... 56 5.2 Materials and methods...... 57 5.2.1 Plant material 57 57 5.2.2 ExÛ¿ction .... -. -.. - - - -. 5.2.3 HPLC analysis...... 57 5.2.4 GC-MS analysis.... 58 5.3 Results...... 59 5.4 Discussion 64 5.5 Conclusion 66 56

5.1 Introduction

The colour of berries is an obvious distinguishing factor between grape cultivars and clones and, in coloured types, is mainly caused by differences in type and amount of anthocyanins in berry skin. rWulf and Nagel (197S) separated and quantified about twenty anthocyanin compounds by high-performance liquid chromatography (HPLC) analysis. This technique is still the standard and has been used by many to gain information about the nature of the colour differences between cultivars (Baker and Timberlake, 1985,'Wenzel et al., 1987, Mattivi et al., 1989, Cravero et al., 1994) and berry ripening stage (Somers, 1976, Crippen and Morrison, 1986, Ferna'ndez-Lo'pez et al., 1992). Recently, Watanabe et al. (1991) used it to classiff grape cultivars into five types based on their 3- glucosides: (a) mostly cyanidin (b) peonidin plus cyanidin (c) delphinidin plus cyanidin (d) four together - cyanidin, peonidin, delphinidin and petunidin and (e) mainly malvidin. Types (a) and (b) have skin colours with high hue angle (rose) and type (e) have low hue angle (blue-black). The muscat group of grape cultivars is known for its large array of berry colours. Collectively, muscats are characterised by similar leaf morphology and a distinctive and widely-appreciated berry flavour and aroma (Galet, 1958, 1962), attributable largely to their content of a number of oxygenated monoterpene compounds (Ribereau-Gayon et al., 1975). The most widely planted member of the group, cv. Muscat Gordo Blanco, has been the subject of much physiological and biochemical research; however this research has not included a studv of anthocyanins within its berry skin-not surprising considering that it is a white cultivar. During the studies on the biosynthesis of flavour compounds in berries of Muscat Gordo (Chapter 4) it was frrst noticed that very ripe Muscat Gordo berries on own- rooted vines or on bunches which were grafted onto Shiraz vine shoots developed a pink or rose coloration. The latter benres rvere fed by leaves from Shiraz grapevine which has coloured berries. The question arose as to whether this colouration of normally white berries was due to the effèct of leaves and whether the precursors of anthocyanins were translocated from leaves to the berries. To gain an understanding of this phenomenon, the components of skin colour, the levels of polyphenolic compounds and juice aroma were studied. Muscat Gordo berries at different stages of ripening and development were used and the colour pigment of the 57 berries that accumulates occasionally in over-ripe berries was extracted, analysed and identified.

5.2 Materials and methods

5.2.1 Plant materials

Muscat Gordo grape bunches were picked from two locations: the Waite experimental vineyard, which were own-rooted; and Penfold's vineyard at Waikerie, Riverland S.4., which were grafted on to cv. Mataro (syn. Mouvedre), and also from bunches grafted onto Shiraz vines and grown in the glasshouse (method in chapter 4.2). Two types of bunches were taken (l) those with berries coloured pink or rose (Fig. 5.1) and (if) those without any pigmented berries. From about ten bunches of each type, a pool of 221 berries were picked at late ripening (E-L stage 39, Coombe, 1995) and mixed.

5.2.2 Extraction

All benies were individually weighed, berry length and two diameters measured to enable skin surface area to be calculated, and the total soluble solids (TSS) of the juice was determined as "Brix after peeling the skin. The skin of each berry was placed in a test tube containing 50% ethanol, acidified to pHcl with HCl, the tubes mixed every hour for 4 hours and then left to stand overnight. After centrifuging at 4000 rpm (1500 g) for 10 minutes, the clear supernatants were used for spectrophotometric measurements (using a Varian/DMS 200 UV Visible spectrophotometer) at 520 and 280 nm. and anthocyanin and total polyphenol content of the skin calculated per berry, and per gram fresh weight of skin. Shiraz berries grafted onto Muscat Gordo and non-grafted Shiraz berries were extracted for anthocyanin pigments and spectrophotometric measurements were preformed in the same way.

5.2.3 HPLC analysis

HPLC analysis was done in triplicate on the extracts of each type of samples after being concentrated b_'" evaporating the ethanol under nitrogen flow at 35'C. HPLC analysis for berries from own-rooted vines, from vines grafted on to cv. Mataro (syn. Mouvedre), and also from bunches grafted onto Shiraz vines (glasshouse experiment) were done separately. 58

The separation of anthocyanins was made on a C18 column (LC.I. ODS 2; 5 pm, 250 mm x 4.6 mm) with precolumn (Cl8 cartridge 10 mm). Elution was with two solvents, solvent gradients were varying proportions of (A): l00Yo methanol, and (B) : O-6 Vo aqueous perchloric acid, pH 1.3 to 1.4. The flow rate was 1 .5 mU min., and run time; 35 minutes. The following elution protocol was used: A;20Yo,0-10 minutes, A; 20o/oûo 35o/o,10-15miuutes, A;35%-70yo,15-30minutesandA;70%-I00yo,30-35minutes. The column was then washed, first for 40-50 min with A-100% and then for 50-60 min with A-20% thus restoring it to its original condition before each run. The eluant was monitored at 520 nm using a diode array detector. The peaks of the chromatograms were identified by comparing retention times with those from known components in Shiraz grape juice, Shiraz wine and authentic cyanidin- 3-glucoside. Quantification was based on peak areas of unknowns against that of known amounts of cyanidin-3 -glucoside.

5.2.4 GC-MS analysis

Monoterpene compounds were determined by GC-MS analysis of 20-25 g of berries according to the method described in chapter four (4.2.4).

Fig. 5.1 Muscat Gordo berries showing pink colour 59

5.3 Results oBrix, Un-pigmented berries of Muscat Gordo, at 18 showed no detectable level of anthocyanins in berry skin, but the sample of pigmented berries (which had a high Brix level of 28.5") had a significant though low level of total anthocyanins, together with a high content of total phenolics (Table 5.1). The compounds contributing to the colour were predominated by cyanidin-3-glucoside, plus minor amounts of peonidin- and delphinidin-3-glucosides. The ch¡omatograms shown in Fig. 5.2 present the separation profile of anthocyanins by the method described above with the Muscat Gordo berry skin extracts. The major peak was identified by comparison with cyanidin-3-glucoside as a standard. The remaining peaks corresponding to other anthocyanins, contributed less than l0% of the total pigments in the berries (Table 5.1).

Table 5.1 Total polyphenolics and anthocyanins in the skin of Muscat Gordo Blanco in two groups: white (non-pigmented) and pigmented (showing a light rose or pink colour)' The fruits were from vines grown in the field.

White Pismented

Juice oBrix 18.0 28.5 Concentration mgikg skin fr. wt. l¡t Total phenolics (GAE*) 930 r570 ** Total anthocyanins n.d. 7.7

Anthocyanin %n of total

Delphinidin-3 - glucoside n.d. 2.4 Cyanidin-3-glucoside n.d. 94.0

Petunidin-3 - gl ucoside n.d. n.d. Peonidin-3-glucoside n.d. 3.6 Malvidin-3-glucoside n.d. n.d. p-Coumarate esters n.d. n.d. Acetate esters n.d. n.d. Total 100 n.d.: not detected

+GAE: gallic acid equivalents

** Values are means of three: standard elrors were +12-l4Yo 60

oBrix The absorbance, fresh weight and of an unselected population of 221 individual berries harvested at random from ten bunches from both sites a¡e shown in Fig' 5.3. The absorbances at 520 nm of skin extracts show that pigment only occurred in berries with oBrix oBrix above 2a Gig.5.3 a); those with values of 24 or less gave low (< 0.18 units) absorbance readings which were shown by spectrophotometer scanning not to be due to specific 520 nm absorbance. Skin extract absorbance at 520 nm is compared with berry weight in Fig. 5.3 b. It is clear that pigmentation occurred in a wide array of berry sizes, oBrix' but especially in smaller berries. Fig. 5.3c shows that small berries have higher in oBrix fact, berries of < 2.3 g all had juice with > 25. No study was made of the relationship of pigmentation to seed development. Regarding the absorbance values at 520 nm, the highest total anthocyanin content of an individual Muscat Gordo ber{ recorded in this experiment was 33.09 t 1.06 Pg per berry. By comparison, the anthocyanin content of Shiraz berries grafted onto Muscat Gordo shoots and non-grafted : Shiraz berries were I .33 mglg berry ("Brix : 23) and 1 .26 mglg berry' ("Brix 20) respectively. Examination of berry surfaces under the mic¡oscope showed that pigmented skin had a mix of coloured and uncoloured cells (data not shown) and suggested that greater pigmentation was due more to a higher proportion of coloured cells than to a deeper colour per cell. Values for the concentration of monoterpenes in homogenates of whole berries or in the berry juice are shown in Table 5.2. The increasing values of TSS inthe three samples used for whole berry extraction showed a cor¡elated increase in all 16 monoterpenes, in many cases showing a ten-fold range. However their relative proportions remained generally similar. 61

o.o? c

0 t0 o.0a b

o.ol

0 ra 0-06 0.02 a

0. 12 0.01 0. oa

0.10 I 0.00

0.02 0.o3 È 1 20. oo ¡0.00

0.06 0.00 o-oó

I o. 02 '- 00 3,.00 !o.00 : i

0.00

10.00 20.00 10.00 Xlôuced

Fig. 5.2 HPLC chromatograms of anthocyanin pigments profile of the extracts for Muscat Gordo berries skin (a) from own-rooted vines, (b), from bunches grafted onto Shiraz vines. chromatogram (c) shows authentic cyanidin-3-glucoside. Chromatogram (b) showed a faster elution profile due to different instrument conditions. 2.0 a 62 cE 1.5 o ôl tf¡

G o 1.0 .J tr ' . "'a' (ú ì.. ¡¡ a at ' o . . ..t...tr at, ¡¡ 0.5 rl.d Ë1.Â:t^ -'¡l a -]l¡ " l.ì;. ..rrl" I^, 1¡ -.' .'j ld". 0.0 t0 15 20 25 30 35 Ëflx

2.0 b

E c 1 .5 o ôt ¡at o o a) .0 tr '.{... 6 lt t.. o ¡ l¡at 0.5 ?t' ¡ .a ,...Âfj.É i d. ï:.¡ t f :rit î,. 0.0 012 345 678 Berry Weight g

40 c

30 x dl fD 25

.t ¡t¡¡ ¡ 20

't5

10 012 345 678 Berry Welght 9

Fig. 5.3 Relationship for 221 Muscat Gordo berries of: (a) total soluble solids ofjuice ('Brix) and absorbance of skin exûacts at 520 nm, (b) berry weight and absorbance of skin extracts at 520 nm and (c) berry weight and oBrix. 63

Table 5.2 Concentrations of monoterpene compounds in whole non-pigmented (White) and rose- or pink-coloured (Colour) Muscat Gordo berries and white juice' Compounds are numbered as in Figure 4.3 of chapter four'

Whole berries Juice White Colour Colour White "Brix 17.8 26.2 28.0 22.0 nmoVet nmoVmLl No Compounds

I Furan linalool oxide, 2 0.18 0.91 1.8 0.04 2 Linalool 1.53 2.70 7.4 1.81 3 Hotrienol 0.16 0.68 0.8 4.7r

4 Pyran linalool oxide isomer 0.58 2.t3 6.6 2.r6 5 a-Terpineol 0.03 0.05 0.4 0.13

6 Pyran linalool oxide isomer 0.03 0.11 0.5 0.1 I 7 Nerol 0.41 0.76 3.4 0.64

8 Geraniol 1.8 r 4.37 24.8 t.74

9 2,6-Dimethylocta-3,7-diene-2,6-diol 4.50 1 5.15 34.5 5.3 3 l0 3,7-Dimethylocta-I.6-diene-3,5-diol 0.10 0.67 0.9 0.22

1 I 2,6-Dimethylocta-1,7-diene-3,6-diol 0.24 0.68 4.9 1.66 12 Z-2,6-Dimethylocta-2.7-diene- 1.6-diol 0.44 t.44 6.7 1.68

I 3 E-2,6-Dimethylocta-2.7-diene- 1 .6-diol 0.58 0.94 3.6 0.36

l4 3,7-Dimethylocta-2-ene-1.7-diol 1.0 0.59 8.8 0.r6

15 p-Menth- 1 -ene-6,8-diol 0.01 0.02 0.3 nd 16 p-Menth-1-ene- 7.8-diol 0.22 0.21 5.3 0.31

Total tt .52 31.43 I 10.7 21.06 + t nmoVg of berry homogenate +nmol/ ml ofjuice 64

5.4 Discussion

The first observations of Muscat Gordo pigmentation were made during the experiments described in Chapter four in which bunches of this cultivar were grafted onto Shiraz shoots to study biosynthesis of flavour compounds. Samples from that experiment showed that the anthocyanin profile for Muscat Gordo berries grafted onto Shiraz shoot was the same as nongrafted berries (Fig. 5.2). Also Shiraz berries developed full berry colour (1.33 gm/ g berrl') when grafted on to Muscat Gordo shoots compared with the non-grafted berries (1.26 mgl g berry). The slightly higher value for non-grafted berries was probably due to smaller berry in size with higher "Brix. Similarly, there were no differences between the pigment profile of berries from own-rooted Muscat Gordo and berries from vines grown on the roots of the black cultivar Mataro (chromatogram is not shown). All of these results support the view that berries are independent of leaves and rootstock for synthesis of anthocyanins as well as other secondary metabolites, as was suggested earlier in Chapter 4. The results show that light pink or rose pigmentation is able to develop in the normally white (un-pigmented) berries of cv Muscat Gordo. The amounts are small - in one test we found a concentration of total anthocyanins of 8 mg per kg skin fresh weight which is lower than the smallest concentration of the range (varying from 28 to 685 mg/kg) found by Cravero et al. (1994)in skins of 22 coloured muscat cultivars. Distinctive features of berry coloration in Gordo are its late development---coloured berries had juice of 24 "Brix or higher and the hot conditions under which it developed. This contrasts with some malvidin-3-glucoside-accumulating, black cultivars such as Cabemet Sauvignon and Shiraz in which colouring develops at about l0 oBrix (e.g. Roggero et al., 1986) and is more intense if the weather is cool (Kliewer, 1970c). It is possible that the Muscat Gordo pigmentation was augmented by the lateness of its development in autumn when nights were cooler. Cyanidin-3-glucoside was the predominant anthocyanin (Fig. 5.2), making up 94 per cent of the total (Table 5.1). According to the classification made by Cravero et al. (1994), this result puts it in the same range as th¡ee other muscat cultivars - Perle de Csaba rouge, Siegerrebe and Marangos. These authors also showed that four other cultivars, which had about 70 per cent of the total as cyanidin-3-glucoside, belonged to the general group of rose-coloured, small-seeded muscats; these are coloured variants of 'Muscat a petit grains' (syn., Muscat de Frontignan, Muscat Canelli, or, in Australia, Frontignac). Of the 22 coloured Muscats examined by Cravero et al. (1994), four had a predominance of peonidin-3-glucoside while, in half of them, malvidin-3-glucoside predominated; many of these two groups of cultivars have blue or black berries. They noted that those cultivars with a high level of cyanidin-3-glucoside in their berry skin also 65 have a high level of monoterpenes in their juice and are commensurately aromatic (in the style of the small-seeded muscat group). Values of monoterpenes in juice of Muscat Gordo (column 4 of Table 5.2), while difficult to compare with the data of Cravero et al. (1994)(which lack units), nevertheless show similarities with, for instance, cv. Muscat de Frontignan rouge in the predominance of diol No. 9 (: their DDI) plus high levels of linalool (Muscat Gordo also had high levels of terpene No.3, hotrienol). The higher concentration of total monoterpenes in whole berries at high "Brix (Table 5.2) is reminiscent of the same phenomenon shown by total phenolics and anthocyanins in skin (Table 5.1 and Fig.5.3 a). Although different biosynthetic pathways would be involved, Cravero et al. (1994) suggested that the relationship between juice monoterpenes and skin anthocyanins warranted further investigation. Any connection to seed development appears unlikely in view of the difference in seed size of Muscat Gordo and the small-seeded muscats, but this aspect also needs exploration. The compound cyanidin-3-glucoside is regarded by Roggero et al. (1986) as a primitive pigment in the biosynthetic pathway of anthocyanins (along with pelargonodin-3- glucoside, which does not occur in Vitis vinifera). This less stable and primitive grape pigment culminates in the stable pigments peonodin-3-glucoside through hydroxylating and methylating enzymes. Only peonidin- and delphinidin-3-glucosides were found in Muscat Gordo and these at only a few per cent. Thus the pathway in Muscat Gordo, under the specific conditions of very high levels of TSS and warm weather during berry development, enabled biosynthesis of cyanidin (or its glucoside) from its flavanone, but it seems that the further development and transformation fails to function properly and the enzymes for further metabolism (flavonoid 3'-5'-hydroxylase and methyl transferase) were relatively inactive. It is interesting that Watanabe et al. (1992) found, from cosses utilising one parent with white fruit colour and one with coloured fruit, that Muscat Gordo has a latent but dominant gene for methylation of the B ring of anthocyanins. The unusual and specific conditions under which anthocyanin synthesis has been found to be triggered in Muscat Gordo berries might be useful for future studies of anthocyanin biosynthesis in grape berries. The results presented in this chapter support the idea that the pigmentation which occurs in Muscat Gordo berries is genetically controlled by the berry itself. This phenomenon occurs onl1, in the over-ripe berries and there is no relationship between grafting practice and this type of berry pigmentation. Feeding from leaves of vines with coloured fruit was not the cause of the pigmentation, and the colour of Shiraz berries was not significantly decreased when grown on shoots of a white-fruited cultivar. As was concluded from the experiments of flavour compounds in chapter four, the benies appear to be independent of leaves in the synthesis of anthocyanin pigments. bÞ

5.5 Conclusion

Pink pigmentation can develop in the skin of the normally white berries of cv Muscat Gordo. The principal causal anthocyanin is cyanidin-3-glucoside. The colour develops only in berries with juice "Brix > 24 most of such berries being small. Their composition of anthcyanin and monoterpene resembles that of berries of 'Muscat a petits grains' cultivars. Bunch grafting showed that leaves of black berried cultivar (Shiraz) did not alter the phenomenon of pigmentation in Muscat Gordo berries. Similarly, the anthocyanin pigments of Shiraz berries were not altered by being grafted onto shoots of Muscat Gordo. As was concluded from the lack of differences in monoterpene composition in Chapter four, berries appeff to be independent of leaves in the synthesis of anthocyanin pigments as well. 67

Chapter 6

Assimilate transport in the grapevine ; carbohydrates, amino acids and potassium

Contents Page

6.1 Introduction ...... -.... 6.2 Materials and methods..... -......

6.2.1 Plant materials. . . 6.2.2 Other materials...

6.2.3 Exudate collection techniques.. -. -... -......

Exudate collectionfrom the peduncle .. Exudate colle ction from the leaves.. -...... Exudate collection fr om excis ed peduncle.. Girdling.... Controls of bufer solutionwithout EDTA.

S ampling t o e s tabli sh diurnal patterns...... 6.2.4 Sugar analysis..... 6.2.5 Amino acid anaIysis...... 6.2.6 Mineral elements measurement ..-...... 6.3 Results 6.3.1 Grapevine phloem sap composition...... 6.3.2 Seasonal changes in the phloem sap composition 6.3.3 Validation of the sap collecting technique,...... 6.3.4 Phloem sap diurnal patterns..... Pattern of sucrose in exudøles during a day (before veraison)...... Dalt snd night diurnal patlerns (after veraison)... Experimentl .....

Experiment II ...... 6.3.5 Grapevine leaf exudates.... 6.4 Discussion...... 6.4.1 Validation of the sap collecting method...... 6.4.2 Grapevine exudate composition; comparison with tissue composition

;.;.; ;;- ,"0 n". ,"t"," u.."*"i",t";t;;;";; ;;;t. "^o 6.4.4 Seasonal changes in exudate flux.-......

6.4.5 Diurnal chances------o- in exudation 6.4.6 Grapevine leaf exudate...... 6.5 Conclusion...... 68

6.1 Introduction

Photosynthesis in the mature leaves of higher plants provides the metabolites for the sink organs and maintains the assimilates required for plant growth and development. About 80% of photosynthetically fixed carbon is used for carbohydrate synthesis (cellulose, starch, sucrose in higher plants, Gordonet al., 1980) and the remainder is used in synthesis of amino acids, organic acids, and other primary and secondary metabolites (Huber et al. 1992). The mature leaves produce assimilates such as sucrose and certain amino acids from inorganic carbon and nitrogen, which then can be transported to and utilised in other parts of the plant. Phloem tissue conducts the solute transport from source tissues to sinks. Many inorganic ions and organic compounds such as sugars, amino acids, organic acids, and plant hormones as well as macromolecules such as nucleic acids and proteins have been detected in phloem exudates collected from various plants (Moss and Rasmussen, 1969; Hall and Baker,1972;Ziegler, 1975; Fondy and Geiger, 1977; Giaquinta and Geiger, 1977, Barlow and Randolph, 1978; Kawabe et al.' 1980; Delrot, l98l;Hayashi and Chino, 1986; Chino etal., 1991; Girousse, l99l: Jeschke and Pate. 1995;Nakamura et al. 1995; Wijayanti et al., 1995), and from Vitis vinifera by Glad et al. (1992a). The concentration of dry matter in phloem sap varies but values of 25o/" w/v have been measured (Pate, 1975). The concentration of total solutes inphloem sap of trees is usually 0.4 to 0.6 Molar (Zimmermann, 1960; Nobel, 1991). The types and concentrations of solutes exhibit daily and seasonal variations, and also depend on the tissues that the phloem solution is flowing toward or away from (Nobel, l99l). The amounts of phloem sap assimilates allocated to fruits over the whole period of their development are important factors affecting the quality and quantity of the fruit. A large amount of glucose and fructose. and somewhat lesser amounts of tartrate, malate, potassium and a wide variety' of other compounds accumulate in grape berries during berry development and ripening (Kiiewer. 1967a; Harris et al., 1968; Peynaud and Ribéreau- Gayon, 1971; Coombe and Matile, 1980; Coombe, 1975,1989, 1992). The mature and well developed leaves on the shoot are the main sources of sugar supplied to the berries and to storage tissues (Candolfi-Vasconcelos and Koblet, 1990). The surplus energy from photosynthesis in source leaves can be stored temporarily in the chloroplast either as carbohydrate or as lipid depending on temperature (Buttrose and Hale, l97l). Starch is an important end product of photosynthesis that serves as a temporary reserve form of 69 reduced ca¡bon stored in leaves (Huber et al., 1992), and also in roots (Hunter et al., 1995). The grape berries have a directional influence on translocation in grapevines and many studies showed a predominant movement of photosynthates towards the grape berries once a certain fruit size is attained (Hale and Weaver,1962; Quinlan and Weaver,1970; Kriedemann , 1977; Hunter and Visser, 1988 Coombe, 1989; Motomora, 1993). Removal of grapes from the vine, and also girdling of the phloem, results in rapid decrease in the photosynthesis rate and an increase in the stomatal diffusive resistance, but topping does not, thus demonstrating the role of berries as a strong sink in activating transport from leaves and suggesting that the growing point of the shoot at this stage of its development is not as strong a competitor for photosynthate metabolites (HofÌicker, 1978). Sucrose is the major carbohydrate produced in the leaves and transported into grape berries (Swanson and El-Shishini, 1958; Koblet, 1977) and accounts for 70%o of carbohydrates of grapevine phloem sap (Glad et al., 1992a). Sucrose translocating in the grapevine could originate from any current photosynthesis or any reserve tissues in the leaf or from stem, wood, roots or other perennial parts of the vine (Matsui et al., 1979; Coombe, 1989; Williams and Biscay, 1991; Candolfi-Vasconcelos et a1.,1994). The levels for hexoses were reported to be low in phloem sap exudates of grapevine analysed by Glad et al., (1992a). Besides sucrose, amino acids and amides are major constituents of the grapevine sieve tube sap (Pate, 1980; Glad et a1.,7992a; Rubelakis-Angelakis and Kliewer, 1992). The transport of nitrogen in the phloem seems to follow the bulk flow theory for carbohydrate movement (Bidwell 1979). Glad et al. (1992a} reported the nitrogen solute composition of grapevine (Vitis vinifera L.), cv. Pinot Noir phloem saP, glutamine being the major (60%) transported form of nitrogen and proline the second (10%) at flowering. Potassium is the most prevalent cation in ripe grape berries and accumulates in considerable amounts after veraison and during ripening of the berry (Hale, 1977; Coombe, 1987a Iland and Coombe. 1988; Monison and lodi, 1990; Williams and Biscay, 1991; Guatiérez-Guanda and Morrison. 1992: Creasy et al., 1993; Boselli et al., 1995). The phloem sap is the source of metabolites accumulated into the grape berries during berry growth and development. most extensively after the inception of ripening. Yet little is known about the composition of grapevine phloem sap. In this study, phloem sap was collected from the peduncle of different grapevine cultivars and analysed for sugars, amino acids and potassium in order to reveal the grapevine phloem sap composition and the seasonal and diurnal changes in the levels of metabolites present in the phloem sap and also to check the validity of the technique for grapevine phloem sap collection. 70

6.2 Materials and methods

6.2.1 Plant materials

Mature grapevines of Vitis vinifera L. cultivars; Muscat Gordo Blanco, Sultana, and Shiraz were used from the Waite Institute experimental vineyards at Glen Osmond, South Australia. The samples were collected at three times during berry growth and development being before veraison (three to four weeks after fruit set), at veraison, and at ripening time. Samples also were collected from Muscat Gordo from early veraison up to ripening time.

6.2.2 Other materials

HEPES {N-[2-hydroxyethyl] piperazine-N'[2-ethansulfonic acid]] buffer solution was made of 10 mM HEPES dissolved in distilled water and EDTA (ethylenediaminetetraacetic acid) was added to the solution at the concentration of 10 mM (final solution 10 mM of each compound) and then the pH was adjusted to 7.5 using sodium hydroxide (1.0 N). Eppendorf tubes filled with 1.5 ml of this solution were used for phloem sap collection.

6.2.3 Exudate collection techniques

The facilitated exudation technique (King andZeevaart,1974; Hanson and Cohen 1985; Glad et al.,1992a) was used to collect phloem sap from severed peduncle laterals and from the cut end of leaf petioles.

Exudate collection from the peduncle The phloem exudate from peduncle was collected in situ. The first lateral on the peduncle was excised under a drop of buffer solution and immediately the cut end was dipped into an Eppendorf tube filled with buffer solution (Fig. 6.1). The tube was fixed to stay on the bunch using blue-tack pliable adhesive [Bostik, (Australia) Pty. Ltd.]. This material was found to have no harmful effect on the plant tissue. The cut end of the peduncle laterals were rinsed with the buffer solution for about l0 minutes and then the tubes were replaced with tubes containing fresh buffer solution which were left to stay for all of the sampling period. The sampling time course varied between experiments. Sampling on each day started from 9 am until 5 pm particularly when a large number of tubes were collected. The tubes were transfered into ice after being removed from the vines and then transfered to -20 "C until chemical analysis. 71

Fruits from cut laterals were collected and weighed, and mean berry weight was determined. "Brix ofjuice from a pool of fruits was measured with a hand refractometer.

Exudate collectionfrom the leaves Mature, healthy leaves from nodes 4 to 6 on the shoot were used for phloem exudate collection. The leaves were severed from the vines at the basal end of the petiole close to the cane under a drop of buffer solution. They were immediately transferred into buffer solution before taking them to a controlled environment for exudate collection. Petioles were placed in the tubes containing buffer solution (10 mM, HEPES buffer containing EDTA as used for sap collection from bunch laterals), and phloem exudate of the leaves was collected (King andZeevaart,lgT4). The leaves were fixed to stand in a box with a top lid, made of clear perspex (Fig. 6.2). The leaves were sprayed with distilled water to maintain high humidity and so minimise transpiration. The exudate samples collected were kept at -20'C until analysis. The first year sampling (1994) was done under lights, with a light intensity of 400 pE m-2s-' and the temperature near the leaves was between 24-28 "C. The second year sampling (1995) was done in the dark with ambient temperatures of 22-25 "C.

Exudate collection from excised peduncle Exudate was collected from the grape peduncle after being cut off the vine to check the contribution made by this tissue. The first bunch lateral on a grape bunch was cut under a drop of buffer solution (as described above), then a sap-collecting tube was fixed on the bunch. The whole bunch was severed from the vine and all berries, laterals and more than half of the peduncle were removed. The remaining bunch stem with the tube attached to it was fixed on a holder (Fig. 6.3) and left on the bench for 6 hours and the exudate collected and analysed for metabolites. Exudate was collected from three replicate bunchstems with a mean fresh weight of 15.2 + I .05 gr.

Girdling To provide further assessment of phloem sap flux, girdling was done on the shoot above and below the bunch (Fig. 6.a) thus excluding the fruit bunch from phloem flux and stopping phloem sap transport into the berries. The leaf opposite the bunch was also removed and collections were made from the girdled bunches as described above and analysed for metabolites.

Controls of buffer solution witltout EDTA A series of control samples was collected for exudation by excluding EDTA from the buffer solution. The method for collecting sap samples was the same as described above. 72

Fig. 6.1 Phloem sap collection from a peduncle lateral

Fig. 6.2 Phloer-n sap collection from cletached leaves. 73

BL

ET

IIB

Fig.6.3 Exudate collection frompeduncletissue off the vine. Ped: peduncle; BL: bunch lateral; ET: Eppendorf tube; HB: holding block.

G

Ped. L

Fig. 6.4 Girdling the cane above and below the fruit bunch in order to stop phloem sap flux to the bunch, G: girdled area; Ped. L: peduncle lateral from which sap was collected. 74

Sømpling to estøblish díurnal patterns Samples were collected shortly before veraison, after veraison, and at early ripening time. The samples from each of ten separate bunches were collected successively during the day and night at different time course intervals and those from each sampling time were pooled to give an adequate sample size for analyses.

6.2.4 Sugar analysis

Soluble sugars were assayed enzymatically using a Boehringer Mannheim enzyme kit and quantified by spectrophotometric detection. The solution containing phloem exudate was used directly without concentrating but the samples with high sucrose levels were diluted to improve accuracy. The levels of sucrose, glucose and fructose were determined for each sample.

6.2.5 Amino acid analysis

The phloem exudates in 10 mM HEPES buffer solution were freeze dried. Lyophilised samples were redissolved in 100p1 of HCI (0.1 N) containing 0.5 mM norvaline and sarcosine as internal standards. Redissolved samples were centrifuged for 5 minutes at 14000 rpm in an Eppendorf 5415c centrifuge. Analysis was performed with one tll of the supernatant. Amino acid analysis was carried out on a Hewlett-Packard Amino Quant amino acid analyser, consisting of an HP 1090 series II liquid chromatograph controlled by HP Chem Station software. An autosampler was used to derivatise the amino acids which were then separated by reversed-phase HPLC on an Amino quant Cl8 column. Derivatisation was with OPA (ortho-phthalaldehyde) (l'amino acids) and FMOC (9- fluorenylmethyl chloroformate) (2' amino acids) and quantified by U.V. absorbance. Amino acid calibration standards. derivatising regents, and the amino acid analysis column were obtained from Hewlett-Packard.

6.2.6 Mineral elements measurement

Inductively coupled plasma spectrometery (ICPS) enables the determination of total P, K, S, Ca, Mg, Na, Al, Zn, Mn, Fe, Co, Mo, and B. ICPS has the potential to simultaneously determine all the nutritional elements (except nitrogen) with a polychromator in a single digestion. A known volume of exudate in buffer solution was lyophilised and dissolved in approximately 1 ml of nitric acid (Univar A.R.71o/o w/w). The mixture was heated in a water bath for 20 minutes, until brown fumes were no longer 75 given off. Digests were removed from the water bath and allowed to cool. After cooling, the digest was diluted to the initial volume with l% v/v nitric acid and decanted into a polystyrene tube. The digest stood overnight before ICP analysis to allow precipitation of any suspended solids. The ICP instrument used was a 35808 ICP Spectrometer Sim/Seq. with 21 simultaneous channels manufactured by Apptied Research Laboratories (Switzerland). As sodium hydroxide (NaOH ) was used for pH adjustment in the buffer solution and sulphur was one of constituents of HEPES buffer, and also because of the presence of EDTA in the solution, the values for Na, S, Ca and Mg are not presented.

6.3 Results

The grapevine phloem sap was collected from bunch lateral branches in situ for three cultivars at different physiological stages during the growing season and analysed for sugars (sucrose, glucose and fructose), amino acids and minerals.

6.3.1 Grapevine phloem sap composition

Tables 6.1 and 6.2 show the levels of sugars, total amino acids and K in the phloem sap collected from cultivars Muscat Gordo, Sultana and Shiraz before veraison, at veraison and at berry ripeness. The results presented are the concentrations of each compound in the buffer solution used to collect the exudates. This was 1.5 mL per bunch lateral and the values correspond to the concentration of the metabolite in solution taken from the pool of samples collected from 120 bunches (1.5 ml/ bunch) during 4 hours between 9.00 am to 5.00 pm. For all three cultivars. sugars and K were the predominant solutes in the exudates collected during the growing season. Sucrose levels in the sap collecting buffer solution were highest before veraison and the lowest at veraison. The major sugar detected in all samples was sucrose. but glucose and fructose were also present although constantly in low amounts and showing no correlation with the corresponding sucrose levels (Table 6.1). 76

Table 6.1 Sugars, amino acids and potassium measured in phloem sap exudate of three grapevine cultiva¡s during the growing season (p mol I 1.5 rnL of buffer solution)'

Cultiva¡ Sampling time Sucrose Glucose Fructose Potassium Total amino "Brix acids

bef-ore veraison 4.5 3.24 0.1 63 0.165 2.54 0.453 Muscat Gordo veraison 7.4 l.0l 0.325 0.331 4.08 0.016 npe 23.5 2.68 0.1 60 0.242 3.92 0.1 57

before veraison 4.7 9.t9 0.1 63 0.1 65 2.35 0.554 Sultana veraison 8.2 2.17 0.1 60 0.081 t.62 0.326 ripe 24.2 6.48 0.243 0.248 1.65 0.199

before veraison 4.2 2.39 0.081 0.075 2.04 0.054 Shiraz veraison 7.9 0.25 0.080 0.080 l'58 0.102

n D e 2t.6 1.36 0 ,l6l 0.161 2.27 0.121

Exuded for 4 hours. (120 bunches; 1.5 ml / bunch/ 4h.). Data from 1994 samples

Sucrose levels in the exudates from Sultana were higher than for the other two varieties. The phloem sap exudates were analysed for I I minerals but only results for potassium are presented here. The K levels in the exudates were similar in all th¡ee cultivars and did not change greatly during the growing season (Table 6.1). Total amino nitrogen was determined and as shown in Table 6.1. concentration of amino acids was lower than for sucrose and K on a molar basis. There were also seasonal changes and varietal differences in total amino acids but they did not parallel those for sucrose (Table 6.1). For example, while there are parallel changes in the concentrations of sucrose and amino acids in Muscat Gordo phloem exudates. Sultana and Shiraz phloem exudates did not show similar fluctuations in the concentrations of sucrose and amino acids.

6.3.2 Seasonal changes in the phloem sap composition

The predomihant amino acid in phloem sap was glutamine, with smaller amounts of glutamate, aspartate, serine and proline. Glutamine and glutamic acid together represented about 50% - 80% of the amino acid profìle all through the growing season and amino acids levels showed different seasonal patterns in various cultivars (Table 6.2). In Muscat Gordo amino acid levels showed a seasonal pattern similar to that of sucrose whereas Sultana and Shiraz each had different seasonal pattems of amino acids levels in the buffer solution (Table 6.2 and Appendix Tables 1,3,.4, and 5). 77

Table 6.2The seasonal changes in phloem exudate amino acids of Muscat Gordo, Sultana and Shiraz. Values given areo/o of total amino acids before veraison (B), at veraison (V), and at ripening time (R).

Muscat Gordo Sultana Shiraz Amino acids (B) CÐ ß) (B) ry) ß) (B) (v) (R) ASP 3.7 9.9 9.3 1.9 3.7 4.5 7.3 8.5 28.7 GLU 6.4 13 tt.7 3.3 5.8 8.1 10.8 13.5 24.4 ASN 0.7 0 I 0.9 0.9 I 0 0.7 0 SER 1.8 3.8 7) 1.5 J.J 2.9 6.2 3.9 4.4 GLN 66.2 54.6 42.2 76 69.3 52.6 48.4 44.6 14.2 HIS 2.1 0 r.2 I r.2 2.r t.4 t.4 2.5 GLY I 0 2.7 0.9 1.5 1.3 2.4 1.9 4.9

THR r.9 04.8 2 1 1.5 1.3 2.1 2.t 0 ALA 1.6 t4 J.J t.7 t.9 3.8 2.5 2.8 5.3 ARG 6 0 9.4 7 4.4 13 0.9 3.5 7.5 TYR 0.3 0 0 02 0.4 0.4 0.2 0.3 0 CYS-CYS 0.2 0 0.6 0.1 0.3 0.2 0.1 0 0 VAL 0.5 0 0.9 0.4 0.9 0.8 0.2 4.8 0 MET 0 0 0 0.3 0 0.1 0.6 0.5 2.t TRP 0.6 0 0.5 0.2 0.3 0.8 0.2 0 0.2 PHE 0.1 0 0.5 0.1 0.8 0.9 0 2.3 0 ILE 0.3 0 0.4 0.2 0.5 0.4 0 2.9 0 LEU 0.3 0 0 0.3 0.6 0.7 0 2.2 0 LYS 0.3 0 0 0.3 0 0.5 0 0 0 "4.7 PRO 5.3 0 r1.2 1.8 2.4 16.5 4.t 5.8

Total * 0.453 0.016 0.153 0.554 0.326 0.199 0.054 0.102 0.121 "Brix 4.5 7.4 23.5 4.7 8.2 24.2 4.2 7.9 21.6 *Totul values are in ¡rmol amino acids/ l.5mL. of buffer solution/ 4 h.

Abbreviations for amino acids presented ASP Aspartic acid, GLU Glutamic acid, ASN Asparagine, SER Serine, GLN Glutamine, HIS Histidine, GLY Glycine, THR Threonine, ALA Alanine, ARG Arginine, TYR Tyrosine, CYS-CYS Cysteine, VAL Valine, MET Methionine, TRP Tryptophan, PHE Phenylalanine,ILE Isoleucine, LEU Leucine, LYS Lysine, PRO Proline. 78

6.3.3 Validation of the sap collecting technique

To assure the validity of the technique used in these experiments, controls were set up by applying different treatments designed to alter phloem sap exudation. Table 6.3 shows the comparison between the sugar levels exuded from cut bunch stems on detached btmches (Fig. 6.3) and the exudate collected from bunches still connected to the vine. The results show more than 14 times more sucrose exuded from the bunches which were connected to the vine compared to the cut bunch stems. The levels of glucose and fructose were very low under both treatments.

Table 6.3 Comparison of sugars exuded from fruit bunches on the vine with excised bunchstems as in Fig. 6.3, (cv. Muscat Gordo at pre veraison, 'Brix:4.5).

Sucrose Glucose Fructose Treatments ttmol/ L5 ml of buffer solution*

Standard sampling 3.77 + 1.36 ** 0.131 + 0.2 0.131+ 0.04

Bunchstem only 0.26! 0.04 0.04 + 0 0.04 + 0 *The exudation period was 6 h. ++ Values are meant SD of three replicates of l0 bunches in each.

As EDTA was used as a chelating agent to keep sieve tubes open, one of the treatments to obtain control samples to test the continuation of phloem sap flux into the buffer solution was exclusion of the EDTA from the buffer. Table 6.4 presents the results of the samples collected b1, using HEPES buffer containing EDTA (buffer + EDTA) and the buffer without EDTA (buffer - EDTA) and also using water only. In the samples collected in the buffer solution without EDTA, and in water, no sucrose was detected and the levels of amino acids and potassium were also low in comparison with the samples collected in standard buffer solution. 79

Table 6.4 The effect of exclusion of EDTA from the buffer solution used for sap collection on the levels of different metabolites in collected sap samples (cv. Muscat Gordo at veraison, 'Brix :6.4).

Treatment Sucrose Total amino acids Potassium ¡unol/ 1.5 ml ofbuffer solution*

Buffer + EDTA 0.342 0.047 2.58 Buffer - EDTA <0.04 0.016 0.642 Water <0.04 0.009 0.307

*The exudation period was 4 hours. Subsampled from pool of samples collected from 30 bunches

Girdling above and below the fruit bunch on the cane (Fig. 6.4) was done to cut the phloem tissue and thus stop phloem sap transport to the bunch. The phloem sap collected from girdled bunches was analysed for sucrose, amino acids and potassium and the results are presented in Table 6.5.

Table 6.5 The effect of girdling the phloem tissue above and below the bunch on the levels of different metabolites in collected sap samples (cv. Muscat Gordo at ripening, 'Brix: 10.2).

Treatment Sucrose Total amino acids potassium

ttmol/ L5 ml ofbuffer solution*

Non-girdled 0.402 0.257 3.61

Girdled 0.089 0.035 2.t9

*The exudation period was 4 hours. Subsampled from pool of samples collected from l0 bunches.

Girdling treatment reduced the amount of each of the compounds exuded compared with the non-girdled bunches. Amino acid levels in the buffer solution were particularly affected by girdling but sucrose was also reduced to low levels and potassium was decreased. In another sampling with both girdling and exclusion of EDTA from the buffer solution the levels of sugars exudated from phloem was measured. The treatments showed a pronounced effect on the sugar levels exudated into the buffer solution with no sucrose and glucose detected in the buffer solution without EDTA and the levels of sugars exuded from girdled bunches were at the lower limits of detection (Table 6.6). 80

Table 6.6 The effect of girdling and exclusion of EDTA on the phloem sap exudation (cv Muscat Gordo at ripening, 'Brix: 15.8).

Treatments Sucrose Glucose Fructose ¡ttnol/ 1.5 mL of buffer solution*

Non-girdled 1.969 0.089 0.045

Girdled 0.045 0.044 <0.039

Buffer - EDTA <0.039 <0.039 0.089 *The exudation period was 6-7 h , Subsampled from pool of samples collected from 60 bunches.

Later in the season the phloem flux control treatments were also applied to other cultivars. Tables 6.7 and 6.8 show that simila¡ results were obtained using cv. Sultana with relatively high sucrose levels exuded overnight from the bunchstem of ripe bunches. Exclusion of EDTA from the buffer solution greatly decreased the levels of sugars, amino acids and potassium exuded.

Table 6.7 The effect of exclusion of EDTA on the phloem sap exudation (cv. Sultana at ripening "Brix : 18.0).

Treatment Sucrose Glucose Fructose Wol/ 1.5 mL of buffer solution*

Buffer + EDTA r.579 0.044 0.044

Buffer - EDTA <0.039 <0.039 0.044 *The exudation period was 5 h. Subsampled from pool of samples collected from 60 bunches.

Table 6.8 The effect of exclusion of EDTA on the phloem sap exudation (cv. Sultana, ripe "Brix:23.8).

Treatment Sucrose Glucose Fructose Amino acids Potassium Umol/ 1.5 mL of buffer solution+

Buffer + EDTA 39.386 0.658 0.570 1.27 6.08

Buffer - EDTA <0.039 0.088 0.132 0.011 0.538 *The exudation period was l6 h (overnight), Subsampled from pool of samples collected from 60 bunches 81

6.3.4 Phloem sap diurnal Patterns

Pøttern of sucrose in exudates during a day (before veraison) Fluctuations in concentation of sucrose \¡/ere seen in phloem exudates taken during one day from Muscat Gordo grape bunches before veraison ("Brix : 4.5). The samples were taken every 1.5 h. during the day and analysed for sucrose, amino acids and potassium. The results a¡e the mean of three replicates with l0 bunches in each. The exudate samples collected from each group of 10 bunches were pooled and analysed. Sucrose level increased during the day reaching its highest at midday (13.30- 15.00) achieving a maximal level of 2.68 pmoles/ 1.5 mL of the buffer solution and then showed a steady decrease during the afternoon (Fig. 6.5 A). Levels of hexoses (glucose and fructose) were found only in trace quantities in the collecting buffer solution and showed no distinct diurnal patterns (data is not shown). Potassium levels also showed diurnal variation during the sampling period and gained the maximum value of 2.16 pmoles/ 1.5 mL of the buffer solution at 13.30-15.00 h. (Fig' 6.5 C). The pattern of potassium was similar to sucrose but the variation was less pronounced. Amino acid levels were lower than sucrose or potassium. Total amino acids also showed diurnal variation with a similar pattern to potassium. However, the contribution of specific amino acids to the total pool varied during the diurnal cycle (Fig. 6.6). The principal amino acids were glutamine, glutamate, aspartate, arginine, alanine, and proline (Fig. 6.6 A-F). All the principal amino acids showed a diurnal pattern, but alanine stayed high in the afternoon samples (Fig. 6.6 E). The sampling time in this experiment was only during daylight and stopped in the afternoon. 82

Sucro.a A

3.0

2.O

1.0

tr I 0.0 I Amlno rcld¡ B ã o 0.6 o J 0.5 t¡

o 0,¿l

ts 0.3 u?

o2 go o E 0.1 Pol...lum c

22

1.8

1.4

1.0

0.6 o oo o o q o qq o q o N oq o o o o oo o o q o qrì q rì o c; NO 6 (o Sam pllng tlme (hrs.)

Fig. 6.5 The diurnal pattern of sucrose (A), total amino acids (B) and potassium (C) levels in phloem sap exudates collected from pre-veraison grape bunches during the day (vertical ba¡s indicate SE). 83

,r(x¡ 2. A + G¡J{ D ^RG ¿ 350 16 - u? ¡o ) E 300 ) 1a rq E ll o o ?.9 10 l¡ õ o E c õ cE 2ú 6

r50 2

50 11 B E + G.IJ +A¡.À

12 12 .G U? c? t¡4 l0 ) E E I u? I UÌ o 6 t8 ! 6 o o õ E E c c 10 I

2 2

2A 6 c F ASP + PRO 2. E E - 5 uì t¡?

18

) È 4 E ul úì 14 ú o s o 10 o õ E cE c 6

2 o o o o o o o o o ô o o o o o o o o o o o o o o o o o o o o o o o o o Sampllng tlme (hrs.) Sampllng tlm€ (hrs.)

Fig. 6.6 (A-F) The diurnal patterns and the levels of major amino acids detected in the grapevine phloem sap (pre-veraison). 84

Døy and night diurnal patterns (øfter veraison) Experiment I. To investigate the effect of time of cutting off the bunch laterals on the appearance of the sucrose exudation peak, another set of samples with different starting times were collected from Muscat Gordo grape bunches when the berries had just started ripening ("Brix : 10.2). The whole sampling period covered a day, a night, and the next day. Three series of sampling (4, B, and C) were done at two hourly intervals, the first set started at 06.00 hr. Each of the other two sets started four hours later than the previous one. The analysis of collected samples showed that regardless of the time that sampling started, sucrose content in the buffer solution increased from late afternoon and reached its highest during the dark period (Fig. 6.7 A, B, and C). From 22.00 hr. to next midday (12.00 hr) samples were collected for 8, 6, and 6 h periods from the same group of bunches. The results showed that after midnight, sucrose levels decreased to very low levels during the light period (Fig. 6.7 A, B, and C). A series of samples was also collected from the bunches that were pre-treated with girdling as described before in 6.2.3. The results of sugar analysis showed very low levels of sucrose and no diurnal pattern of sucrose in the exudates collected from girdled bunches (Fig' 6.7 D). Experiment II. The samples that were collected after veraison (Fig. 6.7) showed a different sucrose diurnal pattern compared with those from before veraison (Fig. 6.5). To investigate diurnal patterns further, a third series of samples were collected at 6 hour intervals from Muscat Gordo bunches later during ripening ("Brix : 15.6). Each set of sampling series started 12 hours later than the previous one. In line with previous results, the sucrose exudation was highest in the samples collected at night and then decreased in samples which were collected during the day (Fig. 6.3). Each sampling group showed the highest sucrose flux during 24.00 hr to 06.00 hr regardless of the time of beginning of the sampling. The sucrose exudation continued for 48 hours for each sampling group and then decreased. There was no diurnal pattern observed for sucrose levels in the exudates collected from girdled bunches in this experiment (Fig. 6.8). Two series of the samples which had been started first and second, were also analysed for potassium and for amino acids. It was found that potassium levels in both samples analysed showed large variations but in a different pattern to that for sucrose. Potassium exudation was high during the light period when sucrose was low in nearly all samples (Fig. 6.10). Total amino acids also showed a diurnal pattern for the first two days (Fig. 6.9). Individual amino acids displayed different patterns of exudation during the sampling period. In both of the light and dark periods glutamine, aspartate, and glutamate were the principal components of the phloem sap amino acid pool (Fig.6.1l A-C). Levels of arginine showed no significant changes in response to light or da¡k periods (Fig. 6.1I D). Levels of arginine were higher in the later samples than in the samples collected earlier in the hrst two days of sap collection. Arginine changed asynchronously reaching a very 85 high level at the third day whereas exudation of other amino acids ceased by that time (Fig. 6. ll D).

3.s0

3.00 -|-A ¿ ----{-B ô¡ 2.50 ------<,- J c Þ vì 2.00 D (Girdlcd) -ô- o f r.so q

L 1.00 (t):

0.50

0.00 o o o o o o o o o o o c o q o q c q I o o ñ ç @ o c; N (o N o o N (\¡ o o O Õ o o o o o q q o q q q q q ? c I (oc (o @ o N ç @ CO o ñ¡ ÔJ o o N N o Sampling time course

Fig. 6.7 The diurnal pattern of continuous sucrose exudation from post-veraison Muscat Gordo bunch laterals during light and dark periods with different starting times (each value represents mean of l0 bunches), 'Brix = 10.2. A, B, and C represent three samples where the bunch laterals were removed at different times. For girdled sample (D) the canes were girdled the day before sampling. The values are for two hours of each sampling period. (o @

6.00

Started at 6 pm day I (series l)

5.00 Started at 6 am day 2 (series 2)

\o Started at ó pm day 2 Þl 4.00 E Started at 6 am day 3 ra Girdled and started at day I o 3.00 E q) U' o 2.OO ¡r È¡ (â 1.00

0.00 o o o o o o o o o O o o o q q q q q q q o q q q + (o (\l c) $ (o $j @ + (o ôJ co (\¡ o o¡ o C\.1 o r I a I I a t I I I -¡ I o o o o o o o o o o o o q e q q q q q q q q q q @ s (o o¡ @ $ (o o.t @ sf (o o.t C\¡ o F (\¡ o r c! o Sampling time course

Fig. 6.8 The diurnal pattern of continuous sucrose exudation from ripe Muscat Gordo bunch laterals during light and dark periods with diflerent starting times (each value represents mean of l0 bunches), "Brix: 15.6. 87

0.8 to Serles I I Serles 2 E ul 0.6

go o È 04

gao (¡ g o.2 o E E

0.0 o o o o o o o q q q c q c c q ç (o N .D rl (oN @ !l (o N (o o î¡ o N o o o o o o o oo o o o o o q q c q c cq c c q q c TD \t (o N (D rt@ C\¡ @ t (o 6¡ N NO N o Sam P ng tlme

Fig. 6.9 The diurnal pattern of total amino acids measured in ripe Muscat Gordo grapevine exudates during days and nights (samplings started at different times and each value represents mean of l0 bunches), (fìrst and second series of samples presented in Fig. 6.8).

I €(o -_F S€riß 1 J S€ries 2 E ul -

gt0 o 4 E

E

rgl 2 õ o o.

0 o o oo o o qc q q q q @ N @ ç.þ N (D !t (D N @ N r? N o o o o oo o o o o o o o o oo q q 9 o q q CD (o ñ @!t @ N q) (o N N o Fô, o ñJ o Sampllng t me

Fig. 6.10 The diurnal pattern of potassium measured in ripe Muscat Gordo grapevine exudates during days and nights (samplings started at different times and each value represents mean of l0 bunches). 88

¡l(x) 2g A + 6.ttðl D + SaDt I E (ln Ssiö2 4 Arg Scrb¡2 to o 200 a 300 - a - ã ¡ = ¡ 150 o o 200 J E E l¡l ul to0 ì a ! o o r0o o É, E 50 Ê Ê

0

120 50 B + Seds I E + S.rbs I S€rbs 2 G sefis 2 3 Ala (o 100 G¡J o 40 o - 3 ã 3 .o EO Ã 30 o o J J 00 E E ul ul 20 ì 10 a o ! o o E E 10 t 20 c

0

120 12

F Soneg 1 E c Sems 1 i Senes 2 @ A5p SoE 2 Pro 100 o 10 - - o o - t ¡ 80 ¡f o o J J E 60 E 6 ul ul

a 10 ! !6 o o E Ê E 20 E

0 oooooooooo ooooo woooaôooooo çooø NONeN@No N@NPN oo,ooo oo,oo oo o o oooooooooo ooooo @lÔwøroco 6?oo6 o N -NO?-N€-?NO -N N Sampling timc Sampling tlnc

Fig. 6.11 (A-F) The diurnal pattern of individual amino acids measured in ripe Muscat Gordo grapevine phloem sap exudates during days and nights (samplings started at different times and each value represents mean of l0 bunches). 89

6.3.5 Grapevine leaf exudates

The exudate samples collected from grapevine leaves of cvs Muscat Gordo, Sultana and Shiraz were analysed for sucrose, glucose, fructose, amino acids and mineral ions and the results are presented in Table 6.9. The values represent the exudate collected from the cut end of petioles of 15 leaves over 16 hours. Low levels of metabolites were exuded from grapevine leaves compared to grape bunchstems. In leaf samples sucrose was high compared to glucose and fructose which in most samples were below detection levels (Table 6.9). Higher sucrose was measured in samples collected in 1995 (in dark) than those of 1994 (in light). The samples with higher sucrose levels had higher potassium and amino acids as well (Table 6.9). In the leaf exudates the same amino acids predominated as were found predominant in the bunchstem exudate i.e. glutamine, glutamate, aspartate, and alanine (Appendix I.3). When the leaves were kept under light conditions the proline levels were higher and alanine levels decreased under the same conditions (Appendix I.3).

Table 6.9 The levels of sucrose, glucose, fructose, potassium and total amino acids of leaf phloem sap exudated under light and dark, collected at ripening period.

Cultiva¡s Muscat Sultana Shiraz Muscat Suløna Shiraz Gordo Gordo r995 t994 I Compounds ¡rmol/ 1.5 mL of buffer solution/ 16 h

Sucrose 0.307 0.1l4 3.772 <0.04 0.1 06 <0.04

Glucose <0.04 <0.04 0.088 <0.04 0.085 0.042

Fructose <0.04 <0.04 0.t32 <0.04 0.043 <0.04

Potassium 1.35 1.27 1.96 r.23 0.88

Total amino acids 0.043 0.066 0.133 0.023 0.054 0.03 I The 1994 samples were collected in the light and the samples for 1995 were collected in dark. 90

6.4 Discussion

6.4.1 Validation of the sap collecting method

Although the EDTA-facilitated method of phloem sap collection has been used previously (King and Zeevaart, 1974; Hanson and Cohen, 1985; Glad et al., 1992a), a number of controlled experiments were conducted to further validate the method (Tables 6.4 - 6.9) The major constituents of phloem sap exudates were sucrose, potassium, and amino acids (Table 6.1 ,6.2) which is consistent with other analyses of phloem sap (Girousse et al., I99I; V/iener et al., l99l; Glad et al., !992a). When EDTA was omitted from the buffer solution, the amounts of each of those compounds detected in the exudates was greatly decreased (Tables 6.5, 6.7,6.8, and 6.9), indicating that EDTA has a key role in facilitating exudation of phloem sap from the bunch laterals. This has also been observed in other plants (King andZeevaart, 1974 Hanson and Cohen, 1985) and is thought to involve chelation of calcium by the EDTA to stop callose formation in the sieve tubes. Girdling of the cane also reduced the levels of sucrose, amino acids, and potassium in the exudates (Tabls 6.6, 6.7 , Figs. 6.7, 6.8), as would be expected if these compounds were derived from phloem sap entering the bunches. All of these results support the notion that the material collected from bunch laterals by this technique is derived from the phloem sap entering the bunch for distribution to the developing berries.

6.4.2 Grapevine exudate composition; comparison with tissue composition

The data from Table 6.1 shows that in the grapevine phloem sap collected by the facilitated technique, sucrose is the main sugar translocated this was also observed by \ Swanson and El-Shishini (1958), Koblet (1977) and by Glad et al. (1992a). The levels of glucose and fructose were consistently low in all samples collected at different times and this may suggest a non-phloem origin for the hexoses. It is possible that some sucrose molecules are hydrolysed or inverted after moving into the collection solution or diffuse out of sieve tubes during normal translocation and are hydrolysed in adjacent cells to glucose and fructose which are not mobile in the phloem. The values for glucose and fructose were lower than that reported by Glad et al. (1992a), probably due to mo¡e lignified and mature rachis tissue used in this experiment compared to the young tissue (at flowering time) used by Glad and coworkers. In all three cultivars the most abundant amino acid present in the phloem sap was glutamine followed by glutamic acid, aspartic acid, alanine, serine and proline (Table 6.3 and Appenix l). These amino acids have also been reported as a major component of phloem sap in some other plant species (Hall and Baker, 1972; Leckstein and Llewellyn, 1975; Pate, 1980; Hoking. 1980; Simpson and Dalling, 1981; Urquhart and Loy, l98l; 91

Wiener et Fukumorita and Chino, 1982;Fisher and MacNicol, 1986; Girousse etal.,l99I; and in all al., 1991). Since glutamine was predominant all through the growing period grapevine phloem cultivars studied it appears to be a major translocating amino N form in to be a major amide sap which was also reported by Glad et al. (1992a). It is also reported et al'' in grapevine xylem sap (Anderson and Brodbeck, 1989a; Glad et al', 1992b; Glad 1994; Anderson et al., 1995). potassium accumulates in considerable amounts in the grape berries after veraison and during berry maturation and ripening (Somers, 1975; Hale, 1977; Hrazdina and Moskowitz, 1980; Iland and coombe, 1988; Lang and Thorp, 1989; Morrison and Iodi, grape 1990; Witliams and Biscay ,IggI; Creasy et al., 1993; Boselli et al., 1995). In ripe and, berries K constitutes about 70%o of mineral cation content of the berry (Hale , 1977) of the total berry K. nearly half may be located in the skin (Somers, 1975; Iland and Coombe, 1988; Gutiérrez-Granda and Morrison, lgg2). Among the cations, potassium has a role in stimulating the unloading of assimilates (Ho, lgSS) and was the major mineral found in the grapevine phloem exudate, which is in agreement with the results reported by Glad et al. (1992a). The potassium levels in Muscat Gordo sap exudates were always higher than in Sultana and Shiraz (Table 6.2). Although xylem exudate is reported to contain considerable levels of potassium (7300pM) (Glad et al. 1992b), its presence in the phloem sap in relatively high concentration indicates that phloem sap is also an important source of potassium accumulating in the ripening berries. In addition the high levels of potassium and other cations in the phloem sapprobably serves to balance the electrical charge due to the presence of some anions. Glad and coworkers (1992a) reported organic acids in grapevine phloem exudates and the pH of phloem sap to be alkaline (pH= 7.7). Changes in K content in fruits is generally associated with phloem flow and in grapes is usually contrasted with Ca accumulation (Lang and Thorp, 1989; Morrison and lodi, 1990) which is thought to be transported only in the xylem (Marschner, 1983). Further studies will be required to quantif, phloem sap composition at different stages of berry development such data could then be used to determine the flux of carbon and nitrogen into the fruits.

6.4.3 Phloem sap flux and solute accumulation in grape berries

The influx into grape berries of nearly 250 mm3 ¿-l per berry of phloem sap was measured by Lang and Düring, (1991). The authors found thatthe stream showed little indication of sensitivity to the day/night cycle and compared with a rate of transpiration of about 200 mm3 ¿-l per berD,. They also measured backflowof xylem of at least 50 mm3 d-l from the berry to the vine. As a result, the influx of phloem sap is more or less in balance with the efflux of xylem sap and transpirational loss to the air ( Lang and Thorpe, 92

+ the light and 19g9; Lang and Düring, 1991). Phloem sap inflow of 0.52 0.28 mm3 h-l in 0.32 + 0.15 mm3 h-l in the dark at pre-veraison, and 5.79 t 1.99 mm3 h-l in light and (1994). 1.62 + 0.32 mm3 h-l in dark at post-veraison, was reported by Greenspan et al. The net flow for the xylem during sugar accumulation is from fruit to vine ( Lang and Thorpe, 1989), and there are a number of reports of xylem backflow at some stage in fruit development (Hamilton and Davis, 1988; Eh¡et and Ho, 1986; Lang and Thorpe, 1989)' 'Water backflow from the xylem also was observed by Greenspan et al. (1994) after veraison in water stressed berries. With the assumption that after veraison all solutes accumulated in the berry are imported via phloem, estimates were made of the required rate of phloem sap flux into Muscat Gordo berries from the known rate of increase in solute increase rate in the berries over time. I. From what is reported by Coombe (1980), the solute levels per berry of Muscat Gordo were 120 mg at day 70 after flowering (at the beginning of ripening) and reached 520 mg per berry at day 100. This increase in solutes over 30 days occurred when berry volume increased by one cm3 and it is equal to 400mg/ berryl30 days or 13.3 mg/ berryi day = 73.9 pmoles/ berry / day as glucose equivalents. il. The same calculations were done using the data from Coombe (1989). Glucose * fructose concentration in the Muscat Gordo berry increased by 13'4 mgl mL of juice per day between the second day after veraison and day 12. If we assume one mL increment of juice per Muscat Gordo berry this conesponds to an increase of: 13.4 mglberry/ day : 74.4 ¡tmoles /berry/ day as glucose equivalents. III. From data presented in chapter three (Fig. 3.2 graphs 4,5,6) solute increase in Muscat Gordo berries was calculated, using the formula suggested by Coombe (1980). Between day 50 after flowering ('Brix: 8.0), and day 100 ('Brix =22.0) berry volume increased by approximately one cm3. The amount of solutes per berry was249.6 mg at day 50 and increased to 915.2 mg/ berrl'at da1' 100. The amount gained per berry was 665.6 mg/ berry during 50 days or: 13.3 mg/ day/ berry, =73.9 ¡rmoles/ berry/ day as glucose equivalents. ¡y. Another calculation was done using the data recorded during phloem exudate collection between two sampling times. The solute increase on a per berry basis was I 1.56 mg/ berry /day over 14 days when 'Brix increased from 10.2 to 15.6 and berry volume increased 0.6 cm3 during the same time which corresponds to an increase of: 11.56 mg daylberry :64-2 pmoles/ berla/ day as glucose equivalent' The similar values resulting from the use of different sources of data indicates that the method of calculation is reasonable and suggest that the rate of increase in solutes of the berries is relatively constant during ripening and in the range 64-74 pmoles glucose equivalents I berrylday. To compare the values of sucrose and other metabolites measured in phloem exudate collected in this experiment, similar calculations were done for the 93 amount of metabolites exuded from phloem into the buffer solution using the number of berries on the laterals before they were removed to collect phloem sap exudates: Using the data in Fig. 6.8, for first24 h, the total sucrose exuded for the samples of to series I was calculated to be 9.78 pmoles per lateral holding 33 berries. This is equal 0.59 pmoles/ berry/ day glucose equivalent. Also total sucrose exuded for the samples of series 2 (Fig.6.8) was calculated to be pmoles/ berry/ day 1 1.45 pmoles/ lateral holding 25 berries /day. This was equal to 0.92 glucose equivalent. Using data presented in Fig. 6.7.(A) for first the 24 h and with similar calculations and 32 benies per lateral a value calculated was 0.73 pmoles/ berry/ day glucose equivalent. Thus the daily rate of sugar efflux from cut laterals was in the range 0.59 - 0.92 pmoles glucose equivalents I berrylday. The maximum rate of efflux can also be calculated using the peak rates recorded.

Using the maximum values of sucrose exuded during exudate collection the following values were obtained: Fig. 6.8 (Series 1);4.56 pmolesi 4 h: 18.24 pmoles/ dayl 24hl 33 berries: 1'l pmoles/ berry / day glucose equivalent. Fig. 6.8 (Series 2); 5.04 pmoles/ 4 h : 20.16 pmoles/ dayl 24 hl 25 berries : l'6 pmoles/ berry / day glucose equivalent. Fig. 6.7.(C) 3.29 pmoles/ 2h:39.48 pmoles/ dayl 24W 32 berries: 2.5 pmoles/ berry/ day glucose equivalent. The data presented in Table 6.1 for sucrose exuded from Muscat Gordo laterals before : veraison was 3 .24 pmoles/ 4h equal to 19 .44 pmoles/ day I 25 berries 1.5 pmoles/ berryi day glucose equivalent. The data presented in Table 6.3 for sucrose exuded from Muscat Gordo laterals at before veraison was 3.77 ¡rmoles/ 4h equal to 15.08 pmoles/ dayl22 berries: 1.4 pmoles/ berry/ day glucose equivalent. Mean value for the calculations done above was 1.6 pmoles/ berry/ day. Compared with the above calculation of sugar flow in each berry of about 80 pmoles/ berryl day, this suggests that the phloem exudation technique has collected 2o/o of that which is normally accumulating in the berry in situ.

Similar calculations can be done for amino acids and potassium. Amino acids were shown to increase in Muscat Gordo berries during approximately 14 days from 1730 pmoles/ 100 mL of juice to 3230 ¡rmoles/ 100 mL (Kliewer 1969) during the time when 'Brix increased from 18.4 to 21.5 which was assumed to take about two weeks. Then the calculated value of amino acid accumulation would be 1.07 pmoles/ mL /day. If we 94

be equal to assume that berry volume increases by 0.6 cm3 during 14 days then it will 0.643 pmoles / berry/ daY. : From data presented in Table 6.1 amino acids exuded was 2.7 pmoles/ 25 berriesl day 108.7 nmoles/ berry/ daY. From the data given in Fig. 6.9 (series l) the value for amino acids exuded was 38.6 nmoles/ berry / day during first 24 h . From the data given in Fig. 6.9 (series 2) the value for amino acids exuded was 34.4 nmoles/ berry / day during first 24 h and also the value for series I, collected at the same time parallel to series 2 was 65.6 nmoles/ berry/ day Mean value for the calculations done above was 61.8 nmoles/ berra/ day which is equal to l0o/o of amino acids calculating to be accumulating into a berry at the same time.

Using maximum values for amino acids exuded during 24 h Fig. 6.8 (series l) 82.4 nmoles/ berrylday: l3Yo of amino acids accumulating into a berry at the same time. Fig.6.8(series2)68.3nmoles/berry/day= llo/oofaminoacidsaccumulatingintoa berry at the same time.

Boselli et al. (1995), reported that potassium content increased approximately 0.02- 0.03 mg b..ry-' d-r during the entire period of fruit development. = 0.51 '0.17 pmoles/ berry/ day. Potassium concentrations were shown to increase about 1000 mg/ L in berry extract from veraison to ripe stage during 56 days in Chaunac grapes (Hrazdina et al. 1984)' = 0.92 pmoles /berry /day Potassium increased from 1.5 mg/berry'to 4.7 mglberry during 65 days inChardonnay grapes (Possner and Kliewer 1985). :1.26 pmoles/ berry idar'. In Shiraz berries potassium per berry increased from about one mg per berry to about 3.8 mg per berry when "Brix increased from 7 fo 25 (lland and Coombe 1988). If approximately 40 days are required for the ripening process in Shiraz berries then, the potassium increase rate will be; 1.8 ¡rmoleslberryl day. Thus the mean value of potassium accumulation would be 1.05 pmoles/ berry/ day. From data presented in Fig. 6.10, (series 1) for Muscat Gordo phloem exudate, K was exuded into the buffer solution at the rate of: 0.376 pmoles/ berry/ day or 360/o of the calculated accumulation rate. From data presented in Fig. 6.10, (series 2) K was exuded at the rate of: 0.339 pmoles/ berry/ day or 32o/o of the calculated accumulation rate. 95

value Using mæ

6.4.4 Seasonal changes in exudate flux

Considerable amounts of sucrose, amino acids, and potassium were exuded into the collection buffer solution from different grape cultivars and at different stages of growth and development during the season (Tables 6.1). The amount of sucrose exuded during 4 hours of sampling varied with ripening and showed lower levels of sucrose in the samples collected at veraison than both before and after veraison for all three cultivars (Table 6.1). After fruit set the developing berries are powerful sinks with high demands for metabolites while they are rapidly increasing in size (Hale and Weaver, 1962). The strong sink activities of berries declines during early veraison and rises again during sugar accumulation (Coombe, 1989). There are a number of indications that at the onset of ripening in grapes a breakdown of xylem connections to the berry occurs and after this stage, the majority of water that accumulates in benies derives from phloem sap (Coombe and Bishop, 1980; Lang et al., 1986; During et al.. 1987; Findlay et al., 1987; Lang and Thorpe, 1989; Creasy et al., 1993; Creas¡'and Lombard, 1993; Greenspan et al., 1994). The increased sink of activity of berries after the inception of ripening , i.e. after verison (Hunter et al., 1995) suggests an increased phloem sap flux into the berries. The seasonal variation in the phloem sap composition seems to be a natural endogenous phenomenon in grapevine and has also been reported in the other plants Q'{obel, l99l). 96

6.4.5 Diurnal changes in exudation

Diurnal patterns of metabolites were found in the grapevine exudates (Fig. 6.5-6.11). This may result from a native diurnal pattern in phloem sap or it may have been the result of wounding when the lateral is cut. To investigate this further two experiments were conducted to determine the effect of the time of excising the bunch laterals from the bunch for exudate collection. The time of the beginning of the sample collection (cutting the bunch lateral off) had no effect on the sucrose exudation pattern (Figs. 6.7 , 6.8) suggesting that the amount of metabolites exuded are most likely related to the source activities and is not a wound response. It is noteworthy that the diurnal patterns were different at the different growth stage of grapes (compare Figs 6.5, 6.7, 6.8). The xylem is the source of water inflow for grape berries before veraison with phloem playing a substantial but secondary role. This is reversed within only two days after veraison, when the xylem supply appears to diminish greatly, and phloem becomes the major vascular pathway for inflow (Coombe and Bishop, 1980; Lang and During, 1991; Greenspan et al., 1994). It is possible that changes in the berry development and the lag phase before veraison could be the cause of alteration of the diurnal pattern between day and night. However, the actual cause of this change is not known. It is known that during the day photosynthetic carbon assimilation usually exceeds the export of the carbon from the leaf so carbohydrates accumulate in leaf chloroplasts as starch or in the cytosol and vacuole as sucrose. During the night, sucrose is mobilised from the vacuole and is catabolised from starch and exported from the leaves to maintain the translocation rate in the absence of photosynthesis (Candolfi-Vasconcelos et al., 1994; Hunter et al., 1994). There is no information on diurnal changes in the grapevine phloem sap composition but there are some reports about diurnal patterns of carbohydrates in grapevine leaves. Carbohydrate concentrations of grapevine leaves were reported to be highest at l3:00 h.. then decreased during the afternoon and reached low levels at l8:00 h¡. (Hunter et al., 1994). Carbohydrates translocated to sink organs of a grapevine could originate fiom newly fixed carbon, or be derived from starch or from carbohydrate reserves in wood (Coombe. 1989; Aloni et al., l99l ; Williams and Biscay, 1991 ; Glad et al., 1992b; Candolfi-Vasconcelos et al., 1994)). It may also take some time for newly fixed carbohydrates to reach the sink organs such as berries. For example, Hale and Weaver, (1962) reported that the labelled CO2 fed to grapevine leaves was detected in the berries after 6 hours. Chaumont et al. ( 1994) showed that both starch and sucrose accumulate in grapevine leaves during the photosvnthesis period, but while sucrose levels increased at early morning and remained almost constant during the rest of the day with a slight decrease during the night. starch content of the leaves increased at mid afternoon and decreased during the dark period. Johnson et al. (1982), showed that carbon dioxide enrichment 97

grapevines resulted in accumulation of high levels of starch in the leaves of treated carbohydrates compared to the vines grown in ambient carbon dioxide. Accumulation of variations in in excess of export, provides a mechanism by which, in spite of remarkable day and the availability of light and carbon dioxide, leaves tend to ensure a continuous the night supply of assimilated carbon for plant metabolism and export of assimilate to various sink tissues (Geiger, 1987; Ho, 1992; Chaumont et a1.,1994). Although the data presented here are not sufficient for exact determination of the biochemical mechanisms responsible for the changes, they provide an important foundation for future biochemical investigations of assimilate translocation in the grapevine. Further studies are needed to check seasonal patterns and the mechanism period, and responsible for diurnal changes between day and night and during the growing to quantiff C and N inflow into fruits.

6.4.6 Grapevine leaf exudate

The results of analysis of leaf exudates showed that the phloem sap depafing from the leaves towards the sink organs contains sucrose and also amino acids and potassium (Table 6.9). There are reports indicating that hexoses are present at high levels in grapevine leaf tissue (Kliewer, 1965; Kliewer and Nassar , 1966; Sepúlveda and Keliwer, 1986; Hunter et al., lgg4) but the results here indicated that the levels of hexoses in phloem exudates are very low (Table 6.9) suggesting that sugars are not translocating out ofthe leaves as hexoses. In the samples collected from leaves at the berry ripening stage sucrose, potassium and amino acids were exuded at higher levels in samples collected in 1995 (in the dark) compared to those collected in 1994 (in the light). The priority of carbon export from the leaves under low light or in darkness and reduction of the ratio of exporlstarch synthesis under high light has been reported for some plant species (King and Zeevaart 1974; Maillard and Procher, l99l). There are some reports suggesting selective export of amino acids from leaves via the phloem (Weiner et al., 1991; grapevine leaves was Riens et al., 1 991 ). The profile of amino acids in the exudates of slightly different fiom what was detected in peduncle exudates, showing higher levels of alanine, serine and proline. Thus there may be some selectivity in the export of amino acids via the phloem. 98

6.5 Conclusion

method The research presented in this chapter demonstrates that the EDTA-facilitating collected is appropriate for collecting grapevine phloem exudates. The analysis of samples the using this method revealed that sucrose is the main carbohydrate translocating in exudate grapevine phloem sap. Potassium is the major mineral ion in grapevine phloem phloem collected from grapevines. Amino acids are important constituents of grapevine these principale exudates with glutamine being the principal amino acid. The amounts of to flow components collected from the cut peduncle were smaller than amounts calculated acids' into each berry in situ. The composition of grapevine phloem sap (sugars, amino and potassium) is subject to seasonal changes. There is clear diurnal pattern of assimilate translocation of sucrose and other metabolites in grapevines with higher rates of transport at night. 99

Chapter 7

Assimilate transport in the grapevine ; secondary metabolites

Contents Page

7.1 Introduction....-.... -----.101 T.2Watenals and methods...... --'.---.--- 103 7.2.lPlartmaterials... -..-.-----.. 103 7.2.2 Other materials...... --.--. 103 7.2.3 Phloem sap collection...... ---.------103 I 03 Standard method...... Exclusion of EDTA...... --.--.--..-103 Girdting...... -.-.--..--. 103 Vineyard control samples..... --.--.-. 103 Washing solution..... --. 104 7.2.4 Glycosyl glucose determination...... -. 104 7.2.5The secondary metabolite aglycone analysis...... 104 7.3 Results ...... 104 7.3.1 G-G in phloem exudates from peduncle laterals and leaves and comparison with the G-G of whole berry homogenates.....- ...... ,....104 7.3.2 Secondary metabolite aglycones in phloem exudates of grapevines 105 7.3.2.1In phloem exudates from peduncle laterals ...... 105 7.3.2.2lnphloemexudatesfrompetioles...... 106 7.3.2.3 Glycoside analysis of peduncle ...... 108 7.3.2.4 The effect on the composition of secondary metabolites of variations in the exudate collection technique...... 108 Omitting EDTAfrom the collection solution 108 Girdling lreotment...... 108 7.3.2.5 Secondary metabolites present in grapevine tissues 109 Plant tissue surfoce compounds ...... 109 In the wash solution ..-. .-.....109

7 .3.2.6 Effect of the diurnal pattem of phloem flux on the monoterpenes collected.._...... 109 7.4 Discussion...... l l l 100

7.4.1 Glycosylated secondary metabolites.. - -... -... - -. -.

7.4.2 Seasonal variations and varietal differences. - - - - i------7.4.3 Treatments to interfere with the phloem sap flux---.. 7.4,4 Compounds washed from plant tissue surface and wound.... 7.4.5 Peduncle tissue analysis..... 7.4.6 The concentration of secondary metabolite glycosides in the phloem exudate samples with high and low sucrose concentrations----.--.---.-----.."""""" 114 115 7.5 Conclusion -... 101

., 1. i , ,1,,i''i lj i ,, , ,, l',",i,{.,. 1 l'ij -ir,- n .

7.1 Introduction Phloem transport in vascular tissues of plants enables the transfer of photosynthetic assimilates from mature leaves to sink tissues in the rest of the plant. Export of photosynthetic products from source leaves is dependent on many complex metabolic events that control the production of solutes which are translocated in phloem (Stitt and euick, 1989; Geiger and Fondy, 1991). Fruits are strong sinks for assimilates translocating via the phloem (Coombe, 19S9). The assimilates at the first stage of fruit development are used for tissue development, and then accumulate as storage products. Sucrose is reported as a major constituent of phloem sap translocating in almost all plants studied so far, including grapevines (Glad et al.,I992a). In the ripening phase, especially after the lag phase at veraison, grape berries accumulate a wide range of secondary metabolites, including flavour compounds, along with sugars and other metabolites (Wilson et al., 1984; Strauss et al., 1986; Williams et al., 1989; Shure and Acree, 1994). Berry softening, sugar and secondary metabolite accumulation are com.mon ripening changes that are accompanied by an increase in the respiratory quotient of the berries. This is due to a change in respiratory substrates as, during this period, the berry undergoes a considerable biochemical differentiation (Brady, 1987; Pandey and Farmahan, 1e77). Biosynthesis and accumulation of flavour compounds in grapes have been the subject of extensive studies (Williams et al.. 1989; Park et al., 1991;V/illiams, 1993). The most common secondarv metabolites that accumulate in grapes include flavour compounds in floral grapes (Wilson et al.. 1984: Williams et al., 1985; Park et al., 1991), polyphenolic compounds (Considine. 1979: Hard¡ and O'Brien, 1988; Hawker et al., 1972) and anthocyanins in coloured cultivars (Pirie and Mullins. 7976, 7977, 1980). Flavour compounds (e.g. monoterpenes) are trace constituents of grape berries, and even in juice of floral cultivars such as muscats monoterpene flavour compounds are present at concentrations of only 1-2 mglL (Gunata et al., 1985a). Since leaves are the primary source of photosynthetic metabolites translocating to fruits. grapevine leaf composition has been studied; secondary metabolites are reported as important constituents of leaves of several plant species including the grapevine (Wildenradt et al.. 1975: Gunata et al.. 1986, 1985a; Otsuka et al., 1989; Schulz and Stahl- Biskup, 1991; Humpf and Schreier,lggl: Tazaki et al., 1993; Skouroumounis and Winterhalter, 7994 ). Terpenols and aromatic alcohols in free and glycosidically bound 102

leaf tissues forms were reported in Muscat of Alexandria (syn. Muscat Gordo Blanco) and (Gunata et al., 1986). It was further shown that the grapevine leaf was richer in free increased bound aromatic alcohols than the berry, and terpenyl glycosides in the leaves (1975) progressively with grape maturation (Gunata et al., 1986). Wildenradt et al. grapevine leaves of the reported six-carbon compounds and also terpenes at high levels in cultivar'Chenin Blanc' among the 84 different volatile compounds identified in a leaf extract. The observations of several researchers that similar secondary metabolites are present in grapevine leaves and in fruits has led to the suggestion that these compounds are probably being translocated from leaves to the fruits (Gunata et al., 1986; Skouroumounis and Winterhalter, lg94). However, apart from this similarity in composition and other circumstantial evidence, such as that arising from the observations reported in Chapter 3 of this thesis, there is no reliable evidence to support the concept of translocation of secondary metabolites from leaves to fruit. The increase in monoterpene flavour compounds and other secondary metabolites in grape berries during the ripening stage could result from the contribution of phloem components to berry growth and composition, or these compounds may be synthesised in the berry independently. The data in Chapter four strongly suggest that secondary metabolites are synthesised in the berry, also the possibility remains that berries of particular cultivars may have mechanism to selectivly import and accumulate these compounds (see section 4.4). The important question is whether secondary metabolites are actually translocating to grape berries via phloem sap. Furthennore, if translocation of secondary metabolites is occurring in the grapevine, is this a mechanism for the accumulation of flavour compounds alternative to the direct biosynthesis of those compounds in the berry itself? There is a lack of information on the exact site of biosynthesis of flavour compounds and the relationship between leaves and berries regarding secondary metabolite accumulation in the fruit during berry maturation and ripening. Although phloem sap is the most important source of solute accumulation in berries (Coombe et al., 1987b). Phloem sap has not been analysed for flavour compounds in any plants, therefore collection and analysis of phloem sap could help to answer these questions. An understanding of the biosynthesis and accumulation of grape flavour compounds may have considerable practical consequences by providing valuable guidance in the application of viticultural practices to optimise flavour production in fruit. This knowledge also could provide a foundation for genetic engineering of grapevines. To provide more direct evidence of any possible transport of secondary metabolites in phloem sap, the role of assimilates translocating from leaves to the berries has been investigated through analysis of the phloem sap. This has been achieved by determining the prof,rle of the secondary metabolites in grapevine (Vitis vinifera L.) phloem sap 103 exudates collected from both leaf petioles and peduncle laterals at three intervals throughout the growing season, and at six hourly intervals throughout a 48 hour period.

7.2 Materials and methods

7.2.1 Plant materials

Described in chapter 6, (6.2.I).

7.2.2 Other materials

Described in chapter 6, (6.2.2).

7.2.3 Phloem sap collection

Standard method Phloem sap was collected as described in chapter six, (6.2.3). This protocol is called the standard method in this chapter to allow differentiation when the protocol was modified to investigate treatments involving samples collected from non-phloem sources.

Exclusion of EDTA To assess the effect of EDTA on the composition of phloem exudate, the EDTA was excluded from the collection solution used in the standa¡d method and only HEPES buffer (10 mM; pH 7.5) was used for exudate collection as a control treatment.

Girdling The girdling method was described in chapter 6 and illustrated in Figure 6.2.3.4

Sampling ond analysis ortifocts

Samples of the buffèred collection solution, before and after contacting the blue-tack adhesive that was used to hold the collection tubes in place, were enzyme hydrolysed, extracted and subjected to GC-MS analvsis. The peaks observed in the chromatograms of these samples were assigned as sampling and analysis artefacts. These artefact peaks were subtracted from the chromatograms of the phloem exudate analysis.

Vineyard control sampl es

To investigate components contributed from the plant surfaces a series of samples was collected as vineyard controls. ln these samplings different tissues with no cut or wound were dipped into the buffered collection solution for three hours. These solutions were then enzyme hydrolysed. extracted and subjected to GC-MS analysis. 104

ÍYashing solution As described in chapter 6, (6.2.3) before the main sampling tubes containing buffer solution to be attached to the peduncle lateral, the cut end of the laterals were washed with the same buffer solution by rinsing them for 5-10 minutes. This solution was taken for analysis as washing solution to determine the compounds that came out of cut cells and wounds, dwing the washing procedure.

7 .2.4 Gly cosyl glucose determinatio

Glycosyl glucose determination was done as described in Chapter 3, (3.2'5)'

7.2.5 The secondary metabolite aglycone analysis

The secondary metabolite aglycone analysis was done as described in Chapter 4, (4.2.4)and the compounds we¡e identified and quantified using the same method as employed in chapter four.

7.3 Results

7.3.1 G-G in phloem exudates from peduncle laterals and leaves and comparison with the G-G of whole berry homogenates

Phloem exudate was collected from fruits and leaves of grapevine cultivars of Muscat Gordo Blanco, Sultana, and Shiraz as describedin 6.2.3. The phloem sap exudates were analysed for glycosidically bound secondary metabolites (G-G) as described in chapter 3, (3.2.s). Results of G-G analvses of the phloem sap exudates are given in Table 7 .l and the data are compared with G-G values of berry homogenates made on fruit sampled and analysed at the same development intervals. 105

Table 7.1 Glycosyl glucose levels measured in phloem sap exuded into buffer solution from peduncle laterals and leaves, and also in berry homogenates of Muscat Gordo, Sultana and Shiraz at three berry development stages'

nmol G-G/ 1.5 mL * ¡tmol G-G of buffer solution. per g fr.wt.

Cultivars Peduncle exudate Leaf exudate Berry homogenate BVVRBVVR BVVR

Muscat Gordo 6.8 2.8 2.8 2.2 2.7 1.5 0.4 0.4

Sultana 7.0 3.2 1.6 1.3 0.9 1.6 0.2 0.1

Shiraz 1.8 1.8 2.5 - 0.8 1.5 0.3 0.2 0.6

* The exudation duration time for fruit bunches was 4 hours and for leaves was l6 hours

- no data available; BV= before veraison, V= veraison, and R: ripe'

Each value represents mean of three replicates.

7.3.2 Secondary metabolite aglycones in phloem exudates of grapevines

7.3.2.1In phloem exudates from peduncle laterals

After the glycosidically bound fraction of the collected phloem exudates were isolated by retention on Cl8 reverse phase Sep-pak cartridges, the isolated fractions were enzymatically hydrolysed. liquid-liquid extracted, and the aglycones analysed by GC-MS. Complex mixtures of glycosylated secondary metabolites were present in the collected solutions as can be seen from the chromatograms of the aglycones presented in Fig. 7.1' (ABC). Compounds originating from the enzyme and/or solvent were identified by GC-MS analysis of a HEPES buffer control which was enzyme hydrolysed and liquid-liquid extracted in the same way as phloem exudate samples. After eliminating these artefacts from the volatiles recorded in the GC-MS analyses, the remaining compounds were assumed to be of plant origin. Analysis values are shown collectively in Table 7 .2 and the remaining discussion in this section of the chapter concerns these compounds only. Monoterpenes, including several with recognised flavour properties, made up a notable proportion of the volatile aglycones from the bound secondary metabolites in the phloem exudates of grape peduncle. The concentration of monoterpenes in l.5 mL aliquots of the phloem exudate collection solutions obtained in 4 h from fruit peduncle laterals of Muscat Gordo, Sultana, and Shiraz at three ripening stages are provided in Table 7.2.A The 106

are concentrations of C13-norisoprenoids and benzene derivatives in the solutions similarly shown in Table 7.2.4 Muscat From the results in Table 7.2.A it is evident that the exudates collected from than Gordo peduncle were richer in glycosides of most of the monoterpenes observed in were the exudates from either Sultana or Shiraz. This observation parallels the findings Sultana Chapter 4 in that Muscat Gordo berries are richer in monoterpene glycosides than or Shiraz berries.

7.3.2.2In phloem exudates from petioles

The phloem exudate of detached leaves was also collected for Muscat Gordo, Sultana, and Shiraz grapevines. The glycosidically bound fraction of the samples were isolated, enzymatically hydrolysed, and arralysed by GC-MS. Results of these analyses for the monoterpenes and for C13 norisoprenoids and benzene derivatives observed are given in Tables 7.2.8. Of the three varieties, Muscat Gordo leaf phloem exudate appeared to have the highest concentration of monoterpene glycosides. Only the glycoside of geraniol was corilnon to all three varieties. A 107

c

B

MC ¡ ooo 500a Fig. 7.1 GC-MS chromatograms of the aglycones isolated from phloem exudates collected from ripe fruit bunch laterals of Muscat Gordo (A), Suløna (B), and Shiraz (C) grapes. 108

7.3.2.3 Glycoside analysis of peduncle The method of phloem sap collection used in these experiments involved the collection solution being in contact with cut and exposed vine tissue surfaces. It was considered necessary therefore to determine the extent, if any, of contamination of the collected phloem sap by non-phloem vine tissue sources during the sampling processes. Peduncle tissue, with the vascular bundles therein, is a passage for metabolites translocating to the fruits, mainly from leaves and also from other parts of the plant' The peduncle tissue of the three cultivars of this experiment was extracted, the glycosides isolated, hydrolysed, and analysed in the same way that phloem exudates were processed. The results of these analyses a¡e shown in Tables 7.2.C' The peduncle tissues for all th¡ee cultivars were a rich source of some of monoterpene, C13-norisoprenoids, and benzene derivatives aglycones (Table 7.2.C). Va¡ietal differences were evident in the concentrations of many of the compounds observed.

7.3.2.4 The effect of variations in the exudate collection technique on the composition of secondary metabolites

To investigate the occurïence in the collection solutions of compounds of non-phloem origin, some treatments were applied before and during the collection. These treatments were designed to inhibit, or cut ofi the phloem sap flux. The samples collected in this way were analysed for aglycones of secondary metabolites.

Omitting EDTA from the collection solution As was shown in Chapter 6 (6.3.3) excluding EDTA from the collection solution stopped phloem sap flux as indicated through a termination of sucrose flux. The first treatment therefore was an omission of EDTA from the buffer solution used for sap collection. Excluding the EDTA. resulted in a major reduction in the number and concentration of aglycones which were detectable in the hydrolysed exudate solution, in comparison with the standard collection protocol (see Table 7.2.D, Control I). This reduction was evident for samples collected both at veraison and early ripening stages (data not shown).

Girdling treslment The second treatment to stop phloem sap translocating in vascular bundles was girdling. Analysis of the exudates collected from girdled bunches (Figure 6.4) showed that girdling on the cane above and below the fruit bunch, before sampling eliminated phloem sap flux as demonstrated by the absence of sucrose flux. The girdling was done about l8-20 hours before sampling; the samples were collected in the presence of EDTA and then analysed for secondary metabolite aglycones. Data for these analyses were compared with data for 109

the samples taken in parallel from un-girdled bunches with EDTA omitted and by girdled bunches standard protocol. The results of analysis of exudates collected from no effect on revealed that the girdling treatment, despite stopping sucrose flux, had little or the number and concentration of the aglycones compared to the samples collected from non-girdled bunches (see Table 7.2.D, Control II).

7.3.2.5 Secondary metabolites present in grapevine tissues

Plant tissue surface comPounds Another set of samples was collected to determine the possible contribution to the secondary metabolites observed in the collection solutions made by plant tissue surfaces. For this purpose the samples were collected by putting grapevine tissues (organs) such as tendrils, uncut bunch stems or the dried end of canes with old cuts, in buffer solution. In order to get detectable levels of compounds in the sampling solution from these sources, the contac t area between the tissue and the solution was made as large as possible. Therefore, the results of analysis of these samples (see Table 7.2.D, Control III, vineyard) are important because of the presence of the compounds observed rather than their concentration levels. The analysis of these samples showed that most of the compounds detected were similar to those of exudates collected from peduncle laterals.

In the wash solution As previously described(7.2.3), for phloem sap collection in this experiment, the cut end of a peduncle lateral was dipped into the buffer solution immediately after cutting and the cut surface was rinsed for 5-10 minutes to wash compounds from the cut cells before exudate collection. The solutions which were used for washing the cut ends of the laterals were pooled and analysed. The results of this analysis of wash solution are given in Table 7.2.D. Control III. (washing solution). The results showed that some monoterpenes and also some Cl3-norisoprenoids and benzene derivatives were present in the washing solution. The levels of sonte compounds in the washing solution were higher than had been observed in samples of phloem exudates for the same compounds in spite of the differences in time length for washing and exudate collecting (5-10 minutes and 3-4 hours respectively).

7.3.2.6 Effect of the diurnal pattern of phloem flux on the monoterpenes collected

The phloem sap exudates collected during day and night periods showed a diurnal pattern for sucrose flux (see chapter 6). The analysis of secondary metabolites in samples containing high and low levels of sucrose collected from the same bunches during day and 110

night periods could show the contribution of phloem sap flur in the levels of secondary metabolites exuded into the collection solution. A series of collections made during aday, a night and the following day, yielded samples with low levels of sucrose, collected over 12 h during the first daylight period, samples with high levels of sucrose, collected over 12 h during the night period, and samples with low levels of sucrose, collected over 12 h during the second day period. These were taken for hydrolysis and GC-MS analysis of the major aglycones. Data for the concentration of two of the aglycones detected in these samples, i.e geraniol and 3,7-dimethyloct-2-ene-1,7-diol are presented inFig.7.2. The monoterpene aglycones were present in all th¡ee collected samples; those with low sucrose concentrations had higher levels of the compounds detected in the buffer solution than the samples with high sucrose levels.

140 I Geraniol

n 3,7-Dim"thylocÞ2ene.1 ,7 diol 120 t8 ! 16 100 o c 0 4 J o o 2 80 a ET o 0 c J 60 o I o o o 6 40 c E o õ f 4 20 2

0 0 o o o o o q o qo o o o o o N @ ? 6 o o @ ô¡ N @ N o o o ôl @ (\l o o o q o o o o o a q q o o o o o I N @ o N o N @ N @ d (\l @ Sampllng tlme cou¡se

Fig. 7.2 The concentration of secondary metabolite aglycones shown as ba¡s (right axis) in the phloem exudate samples collected over a 36 h period. The sucrose concentration (left axis) in the same samples also shown. 111

7.4 Discussion

7.4.1 Glycosylated secondary metabolites

To investigate the site of biosynthesis of flavour compounds accumulating in grape berries, and the possibility of translocation of these compounds via the phloem sap, grapevine phloem exudate from peduncle and leaves was collected. These solutions were initially analysed for total glycosidically bound compounds, i.e. G-G. Data for these analyses (Table 7.1) show that most of the collection solutions contained glycosides. The peduncle exudates of Muscat Gordo and Sultana, collected before veraison had higher levels of glycosides than the exudates collected at later stages (Table 7.1). This trend was also seen in the G-G values obtained for the fruit homogenates. Shiraz fruit bunch phloem exudates and berry homogenates also showed a parallel development of G-G concentation with berry development, but opposite to that found for Muscat Gordo and Sultana. The exudates collected in 16 h from leaves at veraison and at berry ripeness had concentrations of glycosides higher than were found in the 4 h collection solutions of peduncle exudate. To investigate the individual volatile secondary metabolites that were bound as glycosides, larger volumes of the phloem exudates were collected, enzymatically hydrolysed, and the released aglycones were analysed by GC-MS. The results of these chromatographic analyses showed more than 60 compounds were observable in the enzymic hydrolysates of the exudates collected from fruit bunches of the grapes (Data is not shown). Most of the aglycones detected and identified in these experiments were in one of following categories; monoterpenes, Cl3-norisoprenoids, benzene derivatives, and fatty acid-derived constituents.

7.4.2 Seasonal variations and varietal differences

To study the seasonal variations of the glycosidically bound volatiles, peduncle exudate samples were analysed at three different stages of berry development. Although the concentrations of most observed monoterpenes in the collected exudates varied with the developmental stage, no consistent patterns were evident. Thus, unlike the high G-G concentrations seen before veraison in the peduncle exudate of Muscat Gordo and Sultana, the concentrations of only a few individual monoterpene glycosides, e.g. those of citronellol, 3.7-dimethyl-1,7-octandiol and 3,7-dimethylocta-2-ene-1,7-diol in Muscat Gordo showed such a trend (Table 7.2.A). Several individual monterpene glycosides were maximal at ripeness.

From the data in Table 7 .2.A it can be seen that glycosides of C13 norisoprenoids and benzene derivatives were significant constituents of the phloem sap exudates collected 112 from all three varieties. As was the case with the monoterpenes, no obvious pattern of development was evident in the concentration of these aglycones with ripening. Such inconsistent patterns of secondary metabolite occurrence in the phloem sap exudates with berry development give little support to translocation as the mechanism of secondary metabolite accumulation. The decreasing concentration of secondary metabolite glycosides in the exudates during the growing season is due possibly to the increasing maturity of the peduncle tissue rather than changes in the composition of phloem sap as it has been recorded that young plant tissues normally contain more of these compounds (Stahl-Biskup et al. 1993). There were also varietal differences in the profiles of the secondary metabolites detected in the peduncle exudates. Regarding monoterpene compounds, Muscat Gordo peduncle exudate appeared to contain a higher concentration of these constituents than did Sultana or Shiraz exudates (Table 7.2.A). The profile of C13-norisoprenoid and benzene- derivative constituents of the exudates were similar for all th¡ee cultiva¡s. Fewer monoterpene aglycones were detected in the hydrolysates of leaf phloem exudates of the grapevines than were observed in corresponding hydrolysates of the phloem exudates of the fruit bunches (compare data in Table 7.2.8). Of the three varieties, Muscat Gordo leaf phloem exudate appeared to have the highest concentration of monoterpene glycosides. Only glycosides of the furan linalool oxides and geraniol were common to all three varieties. No such differences in either compound type or glycosides. concentration were evident for the C ¡ 3 norisoprenoid and benzene derivative

7.4.3 Treatments to interfere with the phloem sap flux

In chapter six it was shown that some treatments could stop phloem sap flux. Phloem exudates were collected fiom Muscat Gordo grape bunches after application of these control treatments. The samples u,ere analysed by GC-MS to observe the effects of interfering with the phloem flux on the glycosylated secondary metabolites and thus determine whether these compounds were translocated in the phloem sap. Excluding EDTA from the buffer solution stopped exudation of sucrose into the buffer solution (6.3.3. Control. I.). Two samples collected at different times with EDTA omitted from the collection solution showed diminished levels of secondary metabolite aglycones (see Table 7.2.D). Samples were collected from girdled bunches, using buffer solution containing EDTA, The results of the analyses of exudates collected from girdled bunches compared to those of non-girdled bunches. ie using the standard protocol, showed much smaller differences in both the number and concentrations of the compounds detected than betweetr the samples collected from non-girdled bunches with and without EDTA (seeTable 7.2.D, 113

Control II). These observations demonstrated that although girdling stopped the sucrose flux, the occu¡1ence of the secondary metabolites in the collection solutions was largely unaffected. This result, in combination with the absence of aglycones in collections made with the omission of EDTA, indicated that the presence of EDTA in the collection solutions rather than phloem sap, was responsible for the occurrence of the secondary metabolites in the collections. EDTA can move from solution into the vascular bundles of the vine tissue and exsert its effect inside the vascular cells (King and Zeevaart 1974) causing elution of any secondary metabolites resident in the tissue cells. The parenchymatic cells and the tissue surrounding vascular tissue are possible sources of the secondary metabolites.

7.4.4 Compounds washed from plant tissue surfaces

Control samples for determination of any alternative sources of secondary metabolites were collected by washing uncut Muscat Gordo grapevine tissues surfaces into the buffer solution. Data for the analysis of these samples in Table 7 .2.D, Control III show that the collection solutions made from plant tissue surfaces contained a number of the monoterpenes, C¡3-norisoprenoids and benzene derivatives seen in the exudate collections made under standard conditions. It appears that these compounds in some way have exuded to the surface of the tissues and could be washed off the external surface of the grapevines in the presence of the EDTA. This suggests that secondary metabolites have exuded from the plant cells, through the epidermis, to the tissue surface. Such migration has been shown for compounds that accumulate within grape berries as storage products, these can pass through the cuticle and epicuticular layer and reach the surface of the berry (Padgett and Morrison 1990). The results of the experiments reported here show that this type of secretion of compounds possibly occurs in all aerial parts of grapevines. As another control. the solution used for washing the wounded or cut end of the laterals from which the exudates \À'ere collected, was analysed (Table 7.2.D, Control III ). The results showed that a f'ew monoterpenes, notably geraniol and geranic acid, and also some C13-norisoprenoids and benzene derivatives were present in the washing solution. The levels of some conipounds in the washing solution were higher than had been observed in samples of phloem exudates, despite the differences in the duration of the washing and exudate collection regimens ( 5-10 minutes and 3-4 hours respectively). The relatively high levels of tlie compounds detected suggests that they probably are present in high concentrations in the peduncle tissue cells as well as on the tissue surfaces. 114

7.4.5 Peduncle tissue analYsis

The analysis of peduncle tissue by GC'MS showed that some monoterpenes as well as norisoprenoids and benzene derivatives were highly concentrated in this tissue (Table 7.2.C). Although the compounds present in the peduncle tissue were common to most of the phloem exudate samples some were clearly not. Peduncle tissue was rich in monoterpene compounds such as geraniol, nerol, citronllol, linalool oxides, Z-2,6- cta'2,7 - dimethyloc ta-2,7 -diene- 1 ,6-diol, 3,7-dimethy- I ,7-octandiol, E-2,6-dimethylo diene-1,6-diol, and 3,7-dimethy-1,7-octandiol. However, there were other monoterpenes in samples collected from peduncle lateral exudates which were not in the peduncle tissue extacts e.g. o-terpinol and citronellol. In light of the data discussed in 7.4.3 and7.4.4, and the similarities among the compounds in the phloem exudates and those in the peduncle tissue, it seem likely that the latter tissue is the source of these compounds.

7.4.6 The concentration of secondary metabolite glycosides in the phloem exudate samples with high and low sucrose concentrations

It was assumed that if secondary metabolites are translocating in the phloem sap, the exudates collected at periods of high sucrose flux should contain higher levels of these compounds as well. Based on this assumption the phloem exudates with high and low levels of sucrose that were collected at the peak and trough times of phloem sap sucrose flux were analysed for secondary metabolite aglycones. It was found that the exudates containing higher sucrose had lower amounts of detected aglycones (Fig. 7.2). The absence of a positive correlation between sucrose concentration and the levels of aglycones detected in the same exudates, strongly reinforce the view that the secondary metabolites are not translocated in the phloem. 115

7.5 Conclusion

The experiments described in this chapter were aimed at establishing the presence, and possible translocation in grapevine phloem sap, of glycosides of volatile secondary metabolites. Whilst many volatile secondary metabolites were detected as aglycones in phloem exudate collections, the occurrence of these compounds apþears to be the result of their extraction from vine tissues rather than their presence in the phloem sap. These deductions were drawn from the results of aglycone analyses applied in three experimental approaches. In experiments designed to interfere with the phloem sap flux, the importance of EDTA in the collection solutions as an agent facilitating the extraction of the glycosides from the vine tissues was shown. Other experiments, involving analysis of homogenised peduncle tissues and of collections made from plant tissue surfaces, confirmed the ubiquity of the secondary metabolite glycosides in all tissues examined. Finally, the absence of a correlation between the amounts of two major secondary metabolite aglycones in exudate collections and phloem sap flux, the latter indicated by sucrose concentrations in the collection solutions, showed that the secondary metabolites followed no diurnal pattern like the phloem flow. These results, together with the data in Chapter 4, show that secondary metabolites are not transported into fruit from other parts of the vine. It is concluded that the berry itself is the site of biosynthesis of these compounds. Table 7.2 The aglycones determined in hydrolysates of phloem exudate collections and tissue extracts 116

Tal¡lc 7.2.4 Monoterpene, C¡3-Norisoprenoids and benzene Table 7.2,8 Monoterpene, C¡ Tal¡lc 7.2.C The monoterpene, 3- Tnl¡le 7.2 D The aglycones detected in the samples from control T able 7.2.8 Monoterpene, C t 3 derivative aglycones determined in hydrolysates of phloem Norisoprenoids and benzene derivative Cl 3-uorisoprenoid and benzene lreîtrnents that interfere with phloem flux consisting EDTA, girdlíng and norisoprenoids and benzene derivative sap exuded into a buffered collection solution from fruit bunch aglycones determined in hydrolysates of le¡f derivative aglycones determíned also in tissue surface and wound washings. compounds detected in Muscat Gordo, laterals. Data are in pmol/ I .5 mL of collection solution/ 4h. petiole / exudate exuded into a buffered in peduncle (bunch stem tissue) prnol/ I .5 mL of buffer solution/ 4h, Sultana and Shiraz berry homogenates, lateral, (BV= before veraison, V= veraison, and R= ripe). collection solution. Data are in pmol/ 1.5 homogenates (pmol/ g fr.wt.). þmoU g fr. vrt.). mL of collection solution/ l6h. / Ieaf, (V=

and R=

Peduncle exudate Leaf ptrloem exudate Peduncle tissue extract Muscat Gordo exudate Benies Muscat Sultana M. Gordo Suløna Shiraz M.Gordo Shi¡az Sult¡na Control I II M. Gordo Sultana Shiraz Std,r No srd. No Girdled BV VR BVV Vinyard Washing RBVVR VRVRVRRRR EDTA EDTA solutionr' RR Monoterpenes R

Furan J linalool oxide-l 0.4 4 2t face bacc trace 5 tace face bace 7 913 bace Furan linalool oxide-2 21 tsac€ tracc tracc 4 trac€ tracc tsacc 9 Linalool 270t 39 Hotrienol 678 tsace Myrcenol J 3 E trac€ trace , traße cis-Ociminol 2 5. bacc trace ,,

hans-Ociminol )) 7 tace trace 3 bace þran linalool oxide-l tracc 2127 t0 a-Terpinol 3402832 2 493224 3 52 linalool þran oxide-2 il5 Cironellol 976 4 tsace

Nerol 5 23 5099 trace 331 7æ 32 Geraniol 62955tt745t026?6 t09n2523 l3t8t8 366/. 7539 8l 3 t65 I t86 49 64 4368 il0 tacc Geranic acid 7S 86 t04 142 252 87878 406t 7563 8l 17 396 tace ¡@ 9 t4 2t40 70 2,6-Dimethylocta-3,7-dienc-2,6-diol t5r54 ¡16 3,7-Dimethylocta-1,6-diene-3,5-diol 675 2,6-Dimethylocta- 1,7-diene-3,6-diol 683 Z-2,6-Dimethylocta-2,7-diene- 1,6-diol lt2 )) l4 8 7 4 5 a a 50 -3 3394 268 233 6 o4 0.4 9 103 1441 ?s2 2æ 3,7-Dimethy- 1,7{ctandiol 6 5 4 6 J 4 ¡3 6 I 8 )) il 14il 4242 497 2854 l8 42 ?3 50 E-2,6-Dimethylocta-2,7-diene-1,6-diol t5 I8 l5 l7 6 36 3 9 32 ?o 3568 592 668 23 26 87 940 1532 500 3,?-Dimethylocta-2-ene- 1,7-diol t37 75 55 t8 9t4 4 46 .t0 48614 205 t3308 389 23 3r¡ 7 402 477 590 52 p.Menth- l -ene-6,8-diol t6 ¡3 p-Menth- I -ene-7,8-diol 213 529 70 Total monoterpencs 325 3¡8 287 56 tE 48 76 a 25 29 529 59 284613 9224 60t 27 il5 32495 44 925 8 825 8¡4 79 336¡ I 2685 840 Sucrose 3240 l0t0 2700 9200 2200 65æ 2M 250 t.mo 40 N 855 ¡06 212 40 342 40 t969 39 4S e e o C¡3-Norisoprenoids

3-Hydroxy-7,8-dihydro-b.ionol )) 27 36 5 45t3 58 t3710 4490 89 6l 63 t00 50 t70 3-Oxo-b-ionol l6 t5 25 43 26 ¡9 26 39 l9 trace 3-Oxo- a-ionol ,) 24 26 4 2 II 5 5 6 7023 3349 t7Ø 44 2t 5¡ l2t 35 70 60 3-Hydroxy-5,6-epoxy-Lionone 9 9 4 il t3 t0 il il face trace trace 2644 1648 6t2 30 3 24 Eacc 37 23 bace 30 60 50 Grasshopper ketone 29 53 25 42 t2 57 47 77 35 6t 38 530 70 trâce 57 186 41 2æ 650 277 Benzene deriv¡tives

Trimethoxy benzen metha¡rol 30891229 5 35 t2 3 t099 67 4 33 5t4 7 76 26 tracc 310 40 l,m Trimethoxy phenol 707859t2940 2t 38 t8 I 8 t303 43 874 833 4 752 39 88 26 l0 40 7 6 0.4 6 2 4a7 6 Uacc 6 - : not detected I Std. - Stândsrd mcthod ofphlocm sap collection in bulfcr + EDTA solution. (E notmcasu¡cd Trace = less than 0.I pmol +r The time duration for washing thc wound was 5- 10 minutc¡, One analysis was done for each datum 117 Chapter I

General discussion

Although knowledge of flavour precursors and bound volatile aroma compounds in grape berries has been extended over the last decade, much more remains to be clarified including the site of biosynthesis of these compounds in different tissues of the vine. Secondary metabolites and, most importantly, flavour compounds accumulate in the grape berries in glycosylated forms during berry development but very little is known about where they are made. Glycosylated secondary metabolites can accumulate in almost all living tissues of the plant (Sugisawa et al., 1988; Lukner, 1990), though in remarkably different concentrations. The leaves of plants are the sites of active photosynthesis and carbon assimilation and leaves also accumulate glycosylated secondary metabolites (Tazaki et al., 1993). The secondary metabolites that accumulate in grape berries could either be translocated from photosynthetically active sites such as leaves to other sink organs, or synthesised in the tissue itself. In this work and through data presented in Chapter 3 it has been shown that the grapevine leaves accumulate large quantities of glycosylated compounds including those of volatile secondary metabolites. The immature berries of grapes accumulate glycosylated secondary metabolites during the first phase of growth. The berries of floral grape Muscat Gordo had a different pattern of accumulation of total glycosides to the non-floral Sultana grape berries. Muscat Gordo berries accumulated more secondary metabolite glycosides during ripening than Sultana berries. in fact accumulation of these compounds in sultana almost stopped by the beginning of ripening. This occurs despite the fact that Sultana berries had higher levels of glycosides than Muscat Gordo berries at the early stages of berry growth and development. Thus, metabolically active green tissues of young berries, and also leaves, are rich in secondarl, metabolite glycosides and the accumulation of volatile secondary metabolites that are specific to the flavour of the berry occurs later. This phenomenon, which has not been reported before, shows that differences exist between the metabolic activities of berries of Muscat Gordo and Sultana cultivars. This suggest, firstly, that glycoside accumulation in the berries is not linked to sugar accumulation and does not occur through the same mechanism, although both processes can occur coincidentally. This is in agreement with observations reported by Wilson et al. (1984), Williams et al. (1985) and Park et al. (1991) who concluded that biosynthesis and accumulation of monoterpene glycosides in grapes is independent of sugar translocation, Secondly, the high levels of glycosides in the leaves dose not necessarily increase the levels of glycosides in the berries. 118

From the data presented in Chapter 3 and also that reported by Gunata et al' (1986), Bravdo et al. (1989), Skouroumounis and Winterhalter (1994 ) it is known that many secondary compounds, including glycosides of flavour compounds are present in grapevine leaves. There are two possibilities; synthesis of these compounds in the leaves and transportation to the berries, or their formation in the berries themselves from precursor molecules imported via the phloem sap. In this work the possible translocation of volatile flavour compounds from grapevine leaves to the berries or, altematively, their synthesis in the berries, were investigated using two different methods i.e. bunch grafting between cultivars described in chapter 4, and phloem sap analysis' Monoterpene compounds are important constituents of floral grape varieties such as Muscat Gordo (Williams et al., 1982a, b, Wilson et al., 1984; Gunata et al., 1985a), so they were the main compounds that were analysed in the fruit. The results obtained for monoterpene profiles in the grape bunches after being grafted between cultivars indicated the exclusive role of the berries in accumulation of monoterpene compounds. The same spectrum of monoterpene glycosides accumulated in grape bunches, irrespective of whether the fruit developed on its own vine or following bunch grafting onto genetically different vines. Thus the ability to accumulate monoterpene glycosides resides within grape berries and is independent of the rest of the vine. However, there still remained the possibility that the monoterpene glycosides \ryere imported into the fruit so that Muscat Gordo berries, for example, might have mechanisms to selectively import and accumulate these compounds, while these mechanisms were not active in grapes of the non-floral varieties. The grafting experiments do indicate that the inability of non-floral grapes to accumulate monoterpene glycosides does not result from a lack of these compounds or their precursors in the phloem sap circulating around the vine, since fruit of Muscat Gordo grafted onto these varieties still produced the normal levels of monoterpene glycosides. ln Chapter 5 it was discovered that Muscat Gordo berries accumulate anthocyanin pigments at the over-ripe stage (Gholami and Coombe, 1995). Constantly similar profìles of anthocyanin in the Muscat Gordo berries from grafted and non-grafted bunches and also from vines with different root systems, support the view that berries are independent of leaves and rootstock for synthesis and accumulation of anthocyanins as well as other secondary metabolites, which was concluded earlier from the grafting experiment (Gholami et al., 1995). It seems that the berry itself synthesises monoterpene flavour compounds, and anthocyanins, although precursors of these compounds may be produced in other parts of the vine. The hypodermal cells of the berries are the most probable sites of biosynthesis and also storage of these secondary metabolite compounds as was also suggested by Wilson et al. (1986). The grape berries have a controlling role on metabolism of flavour compounds produced and stored within them. 119

The phloem sap analysis method was used to investigate any possibility of glycosylated secondary metabolites being translocated in grapevines. The samples collected using the EDTA-facilitated method showed that the main carbohydrate translocating in the grapevine phloem sap was sucrose and potassium was the major mineral ion. Amino acids were also important constituents of grapevine peduncle exudates with glutamine being the principal amino acid in the sap. The composition of grapevine phloem sap exudates (sugars, amino acids, and potassium) were subject to diurnal and seasonal changes. There was a clear and constant diurnal pattern of metabolite translocation in grapevines with high rates of assimilate transport at night under normal conditions. Diurnal variations in the concentration of sucrose in phloem sap translocating to the berries, supports the suggestions of Brown and Coombe (1985) that va¡iable phloem sap unloading into the apoplast has the primary control over sugar accumulation in the berry rather than compartmentation within pericarp cells. The composition of the phloem exudates and the experiments in which the canes were girdled and EDTA was omitted from the buffered collection solution, support the idea that the technique provides a sample of phloem sap. However, the calculations presented in Chapter 6 indicate that the amount of sucrose and amino acids collected from the peduncle laterals is insufficient to account for the predicted rates of accumulation of these compounds in the berry. This suggests that cutting the lateral and removing the fruit has a significant effect on the phloem sap flow from the lateral. For this reason, the data may not provide accurate quantitative data on the transport of metabolites into the fruit via the phloem. The results of GC-MS analysis showed that the phloem exudates collected from both fruits and leaves contain substantial levels and complex mixtures of glycosylated secondary metabolites. However, the results obtained in Chapter seven indicate that these compounds are not translocating via phloem sap. Since, from the beginning of grape berry ripening, most of the flon, to the berry is provided by the phloem (During et al., 1987; Findlay et al., 1987; Creasl' et al., 1993; Greespan et al.,1994), the increasing levels of some amino acids such as arginine, alanine and proline, (Kliewer, 1969,1970b) and also of the tripeptide glutathione in ripening grape benies (Adams and Liyanage, 1993) supports the idea that these compounds are synthesised in the berries themselves from precursors imported via phloem sap. and implies that the grape berries are actively controlling their storage pool of carboh¡-drates and amino acids as well as proteins (Tensniére et al., 1994). A consideration of all the pathways for amino acid metabolism and secondary metabolite synthesis indicates that grape berries are metabolically active organs and should not be considered only as storage sites. Although it is difficult to determine whether secondary metabolite molecules in a phloem exudate solution are actually in transit from leaves to sink tissues, the data presented here, although insufficient for exact determination of the biochemical mechanisms responsible for the changes, does provide an important 120 foundation for future biochemical investigations of assimilate translocation in the grapevine. tr'urther studies: Substantial advances have been made in research on flavour compounds in grapes, but biosynthetic studies on whole fruit have been limited. However, low levels of flavour compounds is a major limitation in studies on the biosynthetic pathways of secondary metabolite s in Vitis vinifera species. Understanding the metabolic processes and biosynthetic pathways of flavour compounds in grapes and the mechanisms controlling them is an important aspect. Future research activities on flavour compounds should pay more attention to research at the cellular level particularly on the site of biosynthesis of important flavour compounds in the grapevine. V/hen the site of biosynthesis of flavour compounds (i.e. monoterpenes) becomes known it may be possible to alter varietal characteristics of grapes by application of appropriate techniques such as genetic engineering. Then it should become possible to adjust the vines to produce grapes with desired flavour characteristics. The results of this thesis suggest further work should concentrate on the biosynthesis of these compounds in the berries. Often products are synthesised at particular developmental stages or in a specific organ, suggesting that sophisticated control mechanisms exist to regulate their synthesis (Lukner, 1990). Plant cell culture is of potential value for studies of biosynthesis of secondary metabolites (Charlwood et al., 1988; Constable, 1988; Sugisawa et a1.,1988 and references therein). The use of this technique and grape berry cell suspension would seems to be another method available to investigate the site of biosynthesis of flavour compounds in grapevine and the potential abilities of different plant organs and cells in synthesising those compounds. Further understanding of the mechanisms by which flavour compounds accumulate in grapes will require experimentation to identify the actual sites of synthesis of these compounds and their compartmentation within the berry and the role of the bern' itself. The study of enzymes involved in monoterpene biosynthesis pathways and the role geraniol plays will be extremely helpful in elucidating the biosynthesis of these secondary metabolites. Further studies are needed to check seasonal patterns and diurnal changes at different times of the season and the effect of environmental factors on these patterns especially sucrose diurnal pattern and quantitative measurement of C an N inflow into fruits. The technique used in this work was sufficient to collect samples hopefully representative of grapevine phloem sap. nevertheless the rate of exudation was much less than the normal flow of phloem sap ìnto the berry in situ. Therefore, further studies are required to improve the phloem sap collection technique, testing buffers of differing composition and the effect of other facilitating agents. 121

References

Ackerman, LE., D.V. Banthorpe, W.D. Fordham, J.P. Kinder, and I. Poots, 1989, P-D- glucosides of aroma components from petals of Rosa species: Assay, occurrence and biosynthetic implications. J. Plant Physiol. 134: 567-572-

Adams, D.O., and C. Lyanage,1993, Glutathion increases in grape berries at the onset of ripening. Am. J. Enol. Vitic . 44 333-337.

Aloni, R., A. Raviv, and C.A. Peterson, 199I, The role of auxin in the removal of dormancy callose and resumption of phloem activity in Vitis vinifera. Can. J. Bot. 69 1825-1832.

Anderson, P.C., and B.V. Brodbeck, 1989a, Chemical composition of xylem exudate from bleeding spurs of Vitis rotundifolia Noble and Vitis hybrid Suwannee in relation to pruning date. Am. J. Enol. Vitic. 40: 155-160.

Anderson, P.C., and B.V. Brodbeck, 1989b, Diurnal and temporal changes in the chemical profile of xylem exudate from Vitis rotundifolia. Physiol. Plant. 75:63-70.

Anderson, P.C., B.V. Brodbeck, and R.F. Mizell, III, 1995, Diurnal variation in tension, osmolarity, and the composition of nitrogen and carbon assimilates in xylem fluid of Prunus persica, Vitis hybrid, and Pyrus communis. J. Amer. Soc. Hort. Sci. 120:600-606.

Aronoff. S., 1955. Translocation from soybean leaves. Plant Physiology, 30: 184-185

Baker, J., and C.F. Timberlake, (1985) The distribution of anthocyanins in grape skin extracts of port wine cultivars as determined by high performance liquid chromatograph¡,. Journal of the Science of Food and Agriculttue 36, 1315-1324.

Banthorpe, D.V., 1991, Classification of terpenoids and general procedures for their characterisation. In: " Methods in plant biochmistry" series editors; Dey, P.M. and J.B. Harborne, Volume 7 " Terpenoids" Charlwood 8.V., and D.V. Banthoepe, (Eds.) Academic Press, Ha¡court Brace Jovanovich, Publishers London. pp. l-42.

Barlow C.4.. and P.A. Randolph, 1978, Quality and quantity of phloem sap available to the pea aphid. Ann. Entomol. Soc. Am. 7l:46-48. 122

Barz, W. and J. Koster, 1981, Turnover and degradation of secondary (natural) products. In: Biochemistry of plants. Secondary plant products. Conn, E. E., ed.; Academic: London, vol. 7, PP. 35-84.

Bath, G.I., C.J. Bell, and H.L. Lloyd, 1991, Arginine as an indicator of the nitrogen status of wine grapes. Proceedings of the International Symposium on Nitrogen in Grapes and Wine. J.M. Rantz (Ed.) pp. 202-205, American Society for Enology and Viticulture.

Bernfeild, P. 1967, Biogenesis of Natural Compounds. Pergamon Press Ltd.

Bidwell, R.G.S. lgTg,PlantPhysiology. Macmillan Publishing, New York, pp.643.

Bledsoe, 4.M., W.M. Kliewer, and J.J. Marois, 1988, Effect of timing and severity of leaf removal on yield and fruit composition of Sauvignon blanc grapevines. Am. J. Enol. Vitic.39: 49 - 54.

Boselli, M., B. Volpe, and C. Divaio, 1995, Effect of seed number per berry on mineral composition of grapevine (Vitis vinifera L.) benies. J. Hort. Sci. 70: 509-515'

Brady, C.J. 1987, Fruit ripening. Annual Review of Plant Physiology, 38; 155-178.

Bravdo. B., O. Shoseyov, R. Ikan, and A. Altman, 1989, Free and bound monoterpene content of leaves and berries and their biosynthesis by in vitro grown berries. In: Proceedings of the first International Symposium on the A¡omatic Substances in Grapes and Wines. Scienza, A.; Versini, G., Eds., Michele all'Adige, Italy, Instituto Agrario Provenciale: Trento, Italy, pp 55-62.

Bray. C.M., 1983. Nitrogen Metabolism in Plants. 214 pp. Longman Group Limited, Essex, England.

Brown, S.C., 1981, Sugar accumulation in the developing grape berry, Ph.D. thesis, University of Adelaide.

Brown, S.C. and B.G. Coombe, 1985, Solute accumulation by grape pericarp cells. III. Sugar change in vivo and the effect of shading. Biochem. Physiol. Pflanzen. 180: 371-381.

Buttrose, M.S., and C.R. Hale. 1971, Effects of temperature on accumulation of starch or lipid in chloroplasts of grapevine. Planta, (Berl.) 101: 166-170. 123

Candolfi-Vasconcelos, M.C., M.P. Candolfi, and W. Koblet,1994, Retranslocation of carbon reserves from the woody storage tissues into the fruit as a response to defoliation sfress during ripening period nVitis vinifera L. Planta 192 567-573.

Candolfi-Vasconcelos, M.C., and W. Koblet, 1990, Yeild, fruit quality, bud fertility and sta¡ch reserves of the woodas a function of leaf removalin Vitis vinifera, evidence of compensation and stress recovering. Vitis' 29:199'221'

Carroll, D.E., and J.E. Marcy ,1982, Chemical and physical changes during maturation of muscadine grapes (vitis rotundifulia). Am. J. Enol. vitic.33:168-172.

Cash, J.N.; V/.4. Sistrunk, and C.A. Stutte, 1976, Characteristics of Concord grape polyphenoloxidase involved in juice colourloss. J. Food Sci. 41: L398-1402'

Charlwood, 8.V., and D.V. Banthorpe, 1987, The biosynthesis of monoterpenes, In ,.Progress in Phytocemistry", vol.5, Eds. Reinhold, L. ; J. Harborbe and T. Swain, Perganom press, Oxford, UK, pp 65'125.

Charlwood, B.V., J.T. Brown, C. Moustou, and G.S. Monis, 1988, The accumulation of isoprenoid flavour compounds in plant cell cultures. In: Bioflavour '87, analysis, biochemistry, biotechnology. Sch¡eier (Ed.), Walter de Gruyter & Co., Berlin, pp 303-3 14.

Chaumont, M., J.F. Morot-Guadry, and C.H. Foyer, 1994, Seasonal and diurnal changes in photosynthesis and carbon partitioningin Vitis vinifera leaves in vines with and without fruit. J. Exp. Bot.,45: 1235-'1243.

Charlwood. 8.V.. and K.A. Charlwood, 1991, Monoterpenoids. In: " Methods in plant biochmistry" series editors; Dey, P.M. and J.B. Harborne, Volume 7 " Terpenoids" Charlwood 8.V.. and D.V. Banthoepe, (Eds.) Academic Press, Harcourt Brace Jovanovich, Publishers London, PP. 43-98.

Chino, M., H. Hayashi. S. Nakumura, T. Oshima, H. Turner, D. Sabnis, V. Borkovec, D. Baker, G. Girousse, J.L. Bonnemain, and S. Delrot, 1997, Phloem sap composition. ln: Recent advances in phloem transport and assimilate compartmentation. Bonnemain, J.L., W.J. Lucas and J. Dainty, (Eds'), Ouset Editions, pp.64-73.

Conradie, Vy'. J., 1980. Seasonal uptake of nutrients by Chenin blanc in sand culture nitrogen. S. Afr. J. Enol. Vitic., l: 59-65. 124

Considine, J.4., 1979, Biophysics and biochemistry of fruit growth and development. Ph.D. Thesis, University of Melbourne.

Considine, J.4., and R.B. Knox,lgTg, Development and histochemistry of the cells, cell walls and cuticle of the dermal system of fruit of the grape, Vitis vinifera. Protoplasm a 99 : 347 -366.

Coombe, 8.G., 1995, Adoption of a system for identiffing grapevine growth stages. Australian Journal of Grape and Wine Research 1: 104-1 10'

Constable, 1988 Cytodifferentiation. In: Bioflavour '87, analysis, biochemistry, biotechnology. Schreier (Ed.), Walter de Gruyter & Co., Berlin, pp 161-170.

Coombe,8.G., 1973, The regulation of set and development of the grape berry. Acta Hortic., 34:261-74.

Coombe, B.G., 1975, Development and maturation of the grape berry. The Australian Grapegrower and Winemaker, No: 136.

Coombe, 8.G., 1980, Development of the grape berry. I. Effect of time of flowering and competition. Aust. J. Agric. Res.,3l: 125-131.

Coombe, B.G.,1987a, Influence of temperature on composition and quality of grapes' Acta Horticulture, 206: 23-35.

Coombe, B.G.,1987b, Distribution of solutes within the developing grape berry in relation

to its morphology. Am. J. Enol. Vitic.38: 120-127 '

Coombe, 8.G.,1989. The grape berr_v as a sink. Acta Horticulture, 239:149-158

Coombe,8.G., 1992. Research on development and ripening of the grape berr)'.4m. J Enol. Vitic . 43: l0l -l 10.

Coombe, B.G., and G.R. Bishop, 1980. Development of the glape berry. II. Changes in diameter and deformabilir_v during veraison. Austral. J. Agric. Res. 3 l: 499-509.

Coombe, B.G., and Ph. Matile. 1980, Solute accumulation by grape pericarp cells. I. Sugar uptake by skin segments. Biochem. Physiol. pflanzen. 175:369-381.

Coombe, 8.G., and P.E. Phillips, 1982, Development of the grape berry. III. Compositional changes during veraison measured by sequential hypodermic sampling. International symposium on Grapes and Wine, Jun, 1980, Davis, California, pp. 132-136. 125

Coombe, B.G., M. Bovio and A. Schneider, 1987, Solute accumulation by grape pericarp cells. V. Relationship to berry size and the effects of defoliation. Journal of Experimental Botany, 38: I 789-1 798.

Cordonnier, R., and C. Bayonov, 1974, Biochimie appliquée, Mise en évidence dans la baie de raisin, var. Muscat d'Alexandrie, de monoterpenes liés, révélables par une ou plusieurs enzymes du fruit. C.R. Acad. Sc. Paris, t.278, Série D, 3387-3390.

Cravero, M.C., Guidoni, S., Schneider, A. and Di Stefano, R. (1994) Caractérisation variétal cépages musqués à raisin coloré au moyen de paramètres ampélographiques descriptifs et biochemiques. Vitis 33, 75-80.

Creasy, G.L., and P.B. Lombard,1993, Vine water stress and peduncle girdling effects on pre- and post-veraison grape berry growth and deformability. Am. J. Enol. Vitic. 44: 193-197.

Creasy, G.L., S.F. Price, and P.B. Lombard, t993, Evidence for xylem discontinuity in Pinot noir and Merlot grapes: Dye uptake and mineral composition during berry maturation, Am. J. Enol. Vitic. 44:187-192.

Crippen, D.D.Jr. and J.C. Morrison, 1986, The effects of sun exposure on the phenolic content of Cabemet sauvignon berries during development. Am. J. Enol. Vitic. 37: 243-247. Cronshow, J., and D.D. Sabnis, 1990, Phloem proteins. In: Sieve elements, Behnke, H.-D., Sjolund. R.D. (eds.), Springer. Berlin, pp.257-283.

Croteau. R., 1984. Biosynthesis and catabolism of monoterpenes In " Isopentenoids in plants. Biochemistry' and function", W.D. Nes; G. Fuller and L.S. Tsai, eds. Marcel Dekker. lnc. Neu'York.

Croteau, R., 1987, Biosynthesis and catabolism of monoterpenoids. Chem. Rev. 87: 929- 954.

Croteau, R., and C. Maninkus. 1979. Metabolism of Monoterpenes; Demonstration of (+)-Neomenthil-p-D-glucoside as a major metabolite of (-)-Menthon in Pepperminf ( Mcnrha piperita ). Plant Physiol. 64:169-175.

Darné. G., 1993, Nouvelles hypothèses sur la synthèses des anthocyanes dans les baies et dans feuiles devigne. Olew hypotheses on anthocyanin biosynthesis in berries and leaves of the grapevine). Vitis, 32: 77-85. 126

Delrot, S., and J.L. Bonnemain, 1981, involvement of protons as a substrate for the sucrose carrier during phloem loading inViciafabø leaves. Plant Physiol. 67: 560- 564.

Delrot, S., 1981, Proton fluxes associated with sugar uptake in Vicia faba leaf tissues. Plant Physiol. 68: 706-711.

Dimitriadis, E., and P.J. Williams, 1984, The development and use of a rapid analitical technique for estimation of free and potentially volatile monoterpene flavorants of grapes. Am. J. Enol. Vitic .,32(2) : 66-71.

During, H., A. Lang, and F. Oggionni, 1987, Patterns of water flow in Reisling berries in relation to developmental changes in their xylem morphology. Vitis 26:123-13I.

Dzhakhua, M.J., E.S. Derboglav, and M.A.Dzhapariðze, 1978, Study of the volatile phenol compounds in white wines. Prihl. Biokhim. Mikrobiol. l4: 156-158.

Ehret, D.L. and L.C. Ho, 1986, Translocation of calcium in relation to tomato fruit growth. Ann. Bot. 58: 679-688.

Esau, K., 1965, Plant Anatomy, John Wiley and Sons, Inc, Toppan Company, LTD., Tokyo, Japan. pp. 270-307.

Eschrich, W., and W. Heyser, 1975, Biochemistry of phloem constituents. In: Encyclupedia of plant physiology. N.S. Vol. 1 :Transport in plants. 1. Phloem transport. pp. 101-136, Zimmerman, M.H., Milburn, J.A., (eds.), Springer, Berlin.

Feller, U. and R. Hoelzer. 1991. Transport of amino acids via xylem and phloem in detached wheat shoots. In: Recent advances in phloem transport and assimilate compartmentation. Bonnemain. J.L., W.J. Lucas and J. Dainty, (Eds.), Ouset Editions. pp.9l-95.

Fellows. R.J., and M.J. Dalling. 1974, Structural and physiological changes in sugar beet leaves during sink to source conversion. Plant Physiol. 54:877-885.

Fernández-Lopez. J.4.. Hidalgo, V., Almela, L. and López-Roca, J.M. (1992) Quantitative changes in anthocyanin pigments of Vitis vinifera cv. Monastrell during maturation. J. Sci. Food Agric., 58: 153-155.

Findlay, N.K., J. Oliver. N. Nii, and B.G. Coombe, 1987, Solute accumulation by grape pericarp cells. IV. Perfusion of pericarp apoplast via the pedicel and evidence for xylem malfunction in ripening berries. J. Exp. Bot. 38: 668-679. 127

Fisher, D.B. and P. Mac-Nicol, 1986, Amino acids composition along the pathway during grain filling in wheat. Plant Physiol. 82:1019-1023'

Flowers, T.J., and A.R. Yeo, lgg2, Solute transport in plants. Blackie Academic & Professional ( An Imprint of Chapman & Hall), Glasgow' UK'

Fondy, 8.R., and D.R. Geiger, 1977, Sugar selectivity and other characteristicsof phloem loading in Beto vulgaris I. Plant Physiol. 59:953-960'

Fong, C.H., S. Hasegawa, Z. Herman, and P. Ou, 1991, Biosynthesis of limonoid glycosides in lemon (citrus limon ). J. sci. Food Agric.,54:399-411.

Fraga, B.M., 1991, Sesquiterpenoids, In: "Methods in plant biochmistry" series editors; Dey, P.M. and J.B. Harborne, volume 7 " Terpenoids" charlWOod 8.v., and D.V. Banthoepe, (Eds.) Academic Press, Harcourt Brace Jovanovich, Publishers London, pp. 145-185.

Fukumorita, T. and M. Chino, 1982, Sugar, amino acid and inorganic contents in rice phloem sap. Plant Cell Physiol.,23:273-283.

Galet, P., (1958 and 1962) Cepages et Vignobles de France. Tome II et III. Les Cepages de Cuve, 986 et 1234.lere Partie. Dehan, Montpellier.

Gamalei, Y.V., 1985, Characteristics of phloem loading in woody and herbaceous plants Sov. Plant Physiol. 32:656-665.

Gao. Y., and G.A. Cahoon, 1994, Cluster shading effects on fruit quality, fruit skin colour, and anthocyanin content and composition in Reliance (Vitis hybrid). Vitis,33: 205-209.

Gieger, D.R., 1987. Understanding interactions of source and sink regions of plants. Plant Physiol. Biochem., 25 : 650-666.

Geiger, D.R., and B.R. Fondy, 1991, Regulation of carbon allocation and partitioning: status and research agenda. In: J.L. Bonnemain, S. Delrot and J. Dainty, W. Lucas. (eds.) "Research advances in phloem transport and assimilate compartmentation. Ouest Editions, Nants, France, pp l-9.

Gholami, M., and B.G. Coombe, 1995, Occurrence of anthocyanin pigments in berries of the white cultivar Muscat gordo Blanco (Vitis vinifera L.). Australian Journal of Grape and Wine Research l:67-70. 128

Gholami, M., Hayasaka Y., Coombe, 8.G., Jackson, J.F., Robinson, S.P. and W'illiams' P.J., 1995, Biosynthesis of flavour compounds in Muscat Gordo Blanco grape berries. Australian Journal of Grape and Wine Research, l: 19-24.

Giaquinta, R.T., 1983, Phloem loading of sucrose, Ann. Rev. Plant Physiol., 34: 347-387

Giaquinta, R.T., and D.R. Geiger, 1977, Mechanism of cyanide inhibition of phloem translocation. Plant Physiol. 59: 178-1 80.

Girousse, C., J.L. Bonnemain, S. Delrot, and R. Bournoville,lggI, Sugar and amino acid composition of phloem sap of medicago sativa: a comparative study of two collecting methods. Plant physiol. Biochem.29: 4I'48.

Glad, C., J.-L. Regnard, Y. Quero, O. Brun, and J.-F. Morot-Gaudry,1992a, Phloem sap exudate as a croterion for sink strength appreciation in Vitis vinifera cv. Pinot noir

gapevines. Vitis, 3 1 : 13 1-138.

Glad, C., J.-L. Regnard, Y. Quero, O. Brun, and J.-F. Morot-Gaudry, I992b, Flux and chemical composition of xylem exudation f¡om Chrdonnay grapevines: temporal evolution and effect of recut. Am. J. Enol. Vitic. 43:275-282-

Glad, C., J. Farineau, J.L. Regnard, and J.F. Morot-Gaudry, 1994, The relative contribution of nitrogen originating from two seasonal l5N supplies to the total nitrogen pool present in the bleeding sap and in whole Vitis vinifera cv. Pinot noir grapevine at bloom time. Am. J. Enol. Vitic., 45:327-332.

Goldschmidt, E.E.. and S.C. Huber, 1992, Regulation of photosynthesis by end-product accumulation in leaves of plants storing starch, sucrose, and hexose sugars. Plant

physiol. 99 1 443-1448.

Gonrez, E., A. Martinez. and J. Leancina, lgg4,Localization of free and bound aromatic compounds among skin, juice and pulp fractions of some grape varieties. Vitis 33: t-4.

Gomez, E., and A. Martinez. 1995, Changes in volatile compounds during maturation of

some grape varieties. J . Sci. Food Agric ., 67 : 229-233 .

Goodwin, T.W.. and E.l. Mercer, 1983, Introduction to plant biochemistry; Pergamon: Australia. p.227.

Gordon,4.J., G.J.A. Ryle, and G. Webb, 1980, The relationship between sucrose and starch during'dark'export from leaves of uniculm barley. J. Exp. Bot.31: 845- 8s0. 129

Greenspan, M.D., K.A. Shackel, and M.A. Matthews, 1994, Developmental changes in the diurnal water budget of the grape berry exposed to water deficits. Plant Cell and Environment, l7: 811-820.

Gunata, Y.2., C.L. Bayonove, R.L. Baumes, and R.E. Cordonnier, 1985a, I' Extraction and determination of free and glycosidically bound fractions of grape aroma components. J. Chromatogr. 331 : 83-90.

Gunata, Y.2., C.L. Bayonove, R.L. Baumes, and R.E. Cordonnier, 1985b, The aroma of grapes. Location and evolution of free and bound fractions of some grape aroma components cv.Muscat During first development and maturation. J. Sci. Food Agric. 36:857-862.

Gunafa,Y.Z;C.L. Bayonove; R.L. Baumes, and R.E. Cordonnier, 1986, Changes in free and bound fractions of aromatic components in vine leaves during development of Muscat grapes. Pytochemistry, 25: 943'946.

Gutiérez-Granda, M.J., and J.C. Monison, 1992, Solute distribution and malic enzyme activity in developing grape berries. Am. J. Enol. Vitic . 43:323'328.

Hahlbrock, K., and Scheele, D., 1989, Physiology and molecular biology of phenilpropanoid metabolism. Ann. Rev. Plant Physiol. Plant Mol. Biol. 40: 347- 364.

Hale, C.R. , 1977. Relation between potassium and the malate and tartarate content of grape berries. Vitis. I 6:9-19.

Hale, C.R., and R.J. 'ù/eaver, 1962. The effect of developmental stage on direction of photosynthate in Vitis vinifera. Hilgardia,33: 89-131.

Hall, S.M., and D.A. Baker, 1972. The chemical composition of Ricinus phloem exudate. Planta, 106: 131-140.

Hamilton, D.4.. and P.J. Davis, 1988. Mechanism of export of organic material from the developing fruits of pea. Plant Physiol. 86: 956-958.

Hanson, S., and J. Cohen. 1985, A techniquefor collection of exudate from pea seedlings Plant Physiol. 43 : 224-228.

Hardy, P.J. 1967, Sucrose breakdown and synthesis in the ripening grape berry.Aust. J. Biol. Sci. 20 465-470 130

Hardy, p-J. 1968, Metabolism of sugars and organic acids in immature grape berries. Plant Physiol. 43:224-228.

Hardy, W.J., and T. P. O'Brien, 1988, Considerations of the biological significance of some volatile constituents of grape (Vitis spp.). Aust. J. Bot., 36: I07-ll7.

Harris, J.N., P.E. Kriedemann, and J.V. Possingham, 1968, Anatomical aspects of grape berry development, Vitis, 7 : 106-119.

Haslam, E.IgT5,Natural proanthocyanidins In: Flavonoids. Harborne, J.8., T.H' Marby, and H. Marby (Eds.) pp. 505-559.

Hawker, J.S., M.S. Buttrose, A. Soeffky and J.V. Possingham,1972, A simple method for demonstrating macroscopically the location of polyphenolic compounds in grape

berries. Vitis, 1 1: 189-192.

Hayashi, H., and M. Chino, 1986, Collection of pure phloem sap from wheat and its chemical composition. Plant Cell Physiology, 27 : 1387-1393.

Ho, L.C., 1988, Metabolism and compartmentation of imported sugars in sink organs in relation to sink strength. Ann. Rev. Plant Physiol. Plant Mol. Biol. 39 355'378

Ho, L.C., lgg2, The possible effects of sink demand for assimilate on photosynthesis. In: Murata,N. (ed.) Research in photosynthesis, vol. IV, Dordrecgt: Kluwer, pp. 729-736.

Hofläker, W., 1978, lnvestigation on the photosynthesis of vines influence of defoliation, topping. girdling and removal of grapes. Vitis, l7: 10-22.

Hoking, P.J., 1980, The composition of phloem exudate and xylem sap from tree tobacco (Nicotiana glauca Grah). Ann. Bot., 45: 633-643.

Hosel, W., 1981, Glycosilation and glycosides. In "The Biochemistry of Plants" Vol. 7, Chapter 23, pp.725-753, Academic Press, Inc.

Hrazdina, G., and A. Moskowitz. 1980, Subcellular status of anthocyanins in g¡ape skins. Centenary Symp.on Grape and Wine, Univ. Calif. Davis, July 1980.

Hrazdina, G., G.F. Parsons, and L.R. Mattick, 1984, Physilogical and biological events during development and maturation of grape berries. Am. J. Enol. Vitic .35:220- 227. 131

Huber, S.C., J.L.A. Huber and R.W. McMichael Jr, 1992, The regulation of sucrose synthesi in leaves. In: Carbon partitioning within and between organisms. C.J' pollock, J.F. Farra¡ and A.J. Gordon (Eds.) BIOS Scientific Publishers Limited, UK. pp. l-26.

Humpf, H-V., and P. Schreier, lggl. Bound Aroma compounds from the Fruit and the Leaves of Blackberry (Rubus laciniata L.), J.Agric.Food Chem. 39: 1830-1832.

Hunter, J.J., and J.H. Visser, 1988, Distribution od l4C- photosynthetate in the shoot of Vitis vinifera L. cv. Cabernet Sauvignon. I. the effect of leaf position and developmental stage of vine. S.4fr. J. Enol' Vitic' 9: 3-15.

Hunter, J.J., and J.H. Visser, 1989, The effect of partial defoliation, leaf position and developmental stage of the vine on leaf chlorophyll concentration in relation to the photosynthetic activity and light intensity in the canopy of Vitis vinifera L. cv. Cabernet Sauvignon. S. Afr' J. Enol. Vitic. l0:67-73.

Hunter, J.J., and J.H. Visser, 1990, The effect of partial defoliation on growth characte¡istic of Vitis viniferaL. cv. Cabernet Sauvignon. II. Reproductive growth' s. Afr. J. Enol. Vitic. ll:26-32.

Hunter, J.J., R. Skrivan, and H.P. Ruffner, 1994, Diurnal and seasonal physiological changes in leaves of Vitis vinifera L.: CO2 assimilation rate, sugar level and sucrolytic enzyme activity. Vitis, 33: 189-195.

Hunter, J.J., H.P. Ruffner. C.G. Volschenli, and D.J. Le Roux, 1995, Partial defoliation of Viti.s t,inifÞra L. cr,. Cabernet Sauvignon 99 Richter: Effect on root growth, canopy efficiencl'. grape composition, and wine quality. Am.J. Enol. Vitic. 46:

306-3 I 4.

Iland. P.G. and B.G.Coombe, 1988. Malate, tartarate, potassium, and sodium in flesh and skin of Shiraz grapes during ripening: concentration and compartmentation. Am. J. Enol. Vitic. 39: 7 l -7 6.

Ishiwatari, Y., C. Honda. l. Kawashima, S. Nakamura. H. Hirano, S. Mori, T. Fujiwara, H. Hayashi. and M. Chino, 1995. Thioredoxin h is one of the major proteins in rice phloem sap. Pianta. 195: 456-463.

Jeschke, W.D. and S. Pate. 1995, Mineral nutrition and transport in xylem and phloem of Banksia prionotes (Proteaceae), a tree with dimorphic root morphology. J. Exp. Bot. 46: 895-905. 132

Johnson, J.O., R.J. Weaver, and D.F. Paige, 1982, Differences in the metabolization of assimilates of Vitis vinifera L. grapevines as influenced by an increased source strength. Am. J. Enol. Vitic.33:207'213.

Kanasawud, P., and J.C. Crouzet, 1990, Mechanism of formation of volatile compounds by thermal degradation of carotenoids in aquous medium. l. b- carotenedegradation. J. Agric. Food Chem . 38: 237 -243.

Kawabe, S., T. Fucumorita, and M. Chino, 1980, Collection of rice phloem sap from stylets of homopterous insects by YAG laser. Plant Cell Physiology, 2I: l3l9- 1327.

King, R.W., and J.A.D . Zeevaart, 1974, Enhancement of phloem exudation from cut petioles by chelating agents. Plant Physiology. 53: 96-103.

Kliewer, W.M. 1965, The sugars of grapevines. II. Identification and seasonal changes in the concentration of several trace sugars in Vitis vinifera. Am. J. Enol. Vitic. l6:

I 78-1 88.

Kliewer, W.M., 1967a, The glucose-fructose ratio of Vitis vinifera grapes. Am. J. Enol Vitic. l8:33-41

Kliewer, W.M., 1967b, Annual cyclic changes in the concentration of free amino acids in grapevines. Am. J. Enol. Vitic. 18: 126-137.

Kliewer, W.M., 1969. Free amino acids and other nitrogenous substances of table grape varieties. L. Food Sci. 34: 274-278.

Kliewer, W.M.. 1970a. Effect of time and severity of defoliation on growth and

composition of Thompson Seedless grapes. Am. J. Enol. Vitic .21: 693-697 .

Kliewer, W.M., 1970b. Free amino acids and other nitrogenous fractions in wine grapes. J Food Sci. 35:17-21.

Kliewer, W.M.. 1970c. Effect of da¡,temperature and light intensity on colorationof Vilis vinifera L. grapes. Journal of the. American Society for Horticultural Science 95, 693-7.

Kliewer, W.M.. and A.J. Antcliff. 1970, Influence of defoliation, leaf darkening, and cluster shading on the growth and composition of sultana grapes. Am. Enol. Vitic. 2l:26-36. 133

Kliewer, W.M.,and A.L.Y.R. Nassar, 1966, Changes in concentration of organic acids, sugats, and amino acids in grape leaves. Amer. J. Enol. Vitic. 17:48-57.

Koblet, W., Ig77, Translocation of photosynthate in grapevines. Proc. Intern. Symp' euality Vintage, 14 -21 Feb. 1977 ,45-51. Cape Town, Republic of South Africa.

K¡iedemann, P.E., 1977, Vine leaf photosynthesis. Proc. Intern. Symp. Quality Vintage, 14-21 Feb. 1977,67-87 . Cape Town, Republic of South Africa'

Kursanov, 4.L., 1984, Assimilate transport in plants. Elsevier Science Publishers B.V., Amesterdam. pp. 660.

Kurt, B. and G. Torssell, 1983, Natural product chemistry. John Wilry and sons limited, N.Y. pp.40l.

Lang, A. and H. During, 1991, Partitioning control by water potential gradient: Evidence for compartmentation breakdown in grape berries. J. Exp. BoL 42 (242): lllT- rt22.

Lang,4., and M.R. Thorpe, 1989, Xylem , phloem and transpiration flows in a grape: Application of a technique for measuring the volume of attached fruits to high resolution using Archimedes principle. J.Exp. Bot. Vol. 40 (219):1069-1078.

Lang, 4., M.R. Thorpe, and W.R.N. Edwards, 1986, Plant water potential and translocation. In: J.L. Bonnemain, S. Delrot and J. Dainty, W. Lucas, (eds.) "Research advances in phloem transport and assimilate compartmentation. Ouest Editions, Nants. France, pp. 193-194.

Leckstein, P.M., and M. Llewellyn, 1975, Quantitative analysis of seasonal variation in the amino acids in phloem sap of Salix alba L., Planta, 124: 89-91.

Lukner, M., 1972, Secondary metabolism in plants and animals. Translated by T.N Vasudevan, Chapman and Hall Ltd.London, pp. 404.

Lukner, M., 1990, Secondary metabolism in micro-organisms, plants and animals.3rd edition.Springer Werlag, Berlin.

Lunn. J.E., and M.D. Hatch, 1995, Primary partitioning and storage of photosynthate in sucrose and starch in leaves of Ca plants. Planta, 197:385-391.

Madore, M.4., 1991, Diurnal oligosaccharide export patterns in Cucurbit sou¡ce leaves. In: Recent advances in phloem transport and assimilate compartmentation. Bonnemain, J.L., W.J. Lucas and J. Dainty, (Eds.), Ouset Editions, pp 23-26. 134

Mailla¡d, P., J.P. Procher, 1991, Effect of light on the regulation of carbon partitioning in mature leaves of Zea mays L.In: Recent advances in phloem transport and assimilate compartmentation. Bonnemain, J.L., W.J. Lucas and J. Dainty, (Eds.), Ouset Editions, pp. 27'32.

Mann, J., 1987, Secondary metabolism. Atkins, P.W., J.S.E' Holker and A.K. Holliday, (Eds.), Oxford Chemistry Series, Oxford University Press, Oxford England.

Marinos, V.4., 1992, Application of Fast Atom Bombardment Mass Spectrometry to the examination of glycoconjugates in grape juice. Ph. D. Thesis, The University of Adelaide, S.A.

Maris, J. 1983, Terpenes in the aroma of grapes and wines. S. 4fr. J. Enol. 4:49-60.

Marschner, H., 1983, General introduction to the mineral nutrition of plants. In: "Encyclopedia of plant physiology" Lauchli, 4., and R.L. Bieleski (eds.), New series, Volume l2,lnorganic plant nutrition. Berlin: Springer-verlag, pp. 5-60.

Matsui, H., E. Yuda, and S. Nakagawa,1979, Physiological studies on the ripening of Delaware grapes. I. Effect of the number of leaves and changes in polysaccharides or organic acids on sugar accumulation in the berries. J. Japan. Soc. Hort. Sci. 48: 9-1 8.

Mattivi, F., Scienza, 4., Failla, O., Villa, P., Anzani, R., Tedsco, G', Gianazza, E. and Righetti. P. (1989) Vitis vinifera - a chemotaxanomic approach: anthocyanins in the skin. Proceedings of the 5th International Symposium on Grape Breeding, Vitis 1990 special issue, pp I l9-133.

Milburn, J.A. 1970. Phloem exudation from castor bean: induction by message. Planta, 95 272-276.

Morrison, J.C., and M. Iodi. 1990, The influence of water berry on the development and composition of thompson seedless grapes. Am. J. Enol. Vitic.4l:301-305.

Moss, D.N., and H.P. Rasmussen, 1969, Cellular localization od CO2fixation and translocation of metabolites. Plant Physi ol. 44 I 063- 1068.

Motomora, Y., 1993, lac-assimilate partitioning in grapevine shoots: effect of shoot pinching, girdling of shoot, and leaf-halving on assimilate partitioning from leaves into clusters. Am.J. Enol.Vitic. 44:1-7. 135

Mullins, M. G., and Rajasekaran K. (1981). Fruiting cuttings: Revised method for producing test plants of grapevine cultiva¡s. Am. J. Enol. Vitic.32: 35-40'

Murphey, J.M., S.E. Spayd, and J.R. Powers, 1989, Effect of grape maturation on soluble protein characteristics of Gewürztraminer and White Riesling juice and wine. Am. J. Enol. Vitic. 40:199-207.

Nakamura, S., H. Hayashi, S. Mori, and M. Chino, 1995, Detection and characterization of protein kinases in rice phloem sap. Plant Cell Physiol.36; 19-27.

Nassar, 4.L., and W.M. Kliewer, l966,Free amino acids in various parts of Vits vinifera at different stages of development. Proc. Am. Soc. Hortic Sci., 89: 281-194-

Nii, N., and B.G. Coombe, 1983, Structure and development of the berry and pedicel of the grape Vitis viniferaL' ActaHortic' 139:129-140'

Nobel, p.S.,1991, Physicochemical and environmental plant physiology. Academic press' Inc. Harcourt Brace Jovanovich, publishers. San Diego Calfomia.

Noble, A.C., 1994, Wine flavour. In: understanding natural flavours. Piggott, J.R. and paterson, A. (Eds.), Blackie Accademic and professional, an imprit of Chapman & Hall. Glasgow. PP.228-242.

Noble, A.C.; C.R. Strauss; P.J. Williams and B. Wilson, 1988, Contribution of terpene glycosides to bitterness in Muscat wines. Am. J. Enol. Vitic.39 129-131.

Ohloff. G.. 1978. Naturally occurring aroma compounds. In: 'Progress in the chemistry of organic natural products'. Vol.35, W. Herz, H. Grizebach and G.W. Kirby (eds.), pp. 43 1-527 . Springer-Verlag. Wien.

Otsuka, H., K. Yamasaki. and T. Yamaguchi. 1989, Alangifolioside, a diphenylmethylene derivative. and other phenolics from the leaves of Alangium plantifolium var. Trilobum. Phytochem. 28( I I ): 3197-3200.

Padgett, M., and J.C. Morrison, 1990, Changes in grape berry exudates during fruit development and their effect on mycelial growth of Botrytis cinera.J. Am. Soc' Hort. Sci. 115: 269-273.

Pandey, R.M., and H.L. Farmahan, 1977, Changes in the rate of photosynthesis and respiration in leaves and berries of Vitis vinifera grapevines at various stages of

berry dcvelopment. Vitis, l6: 106- I 1 1. 136 park, S.K., J.C. Monison, D.O. Adams, and A.C. Noble, 1991, Distribution of Free and Glycosidically Bound Monioterpenes in the Skin and Mesocarp of Muscat of Alexandria Grapes During Development. J. Agric. Food Chem. 39: 514-518. park, S.K., and A.C. Noble, 1993, Monoterpenes and monoterpen glycosides in wine aromas; analysis, characterization, and technological advances. In: "Beer and Wine Production". Gump, 8.H., (ed), ACS Symposium Series 536'

Pate, J.S., P.J. Sharkey, and O.A.M. Lewis, 1974, Phloem bleeding from legume fruits; a technique for study of fruit nutrition. Planta, 120:229-243'

Pate, J.S., 1975, Exchange of solutes between phloem and xylem and circulation in the whole plant. In: Zimmernann and J. Bilburn (Eds.) Transport in plants. Springer, Berlin. pp.45I-473.

Pate, J.S. 1980, Transport and partitioning of nitrogenous solutes. Ann. Rev. Plant Physiology, 3 I : 3 13-340.

Pate, J.S., 1983, Patterns of nitrogen metabolism in higher plants and their ecological significanc e. P.225-254, In: "Nitrogen as Ecological factor", J.A. Lee, S. Mc Neill, and I.H. Rorison (Eds.) , Blackwell, Boston.

Peynaud, E., and P. Ribéreau-Gayon, 1971, The grape. In: "The biochemistry of fruits and their products". Vol. 2, Hulme, A.C.(Ed.), Academic press.

Pirie,4., and M.G. Mullins 1976, Changes in anthocyanin and phenolics content of grapevine leaf and fruit tissues treated with sucrose, nitrate, and abscisic acid. Plant Physiol. 58: 468-472.

Pirie. 4., and M.G. Mullins 1977 In¡enelationships of sugars anthocyanins, total phenolsand dry weight in the skin of grape berries during ripening. Amr. J. Enol. Vitic. 28;204-209. Pirie, 4., and M.G. Mullins, 1980, Concent¡ation of phenolics in the skin of grape berries during final development and ripening. Am. J. Enol. Vitic., 3l:34-36.

Possner, D., and W.M. Kliewer. 1985, The localisation of acids, sugars, potassium and calcium in developing grape berries. Vitis, 24: 229-240.

Possner, D., H.P. Ruffner, and D.M. Rast, 1983, Regulation of malic acid metabolism in berries of Vitis vinifera. Acta Hortic. 139: ll7-122. 137 pueyo, E., M. Dizy, and M.C. Polo, 1993, Varietal differentiation of must and wine by means of protein fraction. Am. J. Enol' Vitic. 44:255'260. euinlan, L.D., and R.J. Weaver, !970, Modification of pattern of the photosynthate movement within and between shoots of Vitis vinifera I. Plant Physiol. 46:527- 530.

Ramaswami, S.K., P. Briscese, R.J. Garguilllo, and T. von Gelder, 1986, Sesquiterpene hydrocarbons: from mass confusion to orderly line-up. In: Flavours and Fragrants: A world perspective. Proceedings of the lOth International; Congress of Essential Oils, Fragrances and Flavours, Lowtence, B.M., B.D. Mookherjee, and B.J. V/illis, (Eds.), 16-20 November 1986, Washington, DC, USA, 951-980.

Rapp , 4., H. Mandery, and H. Ullemeyer, 1983, 3,7-Dimethyl-1,7-octan-diol-eine neue Terpen-Komponente des Trauben-und wine-aromas. vitis, 22: 225 -230.

Rapp, 4., 1988a, Wine aroma substances from gas chromatographic analysis. In: Modern methods of plant analysis. Vol. 6, Wine Analysis; Linskens, H.F. and J.F. Jackson, (Eds.) Springer Verlag, Berlin, pp.29-66'

Rupp,4., 1988b, Studies on terpene compounds in wines. In; "Frontiers of flavour", proceedings of the Fifth Intemational Flavour conference. Charalambous, G. Ed., l-3 July, 1987, porto Karras, Calkidiki, Greece; Elsevier Science, Amsterdam, pp. 799-813.

Razungles, 4.. C.L. Bayonov, R.E. Cordonier, and J.C. Sapis, 1988, Grape carotenoids: changes during maturation period and localization in mature berries. Am. J. Enol. Vitic. 39:44-48. Ribéreau-Gayon. P.. 1982. The anthocyanins of grapes and wines. In: Anthocyanins as Food Colors. Markakis.P.. Ed.. Academic Press, New York,209.

Ribéreau-Gayon. P., J.N. Boidron, and A. Terrier, I 975, Aroma of Muscat grape varieties J. Agric. Food Chem.23(6): 1042-1047.

Richardson, P.T.. D.A. Baker. and L.C. Ho, 1982, The chemical compositionof cucurbit

vascular exudates. J.Exp. Bot.. 33: 1239-1247 .

Ridge, 1., 1991. Plant physiology. Biology: Form and Function, Hodder and Stoughton, The Open University. 138

Riens,8., G. Lohaus, D. Heineke, and H.W. Heldt, 1991, Amino acid and sucrose content determined in the cytosolic, chloroplastic, and vacuolar compartments and in the phloem sap of spinach leaves. Plant Physiol. 97: 227'233 '

Roggero, J.P., Coen, S. and Ragonnet, 8., 1986, High perforrnance liquid chromatography suvey on changes in pigment content in ripening grapes of Syrah. An approach to anthocyanin metabolism. Am. J. Enol. Vitic. 37:77'83.

Rubelakis-Angelakis, K.4., 1991, Amino acid and protein metabolism in Vitis spp', 'Wine' Proceedings of the International Symposium on Nitrogen in Grapes and J.M. Rantz (Ed.) pp. 52-61, American Society for Enology and Viticulture.

Rubelakis-Angelakis, K.4., and'W. M. Kliewer,lgg2,Nitrogen Metabolism in Grapevine' Horticultural Reivew, Vol. 14: 407-452.

Ruffner, H.P., and J.S. Hawker, 1977, Control of glycolysis in ripening berries of Vitis vinifera. Phytochemistry, 73 : 45-55.

Ruffne¡, H.P., and W.M. Kliewer, 1975 Phosphoenolpyruvate caroxykinase EC-4.1.1.32 activity in grape berries. Plant Physiol.56:67-71'

Ruffner, H.P., W. Koblet, and D. Rast, 1975, Gluconeogenesis in the ripening fruit of Vitis

vinifera. Vitis, I 3 : 319-328. Salles, C., H. Essaied, P. Chalier, J.C. Jallageas, and J. Crouzet, 1988, Evidence and characterisation of glycosidically bound volatile components in fruits. In: Bioflavour 87. analysis. biochemistry, biotechnology. Schreier (Ed.), Walter de Gruyter & Co.. Berlin, pp 145-160.

Schreier, P., F. Drawert aand A. Junker, 1976, Identification of volatile constituents from grapes. J. Agric. Food Chem. ?4:331-336.

Schuls, G. and E. Stahl-Biskup, l99l . Essential Oils and Glycosidic Bound Volatiles from Leaves, Stems. Flowers and Roots of Hyssopus officinalis L. (Lamiaceae), Flavour Fragr. J. 6: 69-73.

Sefton, M.4., G.K. Skouroumounis. R.A. Massey-V/estropp, and P.J. Williams, 1989, Norisoprenoids in Vitis vinifera white wine grapes and the identification of a precursor of damascenone in these fruits. Aust. J. Chem. 42:2071-2084.

Sefton, M.4., Francis, I.L., and Williams P.J., 1993, The volatile composition of Chardonnay juices: a study by flavour precusor analysis. Vitis, 44:359-370. 139

Sefton, M.4., I.L. Francis, and P.J. Williams, 1994, Free and bound volatile secondary metabolites of Vitis vinifera grape cv. Sauvignon Blanc. Journal of Food Science, 59: 142-147.

Sepúlveda, G. and W.M. Kliewer, 1986, Effect of high temperature on grapevines (Vitis vinifera L.).il. Sistibution of soluble sugars. Am. J. Enol. Vitic.37:20-25'

Shoseyov, O., B.A. Bravdo, D. Seigel, A. Goldman, S. Cohen, L. Shoseyov, and R. Ikan, 1990, Imobilised endo-ß-glucosidase enriches flavour of wine and passion fruit juice. J. Agric. Food Chem. 38: 1387-1390.

Shure, K.B., T.E. Acree, 1994, Changes in the odor-active compounds in Vitis labruscana cv. Concord during growth and development. J. Agric. Food Chem. 42:350-353.

Sidahmed, O.4., and W.M. Kliewer, 1980, Effect of defoliation, gibberellic acid and 4- chlorophenoxyacetic acid on growth and composition of Thompson Seedless grape berries. Am. J. Enol. Vitic.31: 149-153.

Simpson, R.J., and M.J. Dalling,l98l, Nitrogen redistribution during grain growth in wheat (Triticum aestivum L.) ilI. Enzymology and transport of amino acids from senescing flag leaves. Planta, l5l:447-456.

Skouroumounis, G.K., and P. Winterhalter, 1994, Glycosidically bound norisoprenoids from Vitis vinifera cv. Reisling leaves. J. Agric. Food Chem.,42: 1068-1072.

Somers, T.C. l975,ln search of quality for red wines. Food technology in Australia,2T; 49-56.

Somers, T.C., and G. Ziemelis,1973, Direct determination of wine proteins. Am. J. Enol Vitic. 24:47-50.

Somers, T.C., 1976, Pigment development during ripening of the grape. Vitis 14, 269-277.

Stahl-Biskup E., F. Intret, J. Holthuijzen, M. Stengele, and G. Schulz, 1993, Glycosidically bound volatiles- a review 1986-1991. Flavour and Fragrance Journal. 8: 6l -80.

Stitt, M., and W.P. Quick, 1989, Photosynthetic carbon partitioning: its regulation and possibilities for manipulation. Physiol. Plant. 77: 633-641. 140

Strauss, C.R., B. Wilson, P.R. Goodley, P.J. Williams, 1986, Role of monoterpenes in grape and wine flavour. In: "Biogeneration of aromas", symposium sponsored by the division of Agriculnre and Food Chemistry at 190th meeting of the American Chemical Society; Parliment T.H. and R. Croteau, (Eds.) 8-13 Septamber 1985, Chicago, Illinois, ACS symposium series 317, American Chemical Society, Washington, DC, pp. 222'242.

Strauss, C.R., R.P. Goodley, B. Wilson, and P.J. Williams, 1987, Application of droplet counter current chromatography to the analysis of conjugated forms of terpenoids, phenols, and other constituents of grape juice. J. Agric. Food Chem. 35:519-524.

Strauss, C.R.; B.Wilson and P.J.Williams, 1988, Novel monoterpene diols and diol

glycosides in Vitis vinifera grapes. J. Agric. Food Chem. 36: 569-573 '

Sugisawa, H., K. Miwa, T. Matsuo, and h. Tamura, 1988, Volatile compounds produced from the cultured cells of Thyme (Thymus vulgaris L.). In: Bioflavour '87, 'Walter analysis, biochemistry, biotechnology. Schreier (Ed.), de Gruyter & Co., Berlin. pp 327-340.

Swanson, C.4., and E.D.H. El-Shishini, 1958, Translocation of sugars in the Concord grapes. Plant Physiology, 33: 33-37.

Taiz. L. and E. Zeiger. 1991, Plant Physiology, the Benjamin/cummings publishing company. Inc. pp 565.

Tazaki.H., N. Ohta. R. Hori. H. Okuyama, M. Okumura, T. Fujimori, and, K. Nabeta, 1993, The glycosides of oxygenated lower terpenes in plant leaves. In: Food flavours. ingredients and composition. Charalambos, G. (ed.), Elsevier Science Publishers B.V.

Tesniére C.M., C. Romieu. and J.-P. Robin, 1994, Grape invivo protein synthesis changes in response to anaerobiosis. Am. J. Enol. Vitic. 45:267-272.

Turgeon, R., J.A. Webb, 1973. Leaf development and phloem transport in Cucurbita pepo: transition from import to export. Planta I l3: 179-191.

Turgeon, R., J.A. Webb. R.F. Every, 1975, Ultrastructure of minor veins in Cucurbita pepo leaves. Protoplasma 83 : 217 -232.

Urquhart, A.A. and K.S/. Loy. 1981, Use of phloem exudate technique in the study of amino acid transport in pea plants. Plant Physiology, 68:750-754. 141

Van Bel, 4.J., 1993, Strategies of phloem loading. Ann. Rev. Plant Physiol. 44:253-281'

Versini, G., A. DalaSeea, M. Dell'Eva, A, Scienza, A. Rapp, 1988, Evedence of some glycosidically bound monoterpenes and norisoprenoids in grapes. In: Bioflavour '87, analysis, biochemistry, biotechnology. Schreier (Ed.), Walter de Gruyter & Co., Berlin, pp 161-170.

Warlaw, I.F., 1990, The control of carbon partitioning in plants. New Phytol. 116: 341- 381.

Watanabe, Y. and Shiriashi, S.I., 1991, Relationship between skin color and anthocyanin of grapes. Science Bulletin of The Faculty of Agriculture, Kyushu University,46:3' 38.

Watanabe, Y. K., Sihara, I. and Shiriashi, S.I., 1992, Variation of anthocyanin composition in fruit skin of F-l hybrid grapes. Science Bulletin of the Faculty of Agriculture, Kushu University, 46: 223 -230.

Weibull, J.; F. Ronquist, and S. Brishammar, 1990, Free amino acid composition of leaf exudates and phloem sap. Plant Physiol., 92:222-226.

Weiner, H., S. Blaechschmidt-Schneider, H. Mohme, W. Schrich, and H.W. Heldt, 1991, Phloem transport of amino acids. Comparison of amino acid contents of maize leaves and of the sieve tube exudate. Plant Physiol. Biochem .29: 19-23.

Wenzel. H., H. Ditrich, and M. Heimfarth, 1987, Die Zusammensetzung der Anthocyane in den Beeren verschiedener Rebsorten. Vitis, 26:65-78.

Wermelinger, 8.. 1991. Nitrogen dynamics in grapevine: physiology and modelling Proceedings of the International Symposium on Nitrogen in Grapes and Wine J.M. Rantz (Ed.) pp. 23-31. American Society for Enology and Viticulture.

West, C.4., 1990. Terpene biosynthesis and metabolism. In Plant Physiology, Biochemistry and Molecular Biology, eds. D.T. Dennis and D.H. Turpin. Longman Scientific Technical. John Wiley & Sons, New York. pp,353-369.

Wijayanti, L., M. Kobayashi, S. Fujioka, K. Yoshizawa, and A. Sakurai, 1995, Identification and quantification of Abscisci acid, Indole-3-acetic acid and Gibberellins in phloem exudates of Pharbitis nil. Biosci. Biotech. Biochem. 59: 1533-l 535. 142

Wilden¡adt, H.L., E.N. Christensen, B. Stackler, A. Caputi, Jr. K. Slinkard, and K. Scutt, 1975, Volatile constituents of grape leaves. I. Vitis vinifera variety 'Chenin Blanc'. Am. J. Enol. Vitic.26:148-153-

Williams, L.E., and P.J. Biscay,IggI, Partitioning of dry weight, nitrogen, and potassium in Chabernet Sauvignon grapevines from anthesis until harvest. Am' J. Vitic. Enol. 42: ll3-117.

Williams, P.J. 1993, Hydrolytic flavour release in fruit and wines trough hydrolysis of non-volatile precursors. In: Flavour Science, sensible principal and Techniques; Acree,T.E., Teranishi, R., Eds.; American Chemical Society: V/ashington, DC.

Williams, P.J., C.R. Strauss, and B. Wilson, 1980, New linalool derivatives in Muscat of Alexandria grapes and wines, Phytochem. 19: I I37-1I39.

Williams, P.J.; Strauss ,C.R.; Wilson, B. and Massy-V/estropp, R.A., 1982a, Novel monoterpen disaccharide glycosides of Vitis vinifera grapes and wines. Phytochemistry, 2 1, 2013 -2020.

Williams, P.J.; Strauss, C. R.; Wilson, B. and Massy-Westropp, R.4., 1982b, Studies on the hydrolysis of Vitis vinifera monoterpen procursor compounds and model mono terpen p-D- glucosides rationalising the mono terpen composition of grapes. J. Agric. Food Chem.30: 1219-1223.

Williams, P. J.; Strauss, C.R. and Wilson,8., 1982c, Use of Cl8 reversed-phase liquid chromatography for the isolation of monoterpen glycosides and nor-isoprenoid precr¡rsors from grape juice and wines. J. Ch¡omatog.235 471-480.

Williams. P.J., C.R. Strauss. B. Wilson, and R.A. Massy-Westropp, 1983, Glycosides of 2-phenylethanol and benzyl alcohol in Vitis vinifera grapes. Phytochem. 22: 2039-2040.

Williams, P.J., Strauss, C.R., V/ilson, 8., and Dimitriadis, E.,1985, Origin of some volatile monoterpenes and nor-isoprenoids in grapes and wines - biosynthetic and biogenetic considerations. Berger, R.G.; Nitz, S.; Schreier, P., eds. Topics in flavour research: proceedings of the international confernce; l-2 Aplil 1985; Freising-Wrihenstephan, FRG. marzling-Hangenham, FRG: H. Eichron;335-352.

Williams, P.J., C.R. Strauss, A.P. Aryan, and B. Wilson, 1987, Grape flavour - a review of some pre- and postharvest influences. In: Proceedings of the sixth Australian wine industry technical conference, Lee, T.H., (Ed.),14-16 July 1986, Adelaide, S.A. Australian Industry Publishers, I l1-ll7. 143

Williams, P.J., M.A. Sefton, B. Wilson, 1989, Nonvolatile conjugates of secondary metabolites as precursoes of varietal grape flavour components. In Flacour Chemistry: Trends and developments; Teranishi, R.; R.G. Buttery; F. Shahidi, eds.; ACS Symposium Series No. 388; American Chemical Society: Washington DC, pp 35-48.

V/illiams, P.J., M.A. Sefton, and I.L. Francis, 1991, Glycosidic precursors of varietal grape and wine flavour. In: Thermal and Enzymatic Conversions of Precursors to flavour compounds. ACS Symposium series.

V/illiams, P.J., M.A. Sefton, and V.A. Marinos, 1993, Hydrolytic flavour release from non-volatile precursors in fruits, wines and some other plant derived foods. In: Recent developments in flavour and fragrance chemistry. Proceedings of the Third Haarman and Reimer Symposium; Hopp, R' Mori, K. Eds.; VCH Verlagsgesellschaft : Weinheim, pp. 283 -290.

Williams, P.J.; W. Cynkar; I.L. Francis; J.D. Gray; P.G. Iland, and B.G. Coombe, 1995, Quantification of glycosides in grapes, juice, and wines through a determination of glycosyl glucose. J. Agric. Food Chem. 43 l2l-I28.

'Williams, Wilson, 8., C.R. Strauss, and P.J. 1984, Changes in free and glycosidically bound monoterpenes in developing Muscat grapes. J. Agric. Food Chem.32:919- 924.

'Wilson, 8., C.R. Strauss, and P.J. Williams, 1986., The distribution of free and glycosidically-bound monoterpenes among skin, juice, and pulp fractions of some white grapr varieties. Am. J. Enol. Vitic. 37:107-llI.

Winkler. 4.J.. J.A. Cook. W.M. Kliewer, and L.A. Lider. 1974. General Viticulture Univ.of California Press. Berkeley, Clifornia.

Winterhalter, P., and P. Schreier. 1988a, Free and bound C13 norisoprenoids in Quince (Cydonia oblangu Mill. ) fruit. J. Agric. Food Chem. 36: 125l-1256.

Winterhalter, P., and P. Schreier, 1988b, Studies on C13-norisoprenoid precursors. In: Bioflavour'87, analysis, biochemistry, biotechnology. Sch¡eier (Ed.), Walter de Gruyter & Co.. Berlin, pp 255-273.

Winterhalter, P. and P. Schreier,1994, C13-Norisoprenoid glycosides in plant tissues: an overview on their occurrence, composition and role as flavou¡ precursors. Flavour

and Fragrance Journal, 9 : 281 -287 . 144

V/interhalter, P., M.A. Sefton, and P.J. Williams, 1990a, Two-dimensional GC-DCCC analysis of the glycoconjugates of monoterpenes, norisoprenoids, and shikimate derived metabolites from Riesling wine. J. Agric. Food Chem. 38: 1041-1048.

Winterhalter P., M. A. Sefton, and P.J. rù/illiams, 1990b, Volatile C13-norisoprenoid compounds in Riesling wine are generated from multiple precursors. Am. J. Enol. Vitic. 4I:277-283.

Woding, F.B.P., 1978, Phloem. Carolina Biology Readers, J.J. Head (Ed.), Carolina Biological Supply Company, Oxford university Press, 2nd ed.pp. 16. Wulf, L.V/. and C.V/. Nagel, 1978, High pressure liquid chromatography separation of anthocyanins oî Vitis vinifera. Am. J. Enol. Vtic. 29: 42-49.

Ziegler, H., 1975, Nature of transported substances. In: Encyclopedia of plant physiology, New Series. Vol. 1, Transport in plants, pp. 171-195, Zimmernann M.H. and J.A. Milburn, eds. Springer Verlag, Berlin.

Ziegler, H., 1956, Utersuchungen iiber die leitung und sekretion der assimilate. Planta,47 447-500.

Zimmermann, M.H., lg5T,Translocation of organic substances in trees. I. The nature of the sugar in the sieve tube exudate of trees. Plant Physioloty, 32:288'291.

Zimmermann, M.H.. 1960, Transport in the phloem, Ann. Rev. Plant physiol. ll: 167- 190 145

Appendix

Table 1 The seasonal changes in phloem sap amino acids of Muscat Gordo, Sultana and Shiraz. Values given areYo of toøl amino acids before veraison (B), at veraison (V), and at ripening time (R).

Muscat Gordo Sultana Shiraz Amino acids (B) (V) (R) (B) ('v) (R) (B) CV) (R) ASP 3.9 8.3 8.1 2.9 2.6 6.9 8.3 1l 18 GLU 5.9 9.6 23.5 r0.2 4.7 t0.4 6.8 t2.2 18.6 ASN 0.8 0.8 0.3 0.3 0.5 0.7 0.3 0.7 0.7 SER t.7 2.9 4 3.6 2.8 2.4 4.5 6.2 3.9 GLN 68.2 56.2 42.r 70.8 68 5t.7 68 49.9 37.4

HIS 2 1 I 0.8 r.2 1.6 0.8 1.3 1.8 GLY 0.9 1.8 1.4 I 0.9 1.3 1.7 J ).¿ THR 0.9 1.3 1.9 1.3 1.6 t.2 1.5 2.4 1.9 ALA 2.5 3.2 9.6 4.2 2.9 4.6 t.7 3.2 3.4 ARG 8.4 9.2 0.4 1.5 4.9 10.1 r.6 4.1 4.1 TYR 0.3 0.3 0.3 0.2 0.3 0.5 0.6 0.8 0.4 CYS-CYS 0 0 0 0 0, 0 0 0 0 VAL 0,8 1.1 1.9 0.6 1.9 0.9 0.8 I t.2 MET 0.4 0.7 0.4 0.3 0.4 0.7 0.6 0.4 0.6 TRP 1.3 0.4 0.4 0.2 0.2 0.7 0.2 0.2 0.2

PHE 0.2 0.5 1 0.4 1.2 t.6 0.3 0.9 0.9 ILE 0.3 0.6 l.l 0.2 0.9 0.5 0.4 0.6 0.6 LEU 0.3 0.7 1.2 0.3 I 0.7 0.4 0.7 0.8 LYS 0.3 0.5 0.1 0.2 0.3 0.5 0.4 0.6 0.9 PRO 0.8 0.8 1.2 t.l 3.7 2.8 l.l I 1.7 0.383 0.146 0.284 0.481 0.392 0.170 0.158 0.084 0.071 Total * Brix 4.9 8.0 23.8 5 7.7 23.4 4.2 9 24.4 tr Total values are in pmoles amino acids/ 1.5 mL of buffer solution/ 4 h

The data are ÍÌom ( 1995). 146

Table 2Thedifferences in amino acid profiles of leaf exudates in light and dark collected at berry ripening time (% of total).

Amino acids Muscat Sultana Shiraz Muscat Sultana Shiraz Gordo Gordo

Dark (1995) Lisht (1994)

ASP 7 3.2 13.9 10.7 9.0 17.9 GLU 19.4 15.7 27.3 20 13.5 24.r ASN 0 0.5 0.3 0 0 0 SER 5.2 8.7 7.9 7.2 4.1 4.9 GLN 22.1 30.8 24.9 0.1 32.0 8.6 HIS 0 0 0 0 0 0 GLY 2.5 3.4 2.9 5.2 0.8 9.2 THR 2.4 3.2 J 0 2.3 1.0 ALA JJ.Z 22.3 14.4 7.6 5.7 6.4 ARG 0 0.6 0 0.3 0.4 3.1 TYR 0 0.7 0.3 0 0 0 CYS-CYS 0 0 0 0 0 0 VAL 1.7 0 0 0 2.r 0 MET 0 1.3 0.7 1.7 0.9 1.6 TRP 0 0 0 0 0 0 PHE 1.5 I 0.6 0 t.7 0 ILE 0.5 0.8 0.4 0 1.4 0 LEU r.3 1.3 0.7 0 2.1 0.0

LYS 1.1 1.1 0.7 0 0 0 PRO 2 5.4 2 47.2 23.9 23.r

* 0.023 0.054 0.030 Total 0.041 0.066 0.133 *Total values are in ¡rmoles amino acidsi 1.5 mL of buffer solution exuded fiom l5 leaves /16 h 147

Table 3 The seasonal changes in amino acid prohles (% of total) of Muscat Gordo peduncle exudates.

Sampling date

Amino acids 3l1l9s rvv95 l9lrl95 t2l2l9s t5,1613195 14/4/95 ASP 7 3.9 8.3 t4.l 17.2 8.1 GLU 10.5 5.9 9.6 t4 12.2 23.5 ASN 0.8 0.8 0.8 1.3 0.7 0.3 SER 4.t t.1 2.9 4.5 5.1 4 GLN 60.8 68.2 56.2 35.4 46.7 42.1 HIS 0.5 2 I 0 0.6 I GLY 2.4 0.9 1.8 4.3 2.6 t.4 THR t.4 0.9 1.3 2.1 t.9 1.9 ALA 4.8 2.5 3.2 7.7 1.5 9.6 ARG 1.6 8.4 9.2 7.7 6.8 0.4 TYR 0.7 0.3 0.3 0 0.2 0.3 CYS-CYS 0 0 0 0 0 0 VAL 1.1 0.8 1.1 2.1 I 1.9 MET 0.8 0.4 0.7 0.7 0 0.4 TRP 0.4 1.3 0.4 0 0.3 0.4 PHE 0.6 0.2 0.5 1.1 0.6 I ILE 0.6 0.3 0.6 t.2 0.4 1.1 LEU 0.7 0.3 0.7 1.5 0.6 r.2 LYS 0.7 0.3 0.5 1.2 0.6 0.1 PRO 0.5 0.8 0.8 I t.l t.2 * Total 0.10 0.383 0.146 0.05b 0.1 08 0.284 "Brix 4.1 4.9 8.0 14.5 2t.r 23.8

*Total values are in pmoles anrino acidsi 1.5 mL of buffer solution/ 4 h. 148

Table 4 The seasonal changes in amino acid profiles (% of total) of Sultana peduncle exudates.

Sampling date

Amino acids 29n2194 r2ly95 912195 tr/3t95 ASP 2.9 2.6 4.2 6.9 GLU 10.2 4.7 8.5 10.4 ASN 0.3 0.5 0.2 0.7 SER 3.6 2.8 J.J 2.4 GLN 70.8 68 62.r 51.7 HIS 0.8 t.2 1.6 1.6 GLY I 0.9 1.5 1.3 THR 1.3 1.6 t.4 t.2 ALA 4.2 2.9 4.7 4.6 ARG 1.5 4.9 8.2 10.1 TYR 0.2 0.3 0.6 0.5 CYS-CYS 0 0 0 0 VAL 0.6 1.9 0.7 0.9 MET 0.3 0.4 0 0.7 TRP 0.2 0.2 0.3 0.7 PHE 0.4 1.2 0.8 1.6 ILE 0.2 0.9 0.2 0.5 LEU 0.3 I 0.4 0.7 LYS 0.2 0.3 0.5 0.5

PRO 1.1 3.7 0.9 2.8 * Total 0.481 0.392 0.138 0.1 60 Brix ì 7.7 l8 23.4 *Total values are in ¡rmoles amino acids,' I.5 mL of buffer solution/ 4 h 149

Table 5 The seasonal changes in amino acid profiles (% of total) of Shiraz peduncle exudates.

Sampling date

Amino acids 221t2194 17/|95 tv3lgs ASP 8.3 ll t8 GLU 6.8 12.2 18.6 ASN 0.3 0.7 0.7 SER 4.5 6.2 3.9 GLN 68 49.9 37.4 HIS 0.8 1.3 1.8 GLY t.7 J 3.2 THR 1.5 2.4 t.9 ALA 1.7 3.2 3.4 ARG 1.6 4.t 4.t TYR 0.6 0.8 0.4 CYS-CYS 0 0 0 VAL 0.8 I 1.2 MET 0.6 0.4 0.6 TRP 0.2 0.2 0.2 PHE 0.3 0.9 0.9 ILE 0.4 0.6 0.6 LEU 0.4 0.7 0.8 LYS 0.4 0.6 0.9

PRO 1.1 1 1.7 * Total 0.1 58 0.084 0.071

"Brix 4.2 9 24.4 *Total values are in ¡rmoles anrino acidsi 1.5 mL of buffer solution/ 4 h 150

a Fig. 1 The inflorescence grown from dormant bud of Muscat Gordo bud-grafted onto Shiraz cane.

Gholami, M., Hayasaka, Y., Coombe, B.G., Jackson, J.F., Robinson, S.P. & Williams, P.J. (1995). Biosynthesis of flavour compounds in Muscat Gordo Blanco grape berries. Australian Journal of Grape and Wine Research, 1(1): 19–24.

NOTE:

This publication is included in the print copy

of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at:

http://dx.doi.org/10.1111/j.1755-0238.1995.tb00073.x

Gholami, M. & Coombe, B. G. (1995). Occurrence of anthocyanin pigments in berries of the white cultivar Muscat Gordo Blanco (Vitis vinifera L.). Australian Journal of Grape and Wine Research, 1(2): 67–70.

NOTE:

This publication is included in the print copy

of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at:

http://dx.doi.org/10.1111/j.1755-0238.1995.tb00080.x