Biosynthesis of Flavonoids in
Grapevines (Vitis vínifera L.)
Thesis submitted for Doctor of Philosophy
Mark Downey
2004
University of Adelaide
Department of Horticulture, Viticulture and Oenology
TTE ISS+EHSTY otÆËt åffitE ESIBO Art Plant Industry Abstract
Flavonoids in grapes contribute to wine quality and to the health benefits derived from grapes and wine. This project sought to examine the spatial and temporal distribution, as well as the composition, of the flavan-3-ols, proanthocyanidins and flavonols in winegrapes throughout berry development. In addition, we sought to explore how modiffing bunch exposure to light would impact upon flavonoid biosynthesis in the grape berry by comparing the flavonoid composition of shaded and exposed grapes. The expression of key genes in the flavonoid biosynthetic pathway was also investigated to understand how the pathway is regulated during berry development.
The flavan-3-ol and proanthocyanidin composition of both seeds and skin of grapes \¡/as determined by reversed-phase HPLC after acetone extraction and acid-catalysis in the presence of excess phloroglucinol. The main period of proanthocyanidin accumulation in grape seeds occurred immediately after fruit-set with maximum levels observed around veraison. Over two seasons there was variation in both the timing and content of proanthocyanidins in seeds. In skin, proanthocyanidin accumulation occurred from fruit set until 1-2 weeks after veraison. Proanthocyanidin subunit composition rwas different in seeds and skin and changed during berry development. The mean degree of polymerisation of the tannin polymers in skins was higher than in the seeds at all stages of berry development. Proanthocyanidin levels in both seeds and skin decreased between veraison and harvest. Additional proanthocyanidin subunits were released when the residues remaining after acetone extraction were subjected to direct acid-catalysis in the presence of phloroglucinol. In the seeds, these accounted for much of the post-veraison decrease, but not in grape skin. At harvest, 75%o of extractable berry proanthocyanidin was in the seeds. Accumulation of proanthocyanidins in the seeds appears to be independent of that in the skins but in both tissues, synthesis occurs early in berry development and maximum levels are reached around veraison.
Flavonols were identified in the skin of developing berries, buds, tendrils, inflorescences, anthers and leaves of the grapevine. The dominant flavonols were quercetin glucosides with trace amounts of kaempferol glucosides detected in Shiraz flowers but not in developing berries. Flavonols were high in flowers and decreased toward berry-set then remained relatively constant through berry development. Total flavonols per berry were low until pre-veraison then increased during berry development, particularly before ripening, in Chardonnay and during ripening in Shiraz. Two genes encoding the enzyme flavonol synthase (FLS) were isolated from Shiraz flowers. The first of these, VvFLSl, was expressed in leaves, tendrils, pedicels, buds and inflorescences as well as in developing grapes. Expression was highest between flowering and fruit set then declined, increasing again during ripening coincident with the increase in flavonols. Expression of
VvFLS2 was much lower than for VvFLSI and did not change during berry development.
The results indicate that two distinct periods of flavonol synthesis occur in gtapes, the first around flowering and the second during ripening of the developing berries.
To explore the influence of bunch exposure on grape, and thereby wine, quality a shading ffeatment comprising opaque boxes was applied to bunches of Shiraz grapes prior to flowering. The boxes were designed to maintain airflow while excluding light, to minimise changes in temperature and humidity. The effects of light exclusion on berry development, flavonoid accumulation and the expression of genes involved in flavonoid biosynthesis were assessed throughout berry development. There was no significant effect of shading on sugar accumulation and in two of the three seasons studied therc was
u no effect on berry weight. Chlorophyll concentration was much lower in the shaded fruit, which appeared pale yellow until veraison. The fruit coloured normally in the shaded bunches and in two of the three seasons there was no significant change in anthircyanin content. Expression of the gene encoding UDP-glucose flavonoid-3-O-glucosyl transferase (UFGT), a key gene in anthocyanin biosynthesis, increased after veraison and was similar in both shaded and exposed fruit. Anthocyanin composition was altered in the shaded fruit, which had a greater proportion of the dihydroxylated anthocyanins, the glycosides of cyanidin and peonidin. Shading had no significant effect on the levels of condensed tannins in the skin or seeds of ripe fruit. Shading significantly reduced the levels of flavonols in the grape skin. In the exposed fruit, flavonol concentration was highest around flowering then declined as the berries gtew, but there was an increase in 'When flavonols per berry during ripening. the boxes were applied before flowering, shaded fruit had much lower levels of flavonols throughout berry development and at harvest the level of flavonols were less than lÙYo of that in exposed fruit. A gene encoding flavonol synthase (FLS) was expressed at flowering and during ripening in exposed grapes but its expression was greatly reduced in shaded fruit. The results indicate that shading had little effect on berry development and ripening, including accumulation of anthocyanins and tannins, but significantly decreased flavonol synthesis.
This research has contributed to our knowledge of grapevine biochemistry and physiology, partþularly in relation to the biosynthesis of flavonoids. The pattern of accumulation of flavonoids has been determined in the seeds, skin and flesh of gtapes during berry development. These investigations have included study of both the content and composition of the flavonoids, condensed tannins and anthocyanins during berry development. The results indicate that while anthocyanins, tannins and flavonols are all products of the flavonoid pathway, they are synthesised at different times during
lll development and in different tissues of the berry. These observations are consistent with the pattern of flavonoid pathway gene expression reported here and elsewhere and indicate that viticultural management to improve fruit quality needs to consider'the different timing of synthesis of these important compounds.
This research has also revealed much about the influence of bunch exposure on flavonoid metabolism in grape berries. While there were significant yet subtle shifts in the composition of anthocyanins and proanthocyanidins, only the synthesis of flavonols was shown to be light dependent. Whether flavonoid accumulation or composition can be effectively managed through viticultural practices, to what extent they can be modified or manipulated, and how such changes might impact upon winemaking are all fields ripe for future investigation.
Keywords: Grapevine, vitis Vinifera L., Shiraz, Chørdonnay, flavonoid, anthocyanin, condensed tannin, proanthocyanidin, flavan-3-ol, flavan-3,4-diol, flavonol, gene expression, flavonol syntltase, BA^WJLS, LAR, chlorophyll, phloroglucinol, acid catalysis, berry development, seeds, skin, leaves light, shading
lv Author's note. Much of the research presented here, primarily Chapters 3, 4 attd 5 has been published in the following papers:
Downey, M.O., Harvey, J.S. and Robinson, S.P. (2003) Analysis of tannins in seeds and skins of Shiraz
grapes throughout berry development, Australian Journal of Grape and Wine Research 9; 15-27 .
Downey, M.O., Harvey, J.S. and Robinson, S.P. (2003) Synthesis of flavonols and expression of flavonol
synthase genes in the developing grape berries of Shiraz and Chardonnay (ltitis vinifera L.), Australian
Journal of Grape and Wine Research 9; 1 10-121.
Downey, M.O., Harvey, J.S. and Robinson, S.P. (2004) Effect of bunch shading on berry development and flavonoid accumulation in Shiraz grape berries, Australian Journal of Grape qnd Wine Research l0; 55-
73.
v IA Declaration
This work contains no material that has been accepted for the award of any other degree or diploma in any university or other tertiary institution. To the best of my knowledge, no material presented here has been written or published by another person, other than where due reference has been made in the text.
I consent to the loan and/or photocopying of this thesis upon lodgement with the
University of Adelaide Library.
SIGNED:
Mark Downey
v11 IIþ Table of Contents
Abstract.... I
Declaratíon...... vtt
Table of Contents tx
Ac kn owle dgem ents...... xrv
Chapter L. General Introduction...... 1
1.1. The Grapevine ...... 1
1.2. Grapegrowing and Winemaking...... 2
1.2.1. The Wine Industry 3
1.3. The Flavonoids 4
1.3.1.2. Tarurins 7
1X 1.4.2. Factors Affecting Flavonoid Biosynthesis t9
1.5. Conclusions and Project Aims.... 24
Chapter 2. Grape Berry Development...... 27
2. 1. Introduction...... 27
2.1. l. Shading Treatments .28
2.2. Múerials and Methods 29
2.2.2. Timing and Application of Treatments ...... 34
2.2.3. Sampling of Grape Berries 36
2.2.4. Analysis of Berry Development 40
2.2.5. Environmental Differences between Seasons ...... 43
2.3. Results 43
2.3.1. Chardonnay Berry Development 1999-2000..-...... 43
2.3.2. Shttaz Berry Development 1999-2000...... 46
2.3.3. Shiraz Berries Shaded One Week Post-Flowering 1999-2000...... 48
2.3.4. Shiraz Berries Shaded Five Weeks Post-Flowering 1999-2000...... 49
2.3.5. Shiraz Berry Development 2000-2001...... 51
2.3.6. Shiraz Berries Shaded Five Vy'eeks Pre-Flowering 2000-2001...... 51
2.3.7. Shiraz Berries Shaded at Bud-burst 2000-2001 and Comparison of Shading Treatments ..54
2.3.8. Shiraz Berry Development 2001-2002...... 55
2.3.9. Shiraz Berries Shaded Pre-Flowering 2001-2002...- ...... 57
2.3.10. Seasonal Differences in Shiraz Berry Development...... 59
2.3.11. Seasonal Differences in Temperature ...... ,...,,,...... 62
2.3.12. Treatment Differences in Temperature...... 66
2.3.13. Treatment Differences in Chlorophyll .dccumu1ation...... 68
72
Chapter 3. Grape Berry Development...... 81
x 3.1. Introduction...... 81
3.2. Materials and Methods 84
3.2.1. Sampling of Grape Tissues u
3.2.2. Extraction of Flavan-3-ols and Proanthocyanidins...... 85
3.2.3. Analysis of Tannins. 87
3.2.4.HPLC Method Development & Separation of Proanthocyanidins 90
3.3. Resutts...... -...... 92
3.3.1. Identification of Flavan-3-ols and Proanthocyanidin Subunits ...... '...... 92
3.3.2. rWhole Berry Analysis of Tannins 1999-2000'...... '....'...... '.....97
3.3.3. Tannin Composition of Whole Berries 1999-2000.'...... '...... 100
3.3.4. Shiraz Seed Tannins 1999-2000..... '..'...... 104
3.3.5. Shiraz Seed Tannin Composition 1999-2000...... '.'..."...... 105
3.3.6. Shiraz Seed Tannins 2000-2001 '...... '.107
3.3.7, Shi¡az Seed Tannin Composition 2000-2001 "...... 109
3.3.8. Shiraz Skin Tannins 2000-2001...... "'.'...... 111
3.3.9. Shiraz Skin Tannin Composition 2000-2001 ...... '...... '.'.'.113
3.3.10. Shiraz Flesh Tannins 2000-2001..'...... 115
3.3. I I . Direct Acid-Catalysis of Seed & Skin Residues 2000-2001 ll6
3.3.12. Polymer Length (mDP) in Grape Berries lt8
3.3.13. Extractable Tannins at Commercial Harvest....." ..'...'...... 119
3.3.14. Tannins in Grapevines Leaves and Tendrils t2t
3.4. Discussion...... 125
Chapter 4. Flavonol Synthesis and Expression of Flavonol Synthase Genes in Vitis
vinifera L...... 137
4.1. Introduction......
4.2. Materials and Methods 139
4.2.1. Sampling of Grape Tissues 139
xl 4.2.2. Extt action and Identihcation of Flavonols 140
4.2.3. Cloning of Flavonol Synthase...... t42
4.2.4.Expression of Flavonol Synthase Genes t44
4.2.5. Southern and Northern Blot Analysis 145
4.3. Results...... 146
4.3.1. Cloning of Flavonol Synthase Genes...... 146
4.3.2. Southern Blot Analysis ...... 147
4.3.3. Identification of Grapevine Flavonols...... 148
4.3.4. Berry Development, Flavonol Accumulation and FLS Expression in Chardonnay Berries ...
150
4.3.5. Berry Development, Flavonol Accumulation and FLS Expression in Shiraz Berries...... 152
4.3.6. Seasonal Variation in Flavonol Accumulation in Shiraz...... 155
4.3.7 . Fravonol Accumulation and FLS Expression in other Tissues of Shira2...... 157
4.4. Discussion.. .160
Chapter 5. Effect of Shading on Flavonoid Biosynthesis in Vitis vinifera L. cv. Shiraz
169
5. 1. Introduction...... 169
5.2" Materials and Methods t7t
5.2.1.3. Cene Expression Studies in Shaded Beries...... 173
5.3. Results...... t74
5.3.1.1. Effect of Shading on Berry Development... 174
5.3.1.2. Anthocyanin Accumulation in Shaded Grapes 176
5.3.1.3. Tannin Accumulation in Shaded Grapes 182
5.3.1.4. Flavonol Accumulation in Shaded Grapes 190
xll 5.3.1.5. Expression of Genes Involved in Flavonol and Anthocyanin Biosynthesis in
Shaded Berries ..'...... '...... '...192
5.3.1.6. Induction of Flavonol Biosynthesis ...'...... '...... 194
5.4. 1. l. Shading Treatment...... 195
5.4.1.2. Berry Development in Shaded Grapes...... '...' 195
5.4.1.3. Anthocyanins in Shaded Grapes 197
5.4.1.4. Condensed Tannins in Shaded Grapes...... 201
5.4.1.5. Flavonols in Shaded Grapes 204
5.4.1.6. UFGT and FLS Gene Expression in Shaded Grapes...... '...... '."'205
5 .4.1.7 .Induction of Flavonol Biosynthesis ...... '...... '206
Chapter 6. General Discussion and Future Directions ...... 209
6.1 . Tannin Accumulation in Shiraz Berries ...."'...... '209
6.2. Flavonol Accumulation in Shiraz Berries ...... 215
6.3. Effect of Shading on berry Development and Flavonoid Biosynthesis ...'....'..217
6.4. Implications for Winemaking. """"'221
Bibliography
xlll CSIBA Acknowledgments Plant krdustry
I owe a debt of gratitude to many friends and colleagues both getting me to the start, as well as the end, of this project. To everybody at CSI Plant Industry in Adelaide, thanks for your support and assistance. Specifically,I'd like to thank:
Dr. Simon Robinson (CSIRO Plant Industry) - My primary supervisor, firstly for cloning FLS36; mainly for his guidance, support and encouragement and advice on all matters, but primarily for exploring so many of my ideas, crazy and otherwise.
Dr. Graham Jones (University of Adelaide) - My University supervisor, who provided the Merck ChromolithrM columns used to analyse flavonols and introduced me to the ad#eafi;r€s perils of undergraduate teaching.
Dr. John Harvey (CSIRO Plant Industry; CRC for Viticulture) - Firstly for exceptional contributions to the molecular aspects of this project, but mostly for endless (futile) and to the Alma Hotel, Willunga.
Dr. Greg Tanner (CSIRO Plant Industry) - Who first introduced me to an HPLC, for better orworse, and with whom I have had many valuable discussions on the f,rner points of tannin biosynthesis and analysis.
Dr. Jim Kennedy (Oregon State University) - Fo er post-doctoral fellow at the University of Adelaide, with whom I had many discussions on tannin method development and the application of the phloroglucinol method, for the donation of st¿ndards of extension subunit-phloroglucinol adduets, anel eonstru-etive debates on tannin chemistry and biochemistry.
Many thanks also to Karin Sefton (CSIRO Plant Industry) for her untiring technical assistance.
John and Di Harvey (Slate Creek Vineyard, Willunga) for their exceeding generosity with grape samples and Russian Caravan tea and many insights into the grape and wine indusby.
xlv Dr. Mandy V/alker (CSIRO Plant Industry) - for support, encouragement and always being ready for a beer, lunch or a chat.
Dr. Paul Downey (Australian National University; CSIRO Entomology; CRC for Weeds) - Comic relief, critic, devil's advocate, brother and friend.
Chi (CSIRO Plant Indusry) e Sole (University of Bologna) for more than adequate distractions, road-trips, dinners and good company.
thinking.
Mel & Luke (CSIRO Plant Industry) for taking up the Beer & Schnitzel Challenge with me at the Earl of Leicester when I should
have been writing-up.
Thanks also to Allison Hogg (Primary Industries Research, Vic.) for invaluable assistance in preparing this manuscript.
Ingridschmechtig @ A special thank-you, "...it hasn't been all beer and skittles, there was some madness in the middle..."
Finally, to Gregg and Ian. Whose company keeping me sane these past 4 years; who let time-trials and hung on grimly as I blasted who dragged me up endless hills in retribution; then somehow coaxed me into sprinting for the next pedestrian crossing even after a 100 km of agony.
This researoh was funded by an Australian Post-graduate Award, CSIRO Plant Industry (Adelaide), the Grape and Wine Research Development Corporation and the CRC for Viticulture.
Grape and Wine Research and Development Corporation
xv I^X for Igg
xvll ITIAX Chapter L. General Introduction
An ancient association exists between man and the grapevine and the wine produced from its fruit. This association has, at times, taken on religious and cultural significance, the wine variously worshipped and vilified, considered an elixir and a poison. As much today as ever, these views persist and while excess is indubiøbly incompatible with healtþ living, there is accumulating evidence of nutritional and medicinal benefit associated with moderate consumption of wine. Much of this benefit, and a significant component of wine quality, is attributable to a class of plant secondary metabolites known as flavonoids. Although extensively studied in relation to their role as pigments in wine, the biosynthesis of flavonoids in the grapevine has, until recently, attracted little interest.
1.1. The Grapevine
The production of winegrapes, table grapes and raisins is based primarily on cultivars of the species Vitis vinifera of the family Vitaceae (Antcliff, 1988; Mullins et a1.,1992). The
Vitaceae are mainly woody shrubs, predominantly intertropical in distribution and characterised by pentamous flowers and the occurrence of tendrils and inflorescences opposite leaves (Mullins et al.,1992).
There are approximately 10,000 known cultivars of Vitis vinifera classified into three main groups, or proles: Proles pontica, Proles orientalis and Proles occidentalis (Negrul,
1936;Galet,l979;Mullins et a1.,1992; Bourquinet a1.,1993). Proles pontica comprises the oldest cultivars, which arose around the Black Sea (Negrul, 1936; Antcliff, 1988;
Mullins et al., 1992). Proles orientalis arose around the Caspian Sea and gave rise to
Muscat wine grapes and many table grapes (Mullins et al., 1992). Neither Proles pontica noÍ orientalis are suited to the cool climates of Northern Europe and the better-adapted
Proles occidentalis were cultivated in these areas. Cultivars of Proles occidentalis now comprise most of the better-known wine grape varieties (Galet,1979; Antcliff, 1988;
Mullins et a|.,1992).
1.2. Grapegrowing and Winemaking
The domestication of the grapevine began around 5,000 BC along the shores of the Black
Sea. Cultivation spread rapidly throughout the Middle East, Europe and North Africa, finding a prominent role in Greek, Roman and Egyptian civilisations (Hyams, 1965;
Mullins et al., 1992). The advent of winemaking ìù/as likely a serendipitous event resulting from the fermentation of stored grapes. The fruit of the grapevine being an ideal substrate for fermentation and the surface of the berry and bunch morphology provide a favourable habitat for yeasts (Hyams, 1965; Mullins et aL.,1992).
Briefly, the winemaking process consists of the extraction of juice from the fruit at various levels of maturity determined by the style of wine to be made and the fermentation of that juice with a cultured yeast (Berry,1987; Ough, 1992; Boultonet al.,
1998; Rankine, 1998). Both the extraction and fermentation techniques are hugely variable, as are the varieties of grape and strains of yeast, all of which contribute to a rli.rarca ro--o .',ino o+',lo lÞ^',1+^^ 1fìôf . D^-1-:-^ f ôôo\ lrr:-^ s¡tvruv rq¡óv vr^f vv¡r¡w ùrJrw qr¡u^-'{ yq4rrtJ^"^l;i-' \uvurlvrr vL^+ Lt.L,,^l r 77u, l\¿ru\tltg) tzto). yy lllç quality is principally determined by the composition of the grapes in respect of colour, acid, sugar and phenolic content and the management of these parameters throughout vinification and maturation (Winkler et al., 1974; Timberlake & Bridle, 1976; Haslam,
1980; Somers & Verette, 1988; Boulton et al., 1996; Saint-Cricq de Gaulejac et al.,
1e98).
2 1.2.1. The Wine Industry
International wine production is around 30 billion litres annually, 75o/o of which is produced in Europe and the former Soviet Union (Mullins et a1.,1992; Australian Bureau of Statistics,2001). Australia generates only 3% of world production, with almost half of
Australian wine exported, a figure that has steadily grown since the mid-1980s when
Australia was a net importer of wine (Australian Bureau of Statistics, 2001). The greatest export market has traditionally been the U.K. and this market still accounts for 50% of
Australian wine exports (Australian Bureau of Søtistics, 2001). The continued success of the Australian wine industry is dependent upon ongoing production of quality, rather than quantity wines (Osmond & Anderson, 1998).
l.2.2.The Health Benefits of Wine
Much of the recent literature on wine concerns the antioxidant properties of phenolic components and their role in reducing mortrality from free radical associated diseases such as coronary heart disease, atherosclerosis, stroke and some cancers (St. Leger et al.,
1979; Armand et al., 1988; Nagabhushan et a1.,1988; Renaud & de Lotgenl, 1992;
Artaud-Wildet a1.,1993; Hertog& Hollmar¡ 1996; Criqui & Ringel 1994; McDonald et al., 1998). Numerous compounds of the broad general class of secondary metabolites known as flavonoids have been identified as antioxidants in wine. Prominent among these are the anthocyanins, flavonols and flavan-3-ols, as well as polymeric flavonoids (Bors &
Saran, 1987; Pace-Asciak et al., 1995; Sato et al., 1996; Yamasakiet al., 1996; B,aldi et a1.,1997; Fauconneau et a1.,1997; Plumb et a1.,1998; Saint-Cricq de Gaulejac, 1999;
Bors ¿/ a1.,2001).
J L.3. The Flavonoids
The flavonoids are a group of compounds based Flavan 5 on the polyphenolic flavan skeleton (Figure 1.1), 4', variously substituted with hydroxyl, metþI, J 7
2', galloyl, glucosyl and acyl moieties. In addition, 6 3 flavonoids can form complexes with other 5 4 flavonoids, metal ions and numerous other molecules (Asen et al., 1972; BrotlJ.llard, et al.,
1989; Strack & Wray, 1993; Haslam, 1998).
Flavonoids also form oligomers and extensive Isoflavan polymers under arange of biological and chemical Figure l.l. The flavan and isoflavan skeleton of flavonoids; showing the conditions (Harborne, 1967; Delcovt et al., 1983; numbering of carbon atoms and ring labels. Numbering of isoflavans is the Haslam, 1998). same as flavans (Harborne, 1967)
There are more than 4000 known flavonoids of diverse physiological function (Stafford,
1991; Koes et al., 1994; Shirley, 1996). Flavonoids occur in all vascular plants and in most mosses, but have not been detected in algae or bacteria and are only rarely found in fungi (Markham, 1982; Stafford, 1991; weiss et al., 1999). Historically, flavonoids, particularly anthocyanins, have been extensively studied in their role as floral pigments and various colour mutants have aide
(Harborne, 1958; Harborne, 1967; stafford, 1990). studies of these compounds in numerous species, including grapes, have been instrumental to our understanding of the genetic basis of pigmentation (Bateson, 1901; Lawrence, 1950; Barritt & Einset, 1969;
Sparvoli et a1.,1994; Holton & Cornish, 1995; Boss et al., r996a; Boss et at. 1996b). The chromophoric nature of the flavonoid ring structure results in the absorption of light in
4 both the ultraviolet and visible spectra (Markham, 1982). Absorbing ultraviolet light protects plants from UV irradiation and this has been proposed as the archetypal role of flavonoid compounds (Koes et al., 1994; Shirley, 1996; Smith & Markham' 1998).
Flavonoids that absorb both UV and visible light, including pigments such as aurones, chalcones and anthocyanins, also act as attractants for pollination and seed dispersal
(Saito & Harborne,1992; Koes ef al., 1994; Shirley, 1996). Other physiological roles athibuted to this diverse class of compounds include facilitating conditional male fertility in pollen (Ylstra et al., 1992; Taylor, 1995) and the establishment of symbioses with nitrogen fixing bacteria in the rhizosphere (Djordjevic et al., 1987; Clarke et al., 1992;
Recourt et al., 1992; Koes et al., 1994; Shirley, 1996). Flavonoids also have demonstrated roles as deterrents to herbivory through the bitter and astringent nature of the flavan-3-ols and condensed tannins (Horowitz, 1964; Shaver & Lukefair, 1969;
Feeny, 1976;Elliger et a1.,1980; Dreyer et a1.,1981; Nef, 1988; Wagner, 1988; Mole,
1989; Elmqvist et a\.,1991; Harborne & Grayer, 1993; Lu & Bennick, 1998). In addition, many flavonoids and isoflavonoids have phytoalexin activity with roles in plant defence against fungal and bacterial pathogens (Skipp & Bailey, 1977; Gnanamanickam & Smith,
1980; Smith, 1982; Dixon &Lamb,1990; Dakotaet a1.,1993).
In contrast, the absence of flavonoids, particularly tannins, from pasture crops can result in the formation of protein foam during rirmen fermentation leading to pasture bloat and death. The presence of dietary flavan-3-ols and proanthocyanidins precipitate the protein preventingfoamformationandtheonsetof bloat (Tanner etal., 1995;Lietal., 1996;
Bavage et al., 1997). The precipitation of protein by condensed tannin has been used to
quantiff the tannin content of tissues (Hagerman & Butler, 1980; Harbertson et al.,
2002). Further, it is this ability to complex with proteins that led to their use in the
5 tanning of hides and gave this class of flavonoid compounds their generic name, tannins
(Haslam, 1989). The mechanism by which tannins prevent pasture bloat and tan leather is not dissimilar to the ihteractions of tannins in wine with human salivary proteins and polysaccharides creating the mouthfeel that tannins contribute to red wine (Thorngate,
1997; Gawel, 1998; Lu & Bennick, 1998).
1.3.1. Flavonoids in Wine
1.3.1.1. Anthocyanins
Much of the visual character and taste of wine is due to the presence of flavonoid compounds derived from the seeds and skin of the grapes (Ribereau-Gayon, 1974;
Glories, 1988). The classes of flavonoids that contribute significantly to wine are the anthocyanin pigments, the tannins and related monomeric flavan-3-ols, and to a lesser extent, the flavonols. Anthocyanins, extracted from the grape skin during crushing and fermentation, are fundamentally responsible for all of the colour observed in young red wines (Ribereau-Gayon & Glories, 1986; Mazza & Miniati,1993; Mazza, 1995). Free anthocyanins can contribute as much as 99% of the colour of young red wines, with 30-
50%o of these anthocyanins held in copigment associations (Somers & Evans, 1977;
Somers, 1978; Revilla et a1.,1999; Boulton, 2001). However, much of this initial colour is lost during fermentation due to the inherent instabilif of the anthocyanin molecule
(Somers & Evans, 1977; Somers & Evans, 1979; Somers & Verette. 1983). Around half of the anthocyanins originally present in the red wine are converted into polymeric pigments during the first year of storage (McCloskey & Yengoyan, 1981; Somers & verette, 1988; Cheynier et al., 1997). As red wines age, pigment composition becomes progressively more complex, with the colour of old wines being predominantly due to
6 polymeric pigments formed through complex associations between anthocyanins and tannins (Somers, 1971; Timberlake & Bridle, 1976; Nagel & Wult 1979;McCloskey &
Yengoyan, 1981; Bakker et a1.,1986b; Somers & Verette, 1988; Brouillard et al., 1997).
1.3.1.2. Tannins
Tannins are an important component of gustatory impact of wine with flavan-3-ols and proanthocyanidin oligomers and polymers contributing to the body and mouthfeel of the wine (Glories, 1988; Thorngate, 1997; Gawel et a1.,2000). Astringency is a feature of young wines and is usually correlated with flavan-3-ol content (Goldstein & Swain, 1963;
Fulcrand et al., 1996; Czochanska et al., 1979). Many polyphenolic compounds, including the flavan-3-ols and related dimers and oligomers, have a harsh astringent taste and produce in the palate a feeling ofroughness, dryness and constriction caused by the precipitation of proteins and mucopolysaccharides in the saliva (Somers, 1971; Glories,
1988; Jackson & Lombard,1993; Thorngate, 1997; Haslam, 1998). The flavan-3-ols also have a bitter sensory character and a significant difference in perception exists between different monomers, with epicatechin perceived as more bitter for a longer duration than catechin (Noble, 1994; Thorngate, 1997). In addition, the presence of ethanol enhances bittemess intensity and duration (Noble, 1994). The bitterness and astringency of flavonoids decreases as the molecular weight increases with polymerisation (Noble,
1994; Bross aud et a1.,2001 Vidal et al., 2002). The loss of astringency occurs with the formation of large polymeric compounds through the oxidative cross-linking of proanthocyanidin oligomers (Wildenradt & Singleton, 1974; Timberlake & Bridle, 1976;
Somers & Wescombe,1987; Fulcrand et al., 1996; Thorngate, 1997; Haslam, 1998).
These large molecules are insoluble and tend to precipitate from the wine. However, the
formation of tannin complexes with anthocyanins increases the solubility of large
7 molecular weight polymers (Cheynier et al., 1997). The t¿nnins readily identified in red wines include the flavan-3-ols catechin, epicatechin and epicatechin-gallate, as well a variety of dimers, oligomers and extensive'heterogeneous polymers (Salagoity-Auguste
& Berhand,1984; Singleton, 1988; Fulcrand et a1.,1999).
1.3.1.3. Flavonols
In addition to anthocyanins and tannins, glycosides of the flavonols myricetin, quercetin and kaempferol are commonly found in wines. While tannins and anthocyanins in wines tend to form large polymeric compounds during vinification and maturation, flavonol glycosides are hydrolysed by acidic conditions and consequently, the flavonol aglycones are also commonly found in wines (Somers & Verette, 1988; Blanco et al., l99B;
Goldberg et al., 1998; McDonald et aL.,1998).
The phenolic profile of white wines is dominated by gallic acid and the hydroxycinnamates, caftanc acid and coutaric acid (Ong & Nagel, 1978a; Singleton er al., 1978). There are virhrally no flavan-3-ols or tannins in white wine due largely to the method of winemaking in which skins and seeds are removed after crushing (Singleton &
Trousdale, 1983; Romeyer et al., 1985; Somers et a1.,1987; Somers & Verette, 1988).
Flavonols in white wines may be derived from the skin during pressing, or the result of cottl¿tttittatiorr with leaf material. Flavonol glycosides and their breakdown products produced during winemaking have low solubility in aqueous solution and can cause a hazing effect in white wine (Somers &, Ziemelis, 1985).
While the conditions of winemaking and aging significantly effect the final concentration of flavonoids in wine, it is the amount and composition of flavonoids in the fruit at harvest that will largely determine the flavonoid profile of the mature wine. Around 65%
8 of the phenolics in wine come from the skins and the remaining 35%o from the seeds.
However, much higher levels of phenolic compounds are reported in grape seeds reflecting the lower extractability of phenolið compounds from seeds, which is thought to be due to lignification of the seed (Singleton & Draper, 1964; Meyer & Hernandez,l9T0;
Ribereau-Gayon & Glories, 1986; Singleton, 1988; Thorngate & Singleton, 1994; Revilla et al.,1997).
1.3.2. Flavonoids in Grapevines
1.3.2.1. Anthocyanins
Anthocyanins are probably the most studied phenolic component of grapes because of their contribution to the colour of red wines (Mazza & Miniati, 1993; Mazza, 1995).
Anthocyanins occur in the vacuoles of dedicated cells in the epidermal layers of the skin of red grapes and in the pulp of red fleshed varieties (Mueller & Greenwood, 1978;
Pecket & Small, 1980; Moskowitz & IJrazdina, 1981; Sparvoli et al., 1994). The anthocyanins in grapes are the -3-O-glucosides, -3-O-acetylglucosides and -3-O-p- coumaroylglucosides of cyanidin, petunidin, peonidin, delphinidin and malvidin (Figure
1.2). Malvidin is the major anthocyanin in red grapes, with malvidin glucosides accounting for around 60Yo of total pigment (Crippen & Monison, 1986b; Mazza &
Miniati, 1993b;Mazza,1995). Anthocyanin pigments are also observed in the young vine leaves of several varieties but not generally in mature leaves (Pirie & Mullins, 1976;
Darne, 1993; Darne,1996;Ezzlli et al.,1999). White grapes do not produce anthocyanins because of mutations that result in their failure to express the final steps in anthocyanin
synthesis (Boss et al.,l996b).
9 Clanidin Delphinidin OH @ @ H Malvidin
OH OH cHs OH Peonidin OH Petunidin OH @ o H Malvidin-3 O-glucoside OH H3 OH OH @ OH H o-cH3 Malvidin-3 -O-ac etylglucoside Malvidin-3O-p-coumaroylgluc osi de H OH CHs o o. @ H H H o-cH3 H3
H H OH OH H
H H
OH OH H OH
Figure 1.2. Anthocyanins inVitis vinifera Shiraz are based on the anthocyanidins cyanidin, peonidin, delphinidin, petunidin and malvidin occuning in three glycosidic forms, the glucoside, acetylglucoside and coumaroylglucoside, shown here as the glycosides of malvidin
L3.2.2. Flavonols
Several studies have been made of flavonols in grapes focussing on the induction of flavonols by UV light and exploring the apparent correlation between this phenomenon and the qualify of grapes and subsequent wines (Morrison & l.loble, i990; Price et al.,
1995). The glucosides of quercetin (Figure 1.3) are the dominant flavonols in grapes, although glycosides of myricetin and kaempferol have also been reported, albeit at significantly lower levels (Cheynier & Rigaud, 1986; Spanos & Wrolstad, 1990; pnce et al., 1995). Flavonols occur in the upper epidermis of the grape berry, consistent with a role in IfV screening and have not been reported in the seeds or flesh of the berries (Flint
l0 (+)-Kaempferol (+)-Quercetin (+)-Myricetin ot-l o{-t
t-to l-to t-to ol-t
l-o o t-to o t-to o
(+) Quercetin-3 4-glucoside (+)-Querc etin-3-O-glucuronide OH OH
l-to t-{o o
H o H o{-l o o. H H H
ol-l oFl t-{o t-to H OH H oÈl
Figure 1.3. Flavonols found in Shiraz are based on the aglycones kaempferol, quercetin and myricetin. The most prevalent flavonols being quercetin glucoside and quercetin glucuronide et a1.,1985; Beggs et a1.,1987; Somers & Verette, 1988; St¿fford, 1990; Haselgrove,
1997). Glucosides of kaempferol, quercetin and myricetin also occur in the leaves of Vitis spp., at levels considerably greater than those in the fruit (Somers & Ziemelis, 1985;
Darne, 1996; Hmamouchi et a1.,1996). Stems of grapevines have also been reported to contain flavonols, primarily glucosides of quercetin (Souquet et a1.,2000).
1.3.2.3. Tsnnins
The flavan-3-ol content of grapes has also been extensively studied due to the contribution of the related tannins to wine quality and more recently because of their nutraceutical potential. In the seeds and skin of gfapes, the flavan-3-ols catechin,
epicatechin and epicatechin-gallate are the colnmon monomers (see Figure 1.4) with over
twenty dimeric and trimeric proanthocyanidins also identified (Lea et a1.,1979; Romeyer
et a1.,1986; Escribano-Bailon et a1.,1992; Prieur et a1.,1994; Santos-Buelga et al., 1995;
ll Cheynier et a1.,1997; Fuleki & Ricardo da Silva, 1997). Of Catechi n the oligomers, only the B-type proanthocyanidins are present
in grapes (Figure 1.5). These mainly have 4p-8 linkages with
H small amounts of dimers and trimers containing 4þ-6 Epi catechin
linkages (Escribano-Bailon et al., 1992; Cheynier et al.,
1997; Fuleki & Ricardo da Silva, 1997). H
Epigallocatechin In the grape seeds, flavan-3-ols and proanthocyanidins are
found in the seedcoat. The levels of catechin and epicatechin
are reputedly similar with lesser amounts of epicatechin- Epicatechin-gallate gallate. In addition, all of the B-type dimers (Figure 1.5), and
the C2 trimer have been reported in grape seeds at varying
H levels (Lea et a1.,1979; Romeyer et al.,1986; Prieur et al.,
1994; Thorngate & Singleton, 1994; Santos-Buelga et al., o; tÏJff;3-"iï lltJ,l,"J 1ee5; cheyn ier et at., teel;Katatinic & Males, 1ee7; saint- Shiraz grapevines, showing the 2,3-trans and 2,3-cis Cricq de Gaulejac et a1.,1997; de Freitas & Glories, 1999). stereochemist4r of catechin and the epi- enantiomers The dominant stereochemistry of grape seed oligomers and polymers is epicatechin with the ratio epicatechin:catechin increasing fuom 2:l to I2:l with increasing polymerisation with catechin relatively more abundant as the terminal subunit unit (Prieur et al., 1994; Souquet et al.,1996a; Cheynier et al.,1997\.
The difference between seed and skin tannins is primarily the presence of the trihydroxylated prodelphinidins in the skin and a higher degree of polymerisation @oo &
Porter, 1980; Souquet et al., 1996a; Cheynier et al., 1997). The dominant flavan-3-ol monomer in grape skin is catechin (Escribano-Bailon et al., 1995) with the trihydroxylated products, gallocatechin, epigallocatechin and epigallocatechin-gallate
12 detected in addition to catechin, epicatechin and epicatechin-gallate @oo & Porter, 1980;
Foo & Porter, 1983; Souquet et al.,1996a; Haslam, 1998). In skins, only the dimers B1 to
B4 are commonly reported, although trace amounts of the B5 dimer, which is one of the
most prevalent in the seed, have also been reported (Santos-Buelga et a1.,1995; Souquet
et al., 1996a; Saint-Cricq de Gaulejac et al., 1997; de Freitas & Glories, 1999). In
addition, proanthocyanidins in grape skin are reported to form complex associations with
other cellular components, including cell wall polysaccharides, lignins and proteins
(Amrani-Joutei et al., 1994; Escribano-B aílon et al., 1995; Cheynier et al., 1997; Saint-
Cricq de Gaulejac et al., t997).
Flavan-3-ols are not detected in the pulp of either red or white grape berries, although
they have been reported in the vascular traces of the fruit (Lea et al., 1979; Lee &
Jaworski, 1989; Ricardo da Silva et al., l99la; Souquet et al., 1996a). Flavan-3-ols have
also been detected in the stems of the grape bunch, with the composition of stems similar
B1 B2 B3 B4 O-Epica (-)-Epicatechin (+)-Catechin (+){atechin
(+)-Catechin (-!Epicatechin (+)-Catechin (lEpicatechin
B5 B6 B7 B8 (-)-Epicatechin (+){atechin (+)-Epicatechin (+)-Catechin
(-)Epicatechin (+)-Catechin (+)-Catechin (-lEpicatechin
Figure 1.5. Proanthocyanidin dimers occurring irLl/itis vinifera Shiraz. All dimers shown have been detected in grape seeds, while only dimers Bl-84 with traces of B5 are reported in grape skin.
13 to that of the fruit (Ricardo da Silva er al., l99lb). Tannins have also been observed in the leaves of Vitis spp. (Darne, 1993). The tannin composition of the seeds and skin of white grapes is similar to that of red grapes, although there is considerable variation in the relative proportions between varieties (Lea et al., 19791' Romeyer et a\.,1986; Ricardo da Silva et al., 1991b; Amrani-Jouteiet a1.,1994; Cheynier et al., 1997).
1.4. Biosynthesis of Flavonoids
Flavonoid biosynthesis is the culmination of two metabolic pathways, the shikimate pathway and the phenylpropanoid pathway. The phenylpropanoid pathway synthesises flavonoids from carboxylated acetyl-CoA (malonyl-CoA) and the amino acid phenylalanine, which is produced in the shikimate pathway (Dewick & Haslam, 1969;
Heller & Forkmann, 1993). Under normal growth conditions, aboú 20Yo of the carbon fixed by plants flows through the shikimate pathway and about 2Yo of all carbon flows from the shikimate pathway into phenylpropanoid metabolism (Markham, 1982;
Herrmann, 1995).
The phenylpropanoid pathway is generally considered to culminate in anthocyanin synthesis (Figure 1.ó). Along the way precursors for a number of other products, are formed including lignin, lignan, aurones, flavones, isoflavonoids, flavonols and flavan-3- ols (Harborne, !967; Gerats & Martin, 1992; Haslam, 1998). Much of the biosynthetic pathway of the flavonoids has been elucidated through characterisation of colour mutants and genes encoding many enzymes in the pathway have been cloned @orkmanr¡ 1993;
Helariutta et al., 1993; Holtonet al., I993a; Diallinas & Kanellis, 1994; Sparvoli et al.,
1994; charner et a1.,1995; Tanaka et al., 1996; Gong et al., 1997; Tanner et al., 2003;
Xie et a1.,2003).
t4 ere¡Âdd
GH + + c¿A ¡cl, CHR
MùsednGdcffi lsl¡SjdÈgenin C¡Mmde
CH frI + f!5H <- F¡H
Uqu¡dis¡¡¡ P okhyturyftvomne Eóölyol MdnFin X \ tFs F3H FIH
D3þeh
R5'H F3H
IOMT Dhyftryicdin Dhydqwrcdin \^ DN DR DR
Mriær h Qlercdin
tr5,H f3H <- <- IFOH
kwdphinidn \T \s 2 llvd Dxyfm Éi¡ Lmx Lmx
PD&ph¡nidh 1 \ 1 ø a Pdþhinìdn líEr .Çåy + CFild' Ddph¡ndh * \. Pß I WGT + "_,"i**
+ EpiGl@hin + t o + Édc¡ph + + & DdpHnidhJ-slucodè Mlphindh plrer Cye¡d¡n-3-sbosi& PDryon¡ù plFq Èlryoddn-l-glu@d&
Figure 1.6. The general phenylpropanoid pathway involved in the biosynthesis of flavonoids. PAL - phenylalanine ammonia lyase; C4H - cinnamate-4-hydroxylase; C3H -p-coumarate-3-hydroxylase; 4CL - 4-coumarate CoA ligase; CCoA3H -p-coumaroyl-CoA-3-hydroxylase; StSy - stilbene synthase; CHS - chalcone synthase; CHR - chalcone reductase (in concert with CHS); CHI - chalcone isomerase; F3H - flavanone-3p-hydroxylase; F3'H - flavonoid-3'-hydroxylase; F3'5'H - flavonoid-3'5'-hydroxylase; FLS -flavonol synthase; DFR - dihydroflavonol reductase; LAR - leucoanthocyanidin reductase; LDOX -leucoanthocyanidin dioxygenase (a.k.a. ANS - anthocyanidin synthase); ANR - anthocyanidin reductase UFGT UDP glucose: flavonoid-3-O- glucosyltransferase; IFS - isoflavone synthase; IOMT - isoflavone-4'4-metþltransferase; IFOH -isoflavone-2'- hydroxylase; IFR - isoflavone reductase; PTS -pterocarpan synthase
15 1.4.1. Flavonoid Biosynthesis in Grapevines
1.4.1.1. Anthocyanins
Flavonoid biosynthesis in grapevines has to date only been studied in the skin of the berry and has primarily focused on anthocyanin biosynthesis (Boss er al.,l996a,Ford, et al.,
1998). In red grapes, anthocyanin accumulation commences with veraison, the onset of ripening (somers, 1976; Mullins et al., 1992; Boss et al., 1996a). Initially, only glucosides of the dihydroxylated anthocyanins, cyanidin and peonidin accumulate, followed by the trihydroxylated anthocyanins delphinidin, petunidin and malvidin
(Mazza & Miniati, I993b; Boss e/ al., 1996b; Katalinic & Males, 1997; Keller &
Hrazdina, 1998). The accumulation of anthocyanins in the skin of red grapes coincides with expression of the gene encoding the final step in anthocyanin biosynthesis, UFGT
(Boss er al., I996a).In white grapes, there are no anthocyanins present in the skin and the gene encoding UFGT is not expressed (Boss et al.,l996b).
While genes encoding many flavonoid biosynthetic en-4¡mes have been cloned, it is relatively rare for all of them to be cloned from the same species (Helariutta et a1.,1993;
Charrier et a1.,1995; Liew et a1.,1995; Pelletier & Shirley, 1996; Fukada-Tanaka et al.,
1997; Gong et al., 1997; Rosati et al., 1999; Saito er al., 1999). Fortunately, Vitis viniferø is one such species with phenylalanine arrunonia lyase (PAL), chalcone synthase (CHS), chalcone isomerase (CHI), flavanone-38-hydroxylase (F3H), dihydroflavonol reductase
(DFR), leucoanthocyanidin dioxygenase (LDOX a.k.a. ANS - anthocyanidin synthase) and UDP glucose: flavonoid-3-O-glucosyltransferase (UFGT) having been cloned
(Sparvoli et al., 1994). The expression studies during berry development showed all of these genes except UFGT to be expressed from flowering until just prior to veraison
l6 when transcript levels decreased. At veraison, expression of these genes plus UFGT was observed (Boss el al., 1996a). The enigma of these results is that the pre-veraison expression of most of the flavonoid pathway genes occurs in the absence of anthocyanin synthesis. The accumulation of other flavonoids such as the tannins and flavonols could explain the early expression of genes in this pathway.
1.4.1.2. Tannins
Tannins appear to accumulate early in berry development reaching a peak around veraison and then decreasing towards harvest, with similar patterns of accumulation reported in both the seeds and skin of the grape berry (Somers, 1976; Czochanska et al.,
1979;Dame,1982; Romeyer et al., 1986; Lee & Jaworski, 1989; Katalinic & Males,
1997; de Freitas & Glories, 1999; Kennedy et al., 2001). Recent work in this area suggests the peak in seed tannin accumulation may actually occur several weeks prior to veraison (Kennedy et a1.,2000b). Grape tannins are primarily polymers comprised of flavan-3-ol terminal subunits and flavan-3,4-diol extension subunits (Figure 1.7) . Prior to veraison the composition of these polymers is primarily the dihydroxylated subunits catechin and epicatechin, whereas post-veraison the trihydroxylated epigallocatechin is also detected in the grape skin (Czochanska et a1.,t979;Katalinic & Males, 1997; Foo &
Porter, 1980).
The enzymatic step involved in the biosynthesis of flavan-3-ols appears to be catalysed by two enzymes accounting for the 2,3-trans and 2,3-cis stereochemistry of the products catechin and epicatechin (Figure 1.4). The first of the two enzymes involved in flavan-3- ol biosynthesis is leucoanthocyanidin reductase (LAR, see Figure 1.6), which converts the leucoanthocyanidins to the corresponding flavan-3-ol with 2,3-trans stereochemistry,
for example leucocyanidin to catechin (Stafford & Lester, 1985; Kristiansen, 1986;
t7 Stafford, 1990; Joseph et al., 1998). The second enz.lrne, anthocyanidin reductase (ANR, see Figure 1.6), converts the anthocyanidins to their respective 2,3-cis flavan-3-ols, for example cyanidin to epicatechin (Albert et al., 1997; Devic et al., 1999; Xie et a\.,2003).
The mechanism of tannin polymer formation is uncerüain. While several enzymatic mechanisms have been postulated there is little convincing evidence to support them
(Haslam, 1998). In contrast, extensive investigations in wine and model solutions have
3
4', 7 5' Rl 2 6
Extension Subunit
R
InterflavanBond €
Terminal Sr¡bunit
Flavan-3-ol 2-3 Stereochemistry Rr R, Catechin trans H H Epicatechin cls H H Epicatechin-gallate cis H Gallic Acid Epigallocatechin cis OH H
Figure 1.7. General proanthocyanidin structure showing terminal flavan-3-ol subunits, flavan-3,4-diol extension subunits and interflavan bond linking polymeric subunits, as well as carbon numbering of the flavan skeleton and nomenclature of monomers
18 demonstrated several non-enzyme catalysed mechanisms for polymer formation (urd,
1972;Delcour et al., 1983; Bakker & Timberlake, 1997; Saucier et al., 1997; Mateus et al., 2002b; Y idal et al., 2002).
1.4.1.3. Flavonols
The pattem of flavonol accumulation in grape berries has been little studied to date
(Haselgrove,1997). Flavonol accumulation appears to occur prior to veraison as the level reported at veraison was the maximum reported between that time and harvest
(Haselgrove et a1.,2000). Biosynthesis of flavonols from dihydroflavonols (see Figure
1.6) is catalysed by flavonol synthase (FLS) (Forkmann et a1.,1986). While the FZ^S gene has not been cloned in Vitis vinifera, it has been cloned in Petunia, Arabidopsis, Solanum and Citrus (Holton et a1.,1993b; Pelletier et a1.,1997; van Eldik e/ al., 1997; Moriguchi et a1.,2002).
1.4.2. Factors Affecting Flavonoid Biosynthesis
The factors reputedly affecting flavonoid biosynthesis in plants are myriad and include light, temperature, altitude, soil type, water and nutritional status, microbial interactions, pathogenesis, wounding and defoliation, plant growth regulators and developmental processes. These factors are also closely intertwined, for example in forest clearings where light intensþ is greater than inside the forest, plants have higher levels of tannins.
Initially, this appeared as an effect of light on tannin accumulation. However, it was also observed that in these clearings, the incidence of defoliating insects was significantly
greater due to the presence of a high proportion of juvenile plant material (Mole et al.,
1988; Mole & Waterman, 1988). In this case, tannin accumulation was induced by
t9 herbivorous attack rather than light, a widely reported phenomenon (Feeny, 1976;
Walters & Stafford, 1984; Wagner, 1988; Elmqvist et al.,I99I).
The effect of incident radiation upon flavonoid accumulation in the plant is also intertwined with changes in photosynthesis and a number of investigations into the effects of sun exposure and shading on both development and flavonoid metabolism have been eonducted on grapevines. Where whole canopies have been shaded, the flavonoid content of grapes was observed to decrease significantly @uthose et al.,l97I;' Kliewer,
1977; Cnppen & Morrison, 1986b; Smart et al., 1988). However, there was also a general decrease in carbon fixation in these grapevines, which resulted in lower levels of other metabolites such as sugars and organic acids (Kliewer et al., 1967; Kliewe¡ 1977; Smart et al., 1988; Rojas-Lara & Morrison, 1989). Although, these differences have also been attributed to greater water loss from exposed berries than from shaded fruit (Crippen &
Morrison, 1986a).
Furthermore, sun-exposure results in an increase in temperature in the plant, either through direct heating of the plant from incident radiation or increases in the ambient air temperature (Jones, 1992). As the temperature of a biological system increases, so too does metabolism with an associated increase in development (Hawker, 1982; Jones,
1992: Ebadi et al., r995a; Dokoozlian & Kliewer, 1996). However, at higher ierrrperatures metabolism ceases (jones, 1992), in grapevines this temperaturc is around
30'C (Coombe, 1987b). Similarly, decreasing temperature slows metabolism and results in reduced carbon fixation as well as reduced production of secondary metabolites (Jones,
1992)" In grapevines, temperatures below 9-10'C are generally considered to slow metabolism to the point where little or no growth occurs (winkler et al., 1974; Jackson & spurling, 1988; Jackson & Lombard, 1993; Boulton et al., 1998). Below a certain
20 temperature metabolism ceases and at some point chilling itju.y will occur (Jones, 1992).
In grapevines, freezing injury to the wood or dormant buds occurs at around -15oC, although this is rarely a problem in Australia (JackSon & Spurling, 1988). Of greater concern is the effect of low temperature around the time of flowering, which has been shown to significantly impact upon fruit-set (Buttrose & Hale, 1973; Ebadi et al.,I995a;
Ebadi et al.,I995b). Both high and low temperatures have also been shown to effect the anthocyanin and tannin content and composition of many plants including grapevines
(Kliewer, 1970; Buthose, 1970; Lees et al., 1994; Bel-Tal & King, 1997; Morris &
Robbins, 1997; Leng et a1.,2000). Unfortunately, the application of shading treatments has also been shown to substantially alter the temperature and humidity within grapevine canopies (Kliewer et a1.,1967; Rojas-Lara & Morrison, 1989; Haselgrove et a1.,2000).
Increased humidity in the canopy has a two-fold effect, by lowering the vapour pressure deficit, transpiration and photosynthesis are decreased, reducing growth, and possibly also flavonoid accumulation (Jones, 1992).In addition, high humidity increases the risk of pathogenesis through fungal or bacterial infection (Emmett et al., 1992), which may cause a wound response inducing flavonoid accumulation (Mehdy & Lamb, 1987; Vogt et a1.,1994).
Chilling injury can also effect flavonoid accumulation (Leyva et al., 1995), but is likely part of a general wound-response that has also been documented to increase flavonoid biosynthesis (Chalmers & Faragher,1977;Yogt et al., 1994).In some instances this is
related to defoliation as described above, while in other cases, pathogenic inducers effect
flavonoid accumulation (Langcake & Pryce, 1976;Langcake & Pryce, 1977b; Harrison &
stickland, 1980; Pool et al., l98l; Marschner, 1995; HipsÞind et al., 1996; Lo &
Nicholson, 1998). In addition to pathogens, soil microorganisms, both mycorrhizal
2l (Chabot et a1.,1992;Harnson & Dixon, 1994) and rhizobial (Clarke et al., 1992), have been shown to affect flavonoid accumulation. The induction of flavonols in the formation of symbioses with nitrogen-fixing bacterium occurs in response to low soil nitrogen (Zaat et a1.,1989; Recourt et a1.,1992). In the presence of high soil nitrogen, a different group of flavonoids are exuded to inhibit nodulation f irmin et a1.,1986; Djordjevic et al.,
1987; Dakora et al., 1993). Thus, what appears to be microbial induction of flavonoid accumulation is in fact a response to nutrient st¿tus. Nevertheless, the impact of wounding and microbial association upon phenylpropanoid biosynthesis in grapevines has been extensively investigated in relation to the accumulation of the stilbene ph¡oalexins resveratrol and viniferin pangcake & Pryce, 1976; Langcake & pryce, r977a; Langcake & Pryce, r977b; Langcake & Mccarthy, 1979; pool et al., rgïl; creasy & coffee, 1988; Dercks & creasy, 1989; Jeandet et al., 1995). However, this work has yet to be extended to flavonoids such as flavonols, tannins and anthocyanins.
Nutrient availability has a fundamental influence upon plant growth (Russell, 1961;
Marschner, 1995; Keller et al., 1998). Nutrient status has been shown to effect the flavonoid composition of plant tissues. Both low and excessively high levels of nitrogen fertiliser have been shown to decrease colour in grape berries ((liewer, 1977; Keller &
Iftazdina, 1998) and high potassium has been reported to decrease colour in grapes
(Jackson & Lombartl, 1993). While in forage legumes, high phosphate levels inhibited tannin accumulation (Morris & Robbins, 1997), the effect on grape tannins has not been examined. Large increases in N, P and K have been shown to decrease phenolic content in plants. However, while increasing fertilisation generally had a positive effect on flavonoid content, it can lead to excessive vigour resulting in decreased flavonoid accumulation.
22 In addition to the chemical characteristics of a soil, amended or otherwise, physical characters can also effect flavonoid accumulation (Jackson & Lombard,1993; McDonald et a1.,1998). Such characters as the parent material and the age of the soil that largely determine the micronutrient pool, structure and texture have a significant effect on plant growth (Russell, 196l; Northcote, 1992; Marshner, 1995). However, the major consequence of soil type is the capacity of the soil to hold water while remaining sufficiently well-drained to avoid waterlogging (Russell, 196l; Northcote, 1992).
Irrigation alleviates reduced development and growth through water stress, although some reports suggest water deficit increases tannin and anthocyanin content in grapevines
(Nadal & Arola, 1995; Dry et a\.,1993). However, other workers suggest excessive water decreases tannin content (Kennedy et a1.,2000a), and that water deficit has no effect on tannin or anthocyanin accumulation (Kennedy et al., 2000a; Stoll, 2000). In contrast to the vines, anthocyanin biosynthesis in grape cell cultures is extremely sensitive to osmotic stress (Do & Cormier, 1991; Suzuki, 1995).
The difficulty with interpreting water deficit treatments is the range of other effects that occur throughout the plant, that are regulated by plant hormones in response to stress.
Stomat¿l closing in response to water deficit reduces photosynthesis, thereby reducing all metabolite accumulation (Jones, 1992). Furthermore, plant growth regulators such as, abscisic acid, ethylene, cytokinins, gibberellins and auxins have all been reported to influence the flavonoid content of plants, including grapevines (Matsushima et a1.,1989;
Woltering & Somhorst, 1990; Deikman & Hammer,1995; Moalem-Beno et al', 1997;
Morris & Robbins, 1997; Stoll,2000).
Generally, a large number of variables have been shown to impact upon flavonoid
composition in many plants, including grapevines. In addition, many of these factors are
23 closely related and difficult to isolate experimentally. Despite the broad variation observed with many of these parameters, the greatest affectors of flavonoid content in grapevines are site and season @akker et al., 1986a; Gonzalez-San Jose et al., 1990;
Revilla et a1.,L997;McDonald et a1.,1998; de Freitas & Glories, 1999; Guidoni et al.,
2002).
1.5. Conclusions and Project Aims
This project grew out of the original work of Paul Boss (1998), who examined the expression of flavonoid pathway genes in Vitis vinifera L. cv. Shiraz during berry development. While examining the biosynthesis of anthocyanins in grape berries, he observed that the genes encoding the whole biosynthetic pathway with the exception
UFGT, the gene encoding the final step in anthocyanin biosynthesis, were expressed throughout most of berry development. UFGT was only expressed post-veraison when anthocyanin synthesis commenced in the berry skin. What purpose then did expression of the pathway prior to veraison serve? The most likely explanation is that the pathway is producing other flavonoid compounds early in berry development, likely candidates being the flavonols and flavan-3-ols, two classes of compounds that diverge from the pathway before anthocyanin synthcsis (Figure 1.6).
It is rare that the whole or even e substantial part of the flavonoid biosynthetic pathway has been examined in a single plant and these generally at a single point in development.
Vitis vinifera is an ideal system to examine flavonoid biosynthesis as it produces all of the major classes of flavonoids formed by branchpoints in the pathway. An examination of the spatial and temporal pattern of accumulation of these compounds, and the expression
24 of genes involved in their synthesis would substantially increase our understanding of fl avonoid biosynthesis.
While many studies have been conducted on grape flavonoids these have also tended to focus on flavonoid composition at harvest or changes during ripening. In addition, the bulk of this work has been conducted on important European or North American cultivars grown under the prevailing conditions at those sites. As wine quality reflects that of the fruit, and flavonoids are a significant component of wine and grape quality, defining the pattern of flavonoid synthesis in winegrapes would also be a useful contribution to the production of high quality wine.
This project sought to make a comprehensive examination of tannin, anthocyanin and flavonol accumulation in tissues of the major Australian winegrape varieties, Shiraz and
Chardonnay, throughout the growing season and in seeds and skin during berry development over successive seasons. In addition, it was intended to clone the genes encoding the determinant enzymes in the biosynthesis of flavonols (FZ,f) and flavan-3-ols
(LAR). Moreover, the prevailing zeitgeist suggests light has a strong influence on all aspects of flavonoid metabolism, although much of this work has been complicated by microclimate affects such as temperature and humidity. Therefore, flavonoid composition and gene expression studies would be extended to a shading treatment of the fruit that
sought to eliminate these effects.
Thus, in summation, the aims of this project were:
a To determine when and where flavonols are synthesised during berry development.
To determine when and where flavan-3-ols and proanthocyanidins (tannins) are
synthesised during berry development.
25 o To determine the influence of light on flavonoid biosynthesis in grapes.
o To clone and characterise key genes in flavonol and ønnin biosynthesis in grapes and
to determine their expression pattem and correlate it with biosynthesis.
a To investigate the basis of pre-veraison expression of most of the flavonoid
biosynthetic pathway as observed by Boss et al. (I996a).
26 Chapter 2. Grape Berry Development
2.l.lntroduction
The fruit of the Vitaceae are berries, the archetypal grape being a black berry with
rWhite anthocyanins in both the flesh and skin (Mullins et al., 1992). and pink skinned
grapes are widely cultivated, as are several pink-fleshed varieties, although the flesh of
most extant grape cultivars is not pigmented (Somers, 1976; Mullins et al., 1992). Berry
development is bimodal, the initial growth phase being mainly cell division followed by a
lag period before the second phase of growth, which is predominantly cell expansion
(Cawthorn & Morris, 1982; Coombe,1992; Mullins et a1.,1992; Coombe & McCarthy,
2000). Ripening occurs during the second phase of development (Figure 2.1) and is
characterised by softening of the berry, accumulation of sugars and met¿bolism of
organic acids (Winkler et al., 1974; Pftie & rFqr Ser VERAISoN RIP+sS
Mullins, 1980; Coombe, 1988; Nunan et al.,
1998). In addition, pigments, flavour and
aroma compounds accumulate during
ripening @irie & Mullins, 1980; Coombe,
1988; Coombe, L992; Mullins et al., 1992;
Coombe & McCarthy, 1997). The relative
ÞU@AR contribution of each of these components to
the grape and wine is influenced considerably
by site, season, viticultural practice and the
timing of harvest. Time afterflowering Figure 2.1. Schematic representation of grape berry development
27 For a given site in an irrigated vineyard, it is assumed that characteristics such as soil will remain invariant, nutrition will be adequate and viticultural practices will not vary greatly from year to year. Thus, the primary seasonal difference at such a site will be climatic, predominantly temperature and light, which have demonstrated impacts on berry development and flavonoid metabolism.
As this project sought to compare flavonoid composition in both exposed and shaded fruit, between cultivars and across seasons, it was necessary to quantiff fruit development in each treatment to enable valid comparisons. Failure to do this might athibute differences in flavonoid content directly to treatment effects where in fact the cause may have been differences in fruit development.
2.1.1. Shading Treatments
A large body of historical data exists on the effects of sun-exposure on grape berry composition. Some of these investigations involved the application of physical shade treatments such as plastic sheeting (Kliewer et al., 1967), shade-cloth (Smart et al.,
1988), bags (Kliewer & Antclifl 1970) and foil-clad cages (Flaselgrove et a1.,2000).
Others have tried binding canes under the canopy (Archer & Strauss, 1989), sampling different parts of the canopy where the light regimen was perceived to be different (Price et al., i995; Haselgrove et a1.,2000; Bergvist et a1.,2001) and defoliation (Kliewer &
Antcliffi 1970; Hunter et a1.,1995). In addition, treatments have been applied at different developmental stages from fruit-set to veraison, with samples taken primarily during ripening, or at harvest. In many of these studies, the treatments have altered other aspects of vine physiology, such as cane length, leaf number and foliage exposure as well as bunch exposure. These approaches have also produced considerable variation in
28 lighldark levels between studies, as well as a range of microclimate effects, notably changes in temperature with concomitant impacts on photosynthesis and metabolism.
Thus, it has sometimes been difficult to discriminate between the effects of treatments and other microclimate influences. To add¡ess this, it was proposed that a shading treatment be applied that would eliminate other microclimate effects while achieving light exclusion.
2.2. Materials and Methods
2.2.1. Design of Light-Proof Boxes
Of the previous strategies (Section 2.1.1) the application of bags to developing bunches seemed an economic and straightforward approach. A preliminary investigation into the temperature difference between the ambient air and the inside of an aluminised paper bag was conducted. Temperature was logged over several days using a Tinytag PlusrM logger
(Hastings Data Loggers, Australia) with an external temperature probe inserted into the bag. A significant difference in temperature was recorded, particularþ in full sunlight, with the inside of the bag significantly warmer (data not shown). In addition, the inside of the bag became extremely moist from condensation indicating high humidity within the bag. These observations indicated an opportunity for aradical new approach to shading.
To shade bunches of grapes, the method of shading needed to fulfil several requirements.
It must be light-proof while maximising airflow to maintain the ambient vineyard
atmosphere within the box. In addition, the design had to enclose the bunch without
impacting upon grapevine physiology and, being in a commercial vineyard, needed to be
sufficiently unobtrusive yet robust, that normal viticultural operations such as spraying
29 and tri ing eould eontinue. The chosen material was polyprooylene sheet (0.6 mm) to make a lighrweight box that could hang in the canopy without damaging the vines, which would produce a wound response (vlehdy & Lamb, 1987; vogt et al., 1994; Dixon
&'Paiva, 1995). The box was designed such that it could be assembled easily, but pack flat for storage and transportation. Application of the box enclosed a bunch upon the cane as the box was closed, then the box was attached to a neighbouring cane to support some of the weight and prevent excessive movement in high wind (Figure 2.2). The hanging points also acted as a closing mechanism for the box. The box design employed a darkroom-door type arrangement on each side with the inside of the box painted black to prevent reflected light entering the box (Figure 2.2). The outside of the box was white to maximise reflection and prevent black-body heating (Smart & Sinclair, 1976; Jones, teez).
Initially two prototype boxes were made from polypropylene sheet, one black, the other white. A trial to assess the temperature inside both boxes was set up in the Waite
A B \ I \
L
D:^,,-^ 11 n^^:^- ^a--r Ílgittc ¿.¿. Lrestgn anc^-r assemb'ly^^^^-Lr-. oi polypropyleits i-rr.rxcs uscrl ir-r shade grape beny bunches. A) Opened box as for application to the vine showing foam padding around slots to prevent cane damage. B) Closed box as applied to the vine with tabs for securing the box to the cordon or trellis visible at top.
30 Temperature'C Temperature oC J N C¡) J N) C¡) @ O) Þ N @ 6) À N) (/lÊÈ gtrËË 10:30:00 æ: äF ,,ts.õBÉ 22:30:00 EgEä rã69=(AFr fiÊ)ÞoôE E q F* ãräx 10:30:00 aõ=x'Ê 5 -(D 8gF 22:30:00 ÃØ F BÈ F EãË; =oo(DllF+) € ï öiã 10:30:00 Flê $ 3Ë *ë 22:30:00 v)F q ='3. H F ië $ å å f E 10:30:00 o-(D 3 q I ã'ã ã ôÞË. Fg S: Þ d5 22:30:00 +o:ef Ëa î it ÈJ v, #Ë F; f Ê 10:30:00 (Dãut Eqõ-ÈA 309: 22:30;00 îEáco+ É - î F ã Põìå(t)ô o- 10:30:00 8o& qq ÉE É.E '=< s= f öPäg 22:30:00 ¡u)(D E6ã.gd 10:30:00 õãn'8ç E E'qEo I F.t 'ûao5(D rí5I 22:30:00 {8.$vL. SqBgv)'À E rs 3 ''r õ- 10:30:00 (Dé^é Eåãolfgin IÉ Ë ã AË' =!(DË õÞ__ 10:30:00 üÈrgÉrqqfÐ-oe e¡!"ä Ëç F HØJ åtoã 22:30:00 RÑË gs qä 8- dã9 !ø þAo(AF¡ 10:30:00 ãv)(ld H8 ì/Fõ4.t, f s'Ë P '(D ;x3. 22:30:00 g åx.s I 10:30:00 u,6'g ;<Êä # €H]E 22:30:00 Ë: ã )I =.sf 10:30:00 À¡lJ(D e;---¡ç F s: g (JJÊ Aô(Dã ËFë(l 22:30:00 g t"ir =(D õü s 10:30:00 Ëä å 4P€ K g É o. (D (rt *ÊeE r 22:30:00 ]P oé6FÐ s (Da)Fl oä 10:30:00 o¡vFt{Fg É ='Þ ítÉ(D ã(DË oa Éaâ,8==9 22:30:00 ofrd -Þ4.€ :J E.Eä' doa 5(D(D ë(D5ii<é 10:30:00 gäE' otl \od(DÞ--t- -'Ë !f¡ ¿áo \o='À=l(9! 22:30:00 OoiF. O.qrÞ våöd - , .P-uä * 10:30:00 between the Stevenson Screen and the white box (Figure 2.38), with the temperatures virtually the same, except for a slightly lower temperature in the box at night. In addition, the fourth and fifth days showed a slightly higher temperature inside the box.
Meteorological data for these days (04 September and 05 September 1999) showed a front passing through with associated high winds (Bureau of Meteorology data, not presented). This indicates that there was better airflow through the Stevenson Screen than through the light-proof box, although it is important to remember that the Stevenson
Screen is not light-proof.
Nevertheless, perfect airflow was not achieved as the box afforded some protection from the wind resulting slightly higher temperature inside the box. Between the box and the outside air temperature, the box was cooler at night and warmer in high wind (Figure
2.3C). There was a marked difference in temperature between the box made from black polypropylene and that made from white polypropylene with the inside painted black
(Figure 2.3D). The results of these trials, indicated that the white box would be suitable for the purpose of this investigation as the temperature regime was similar to that of the ambient air temperature and performed similarly to a Stevenson Screen.
Airflow through the box is important not only for rapid temperature equilibration between the interior of the box and the ambient atmosphere, but also for maintaining the same relative humidity inside and outside of the box. Thus, relative humidify was also logged with two TinyTag Plu{M relative humidity/temperature loggers (Hastings Data
Loggers, Australia) to determine whether the shading treatment would impact upon this aspect of bunch microclimate. Readings were made at 5 minute intervals and the hourly average was calculated. Relative humidity readings in and out of the box for a month were compared using Students T-test. While there were differences in relative humidity,
32 â. l, 11 Relative Humidity (%) Ê?E lsû)s('ìo)\¡@ (oo (e t'(D o oooooooo oo +é ÍB's Ër oa å l.) iÉ8 f Ëã s 9ã¡ 00:58:59 05:58:59 €o Fpi';3 iä .ã'õ å. 10:58:59 ãrFí3 E<á 'o55 15:58:59 åÊ rEHFå ;-À) B3E+ä+; E = 20:58:59 õ o.=ä. i¡. 5 FËe 01:58:59 e ã 2 t*
In addition, transmission of light through the box material was determined using a
ShimadzurM (Japan) spectrophotometer. Transmission through the white polypropylene sheet with one side painted black was less than 0.1% between wavelengths of 220 and
800 nm, less than 0.01% below 400 nrn
Furthermore, as a physiological measure of the light environment inside the boxes, chlorophyll in both shaded and exposed fruit was measured. Chlorophyll synthesis is light induced and in the absence of light no synthesis occurs, thus shaded fruit would have less chlorophyll (Zucker, 1972;Raven et al., 1992). Chlorophyll was measured by the method of Harborne (1998). Briefly, chlorophyll was extracted in 80% acetone from developing inflorescences and grape berry and skin throughout the collection series (1999-2000 and
2000-2001), from exposed and shaded fruit. The absorbance of the extract was recorded at646 and663 nm and total chlorophyll was determined according to the function:
Total chlorophyll (mg/L) : I7.3&¿,a + 7.18,\o¡
2,1.2. Timing and Application of Treatments
In the first season (1999-2000), two shading treatments were trialed. The first shading experiment applied 56 boxes to inflorescences at the onset of cap-fall to the north and south sides of two rows. The second experiment applied 21 boxes after fruit-set. Boxes applied to canes can be seen in Figure 2.5.
Based on data collected during the 1999-2000 season, two shading experiments were conducted during the second experimental season. The first shading treatment in 2000-
34 Figure 2.5. Light-proofboxes used to shade bunches of Yitis vinifera L. cv. Shiraz berries in the freld at the time of application one week post-flowering during the 1999-2000 season in Willunga, S.A.
2001 applied 45 boxes to buds in two rows just prior to bud-burst so that fruit would develop from inflorescences that had not been exposed to light. When the developing canes \¡/ere beginning to fill the boxes, and adjacent exposed shoots were clearly showing inflorescences, the boxes were opened at night and repositioned to enclose only the
inflorescence. The remainder of the cane developed in light. The second shading
treatment during the 2000-2001 season repeated that of the first season on a larger-scale
with72 boxes applied across two rows as inflorescences became apparent.
During the 2001-2002 season, a single large-scale shading experiment was conducted to
enable winemaking from the shaded and exposed fuit. Based on the results of the trials
35 conducted in 1999-2000 and 2000-2001, shading was applied to 425 inflorescences prior to flowering, in the same block as previously. A two-panel and two-row buffer was employed for all experiments to avoid end effects. Boxes were applied to both the north and south sides of rows. Boxes were examined regularly and where cane growth resulted in bunches emerging from boxes, these were discarded. The winemaking component of this research comprises a separate research project (R. Ristic, University of Adelaide) and will not be considered here.
2.1.3. Sampling of Grape Berries
Tissues of Vitis vinifera L. cultivars were collected from Shiraz and Chardonnay vines glown on modified Scott-Henry trellises in a commercial vineyard in Willunga, South
Australia (34 46' south, 138" 32' east). The region has a maritime climate with a mean daily temperature of 23"C during berry development (Bureau of Meteorology data, not presented). Sampling dates for Chardonnay and Shiraz over three seasons are presented in
Table 2.1. Shiraz samples were collected at weekly intervals from flowering until harvest during the 1999-2000 and 2001-2002 seasons and from bud-burst until harvest in 2000-
2001. Chardonnay fruit was collected from flowering to harvest during the 1999-2000 season. Flowering was defined as the time when approximately 50% of the opercula had fallen from the flowers (50% cap-fall).
Bunches were collected randomly from 180 Shiraz vines spread across six rows and ten panels of a single block and from 72 Chardonnay vines across four rows and six panels.
A minimum of twenty bunches were collected on each date and the berries removed and bulked. A sub-sample of approximately 100 berries was taken for measures of grape
36 berry development. The remaining berries were immediately frozen in liquid nitrogen then stored at -80'C for later analysis of metabolites and RNA extraction.
During the first season (1999-2000), seeds and whole berries of both Shiraz and
Chardonnay were collected. In the following season (2000-2001), only Shiraz tissues were collected and berries were dissected into seeds and skin in the field, on several occasions during berry development, grape flesh was also collected. In the third and final season (2001-2002), only Shiraz skin was collected during berry development with a sample of seeds made at harvest.
Sampling of grape tissues from shading experiments was similar to that of the exposed fruit. At each sampling date during the 1999-2000 and 2000-2001 seasons (Table 2.2),5-
6 boxes were opened and bunches removed for analysis. At harvest all remaining boxes were opened and their contents removed for analysis. In the final season (2001-2002), to retain the maximum number of shaded bunches for winemaking, sampling was limited to removing single berries from shaded and exposed bunches. Exposed bunches sampled on these occasions had been tagged at the time of applying the shading treatment. Boxes applied to bunches in the canopy and øgged control bunches can be clearþ seen in
Figure 2.2.Between 50 and 100 benies were collected on each sampling date. Seeds and flesh were separated from the skin of the berry by rolling between thumb and forefinger to disrupt cell wall integrity in the flesh of the berry and the contents expelled by
squeezing. At harvest, the fruit was distributed into five replicates for winemaking and
each replicate was sampled for later analysis of metabolites and RNA extraction.
37 Table 2'l Grape berry sampling of Chardonnay 1999-2000 and Shiraz over th¡ee seasons 1999-2000, 2000- 2001 and 200l-2002
Weeks Post- 1999-2000 2000-2001 2001-2002 1999-2000 Veraison Shiraz Shiraz Shiraz Chardonnay -21 17-Aug-00 -20 -19 31-Aug-00 -18 -17 12-Sep-00 -16 -15 26-Sep-00 -14 05-Oct-O0 26-Oct-01 -13 12-Oct-00 -12 13-Oct-99 19-Oct-00 -11.5 23-Oct-O0 -11 21-Oct-99 27-Oct-00 16-Nov-01 -10 28-Oct-99 02-Nov-00 23-Nov-O11 13-Oct-99 -9 04-Nov-99 r 09-Nov-00 21-Oct-99 t -8 12-Nov-99 16-Nov-00t 07-Dec-01 28-Oct-99 -7 18-Nov-99 23-Nov-00 04-Nov-99 -6.5 30-Nov-00 -6 25-Nov-99 30-Nov-00 21-Dec-01 12-Nov-99 -5 02-Dec-99 07-Dec-00 18-Nov-99 4 10-Dec-99 14-Dec-00 04-Jan-02 25-Nov-99 -3 17-Dec-99 21-Dec-00 02-Dec-99 -2 23-Dec-99 28-Dec-00 19-Jan-02 10-Dec-99 -1 30-Dec-99 04-Jan-01 25-Jan-02 17-Dec-99 t 0 06-Jan-00 11-Jan-01 r 31-Jan-02 t 23-Dec-99 t 1 13-Jan-00 18-Jan-01 08-Feb-02 30-Dec-99 2 19-Jan-00 25-Jan-01 15-Feb-02 06-Jan-00 3 27-Jan-00 01-Feb-01 22-Feb-02 13-Jan-00 4 03-Feb-00 08-Feb-01 28-Feb-02 19-Jan-00 5 10-Feb-00 15-Feb-01 08-Mar-02 26-Jan-00 b 17-Feb-00 22-Feb-01 15-Mar-Q2 ø 03-Feb-00 ø 7 24-Feb-00 ø 01-Mar-01 I 08-Mar-01 9 13-Mar-01 ø Flowering; Veraison; " Haryest
38 Table2.2 Sampling of Shiraz berry shading experiments over three seasons L999-2000,2000-2001 and 2001-2002, with two shading treatments applied in both I 999-2000 and 2000-2001 .
Weeks Post- 1999-2000 1999-2000 2000-2001 2000-2001 2001-2002 Veraison Experiment 1 Experiment 2 Experiment 1 Experiment 2 -19 31-Aug-00s -18 -17 -16 -15 26-Sep-00 -14 26-Oct-01$ -13 19-Oct-00$ -12 -11.5 23-Oct-00 -11 27-Oct-00 -10 -9 t r t -8 12-Nov-99 $ -7 23-Nov-00 23-Nov-00 -6 25-Nov-99 21-Dec-01 -5 4 10-Dec-99 10-Dec-99 $ -3 21-Dec-00 -2 23-Dec-99 19-Jan-02 -1 30-Dec-99 04-Jan-01 0 06-Jan-001 + 31-Jan-021
1 18-Jan-01 2 19-Jan-00 19-Jan-00 15-Feb-02 3 01-Feb-01 4 03-Feb-00 28-Feb-02 5 10-Feb-00 15-Feb-01 08-Mar-02 6 15-Mar-02 " 7 24-Feb-00 24-Feb-00 01-Mar-01 I I 13-Mar-01 ' 13-Mar-01 ' * " Shading treatment applied; ' Flowering; Veraison; " Harvest
Note: In 1999-2000, veraison in Experiment 1 (shading treatment applied one week post-flowering, -9V) was delayed by one week in the shaded fruit
39 2.1.4. Analysis of Berry Development
During the 1999-2000 and 2000-2001 seasons, measurements of berry deformability and total soluble solids ('Brix) were made in the field. Deformability was determined using skin-fold callipers applied to berries prior to their removal from the bunch Çoombe,
1992). Each berry thus measured was then removed and tot¿l soluble solids fBrix) for that berry measured by hand-held refractometer. A sub-sample of 100 berries was taken back to the lab for dissection and measurement of seed and skin weight, seed number and bulk weight of berries. The weight of each component was determined by dividing the bulk weight of seeds or skin by berry number and for seeds also by seed number.
At the end of the first season (1999-2000), pH and titratable acidity were also determined as the fruit approached harvest. This was considered particularly important during that season as a pronounced occurrence of"hen and chicken" (Stannard et al.,1974; Cholet et al., 1998) resulted in many small, under-ripe berries at harvest. This phenomenon was not observed during the two successive seasons. Titratable acidity and pH were determined by the methods outlined in Iland et al. (1993).
During the 2001-2002 season, Shiraz skin was collected in the field and frozen immediately in liquid nitrogen. The flesh and seed portion was retained and this was used to determine total soluble solids ('Brix). Because of this approach, no standard error for
"Brix was available for the 2001-2002 data, except for the harvest sample. Bulk berry weight and skin weight were determined in the field. Seeds were only collected at harvest during the 2001-2002 season and at that time seed weight and seed number were determined. Berry deformability was not measured during the 2001-2002 season.
40 Immediate dissection and freezing of samples was important to limit changes due to sampling and transportation, particularly for subsequent RNA extractions. Snap freezing minimises RNAse activity (Sambrook et al., 1989) in the excised berry. Furthermore, immediate fteezing prevents expression artefacts caused from sampling, such as gene expression related to wound response (Mehdy &Lamb,1987; Vogt et al., 1994; Dixon &
Paiva, 1995).
Berry weight, skin weight and seed weight were determined by measuring bulk weight of
50-100 berries divided by berry or seed number and thus no statistical analysis was possible. Total soluble solids ('Brix) and deformability we determined for 50-100 individual berries and thus st¿tistical comparisons were possible. Figures representing these data are shown with standard error bars, often these are smaller than the symbols for each data point.
The tendency of many workers towards analysis of freeze-dried samples has been resisted here. It was considered that the main beneficiaries of this research would be grapegrowers and winemakers, who deal with fresh fruit. Trying to relate data from dry weight back to fresh weight can at best be irksome, otherwise, prone to error. All measurements are expressed here as mg/g fresh weight of tissue or mg per whole fresh berry.
In addition, much research on grape berry development and the changes in metabolites
during grape berry development has, in the past, been presented with reference to the date
each sample was taken. The reason for this is uncertain given that the dates for flowering,
veraison and harvest are highly variable between sites and between seasons at the same
site (as shown below, Section 2.3.11). This makes sample dates an arbitrary scale of little
4l practical value, an observation that becomes painfully obvious when referring to studies conducted in the Northern Hemisphere. A reference point in berry development that is meaningful to researcher, grapegrower and winemaker alike is veraison, thus all samples analysed here have been referred to as weeks pre-veraison or post-veraison.
Abbreviations such as +6V or -3V have been used to denote six weeks post-veraison or three weeks pre-veraison respectively. Actual sampling dates are presented in Table 2.1.
Veraison is the term used by viticulturalists to denote the onset of ripening. Historically, this has referred only to colour change, although it is now commonly used to refer to berry softening, resumption of growth, sugar accumulation, decreases in titratable acidity, increasing pH and loss of chlorophyll Çoombe, 1973; Mullins et at., 1992). There appears no clear pattern to the order in which these process occur, although berry softening, deformability, has been suggested as a useful objective measure Çoombe,
1973; Coombe & Hale, 1973). While sugar accumulation, softening, colouration and increasing size collectively constitute veraison, it is sugar accumulation in the flesh of the berry that increases most dramatically and is most readily measured, which is the best indication of the onset of ripening Çoombe, 1989; Coombe & Mccarthy, 2000). In comparison, colour development can spread across several weeks @irie, 1977; Pirie &
Mullins, 1977), with small changes in anthocyanin concentration not readily detected by the naked eye, particularly when masked by chlorophyll pigments in the berry. Here we have used sugar accumulation in the berry to determine veraison. This determination was made based on total soluble solids (Brix) measured in 50 individual berries and the average Total Soluble Solids (Brix) calculated. Veraison (0V) was then defined as the sample prior to an increase in Total Soluble Solids ("Brix), having established base level prior to veraison.
42 2.1.5. Environmental Differences between Seasons
In order to estimate the environmental differences between experimental seasons, temperature in the vineyard was logged from before flowering until harvest over the three experimental seasons, 1999-2000,2000-2001 and200l-2002 using a TinyTagrM data- logger (Hastings Data-Loggers, Australia). The temperature logger was suspended in the canopy and upon application of shading treatments, a temperature probe was attached to the dataJogger, which was inserted into a box adjacent to the logger. Temperature recordings were made at 15 minute intervals throughout the season and these data were used to calculate the mean daily temperature as well as the daily minimum and maximum, both in the canopy and inside the box.
2.2. Results
2.2.1. Chardonnay Berry Development 1999-2000
in Sampling of Chardonnay 1.40
1.20 1999-2000 commenced prior to o 't.00 E .9 0.80 flowering, which occurred o =È 0.60 6o around 21 October 1999 (-9V), 0.40 0.20 and continued at weekly 0.00 -10-9 -8 -7 -6 ó -4 -3 -2 -1 0 1 2 3 4 5 6 Weeks Post-Veraison intervals until harvest. Berry Figure 2.6. Chardonnay berry weight (g) during 1999-2000 weight increased in a double season, Willunga S.A.
sigmoidal pattern considered typical (Cawthorn & Morris, 1982) of grape berry
development (Figure 2.6).Berry weight increased slowly from flowering through fruit-set
(8V) and then more rapidly until one week pre-veraison. A brief lag phase was observed
43 for a week, whereafter berry growth recoÍrmenced. During this phase of growth, berry weight increased for three weeks from veraison, then remained static for the following two weeks followed by a decline in the final week until harvest.
The pattern of sugar 25 accumulation (Brix) in the 20
'15 'tr berry followed a similar c0 10 pattern to berry weight in 5 the period post-veraison 0 -10 € -8 -7 -6 { -4 -3 -2 -'t 0 1 2 3 4 5 6 Weeks Post-Veraison (Figure 2.7). Veraison, Figure 2.7. Changes in total soluble solids (oBrix) for Chardonnay defined as the onset of berries, 1999-2000, with standard errors (n:50). ripening, is usually determined by an increase in sugar in the berry. Here we have designated veraison as the week prior to an increase in sugar concentration. This phase of sugar accumulation closely matched the second phase of berry growth from veraison to three weeks post-veraison. As with berry weight, sugar accumulation then remained stable and as berry weight decreased in the last week total soluble solids increased, suggesting water loss from the berry.
As the berry begins to ripen, '1.6 numerous processes within the 1.4 ;É 1.2 Ë 1.0 berry are responsible for the ¡6 0.8 eE too breaking down of the cell walls 0.4 (oh 0.2 in the flesh (Coombe, 1976; 00 10-9 € -7 -6 -5 -4 -3 -2 -1 0 I 2 3 4 5 6 Nunan et al., 1998). Thus, the Weeks Post-Veraison
Figure 2.8. Changes in Chardoruray berry deformability during berry begins to soften and this 1999-2000 season with standard errors (n:50).
4 can be measured as the deformability of the berry (Coombe & Bishop, 1980). In
Chardonnay berries during 1999-2000, berry deformability was seen to increase post- veraison towards harvest (Figure 2.8). While deformability is a strong indication of ripening, it is highly susceptible to irrigation and rainfall events increasing berry turgor, shown as a decrease in deformability four weeks post-veraison (Figure 2.8).
were Chardonnay seeds 0.08
0.07 collected the time from they ç 0.06 Ë 0.05 were easily separated from the ã 0.04 : 0.03 0.02 remainder of the berry under ,H 0.01
0.00 freld conditions, during the -10-9 -8 -7 6 ó -4 -3 -2 -1 0 1 2 3 4 5 6
Weeks Post-Veraison 1999-2000 season this was Figure 2.9. Seed weight per berry (g) for chardonnay grapes during the 1999-2000 season' rwillunga s'A' possible at three weeks pre- veraison. From this time, seed weight per berry increased over the following two weeks, thereafter seed weight per berry remained relatively stable (Figure 2.9). At four weeks post-veraison an increase \ryas observed in seed weight followed by a marked decline in the final week before harvest. The increase observed late in berry development when the seed had hardened off is difficult to rationalise other than as sampling variability. The most likely explanation being that slightly larger bunches or berries were inadvertently sampled at this point. Larger berries having been shown to have more, and/or larger seeds
(Cawthorn & Morris, 1982; Boulton et a1.,1998). However, seed number in Chardonnay was stable (1.49310.031 seeds per berry, n:9 weeks), and berry size remained constant
during this late increase in seed weight (Figure 2.6). These data indicate that the main
period of seed growth in Chardonnay occurred prior to the first sampling at three weeks
pre-veraison.
45 2.2.2. Shiraz Berry Development 1999-2000
Changes in Shiraz berry weight during the 1999-2000 season 1.6 1.4
1.2 (Figure 2.10) showed the 6 ; r.o b) ou sigmoidal I classic double 0.6 óñ 04 pattern of development o2 0.0 't 2 3 4 s 6 7 (Cawthom &, Morris, 1982; w""*"r1",_u1.,1""0
berry (g) Coombe, 1992; Mullins et al., irtrå3¡tl*îiä:ï.î*iraz weight during the 1992). Berry weight increased from flowering (9V) until two weeks pre-veraison, followed by a lag until one week pre-veraison then berry weight increased towards harvest. A decline in berry weight was observed from four weeks post-veraison as the berry began to shrivel however a rainfall event immediately prior to harvest resulted in an increase in berry weight at this time.
These changes in berry weight 30
25 \Mere reflected in sugar 20 accumulation (Figure 2.Il). Frs
'10 Total soluble solids increased 5 steadily from veraison towards 0 -12 l1 -10 -0 -8 -7 4 -5 -4 -3 2 1 0 't 2 3 4 5 6 7 Wæks Posl-VeEison harvest. At the time when berry Figure 2.11. Changes in total soluble solids fBrix) during weight began to decrease berry development, I 999-2000 (n:50).
(Figure2.10)"Brix continued to increase, suggesting an increase in concentration due to water loss. Further, at harvest when rain resulted in an increase in berry weight, oBrix declined.
46 Shiraz berry deformability 1.8 't.ô increased with ripening from ts l.il é,, € t.o E veraison towards harvest ä 0.8 ôt è 0.6 b 0.4 (Figure 2.12). The initial @ 0.2 0.0 period of increasing -'t2-11-10 -9 -8 -7 € -5 4 -3 -2 -1 0 I 2 3 4 5 6 7 Weeks Posl-Veraìsn deformability was slow, but Figrre2.l2. Changes in Shiraz berry deformability 1999-2000 with standard errors (n:50). from two until four weeks post-veraison there was a sharp increase in deformability. Following this, deformability remained stable for two weeks, then decreased at harvest consistent with an increase in turgor due to rainfall.
Shiraz seed weight per berry 010
^0.08 1999-2000 increased È in ¡ Þo.oo sharply from the first sample .gl $ o.or
Ø at five weeks pre-veraison for o.o2
0.00 two weeks, reaching 80% of -12-'t't-'to -9 .8 -7 S 6 4 -3 -2 -1 0 I 2 3 4 5 6 7 wssks Post-Vsraison the observed maximum at Figure 2.13. Weight of seeds (g) per Shiraz berry during the 1999-2000 season. three weeks pre-veraison
(Figure 2.13). Seed weight per berry increased slowly from that point reaching its maximum (0.093 g/berry) one week post-veraison, whereafter seed weight per berry began to decline towards harvest. The decline in seed weight per berry was likely the result of decreased weight of individual seeds as the seedcoat lignified and the seed dehydrated (Raven et a1.,1992; Kennedy et a1.,2000b). Seed number per berry in Shiraz
during the 1999-2000 season was 1.538+0.033, calculated from samples of 50-100 berries per week over 13 weeks.
47 During the 1999-2000 season, there was a large proportion of very small berries in
bunches of Shiraz. These small berries, upon dissection, contained only seed traces. Berry
development was retarded with no sign of colouration in these small berries as normal
fruit passed through veraison. However, at around four weeks post-veraison in the normal
berries, the small berries began to colour. At five weeks post-veraison a sample of these
small berries was taken and total soluble solids fBrix), titratable acidity (TA) and pH
was compared with that of the normally developing fruit (Table 2.3). The smaller fruit
had lower sugar (oBrix) and higher levels of acidity, both as lower pH and a higher level
of titratable acidity expressed as tartaric acid equivalents (Table 2.3).
Table 2.3. Comparison of berry weight (g), total soluble solids ("Brix), pH and titratable acidity (TA; glL ârtaric acid equivalents) in normally developing berries and smaller underdeveloped berries of I/itis vinifera cv. Shiraz during I 999-2000 at five weeks post-veraison.
Normal Berries Small Berries Weisht (s) 1.41 0.26 oBrix 22.1 20.7 pH 3.24 2.98 rA (s/L) 9.98 14.10
2.2.3. Shiraz Berries Shaded One Week Post-Flowering 1999-2000
Shading of Shiraz bunches 1.6 -a-Expos8d .-.4 .shadsd I 1.4 one week post-flowering ^o 1.2 Ëo 1.0 delayed veraison one o.e by Ë l. ^ Ë'0.. @ week compared to the o.4 0.2 exposed fruit (Table 2.2). 0.0 -'12-'11 -10 -9 4 -7 -6 -5 -4 -3 -2 _1 0 1 2 3 4 6 6 7 Wesks Post-Veraison Shading also had a Figure 2.14. Effect of shading treahnent applied one week post-flowering noticeable effect on berry on berry weight (g) in 1999-2000
48 weight (Figure 2.14). At veraison, shaded berries weighed less than the exposed fruit as a result of reduced berry growth post-veraison. At harvest, shaded berries weighed 35% less than exposed berries.
30 Despite the difference in #Exposed ..^-.Shaded I I berry weight, shaded n 6rs berries ripened at a similar 10
5 rate to the exposed fruit 0 -12-1't-',t} -9 -8 -7 -6 -5 4 3 -2 -',| 0 I 2 3 4 5 6 7 and reached the same level W€eks Post-V€ra¡son
Figure 2.15. Effect of shading treatment applied at flowering on soluble solids at of total sugar accumulation (oBrix) 1999-2000 with standard errors (n:s0). harvest, around 23.5'Brix
(Figwe 2.t5).
2.2.4. Shiraz Berries Shaded Five Weeks Post-Flowering 1999-2000
Application of the shading 1.ô *Expæed o shaded 2 1.4 post- 1.2 --'o treatment five weeks o --D- '1.0 o flowering in the 1999-2000 3 0.8 Þ 0.6 6o 0.4 season also resulted in reduced 0.2 0.0 berry weight (Figwe 2.16). -12-11-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 W€êks Post-Veraison
However, the impact of shading Figure 2.16. Effect of shading applied at fruit-set on Shiraz berry weisht (g), 1999-2000. on berry weight was less in berries shaded five weeks post-flowering than observed in berries shaded one week post-
flowering (Section 2.3.4). Berries shaded five weeks post-flowering were only around
20%lighter at hawest than the exposed fruit.
49 There was no significant 30 {-Exposed - ô- Shaded 2 25 difference observed in the 20 pattern of accumulation of iE tu l0 total soluble solids fBrix) in 5
0 fruit shaded from five weeks 12-11-10 -9 -8 -7 6 -5 4 -3 -2 -'t 0 I 234567 Weeks Post-Veraison post-flowering and the Figure 2.17. Effect of shading treatment applied five weeks post-flowering on total soluble solids (oBrix), 1999-2000, exposed fruit (Figure 2.I7). with standard errors (n:50)'
Although at harvest, the level of toøl soluble solids fBrix) was slightly higher in the shaded fruit.
2.2.5. Shiraz Berry Development 2000-2001
During the 2000-2001 season, the pattern of Shiraz berry growth, as described by changes in berry weight, did not exhibit 1.6 14 'l 2 the typical sigmoidal pattern of o s 'l.O Èt I o.u development (Figure 2.18) so È o.o d o4 apparent in the previous season o2 00 4 5 6 7 I e (Figure 2.I0). During the 2000- *""*"1o",-lr"olont Figure 2.18. Changes in Shiraz berry weight (g) during the 200I season, berry weight 2000-2001 season, ïúillunga S.A. increased steadily from fruit-set 30
25
(-7V) towards veraison, when a 20
.x ò 15 lag phase pre- at two weeks '10
5 veraison was observed for one 0 -8 -7 -6 -5 -4 -3 -2 -'t 0 t 2 3 4 5 6'789 week. Berry weight then Weeks Post-Veɡson Figure 2.19. Changes in total soluble solids fBrix) in Shiraz increased steadily until five berries during 2000-2001, with standard errors (n:50).
50 weeks post-veraison, after which time berry weight began to decrease. As seen in Figure
2.I9,total soluble solids ('Brix) increased steadily during this time. Sugar accumulation from five weeks post-veraison until harvest appears to be the result of water loss from the ripe berry as berry weight decreased over this period (Figure 2.18).
Berry deformability, being so susceptible to changes in berry turgor, as well as large diurnal fluctuations (Coombe & Bishop, 1980), its value beyond determining the onset of ripening is of little value. While this data was collected, it will not be presented here.
Seed weight per berry in the o.14
o.12
2000-2001 season increased 0.10 E ö 0.08 þ steadily from a week after fruit- I 0.06 r 0.04 set (-6V) for two weeks (Figure 0.02
0.00 -8-? 6 4 4 -3-2-1 0',| 2 3 4 5 6 7 E I 2.20). From four weeks pre- Wæks Pæt-VoÈlsn Figure 2.20. Changes in Shiraz seed weight (g) per berry veraison there was a two week during 2000-2001 lag followed by further increase in seed weight per berry. This level (0.13 g/berry) began to decline steadily from one week post-veraison towards harvest, losing around 35%o of the maximum weight recorded. Seed number during the 2000-2001 season was 2.93+0.04 seeds per berry (n:16 weeks).
2.2.6. Shiraz Berries Shaded Five Weeks Pre-Flowering 2000-2001
In the second season (2000-2001), the difference in berry weight between shaded and
exposed fruit (Figure 2.21) was less noticeable than that observed during the 1999-2000
season (Figure 2.14). The shading treatment was applied four weeks before flowering (-
13V) and samples taken seven, three and one week pre-veraison (Table 2.2). Beforc
5l veraison, berry weight in the '1.6 -tl-Exposed ..'l'.Shaded '1.4 shaded fruit was lower than that 1.2 ^ r.o E9 .9, in the exposed (Figure 2.2I). I 0.8 b 06 t' 6 ¡ However, post-veraison o4 o.2 increases in the berry weight of 0.0 -8 -7 -6-5 -4 3 -2 -,t 0 I 2 3 4 5 6 7 I I Wêêks Post-VeEison shaded fruit resulted in similar Figure 2.27. Effect of shading treatment on berry weight in weights by five weeks post- 2000-2001'treatmentappliedpriortoflowering' veraison. At harvest, berr¡' weights were similar between shaded and exposed fruit
There was also virtually no 30 #Exposed ¡ Shadêd di rence in total soluble solids 25 -a 20 between shaded and exposed i5 ,u
10 fruit throughout ripening during 5 the 2000-2001 season (Figure 0 -8 -7 6 5 -4 -3-2-'t 0 1 2 3 4 5 6 7 I 9 Weeks Post-V€Eison 2.22). This was similar to the Figure 2.22. Effect of shading treatment on sugar accumulation ('Brix) in Shiraz berries shaded prior to trend observed the previous year flowering with standard errors (n:50)
(Figure 2.15).
0.45 +É(pced ¡ Shaded Drrring the 2000-2001 season, 0.40 0.35 O o.so I grape skin was collected from o o.25 ^. o L 3 0.20 t -^ both shaded and exposed fruit Ø 0.'15 { 0.10 ^ 005 (Figure 2.23). Skin weight in the 0.00 -8-76Ã-4-3-2-'t0t23 456789 W€eks Post-Vemison shaded fruit was lower than that Figure 2.23. Effect of shading treafinent on the skin weight of Shiraz berries during 2000-2001. Shading treatment applied recorded for the exposed fruit. prior to flowering.
52 Skin weight was lower prior to 014 *Expæed oj2 ^ ^t t veraison, but increased to o.ro ..-^ ëó t Eo o.oe o levels comparable with those 006 =! .H o.* in the exposed fruit at around 002
0.00 -8 -7 -6 -5 -4 -3 -I*'*. t 4 5 6 7 three to four weeks post- å,ur1",",1"" on Shiraz seed veraison. After that time, skin i'j,'rï ¿,i. iiffïï;1ïiå:rJ""t"-ent weight of the shaded fruit began to decline towards harvest. A similar decline was observed in the skin weight of exposed fruit, but this decrease occurred later (+5V) and was lesser in magnitude. Shaded berry skin decreased around 25%;o between the peak three weeks post-veraison and harvest, while exposed skin decreased around 15% from seven weeks post-veraison until harvest.
During the 2000-2001 season, seeds were collected from shaded as well as the exposed fruit. Seed weight per berry in shaded fruit was slightþ lower than in the exposed fruit before veraison, but there was no difference in seed weight per berry at harvest (Figure
2.24). Around veraison, seed weight per berry was slightly higher in the exposed fruit.
Immediately post-veraison, the weight of seeds per berry in exposed fruit began to decrease and reached a similar level to the slowly increasing seed weight per berry in the shaded fruit. Seed weight per berry continued to decrease until around three weeks post- veraison whereafter seed weight per berry in exposed fruit was more or less constant.
From around one week post-veraison, seed weight per berry in the shaded fruit also began
a slow decline towards veraison, mirroring the values of the exposed fruit in the last few
weeks to veraison (+7V, +9V). Differences in seed weight per berry between treatments
could be due to differences in seed number per berry. While seed number in the shaded
fruit (2.7010.08 seeds per berry, n:7 weeks) was slightly lower than that in the exposed
53 fnlit (2.93+0.04 seeds per berr¡', n:17 weeks), this difference was not significant in a
Student's T-test (p:0.1 I 17, a:0.05).
2.2.7. shiraz Berries shaded at Bud-burst 2000-2001 and comparison of
Shading Treatments
The second shading experiment conducted during the 200Q-2001 season arose from
observations of berry development and flavonoid accumulation made during the lggg-
2000 season. These observations (Chapter 5) prompted the question of what effect would
complete light exclusion throughout shoot, inflorescence and berry development have on
flavonoid composition. Complete light exclusion of the whole cane in the field was
impractical at this time given the constraints of the shade box. Nevertheless, buds were
boxed before bud-burst and later the boxes were rearanged in darkness to enclose only
the inflorescence. Initially, 45 boxes were applied however, practical difficulties greatly
reduced this number. While it was possible to take several small samples during the
growing season (Table 2.2), obtaining a representative sample at harvest rather than a
developmental series was considered more valuable. Nonetheless, at harvest only five
boxes from the original experiment had survived.
At harvest, the weight per berry of the exposed fruit was 0.96 glberry based on a sample
of 100 berries. A similar sample of berries from the shading treatment applied before
flowering showed berry weight to be 0.87 gl berry (n:90), while those from the shading
treatment applied prior to bud-burst weighed 0.95 g/berry (n:100).
The levels of tot¿l soluble solids in the exposed fruit and fruit from each of the shading experiments conducted in 2000-2001 are presented in Table 2.4. At harvest, there was no
54 significant difference in tot¿l soluble solids fBrix) between the exposed fruit and the fruit from the shading treatment applied prior to bud-burst (Experiment 1) using
Student's T-test (p:0.3672, o:0.05). However, there was a significant difference in total soluble solids between the exposed fruit and that from the shading treatment applied prior to flowering (Experiment2) (p:0.0001, cr:0.05). No significant difference was found in the level of total soluble solids ("Brix) between shading treatments (p=0.6243, cr:0.05).
Table2.4. Difference in total soluble solids ("Brix) at harvest between exposed fruit and fruit from two shading experiments, showing standard deviation (SD) and number of berries sampled (n). Experiment 1:.Shading treatment applied to canes prior to bud-burst; Experiment 2: Shading treatment applied to inflorescences prior to flowering (Table 2.2).
Exposed Experiment I Experiment 2 oBrix 29.30 27.83 27.35 SD 2.29 2.39 1.52 n 30 15 50
2.2.8. Shiraz Berry Development 2001-2002
The change in Shiraz berry 16 weight during the 2001-2002 't4 1'2 t' season is shown n Figarc 2.24. t g 0.8 ou In the third experimental season F 04 at the Willunga site (2001- 02 0'0 6 -5 -4 -3 -'*""l,ro"T-u*",1on 3 4 2002), the classic pattern of ' berry growth was much less Figure 2.24. Changes in Shiraz berry weight (g) during the 2001-2002 season, rù/illunga S.A. pronounced than in either of the previous two seasons (Figures 2.10 &,2.18). Berry weight increased slowly from the first
collection two weeks after fruit-set (-8V), which had occurred tr¡¡o weeks after flowering
55 C10V). After veraison, berry weight began to increase more steeply reaching a maximum
(1.37 glberry) four weeks post-veraison. Berry weight then declined slightly towards harvest (1.16 g/berry), appearing at first to drop then rise again. This may have been related to an irrigation event as no rainfall was recorded at this time. This final sample represented a five-fold increase in sample size with five replicates of 50 berries rather than a single sample of 50 berries.
Total soluble solids fBrix) v/ere measured from three weeks after fruit-set during the
2001-2002 season (Figure 2.25). The level remained virtually unchanged before veraison, but increased sharply from veraison towards harvest. As seen in both of the previous seasons (1999-2000, 2000- 30
2001), sugar accumulation in 25
20 the weeks immediately prior to iE,s harvest was coincident with 10
5 decreases in berry weight due 0 -6ó-4-3 -2-101 23456 to water loss. Wseks Post-Veraison
Figure 2.25. Changes in total soluble solids (oBrix) in Shiraz berries during the 2001-2002 season, Willunga S.A (n:50). Apart from berry weight and oBrix, was only skin weight 0.5 monitored throughout berry 04 o E 03 development in the 2001-2002 .9) Io 802 season (Figure 2.26). Skin q 01 weight increased slowly from 00 -6-54-32-'t01 23456 the initial sample towards Weeks Post-Veraison
Figure 2.26. Changes in the skin weight (g) of Shiraz berries, veraison, whereafter skin 200t-2002.
56 weight increased sharply until three weeks post-veraison. Skin weight then remained stable, although an increase was observed in the final week to harvest.
Seeds were collected at harvest from 450 berries for each treatment. Total seed weight per berry at harvest was 0.046+0.002, with 1.78+0.034 seeds per berry, each seed weighing 0.02610.00 I grams.
2.2.9. Shiraz Berries Shaded Pre-Flowering 2001-2002
The effect of shading on berry weight during the third experimental season (2001-2002) is shown in Figure 2.27. There was virtually no difference in berry weight throughout development during the 2001- 1.6 "l'shaded 2002 season with berry weight 1.4 {ÞExposed a|'2 increasing steadily from fruit- ã r.o Ë o.e set towards harvest in both the b o.o d¡ 0.4 exposed and shaded fruit. The o.2 0.0 6 -5 4 -3 -'*"*l"o"lu","L"n 3 4 s 6 increase in the weight of the ' FiCtxe 2'27 ' Effect of shading treatment on Shiraz berry exposed berries between three weight (g) duringthe 2001-2002 season. and four weeks post-veraison 30 BExposed -+shaded was not observed in the shaded 2s
20 fruit and suggests a sampling Ë'u error in the exposed sample. 10
5
Similarly, there was no 0 { -5 4 -3 -2-1012 3456 Weeks Post-V€raison discernible difference in the Figure 2.28. Effect of shading treatment on sugar level of total soluble solids accumulation("Brix),2001-2002(n:50).
57 ("Brix) between the exposed control and shaded berries (Figure 2.28) at any time during berry development in the 2001-2002 season. At harvest, there was no significant difference in total soluble solids between the shaded and exposed it (Table 2.5).
Skin weight throughout berry 0.5 €ÞExposed *shaded development was monitored for 0.4 o c 03 .9 both the exposed and shaded o
=.E 0.2 fruit (Figure 2.29). Towards U) 0.1 harvest in the third season 0.0 €-54-3-2-10123456 (2001-2002) there were slight weeks PosþVeraison atment on the skin weight of differences in skin wei
increased slightly more than that of the exposed fruit, but then declined to a level lower than the exposed fruit. The practice of measuring bulk skin weight rather than individual weights during development, while expedient, precluded statistically testing differences in skin weight. At harvest, the replicates for winemaking purposes were sampled and skin weight, seed weight and total soluble solids determined for both the shaded and exposed fruit. This allowed st¿tistical analysis of the differences observed. At harvest, skin weight in the shaded fruit, while slightly lower, was not significantly different from that of the exposed it (Table 2.5).
At harvest, there was a significant difference in seed weight between the shaded and exposed fruit (Table 2.5). Seed weight per berry was significantly higher (p:0.007) in the shaded fruit than it was in the control bunches, which had been exposed to normal light levels. The difference in seed weight per berry was due partly to the significantly greater
58 weight of each seed þ:0.048), but also a higher seed number per berry in the shaded fruit
(p:0.024).
Table 2.5. Harvest data for exposed and shaded fruit 2001-2002 showing average, standard deviation and standard enors with p-value for Student's T-test for difference between means (a:0.05). All weights in grams (g); number of samples (n) represents number of 50 berry sub-samples from each treatment
Berry oBrix Skin Seed Weight Seed Seed Weiqht Weiqht per Berry Weight Number Exposed Average 1.161 26.356 0.443 0.046 0.026 1.782 SD 0.069 1.109 0.078 0.007 0.003 0.101 SE 0.018 0.370 0.026 0.002 0.001 0.034 n 15 I 9 I I I
Shaded Average 1.101 26.489 0.379 0.066 0.034 1.987 SD 0.129 0.824 0.091 0.013 0.009 0.169 SE 0.033 0.275 0.030 0.004 0.003 0.056 n 15 I 9 I I I T-test 0.122 0.786 0.185 0.007 0.048 0.024
2.2.10. Seasonal Differences in Shiraz Berry Development
Table 2.6 shows the flowering, veraison and harvest dates for Shiraz in a commercial
vineyard in Willunga, SA. Considerable variation was observed in the timing of
flowering over the three seasons examined here (1999-2000, 2000-2001 8.2001-2002).
Data collected from this site during the past decade indicates that flowering usually
occurred in the second or third week of November (Table 2.6) and the period from
flowering to veraison was generally nine weeks. Harvest commonly occurred around 16
weeks post-flowering, although this varied with season and other unrelated factors such
as winery capacity. Thus, 16 weeks post-flowering has been used here to compare
ripeness, as measured by sugar accumulation ('Brix), between seasons.
59 In the first season (1999-2000), flowering occurred earlier than previously reported, the
time from flowering to veraison was nine weeks, thus veraison was also earlier. However, oBrix by 16 weeks post-flowering, indicates that these fruit had reached a similar level of
ripeness to fruit in other seasons at this site (Table 2.6).ln the third season (2001-2002),
flowering occurred relatively late and the time to veraison was longer than in previous oBrix years. However, by 16 weeks post flowering, was the highest reported at that time.
The highest sugar level recorded was 29.3"8nx at harvest in 2000-2001.
Table2.6. Timing of flowering, veraison and harvest over the three experimental seasons of this project as well as the level of total soluble solids ("Brix) at 16 weeks post-flowering and harvest. In addition, the dates of flowering and veraison and total soluble solids at 16 weeks post-flowering between 1993 and 1998 are included for comparison.
oBrix season Flowering weeks to Veraison at 16 weeks weeks to Harvest oBrix at Veraison Post-Flowerino Harvest Harvest 1993-1994 15-Nov-93 9 20-Jan-94 20 31-Mar-94 27.7
1994-1995 23-Nov-94 I 19-Jan-95 24 27-Mar-95 24.0
1995-1996 16-Nov-95 I 20-Jan-96 20 25-Mar-96 23.2
1996-1997 21-Nov-96 I 31-Jan-97 20 29-Mar-97 23.8
1997-1998 1B-Nov-97 9 23-Jan-98 21
1998-1999 28-Jan-99
1999-2000 04-Nov-99 9 06-Jan-00 23.5 7 24-Feb-00 23.5
2000-2001 16-Nov-O0 8 1 1-Jan-01 25.O 9 13-Mar-01 29.3
2001-2002 23-Nov-01 10 31-Jan-02 26.4 6 15-Mar-02 26.4 Lrata collected ty'J to ryyð courtesy oÌ l)r. Ljhns Lravres, Lòu(u rlant lndustry, Adelalde Laboratory
Shiraz berry development as measured by berry weight for the three seasons 1999-2000,
2000-2001 and200l-2002 is shown in Figure 2.30.Berry weight in each season followed
a similar pattern, although in the third season (200I-2002) berrl, weight was lower than in
the two previous seasons.
60 '1.6 +1999-2000 1.4 ---^'-- 2000-2001 1.2 L l. o -lÉ-2001-2002 _c l.o ^' .o ^ I 0.8 0.6 ) Æ t o.4 o.2 A o.o -8-7-6-54-3-2-10123 456789 Weeks Post-Veraison Figure 2.30. Comparison of Shiraz berry weights across three seasons at the same site in Willunga, SA
Despite berry weight being lower in 2001-2002 than in previous seasons, the pattern and level of sugar accumulation fBrix) was virhrally identical (Figure 2.31). The final harvest level of total soluble solids fBrix) in the second season was higher than in other seasons. However, at six weeks post veraison, the harvest date for the third season, the levels were similar across the three seasons. Nevertheless, a Student's T-test of the mean values for total soluble solids at 16 weeks post-flowering showed a significant difference between each season (1999-2000 23.47r0.24"Brix; 2000-2001 24.96+0.23oBrix; 2001-
2002 26.39+0.3 7' Brix).
I 30 4l-1999-2000 25 ..-^--. 2000-2001 ¡. 1(-2001-200.2 20 *' ¡È õts.x
10
5
o -8 -7 -6 -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 I I Weeks Post-Veraison
Figure 2.3 l. Comparison of changes in sugar accumulation ('Brix) in Shiraz grape berries across three seasons, 1999 -2000, 2000-200 I and 200 I -2002 (+SEM; n:50).
6l Shiraz skin, collected during two seasons (2000-2001 and 2001-2002), showed a similar pattern of increasing skin weight. Prior to veraison, both berry weight and skin weight were higher in the 2000-2001 season. Towards harvest, berry weight was slightly higher in 2000-2001, while skin weight was lower.
Shiraz seeds were collected throughout berry development during the 1999-2000 and
2000-2001seasons and at harvest in 2001-2002. Seed weight per berry was lower during the 1999-2000 season than it was during 2000-2001, however at harvest seed weight per berry was lowest in 2001-2002. Despite lower seed weight per berry during 1999-2000, individual seeds weighed more than in 2000-2001, thus the lower seed weight per berry in 2001-2002 was due to smaller rather than fewer seeds. Based on samples of 50 berries per week, average seed number in 2000-2001(2.93+0.043, n:17 weeks) was higher than during the previous season, 1999-2000 (1.5410.033, n:13 weeks), or the following season, 200I-2002 (1.7810.043, n:9 replicates of 50 berries at harvest).
2.2.11. Seasonal Differences in Temperature
Temperature in the vineyard was recorded throughout each of the three experimental seasons. The data logger (IinyTagrM) was attiached to the vine trellis on the first visit to the site each season. The temperature was logged at 15 minute intervals and the readings for each 24 hourr period from midnight were used to determine the mean daily temperature and the daily maximum and minimum. Figure 2.32 shows these data for the
1999-2000 season in Willunga, SA with flowering and veraison indicated by red data points.
62 45 40
35 rl 9.0 Lr I I I Ëru E il .t I ll, ¡l 920 E Prs 10
5
0 qööõõ üq se q 98994 A ÄA A 55 E 5 ee e 4 B ì i\ i ë çq ç Ë q ç* q qT i \ qç ì i Ê Ë É E É Ë E É É Ë Ë Ë Ë Ë E É F F E Ë É F
Fígure2.32. Mean daily temperature in Willunga, SA during the 1999-2000 season, vertical bars show daily mãximum and minimum. Red data points at 04 November and 06 January represent flowering (-9V) and veraison (V) respectively of Shiraz. Flowering in Chardonnay occurred on 2l October 1999 and veraison on 23 l)ecemher 1 999
It is interesting to note that there were several days during the 1999-2000 season when
the daily mean approaches the minimum temperature at which grcpes metabolise, 10oC
(Winkler et a1.,1974; Jackson & Spurling, 1988; Jackson & Lombard,1993; Boulton er
al., 1998). The lowest daily mean was 9.8"C on 08 November 1999, just four days after
flowering. In addition, during the early part of the season, daily minima were often
recorded below 9'C with several of these occurring in the period immediately post-
flowering (Figure 2.32). On a large number of days, the daily maximum exceeded 30"C
above which grape metabolism is also reduced (Coombe, 1987b), with some daily
maxima exceeding 40'C. Post-veraison, there \¡/ere several days when the mean daily
temperature was above 30"C.
In the second season, 2000-2001, sampling commenced with bud-burst in mid-August
(Figure 2.33). At the time of flowering, 16 November 2000, thc mean daily temperature
was around 23'C with daily maxima at that time frequently exceeding 30'C. This was in
63 contrast to the previous season when temperatures were low around flowering. While
daily minima below 9oC were recorded during berry development, the first of these did
season, numerous days had mean temperatures above 30"C with daily maxima frequently
exceeding 40'C.
45 40
35 9.0 9zs E g_ 20 E ,o 15
10
5
0 iåä'F't'F8q q E Ë fi ã Ê ñ Ë Ë $ ËË Ë i Ë fi $ Ë $ $ f Ë Ë ¡ÈË ËËËË ËËËËEEËEEËËFËÊËËËÊEË ==EË Figure 2.33. Main daily temperature in Willunga, SA during the 2000-2001 season, vertical bars show daily maximum and minimum. Red data points at 16 November and ll January represent flowering (-SV) and veraison (V) respectively.
In the third experimental season, 2001-2002, the temperature at the time of flowering, 23
not as low as during 1999-2000 (Figure 2.32). Daily minima were frequently less than
9"C and maxima greater than 30oC, but at no point during the season did the mean daily
temperature fall below 9oC, rarely in fact did it fall below 15'C. FurtherTnore, while the
daily maxima exceeded 40oC on several occasions, the daily mean never exceeded 30"C
and often did not exceed 25"C.
& 45
40 I 35 I I I 930 I lr I ltt¡rl I l,rl tl l I I (EËru g_ 20 l1 E ,o 15 rrlrlt rltrtrlll I rl il 'l' I ì rÌ llrrlll il "l"ll 10 rl 5 0 Ë ZZà ð ZZà ð Zà õI õI õI õI ìE iE6 } f66€ f È È8€€ È È -E õ õñ õ ã ãE E oö o ñ + É É + + ó =É =+ =ÞÊîIìËïÍÌEEEÈEEÈÈrEE E trtrtrtrtrtrtrtrtrF1-FFr-FFFF.e E .¿ ? .e E .e,¿ .e F E .E =E E E E E É F
Figare 2.34. Main daily temperature in Willunga, SA during the 2001-2002 season, error bars show daily maximum and minimum. Red data points at 23 November and 31 January represent flowering (-10V) and veraison (V) respectively.
That 2001-2002 was a milder season than the two previous seasons is clear in Figure
2.35,wherc the mean daily temperature for each of the three seasons has been overlaid.
Throughout the 2001-2002 season (black line, Figure 2.35), daily means were lower than
in the previous seasons. Of the three seasons, the second, 2000-2001 (yellow line), was
35
30
Ozs e-
Lo E20= 0) o- E Êrs
10
5 6 6 -o 6 ó o o o c¡ õ õ õ õ õ o o o oSg gEEñREëHñBÈÉñtÉÉ*RÉsz z z 1 1 õ õ o o ? ? ? ? ? l+ r+ r+ r+ ¿ ¿.
Figure 2.35. Mean daily temperature over three seasons, 1999-2000 (light blue), 2000-2001 (yellow) and 2001- 2002 @lack) at Willunga, SA. Red data points in the early part of each season denote the time of flowering while those mid-season represent veraison.
ó5 the warmest with more hot days both during early berry development and post-veraison, with even the cooler days not recording temperatures as low as those experienced during
1999-2000 a¡d 2000-2001. The first season, 1999-2000 (blue line), recorded more cool days and generally colder days during berry development than either of the two successive seasons.
2.2.12. Treatment Differences in Temperature
Temperature was also recorded in the boxes applied for the shading treatments during the three seasons, 1999-2000, 2000-2001 and 2001-2002. In the first season, 1999-2000
(Figure 2.36), box temperature was measured early in the season to test the box under field conditions. The data logger and a box were placed in the canopy and the temperatures recorded. The box \¡/as removed and the data downloaded approximately one week pre-flowering and then reapplied at flowering (04 November 1999).
During berry development, the temperature difference between the shaded box and in the canopy was approximately 0.5oC, with the box being slightly cooler. Unfortunately, temperature recording inside the box did not commence immediately after the application of boxes at flowering. The fault was detected during the first data transfer two weeks later and rectified.
In the following seasons, 2000-200I and 2001-2002, the temperature inside the boxes was also lower than the ambient canopy temperature. Given the daily temperature rartge, the difference in mean daily temperature between the canopy and the box was not significant, however the difference in mean temperature was tested over three hour periods (n:I2) during the coolest and warmest periods of the 1999-2000 season. On the morning of 23 October 1999 from 3:30 AM there was a significant difference (p<0.0001,
66 35
30 p25
o) L Ë20f oL o. E Ê15
10 '+Exposed -.'- . Shaded
5 cD o, o, o, o, o, o) o, o)q5 o, o, o, o, g) a õo óo oa oa oa oa ao qíÅÉù*€>> Ei E Ei qr E E E qt õt ót ot õ Õ EA ðÉ €+ ö+ ö@ t t 2 2 s s I I I + + 5 s 4 + 4 4 äoc!c\¡òr=Ñöö=Ñõo + : ó ñ ñ ó,¿ ô è I P R N I p Þ ñ
Figure 2.36. The effect of shading treatment on temperature in the field during the 1999-2000 season at Willunga. SA. Note: A preliminary trial commenced 20 September 1999 removed a week prior to flowering (04 November 1999), shading treahnents were applied to berries at flowering, but temperature logging in the shade box experienced technical difhculties.
cx:0.05) in the temperature between the ambient canopy temperature (5.90+0.22'C) and
the temperature inside the box (5.10ú.27"C). Later in the season, on 04 February 2000
between noon and 3:00 PM, the temperature difference between the canopy air
temperature (38.7ú.26'C) and the box (37.7+0.27'C) was also significantly different
þ<0.0001, a:0,05) using Student's T-test.
During both the 2000-2001 and200l-2002 seasons, the maximum temperature difference
between the inside of the box and the ambient temperature in the canopy was less than
1.0'C. This difference was observed consistently throughout the season. Furthermore, this
difference was consistent throughout each day despite large daily temperature ranges as
seen during the 2000-2001 season or during extended hot periods (1999-2000) or cool
periods (2001-2002).
67 These observations prompted a trial of the data logger under a range of temperatures.
These included refrigeration (3.25+0.034'C), ambient indoor temperature (17.91+0.19"C) and ambient outdoor temperature (19.06+0.61'C) each nill over a 48 hour period. In each case there was a significant difference 0<0.00001) between the logger and probe temperafures, which ranged from 0.5oC indoors, to 0.8oC under refrigeration. However, the greatest differences were recorded at high temperatures outdoors when during part of the day the logger and probe were exposed to direct sunlight. The difference between the mean temperafures over two days was around 3'C, while the difference for a given recording ranged from 15oC at the hottest part of the day to virtually zero when the ambient temperature was around l1-13"C and conditions were overcast. Thus, a significant warming effect was observed when the logger was in full sun. Except for those times, the differences between the logger and the thermocouple in the probe rarely exceeded 1.0'C throughout the trial. This trial of the logger and thermocouple probe also showed that the logger was less sensitive to temperature changes than the thermocouple, heating and cooling more slowly than the probe. Consequently, the greatest differences occurred when the temperature was highly variable. Nevertheless a statistically significant difference was obseryeci in the field between the temperature recorded by the data logger and that recorded by the probe located inside the boxes.
2.2.13. Treatment Differences in Chlorophyll AccumulatÍon
Chlorophyll levels of shaded fruit and fruit exposed to normal levels of light were determined for grape berries sampled throughout the 1999-2000 and 2000-2001 seasons.
Figure 2.37 A shows the concentration of total chlorophyll in developing Shiraz berries from the expanding inflorescence (-12V) until harvest (+6V). During the 1999-2000
68 season, chlorophyll concentration ^ 0.05 A o -Exposed whole berries was seen to 0.04 in €o +Shaded E I o.'os decrease as the berry expanded. Eq I nn2 Figure 2.378 shows total = I o.or 'o chlorophyll per berry increasing 0.00 12 -10 -8 €4-20 246 Weeks Poslveraison with berry development. The main
0.50 +Exposed period of chlorophyll synthesis q o) 0.40 E -gShaded 0.30 during berry development occurred E o. e o.20 B -9 from six to three weeks pre- ()E (s 0.10 o veraison, which was coincident F 0.00 -12 -10 -8 46 with the main period of berry Weeks Post-Veraison growth (Figure 2.18). As the berry Figure 2.37. Total chlorophyll levels in shaded fruit (light green line) and fruit exposed to normal light (dark green in began to increase in size, line) in Shiraz berries during the 1999-2000 season V/illunga, SA. A) mg per berry; B) mg per gram fresh weight of berry (+SEM; n:3). chlorophyll per berry increased substantially in the exposed fruit, but only slightly in the shaded fruit. Field observations reflected these results with berries noticeably reduced in chlorophyll pigmentation (Figure 2.38).
During the second experimental
season, chlorophyll levels were Figure 2.38. Shiraz berries pre-veraison showing differences in chlorophyll accumulation between bunches developing again determined in Shiraz berries inside boxes (left) and those in the canopy exposed to normal levels of light (right) from the 1999-2000 soason in Willunga, from Willunga, SA over the main SA.
69 period of berry development (Figure 2.39). In the exposed fruit the total chlorophyll per
berry increased sharply with the initial period of berry growth, up until three weeks pre-
veraison (Figure 2.39A). From one week pre-veraison until two weeks post-veraison,
chlorophyll per berry in the exposed fruit began to decline as the berries increased in size.
From three weeks post-veraison towards harvest chlorophyll again appeared to increase.
In the shaded fruit, the total chlorophyll per berry was very low pre-veraison, but
increased steadily from three weeks post-veraison until harvest. In both the shaded and
exposed fruit, the concentration of chlorophyll per gram fresh weight of berry decreased
with increasing berry size (Figure 2.398).In the shaded fruit the minimum was reached
three weeks pre-veraison, while in the exposed fruit this decline continued until around
three weeks post-veraison. From three weeks post-veraison the chlorophyll concentration
increased towards harvest in both shaded and exposed fruit.
005 0 't2 +Exposed -€FShaded ç E +Exposed A o +lshaded oo4 ô B Ê 009 ê E P oos o ;= E 006 .. 002 P .9 I 003 6 ool cs o 000 000 -7-6-5-43-2-1 0 1 2 3 4 5 6 7 I I 7.6-5-4-3-2-101 234 56 769 Weeks Post-Vorsison Weeks Post-Vera¡son
Figure 2.39. Chlorophyll accumulation in shaded (light green line) and exposed (dark green line) Shiraz berries in willunga, SA during the 2000-2001 season, shown as A) mg per berry; and B) mg per gram fresh weight of berry (+SEM; n:3).
The apparent increase in chlorophyll post-veraison was unexpected. Subsequently, an
opportunity arose to repeat this analysis on similarly heated fruit during the 2002-2003
season. Fruit of the cultivars Chardonnay and Shiraz at a field site in Sunraysia, Victoria
(35" 74' south, 142" 77' east) were shaded using the same methods described above
(Section 2.2.2). The shading treatment was applied around flowering and samples were
70 collected at weekly intervals throughout berry development from the exposed (control) fruit. Samples were collected from the shaded fruit less frequently pre-veraison to preserve the number of boxes available for sampling later in berry development. In
Shiraz, the shading treatment was applied one week before flowering C7V). Boxes were applied to Chardonnay bunches on the same date as Shiraz (29 November 2002), which was the time of flowering for Chardonnay C6Ð.
In Shiraz berries from the 2002-2003 season (Figure 2.40A), chlorophyll was lower in the shaded fruit than in the exposed pre-veraison. The concentration in the exposed fruit approached a level similar level to that in the shaded fruit at veraison and remained the same in the period immediately post-veraison when the berries began to expand (data for berry weight not presented). Post-veraison there was again an apparent increase in chlorophyll concentration in both the shaded and exposed Shiraz fruit commencing from around three weeks post-veraison. In Chardonnay, the level of chlorophyll was also lower
0.14 + Exposed ! A Ø o.12 --+ Shaded Ë o 0.10 o 0.08 E 0.06 o 0.04 -9 oE 0.02 0.00 -54-3-2 -1 01 2 4 5 Weeks Post-Veraison
0.07 +Exposed q B 0.06 +Shaded = E 0.05 g o 0.04 E 0.03 o o.o2 e -9 0.01 o 0.00 -5 4 -3 -2 1 2 3 4 5 w""1r"po"t-vIr"¡ron
Figure 2.40. A). Chlorophyll concentration in shaded and exposed Shiraz berry skins during the 2002-2003 season in Sunraysia, Vic. B). Chlorophyll concentration in shaded and exposed Chardonnay berry skins during the 2002-2003 season in Suruaysia, Vic. (+SEM; n:3).
7l in the shaded fruit pre-veraison. The level of chlorophyll in the exposed fruit decreased towards veraison to reach a level similar to that in the shaded fruit by one week post- veraison (Figure 2.408). Post-veraison, the chlorophyll concentration in both the shaded and exposed fuit of chardonnay continued to decrease towards harvest.
2.3. Discussion
The work described here defines developmental differences in the fruit of Vitis vinifera between the cultivars Shiraz and Chardonnay as well differences in Shiraz berries between successive seasons. In addition, differences in berry development between
Shiraz fruit exposed to normal levels of light within the canopy and artificially shaded fruit over three seasons have been examined and documented.
To achieve the shading treatment for these experiments a polypropylene box was designed to maximise airflow for constant temperature and humidity while excluding light. Measurements of the temperature between the shaded and exposed treatments showed that there was as much as 1.0oC difference in mean daily temperature between the boxes and the ambient temperature in the canopy. For a single reading, differences could be as great as 5.0"C, but rarely exceeded an average of l-2"C over a three hour period at any time during the season, although these differences were significant.
However, the temperature differences between the data logger and its external thermocouple under controlled conditions (Section 2.3.14) were commensurate with differences in field observations between the box and canopy. Thus, it was concluded that there was not a significant temperature eflect created by the shading treatment.
72 Similarly, relative humidity was measured in and outside of the boxes to determine if the shading treatrnent effected humidity. While differences were observed between individual readings, there was no significant difference in relative humidity with shading when these readings were calculated as hourly averages. Thus, it was concluded that the boxes designed for the shading treatment had no significant impact on relative humidity.
While application of the shading treatment did not substantially alter temperature or humidity, the light regime was significantly changed. The box material transmitted less than 0.1%o of light and there was not a significant level of PAR recorded inside the box.
As a physiological test of light exclusion, chlorophyll was measured during berry growth over several seasons, 1999-2000, 2000-2001 and2002-2003. In both shaded and exposed
Shiraz berries chlorophyll was seen to decrease towards veraison, with the initial level of chlorophyll lower in the shaded fruit. This decline in chlorophyll pre-veraison was consistent with increases in berry size during that time and with the conclusions of
Pandey and Farmahan (1977), that berry photosynthate makes a negligible contribution to berry growth. Post-veraison, chlorophyll appeared to increase in both shaded and exposed berries. This increase was unexpected as previous research has shown chlorophyll biosynthesis to be light induced (Zuker, 1972) and that the chlorophyll content of grape berries declines during ripening @ougère-Rifo\ et al., 1995b). This suggests that a process other than de nuovo chlorophyll biosynthesis was occurring post-veraison- It was considered possible that this observation could be attributed to chlorophyll degradation products that also absorb in the wavelengths measured for total chlorophyll. However, the
molar absorbtivities of these compounds are lower than those of the chlorophylls (Joslyn
& Mackinney, 1938; Vernon, 1960; White et al., 1963). Thus, while the full extent of
73 chlorophyll degradation might be masked by the presence of these compounds, their
presence would not account for the apparent increase in chlorophyll post-veraison.
Another possibility that might account for the apparent increase in chlorophyll post-
veraison is the presence of anthocyanins in the skins of Shiraz berries during ripening.
Observation of absorbance spectra of the 80% acetone extract from grape skins show a
broad Aszo Peak in absorbance that extends into the nnge at which chlorophylls absorb,
vis 646 and 663 nm. In addition, a commercial standard of malvidin-3-glucoside
(Extrasynthese, France) also showed absorbance at 646 and 663 nm. Furthennore, when
the chlorophyll level was determined in the skin of shaded and exposed berries from the
Chardonnay cultivar, there was no increase in chlorophyll concentration post-veraison. In
fact, the chlorophyll concentration in both shaded and exposed fruit continued to decrease
post-veraison, even after berry expansion ceased. This confirms chlorophyll degradation
in grape skins post-veraison, and is consistent with recent research on chlorophyll
fluorescence in grapes (During & Davtyan, 2002). Furthermore, the decrease in
chlorophyll in Chardonnay grape skins post-veraison supports the supposition that
chlorophyll degradation products are unlikely to account for the increased absorbance at
646 and 663 nm in Shiraz skin post-veraison. Comparison of absorbance spectra from
Chardonnay and Shiraz extracts post-veraison and the absorbance spectrum of
anthocyanins are convincing indications that the apparent increase in chlorophyll in
Shiraz post-veraison is an artefact of anthocyanin accumulation in the ripening fruit.
Thus, the lower level of chlorophyll in shaded fruit is further evidence that the light
environment inside the boxes was below photosynthetically active levels.
The development of both Chardonnay (1999-2000) and Shiraz berries (1999-2000 ,2000-
2001 and 200I-2002), was seen to follow the classic pattern of berry development
74 (Winkler et al.,1974; Coombe, 1992; Mullins et al., 1992; Coombe & McCarthy' 2000).
Excluding light from developing grape berries had little impact on sugar accumulation.
The exception \ryas a delay in ripening observed with shading at flowering during the first season. Dokoozlian and Kliewer (1996) reported such a delay, however in successive seasons no such delay was observed with sugar accumulation the same between treatments.
The effect of shading on berry development was less than initially anticipated with the greatest difference observed in berry weight during the first season (1999-2000). This result suggested that substantial differences would be observed in successive seasons, however this was not the case. During the first season, berry weight in both the fruit shaded at flowering and at fruit-set was lower than the exposed fruit, also the fruit shaded at flowering was lower in weight than that shaded at fruit-set. Furthermore, in the first season, veraison in the fruit shaded at flowering was delayed by approximately one week with respect to the exposed fruit and the fruit shaded at fruit-set. While shading has previously been reported to delay fruit ripening (Bureau et a1.,2000), the delay in sugar accumulation, or ripening, and reduced berry size observed here was complicated by the incidence of "hen and chicken", a phenomenon that also exhibits these characteristics
(Cholet et al., 1998). During the first season, 1999-2000, the phenomenon known variously as millerandage (Fougère-Rifot et al., 1996) in France, "hen and chicken"
(Stannard et a1.,1974) in Australia or as shot berries elsewhere (Scienza, 1976; Kharjuria
& Bakhshi, 1988), resulting in a high proportion of small under-developed berries, was
observed in Shiraz. The cause of this condition is uncertain, although it has been shown
that low temperature around flowering is a significant contributor to reduced fruit-set and
abnormal ovule development @badi et al.,1995a; Ebadi et a1.,1995b). At the time of
75 flowering in the 1999-2000 season, the temperature in the field was low with several minima below 5.0'C recorded in the weeks either side of flowering (Figure 2.32) and this may have contributed to the phenomenon. The suggestion is that a proportionally greater number of the smaller berries may have been sampled from within the boxes during the
1999-2000 season contributing to the apparent difference in development, particularly in the absence ofsuch an effect in subsequent seasons.
In addition, the observation that shading had no significant effect on berry development is in agreement with that of Haselgrove (1997), which reported no significant difference in sugar between exposed and shaded fruit. These results and previous reports indicate that sugar accumulation in the berry is largely independent of other process in the berry and the vine in general. Sugar levels across cultivars were similar throughout ripening with both Chardonnay and Shiraz berries reaching the same level of ripeness at the time of harvest (around 23.5oBrix, Figure 2.I4). This level was consistent with reported values for the production of table wines (Boulton et al., 1998). Berries of the Chardonnay cultivar weighed less than those of Shiraz and seed weight per berry was also lower in
Chardonnay. This was partly due to smaller seeds (by weight) and to lower seed number per berry. This relationship between seed number and berry weight has previously been documented in grapes Çawthorn & Morris, 1982; Boselli et a1.,1995; Boulton et al.,
1998). Further, Cawthorn and Morris (1982) reported that while increasing seed number resulted in larger fruit, it did not substantially impact upon the accumulation of sugar in the fruit. This observation is consistent with the results presented here and suggests that sugar accumulation is independent of berry size and seed number and that the sink shength of the ripening fruit is considerable. In fact, it has been previously reported that, once begun, sugar accumulation is "massive and undeviating" unaffected even by
76 substantial defoliation, presumably due to mobilisation of carbohydrate reserves from roots or wood (Coombe, 1989).
Differences in berry weight between Shiraz and Chardonnay berries showed that
Chardonnay berries developed slowly at first then more quickly towards veraison. In
Shiraz, the more rapid period of growth occurred earlier in development and slowed towards veraison. This pattern of berry growth may be characteristic of the respective cultivars, or may reflect cooler temperatures earlier in the season (Figure 2.32),
Chardonnay having flowered two weeks before Shiraz. That Shiraz betry growth is initially rapid and slows toward veraison is apparent from a comparison of berry weights over several seasons (Figure 2.30), suggesting that this is indeed the pattern for this cultivar. Increases in berry weight post-veraison exhibited the same pattern in both cultivars. This indicates that although Chardonnay berries ,were expanding more slowly, they accumulated sugar more quickly resulting in the same increase in weight as Shiraz.
Over the three seasons of sampling of Shiraz berries, while the patterns of berry development were similar, differences in development were observed. By weight, berries were larger in the first season (1999-200) than either of the two following seasons with berryweightlowestinthethirdseason(2001-2002).Berrl,weightincreasedmostrapidly in the second seasons (2000-2001) and most slowly in the third season (2001-0220). As postulated with respect to the differences between Shiraz and Chardonnay, this more
rapid increase in berry weight in 2000-2001 was possibly due to warmer temperatures in
the second season (Figure 2.35). The generally lower temperatures during the third
season (2001-2002) coincided with slower increases in berry weight, which flrther
suppoß this notion.
77 Where deformability was measured during berry development in Shiraz and Chardonnay berries it was observed that the initial increase in deformability coincided with veraison, as defined by changes in sugar accumulation. In this regard, deformability is a valuable indicator of ripening, where the timing of changes in other parameters, such as berry weight, are not always synchronised with veraison. Apart from indicating veraison, measures of deformability during ripening are of little value as berry deformability is extremely sensitive to the water status of the vine. In addition, large diurnal fluctuations in grape berry deformability have been reported (Coombe & Bishop, l9S0). Thus, berry deformability post-veraison seems to be a better indicator of turgor pressure than the progress of ripening.
As previously observed by other authors (Cawthorn & Morris, 1982; Boselli et at., 1995;
Boulton et al., 1998) and as described above, there was a strong correlation between berry weight and seed weight and number. In the third experimental season examined here (2001-2002), berry weight was lowest and this corresponded to the lowest individual seed weight and lowest seed weight per berry. Of the first two seasons (1999-2000 and
2000-2001), seed weight was higher in the first year, but seed weight per berry lower as a result of lower seed numbers, nevertheless berry weight was greater. In the second year
(2000-2001) the seeds were smaller than in the first year, but seed number was higher resulting in a higher seed weight per berr¡', yet berry weights were lower. These data indicate that individual seed size has a greater influence on berry weight than seed number. A few large seeds impact more on berry growth than many small seeds.
Seeds increased rapidly in weight from fruit-set towards veraison with as much as 80% of the final seed weight achieved three to four weeks pre-veraison. Seed weight then increased slightly towards veraison and decreased from veraison towards harvest. In the
78 first season (1999-2000), seeds did not reach their maximum weight until one to two weeks later than was observed in the following season (2000-2001) suggesting that season can effect seed development. The decrease towards harvest was consistent with the lignification and dehydration of seeds generally observed with ripening in many species and previously reported in grapevines (Raven et al., 1992; Kennedy et al.,
2000b).
Despite significant differences in berry weight between seasons, the pattern and level of sugar accumulation ("Brix) in each season was similar. This observation lends support to the notion that sugar accumulation in the fruit is independent of other processes occurring in the berry during ripening, particularly other factors affecting berry growth. This is further reinforced by the observation that the timing of the second phase of berry development does not always coincide with the onset of sugar accumulation as seen with
Shiraz berry weight during 1999-2000 (Figures 2.10 and 2.ll).
Interestingly, despite berries being smaller in the third season (2001-2002) skin weight per berry was greater than in the previous season. It has previously been suggested that high light and temperature contribute to thicker skin in grcpes (McDonald et a1.,1998). In
2000-2001 the skin represented 28.3% of berry weight, while in 2001-2002 the skin accounted for 38.2%o of toøl berry weight, these levels were relatively constant throughout the season from fruit-set until harvest. Such a calculation is sensitive to variations in grape water potential affecting berry weight, thus comparisons were also made of fruit at similar weights (=1.15 g) at around 24"Bnx. Percent skin weight, as well
as absolute skin weight, was higher in the 2001-2002 season than in the previous season
(2000-2001). This is inconsistent with the suggestion that high light and temperature
contribute to thicker skins because monthly average temperature and solar radiation
79 levels were both lower in 2001-2002 (Bureau of Meteorology data, not presented). In the second season (2000-2001), shading applied pre-flowering resulted in lower berry weights early in development, but little difference in berry weight or seed weight at harvest. Neither the timing of veraison, nor the accumulation of sugar appeared to be effected by shading during the second season. The only parameter that was noticeably and consistently effected by shading was skin weight, which was lower in the shaded fruit in both the 2000-2001 and 2001-2002 seasons. In the 2000-2001 season, skin from the shaded fruit represented 24.3%o of berry weight, while skin from the exposed fruit accounted for 28.3o/o of berry weight. In the following season, shaded berry skin was
33.9% of berry weight and exposed skin 38.2o/o. These observations are consistent with the report of McDonald et al. (1998) that the skin of exposed fruit was thicker, as well as those of Haselgrove (1997) that skin weight was greater in exposed compared to shaded fruit. While this was notreflected in the meteorological data for 2000-2001 and Z00l-
2002 as described above, the empirical data presents a convincing argument for a strong positive relationship between light exposure and skin weight or thickness.
Here we have shown that grape berry development was largely unaffected by the light environment in the vicinity of the fruit with growth and ripening occurring normally in the absence of light. The main factors affecting berry development were cultivar and season, however shading did result in reduced skin weight. The significance of these observations will be discussed further in relation to flavonoid biosynthesis in Zir¡s vinifera fruit in the absence if light (Chapter 5).
80 Chapter 3. Analysis of Grapevine Tannins
3.1. Introduction
The term tannin, in grapes and wine, describes a range of poþhenolic compounds that
includes flavan-3-ol monomers and proanthocyanidins. Proanthocyanidins are polymers
comprised of flavan-3-ol terminal subunits and flavan-3,4-diol extension subunits as
shown in Figure 1.7 (Goldstein & Swain, 1963; Geissman & Dittmar, 1965; Czochanska
et a1.,1979; Botha et a1.,1982; Delcolü et al., 1983; Haslam, 1998). These compounds
have been reported in the seedcoat and skin of the grape berry @ougère-Rifot et ø1.,
1993; Amrani-Joutei et a1.,1994; Thorngate & Singleton, 1994) and their structures and
composition in grapes and wines has been studied extensively. However, most of these
studies have focused on the flavan-3-ol and proanthocyanidin composition at harvest
ffantz & Singleton, 1990; Prieur et a1.,1994; Escribano-Bailon et a1.,1995; Souquet ef
al., 1996a, 1996b; Fuleki & Ricardo da Silva, 1997; de Freites et al., 1998)' or changes
dwing grape ripening (Czochanska et a\.,1979; Romeyer et al., t986; Katalinic & Males,
1997; Saint-Cricqde Gaulejac et a1.,1997;de Freitas & Glories, 1999; de Freitas et al.,
2000; Kennedye/ al., 2000a).
The flavan-3-ols, catechin, epicatechin and epicatechin-gallate (see Figure 1.4) ate
readily detected in grapes and occur as both free monomers and terminal subunits in the
seeds and skin of the grape berry. In grape seeds, catechin, epicatechin and epicatechin-
gallate all commonly occur as terminal subunits, while in the skin the terminal subunit is
primarily catechin (Romeyer et a1.,1986; Prieur et al., 1994; Escribano-Bailon et al.,
1995; Souquet et al., 1996a). The extension subunits catechin, epicatechin, epicatechin-
8l gallate and epigallocatechin are all detected in grapes, although epigallocatechin extension subunits have only been reported in the grape skin and not in the seedcoat
(Tsai-Su & Singleton,196g; Santos-Buelga et al., 1995; Escribano-Bailonet al., 1995;
Moutounet et a|.,1996; Souquet et al., I996a; Cheynier et al.,1997). Total t¿nnin content is reported to be significantly higher in seeds than skins, although polymer length is generally several-fold lower in the seeds (Prietr et al., 1994; Escribano-Bailon et al.,
1995; Moutounet et a1.,1996; Cheynier et al., 1997).
Previous studies of grape berry tannins have observed that substantial amounts of flavan-
3-ols and proanthocyanidins are present in both the seeds and skins of the berry at veraison (Katalinic & Males, 1997; de Freitas & Glories, 1999; Kennedy et a1.,2000b;
Kennedy et a1.,2001). In many such studies, determination of monomers, dimers and trimers has been made but the polymeric fraction was not analysed (futalinic & Males,
1997; Jordão et al.,200Ia; Mateus et al., 2001). In addition, molecular studies of anthocyanin biosynthesis have shown that genes encoding enzymes required for both anthocyanin and tannin biosynthesis are expressed prior to veraison, when anthocyanins are not synthesised (Boss et al.,I996a). This suggests that the expression of these genes might be related to the accumulation of other flavonoid products, such as flavan-3-ols and proanthocyanidins in the developing fruit.
Tannins have also been reported in other grapevine tissues. Tannins in the internodes of
Vitis vinifera L. cv. Ugni blanc grapevine canes are reportedly highly condensed @arne,
1982). Phenolic content was examined at flowering, veraison and leaf-fall with the highest levels observed at veraison. The lower phenolic content in the mature and lignified stems was likely a result of decreased extractability. In grapevine shoots of the cultivar Carignan, mainly monomers and oligomeric proanthocyanidins have been
82 reported @oukharta et a1.,1988; Boukhafia et al., 1994). The dimers observed included both 4-6 and 4-8 linkages in the dimers Bl to 84 and B7 as well as two trimeric
pîoanthocyanidins. Upon hydrolysis, both catechin and epicatechin appeared as terminal
subunits with the extension subunits primarily epicatechin and epicatechin-gallate
the @oukharta et al., lg94). However, Bourziex et al. (1936) had earlier reported
terminal subunits in stems to be aroundg4% catechin in the cultivar Alicante-Bouschet,
considering the tannin composition of the stems analogous to that of the berry skin. This
was consistent with the observations of Souquet et al. (2000) for the cultivar Merlot, with
condensed tannins in stems around 80% catechin terminal subunits with >15%
epicatechin and -5%o epicatechin-gallate. Extension subunits were about 75o/o epicatechin
with epicatechin-gallate around l7o/o, catechin around 6.5% and the remainder
epigallocatechin (Souquet et a\.,2000). In grape stems, the average polymer length across
anan1e of cultivars was around 8-10 subunits compared to around 10 in seeds and 30 in
skins (Souquet et aL.,2000).
Tannins are also reported in leaves of Vitis vinifera L. with catechin, epicatechin and
epicatechin-gallate detected as well as the dimers B1 to 84, a trimer and tetramer
(Bourziex et al., 1986; Boucheny & Brum-Bousquet, 1990)' The most abundant
compound in grapevine leaves was the 81 dimer (see Figure 1.5), with proportionally
more of the monomers and terminal subunits being catechin. These authors also reported
that the amount of proanthocyanidin material extracted from grapevine leaves decreased
with increasing leaf maturity (Bourziex et a1.,1986; Boucheny & Brum-Bousquet, 1990).
To date there have been no detailed studies of the flavan-3-ol and proanthocyanidin
composition during early berry development and no investigation of both seeds and skins
of the same fruit throughout development and few studies of the proanthocyanidin
83 composition of other grapevine tissues. This project examines the accumulation of flavan-3-ols and proanthocyanidins in vegetative and reproductive tissues throughout the growing season until commerci¿il harvest in the major Australian winegrape cultivars,
Shiraz and Chardonnay.
3.2. Materials and Methods
3.2.1. Sampling of Grape Tissues
Grape berry tannins were analysed in whole berries of Shiraz and Chardonnay, as well as
Shiraz seeds during the first experimental season, 1999-2000.In the second season, 2000-
2001 only Shiraz berries were analysed and these were dissected into seeds and skin and each component analysed separately. Tannins were also analysed in a range of vegetative tissues in Shiraz grapevines collected during the second season. Chardonnay and Shiraz berries were collected as detailed in Section 2.3.3. In the first season (1999-2000) approximately 50 whole berries each of Shiraz and Chardonnay were analysed from each sampling date (Table 2.1). These were frozen in liquid nitrogen in the field and stored at
-80"C until analysed. In addition, seeds were dissected out of the Shiraz berries in the field from fruit-set until harvest during the 1999-2000 season. In the second season
(2000-2001) Shiraz berries were dissected into seeds and skins in the field and each component frozen separately liquid in nitrogen and stored at -80oC until analysed. A a sample of Shiraz berry flesh was also collected after veraison during the 2000-2001 season (01 February 2001; +3V).
Shiraz leaves were collected at five stages of development. The first stage was a newly emerged leaf less than 2 cnf and 0.09 g, densely villous with pigmented leaf margins.
The second Stage was the next oldest leaf after Stage 1, approximately 5-6 cm2 and
84 weighing 0.22 g, becoming glabrous on the dorsal surface of the leaf lobes. Shiraz leaf
Stage 3 was approximately 15 crrl and 0.44 g, with noticeably reduced trichome density and a faint trace of anthocyanin pigment iemaining in leaf epidermis. Stage 4 was an expanded leaf around 40 cr* and 0.99 g with glabrous dorsal, but villous ventral surface with a translucent appearance and no trace of pigmentation. The final leaf stage, Stage 5, was a fully expanded mature leaf approximately 160 cn? in area, weighing 3.62 g and no longer translucent. In addition, tendrils were also collected from Shiraz vines at flowering during the 2000-2001 season (16 November 2000). All tissues were frozen in liquid nitrogen and stored at -80oC until analysed.
3.2.2. Extraction of Flavan-3-ols and Proanthocyanidins
A number of methods have been proposed for the extraction of tannins from plant material. The solvation and diffusion of these compounds from the plant tissues is greatly increased if they are very soluble in the extractant, the concentration in the solvent relatively low and the temperature high. Further, extraction is generally slower from dried samples than from fresh material (Waterman and Mole, 1994). Several factors affecting the yield of tannins from plant material ,üere examined preliminary to the analysis of experimental samples. As discussed above (Section 2.2.3) all analyses here were conducted on fresh samples and to increase extractability all samples were homogenised.
In addition, the influence of solvent, time and temperature on yield were investigated.
A variety of solvents have been previously used to extract phenolic compounds from
plant tissues, these have tended to utilise organic solvents such as etþl acetate and
aqueous diethyl ether @ernandez de Simon et a1.,1990; Pekic et al., 1998), but also
alcoholic solvents such as ethanol (Kantz & Singleton, 1990; Alonso et al., 1991) and
methanol (Cork & Krockenberger, 1991; Constantinides & Fownes, 1994; Sun et al.,
85 1998, Jordão et al., 200Ia).In a survey of different solvents, Kallithraka et al. (1995) showed that methanolic extracts yielded proportionally more monomers, while ethanol tended to extract gallic acid. However, the most efficacious solvent system for the extraction of total tannins appears to be aqueous acetone ranging from 50% to 80% with the most popular composition being 70o/o acetone. This solvent has been used extensively in the analysis of grape proanthocyanidins (Cork & Krockenberger, 1991; Kallithraka et a1.,1995; Kennedy et aL.,2000a; Souquet et aL.,2000; Kennedy et al.,20OI).
Several durations of extraction were trialed as $/ere several combinations of re-extraction and heating or cooling. Extraction for 12 and 24 hours in 70%ó acetone yielded the highest levels of both proanthocyanidins and catechins from all grape tissues. Subsequent extraction of the residue in fresh solvent yielded less than l}Yo of the initial 24 hotx extraction, a third extraction of the same material yielded no detectable tannins.
Extraction for 2 hours yielded around 70-80% of the yield from a 24 hotx extraction. A second extraction for 2 hours of that material yielded approximately l0%o of the yield from the initial extraction as did the third extraction. At this point repeated short extractions were considered too labour intensive and abandoned in favour of a single overnight extraction (15-20 hours). These extractions were carried out in a ten times volume (v/w) of solvent. The minimal yield with repeated extraction of the grape tissue residue after a 24 hotr extraction suggested this to be a suitable volume to achieve a sufficiently steep diffusion gradient and was consistent with volumes reported elsewhere
(Waterman & Mole, 1994).
The effect of temperature on the extraction of tannins from grape seeds was also tested.
Commonly, tannin extracts have been made at room temperature, although both high and low temperature conditions have been utilised (Waterman & Mole, 1994). Extraction at
86 low temperature has been shown to increase the stability of extracted tannins (Cork &
Krockenberger, l99l). Here, the frozen tissue (0.1 g) was extracted in 1.0 nI- of 70%o acetone overnight in a coolroom (4"C),'this ffeatment made no difference to the yield of catechins and proanthocyanidins from grape seeds or skins. Further frozen samples were then heated to 50'C for 20 minutes followed by overnight incubation at room temperature. This method was shown to reduce catechins and proanthocyanidins by as much as 50% and resulted in large variations between samples during the subsequent
HPLC analysis. In light of these observations, further temperature treatments seemed of dubious value.
In addition, light has also been reported to decrease tannin stability (Cork 8L
Krockenberger, 1991) so extractions are commonly made in darkness. Some mixing has also been employed, almost universally, in the extraction of flavonoids and these practices were adopted here, as was the addition of an anti-oxidant in the form of 0.1%
(w/v) sodium ascorbate (Kallithraka et a1.,1995).
Each of these parameters was shown to effect the yield of proanthocyanidins from plant tissues. In addition to these considerations, it was also necessary to be able to compare results arising from this research with those of other workers in the field. For these
reasons, and consistent with the investigations described above, all samples analysed here
were extracted overnightin7}o/o acetone (v/v) plus 0.1olo ascorbate (w/v) on a spinning
wheel operating at 30 revolutions per minute in darkness at room temperature.
3.2.3. Analysis of Tannins
The acetone extract of grape seeds and skin contains both flavan-3-ol free monomers and
the polymeric proanthocyanidins. The concentration of free monomers was determined
87 by HPLC analysis of the acetone extract. Polymeric proanthocyanidins have, in the past, been determined by acid hydrolysis of the proanthocyanidins to yield anthocyanidins these measured by visible absorbance spectroscopy (Swain & Hillis, 1959 Czochanska e/ al., 1980). While this method can differentiate procyanidins from prodelphinidins, its limitation is that both catechin and epicatechin generate cyanidin upon acid hydrolysis, while epigallocatechin yields delphinidin (Swain 8. Hillis, 1959; Butler, 1982;
Czochanska et al., 1980). A less stringent degradation of the proanthocyanidin polymer releases the terminal and extension subunits, however the extension subunits are then present as reactive carbo-cationic species able to reform polymers or undergo separate side reactions (Botha et aL.,1982; Stafford & Lester, 1984; Ricardo ad Silva et al.,I99lb;
Haslam, 1998). Thus, the cleavage of polymers has been conducted in the presence of a strong nucleophile. This has commonly involved acid catalysis with benzyl mercaptan
(Hemingway et al., 1982; Porter, 1988; souquet et al.,l996a; Sun e/ al., r99B). However, use of this thiolysis method has declined in many laboratories because of health and safety issues associated with the use of mercapto derivatives. In addition to the cr-thiol, phloroglucinol is a strong nucleophile that has been employed in this field and a method was recently optimised for use with grape extracts (Kennedy & Jones, 2001). This is the method that has been used here to determine the subunit composition of proanthocyanidins.
Acid-catalysed cleavage of the proanthocyanidins in the acetone extract released the terminal subunits from the proanthocyanidin polymer (Figure 1.7) adding to the pool of monomers. The concentration of terminal subunits was determined by subhacting the concentration of monomers in the uncleaved sample from the monomers in the cleaved sample. Acid-catalysed cleavage of the proanthocyanidin polymer in the presence of
88 excess phloroglucinol (Kennedy & Jones, 2001) replaces the interflavan bond with a phloroglucinol moiety at the 4-position of the C-ring (Figure 1.7), from which the coricentration of extension subunits was determined. The rnean degree of polymerisation
(mDP) was determined from the ratio of the total terminal plus extension subunits to terminal subunits as described by Butler (1982) and Guyot et al' (1999).
Thus, to extract and identiff proanthocyanidins in grape tissues, frozen samples were ground to a fine powder under liquid N and three separate 0.1 g aliquots (replicates) were extracted in darkness in2 mL screw-top Eppendorf tubes for 24 hours withT0o/o acetone (1.0 mL) containing 0.1olo ascorbate to prevent oxidation of polyphenols
(Kallithraka et a1.,1995). Samples were then centrifuged (15 min, 13,000 x g) and trvo
200 ¡tL aliquots of the supernatant of each replicate were transferred to fresh tubes and dried down under vacugm at 35"C for 60 minutes. One of these aliquots was resuspended in 100 pL methanol acidified with 1% HCI then neutralised with 100 pL sodium acetate
(200 mM, pH 7.5) for the analysis of free monomers. The other aliquot underwent acid- catalysed cleavage of the proanthocyanidins in the presence of excess phloroglucinol following the method of Kennedy & Jones (2001). Briefly, the dried sample was resuspended in 100 ¡rL of phloroglucinol buffer (0.25 g ascorbate, 1.25 g phloroglucinol,
215 ¡tL conc. HCl, 25 rnL methanol) and incubated at 50"C for 20 minutes, then neutralised with 100 pL of sodium acetate (200 mM, pH 7.5) and centrifuged (15 min,
13,000 x g).
To determine the amount of proanthocyanidin remaining in the seed and skin residues
after acetone extraction, the residue also underwent acid-catalysed cleavage in the
presence of excess phloroglucinol (Matthews et a1.,1997; Kennedy & Jones, 2001). The
remainder of the supernatant from the acetone extraction was removed and the residue
89 dried under vacuum at 35"C for 60 minutes. The residue was then resuspended in 500 pL of the phloroglucinol reagent, incubated as above, neutralised with 500 pL of sodium
acehte and centrifuged (15 min, 13,000 x g). The supernatant after centrifugation was then used for the determination of flavan-3-ols and proanthocyanidin subunits.
3.2.4. HPLC Method Development & separation of Proanthocyanidins
Historically, chromatography on both paper and cellulose thin layers has been used to determine the composition of flavonoids in plant extracts (Bate-Smith,1954; Markham,
1982; Anderson & Francis, 1985). While good separation and identification have been possible, difficulty has arisen in quantiffing components using these methods. While quantitation has been possible using absorbance spectroscopy, high-performance liquid- chromatography (HPLC) employs both chromatographic separation and spectral determination of compounds and has been used extensively in the study of flavonoids
(williams et al., 1978; Mccloskey & Yengoyan, 1981; Daigle & conkerton, l9g3;
Nagel, 1985;Lazarus et a1.,1999; Peng et a1.,2001).
Numerous published separations using a range of solvents are available and several of these have been trialed here. Attempts at separation using acetic acid with a methanol gradient @oo & Porter, 1983) proved relatively efficacious, but it took some time to develop a suit¿ble method. Initially, a VydacrM (USA) C-18 reversed-phase column
(5pm, 250 mm x 4 mm) was used, however after several months resolution decreased and repeatable results could not be achieved despite extensive efforts varying acid concentration (pH) and type (acetic, phosphoric, formic), temperature conditions, flow rate and mobile phase (methanol, ethanol, acetonitrile). At this point a Merck
LiCrospherrM column (5pm, 250 mm x 4 mm) was borrowed from another research
90 group working on plant polyphenols and showed good potential for separation. A
LiCrospheflM column was purchased and method development recoÍtmenced. A suitable separation for the flavàn-3-ol terminal subunits and the phloroglucinol adducts of the extension subunits, produced during acid-catalysed cleavage of proanthocyanidins, was developed.
However, this method proved unsuitable for the separation of free monomers in the acetone extract as the polymer fraction routinely coeluted with epicatechin-gallate.
Alternate conditions of acid, mobile phase and temperature were tested culminating in a method that employed phosphoric acid with an acetonitrile gradient. Baseline separation of the monomers from the polymer fraction was achieved with the aim of quantiffing the unhydrolysed polymer fraction in the acetone extract. It was envisaged that this would enable an estimation of the percent conversion during acid-catalysed cleavage of the polymer and determination of the fraction resistant to cleavage under the conditions used here. This turned out to be fraught with difficulty as the polymer peak was later determined to contain a significant level of contamination with other phenolic compounds. At this point it was considered that the benefit in knowledge did not justiff the effort to identitr and quantiff all of the components of the polymer fraction as the objective of the project had been to determine flux through the pathway with respect to the expression pattern of genes encoding biosynthetic enzymes observed by Boss et al.
(1996a). Nevertheless, the method developed here was used to identiff and quantiff free monomers in the acetone extract of grapevine tissues.
A 200 pL aliquot each of the acetone extract, the acetone extract after acid-catalysed
cleavage and, where applicable, the acid-cleaved residue of samples, was transferred to
Sunbrokerru 250 ¡rL reduced volume HPLC autosampler vials. Samples were nrn on a
9l Hewlett Packard HP 1100 high-performance liquid-chromatograph (HPLC) using a
Merck (Germany) LicrospherrM c-18 (5pm, 250 mm x 4 mm) anal¡ical column protected by an SGE (Australia) C-18 guard'column.
A separate reversed-phase HPLC method was developed for the cleaved and uncleaved proanthocyanidin samples. Uncleaved samples: solvent A, 0.2o/o (v/v) phosphoric acid, solvent B, 4:l (viv) acetonitnle:0.2%o phosphoric acid (gradient of solvent B: zero min,
0%;5 min,l0o/o;40 min, I0%;55 min,lTYo;65 min, I9%;75 min,19o/o;80 min, 100%;
85 min, I00%;86 min, 0o/o). Cleaved samples: solvent A,0.2yo (vlv) acetic acid, solvent
B, methanol (gradient of solvent B: zero min, lo/o;40 min, r%; r20 mn 30%o; 120.1 min,
100%; 125 min, l00o/o; 126 min,l%). For both methods, 25 ¡rL of sample was injected and run at 25"C with a flow rate of 1 mllminute. Concentrations of free monomers and hydrolysed terminal subunits were determined from standard curves prepared from commercial standards of catechin, epicatechin and epicatechin-gallate obtained from
ExtrasyntheserM (France). The concentration of extension subunit-phloroglucinol adducts was calculated from published molar extinction coefficients (Kennedy & Jones, 2001).
Polymer length was calculated as the mean degree of polymerisation (mDP), whereby terminal and extension subunits were summed and divided by terminal subunits (Butler,
1982; Guyot et a1.,1999).
3.3. Results
3.3.1. Identification of Flavan-3-ols and Proanthocyanidin Subunits
Free monomers from the acetone extract of both seeds and skin from grape berries collected during the second experimental season (2000-2001), as well as whole berries from the first experimental season (1999-2000) were identified by comparison of their
92 c) o €o o at f E 400 NI
o 300 o E, c _Ê. (5 .E o -o E (¡) o o o U' c) .9 -o (ú o O t.rJ
0 l0 30 40 60 80 Time (min)
Figure 3. I . HPLC separation of free monomers, catechin, epicatechin and epicatechin-gallate from Shiraz seeds one week pre-veraison (04 January 2000).
elution time and absorbance spectra with those of commercial standards. Commercial
standards were purchased from ExtrasyntheserM (France) and chromatographed
separately to determine elution times and to generate a library of absorption spectra.
Figure 3.1 shows the separation of an acetone extract from grape seeds collected one
week pre-veraison. The elution times of the free monomers catechin, epicatechin and
epicatechin-gallate were consistent with those of the commercial standards. The peak
eluting at 80 minutes contains polymers plus other phenolic compounds.
Following acid-catalysed cleavage of the proanthocyanidins in the grape seed,
phloroglucinol adducts of the proanthocyanidin extension subunits were also observed in
the HPLC analysis. Figure 3.2 shows peaks corresponding to the commercial standards of
monomers and represent the sum of free monomers and terminal subunits released by
acid-catalysed cleavage of proanthocyanidins. In addition, peaks corresponding to the
phloroglucinol adducts of the extension subunits were also present (Figure 3.2). These
93 o o o J o f Þ o ñ E Þ õ -o D(! (! o) (Jo o I (t c .= ! s L C) (ú o c) o) o / o I o õ .9 .= .o 'õc o- E o- l (J uJ o t¡J =o, (s o () E 'õ_ o o co -c uJ .c È ! o o / o c) / l .E o o Þ ! c E o .o (5 o o) o- -o o LrJ L .tr O o ! ú, o -o o oo
20 40 60 100 120 Time (min)
Figure 3.2. HPLC separation of Shiraz seed proanthocyanidin subunits one week pre-veraison (04 January 2000). Terminal and extension subunits after acid-catalysed cleavage in the presence of excess phloroglucinol. Terminal subunits catechin, epicatechin and epicatechin-gallate and phloroglucinol adducts of extension subunits wcre identified by their retention times with respect to the terminal subunits and by comparison of their absorbance spectra and retention times with those of purified phloroglucinol adducts provided Dr. Jim Kennedy (University of Adelaide; Kennedy &
Jones,2001).
In the separation of both the free monomers and proanthocyanidin subunits, the ascorbate added to the extraction solvent was observed to elute immediately after injection of the sample. In addition, excess phloroglucinol from the acid-catalysed cleavage was also observed to elute early in the separation (Figures 3.1 & 3.2). Atthe end of the separations of both the monomers and proanthocyanidin subunits alarge peak was observed. It was thought that in the uncleaved sample this peak represented the polymeric fraction of the acetone extract, while in the samples following acid-catalysed cleavage this peak represented proanthocyanidins that were resistant to cleavage under the conditions used here. Consistent with this notion was the observation that the 'polymer' peak is
94 substantially larger in the uncleaved sample. Further, the HPLC method of Peng et al.
(2001) utilised a similar solvent system and gradient to that employed here to separate the
free monomers and identified this late eluting peak as cbntaining 'þredominantly
polymeric procyanidins". Extensive discussions on this matter with Dr. Jim Kennedy
concluded that with the method of extraction and sample preparation used here it would
be difficult to conclude that this peak represented only proanthocyanidin polymers. The
main contaminants likely to be present in this fraction were thought to be
hydroxycinnamates and
may represent a substarLtial I zsoo A I zæo d, fraction of this late eluting rsoo (!P 6 I rooo ¡tt peak.
0 'When a number of Wavelength þm) absorbance spectra are 5 B * ,ooo taken through this late í) I 1500 o ¡ peak, apparent þ rooo eluting it is ¡o this does not represent a 0 400 pure compound (Figure Wavelength (nm)
Figure 3.3. Absorbance spectra through the late eluting peaks from 3'3)' Figure 3'34 shows (A) uncleaved acetone extract from Shiraz seeds and @) after acid catalysedcleavageofproanthocyanidins five spectra through this
late eluting peak in the
uncleaved sample. Differences between these spectra suggest several compounds coelute
under the chromatographic conditions at this time. Figure 3.38 shows spectra through the
late eluting peak after acid-catalysed cleavage of proanthocyanidins. Differences between
the spectra in Figure 3.34 and 3.38 suggest that some of this material has been modified
95 o (E € o <õah '6L f : oo) E _9 o ! o @ À () J N f E ! E o Eo o o d) c -q o G -oL () o f c? at, E -o E !o o o) L () (E ! () o J 'õ. 0) E ! uJ ()(U o õ-9 'õ-o, uJ 0
20 40 60 80 't00 120 Time (min)
Figure 3'4. HPLC separation of Shiraz skin proanthocyanidin subunits one week pre-veraison (04 January 2000). Terminal and extension subunits after acid-catalysed cleavage in the presence of excess phloroglucinol. Terminal subunits catechin, epicatechin and epicatechin-gallate and phloroglucinol adducts ofextension subunits
during catalysis of the proanthocyanidins. Similarities between the spectra in this peak before and after hydrolysis support the notion that some of this material may be resistant to cleavage under the conditions employed here.
The chromatographic separation of skin tannin extracts showed a similar pattern to that of the seed extracts with the exception of the acid-cleaved samples. In the cleaved skin samples, phloroglucinol adducts of epigallocatechin were also detected (Figure 3.4). As with the analysis of the seeds. Peaks representing ascorbate and phloroglucinol were observed to elute early in the separation, while a large peak eluting near the end of the mn, may represent those elements of the grape skin polymer that were resistant to acid- catalysed cleavage by the method utilised here.
96 3.1,2. Whole Berry Analysis of Tannins 1999-2000
Initially, it was envisaged that a rapid whole berry analysis of tannins would be a useful tool for industry application. Thus, in the first experimental season, tannins were extracted from homogenised whole berries as described above (Section 3.2.2). Further, one of the aims of this project was to clone and characterise the gene encoding the biosynthetic enzyme leucoanthocyanidin reductase (LAR) that catalyses the reduction of leucoanthocyanidins to flavan-3-ols (Figure 1.6). Thus, the HPLC analysis primarily sought to identiff products of LAR activity. Our current understanding of flavonoid biosynthesis suggested that only the flavan-3-ol free monomers and proanthocyanidin terminal subunits were products of LAR activity, thus the flavan-3-ols identified by
HPLC subsequent to acid-catalysed cleavage in the presence of excess phloroglucinol would represent the sum of flux through this part of the pathway. Establishing the temporal pattern of accumulation of the products of LAR would focus efforts to clone the gene encoding this etlzyme, assuming that this part of the pathway was transcriptionally regulated as appears to be the case with the expression of other genes involved in anthocyanin biosynthesis (Martin & Gerats, 1993; Sparvoli et al., 1994i Boss et al.,
1996a).
Thus, the first analyses of tannins in this project, which were conducted on whole berries
of the Chardonnay cultivar, used only acid-catalysed cleavage to determine the products
of LAR. It readily became apparent, that rather more information regarding the general
nature of tannin biosynthesis, accumulation and composition could be gleaned by
additionally separating free monomers in the acetone extract thereby enabling a
determination of mDP as well as giving an indication of the nuances of polymer
formation during berry development. Furthermore, the distribution of flavan-3-ols
97 between monomers and polymeric terminal subunits has an organoleptic impact on the
final wine, particularly with respect to bitterness and astringency (Horowitz, 1964; Noble,
1984; Glories, 1988; Thorngate, 1997).
Figure 3.5 shows the pattem of total extension subunit and flavan-3-ol accumulation in
Chardonnay berries collected during the 1999-2000 season. On a per berry basis
extension subunits accumulated steadily from flowering (9V) towards veraison reaching
approximately 75o/o of the maximum level two weeks pre-veraison (Figure 3.54). This
was followed by a further increase subsequent to veraison reaching a maximum one to
two weeks post-veraison, then declining towards harvest. When expressed on a per gïam
fresh weight of berry basis, there was an increase in the concentration of extension
subunits immediately following flowering (Figure 3.58) with a decline from around fruit-
set towards harvest, the greatest decrease occurring during the main period of berry
growth (Figure 2.I3). Total flavan-3-ols followed a similar pattern to that of the extension
8.0 È o o ¡ 7.0 o n B o É E 6.0 5.0 15 .E 4.0 ô 4 J I @ l0 Ø 3.0 c .9 .9 2.O c 5 '2 @ o 10 u U 00 0 -10-9-8 -7 -6 -5 4 -3-2 I 0't 2 3 4 5 6 -10-9 €-7 6-5 4 3 -21 0 1 2 3 4 5 6 Weêks Post-Vera¡son Weeks Post-Veraison 3.0 6.0 E c I 2.5 g s.o D o g '9 2.O p 4.o .9o s ¿ 1.5 30 ! o¿ o È 1.0 ,.0 I $ E o.s Ë 1.0 Fo F 0.0 00 -10.9 € -7 -G -6 -4 -3 -2 -1 0 1 2 3 4 5 6 -104 -B -7 -05 -4 3 -2-1 0 1 2 3 4 5 6
Weeks Post-Veraison W6êks Post-Veraison
Figure 3.5. Tannins extracted from Chardoruray whole berries (1999-2000) with SEM (n:3). A) Total extension subunits (mg/berry); B) Total extension subunits (mglg fresh weight of berry); C) Total free monomers plus terminal subunits (mglberry) referred to as total flavan-3-ols; D) Total free monomers plus terminal subunits (mg/g fresh weight of berry)
98 subunits, albeit at a lower level. Flavan-3-ols also reached a maximum per berry one to
two weeks post-veraison followed by a decline (Figure 3.5C) with the concentration of
flavan-3-ols highest around fruit-set (Figure 3.5D).
The pattern of total extension subunit and flavan-3-ol accumulation in Shiraz berries
collected during the 1999-2000 season is shown in Figure 3.6. The overall trend in tannin
accumulation in Shiraz was the same as that observed in the Chardonnay cultivar.
Extension subunits per berry began to increase post-flowering (-9V) reaching a plateau
around four weeks pre-veraison, which was slightly earlier than observed in the
Chardonnay berries. This level remained relatively constant until veraison when a further
increase in extension subunits was observed (Figure 3.64). This level then decreased
8.0 30 7.0 A Qzs B 6.0 g 5.0 P20 E 4.O €15 Ø 3.0 .9 10 2.0
1.0 ñ5 0.0 0 -12-11-10 -9 -8 -7 4 -5 4 -3 -2 -1 0 1 2 3 4 s 6 7 -12-1't-10 -9 -8 -7 { -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 Wgeks Post-V€Elson Weeks Post-Verâlson 3.0 5.0 c 4.5 D 2.5 4.0
2.0 8.o 9 3.0 I 1.5 I 2.5 ¿ 2.O 1.0 à 1.s Í 1.0 0.5 0.5 0.0 0.0 -12-11 -1O -9 4 -7 € -5 .4 -3 -2 -1 O 1 2 3 4 5 6 7 -12-11-10 -9 { -7 -6 -5 4 -3 -2 -1 0 1 2 3 4 5 6 7
Figure 3.6. Tannins extracted from Shiraz whole berries (1999-2000) with SEM (n:3). ,a.). Total extension subunits (--glb"rÐ'); B). Total extension subunits (mgig fresh weight of berry); C). Total free monomers plus terminal subunits (mg/berry); D). Total free monomers plus terminal subunits (mg/g fresh weight of berry)
post-veraison to a level similar to that immediately pre-veraison. This level was
maintained over the following three weeks then decreased towards harvest. The
concentration of extension subunits in Shiraz over the 1999-2000 season showed the
99 same pattern as seen in Chardonnay, a steady increase following flowering reaching a maximum concentration seven weeks pre-veraison then declining rapidly during the first phase of berry growth (Figure 3.68). This level was observed to vary slightly during the remainder of the season, but the overall trend was a decline in extension subunits towards harvest.
Similarly with the level of flavan-3-ols per berry, in Shiraz the pattern of an increase to around four weeks pre-veraison reflected that observed in Chardonnay. However, while
Chardonnay berries continued to steadily accumulate flavan-3-ols from four weeks pre- veraison towards a maximum one week post-veraison, in Shiraz the level was more or less constant until veraison (Figure 3.6C). In the week following veraison, a sharp increase in flavan-3-ols per berry was observed in Shiraz, also reaching a maximum one week post-veraison. This level was relatively stable for the following two weeks then declined towards harvest. In Figure 3.6D the decline in flavan-3-ols per gram fresh weight of berry from seven to three weeks pre-veraison also reflects the first phase of berry growth (Figure 2.13). Post-veraison, the dramatic increase observed on a per berry basis is also observed as an increase in flavan-3-ol concentration. This increase occurs during the second phase of berry growth immediately post-veraison and must represent a period of intense flavan-3-ol synthesis at this time.
3.1.3. TannÍn Composition of Whole Berries 1999-2000
Extension subunit composition in Chardonnay berries throughout berry development was predominantly epicatechin. Around veraison 70o/o of total extension subunits were epicatechin with epicatechin-gallate accounting for a further 25Yo andcatechin making up the remaining SYo of total extension subunits (Figure 3.7A). While analysing samples collected during 1999-2000, epigallocatechin extension subunits were not observed. This
100 was partly due to the absence of a suitable standard, which when rectified indicated that
these compounds were coeluting with other compounds requiring a modification of the
original niethod to separate the epigallocatechin-phloroglucinol adilucts from these non-
tannin contaminants in the following season. In contrast, the composition of flavan-3-ols
in Chardonnay berries showed similar levels of catechin, epicatechin and epicatechin-
gallate throughout the season, although a slight increase in the proportion of epicatechin-
gallate was observed post-veraison (Figure 3.78).
6.0 ò A oL -o 5.0 o) E 4.0 Èat, L f 3.0 -c¡f U) c 2.O .9 at oL x 1.0 IU 0.0 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 't23456 Weeks Post-Veraison 1.2 rCatechin B tr Epicatechin à 1.0 L I Epicatechin-gallate o) -o o, 0.8 E .9, 0.6 (')? cI (ú 0.4
tL-g 0.2
0.0 -10 -9 -8 -7 -6 -5 4-3-2-'10 123456 Weeks Post-Veraison
Figure 3.7. Subunit composition of Chardonnay whole berries (1999-2000) after acid-catalysed cleavage of proanthocyanidins. A). Extension subunit composition (mg/berry) B). Flavan-3-ol cornposition comprising free monomer and terminal subunits (mg/berry)
l0l In berries of the Shiraz cultivar the composition of extension subunits was similar to that
observed in Chardonnay berries, with epicatechin comprising the bulk of extension
subunits (Figure 3.84). Epicirtechin subunits accounted for more than 80% of total
extension subunits at veraison, with epicatechin-gallate extension subunits around l5olo
and catechin less than 5o/o of the total at that time. In contrast to the Chardonnay berries,
the flavan-3-ol composition in Shiraz, while equally catechin, epicatechin and
epicatechin-gallate prior to veraison, exhibited a marked increase in the proportion of
epicatechin post-veraison (Figure 3.SB). While all flavan-3-ols were observed to increase
7.0
6.0 A Eo -o 5.0 ÞE 4.0 Ë5 ! õ 3.0 c o -e 2.0 o x rtr 1.0
0.0 -12-11 -10 -9 4 -7 € -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 Weeks Post-Veraison 1.6 r Catechin 1.4 B tr Epicatechin Þ 't.2 o -o X Epicatechin-gallate ct) 1.0 E .9,o 0.8 (ÐI I (úc 0.6 o tr 0.4 0.2
0.0 -12-11-10 -9 € -7 -6 -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 Weeks Post-Veraison
Figure 3.8. Proanthocyanidin subunit composition of whole Shiraz berries (1999-2000). A). Extension subunit composition (mg/berry); B). Flavan-3-ol composition, includes free monomers and terminal subunits (mg/berry)
102 post-veraison, the increase in epicatechin in the first week post-veraison represented an eight-fold increase compared with a three- to five-fold increase in catechin and epicatechin-gallate in the week following veraison
While the analysis of Chardonnay berries from the 1999-2000 season only identified total flavan-3-ols and extension subunits following the cleavage of proanthocyanidins, free monomer content and composition was analysed in Shiraz berries collected during that season (Figure 3.9).
This analysis showed that the increase in total flavan-3-ols post-veraison was due to an increase in both the terminal subunits and the free monomers (Figures 3.94 & 3.98).
Strikingly, the pattern of free monomer accumulation \Mas virtually identical to that of total flavan-3-ols
1.0 (Figure 3.8). In both 0.9 ^à A ob 0.8 the terminal subunits Þt o.z ã o.o and free monomers, -E o.s ã o.+ E o.s the majority of the E t-b 0.2 0.1 total increase is due 0.0 -12-',t1-10 -9 € -7 6 -5 -4 -3 -2 -'l 0 1 2 3 4 5 6 7 to an increase in Weeks Post-Veraison 0.7 lCatechin epicatechin, ç 0.6 B E trEpicâtechin o å o.s n Epicatechin-gallate although catechin E ó o.+ co) monomers were o.s free oÞ ão, E also seen to increase * 0.1
0.0 substantially post- -12-'11-10 -9 -8 -7 6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 Weeks Post-Veraison (Figure veraison Figure 3.9, Accumulation of terminal subunits released during acid- catalysed cleavage of proanthocyanidins (A); and free monomers (B) 3.9B). What is not present in the acetone extract ofwhole Shiraz berries (1999-2000)
103 apparent from this analysis is the relative contribution of the different tissues in the grape
berry that contain tannins to this overall picture of tannin accumulation and composition
in the whole berry.
3.1.4. Shiraz Seed Tannins 1999-2000
During the 1999-2000 season, seeds 3.0 A Ê were dissected out from Shiraz Ë z.s ¡l o) s2.0 berries and the tannin composition ø '5 t.s -o f was determined from the acetone ã0.6(l) accumulated in two phases, initially E o.¿ oc > 0.2 from fruit-set until three weeks pre- 0.0 -54-3-2-10't234567 veraison, then a further phase Weeks Post-Veraison occ d from one to two weeks pre- Figure 3.10. Total flavan-3-ols and proanthocyanidin subunits in Shiraz seeds (tSEM; n:3), 1999-2000. A. veraison until two weeks post- Extension subunits (mglberry); B. Terminal subunits (mg/berry); C. Free monomers (mg/berry) veraison (Figure 3.104). From two 104 weeks post-veraison until harvest there was a decline of approximately 20o/o of toøl extension subunits. This general pattern early in seed development parallels that of the whole berry (Figure 3.6A), however seed extensiori units do not increase as dramatically post-veraison suggesting this increase occurs in the skin of the berry. Post-veraison the seed extension subunit decrease appears greater than that of the whole berry and this also suggests extension subunits may be synthesised in the skin at this time. The pattem of terminal subunit composition was similar to that of extension subunits, however the initial phase of accumulation represented a smaller increase in subunits and the later phase a proportionally greater increase (Figure 3.108). The highest level of terminal subunits, on a per berry basis, occurred two to three weeks post-veraison, followed by a decline towards harvest. Free monomers in grape seeds in the 1999-2000 season increased slowly from fruit-set until two weeks pre-veraison when the rate of accumulation increased markedly (Figure 3.10C). Free monomers increased to a peak three weeks post-veraison whereafter a steady decline until harvest was observed. When considered together, the increase in terminal subunits and free monomers in the seed immediately post-veraison account for the majority of the increase in flavan-3-ols in the berry at this time. 3.1.5. Shiraz Seed Tannin Composition 1999-2000 Extension subunit composition in Shiraz seeds from the L999-2000 season showed that epicatechin was the main extension subunit (Figure 3.114), which was similar to the pattem observed for the whole berry (Figure 3.84). Extension subunits in the seed represent only 30 to 40Yo of the total in the berry around veraison suggesting polymers in the seed to be shorter than those in the skin. 105 2. 5 ò A 0) -o 2 .0 o) E at .5 =C J ¡f U) .0 E .9 ît c 0) 0 5 X IJJ 0.0 -5 -4 -3 -2 -1 012 34567 0.7 Weeks Post-Veraison o.o E(¡) -o È, 0.5 g B 0.4 IE f € o.o ct) E o.z E .o 0.1 0.0 -5 -4 -3 -2 -1 012 34567 0.6 Weeks Post-Veraison r Catechin E o.s €o) tr Epicatechin 9) 0.4 È r Epicatechin-gallate ; C Þ o.s E co s 0.2 o) ,Ë o.r 0.0 -5-4-3-2_101234567 Weeks Post-Veraison Figure 3.11. Shiraz seed proanthocyanidin subunit composition (1999-2000) with SEM (n:3). A). extension subunits (mg/berry); B). Terminal subunits (mg/berry); c). Free monomers (mg/berry) While terminal subunit analysis of the seed also showed epicatechin to be the predominant subunit, these were at most two-fold the level of catechin and epicatechin- gallate. Compare this with the ratios in the whole berry @igure 3.8) and it becomes apparent that much of the disproportionate increase in epicatechin terminal subunits must 106 occur in the skin. Although it must also be considered that terminal subunits in the seed, at around 5-6 mgiberry two weeks post-veraison (Figure 3.98), account for the bulk of terminal subuníts in the whole berry, around I mg/berry two weeks post-veraison (Figure 3.88). These observations further suggest that polymers in the seed must be substantially shorter than those in the skin, through a combination of fewer extension and more terminal subunits. Similarly with the free monomer composition in the seeds of Shiraz berries, the post veraison increase in free monomers occurs largely in the skin with the level of free monomers detected in the seed around 80% of total free monomers in the whole berry (Figure 3.8C & Figure 3.9C). Also in common with the observed pattern in terminal subunits, the increase in epicatechin monomers was proportionally less than for the whole berry suggesting some epicatechin monomer formation in the skin at this time. 3.1.6. Shiraz Seed Tannins 2000-2001 In the second experimental season, the general pattern of seed tannin accumulation was similar to that of the previous season (1999-2000). Figure 3.12 shows the level of extension and terminal subunits as well as free monomers calculated as mg/berry and mg/g fresh-weight of seed for the 2000-2001 season. Extension subunits per berry showed a rapid increase from fruit-set until four weeks pre-veraison (Figure 3.12A), whereafter the level remained relatively constant until two weeks pre-veraison when a further increase was observed over the following week. This was the highest level of extension subunits observed (5.07 mg/berry) and was around three times higher than either terminal subunits or monomers. There was a slight decrease in extension subunits at veraison whereafter the level remained relatively constant until eight weeks post veraison (3.16 mg/berry) with a decline to 1.57 mg/berry in the final week (Figure t07 3.12A). When viewed on a fresh-weight basis, extension subunits increased substantially in the first week after fruit-set and reached a maximum four weeks pre-veraison, then gradually declined until harvest (Figure 3.128). A similar pattern was observed for terminal subunits per berry with an increase immediately following fruit-set (Figure 3.12C), with a second period of accumulation was observed from two weeks pre-veraison 50 s.o A o B Eo) ô õ) 40 E 9 4.0 c Ø = 30 ãc ,.0 ôf = f o a 20 õ 2.0 c c .9 cØ t.o o 10 3 x o) t¡J ñ o.o 0 -7 6-5 4-3-2-1 0 1 2 3 4 5 6 7 8 I -7 6 -5 4 -3 -2 -.t 0 1 2 3 4 5 6 7 8 I Weeks Post-Veraison Weeks Post-Veraison 2.0 ^ 18 è c D oo 416 gÈ, 1.5 914 g p12 1.0 5to o o =f õ8 U) õ6 E o.s É E otE4 Fo t- o.o 0 -7 6-5 -4-3-2-1 0 1 2 3 4 5 6 7 8 I -7-6-5 4-3 -2-1 0 1 2 3 4 s 6 7 B I Weeks PoslVeraison Weeks Post-Veraison 20 '18 E F 16 ò 14 c) 5 o o o 't2 o E E 10 0 p 6 c) o I oE oE c 6 q 0.5 o o 4 2 0.0 0 -7 s-5 4-3-2-1 0 1 2 3 4 5 6 7 I I -7 6 -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 I 9 Weeks Post-Veraison Weeks Post-Veraison Figure 3.12. Total flavan-3-ols and proanthocyanidin subunits in Shiraz seeds 2000-2001 A. Extension subunits (mg/berry); B. Extension subunits (mglg fresh weight of seed); C. Terminal subunits (mg/berry); D. Terminal subunits (mg/g fresh weight of seed); E Free monomers (mg/berry); F. Free monomers (mg/g fresh weight of seed). Three separate aliquots of each sample were extracted, cleaved and analysed Uy ÈrtC. rne data presented is the mean (+SEM) of the triplicate extractions 108 until veraison, thereafter declining towards harvest. The pattern per gram fresh-weight showed a more gradual increase towards veraison, after the initial sudden increase immediately following fruit-set (Figure 3.12D). The brief increase in terminal subunits around veraison was exaggerated in the pattern of free monomer accumulation both on a per-berry and per-gram fresh-weight basis (Figure 3.128,3.12F). The observed pattern of monomer accumulation was a slow increase until three weeks pre-veraison, when an eight-fold increase was observed between that time and the peak of monomer accumulation one week post-veraison. This was followed immediately by a sharp decline over the succeeding three weeks and a steady decline from that point until harvest. 3.1.7. Shiraz Seed Tannin Composition 2000-2001 Figure 3.13 shows the proanthocyanidin composition of seeds, on a per berry basis, throughout berry development. In grape seeds epicatechin was the major constituent representing around 650/o of extension subunits, with epicatechin-gallate accounting for around 25Yo of extension subunits, while catechin represented less than l0o/o (Figure 3.134). The relative proportion of each extension subunit remained relatively constant throughout berry development. In comparison, there was a noticeable change in the relative proportions of catechin, epicatechin and epicatechin-gallate terminal subunits in the seed during berry development (Figure 3.138). During the period of apparent synthesis of terminal subunits immediately prior to veraison, proportionally more catechin was produced than either epicatechin or epicatechin-gallate. At veraison, epicatechin represented 20%o of tot¿l terminal subunits with catechin and epicatechin-gallate both around 40%. Catechin decreased post-veraison while epicatechin increased, after which the relative proportions 109 4.0 È A c) -o o, 3.0 E o =c 5 2.0 -o' f U) .9tt 't.0 c 0) x lrJ 0.0 7ô-54-3-2 -1 0't 23 456789 Weeks Post-Veraison 1.0 È o) B -o o) 0.8 E Ø Ëc 0.6 f -o f @ 0.4 (u .c E 0.2 E Fo 0.0 I -7 -6 -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 I 9 Weeks Post-Veraison 1.0 c I Catech¡n È 0.8 o Epicatechin o -o Epicatechin-gallate o) ¡ E 0.6 6 o 0.4 Eo c, o 0.2 0.0 -7 -6 -s -4 -3 _2 _1 0 I 2 3 4 5 6 7 I I Weeks Post-Veraison Figure 3.13. 2000-2001 Proanthocyanidin composition of Shiraz seeds between fruit-set and harvest showing levels of individual subunits expressed as mg/berry (tSEM; n:3). A). Extension subunit composition; B). Terminal subunit composition; C). Free monomer composition of terminal subunits varied little with epicatechin accounting for 40Yo and catechin and epicatechin-gallate both around 30% of toøl terminal subunits. In the early period of seed development, the free monomer composition was mainly catechin. By three weeks pre-veraison, seed monomer composition was equally catechin, epicatechin and epicatechin-gallate (Figure 3.13C). From one week pre-veraison the level ll0 of all monomers began to increase, with a four-fold increase in epicatechin and a three- fold increase in catechin, but only a two-fold increase in epicatechin-gallate to a maximum one week post-veraison. After this peak, the concentration of all monomers declined. Both catechin and epicatechin-gallate decreased by 50% over the following two weeks, however epicatechin decreased more slowly and there was proportionally more epicatechin than other monomers until around eight weeks post-veraison. For the last two weeks until harvest the levels of catechin and epicatechin were the same, with epicatechin-gallate around 5% of total monomers. 3.1.8. Shiraz Skin Tannins 2000-2001 On a per berry basis, extension subunits in skin showed a steady increase from fruit-set until two weeks pre-veraison at which point a sharp increase was observed (Figure 3.144). The greatest increase in extension units was recorded one to two weeks before veraison, when the level doubled from I.7 to 3.4 mglberry. Extension subunits per berry then remained static for the following two weeks when there was a further increase to the maximum level observed, around 4.0 mglberry. This was followed by a steady decline to around five weeks post-veraison, after which the level was around 1.5 to 2.0 mgherry. Considered on a per gram fresh weight basis, a substantial concentration of extension subunits was present in the developing berries at fruit-set (Figure 3.148). The level of extension subunits per gram declined immediately after fruit-set, then increased steadily for the next four weeks with a sharp increase two weeks pre-veraison. A steady decline in extension subunits per gram was observed from one week pre-veraison until harvest. lll Terminal subunits per berry exhibited a steady increase from fruit-set until two weeks pre-veraison when a sharp increase was seen, doubling terminal subunit level to 0.09 mg/berry. Subseqüently there was a lag phase when the level of terminal subunits remained static for a two-week period followed by a further increase in terminal subunits to the maximum of 0.13 mg/berry. From this point terminal subunits declined steadily to 5.0 25.0 A 4.5 Ë 4.0 o Ê s.s o E g ø 3.0 15.0 - Ec = f 2.5 o ¡= = 8 2.0 ! ro.o 1.5 .9 .E c 5 1.0 s.o lrl* r¡J 0.5 0.0 0.0 -7 -6 -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 I I -76-5-4-3-2-,t0 1 2 3 4 5 6 7 8 I Weeks Post-Veraison Weeks Post-Veraison 0 5 0.8 c D È 0.7 (¡) Q o o.u CD € E 0 0 0.5 .gc =c f oo 6 ë 3 o.s U)= 0.05 o E 'EE .E 0.2 E Ê o.t Fo 0.00 0.0 -7 € -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 I I WeeksPost-Veraison Weeks PosfVeraison 0.05 0.6 E F 0.5 0. 04 à o) o) o) t E 04 o) 0. 03 E Ø q) 03 e E o 0. 02 co oE o 02 o 001 0 0.00 0.0 -7 s -5 4 -3 -2 -1 0 't 2 3 4 5 6 7 8 I -7-6-5 4-3-2-1 0't 2 3 4 s 6 7 8 I Wee ks Post-Ve ra iso n Weeks Post-Veraison Figure 3.14. 2000-2001 Proanthocyanidin composition of Shiraz seeds between fruit-set and harvest showing levels of individual subunits expressed as mg/berry (tSEM; n:3). A). Extension subunit composition; B). Terminal subunit composition; C). Free monomer composition tt2 around 50% of the maximum, thereafter remaining relatively constant at around 0.06 mg/berry until harvest (Figure 3.14C). As with the extension subunits, there was a substantial concehtration of terminal subunits at fruit-set. This point Ïepresents the highest concentration of terminal subunits per gram of skin at any point during berry development (Figure 3.14D). There was a dramatic decrease as the berry started to develop, followed by a steady increase towards veraison. Another noticeable increase was observed immediately prior to veraison, whereafter the level remained relatively constant at around 0.45 mglg fresh weight until three weeks post-veraison. The concentration of terminal subunits in the skin then declined to approximately 0.25 mglg fresh weight at five weeks post-veraison, remaining constant around that level until harvest. Monomer accumulation in the skin showed a steady increase from fruit-set to a peak of 0.04 mg/berry two weeks pre-veraison declining to around 0.02 mgiberry at four weeks post-veraison after which the level was relatively constant (Figure 3.148). On a per gtam fresh weight basis the highest level of monomers in skin also occurred at fruit-set followed by a sharp decrease from seven to six weeks pre-veraison. A less dramatic decrease was observed over the following week and then a small increase (Figure 3.14F). From four weeks pre-veraison the level was constant for two weeks then began a gradual decline until four weeks post-veraison, at which level it remained static until harvest. 3.1.9. Shiraz Skin Tannin Composition 2000-2001 In grape skins, epigallocatechin extension subunits were observed in addition to the catechin, epicatechin and epicatechin-gallate subunits detected in the seeds (Figure 3.154). Extension subunit composition in grape skins was primarily epicatechin and epigallocatechin with the former slightþ in excess and levels of both some twenty-fold greater than either catechin or epicatechin-gallate. The levels of epicatechin and 113 epigallocatechin showed a two-fold increase during the main period of synthesis one to two weeks prior to veraison. Post-veraison, both epicatechin and epigallocatechin extension subunits declined by 50% tówards harvest, while the levels of catechin and epicatechin-gallate were virtually unchanged throughout berry development. 3 0 ð A o) -o 2 .5 o) E 2 .0 tt, Ëc f I .5 -o f Ø c 1 0 .9 cat, o 0. 5 x l¡J 0. 0 -7-6-54-3-2 -1 0123 456789 Weeks Post-Veraison 0.1 5 à B (¡) -o o) g 0.10 Ø .E f -o f Ø o.o5 o .E oE t-- 0.00 7-6-54-3-2 -1 0123 456789 Weeks Post-Veraison 0.04 c r Catechin o Epicatechin 0.03 Epicatechin-gallate ot r -o o) I Epigallocatechin E 0.02 g. o E co 0.01 o 0.00 -7-6-54-3-2 -1 012 3456789 Weeks Post-Veraison Figure 3.15. Shiraz skin composition between fruit-set and harvest showing levels of individual subunits and monomers expressed as mg/berry (tSEM; n:3). A). Extension subunit composition; B). Terminal subunit composition; C). Free monomer composition tt4 The composition of terminal subunits in the skins was primarily catechin, with low levels of epicatechin and epicatechin-gallate observed (Figure 3.158). There was a two-fold increase in catechin during the period'of synthesis commencing just prior to veraison and little change in the amount of epicatechin and epicatechin'gallate. In the skins of the grape berry, the free monomer composition was also predominantly catechin (Figure 3.15C). Catechin accumulated steadily from flowering reaching a peak around two weeks prior to veraison. Subsequently, a decline accounting for around 50% of catechin monomers was observed. While gtape skin monomers between fruit-set and two weeks post-veraison were mainly catechin, after this time there were similar levels of catechin and epicatechin due to a two-fold increase in epicatechin free monomers. The level of epicatechin-gallate monomers was low and remained relatively constant throughout berry development. The decrease in catechin free monomers in the grape skin coincided with an increase in catechin terminal subunits. 3.1.10. Shiraz Flesh Tannins 2000-2001 Tannins were extracted from a sample of grape berry flesh after veraison (01 February 2001; +3V) during the 2000-2001 season. Generally the levels of tannins detected in the flesh were lower than the level detected in the skin. While the levels of flavan-3-ols were similar between the flesh and skin, extension subunits and toøl tannins were lO-fold lower in the flesh than the skin and almost 100-fold lower than in the seed on a fresh weight basis. ll5 Table 3. l. Tannins in the flesh of Shiraz grapes three weeks post-veraison (01 February 2001). Flavan-3-ol subunits, comprising proanthocyanidin terminal subunits and free monomers (mg/g fresh weight of flesh + sEM; n:3), and proanthocyanidin extension subunits (phloroglucinol adducts) are shown, as well as total flavan-3-ols and extension subunits and total tannins in the flesh (mgig fresh weight of flesh+ SEM; n=3) mq/q fresh we¡qht Catechin 0.04910.013 Epicatechin 0.229!0.026 Epicatechin-gallate 0.014t0.001 Gatech in-ph loroglucinol 0.010r0.001 Epicatech in-phlorog luc¡nol 0,326+0.066 Epicatechin-gal late-ph lorogluci nol 0.03010.004 Epigal locatechin-phlorog lucinol 0.040r0.008 Free Monomer + Terminal Units 0.292!0.017 Extension Units 0.406r0.060 TotalTannins 0.698r0.044 Table 3.1 shows the content and composition of tannins in the flesh of Shiraz berries post-veraison. The sum of free monomers and terminal subunits, as products of LAR, is shown and of these the dominant compound was epicatechin accounting for arowtd, TSYo of tot¿l flavan-3-ols. Extension subunit composition in the flesh was also predominantly epicatechin with around 80% of extension subunits being epicatechin. Epicatechin-gallate and epigallocatechin extension subunits were present at more or less equal levels, combined accounting for around I7.5% of the total, while catechin made up the remaining 2.5%o of total extension subunits (Table 3.1). al 3.1.11. Direct Acid-Catalysis of Seed & Skin Residues 2000-2001 The post-veraison decrease in tannin, in both seeds and skins, may have been the result of less efficient extraction from these tissues as the grapes ripened. While repeated extraction of the residues with acetone released only trace amounts of additional ll6 proanthocyanidins (data not shown), direct acid-catalysed cleavage of the residue did yield significant amounts of proanthocyanidins. The amount of extension and terminal sübunits released with direct acid-catalysed cleavage of the seed residue increased from veraison to harvest (Figure 3.164, 3.168). Acid-catalysis of the seed residue showed a three-fold increase in terminal subunits and an eight-fold increase in the yield of extension subunits from veraison to eight weeks post-veraison. The level of extension subunits in the residue at veraison was around 3o/o of that in the extract. By eight weeks post-veraison, levels of extension subunits in the residue were equivalent to 30% of the ^ 4.5 Seed A 1.80 Seed B t 1.60 Ë ¿.0 o) -o I -o È, 3.5 o) 1.40 F ! E e 3.0 1.20 Ît, I a c. ¿.c =c 1.00 a Ê= 2.0 -o) 0.80 Ø U) 0.60 c 1.5 (ú o c 'õ 1.0 'E 0.40 c L * o.s t-o 0.20 t o.o 0.00 0 6 I 0 6 I Weeks Post-Veraison Weeks Post-Veraison ^ 4.5 Skin c 0.14 Skin D È Ë 4.0 (¡) o.12 -o -o 3.5 o, è,c 0.10 e g.o I E U' 9_ 0.08 c. ¿.c =c l f -o € 2.0 J 0.06 U) I U) c 1.5 (5 0.04 '8 1.0 .c c E o.o2 g o.s (¡) r,rJ F 0.0 0.00 0 6 I 0 6 8 Weeks Post-Veraison Weeks Post-Veraison Figure 3.16. Extension and terminal subunit levels present in the acid-cleaved acetone extract (n -SgVt) ot Shiraz grape seeds and skin and released from the residue by direct acid-catalysed cleavage following acetone extraction ( I +SEM), at veraison, six and eight weeks post-veraison. A. Seed extension subunit yield from the acetone extract after acid-catalysis and from direct acid-catalysis of the residue; B. Seed terminal subunits; C. Skin extension subunits; D. Skin terminal subunits tt7 exhactable proanthocyanidins in grape seeds. Terminal subunits in the residue were also around 3Yo of the level in the extract at veraison, but equivalent to 20% by eight weeks post-veraison. Conversely, acid-catalysis of Shiraz skin residues showed a decrease in the yield of proanthocyanidins with increasing maturity (Figure 3.I6C,3.16D). Direct acid-catalysis of the skin residue at veraison and eight weeks post-veraison showed a three-fold decrease in extension subunits in the residue and a two-fold decrease in terminal subunits in the residue during this time. In the skin at veraison, the level of extension subunits in the residue represented approximately 25o/o of that in the acetone extract while the level of terminal subunits was almost 50%o of that in the acetone extract. By eight weeks post- veraison extension subunits in the residue were equivalent to l5%o of the level in the acetone extract and terminal subunits had decreased to around20%. 3.1.12. Polymer Length (mDP) in Grape Berries In Shiraz seeds, mean proanthocyanidin polymer length (mDP) was relatively constant throughout berry development (2000-2001) at around five subunits (Figure 3.174). From fruit-set to one week pre-veraison, polymer length remained between five and six subunits. An increase in terminal subunits at one week pre-veraison temporarily exceeded the accumulation in extension subunits resulting in a slight fall in polymer length to around four subunits between one week pre-veraison and veraison either through increased extension subunits, or through loss of terminal subunits, as appeared to be the case (Figure 3.13C). Following this decrease, mDP increased to five subunits and this level was maintained from two weeks post-veraison until harvest. In the lggg-2000 season, seed proanthocyanidin polymer length was slightly higher pre-veraison at around 118 I 6.5 subunits and declined towards A. Seed tr 1999-2000 +-2000-2001 7 veraison (Figure 3.17A). From that 6 Fe 5 E fL time until harvest polymer length OA .tr E E.E' 3 remained between 3.5 and 4.0 2 1 subunits. 0 -7-6-5-4-3-2-1 0 1 2 3 4 5 6 7 I I weeks post-Veraison Shiraz skin polymers (2000-2001) 60 B. Skin +2000-2001 increased in length during the phase 50 40 of berry development immediately IL Q30 b after fruit-set from around 25 to 40 20 subunits. Mean degree of 10 0 polymerisation then remained -7-6-5-4-3-2-1 0 'l 2 3 4 5 6 7 8 9 Weeks Post-Veraison relatively constant at around 30-40 Figure 3.17. Mean degree of polymerisation (mDP), (Figure derived from the sum of terminal and extension subunits subunits until veraison divided by terminal subunits. A). Grape seed proanthocyanidin polymer length over two seasons. 1999- 3.178). After veraison, mDP was seen 2000 (fl ) and 2000-2001 (O); B). Grape skin polymer length,2000-2001 (O) to decrease gradually towards 30 subunits at four weeks post-veraison. From that time, polymer length was around 25 to 30 subunits dropping to 20 subunits at harvest. 3.1.13. Extractable Tannins at Commercial Harvest Commercial harvest of winegrapes normally occurs between 20-24"Bix(Boulton et al., 1998). In the Willunga vineyard at the end of the 1999-2000 season, the Chardonnay grapes were harvested six weeks post-veraison at 23.2"8rix, while the Shiraz was harvested seven weeks post-veraison at 23.5oBrix. At harvest total tannins in the 119 Chardonnay fruit was around 5.63 mg/berry (Table 3.2) with extension units comprising around 78%o of the total. In the Shiraz grapes the total acetone extractable tannins in the berry was 4.88 mg/berry also with exterision subunits accounting for 78%o of the tot¿l (Table 3.1). Of the total extractable tannins 59%o was extracted from the seed where extension subunits only accounted for around 68%o of the total. With the breakdown of flavan-3-ols into free monomers and terminal subunits it is apparent the latter represent the bulk of these resulting in a mean polymer length in the seed at this time of around 4 subunits (see also Figure 3.174). In Shiraz grape berries in the following season, commercial harvest occurred nine weeks post veraison (29"Brix), which was approximately a week later than was usual for this vineyard and the fruit were noticeably shriveled. One week earlier at eight weeks post- Table3.2. Extractable tannins at commercial harvest in Chardonnay 1999-2000 (+6Y; 23.2"Brix); Shiraz 1999-2000 (+7Y; 23.5"Brix); Shiraz 2000-2001 (+8V; 24.9"Brix), showing total extension and terminal subunits, free monomers and the sum of terminal subunits and free monomers presented as total flavan-3-ols (mg/berry) in seeds and skins of Vitis vinifera L Chardonnay Shiraz Shiraz 1999-2000 I 999-2000 2000-2001 Whole Berry Extension Subunits 4.36 3.78 4.47 Terminal Subunits 0.62 0.94 Free Monomers 0.49 0.30 Total Flavan-3-ols 1.27 1.11 1.24 Total 5.63 4.88 5-71 Seed li Extension Subunits 1.95 3.16 Terminal Subunits 0.64 0.88 Free Monomers 0.28 0.28 Total Flavan-3-ols 0.93 1.16 Tofal 247 4.32 Skin Extension Subunits 1.31 Terminal Subunits 0.06 Free Monomers 0.o2 Total Flavan-3-ols 0.08 Total 1.39 120 veraison, total soluble solids were 24.9"Brix and total extractable tannin in the berry was 5.71mgherry . Once again 78%o of this total was attributed to extension subunits, with terminal subunits comprising approximately l7o/o and free monomers 5% (Table 3.2). However, in the second season (2000-2001) seed tannins accounted for 75o/o of total tannins. From the seed derived material the average polymer length was 4.6 subunits, while grape skin polymer length at this time was 22.8 subunits, with extension subunits in the skin accounting for almost 95% of the extractable tannin in that tissue (Figure3.17). 3.1.14. Tannins in Grapevine Leaves and Tendrils Tannins were also extracted 1.4 A 1.2 ot from the leaves of the È, 1.0 E grapevine (Vitis vinifera L. .!2I 0.8 í)I 0.6 oC Shiraz) with acetone 0.4 cv. IL-q 0.2 (70o/o vlv) and this acetone 0.0 Stage Stage 2 Stage 3 Stage 4 Stage 5 1 extract was subjected to 1.2 I Catechin B acid catalysed cleavage of B 1'o I Epicatechin EoB ¡ Epicatechin-gallate the proanthocyanidins in ø õ 0.6 Cå the presence of a molar Ë o.¿ (U tr 0.2 excess of phloroglucinol. 0.0 Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 The composition of free terminal Figure 3.18. Composition of flavan-3-ols in Shiraz leaves at hve monomers, and different stages of development as described in Section 3.2.1. A). Free monomers and terminal proanthocyanidin subunits subunits released through expressed as flavan-3-ols (mg/g fresh weight of leaf +SEM; n=3); B). Free monomers and terminal subunits expressed as flavan-3- acid-catalysed cleavage of ols (mg/leaf +SEM; n=3) proanthocyanidins in the acetone extract, collectively termed flavan-3-ols, is presented in Figure 3'18 t2l Composition of flavan-3-ols in newly emerged and expanding leaves was predominantly epicatechin, while in the more mature leaves, Stages 3 and 4 as well as the fully expanded matufe leaf, Søge 5, there were similar levels of each of the flavan-3-ols, catechin, epicatechin and epicatechin-gallate. The concentration of flavan-3-ols was highest in the youngest leaves, decreasing with leaf age (Figure 3.184). However the total per leaf was seen to increase throughout leaf expansion suggesting ongoing synthesis (Figure 3.188). Extension 7.0 subunits released o) o) 6.0 A E from Shiraz leaf extracts eø 5.0 5c -o 4.0 during acid-catalysed U) c 3.0 o at, cleavage had a similar E 2.0 o x 1.0 lrJ co osition to that of the 0.0 Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 skin of Shiraz berries with 10.0 I Catechin Êo o 9.0 epigallocatechin present in I Epicatechin B o) 8.0 E 7.0 ! Epicatechingallate Ëtt addition to catechin, c 6.0 J n Epigallocatechin ¡ 5.0 U)= epicatechin and epicatechin- c 4.0 .9 3.0 at, c) 2.0 gallate (Figure 3.19), On a x 1! 1.0 0.0 per gr¿ìm fresh weight of Stage 1 Stage 2 Stage 3 Stage 4 Stage S leaf basis, extension subunit Figure 3.19. composition of proanthocyanidin extension subunits in Shiraz leaves at flrve stages of development as described in Section 3.2.1. A). Extension subunits (mgþ fresh weight of leaf concentration increases with +SEM; n:3);B). Extension subunits (mg/leaf +SEM; n:3) increasing leaf size and age from the newly emergent leaf until leaf Stage 4 (Figure 3.194). However, in the fully expanded leaf the conce ation is lower than at any other time during leaf expansion, likely as a result of massive leaf expansion. The level of extension subunit adducts in Shiraz leaves on a per leaf basis appears very low in the newly emergent leaf compared to 122 the mature leaf (catechin 0.0022+0.0008; epicatechin 0.0223+0.0043; epicatechin-gallate 0.0052+0.0006; epigallocatechin 0.0075+0.0041 mg/Ieaf). While the concentration of exterision subunits was seen to decrease substantially in thê leaf as the leaf matured (Figure 3.194) the total amount per leaf continued to increase (Figure 3.198), indicative of ongoing extension subunit synthesis in the leaves. This general pattern in both flavan-3-ols and extension subunits is confirmed in Figure 3.20, which shows total proanthocyanidins subunits in each of the five Shiraz leaf stages. The amount of both flavan-3-ols and extension subunits in leaves increased as the leaf expanded and matured. The level of proanthocyanidins in the newly emerged leaves (flavan-3-ols 0.012; extension subunits 0.037; total tånnins 0.044 mglleaf) was around 200-fold lower than that in the fully matured leaf. On a per gram fresh weight basis, the mature leaf contained a total of around 3.8 mg of total tannin, in the skin of a ripe grape berry the concentration was around 5.1 mglg fresh weight of skin, while in the seed at commercial harvest the level was around 27.8 mgper gram fresh weight of seed. 1 4.0 p (õ I Flavan-3-ols o ) 1 2.0 o) Extension a E ) 10.0 at - + ' TotalTannins .E p 8.0 a c a (5 o 6.0 ) o ! c 4.0 ) (ú I 2.0 o- 0.0 Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Figure 3.20. Total proanthocyanidin subunits in Shiraz leaves at five stages of expansion from newly emerged to fully mature as described in Section 3.2.l.Total flavan-3-ols, total extension subunits and total tannins (mg/leaf +SEM; n:3) are shown 123 In addition to the leaves of the grapevine, tendrils were also collected and tannins extracted and analysed after acid-catalysed cleavage ofproanthocyanidins. The results of this analysis can be seen in Table 3.3. The flavan-3-ols were largely epicatechìn and epicatechin-gallate, each a little over 40Yo with catechin around l0l5% of the total. Extension subunits in the tendrils were predominantly epicatechin at around 75yo of the total with the other extension subunits more or less equal (catechin 8%; epicatechin gallate 5Yo; epigallocatechin I2.5%). The level of total proanthocyanidin subunits in the tendril was similar to that observed in the skin of the mature berry, although the proportion of free monomers and terminal subunits (flavan-3-ols) was substantially greater. The relative proportion of extension subunits in the acetone extract of the tendrils was substantially less than that observed in either the seeds or skin of the berries or in the leaves. Table 3.3. Tannin composition of shiraz tendrils showing flavan-3-ol subunits, comprising proanthocyanidin terminal subunits and free monomers (mg/g fresh weight of flesh t SEM; n:3). Also shown are proanthocyanidin extension subunits (phloroglucinol adducts) as well as total flavan-3-ols, total extension subunits and total tannins in the flesh of Shiraz grapes (mgig fresh weight of flesh + SEM; n:3) mq/q fresh weight Catechin 0.255 !0.042 Epicatechin 0.835 t0j02 Epicatechin-gallate 0.852 r 0.045 Gatech in-ph loroglucinol 0.336 r 0.029 Epicatech in-phlorogl ucinol 2.986 t 0.326 Epicatechin-gallate-phloroglucinol 0.212 !0.O17 Epigal locatechi n-ph lorogl ucinol 0.514 r 0.025 Total Flavan-3-ols 1.942 t 0.189 Total Extension Subunits 4.048 r 0.396 TotalTannins 5.990 + 0.579 r24 3.4. Discussion Tannins extracted from whole berries of the I/itis vinifera cultivars, Shiraz and Chardonnay showed very similar patterns and levels of accumulation. In both cultivars, the total flavan-3-ols and extension subunits per berry reached their maxima around one week post-veraison, followed by a gradual decline towards the time when the fruit was harvested. The main difference between the two cultivars was that the total levels of tannins were slightly higher in Shiraz berries. In terms of their composition, both Chardonnay and Shiraz extension subunits were primarily epicatechin. In Chardonnay, flavan-3-ols in the whole berry were equally catechin, epicatechin and epicatechin- gallate, while in Shiraz, particularly post-veraison, there was proportionally more epicatechin. The similarity in the pattern and timing of tannin accumulation across the two cultivars studied here suggests that tannin biosynthesis in grape berries is a developmental process and likely the pattern throughout the Vitaceae. However, the differences in total tannins and tannin composition observed here indicates the possibility of considerable variability in composition across the family. While the extent of this diversity has yet to be established, there is present the potential to alter wine tannin profiles through a variety of strategies. These might include blending of wines from cultivars with novel or desirable tannin prof,rles, puriffing tannins from such cultivars for use as a wine additive as well as the strategic breeding or engineering of new varieties with idealised or novel tannin profiles. Much of the quality of red wine derived from its tannin content and composition is dependent on the distribution between monomers and polymers, on the length of polymers and on their associations with other compounds, which effects their stability and solubility in the wine. HPLC analysis of the acetone extract of grape tannins before 125 as well as after acid-catalysed cleavage partitions the flavan-3-ol fraction into polymeric terminal subunits and free monomers, the monomers being significantly more bitter in the subsequent wine (Noble, 1994). This analysis of Shiraz whole berry extracts showed that a large proportion of the flavan-3-ols present pre-veraison were terminal subunits. This analysis also suggested that the post-veraison increase in flavan-3-ols was equally distributed between terminal subunits and free monomers generally. While both of these decrease post-veraison, the inference from previous studies and this analysis is that monomers become incorporated into polymers and that the extractability of polymers from the grape decreases with increasing grape maturity as discussed below. What is not apparent from the whole berry analysis is the relative contribution of the seed and the skin to the pool of tannic material extracted from the whole berry. In grape seeds, tannin biosynthesis and accumulation commenced with seed development. From the f,rrst time that seeds could be readily dissected from the remainder of the berry, one week post-flowering, significant levels of proanthocyanidins could be detected suggesting that tannin biosynthesis commences in the earliest stages of seed development. The tannin composition of grape seeds consisted primarily of epicatechin extension subunits the majority of which were formed in the three weeks following fruit- set. The bulk of tannin synthesis in grape seeds occured immediately after fruit set and was complete five weeks before veraison. Both extension and terminal subunits increased at this time and little free monomer was detected (Figure 3.10). A second phase of tannin accumulation commenced just prior to veraison and the maximum levels of proanthocyanidins were observed at veraison. These observations are consistent with and expand upon previous reports that suggest a substantial amount of proanthocyanidin material is present in grape seeds prior to veraison @omeyer et a1.,1986; Katalinic & 126 Males, 1997; Kennedy et a1.,2000a; Kennedy et a1.,2000b; Jordão et a1.,2001b). The second phase of proanthocyanidin accumulation in seeds was the result of an increase in terminal subunits and monomers. A similar'increase in both terminal subunits and free monomers was reported by Kennedy et al. (2000b) with comparable proportions of individual subunits, although the timing was one to two weeks later than that reported here. If free monomers were being incorporated into polymers, then the level of terminal subunits would be expected to continue to increase as free monomers decreased. Such being the case polymer length would decrease, as observed by Kennedy et al. (2000b). While we observed such a decrease in polymer length in the previous season (1999- 2000), it was not seen during 2000-2001. This suggests that a substantial increase in extension subunits occurred at this time that was sufficient to maintain a constant polymer length. The post-veraison decrease in both terminal subunits and free monomers that was observed over two seasons appears to represent a net loss of flavan-3-ols. Similar increases in monomers around veraison followed by a significant decrease post-veraison have been previously reported (Romeyer et al., 1986; Kennedy et al., 2000b). The decrease in monomers was accompanied by a decrease in proanthocyanidin polymer length suggesting their conversion to terminal subunits. However, at this time both terminal and extension subunits also declined. This decrease in proanthocyanidins post- veraison has been widely reported and while turnover or degradation are possible explanations, no direct evidence of such metabolism has been reported in plants. Several authors have suggested this decrease could be athibuted to reduced exhactability resulting from the conjugation of proanthocyanidins with other cellular components (Amrani-Jottei et al., 1994; Cheynier et al., 1997; Saint-Cricq de Gaulejac et al., 1997 ). 127 Further, Kennedy et al. (2000b) asserted that oxidative cross-linking of polymers would decrease their extractability. The inference drawn from these observations is that a substantiál amount of proanthocyanidin remains in the residue after extraction and the proportion of non-extractable material changes during berry development (Figure 3.16). Direct acid-catalysis of the grape seed residue remaining after extraction with acetone supported this hypothesis as an increase in the yield of both terminal and extension subunits was observed in the residue post-veraison (Figure 3.16). The yield of proanthocyanidin extension subunits from the residue increased with maturity to the extent that the sum of extractable extension subunits and those released from direct acid- catalysis of the residue remained constant after veraison. These results suggest that association of proanthocyanidins with the residue accounts for the decrease in extension subunits post-veraison. However, the sum of extractable and residue derived extension subunits (around 4.0 mg/berry) was at no time equivalent to the maximum level of extension subunits at one week pre-veraison (around 5.0 mg/berry). Thus, assuming the maximum observed at one week pre-veraison represents total extension subunit production in the seed, some 2oo/o of extension subunits remain unaccounted for. While there was an increase in terminal subunits released from the seed residue post- veraison, these represented significantly less than the observed decrease in terminal subunits in the acetone extract. As with the analysis of extension subunits released from direct acid-catalysis of the grape seed residue, a substantial portion of terminal subunits remain unaccounted for. The formation of branched or oxidatively linked polymers might well account for these 'lost' extension and terminal subunits, as well as the decrease in free monomers that do not appear as terminal subunits. Branched polymers would require the formation of 4-6 128 interflavan linkages in addition to the more common 4-8 linked proanthocyanidin dimers. The proanthocyanidin dimer B5 contains the 4-6 dimeric vinculum and has been reported as a significant component of grape seed tannins (Santos-Buelga et a1.,1995; Souquet e/ aI., 1996a; Saint-Cricq de Gaulejac et al.,1997; de Freitas & Glories, t999)' Further, such linkages are more resistant to hydrolysis than 4-8 linkages (Hemingway & McGraw, 1983; Ricardo da Silva et a1.,1991b). In addition, oxidative coupling (Geissman & Crout, 1969) of adjacent proanthocyanidin subunits generates xantþlium salts, also resistant to hydrolysis (urd & Somers, 1970). Xantþlium derivatives (Singleton & Esau, 1969; l*azdina& Borzell, 1971) and the free radicals implicated in the formation of oxidative linkages (Kennedy et a1.,2000b) have previously been reported in grape seeds. Thus, during the acid-catalysed cleavage of seedcoat polymers, 4-6 linked dimers and xanthylium derivatives would be released, but not identified by this analysis as not corresponding to any of the known monomers or phloroglucinol adducts. While several candidate peaks were identified during HPLC analysis, these have yet to be isolated and their structures confirmed by NMR. At harvest, the Shiraz seed mDP was, on average, between four and five subunits across two seasons. Previous studies have reported seed mDP in Cabernet Sauvignon as high as 16 subunits with Cabernet franc and Carignan around 9 subunits (Moutounet et a1.,1996) and Alicante Bouchet at around 6-7 subunits Qrieur et al., 1994). Other workers (de Freitas et a1.,2000;Kennedy et a1.,2}}}a;Kennedy et a1.,2000b) have also observed a decrease in grape seed polymer length during maturation, however this was not observed here to any appreciable extent. The most likely explanation for this observation appears to be seasonal variation in tannin composition. Considerable seasonal variation in the tannin content of seeds has been reported (Katalinic & Males, 1997; de Freitas & Glories, 129 1999). Data collected in the previous season from the same site (1999-2000), and analysed by the same method, showed a higher mDP pre-veraison (6.5 subunits), decreasing with maturity to a lower level than that observed in the 2000-200I season. This represented a decrease of similar magnitude, approximately 40o/o, to that reported by Kennedy et al. (2000b). In the previous season (1999-2000), the same general pattern of accumulation was observed with an initial burst in synthesis occurring a few weeks after fruit-set followed by a second phase around veraison. However, the initial period of accumulation \Mas much smaller, around 45o/o of the maximum observed that season compared with 72% during the 2000-2001 season. The second phase of tannin accumulation in the previous season commenced later, one week pre-veraison, reaching a maximum two to three weeks post-veraison. In the early phase of accumulation, the relative proportion of terminal subunits was lower than in 2000-2001, but the increase during the second phase was proportionally greater, resulting in a higher mDP early in development followed by a noticeable decrease. In addition to the variation in the timing of seed tannin accumulation between seasons, there was also a significant difference in total tannins derived from the seeds as well as seed number per berry. This is supported by the observations of Harbertson e/ al. (2002), which reported variability in total seed tannin being influenced more by seed number per berry than by tannin per seed. A final consideration on seed tannin extractability and differences between the results presented here and elsewhere concerns sample preparation. Tannin analysis of grape seeds was conducted on ground frozen seeds as opposed to whole seeds. It is possible that the homogenisation process exposed grape seed tannins to proteins associated with embryonic material resulting in a degree of protein fining. Such an occuffence has not previously been studied or documented. Two factors likely to impact upon such a 130 phenomenon are the nature of the extraction solvent (70Yo acetone), which precipitates protein (Sambrook et al., 1939). In addition, recent research has shown that t¿nnin extractability from seeds increases with cracking or disruption of the seed and seed'coat (pers. comm. Dr. Steve Price, Ph.D.). The proanthocyanidin content of the pulp or flesh of grape berries has in the past been largely neglected. Generally the flesh has been considered not to contain any proanthocyanidin material (Souquet et al., I996b), although trace amounts of proanthocyanidins have been reported in the vascular traces of the fruit (Lea et a1.,1979). However, flavan-3-ols and proanthocyanidins have previously been reported in the flesh of grape berries, albeit white cultivars (Lee and Jaworski, 1989). Here it is reported that in the flesh of Shiraz berries, a low level of proanthocyanidins has been detected. The total tannic material detected in the flesh was around 0.7 mg per gram fresh weight of flesh (Table 3.1). rWith the weight of berry flesh at that time (+3V) being around 0.23 grams, the amount per berry was approximately 0.16 mg total tannins in the berry flesh. In comparison, total tannins in the seed at this time was around 5.9 mglberry (51.9 mg/g fresh weight of seed) and in the skin 3.2 mg per berry Q29 mglg fresh weight of skin). Thus, the flesh contributed less than 2%o of total berry tannins. It is likely that the levels detected in the flesh are attributable to the berry vasculature or the result of contamination with seed or skin material during dissection in the field. In grape skins the concentration of proanthocyanidin subunits was relatively high at fruit- set. For free monomers and terminal subunits, this was the highest concentration observed during berry development, with the concentration of extension subunits almost as great as that in the seed at this time. This may be partly due to the limitations of field dissection at this early stage as it was difficult to separate skin from flesh, resulting in an 131 arbitrary separation into seed and not-seed. This fraction would have included the ovary wall, a tissue known to contain proanthocyanidins fougère-Rifot, et al., I995a). The pattern of accumulation in grape skins showed that free monomers, notably catechin, appeared prior to veraison, but only at relatively low levels. These began to decrease towards veraison coincident with a substantial increase in catechin terminal subunits that greatly exceeded the pool of catechin free monomers. This strongly suggests an upregulation of the biosynthesis of catechin at this time, although expression studies of the gene encoding the biosynthetic enzyme (leucoanthocyanidin reductase) and assays of enzyme activity will be required to confirm this. Concurrent with this apparent period of flavan-3-ol biosynthesis, was an increase in extension subunits suggesting these too were being synthesised. Thus, there was an accumulation of proanthocyanidins during the pre- veraison phase of berry development when anthocyanin synthesis had not commenced. This could account for the pattern of expression of flavonoid pathway genes reported by Boss el al. (I996a) since the early part of the pathway is required for the synthesis of both anthocyanins and tannins. This suggests that expression of genes in the flavonoid pathway may occur pre-veraison for the synthesis of tannins, and possibly flavonols, whereas expression post-veraison is required for anthocyanin biosynthesis. In grape skins, the increase in both extension and terminal subunits was coincident and the rates of accumulation complimentary, such that polymer length remained virtually unchanged from just after fruit-set until just after veraison. Post-veraison, polymer length gradually declined and this was due to the rate of extension subunit decrease exceeding that of terminal subunits. Such a decrease in proanthocyanidins post-veraison has been previously reported (Czochanska et a1.,1979; de Freitas & Glories, 1999; Kennedy et al., 2001). While Czochanska et al. (1979) observed polymer length to decrease towards 132 harvest, the findings of de Freit¿s & Glories (1999) suggest a decrease in mDP post- veraison in one cultivar (Ugni blanc) and an increase in another (Semillon). Kennedy e/ at. (2001) also reported mDP increasing towards maturity in Shiraz. However, the post- veraison increase observed by Kennedy et al. (2001) was only to a level similar to that reported here at a similar level of ripeness as defined by sugar accumulation ('Brix). The harvest mDP of proanthocyanidins in Shiraz skin reported here to be around 25-30 subunits, is comparable with those reported elsewhere (Moutounet et a1.,1996) Decreasing proanthocyanidin extraction from grape skins post-veraison has been attributed to tannins forming associations with other cellular components such as cell wall polysaccharides, lignins and proteins (Amrani-Joutei et a1.,1994; Escribano-Bailon et a1.,1995; Cheynier et a1.,1997; Saint-Cricq de Gaulejac et al., 1997). The association of grape skin proanthocyanidins with components of the grape skin residue, as seen with grape seeds, might account for the apparent decline in skin tannins. Direct acid catalysis of the grape skin residue realised a substantial yield of proanthocyanidins, however the level of both extension and terminal subunits in the residue decreased during ripening. The formation of 4-6 linkages in grape skins might also be an explanation for the apparent decrease of grape skin proanthocyanidins as mooted for grape seeds. Reports to date suggest 4-6 linkages do not occur in grape skins to any appreciable extent with only trace amounts of the 85 dimer detected in the skin while it was one of the most prevalent in seeds (de Freitas & Glories, 1999). However, Cheynier et al. (1997) reported that proanthocyanidins with 4-6 linkages bound much more readily to proteins than those with predominantly 4-8 linkages. This could account not only for the absence of 4-6linked dimers, but for a substantial portion of the apparent loss of proanthocyanidins in grape skins during ripening. Thus, it would appear that the overall decline in grape skin 133 proanthocyanidins between veraison and harvest could only be accounted for by proanthocyanidins in the skins forming increasingly stable associations with other cellular comþonents. Furthermore, the nature of these associations rehders them resistant to dissociation by the methods utilised here. Around the time of veraison, there was a proportional increase in epigallocatechin extension subunits in the skin (Figure 7A). This increase in flavonoid intermediates that are trihydroxylated on the B-ring (Figure 1), occurs immediately prior to veraison when the anthocyanin pigments accurnulate in the grape skin, these being primarily the trihydroxylated malvidin glycosides (trtazza & Miniati, 1993). This suggests that the accumulation of tannins and anthocyanins in grape skins is a coordinated process, although it has yet to be established whether these compounds occur in the same cells. It is clear that tannins in the seeds and skin of Vitis vinifera differ biochemically, seed proanthocyanidins were short (mDP -5) with subunit composition in both terminal and extension subunits consisting of a mixture of flavan-3-ols. In skin, the proanthocyanidins were longer (mDP -40) consisting mostly of catechin terminal subunits and epicatechin and epicatechin-gallate extension subunits. The results presented here are consistent with much of the published literature for grape seed and skin tannins and confirm the suggestion that the bulk of tannin synthesis occurs prior to veraison. Most of this synthesis occurs in the seeds and it is now apparent that accumulation of tannins in the seeds is independent of that in the skins. It was also observed that early accumulation of flavan-3-ols and proanthocyanidins in the skins was consistent with the reported expression pattern of genes coding for enzymes in the flavonoid biosynthetic pathway prior to veraison. This suggests early expression of the pathway is at least related to the synthesis of extension subunits in the first three to four weeks of berry development. 134 Expression studies of the gene encoding the biosynthetic entf,yme,leucoanthocyanidin reductase, should correlate with the observed pattern of monomer and terminal subunit accumulation presented here. For this purpose, the tiSsue most actively synthesising flavan-3-ols, that of young berries immediately following fruit-set, would present the best opportunity to clone the genes encoding biosynthetic enzymes. 135 9gr Chapt er 4. Flavonol Synthesis and Expression of Flavonol Synthase Genes in Vítís víniferøL. 4.L. Introduction Flavonols are products of the flavonoid biosynthetic pathway, which also gives rise to anthocyanins and to condensed tannins in grapevines (Stafford, 1990; Dame, 1993). Flavonols in plants have physiological functions ranging from microbial interactions to pollen fertility (Koes et al., 1994; Taylor, 1995). However, their most widespread roles appear to be as UV protectants (Flint et al., 1985; Smith & Markham, 1998), and as copigments in flowers and fruit (Ssen et ø1., 1972; Scheffeldt & Hrazdina, 1978). Copigmentation is a non-covalent association between a co-factor (eg. flavonols) and anthocyanins that confers stability on the coloured form of the anthocyanin molecule resulting in increased colour or altered hue (Hoshino et a1.,1980; Osawa,1982). Of the three main flavonols kaempferol, quercetin and myricetin, mainly quercetin-3-O- glucoside and -3-O-glucuronide are found in grape berries (Cheynier & Rigaud, 1986; Pnce et at., 1995). Flavonols have also been reported in the leaves (flmamouchi et al., 1996) and stems of grapevines (Souquet et a1.,2000; Jordão et al.,200lb). Flavonols occur in the upper epidermis of plant organs, consistent with their role in UV protection (Flint e/ a1.,1985; Beggs et a1.,1987), while anthocyanins are present in the cells of the lower epidermis (Asen, 1975). However, for copigmentation to occur, anthocyanins and flavonols would have to be present in the same cells. Hrazdina and Moskowitz (1980; Moskowitz & Hrazdina, 1981) identified both flavonols and anthocyanins in vacuoles of the outer epidermis of the ffape, but were not convinced these were associated as 137 copigments. Nevertheless, flavonols may play a more significant role as copigments in wine. The colour of young red wines is due to anthocyanin pigments extracted from the grape skin, but these compounds are unstable and much of the initial colour is lost during fermentation and maturation (Somers & Evans, 1979; Somers & Verette. 1988). The colour of older wines results from increasingly stable associations between anthocyanins and other flavonoid compounds to form pigmented polymers (.{agel & Wulf, 1979; Mccloskey & Yengoyan, 1981; somers & verette, 1988; Malien-Aubert et al., 2002; Mateus et al., 2002a). Pigmented polymers form slowly in the wine and anthocyanin- copigment complexes are considered intermediates that not only maintain the anthocyanins in the wine, but align them in favorable orientations for the formation of more stable associations (Brouillard & Dangles, 1994; Boulton, 2001). The most stable copigment associations occur between the flavonols quercetin and quercetin-3-O- glucoside and the main anthocyanin in red wines, malvidin-3-O-glucoside @aranac et al., 1997; Lambert,2002). These observations suggest that the critical factor in determining the extent of colour in mafure wine is the concentration of anthocyanins and copigments at harvest (Lambert, 2002). Thus, any process that increases the anthocyanin or flavonol content in grapes would be a valuable contribution to wine quality. Previous investigations into grape flavonols have largely focused on the flavonol composition of the ripe fruit (Cheynier & Rigaud, 1986) and the effects of sun exposure (Price et al., 1995). Haselgrove et al. (2000) examined flavonols earlier in berry development, reporting relatively high levels (0.15 mg/berry) at veraison, the onset of ripening, with little change towards harvest. Furthermore, while many genes involved in flavonoid biosynthesis have been cloned in Vitis vinifera (Sparvoli et al., 1994), this has 138 not been the case for flavonol synthase (FLS), the gene that encodes the biosynthetic enryme converting dihydroflavonols to flavonols (Spribille & Forkmann,1984; Lukacin et a1.,2003). This gene has been cloned in other genera including Petunia, Arabidopsis, Solanum and Citrus (Holton et a1.,1993b; Pelletier et al., 1997; van Eldik et al', 1997; Moriguchi et al., 2002). Here, the cloning and characterisation of genes encoding flavonol synthase from grapevine together with a study of flavonol accumulation in grapes during berry development are presented. 4.2. Materials and Methods 4.2.1. Sampling of Grape Tissues Grapevine tissues from Vitis vinifera L. cultivars Shiraz and Chardonnay \ryere collected from a commercial vineyard in V/illunga, South Australia as described in Section2.2.l. The developing fruit was sampled at weekly intervals (Table 2.1) over three seasons for Shiraz, and a single season for Chardonnay. Leaves from Shiraz grapevines were collected as described in Section 3.2.1 and early vegetative stages from bud-burst towards flowering were collected from Shiraz grapevines during the 2000-2001 season. At bud- burst the emerging buds were excised from the canes. On successive weeks the developing shoots were collected, the growth stages of the developing tissues were determined according to the modified Eichorn-Lorenz (E-L) scheme (Coombe, 1995). Samples corresponded to E-L Stage 4 (buds), E-L 7 (1't separated leaf; -16V), E-L 9 (shoot with 2-3 leaves separated), E-L t2 (shoot with 4-5 leaves separated), E-L 13 (developing inflorescence; -14V), E-L 14 (expanding inflorescence; -13V) and F-L 17 139 (unopened flowers; -10V). All samples were immediately frozen in liquid nitrogen and stored at -80"C for later analysis. Roots of Shiraz vines were collected from glasshouse grown potted vines. Plants were removed from pots and the soil shaken free from the root ball. The roots were then washed to remove the remainder of the soil and blotted dry. Root samples of l-2 cm in length were cut from the root tips, these were frozen immediately in liquid nitrogen and stored at -80oC. 4.2.2. Extraction and Identification of Flavonols In order to determine the pattern of flavonol accumulation in the grape berry and to target cloning of a grapevine homologue of flavonol synthase the flavonol content of the fruit was investigated. The flavonol synthase enzyme catalyses the desaturation of the dihydroflavonol (see Figure 1.6) to produce the flavonols gpribille & Forkmann, l9B4; Forkmann et a1.,1986; Holton et al., 1993b). However, the flavonol aglycones (Figure 1.3) rarely occur in plant tissues, with the majority of flavonols present as glycosides, commonly glucosides and glucuronides (Markham, 1989). Thus, flavonol accumulation was determined as the accumulation of the flavonol glycosides. While the aglycones are relatively non-polar and readily extracted in organic solvents such as methanol, ethanol or acetone (Markham, 1975), the glycosides are comparatively more polar and aqueous combinations of these solvents are more effective in extracting flavonol glycosides (Markham, 1982). While a range of solvents are suitable for extraction of flavonol glycosides from plant material, it was considered prudent to utilise here the method employed by others investigating flavonol composition in grapevines to enable valid comparisons between experimental outcomes. While Cheynier and Rigaud (1986) used 140 methanol (0J% HCI) as a solvent, subsequent investigations (Price et al., 1995; Haselgrove et a1.,2000) have employed acidulated ethanol. Thus, for the HPLC analysis of flavonol composition in grdpevines, samples collected in the field, frozen and stored'at -80'C were ground to a fine powder under liquid nitrogen. Three separate aliquots (analysis replicates) of 0.1 g ìüere extracted in 1.0 mL of acidulated ethanol (Price et al., 1995) for 5 minutes at70"C followed by 20 minutes at room temperature with occasional vortexing. Extracts were then centrifuged (15 min at 13,000 x g) and 200 ¡rL of the supernatant transferred to HPLC auto-sampler vials. A range of extraction durations and temperatures were trialed to determine effectiveness of this method. It was observed that heating for five minutes yielded the same level of flavonol glycosides as a 12 hour extraction at room temperature and considerably more flavonols than 20 or 60 minute extractions at room temperature. Extraction at low temperature (0"C) was not effective, even after 12 hours. Repeated extraction was effective for extracts made at room temperature for 20 and 60 minutes, but no additional flavonols were detected in extracts that had initially been heated at 70'C for five minutes. \When triplicate samples were allowed to return to room temperature after extraction at 70"C and prior to transfer to autosampler vials, more consistent results were observed suggesting some loss of solvent through evaporation. While both Price et at. (1995) and Haselgrove (1997) employed a method for HPLC separation of the flavonol glycosides that utilised phosphoric acid and with an acetonitrile gradient, issues associated with the disposal of acetonitrile waste encouraged the adoption of acetic acid with a methanol gradient Çheynier & Rigaud, 1986). Samples were analysed using a Hewlett Packard HP 1100 high-performance liquid chromatograph (HPLC) with two Merck (Germany) ChromolithrM analytical columns in series protected t4t by an SGE (Australia) C-18 guard column. The separation used 0.2o/o acetic acid (solvent A) with a methanol (solvent B) gradient (gradient of solvent B: zero min,24Yo;10 min, 28o/o; l0.l min, 1007o; 13 min, 100%;13.1 min 24%) with a flow rate of 5 ml/minute. The major flavonol peak was identified by comparison of the elution time and absorbance spectra with a commercial standard of quercetin-3-O-glucoside obtained from ExtrasyntheserM (France). In addition, the main flavonol peak was collected and analysed by mass spectroscopy (ESI-MS). The integrated absorbance at 353 nm was used to determine the concentration of flavonol glycosides in each sample (25 pL injection) expressed as quercetin-3-O-glucoside equivalents. Where present, kaempferol-3-O- glucoside was identified by comparison of elution time and absorbance spectra with a kaempferol-3-O-glucoside standard from ExtrasyntheserM (France). Concentrations of quercetin-3-O-glucoside and kaempferol-3-O-glucoside equivalents were calculated from a standard curve prepared from commercial standards. 4.1.3. Cloning of Flavonol Synthase For the cloning and expression of flavonol synthase genes grapevine material frozen in liquid nitrogen in the field and stored at -80oC, was ground to a fine powder in liquid nitrogen using a pre-frozen mort¿r and pestle. Genomic DNA was extracted from frozen Stage 2 leaf tissue following the method of Thomas et al. (1993). Total RNA was extracted from 1.0 g of the frozen powdered tissue in 7.0 mL of hot borate buffer as described by Wan and Wilkins (1994). Final pelleted RNAs were resuspended in Tris-EDTA (TE) and stored at -40oC in a solution containing 1/10th volume sodium acetate and 3 volumes of 100% ethanol. Two 5.0 pg aliquots of total RNA for each sample \Mere pelleted by centrifugation at 13,000 x g for 20 minutes at 4oC, t42 the pellet was rinsed in 70o/o cold ethanol and allowed to air dry. First strand cDNA synthesis was performed using the 3'-RACE adapter primer and Superscript IIrM (Invitrogen Life Technologies, Aushalia) using niethods recommended by the supplier in afnalreaction volume of 25 ¡tL. After heat inactivation of the reverse transcriptase, two reactions were combined to give a working 50 pL stock and stored at-20oC. cDNA clones encoding flavonol synthase (FLS) were obtained by PCR of cDNA prepared from RNA isolated from Shiraz flowers. Degenerate primers were designed based on FLS sequences for other dicotyledonous plants in the GENBANK database. The forward primer was 5'-GCGAATTCGAA/GAAIGAA/GCAA/GCCIGC-3' (FLS9), designed to the protein sequence EN/KEQPA and the reverse primer was 5'- ACIGGCCAIC/GA/TCATICG/TIGTC/TTT-3' (FLS6), designed to the protein sequence KTRMSWPV. The PCR was carried out according to the method of Frohman et al. (1988) and a predicted band of approximately 800bp was obtained and cloned into the pGEM-T Easy vector from PromegarM (Australia). A number of independent clones were sequenced and one clone (FLS36) was sequenced in both directions and found to encode a 797bp fragment with high homology to FLS seqrlences in the database. This sequence was named as pVvFLSl (GenBank Accession No. 4Y257978). A second gene fragment encoding FLS was identified by PCR from gtapevine genomic DNA. The equivalent cDNA was amplified from the flower cDNA using a gene-specific primer based on the genomic sequence. Flower cDNA was amplified by PCR using the specific primer 5'-AATCCTCCTTCTTACAGGGATGCG-3' (FLSIO) and the oligo-dT adapter primer 826 described by Frohman et al. (1983). This yielded a band of approximately 700bp that was cloned and sequenced as described above to yield a706bp t43 cDNA fragment (FLS26), also encoding FLS. This second sequence was named pVvFLS2 (GenBank Accession No. AY257979). 4.1.4. Expression of Flavonol Synthase Genes Gene expression was determined by real-time PCR analysis. cDNA syntheses for all RNA samples to be analysed were performed simultaneously. Reagents required for the reverse transcription (RT) reactions were made as a master mix and aliquoted into each reaction in a single step to avoid pipetting errors that may have affected the efficiency of the RT reactions in different samples. Quantiøtive real-time PCR was performed on a Rotor-Gene 2000rM (Corbett Research, Australia) real-time PCR machine using SYBR green as the system of detection for double stranded PCR products. Primers homologous to the grapevine cDNAs encoding VvFLSl, VvFLS2 and VvUbiquitinl were designed using primer 3 (http://www- genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi). The primer pairs VvFLSI Forward: 5'-CAGGGCTTGCAGGTTTTTAG- 3', vvFLSl Reverse: 5'- GGGTCTTCTCCTTGTTCACG- 3' and VvFLS2 Forward 5' TGAGCTGGCTATAGGTCCTC- 3' and VvFLS2 Reverse 5' TGCATGTACACTGGAAAAGG- 3'were used to generate PCR products of l54bp and 195bp respectively. The VvUbiquitinl transcript was detected using the primers ID vvubiquitinl Reverse: 5'-AACCTCCAATCCAGTCATCTAC- 3' and vvubiquitinl Forward: 5'-GTGGTATTATTGAGCCATCCTT- 3' which ampliff a product of lS2base pairs. All primer pairs were used under identical conditions, 95oC for 5 minutes followed by 35 rounds of amplification (95oC/30sec, 57oCl30sec, 72oC/30sec). Each PCR reaction 144 contained 1 x SYBR Green PCR Master MixrM (Applied Biosystems, U.K.), 30 nmoles of each primer and 1/50th volume of cDNA stock in a final reaction volume of 25 pL. The VvFLSI and VvFLS2 mRNAs was quantified as a proportion of TvUbiquitinl usrng the Rotor Gene 2000rM software. 4.1.5. Southern and Northern Blot Analysis Southern and Northern blot analysis was performed as described by Sambrook et al. (1939) using 0.4 M and 50 mMNaOH respectively as transfer buffer. Nucleic acids were transferred onto Zeta-Probe'* nylon membranes (Bio-Rad, Australia) and fixed as recommended by the supplier. Membranes were prehybridised in 7o/o SDS, 0.25 M Sodium Phosphate (pH 7.0), 1.0 mM EDTA at 65oC for 2 hours. DNA fragments specific for VvFLSl and VvFLS2 were excised from pGEM T-EasyrM (Promega, Australia) clones by digestion with EcoRI according to the instructions of the supplier. Vector and insert fragments were resolved on lo/o agarose gels and the insert bands removed under tIV light after staining with ethidium bromide. Insert DNA was purified from the gel fragment using a QlAquickrt gel extraction kit (QIAGEN, Australia) according to the manufacturers instructions. Inserts were radioactively labeled using a RediprimeruIl labeling kit (Amersham Biosciences, Australia), added to the prehybridisation solution after heat denaturation and hybridised overnight at 65oC. Membranes were rinsed in2x SSC, 0.1% SDS at room temperature before being washed in 1.0 x SSC, 0.1% SDS for 20 minutes at 65oC and 0.1 x SSC, 0.1%SDS for 20 minutes. Membranes were then rinsed in 2 x SSC at room temperature before being exposed to BiomaxrM (MS) frlm (Eastman Kodak Co., USA) at -80oC with appropriate intensiffing screens. t45 4.3. Results 4.3.1. Cloning of Flavonol Synthase Genes The flavonoid pathway produces a range of secondary compounds including anthocyanins, flavonols and tannins. The first step in the branch of this pathway that leads to the synthesis of the flavonol derivatives is catalysed by the enzryme flavonol synthase (FLS), a dioxygenase that catalyses the desaturation of the 'C' ring of dihydroflavonols (Stafford, 1990). If FLS is hanscriptionally regulated as appears to be the case with other flavonoid pathway enzymes (Boss e/ al., 1996a), it is anticipated that flavonol accumulation would be preceded by expression of the gene encoding FLS. To determine expression levels of the genes encoding FLS in grapevine tissues, we first isolated cDNA clones encoding FLS to provide gene probes. RNA extracted from Shiraz flowers was used to make cDNA by reverse transcription and this was amplified using the polymerase chain reaction (PCR) and the DNA fragments obtained were cloned and sequenced in both directions. This yielded two distinct grapevine FLS clones: pVvFLSl, a 797bp fragment of FLS encoding 265 amino acids in the open reading frame of the gene, and pVvFLS2, a706bp fragment encoding 185 amino acids at the C-terminal end of the protein plus a 3'-untranslated region of l51bp. The two fragments shared a 425bp overlap and pVvFLS 1 and pVvFLS2 had 79% sequence identity at the nucleotide level in this overlapping region. The two grapevine FLS sequences have been deposited in the GenBank database (Accession numbers AY257978 and AY257979). The two grape FLS genes had high protein sequence homology to other FLS genes in the two conserved dioxygenase domains (data not shown), indicating that they are part of the dioxygenase family (van Eldik et a1.,1997; Monguchi et al., 2002). Both grape sequences had high homology to FLS genes in the databases and the nearest matches at the amino acid level r46 were to Petunia FLS (222543;73% identity to pVvFLSt,T4yo identity to pVvFLS2) and Citrus FLS (84436554; 73% identtty to pVvFLSl, 70yo identity to pVvFLS2). A number of other degenerate prirhers were designed based on plant FLS sequences and used in PCR of the oDNA from grape leaves and flowers, however no additional FLS genes were identified. 4.1,2. Southern Blot Analysis Southern blots of Shiraz genomic DNA restricted with HindIII and hybridised with either the VvFLSI or VvFLS2 probes demonstrated that each probe could bind to four distinct bands of the HindIII restricted DNA (Figure 4.1). Analysis of the two cDNA sequences shows that HindIII cuts the VvFLSI clone at least twice and the VvFLS2 clone at least once. As both clones are only partial oDNA sequences it is not known if the remaining coding sequence or the introns of the respective genomic sequences contain further HindIII sites. Given the high stringency washing VvFLS VvFLS2 conditions used in this procedure, it is unlikely that 1 either probe could cross hybridise to the target + < genomic sequence of the other. In the absence of < + further analysis we cannot absolutely confirm that + < the two sequences exist as single copy genes in the < + Shiraz genome. However, given the fact that no other FLS-like clones were identified during the Figure 4.1. High stringency wash (0.1 x SSC @ 65'C) of Southern cloning of the VvFLS cDNAs and that HindIII does blot of Vitis vinifera clones of the flavonol synthase gete, VvFLSI and VvFLS2 restricted with HindIII. cut the FLS sequences, this is the most likely conclusion. 147 4.1.3. Identification of Grapevine Flavonols Flavonols in Vitis vinifera L. Quercetin - glucoside cultivars Chardonnay and Shiraz f, 80 pferol - glucoside E 60 A were identified initially by o) o 40 ([L comparison of their elution cL o 20 att -o times and absorbance spectra 0 with co ercially available 02 46810 12 Elution Time (minutes) standards of quercetin-3:O- 380 glucoside or kaempferol-3-O- g60 B c) glucoside. Figure 4.2A shows (u840 -o (t,b20 the separation ofthese stand s -o using conditions the IIPLC 02 46810 12 Elution Time (minutes) described above. Typically, a 16000 glucuronide extracts (Figure 4.28) showed 14000 Quercetin 12000 479.0 Quercetin two peaks corresponding to the 10000 C 8000 quercetin-3-O-glucoside and 6000 4000 kaempferol-3 - O-glucoside 2000 standards, based on their elution 200 300 400 500 600 700 m/z,amtt times and absorbance spectra. In Figtte 4.2. Identification of flavonols in vitis vinifera by HPLC-MS (A¡s¡) A). HPLC separation of commercial all s les, the peak standards of quercetin-3-glucoside and kaempferol-3-glucoside (25 ¡tL injection,50 pglml-); B). HPLC separation of extract corresponding to quercetin-3- from Shiraz flowers (-8V, l6 November 2000;25¡tL injection); C). Mass spectrum of the major flavonol peak in Vitis vinifera tissues corresponding to quercetin-glucoside; glucoside accounted for most of three ionisation products were apparent, quercetin (M+If, mlz 303.2), quercetin-3-O-glucoside (M+H*, mlz 465.2) and quercetin-3- the flavonols. The separation O-glucuronide (M+tf, mlz 479.0). 148 presented here (Figure 4.2Ã), is that of Shiraz flowers, nine weeks pre-veraison and was the only time during Shiraz berry development when the kaempferol glycoside could be detected. Kaempferol derivatives were not detected in Chardonnay berries or flowers. The major flavonol peak, corresponding to quercetin-3-O-glucoside was collected and analysed by mass spectrometry. The mass spectra from this peak (Figure 4.2C) showed three ionisation products present, corresponding to the aglycone quercetin, quercetin-3-O- glucoside and quercetin-3- O-glucuronide. HPLC analysis did not detect the flavonol aglycone, quercetin. While a peak was present in the HPLC separation at the approximate time that quercetin eluted (elution time), spectral analysis of this peak was unable to identiff any flavonol. The absorbance spectra of a commercial standard of the A 5 1000 flavonol aglycone, g 800 I 600 quercetin is presented C S 4oo The o in Figure 4.34. ts 2oo 0 absorbance spectra of 300 350 400 500 Wavelength (nm) flavonols has a distinctive absorbance l 1200 B E 1000 maxima at 353 nm. o co 800 (E -o 600 Five absorbance oL at 400 -o spectra through the 200 0 late eluting peak in Wavelength (nm) the HPLC separation Figure 4.3. Absorbance spectra of commercial quercetin standard @) and five absorbance spectra through the peak in a grape skin extract that of grape samples does elutes at a similar time to that of the quercetin standard (B). 149 not show any spectra consistent with that of the flavonols (Figure 4.38). 4.l.A.Berry Development, Flavonol Accumulation and FLS Expression in Chardonnay Berries Samples of Chardonnay grapes were collected during the 1999-2000 growing season. The pattern of berry development and the accumulation of flavonols and expression of VvFLSl, the gene encoding a grapevine flavonol synthase, during the 1999-2000 season are presented in Figure 4.4. Flowering occurred in the week of 21 October 1,999 and veraison occurred nine weeks later, around 23 December 1999 (Figure 4.4A). Six weeks after veraison, the Chardonnay was harvested at 23.2"8ix on 03 February 2000. Berry weight increased slowly from flowering until six weeks pre-veraison, however this represented a l2-fold increase in berry weight (Figure 4.4A). From six weeks pre- veraison berry weight increased steadily until one week pre-veraison, then remained static until veraison when berry growth reconìmenced for a further three weeks. During this time the level of sugar in the berry, measured as total soluble solids ("Brix), increased steadily (Figure 4.4A). Berry weight and total soluble solids were constant for the next two weeks, while in the final week before harves! there was a decrease in berry weight and an increase in total soluble solids. This observation was consistent with water loss from the ripe fruit. In Chardonnay berries, only glycosides of quercetin were detected. At flowering there was a high concentration of flavonols (1.46+0.130 mglg fresh weight), however as the berry began to develop there was a 5-fold decrease in flavonol concentration (Figure 4.48). This decline continued until two weeks pre-veraison, after which time the level of flavonols per gram of tissue remained relatively constant until harvest (0.0310.00 5 mg/Ð. 150 During this time, the berry was increasing in size suggesting an overall increase in total flavonols in the berry. This was the pattern observed on a per berry basis (Figure 4.4C), where total quercètin 1.50 30 A 1.25 glycosides per berry were o 1.00 a 20 E l a' .9 r Berry Weight .Ex o 0.75 o relatively constant for the first r 'Brix =à 0.50 l0 o @ few weeks (0.009 mg/berrl') 0.25 0.00 0 followed by a two-fold -9 -8 -7 -6 -5 -4 -3 -2 -1 0 I 2 3 4 5 6 Weeks Post-Varaison 1.60 B increase at five weeks pre- Ð 1.20 o E mdberrl'). veraison (0.019 Io 0.80 oc o This level was maint¿ined until tr 0.40 0.00 one week pre-veraison when -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 Weeks Post-Veraison 0.05 c flavonols began to accumulate E 0.04 oo o 0.03 once more. At one week post- E .9. o c 0.02 o veraison, quercetin glycosides o I 0.01 increased the 0.00 again to -9-8-7 -6 -5 4 -3 -2-1 0 1 2 3 4 5 6 0.4 Weeks Post-VeÍaison maximum level observed in c D .9 0.3 oI the berry during development ôx ul 0.2 (t J (0.035 mdberry). From one 0.1 $ 0.0 week post-veraison the -9 €-7 6 -5 -4 -3 -2 -1 0 I 2 3 4 5 6 Weeks Post-Veraison flavonol level remained Figure 4.4. Berry development, flavonol accumulation and flavonol synthase gene expression in Vitis vinifera L. cv' relatively constant until Chardonnay. Flowering occurred around 2l October 2000 (-9V) with veraison (V) around 23 December 2000 and commercial harvest. harvest on 03 February 2001 (6V). A)' Chardonnay berry development, berry weight (A) and total soluble solids oBrix expressed as C); n). Quercetin glycosides (mg/g fresh weight of skin tSEM); C). glycosides (mg/berry The expression of a grapevine Quercetin ISEM); D). Expression of the flavonol synthase gene (VvFLSI) in developing grape skins relative to the expression of gene encoding the flavonol TvUbiquitinl. 151 synthase biosynthetic enzyme WFLSI) in Chardonnay berries during the 1999-2000 season is shown in Figure 4.4D rclative to an internal control (vUbiquitinl) whose expression was constant throughout berry development (data not shown). In flowers, nine weeks pre-veraison, a high level of VvFLSI expression was observed, which declined markedly after flowering and remained low until veraison. After veraison, the level of VvFLSI expression increased steadily towards harvest with a lO-fold increase in VvFLSI expression dwing this time. The final level of VvFLSI expression in grape berries at harvest was similar to that at flowering. The expression of VvFLS2 was not determined in Chardonnay berries. 4.1.5. Berry Development, Flavonol Accumulation and FLS Expression in Shiraz Berries Primarily, results from the 2000-2001 season are presented here with the previous and successive seasons included for comparison. Shiraz grapes were collected over three seasons; 1999-2000, 2000-2001 and 2001-2002. Table 4.1 shows the flowering and veraison dates as well as the date on which berries attained 24"Bnx for Shiraz grcpes over the three suceessive seasons. Berry weight, skin weight and flavonol levels are also indicated. The patterns of berry development, flavonol accumulation and expression of VvFLSl, a grapevine gene encoding flavonol synthase during the 2000-2001 season are presented in Figure 4.5. Berry growth during the 2000-2001 season was measured as berry weight and showed a six-fold increase from seven weeks pre-veraison (fruit-set) until four weeks pre-veraison when berry growth slowed (Figure 4.54). Berry weight reached a maximum five weeks post-veraison then declined until commercial harvest at nine weeks post-veraison. From veraison, sugars in the berrl', measured as total soluble solids (Brix), also increased steadily reaching 29.3'Brix at commercial harvest. This ts2 Table 4. 1. Seasonal variation in the timing of flowering, veraison and ripening in Vitis vinifera L. cv. Shiraz over th¡ee successive seasons, at Willunga, SA. The date when berries reached approximately 24"Brix is shown. Berry weight, skin weight and levels of quercetin glycosides (flavonols) at24"Btix is shown for each season expressed per gram fresh weight of skin and per berry (+$þ¡4, .:3¡ Season Flowering Veraison =24"Brix Berry Skin Flavonols tt2í"Brix weisht weight (e) (e) mg/g skin mdberry 1999-2000 04-Nov-99 06-Jan-00 l7-Feb-00 1.36 0.04610.001 2000-2001 16-Nov-00 ll-Jan-Ol 01-Mar-01 1.27 0.39 0.08610.002 0.03310.001 2001-2002 23-Nov-01 3l-Jan-02 08-Mar-02 1.16 0.44 0.07110.009 0.02610.003 Note: Shiraz grape skin was not separated from the berry during the first season, 1999-2000. The grapes reached 24"Bnx six weeks post-veraison in 1999-2000, seven weeks post-veraison in 2000- 2001, and five weeks post-veraison in200l-2002. increase in total soluble solids from five weeks post-veraison coincided with a decrease in berry weight, which was consistent with previous reports of berry shrivel in ripe Shiraz (McCarthy, 1999; McCarthy & Coombe, 1999). The concentration of quercetin glycosides during berry development was highest in the unopened Shiraz flowers, nine weeks pre-veraison (1.6010.050 mg/g fresh weight, Figure 4.58). At flowering (50% cap-fall) the following \ /eek, the level had fallen to 1.31+0.069 mglg and in the following week to fruit-set, a substantial decrease was observed to 0.17+0.005 mglg, followed by a smaller decline in the next week. From this time until harvest the flavonol concentration remained relatively constant (0.08+0.004 mg/g fresh weight of skin). On a per berry basis (Figure 4.5C), there was an initial increase in total quercetin glycosides in the developing berries from fruit-set (0.011+0.001 mglberry) until five weeks pre-veraison. Although some fluctuations were observed, this level was maintained until veraison (0.013t0.001 mg/berry). At veraison, flavonols in the Shiraz berry began to increase and continued to do so until six weeks post-veraison 153 (0.04910.001 mg/berry). For the l.ô 30 A o '1.2 following two weeks flavonols 20 c Berry Weight .9 r a o 08 ..¡ .é ì r "Brix t P appeared to decline, however the ð 10 in o J o o.4 ¡' a final weeþ total flavonols in the 0.0 -9-8-7-6-5-4-3-2-.1 0 1 2 3 4 5 6 7 I I Weeks Post-Veraison berry increased again. 1.8 B 6 1.5 Þ '1.2 The expression ofthe gene encoding ã0,c 9 o.e --g the flavonol s¡mthase biosynthetic 0.3 0,0 -9-8-7-6-54-3 -2-1 0 I 2 3 4 5 6 7 8 I enzyme (VvFLSI) during berry Weeks Post-Veraison 0.05 c È 0.04 development in the 2000-2001 oo o 0.03 E o 0.02 season is shown in Figure 4.5D o 0.01 Ll! relative to the expression of 0.00 -9-8-7 € -54-3-2-1 0 1 2 3 4 5 6 7 I I Weeks Post-Veraison 05 VvUbiquitinl whose expression c D '3 o.t remained constant throughout berry E o.e ul ô 0.2 J development (data not shown). In \ o.t 0.0 -9-8-7-ô-54.3 -2 -1 flowers prior to cap-fall, a high level 0 1 2 3 4 5 6 7 I I Wseks PosþVeraison of VvFLSI expression was Figure 4.5. Berry development, flavonol accumulation and flavonol synthase gene expression in Vitis vinifera observed, which declined markedly L. cv. Shiraz. Flowering occurred around l6 November 2000 (8 weeks pre-veraison) with veraison around I I January 2001 and commercial harvest on 13 March 2001 at flowering (eight weeks pre- (9 weeks post-veraison). A). Shiraz berry development, berry weight (1,) and total soluble solids expressed as veraison) and during the following 'Brix [); B). Quercetin glycosides (mg/g fresh weight of skin tSEM); C). Quercetin glycosides (mg/berry week, remaining at a low level from tSEM); D). Expression of the flavonol synthase gene (VvFLSI) in developing grape skins relative to the expression of VvUbiquitin 1 . six weeks pre-veraison until veraison. After veraison, the level of VvFLSI expression began to increase slowly, with a noticeable increase three weeks post-veraison. This was followed by an additional increase from four to six weeks post-veraison. From two to six weeks post-veraison there ts4 was a lQ-fold increase in VvFLSI expression. This was followed by a decline and then a further increase over the final two weeks of berry development until harvest. The final level of VvFLSI expression was marginally higher than that seen in unopêned flowers. VqFLS2 was constitutively expressed in all the developing berry skin samples but at levels 104-10s times lower than that of VvFLSI in corresponding samples. The expression of VvFLS2 did not vary in any of the skin samples that were assayed and was thus incongruent with the observed pattem of flavonol accumulation. These results were confirmed by Northern blot analysis (data not presented). 4.1.6. Seasonal Variation in Flavonol Accumulation in Shiraz Figure 4.6 shows the variation in total soluble solids fBrix) and flavonols (quercetin glycosides) across three seasons,1999-2000, 2000-2001 and 2001-2002 in Vítis vinifera L. cv. Shiraz berries. In the 1999-2000 season, flowering occurred during the week of 04 November 1999 with veraison nine weeks later around 06 January 2000 and commercial harvest seven weeks later on 24 February 2000 (Table 4.1). In the 2000-2001 season, flowering occurred two weeks later than in the 1999-2000 season, during the week of 16 November 2000. Veraison occtuïed eight weeks after flowering, around 1l January 2001, with commercial harvest after a further nine weeks on 13 March 2001. During the 2001- 2002 season flowering occurred later than in either of the preceding seasons, around 23 November 2001. In addition, the time between flowering and veraison was longer than in the previous seasons at ten weeks, veraison occurring at the end of January (31 January 2002). However, in the 2O0t-2002 season the time between veraison and commercial harvest was the shortest recorded during these investigations at six weeks. Shiraz grapes were harvested at 26.4 "Bixon 15 March2002 (Table 4.1). 155 Ripening of the fruit 30 1999-2000 ¡ 25 progressed -.r-.2000-2001 in similar 20 x -#2001-2002 I fashion each seasón with dt 5 0 sugar accumulation fBrix) 5 A 0 increasing from veraison -9-8-7-6-5-4-3-2-1 0 1 2 3 4 5 6 7 I 9 Weeks Post-Veraison 0.06 towards harvest, reaching Là 0.05 c) -o 24"Bix between o) 0.04 E five and E ! al, 0.03 õ ! seven weeks post-veraison oE 0.02 e _g II 0.01 I B (Table 4.1, Figure 4.6A). 0.00 -9-8-7-6-5-4 -3-2-1 0 1 2 3 4 5 6 7 I 9 The pattern flavonol of Weeks Post-Veraison (quercetin glycoside) Figure 4.6. Seasonal variation in the accumulation of sugars and quercetin glycosides in Shiraz ber.,' skin across th¡ee experimental seasons, 1999-2000,2000-2001 and 2001-2002. 1). accumulation Ìvas that Sugar accumulation ('Brix); B). Quercetin glycoside accumulation (mg/berry). observed in Shiraz grape skin during the 2000-2001 season was observed over all three of the seasons studied here (Figure 4.68). When examined on a per berry basis, total quercetin glycosides per berry six weeks post-veraison. Towards harvest, there was some variation in the levels of flavonols between seasons. /â When considered at similar levels of ripeness (24'Brix), the highest level of flavonols per berry was observed in the first season (1999-2000), ile the lowest level was observed in the final season (2001-2002). However, it should be noted that berry weight at this time was also highest in the first season and lowest in the third season (Table 4.1). 156 In the two seasons where skin was collected (2000-2001 and 2001-2002), the level of flavonols (mglg fresh weight of skin) was higher in 2000-2001 than in the following season. To some extent this is reflected in the level of flavonols per berry (Table 4.1), but it should also be noted that in the earlier season skin weight as a percentage of berry weight was lower (30.1%) than in 2001-2002 (37.9%). 4.1.7. Flavonol Accumulation and FLS Expression in other Tissues of Shiraz Figure 4.7 shows the comparative levels of flavonols present in other tissues of the gmpevine and expression of the gene encoding flavonol synthase WFLSI) in the same tissues. Immediately after bud-burst, at the beginning of the growing season for grapevines (E-L stage 4), a sample of the emerging buds was taken with successive developmental stages of young shoots made at weekly intervals. All of these tissues had high levels of quercetin glycosides, but kaempferol glucoside \ryas not detected. The developing inflorescence (E-L stage 13) had the highest flavonol concentration of any stage in berry development (2.11110.063 mglg) and, accordantly, the highest level of VvFLSI expression (Figure 4.78). The decline observed in the level of quercetin glycosides in the developing fruit (Figure 4.58) was more dramatic when the initial level in the inflorescence is considered. A high level of flavonols was also detected in the anthers of Shiraz flowers (1.200+0.025 mdg). Compared with the vegetative tissues, the levels of quercetin derivatives in the berry three weeks post-veraison and at harvest, were low (Figure 4.7A). There were no flavonols detected in either the seeds or the flesh of Shiraz berries sampled three weeks post-veraison, although a trace of VvFLSI expression could be detected in flesh, while there were significant levels in the skin throughout development as described above. Grapevine tendrils and the pedicel of the berry also contained high levels of flavonols, with kaempferol glucoside also detected in these 157 tissues, around 5o/o of total flavonols in tendril and 2%o in the pedicel. The level of flavonols in newly opened buds, tendrils, pedicels and developing inflorescences was significantly higher than at any time during berry development from fruit-set until harvest. Although there were clearly detectable levels of TvFLSI in emerging buds, tendrils and developing flowers, very little VvFLSI expression was detected in the pedicel or root, despite the presence of flavonols in these tissues. There was no detectable expression of VvFLS2 in any of the vegetative tissues analysed here. Shiraz leaves had a high concentration of flavonols, around 2.5 times greater than that observed in the flowers. In berries, almost all of the flavonols were quercetin glucosides, but Shiraz leaves contained both quercetin and kaempferol glucosides (Figure 4.71r^).In leaves, approximately l0o/o of total flavonols were kaempferol glucoside. A high level of flavonols was observed in newly emergent Shiraz leaves (quercetin glycoside 3.16+0.27 mdg; kaempferol glycoside 0.3910.04 mg/g fresh weight). As the leaf expanded this level declined as seen in the second leaf stage (Figure 4.7A). The level of flavonols increased again as the leaf continued to expand reaching a level similar to that in the youngest leaf. However, the level of kaempferol glucoside in the older leaves was lower than that observed in the youngest leaf. Roots from glasshouse grown Shiraz vines contained low levels of flavonols compared to other vegetative tissues, but similar levels to the skin (0.069+0.004 me/e). The VvFLSl, but not VvFLS2, mRNA was detected at all stages of leaf development with the transcript becoming gradually more abundant as the leaves matured. This level of transcript expression \¡/as commensurate with the ongoing accumulation of flavonols in the expanding leaf. 158 VvFLSI Expression Flavonoló (mg/g) P P o o I J J OOJJNN)G)O)Þ ifEå?säs:Ë N b, @ o i..r b ¿n b t¡b i¡ b t¡ b gEËËsä"3si o 5 ã'9P.9ão.qF\ Buds oHØP^Þ-' Buds yF.Þ.+jÈi5 T O ô Y-a Tendril Þ iEËrÈt Tendril tÉ ñ:? uku ËEnieì Pedicel Pedicel lnflorescence ÉãääsäggËgeä lnflorescence '¡ri^^if9çt t¡ * Þì Flower J.995'ã3ô ò Flower : 5 ;E * d {.-JdâÞ¡å^ o v'6 (D f+ not determined Anthers ¿oc)0 Anthers ?EÊãgg Post-Veraison Seed none detected q.v!ø-5=. Post-Veraison Seed äFã ãEåÈ å 3 none detected Post-Veraison Flesh Post-Veraison Flesh iä$ãå3gqH apå$ãs \o u"ã$ì none detected Root i€* Root 4.4. Discussion Flavonols and anthocyanins are predominantly synthesised in the skin of red grapes whereas tannins are present in both the skin and seeds QYIazza & Miniati, 1993; Souquet et al., 1996). No flavonols were detected in the seeds, nor in the flesh, of Shiraz grapes (Figure 4.7) suggesting that flavonols are only present in the skin, which was consistent with previous reports (Souquet et a1.,1996). InVitis vinifera L. cv. Shiraz grapes at harvest over three seasons, the major flavonols detected were quercetin-3-glucoside and quercetin-3-glucuronide (Figure 4.2). In Chardonnay grapes the major flavonols were also quercetin glucosides (data not shown). In neither Chardonnay nor Shiraz, were flavonols other than quercetin glycosides detected at harvest, although kaempferol glucosides were detected in Shiraz at flowering. We did not detect the flavonol aglycones quercetin and kaempferol in any of the grape berry samples analysed. Ribereau-Gayon (1964) reported the main flavonols in grapes as being the 3-glucosides of myricetin, quercetin and kaempferol, as well as the 3- glucuronide of quercetin. In addition to these, Cheynier and Rigaud (1986) also reported the glucuronides of kaempferol and myricetin, the glucosylgalactoside and glucosylxyloside of quercetin, and the galactoside and glucosylarabinoside of kaempferol, as well as the 3-glucoside of isorhamnetin, a 3' methylated quercetin derivative. Of these compounds, the glucoside and glucuronide of quercetin represented by far the greater proportion of flavonols in the fruit of Vitis viniferqL. cv. Cinsault (Cheynier & Rigaud, 1986). Macheix et al. (1990) later quantified the level of quercetin derivatives in grape berries, reporting levels to range from less than 0.01 to 0.1 mg/g fresh weight of skin. Pnce et al. (1995) also reported quercetin glucoside and glucuronide as the largest flavonol components in grape skin, with around 0.065 mg/g fresh weight of 160 skin in Vitis vinifera L. cv. Pinot noir at harvest (23"Brix). Subsequently, Haselgrove et at. (2000) reported levels of around 0.065 mg/g fresh weight of berry (0.056 mg/berry;) in Shiraz at harvest (24"Brix). The results presented here for Chardonnay and Shiraz berries at commercial ripeness are comparable with previous reports. In the 1999-2000 season, the level of quercetin glycosides in Shiraz (0.03410.001 me/g of berry fresh weight; 0.046+0.001 mg/berry; n:3) and Chardonnay (0.03110.003 mele of berry fresh weight; 0.047+0.005 mg/berry; n:3) were not significantly different on either a per gram (p:0.133) or a per berry (p:0.478) basis. These levels are slightly lower than those reported by Haselgrove et al. (2000) for whole Shiraz berries. The level of flavonols in Shiraz skin ranged from 0.071- 0.086 mgig fresh weight of skin at approximately 24"Blnx, which was slightly higher than reported by Price et at. (1995) in Pinot noir, but falling within the general range reported by Macheix et al. (1990). Over three seasons, there were similar levels of flavonol glycosides in Shiraz berries (mg/berry) at a similar level of ripeness (24"Brix, Table 4.1). In both Chardonnay and Shiraz berries, similar patterns of accumulation were observed. High flavonol concentrations (^glg of berry fresh weight) were recorded around flowering, followed by a decrease as the grape berry increased in size. However, after the initial decrease following flowering and fruit-set, the per gram level of quercetin glucosides in the fruit remained constant until harvest despite substantial increases in berry size. This indicated ongoing biosynthesis of flavonols during berry development and this was confirmed by the increase in flavonols per berry. The pattern of flavonol accumulation per berry suggests two main periods of flavonol synthesis, the first occurring around flowering and the second occurritrg after veraison. The post-veraison phase represented the greatest increase in flavonols per berry, with a substantial increase 161 observed in all three seasons commencing three to four weeks post-veraison (Figure 4). Despite minor variations in timing and amplitude, the similarity across seasons establishes a convincing pattem for flaVonol accumulation in grape berries that has not previously been reported. The increase observed in the final weeks of berry development was not observed by Haselgrove et al. (2000), although McDonald et al. (1998) reported that the levels of flavonols were higher in wines made from very ripe fruit, which is consistent with what we have observed here. Towards the end of berry development there was a notable increase in flavonol content during the final weeks of ripening. The post- veraison increase in flavonols occurred after the main period of anthocyanin biosynthesis, which occurred in the 2-3 weeks immediately post-veraison (data not shown). This increase in flavonols post-veraison might be related to copigmentation or to the role of flavonols as UV protectants, or involved in preventing the photobleaching of anthocyanins reported by Yamasaki et al. (1996). At flowering, significant amounts of flavonols were present in both Chardonnay and Shiraz gapes. This led us to examine the flavonol content in developing inflorescences and other vegetative tissues of Vitis vinifera L. cv. Shiraz. The level of flavonols in flowers was 10-15 times higher than that observed in the skin of ripe grapes, but was highest in the developing inflorescence (Figure 4.7A). Such high levels of flavonols would be consistent with a role for flavonols in UV protection and of particular importance in reproductive tissues (Lodish et al., 1995). High levels of flavonols were also observed in the anthers collected from Shiraz flowers and this might account for a considerable portion of the flavonols in the flower. The high level of quercetin glycosides in anthers suggests that flavonols might also play a role in grapevine pollen fertility as reported in tobacco and petunia (Ylstra et al., 1992; Ylstra et a1.,1994). 162 The bunch rachis of Vitis vinifera cv. Merlot has previously been shown to contain primarily quercetin-3-glucoside, around 0.2 mglg fresh weight, with quercetin glucuronide less than l0o/o of that amount and small amounts of the myricetin and kaempferol glycosides $ouquet et a1.,2000). While the relative proportions observed here are similar to those reported in Merlot stems by Souquet et al. (2000), the level of flavonols in Shiraz pedicels was around ¡wo-fold greater Q.a3 mglg) and in tendrils, six- fold higher (1.23 mg/g). Given the proposed role of these compounds in UV protection, the European (France) location of the previous work might well account for lower flavonol content compared to vines grown in South Australia. Previously, the flavonol composition of gtapevine leaves has also been studied. Mature grapevine leaves growing in Morocco were shown to contain all three flavonols, myricetin, quercetin and kaempferol, with quercetin derivatives accounting for 80-100% of the flavonol content (Imamouchi et al., 1996). Leaves of the fout Vitis cultivars, Alicante, Carignan, Cinsault and Grenache noir, were examined with leaves of the latter two containing only trace amounts of myricetin and kaempferol. Total flavonols in grapevine leaves ranged from 8.1 mgß ey weight in Grenache noir, to 38.6 mdg dry weight in Alicante (flmamouchi et a1.,1996). In Shiraz, we observed both quercetin and kaempferol glycosides with total flavonols around 3.5 mglg fresh weight (Figure 4A). Based on a water content of grape leaves of 90o/o or greater (Patakas et al., 1997), the levels reported here in Shiraz are comparable with those reported in cultivars grown in Morocco. The flavonol synthase (FLS) enzyme forms a branchpoint in the flavonoid pathway and genes encoding FLS have now been isolated from a ntrmber of plants. Holton et al. (1993b) confirmed the function of the petunia FLS gene by expressing it in yeast and 163 demonstrating that extracts of the yeast catalysed the conversion of dihydroquercetin to quercetin in the presence of the appropriate cofactors. The grapevine cDNAs VvFLSt and, VvFLS2 both had high sequence homology to the petunia arid other FLS genes indicating that they most likely encode FLS in grapevine. Previously, expression of genes encoding FLS has been detected in flowers but there have been few studies of flavonol synthesis in fruit. The petunia FLS gene was highly expressed in early developmental stages of flowers, but was not detected in leaves of petunia (Holton et al., 1993b). In potato, van Eldik et al. (1997) also found FLS expressed during flower development, particularly in the pistil, anther and petal tissues. Moriguchi et al. (2002) detected citrus FLS expression in flowers and young leaves but also in the juice sacs and peel of the fruit throughout fruit development. While little expression of VvFLS2 was detected in any grapevine tissues, VvFLSI was expressed in the buds, leaves and tendrils of the grapevine as well as in the flowers, developing fruit and the skin of ripening berries of both Chardonnay and Shiraz. The pattern of FLS gene expression in both Chardonnay and Shiraz berries between flowering and veraison was commensurate with the pattern of flavonol accumulation in the berry during that time. Expression of the VvFLSI gene at a high level at flowering suggests a period of intense flavonol biosynthesis coinciding with the high concentrations of flavonols detected in these tissues. The concentration of flavonols was observed to decrease substantially after flowering, due to a 2O-fold increase in berry size by weight, and VvFLSI expression also decreased after flowering. The concentration of flavonols (mg/g fresh weight) was relatively constant after veraison, but the amount of flavonols (mg/berry) increased, suggesting continued synthesis during the secondphase of berry growth. Expression of VvFLSI increased around veraison and particularly during the period from two weeks post-veraison until ripening was complete. 164 This expression of VvFLSI during ripening coincided with the increase in flavonols per berry in Shiraz (Figure 4.5D), although this was less noticeable in ripening Chardonnay beiries (Figure 4.4D). The level of flavonols in newly opened buds, tendrils, pedicels and developing inflorescences of Shiraz was significantly higher than at any time during berry development from fruit-set until harvest. Although there were clearly detectable levels of VvFLSl in emerging buds, tendrils and developing flowers, very little VvFLSI expression was detected in the pedicel or root. No expression of VvFLS2 was detected in these or any other vegetative tissue and it seems most likely that at the time of sampling, flavonol synthesis was complete in the grape pedicel and the root. Therefore, to determine expression of YvFLSI in these tissues, samples would need to be taken at different times during their development. An alternative, albeit less likely, explanation for the absence of VvFLSI expression in these tissues is that another FLS gene exists in grapevines that 1üe are yet to identiff and that expression of this gene accounts for flavonol synthesis in these tissues. This seems less likely as we were unable to ampliff additional gene homologues by PCR with a range of degenerate primers and only two bands homologous to the VvFL,Sl and VvFLS2 probes were detected by Southern blot analysis. Studies of anthocyanin biosynthesis in grapevines showed that genes encoding biosynthetic enzymes for much of the flavonoid pathway were expressed post-veraison, when anthocyanin synthesis occurs, but also early in berry development (Boss et al', I996a). This was considered particularly interesting, as expression of the pathway was not anticipated pre-veraison when anthocyanins were not being produced. Furthermore, Boss et al. (1996b) observed that much of the pathway was also expressed in other grapevine tissues that did not normally accumulate anthocyanins. An explanation 165 consistent with these observations would be that flavonoid compounds other than anthocyanins were being synthesised in these tissues. We have previously shown substantial quantities of tannins (flavan-3-ols and proanthocyanidins) to be prêsent in Shiraz berries from the earliest phase of berry development and increased towards veraison (Chapter 3). It is now clear that a significant level of flavonol biosynthesis is also occurring during berry development. Together, the spatial and temporal pattern of flavonol and tannin accumulation in grapes and expression of the FLS gene form the strongest indication yet that the expression of genes involved in flavonoid biosynthesis reported by Boss et al. (1996a; I996b) is related to flavonol and tannins biosynthesis in glapevrnes. To date, this is the most comprehensive study of flavonols in grapevines in that it has quantified flavonols throughout berry development. The pervasive nature of flavonols in the grapevine, and particularly the very high levels in juvenile and reproductive tissue, as well as the high levels in leaves are strong indications that the role of grapevine flavonols is UV protection. Such being the case, and the correlation between sun exposure and flavonol content reported by Price et al. (1995) and Haselgrove et al. (2000), viticultural practices that increase the sun exposure might significantly increase the flavonol content of the ripe fruit and the subsequent copigment pool in the wine. The level of flavonols in grapes is relatively low compared to the anthocyanins and other potential copigments such as proanthocyanidins, with flavonols comprising as little as one to ten percent of total flavonoids depending on variety, site and season. The role of flavonoids in contributing to colour stability and the organoleptic properties of wine has yet to be established. However, our studies suggest that two periods of flavonol synthesis occur during berry development, around flowering and in the latter phases of ripening. 166 Manipulation of conditions in the vineyard at those times, such as bunch exposure to light, may have significant effects on flavonol synthesis and wine quality. 167 89r Chapter 5. Effect of Shading on Flavonoid Biosynthesis in Vítis vinifera L. cv. Shiraz 5.L. Introduction The flavonoid composition of grapes has attracted much attention as an important contributor to wine quality (Glories, 1988). The three main classes of flavonoids found in grapes and wine are the anthocyanins, flavonols and condensed tannins. Anthocyanins are the pigmented compounds responsible for the colour of red wine and grapes (Ribereau- Gayon & Glories, 1986). Flavonols are thought to act as IJV protectants (Flint et al., 1985; Smith & Markham, 1998) and free radical scavengers (Markham et a1.,1998), and although not red in appearance they are thought to contribute to wine colour as copigments for anthocyanins (Asen et a1.,1972; Scheffeldt & Hrazdina, 1978). While the role of condensed tannins in the grape is uncertain, their astringency is thought to act as a feeding deterrent to herbivorous animals and insects @eeny, 1976; Hatborne & Grayer, Igg3).In wine, condensed tannins contribute mouthfeel to red wines, as well as colour stability by forming complexes with anthocyanins (Iimberlake & Bridle, 1976; Gawel, 1998; Malien-Aubert et a1.,2002; Mateus et a1.,2002b; Vidal et a1.,2002). Of the three classes of flavonoids, tannins are present in the gteatest proportion in grapes, followed by anthocyanins, with flavonols being present at relatively low levels (Souquet et al., 1996). Of the many environmental and management factors that affect the flavonoid composition of grapes, exposure is regarded as one of the major influences (Winkler e/ al., 1974; Archer & Strauss, 1989; Jackson & Lombard, 1993). The consensus has been that low light reduces colour and other flavonoids, while increased light increases t69 flavonoid content (V/icks & Kliewer, 1983; Dokoozlian & Kliewer, 1996). However, some authors have reported no change with light treatments (Hunter et al., 1995), while others had observed high light to result in decreased colour @ergqvist et a1.,2001). Many explanations for these differences have been suggested, including differences in cultivar, site and season as well as different sampling and analytical techniques. In addition, many shading treatments have had an impact upon other microclimate parameters in the fruit zone, particularly temperature, which has been shown to significantly affect flavonoid accurnulation and berry development (Kliewer, ß7A; Jackson & Lombard, 1993). Some treatments have also altered the foliage exposure or the fruit to leaf ratio, affecting photosynthetic carbon assimilation or partitioning of assimilate to the fruit @eterson & Smart,1975; Rojas-Lara & Morrison, 1989; Morrison & Noble, 1990). Where treatments have tried to shade fruit without altering berry temperature or leaf exposure, the anthocyanin content of the fruit was reduced in some cases (Wicks & Kliewer, 1983; Gao & Cahoon, 1994; Dokoozlian & Kliewer, 1996). Other authors have reported no change in the anthocyanin content of the fruit with shading, but a significant decrease in the flavonol content of the fruit and lower anthocyanins in the wine (Price er al., 1995; Haselgrove et a1.,2000). In addition, shading has also been reported to alter the anthocyanin composition, with a shift towards acylated anthocyanins in the shaded fruit (Gao & Cahoon, 1994; Haselgrove et a1.,2000). Thus, while a body of research exists on the effects of sun-exposure on grape flavonoid composition, the diversity of outcomes has failed to clarify the relationship between light and flavonoid biosynthesis in grapevines. In addition, many of these investigations have been limited to ripe or ripening fruit over as little as a single growing season. Recent studies have shown large seasonal variation in t70 flavonoid composition (Revilla et al.,1997; de Freitas & Glories, 1999; Guidori et al., 2002), and indicate that much of the accumulation of tannins in the grape occurs prior to veraison (Kennedy et a1.,2000a; Kennedy et a1.,2000b). High levels of flavonols have also been reported in the fruit at veraison as well as at flowering and early in berry development (Haselgrove, 1997). To gain a gleater understanding of the effect of shading on flavonoid biosynthesis in grapes with a view to managing the flavonoid content and composition of the fruit, and ultimately the wine, it is necessary to examine the entirety of berry development across several seasons. In this work, we sought to separate the effects of light and temperature and to explore the impact of light exclusion on flavonoid biosynthesis in the major Australian red winegrape cultivar Shiraz throughout berry development over three successive seasons. 5.2. Materials and Methods 5. 2. 1. 1. Shading Treatments In addition to the shading treatments described previously (Section 2.2.2), a modest trial also investigated the inducibility of flavonol accumulation and VvFLSI expression. When the boxes were repositioned during the 2000-2001 season to enclose the developing bunches while allowing further development of the shoo! shaded tissue samples were collected. Three days following this operation, samples were also collected of the newly exposed leaves and inflorescences to determine flavonol accumulation and VvFLSj expression in these tissues following exposure to light. l7l 5.2.1.2. Analysis of Flavonoids in Shaded Berries Tannins in whole berries, seeds and skins of grapes were extracted and analysed by HPLC as described in Section 3.2.2 and Section 3.2.3, while flavonols were extracted and analysed by HPLC as described in Section 4.2.2. For HPLC analysis of anthocyanins, the frozen samples were ground to a fine powder under liquid nitrogen. Three separate aliquots (analysis replicates) of 0.1-g were extracted in 1.0 mL of acidified methanol (l% HCI v/v; 3 mLconc. HCL + 97 mL methanol; Harborne, 1958) for one hour in darkness at room temperature with occasional vortexing. Extracts were then centrifuged (15 min at 13,000 x g) and 500 pL of the supernatant transferred to HPLC auto-sampler vials. Samples were analysed using a Hewlett Packard HP 1100 high-performance liquid chromatograph (HPLC) with a Merck (Germany) LiCrospherrM analytical column protected by an SGE (Australia) C-18 guard column. The separation used water (solvent A) with a methanol (solvent B) and perchloric acid 1.5%o v/v (solvent C) gradient. The gradient was: zero min, 25o/o B, rsyo c; l0 min, 30vo B, 15% c:40 min, 35yo B, 15% c; 45 min, 45o/oB.,30% c; 105 min, 55o/oB.,30% c;107 min r00 %8, zero % c; ll2 min, I00yoB, zero o/o C; ll4 min,25o/oB, l5o/o C. The anthocyanin peaks were identified by comparison of their elution order with the published separation of \Vulf and Nagel (1973) and comparison of the elution time and absorbance spectra with a commercial standard of malvidin-3-glucoside (Extrasynthese, France). The integrated absorbance at 520 nm '\À/as used to determine the concentration of individual anthocyanins in each sample (100 ¡rL injection) expressed as malvidin-3- glucoside equivalents calculated ftom a calibration function prepared from the commercial standard. t72 5.1.1.3. Gene Expression Studies in Shaded Berries Gene expression was determined by real-time PCR analysis. For the isolation of RNA, grapevine material frozen in liquid nitrogen in the field was ground to a fine powder in liquid nitrogen using a pre-frozen mortar and pestle and stored at -80oC. Total RNA was extracted from the frozen powdered tissue and oDNA was prepared from the RNA as described in Section 4.2.3. cDNA syntheses for all RNA samples to be analysed were performed simultaneously. Reagents required for the reverse transcription (RT) reactions were made as a master mix and aliquoted into each reaction in a single step to avoid pipetting enors that may have affected the efficiency of the RT reactions in different samples. (Corbett Research, Quantitative real-time PCR was performed on a Rotor-Gene 2000rM Australia) real-time PCR machine using SYBR green as the system of detection for double stranded PCR products. Primers homologous to the grapevine cDNAs encoding VvFLSI (AY257978), VvUFGT (X75968) and í/vUbiquitínl (TC9998) were designed using primer 3 (http://www-genome.wi.mit.edu/cgi-bin/primerþrimer3-www.cgi). The primer pairs vvFLSl Forward: 5',-CAGGGCTTGCAGGTTTTTAG- 3" VvFLSI Reverse: 5'-GGGTCTTCTCCTTGTTCACG- 3' and VvUFGT Forward: 5' TAACACATTGTGGATGGAACTCAT- 3' ANd VVUFGT REVETSE: 5' ACCTTCAATTCTCACTCCAATCTC- 3'were used to generate PCR products of 154bp and 140bp respectively for FLS and UFGT. The VvUbiquitinl transcript was detected using the primers VvUbiquitin Forward: 5'-GTGGTATTATTGAGCCATCCTT- 3' and VvUbiquitin Reverse: 5'-AACCTCCAATCCAGTCATCTAC- 3' which ampliff a product of 182base pairs. 173 All primer pairs were used under identical conditions, 95oC for 5 minutes followed by 35 rounds of amplification (95oC/30sec, 57oCl30sec, 72oC/30sec). Each PCR reaction contained 1 x SYBR Green PCR Master MixrM (Applied Biosystems, UK), 30 nmoles of each primer and 1/50th volume of cDNA stock in a final reaction volume of 25 ¡rL. The VvFLSI andVvUFGZmRNAswerequantifiedasaproportion of VvUbiquitinl usngthe Rotor Gene 2000rM software. Expression of VvFLSI andVvUFGf mRNA is presented as a ratio of expression of Vv F LS I and VvUF GT to expression of VvUbiquitin I and, therefore has no units. 5.3. Results 5.3.1.1. Effect of Shading on Berry Development In the first experimental season (1999-2000) flowering occurred nine weeks pre-veraison, with fruit set occurring one week later (eight weeks pre-veraison). Berry weight increased steadily from fruit-set towards veraison in both the shaded and exposed fruit. At veraison, shaded berries weighed less than the exposed fruit and shading had a noticeable effect on berry weight post-veraison (Figure 5.14) with the weight of shaded berries substantially lower than that of the exposed û:uit. At harvest, shaded berries weighed 35% less than the exposed berries. Despite the difference in berry weight, shaded berries ripened at a similar rate to the exposed fuit and reached the same level of tot¿l soluble solids at harvest, around 23.5"Brix (Figure 5.18). 174 In the second season (2000-2001), flowering occurred eight weeks pre-veraison with fruit-set the following week (seven weeks pre-veraison). The difference in berry weight betwþen shaded and exposed fi:uit (Figure 5.1C) was less noticeable than that observed during the 1999-2000 season (Figure 5.14). Pre-veraison, berry weight in the shaded fruit was lower than that in the exposed fruit (Figure 5.1C). However, post-veraison increases 1.6 30 +ExpGed +Exposed .El'Shaded 25 ^t 2 -.O - Shaded ? 20 E .9 o I 0.8 öis È 10 oo 0.4 A B 0.0 0 -9-8 -7-6-5 -4 -3-2-1 0 1 2 3 4 5 6 7 €-8 -7-6 -5-4 3 -2 -1 0'l 2 3 4 5 6 7 Weeks Post-Veraison weeks Post-Veraison 1.6 +Exposed '+Exposed 25 Et- Shaded 1.2 e Shaded e^ 20 E .9 I 0.8 815 è E o l0 t 0.4 c 5 D 00 0 -9-8-7-6-5-4-3-2-1 0 1 2 3 4 5 6 7 I I -8-7 6ó-4-3 -2-'l 0 1 2 3 4 5 6 7 I 9 Weeks Post-Veraison Weeks Post-Veraison 30 1.ô +É{Poæd -Þ Exposed 25 .B- Shaded E 'Shaded 1.2 G 20 o ë rs tE o.e P ò 10 o o 0.4 E 5 F 00 0 €-5-4-3-2-'t0123456 6ó-4-3-2¡0123456 Weeks Post-Veraison Weeks Post-Verâison Figure 5.1. Grape berry development in Vitis vinifera cv. Shiraz over three seasons. A). Berry weight oBrix for exposed (o ) and shaded (tr) fruit; B) Total soluble solids expressed as for exposed (o ) and shadeã @) fruit in the 1999-2000 season. Flowering occurred during the week of 04 November 1999 (9 weeks pre-veraison) with veraison around 06 January 2000 and commercial harvest on 24 February 2000 (7 weeks post-veraison). C) Berry weight for exposed (l ) and shaded (O) fir¡it; D) Total soluble oBrix solids expressed as for exposed (o ) and shaded @) fruit in the 2000-2001 season. Flowering occurred 8 weeks pre-veraison (16 November 2000), with veraison around 1l January 2001, the fruit was harvested 13 March 2001 (9 weeks post-veraison). E) Berry weight for exposed (r) and shaded (tr) fruit; tr) Total soluble solids expressed as "Brix for exposed (o ) and shaded (tr) fruit in the 2001- 2002 season. Flowering occurred l0 weeks pre-veraison (23 November 2001), with veraison around 31 January 2001, the fruit was harvested l5 March 2001 (6 weeks post-veraison). t75 in the berry weight of shaded fruit resulted in similar weights by five weeks post-veraison and at harvest, berry weights were similar between shaded and exposed fruit (approximately 0.96 glbery). There was also no difference in total soluble Solids between shaded and exposed fruit throughout ripening during the 2000-2001 season (Figure 5.1D), similar to the trend for sugar accumulation observed during the previous season (Figure s.1B). During the third experimental season (2001-2002), flowering occurred 10 weeks pre- veraison with fruit-set occurring two weeks later at eight weeks pre-veraison. The increase in berry weight in the early part of berry development (pre-veraison) was more gradual than observed in previous seasons (Figure 5.1E). There was virhrally no difference in berry weight between the shaded and exposed fruit throughout development during the 2001-2002 season with berry weight increasing steadily from fruit-set towards harvest. Similarly, there was no discernible difference in the level of total soluble solids ('Brix) between the exposed and shaded berries (Figure 5.1F) during berry development in the 2001-2002 season. 5.3.1.2. Anthocyanin Accumulation in Shaded Grapes Figure 5.2 shows the pattern of anthocyanin accumulation in Shiraz berries over three successive seasons (1999-2000, 2000-2001 and 2001-2002) on both a mglg fresh weight and a mglberry basis. The total anthocyanin content was calculated as the sum of the individual compounds determined by HPLC and expressed as malvidin-3-glucoside equivalents. In the first experimental season (1999-2000) anthocyanins accumulated slowly from veraison, followed by a rapid increase in concentration between two and four weeks post-veraison (Figure 5.2A). Anthocyanins accumulated more slowly at first in the t76 exposed fruit than in the shaded fruit, but at commercial harvest there was no significant difference @:0.05; n:3) in the concentration of anthocyanins between shaded and control fruit. On a per beiry basis (Figure 5.28), the pattern was similar to that observed per gram of berry weight. In the second experimental season (2000-2001), there was a significant difference (cr:0.05, n:3) in the concentration of anthocyanins between the shaded and exposed (control) fruit, with the exposed ftiit having a higher level of anthocyanins per gram ç 2.o 2.5 E +Exposed +Shaded +Expos€d +Shaded o È o o E l.s o 2.O o A E B o 't.5 .= * 1.0 o ,¿ 1.0 o o ò o.s c 0.5 €c 0.0 0.0 01 2345 67 01 2345 67 Weeks PosþVeraison Weeks Post-Veraison 15 2.5 +Exposed €-Shaded 4 Exposed +shadsd È oo È 2.O c o D E 10 o o E 1.5 E .E c 1.0 'Ê o o 05 c 0.5 00 0.0 012 3456 789 0'l 23456 789 Post-Vereison Weeks Post-Veraison Weeks 3.0 3.0 + Exposed +Shaded -.-Exposed +Shaded Ê b 2.5 2.5 o è F' Ë E E ; z.u È, 2.0 È, E É 9 ,1.5 1.5 .E .Ec d 1.0 1.0 E o o ¡ G Ë 0.5 0.5 0.0 00 01 234 U 234 Þ Weeks Post-Veralson Weeks Post-Veralson Figure 5.2. Total anthocyanin accumulation in Vitis vinifera cv. Shiraz berries over th¡ee seasons in lVillunga, SA. A) Total anthocyanins (mglg fresh weight berry) in exposed (r ) and shaded (tr) fruit in the 1999-2000 season; B) total anthocyanins (mg/beny) in exposed (o ) and shaded @) fruit in the lggg-2000 season. C) Total anthocyanins (mg/g fresh weight berry) in exposed (o ) and shaded (tr) fruit in the 2000-2001 season; D) total anthocyanins (mg/berry) in exposed (o ) and shaded (tr) fruit in the 2000-2001 season. E) Total anthocyanins (mgig fresh weight berry) in exposed (o ) and shaded (tr) fruit in the 2001-2002 season; F) total anthocyanins (mg/berry) in exposed (l ) and shaded (tr) fn¡it in the 2001-2002 season. 117 fresh weight of skin than the shaded fruit (Figure 5.2C). In the exposed fruit, anthocyanins accumulated steadily from one to two weeks post-veraisor¡ then more rapidly over the following week to three weeks post-veraison when the highest concentration \ryas recorded (Figure 5.2C). This level per gram fresh weight of berry was maintained until harvest. However, on a per berry basis, anthocyanin accumulation continued as the berry increased in size until five weeks post-veraison as shown in Figure 5.2D. On a per berry basis a decline in total anthocyanins was observed from seven weeks post-veraison until harvest. This decline was also observed in the shaded fruit, although the maximum observed was lower. In addition, the peak in anthocyanin accumulation per gram fresh weight in the shaded fruit occurred one to two weeks later than observed in the exposed berries. The level of toøl anthocyanins per berry was also significantly lower in the shaded fruit during the second season (Figure 5.2D). While the pattern of anthocyanin accumulation was similar between the 1999-2000 and 2000-2001 seasons, the level of total anthocyanins per gram fresh weight of berry and per berry was substantially lower than in the previous season, in both exposed and shaded fruit. This level was also lower than that observed in the following season (200r-2002). In the third experimental season Q00l-2002), there was no difference between exposed and shaded fruit in either the amount of anthocyanins per gram fresh weight of berry skin (Figure 5.28), or per berry @igure 5.2F). The pattern of anthocyanin accumulation per gram fresh weight of berry and per berry was similar to that observed in previous seasons with the bulk of accumulation occurring during the first three to four weeks following veraison. However, on a per gram fresh weight basis, anthocyanin accumulation in the shaded fruit continued steadily until harvest whereas in the exposed fruit, accumulation slowed towards harvest. 178 oBnx The anthocyanin composition of Shiraz berries at24 was determined by HPLC and the mean percent composition compared between treatments and seasons for each of the anthocyanins commonly detected in Shiraz using Student's t-test (cr:0.05; n:3) (Table 5.1). In the first experimental season (1999-2000), there was no significant effect of shading on total anthocyanins and there was no apparent change in the relative proportions of non-acylated, acetyl- and coumaroylglucosides. In the second experimental season (2000-2001), there was a significant difference in the level of total anthocyanins in Shiraz berries between treatments (Figure 5.2) as well as substantial shifts in anthocyanin composition with shading (Table 5.1). While all anthocyanins \Mere lower in the shaded fruit, the decrease in non-acylated and Table 5.1. Mean percentage of anthocyanins in Shiraz berries at 24 tsrìx across three seasons. The effect of shading on the level of each anthocyanin is shown; NS indicates no significant difference, while increase or decrease indicate a signihcant difference in the level ofanthocyanin (e0.05; n:3). Superscripts indicate differences between seasons in the level of each of the anthocyanins present in Shiraz berries for shaded and exposed fruit Percentage of Anthocyanins in Shiraz Berries Glucoside Acetylglucoside Coumaroylglucoside Season shaded Exposed Effectof shaded Exposed Effectof shaded Exposed Effectof Shading Shading Shading Malvidin I 999-2000 2406^ 23 80" NS 6.1 3b 7.0f NS I 1.66" t2.62' NS 2000-2001 20 51b 31 54b decrease 7.8 r 13.45b decrease 3l,zlb 24,54b lncrease 200t-2002 34.61 3s.87 dec¡ease s.2t 6.4t decrease 8.9J 9.4t decrease Petunidin 1999-2000 7.0e 7.48" NS 1.50 t.7T NS 3.5 r 4.07 decrease 2000-2001 3.s4 5.17b decrease 0.7'P t.27b decrease 4.6t z.tt rncrease 2001-2002 93q 9.9'.7" decrease 1.2'r l.s9 decrease l.8r 2.0f decrease Delphinidin 1999-2000 6.2fb 7.72^ NS 1.6sb 1.6E NS 2.8f 3.efb decrease 2000-2001 4.3t 4.97h NS 0.8f 0.89b NS ¿.sf 3.6f tncfease 2001-2002 e.0J 10.51 decrease l.1d l.5f decrease 2.08 z.¿* decrease g.g6b Peonidin t999-2000 t4,66^ 9.81" lncfease 4.5t 4.4t NS 8.5/ rncfease 2000-2001 e .o'f 3.64b increase z.o* 1.96h NS 10.73' 4.6'Ì rncrease 2001-2002 15.24^ 10.56" lncrease z.t4 1.83h tncfease s.44 4,3Ú rncrease Cyanidin 1999-2000 2.9t 2.95^ NS 1.36" decrease 2.2s 2.7e NS 2000-2001 o.9d o.sob rncrease 0.1 5b Increase r.o4 0. 5lb lncrease 2001 JOO). 2, sÊ 2.61' NS o2* NS 0.6ú o5ç lncfease Total I 999-2000 54.97u 5l.76" NS 14.8cb 16.29^ NS 30.19" 31.95" NS 2000-200 I 35.68b 45.81h decrease 11.76^ 17.72^ decrease 52.56b 36.47h lncfease 'to ).oo1-2002 o1o ro sR" 060 I l -60b decrease lqol" 18 82" NS t79 acetylglucosides in the shaded fruit was greater than the change in coumaroylglucosides, so the proportion of total coumaroylated glucosides increased, while the proportion of non-acylàted and acetylglucosides decreased. This general patterir was the same for the glucosides of malvidin and petunidin, and while there \¡/as no change in the delphinidin non-acylated and acetylglucosides, the coumaroylglucoside of delphinidin also increased with shading during the 2000-2001 season. In contrast to delphinidin, petunidin and malvidin, there was a general increase in the relative proportions of the glucosides of cyanidin and peonidin, with peonidin non-acylated glucosides and coumaroylglucosides in the shaded fruit consistently higher across all three seasons (Table 5.1). In the third experimental season (2001-2002), while there was no change in total anthocyanins with shading, significant changes in percent composition were again observed (Table 5.1). The general pattern was very similarto that of the previous season (2000-2001), with the proportion of total acetylglucosides decreased in the shaded fruit, although there was no change in total coumaroylglucosides and a slight increase in total non-acylated glucosides. All of the glucosides of malvidin, petunidin and delphinidin were lower in the shaded fruit while the glucosides of peonidin were significantly higher and those ofcyanidin generally unchanged. These observations indicate a clear difference in the response of cyanidin and peonidin based anthocyanins to shading compared to delphinidin, petunidin and malvidin. Cyandin and peonidin differ from delphinidin, petunidin and malvidin in the hydroxylation pattern of the flavonoid "8" ring (Holton &, cornish, 1995). Therefore, the percent of dihydroxylated anthocyanins was compared with the proportion of anthocyanins with three hydroxyl moieties on the "8" ring (Table 5.2). In each of the three seasons there was a significant difference (p<0.005; n:3) between treatments. The shaded fruit were 180 consistently lower in trihydroxylated anthocyanins and there was a higher proportion of anthocyanins with two hydroxyl moieties on the flavonoid "B" ring. Significant differences were also observed between seasons in the mean percentage of anthocyanins (Table 5.1). Total non-acylated glucosides and coumaroylglucosides were significantly different across all three seasons in both shaded and exposed fruit, but there was no consistent pattern in the changes in total acetylglucosides, between seasons. In the second season (2000-2001) when total anthocyanins were significantly lower in the shaded berries, there were proportionally less total non-acylated glucosides, little difference in the level of acetylglucosides and a significantly higher percentage of coumaroylated anthocyanins in the shaded ftiit. Generally, non-acylated and acetylglucosides were lower in the second season, although the pattern of malvidin non- acylated and acetylglucosides was less consistent between seasons. Malvidin acetyl- and coumaroylglucosides were proportionally higher in the second season in both the shaded and exposed fruit. Otherwise, there was not a consistent pattern in the differences in coumaroylated anthocyanins between seasons. Table 5.2. Effect of shading and season on the pattern of "B" ring hydroxyl substitution of anthocyanins in grapes at 24"Brix. Incrèase/Decrease indicates a significant difference in mean percentage of ãilrydroxylated (3',4'-OH; cyanidin, peonidin) and trihydroxylaied (3',4',5'-OH; delphinidin, petunidin, malvidin) anthocyanins (p<0.005; n:3). Superscripts indicate significant difference in mean percentage of dihydroxylated and trihydroxylated anthocyanins between seasons (p<0.0001; n:3) ttBtt Ring Substitution Season Shaded Exposed Effect of Shading 3'-5'-OH 1999-2000 3s.4t 29.87 Increase 2000-2001 2lJ3b I 1.43b Increase 2001-2002 26.3c 20.17 Increase 3450H 1999-2000 64.ss 70.13^ Decrease 2000-2001 78.8f 88.57b Decrease 2001-2002 73.64" 79.83' Decrease 181 There were also significant differences in the proportions of di- and trihydroxylated anthocyanins between seasons (Table 5.2). While there was a significant difference between all three seasons (p<0.0001; n:3), the most obvious difference was the substantially lower proportion of dihydroxylated anthocyanins (cyanidin and peonidin) and greater proportion of trihydroxylated anthocyanins (delphinidin, petunidin and malvidin) in both the shaded and exposed fruit in the second season (2000-2001). In absolute terms, this difference represents a decrease in cyanidin and peonidin glucosides per gram fresh weight of skin atrd an increase in delphinidin, petunidin and malvidin glucosides (raw data not shown). The pattern of methyl substitution was also significantly different between seasons with the levels of the unmetþlated anthocyanins, cyanidin and delphinidin, and peonidin and petunidin, with a single methyl group on the "8" ring, both lower in the 2000-2001 season (Table 5.3). Whereas malvidin was significantly higher in the second season (2000-2001) in both the shaded and exposed (control) fnÌit þ<0.05; n:3). Table 5.3. Effect of shading and season on the pattern of "B" ring methoxyl substitution of anthocyanins in the skin of Shiraz grapes at 24"Brix. Increase/Decrease indicates a significant difference in mean percentage of unmethylated (cyanidin and delphinidin glucosides), 3'-methylated (peonidin glueosides) and 3'5'-d-methylated (petunidin and malvidin glucosides) anthocyanins þ<0.05; n{). Superscripts indicate significant difference between season þ<0.05; n:3). "B" Ring Substitution Season Shaded Exposed EffectofShading Unsubstituted 1999-2000 16.93"" 20.41' NS 2000-2001 t2.4tr 10.65b NS 2001-2002 15.82b 18.03' Decrease 3'-CI{3OH 1999-2000 4t.2t 36jt Increase 2000-2001 2s.0d 19.82b Increase 2001-2002 35.3f 30.26c Increase 3',5'-CH3OH 1999-2000 4l.g5a 43.47 NS 2000-2001 59.53b 69.53b Decrease 2001-2002 48.8t 51.71' Decrease 182 5.1.1.3. Tannin Accumulation in Shaded Grapes The tannin composition of fruit from the first experimental season was determined from whole berries. In the shaded fruit, the level of free monomers per gram fresh weight of berry was initially slightly lower than that in the exposed fruit (Figure 5.34). From veraison onwards there was an increase in monomers in both the shaded and exposed fruit, both reaching similar maxima one to two weeks post-veraison, thereafter declining towards harvest. Expressed on a per berry basis the difference in free monomers between exposed and shaded fruit was more apparent, particularly the greater increase in free ê 1.4 lExposed 1.4 4-Exposed I 1.2 .€FShaded à '1.2 fl-Shaded B A o E ¡ 1.0 I 1.0 Ð c) E g 0.8 0.8 o o E E 0.6 o 0.6 o oE o 0.4 0.4 o o E o 0.2 L 0.2 I 0.0 0.0 -54-3-2-101234567 -5-4-3-2-1 0 1 2 3 4 5 6 7 Weeks Post-Veraison Weêks Post-Vera¡son 1.6 E 4.0 -!Exposed -.> Exposod I c ð D o E ft-Shaded ¡ 1.2 + Shaded o 3.0 o Þ, E E 2.0 0.8 tr ¡a o ct) f r o) .E 04 6 1.0 .g o E t- o F 0.0 0.0 -9-8-7-6 -5-4 -3 -2-1 0 1 2 3 4 5 6 7 -9 -8 -7 6 .5 -4 -3 -2 -',| O 1 2 3 4 5 6 7 Weeks Post-Veraison Weeks Post-Veraison 8.0 -+'Exposed -+ Exposed F E30o E ¡ E ôo +ÞShaded Ê + Shaded è, 6.0 ó E è, 20 g 'E a 4.O Pisc ¡ f a Êro 'õ U) 2.O .9rÉ o U oÉ fto 0.0 -9-8-7 -6-5-4 -3 -2-1 0 1 2 3 4 5 6 7 -9-8 -76 -5-4 -3 -2-1 0 1 2 3 4 5 6 7 Weeks Post-Veraison Weeks Post-Veraison Figure 5.3. Tarurin accumulation inVitisvinifera cv. Shiraz whole berries 1999-2000 in exposed (o) and shaded tn) fruit. A) Free monomers mg/g fresh weight of berry ftSEM); B) free monomers rng/berry (+SEM); C) Terminal subunits mg/g fresh weight of berry (ISEM); D) terminal subunits Ítg/berr)' (tSEM); E) extension subunits mg/g fresh weight of berry (tSEM); F) extension subunits mg/berry (iSEM) 183 monomers in the exposed fruit at veraison (Figure 5.38). This lower total per berry in the shaded fruit was consistent with the lower berry weight in the shaded fruit observed during that season. At commercial harvest, the leVels per gram and per berry in both the shaded and exposed fruit were similar. In the shaded fruit, the level of terminal subunits per gram fresh weight of berry was lower than that observed in the exposed fruit (Figure 5.3C). This difference became apparent from around fruit-set, with the berries shaded from flowering during that season. The maximurn level of terminal subunits in the exposed fruit occurred around six weeks pre-veraison, in the shaded fruit the maximum occurred at around fruit-set, eight weeks pre-veraison. From their respective maxima, the levels in both shaded and exposed fruit declined steadily towards veraison, remaining relatively constant thereafter at similar levels. This was not reflected in the per berry presentation (Figure 5.3D) as the difference in berry weight post-veraison resulted in a large difference in terminal subunits per berry between treatments. Nevertheless, at commercial harvest, the substantially greater level of terminal subunits per berry in the exposed fruit had decreased to a level similar to that in the shaded fruit. This pattern was virtually the same as that observed for the extension subunits between the shaded and exposed whole berries during the 1999-2000 season (Figure 5.3E, 5.3F). On a per gram fresh weight basis the highest concentration of extension subunits occurred around fruit-set and declined towards veraison, remaining relatively constant from that time until harvest (Figure 5.3E). On a per berry basis, extension subunits began to increase with increasing berry weight (Figure 5.1). The increase in extension subunits per berry was greater in the exposed fruit (Figure 5.3F). From six to four weeks pre-veraison extension subunits increased in both shaded and exposed fruit, although the increase was 184 greater in the exposed fruit. After that time the total in the shaded berries remained relatively stable until veraison when a further increase was observed to two weeks post- veraison. This level was maintained until harvest. In the exposed fruit there was a greater increase at veraison than in the shaded whole berries, followed by a substantial decline in extension subunits towards harvest, but again at harvest similar levels were observed (Figure 5.3F). In the second experimental season (2000-2001) tannins were determined in the seeds and skins separately. The levels of free monomers, terminal and extension subunits in the 2.5 +Exposed €'q) ,o +Exposed ø E +Shaded ôo +Shaded Èlo ó 2.O o E ø 1.5 8,, o E e.q8 1.0 E o o c A o 0.5 B E 34 II o 0.0 Ëo 7 I I -76-54-3-2-1 0 I 2 3 4 5 6 7 I 9 -7€-5 4-3-2-1 0 1 2 3 4 5 6 Weeks Post-Veraison Weeks Post-Veraison 4 Exposed 30 'l-Exposed E i.,U o + Shaded ¡ 25 +Shaded Ezo o o E Þ 2.O 915 '- lAç oa 1.5 ro l -3a U) 1.0 tt) o D õ_ C .E .çc 0.5 E ts o o FO F 0.0 -7-6-54-3-2-1 0 I 2 3 4 5 6 7 I I -7 4 -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 I I Weeks Post-Vera¡son a' Weeks Post-Veraison 6.0 -i- Exposed Hso -.F Exposed E s.o + Shad€d g40 Shaded -8 p Þ È I 4.0 oo g Þ'c f ..0 -Ef e= 2Q Ø 2 z.o E ô F 'H ''.0 ç3ro o ño ñ o.o -7-6-5 4-3-2-1 0 1 2 3 4 5 6 7 I I -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 I 9 Weeks Post-Veraison Weeks Post-Veraison Figure 5.4. Tannin accumulation inltitis vinifera cv. Shiraz seeds in the 2000-2001season in exposed (o) and shaded (tr) fruit. A) Free monomers mglg fresh weight of seed ÉSEM); B) free monomers mg/berry (ISEM); C) Terminal subunits mg/g fresh weight of seed (tSEM); D) terminal subunits rng/berry @SEM); E) extension subunits mglg fresh weight of seed (tSEM); F) extension subunits mglberry (ISEM) 185 seeds of both shaded and exposed fruit during the 2000-2001 season are presented in Figure 5.4. During the second experimental season no difference was observed in the level of free monomers in the seeds between treatments on either a pef gram fresh weight of seed or a per berry basis (Figure 5.4A, 5.48). In both shaded and exposed fruit free monomers accumulated slowly at first, then rapidly from around three weeks pre- veraison until one week post-veraison, thereafter declining steadily until around five weeks post-veraison. After this time, the levels remained relatively constant until commercial harvest. Similarly, the pattern of terminal subunit accumulation in the seeds was virtually identical between the shaded and exposed fruit, although on a per gram basis it appeared that there \¡/as a slightly higher level of terminal subunits in the seeds of the shaded fruit (Figure 5.4C). However, this was not apparent on a per berry basis where no difference \¡/as observed in the level of seed terminal subunits between the shaded and exposed fruit (Figure 5.4D). On both a per gram and a per berry basis a substantial level of terminal subunits was present at fruit-set, the level was maintained until around two to three weeks pre-veraison when the level increased until around one week post-veraison. This was followed by a decline over the following two weeks, whereafter the levels remained relatively constant until the fruit was harvested. As with the free monomers and terminal subunits, there was very little difference in the level of extension subunits in the seeds of the shaded and exposed fruit (Figure 5.4E, 5.4F). The level of extension subunits per gram of seed was highest in the exposed fruit by around four weeks pre-veraison thereafter the level declined towards harvest (Figure 5.48).In the shaded fruit samples were not analysed until three weeks pre-veraison when the level was also relatively high, increasing slightly in the next sample (one week pre- 186 veraison) and then also declining towards harvest. On a per berry basis there were minor differences in the level of extension subunits in the seeds between the shaded and exposed fruit, but the trend was the same and the levels comparable (Figure 5.4F). In Shiraz berry skins collected during the second experimental season (2000-2001), the level of free monomers was noticeably different between treatments prior to veraison. On a per gram fresh weight of skin basis the concentration was much lower in the shaded fruit pre-veraison (Figure 5.54). However, an increase in free monomers per gram +Exposed ?Exposed ^c 0.6 0.05 A B tl-Shaded o.s +Shaded È ü oo 0.04 o S o.¿ E E 0.03 0.3 o eo E e co 0.02 6 0.2 o o 0) 0.0r o o E L u0 0.00 -7 -6 -5 4 -3 -2 -1 0 1 2 3 4 5 6 7 8 I -7-6-5 4-3-2-1 0 1 2 3 4 5 6 7 I I Weeks Post-Veraison Weeks Post-Veraison Ê1 * Exposed 0.21 * Exposed C à D 3 o.g + shaded o 0.18 + Shaded È o.g ¡ o o È, 0.7 E 0.15 : 0.6 ø 0.12 Èc '9ç o'5 a ¡l 0.09 -ã o.¿ u, u, 0.3 t 0.06 o.z 'Ë E 0.03 ots E 0.1 F Êo 0 -7-6 -5 4-3-2-1 0 1 2 3 4 5 6 7 I I 7-6-54 -3-2-1 0 1 2 3 4 5 6 7 I I Weeks Post-Veraison Weeks Post-Veraison {- 6 {- Exposed Eâ E Exposed T' à +l-Shaded èn tFShaded oo 5 I o co E 4 .Þ 15 'Ê. = 3 !a 10 f -ãa U' 2 U' É .9 .Es c I c o o x i0 uJ 0 l¡l -7-6 -54-3-2-1 0 1 2 3 4 5 6 7 I 9 -7-6-5 4 -3-2-1 0 I 2 3 4 5 6 7 I I Weeks Post-Veraison Weeks Post-Veraison Figure 5.5. Tannin accumulation ínVitisvinifera cv. Shiraz skin in the 2000-2001 season in exposed (t) and shaded (tr) fruit. A) Free monomers mg/g fresh weight of skin (1SEM); B) free monomers rng/berÐ' (1SEM); C) Terminal subunits mg/g fresh weight of skin (1SEM); D) terminal subunits rnglberry (tSEM); E) extension subunits mg/g fresh weight of skin (tSEM); tr') extension subunits mg/berry (1SEM) 187 between one week pre-veraison and one week post-veraison made the levels indistinguishable. On a per berry basis, this difference pre-veraison was more apparent, but again the levels converged pcist-veraison (Figure 5.58). In the exposed fruit, the level of free monomers per berry reached a maximum around two weeks pre-veraison, while in the shaded fruit, the maximum level observed was recorded five weeks post-veraison. On both a per gram and a per berry basis, the levels of free monomers in the skins of shaded and exposed fruit were comparable at commercial harvest. Skin tannin terminal subunits in the shaded fruit were acfinlly higher per giam fresh weight of skin than in the exposed fruit prior to veraison, with the maximum level observed around twice that in the exposed fruit (Figure 5.5C). However, this difference was not as obvious when the data were expressed on a per berry basis (Figure 5.5D). The level of terminal subunits in the skins of both shaded and exposed Shiraz berries reached a maximum around two weeks post-veraison. This level was comparable between treatments and declined to similar levels towards harvest. The level of extension subunits in grape skins was also noticeably different between shaded and exposed fruit. While the difference per gram fresh weight of skin was relatively small (Figure 5.5E), the differences per berry were much more apparent (Figure 5.5F). The maximum level of extension subunits observed in the skin of the exposed fruit per berry was around twice that observed in the skin of the shaded fruit. However, as was observed with both the free monomers and the terminal subunits, by commercial harvest similar levels were observed between the shaded and exposed fruit. While extension subunits were generally lower in the skin of the shaded fruit, there was a shift in the relative amounts of individual extension subunits based on the hydroxylation pattern of the "8" ring (Table 5.4). There was a proportionally greater decrease in the 188 trihydroxylated extension subunits, epigallocatechin, with shading than there was in the dihydroxylated extension subunits, catechin, epicatechin and epicatechin-gallate þ<0.01; n:3). This decrease represented a 50Yo decrease in the concentration of epigallocatechih (mg/gfivt berry) with a commensurate increase in catechin and epicatechin in the skin of the shaded fuit (tannin compositional data for shaded fruit not presented). Epigallocatechin extension subunits were not detected in Shiraz seeds. Table 5.4. Effect of shading and season on the pattem of "8" ring hydroxyl substitution of proanthocyanidin extension subunits in the skin of Shiraz grapes at 24"Btix. Increase/Decrease indicates a rignif,.unt difference in mean percentage of dihydroxylated (3',4'-OH; catechin, epicatechin, epicatechin- gallate) and triþdroxylated (3',4',5'-OH; epigalloctaechin) extension subunits (p<0.01; n:3). "8" Ring Substitution Season Shaded Exposed Effect of Shading 3'-5'-OH 2000-2001 77.95 64.79 Increase 3'-4'-5'-OH 2000-2001 22.05 35.21 Decrease Of the dihydroxylated procyanidins, there was also a slight, but significant þ<0.001, n:3) shift from catechin to epicatechin extension subunits and a greater shift towards epicatechin terminal subunits (p<0.0001; n:3) with shading. Epicatechin accounted for approximately 25Yo of terminal subunits in the exposed fruit, but almost 50% in the shaded fruit due to a decrease in the amount of catechin per gram fresh weight of berry skin (compositional data for shaded fruit not presented). In addition to compositional shifts between catechin and epicatechin with shading, the effect of shading on galloylation was also examined. The proportion of galloylated terminal subunits were slightly higher in the shaded fruit (1S%) than in the exposed fruit (14%) in the 2000-2001 season, however this difference was not statistically significant (p:0.594; n:3). In contrast, galloylated extension subunits in the shaded fruit were significantly higher (15%) than in the exposed frút (8.7%; p<0.0001; n:3). 189 The small differences observed between shaded and exposed fruit in seed tannins and the larger differences in skin tannins are highlighted by the differences in polymer length (mDP) between treatments (Figure 7.0 6.0 5.6). In Shiraz seeds, polymer length o- 5.0 Ê o.o was very similar between the shaded 3.0 .-Exposed 2.0 * and exposed fruit as suggested by the 1.0 A Shaded 0 -9-8-7-6-5-4 -3-2-1 0 1 2 3 4 5 6 7 I I similar levels of terminal and Weeks Post-Veraison extension subunits (Figure 5.64). 60 .-Exposed 50 *Shaded However, in Shiraz skin, the slightly ofL 40 E 30 higher level of terminal subunits and 20 lower extension 10 B subunit level in the 0 -9-8-7-6-5-4-3-2-10 1 2 3 4 5 6 7 I I shaded fruit combined to lower the Weeks Post-Veraison Figure 5.6 Polymer length (mDP) of condensed tannins mean degree of polymerisation in exposed (r ) and shaded ( ) l¡itis vinifera cv. Shiraz A) seeds and B) skins in the 2000-20001 season in (mDP) in the shaded fruit (Figure Willunga, SA. s.68). 5.1.1.4. Flavonol Accumulation in Shaded Grapes Flavonol accumulation \üas measured as total quercetin glycosides (Section 4.2.2) over three successive seasons (1999-2000, 2000-2001 and 200I-2002) in the skin of shaded and exposed fruit (Figure 5.7). In the first season (1999-2000), when boxes were applied one week after flowering, high levels of flavonols per gram fresh weight of skin were present in berries of both exposed and shaded fruit in the early stages of berry development, declining as the berries grew (Figure 5.74). The level of flavonols per gram fresh weight of skin was noticeably higher in the exposed fruit than in the shaded fruit at all stages of berry development and this was emphasised when the data were 190 presented on a per berry basis (Figure 5.7B). There was an increase in flavonols per berry 2-4 weeks after veraison in both the exposed and shaded fruit. In ripe fruit (-24'Brix), the concentration of flavonols in the skin of shaded fruit was 48%o of that in exposed fruit. In the second season (2000-2001), boxes were applied five weeks before flowering, when significant flavonol accumulation occurs (Section 4.3.5). The pattern of flavonol accumulation was broadly similar to the previous year in exposed fruit but no flavonols were detected in the shaded fruit at any stage of berry development (Figure 5.7C,5.7D). On a per berry basis, flavonol accumulation was again observed after veraison in the 0.06 o.25 + Exposed -+.Exposed 0.05 = 020 €- Shaded è €-Shaded oo E Þt 0.04 6 -ò 015 F E .9. 0.03 B A o .9 o co 0.10 o.02 o tr l! I 005 0.01 000 0.00 .6ó-43-2-101234567 6-54-3-2-10123 4567 Weeks Post-Veraison Weeks Post-V€raison 0.06 -rFExposed 015 +Exposed 0.05 +Shaded E tlShaded E ! o.'12 o ¡ 0.04 E o o É D È) 009 c E L 0.03 o E o c 0.06 o.u2 o o -g Ll L 0.03 0.01 0.00 0 6-5-4 3 -2 -1 0'l 2 3 4 5 6 7 I 9 s-s-4 -3-2-1 0'l 2 3 4 5 6 7 I I Weeks Posl-Veraison Weeks Post-Veraison 0.15 + Exposed 0.04 -è Exposed c +Þ Shaded o12 '+ Shaded à 0.03 Ë o 0.09 E ¡ F E P o02 .9. co 0.06 I o o .q o 001 L 0.03 L 000 0.00 -54-3-2-'10123456 5-4-32.-10 123456 Weeks Post-Veraison Figure 5.7. Flavonol accumulation in Vitis viniferacv. Shiraz skins in exposed ( o ) and shaded (' ) ftuit. A) Toìal flavonols (mg/g fresh weight berry) 1999-2000; B) total flavonols (rrg/berry) 1999-2000; C) Total flavonols (mg/g fresh weight berry) 2000-2001; D) total flavonols (mdberry) 2000-2001 ; E) Total flavonols (mgig fresh weight berry) 2001-2002; F) total flavonols (mg/berry) 2001-2002 l9l exposed fruit (Figure 5.7D) but even at harvest no flavonols could be detected in the shaded fruit. In the third experiment¿l season Q00I-2002) a much larger number of boxes were applied, four weeks before flowering. Flavonols in the shaded fruit were again much lower than in exposed fruit, on both a per gram fresh weight and per berry basis (Figure 5.78, 5.7F). In ripe grapes (-24Bnx), the concentration of flavonols in the skin of shaded fruit was only l0%o of that in exposed ftiit. An increase in flavonols per berry was observed in the exposed fruit later in ripening in all three seasons (Figure 5.7). In the shaded fruit, there was also some increase in flavonols per berry after veraison in the first experimental season (1999-2000) and to a lesser degree in the third season (200I-2002). 5.1.1.5. Expression of Genes Involved in Flavonol and Anthocyanin Biosynthesis in Shaded Berries Expression of genes encoding enzymes involved in anthocyanin and flavonol biosynthesis was determined by real-time PCR. The levels of expression of the genes encoding UDP glucose: flavonoid-3-O-glucosyltransferase (UFGT), one of the key steps in anthocyanin synthesis, and flavonol synthase (FLS), the first step in flavonol synthesis, are presented relative to the expression of a consistently expressed gene, Ubiquitin. The levels, of expression of these genes was compared between thc shaded and exposed ftiit in the 2000-200I season in Shiraz berry skin (Figure 5.S). Expression of VvUFGZ was very low before veraison but increased dramatically in the first 1-2 weeks after veraison, coincident with the onset of anthocyanin accumulation (Figure 5.2C). The level and pattern of expression of VvUFGZ was similar in shaded and 192 exposed fruit, declining at the later stages of ripening as reported previously (Boss et al., 1996c). In exposed fruit, the gene encoding flavonol synthase WFLSI) was expressed early in berry development and again during ripening (Figure 5.8), consistent with the two phases of flavonol accumulation in the berries (Section 4.3.5). Expression of VvFLSI in shaded fruit was extremely low post-veraison, when expression was detected in exposed fruit (Figure 5.8). To determine the effect of light on flavonol synthesis early in berry development, the concentration of flavonol glycosides and expression of VvFLSI was also determined in developing inflorescences, leaves and bunches which had not been exposed to light (Table 5.5). Boxes were applied to spurs prior to budburst and transferred to the developing inflorescence in the dark once the shoot had emerged. 0.5 +Expos€d 20.0 *Exposed +ÞShaded 16.0 {l-Shaded c 0.4 c .9 .9 Ø qE 0.3 o 12.0 x ê t¡J uJ U' 0.2 È- 8.0 J (, lr L l 40 0.'l 0.0 0.0 -9 -8-7-6 -54 -3-2-1 0 1 2 3 4 5 6 7 I -9 -8-7-6 -5-4 3 -2-',| 0 I 2 3 4 5 6 7 I Weeks Post-Veraison Weeks Post-Veraison Figure 5.8. Expression of VvUFGT and VvFLSI genes in Vitis vinifera cv. Shiraz skin in the 2000-2001 season in exposed (r ) and shaded (tr) fruit. VvUFGT and VvFLSI expression is the ratio of expression of each gene divided by the level of expression of an internal control gene (VvUbiquitinl)' Leaves sampled from shoots grown in the dark had less than 3Yo of the flavonol concentration of equivalent exposed leaves and expression of YvFLSI was only 5o/o of that in exposed leaves (Table 5.5). In the developing inflorescence, the bunch one week after flowering and in the skin of ripe grapes gtown in boxes throughout development, the 193 concenfration of flavonols lvas very low or undetect¿ble and the expression of VvFLSj was also greatly reduced compared to equivalent tissues exposed to light (Table 5.5). Table 5.5. Concentration of flavonols and relative expression of the flavonol synthase gene TvFLSI n different tissues of Shiraz. The growth stages of the developing inflorescence were determined according to the modified Eichom-Lorenz (E-L) scheme (Coombe, 1995) and were collected 7 weeks prior to flowerìng (EL#12),4 weeks before flowering (EL#16) and I week after flowering(EL#28). Expanding leaves were at stage 3 as described in Section 3.2.1. Harvest skin is from ripe grapes. All samples were collected during the 2000-2001 season and mean flavonol concentration (+SEM, n:3) is presented. VvFLSI expression is the ratio of expression of VvFLSI divided by the level of expression of an intemal control gene (VvUbiquitinI). Flavonols (mg/g fwt tissue) VvFLSI Expression Shaded Exposed Shaded Exposed LeafStage 3 0.093 + 0.012 3.323 + 0321 0.004 0.080 ß,L #l2Inflorescence 0.056 + 0.016 1.529 + 0.037 0.019 0.1 60 EL #16 Inflorescence 0.281 + 0.021 1.407 + 0.063 0.021 0.1 07 EL#28 Berries 0 0.172 + 0.005 0.003 0.009 Harvest Skin 0 0.1l6 + 0.007 0.036 0.175 5.1.1.6. Induction of Flavonol Biosynthesis Developing inflorescences and grapevine leaves were sampled three days after the shading treatment had been removed. These were analysed by HPLC for flavonol content and by real-time PCR for expression of VvFLSl. In shaded leaves at growth stage 3 (Section 3.2.I) the level of flavonols was less than3%o of that in the exposed leaves. After the shading treatment had been removed the level of flavonols in the induced leaves was around 39%o that of normally exposed (control) leaves, or about l4-fold higher than the level in the shaded leaf (Table 5.6). While the level of flavonols in the newly exposed leaves had not reached the level in exposed leaves, the level of VvFLSI transcript was more than 6-fold greater than the level in the normally exposed leaves and 126-fold higher than shaded leaves. 194 Similarly, in the shaded inflorescence the level of flavonols was only 20o/o of that in the exposed inflorescences, while the level in the induced tissues after three days was around 45o/o of that in exposed inflorescence. Again, it was seen that the level of flavonols in the induced tissue had not reached that of the exposed inflorescence, but the level of VvFLSI transcript was almost 2-fold higher than that of the control and l0-fold higher than the level observed in the shaded inflorescences. Table 5.6. Induction of flavonol biosynthesis in vitis vinifera cv. shiraz tissues. Tissues were shaded for ten weeks from bud-burst and then exposed to normal light for three days. Flavonol andVvFLSl levels shown as percentage of normally exposed (control) tissues. Growth stage of the developing inflorescence deiermined according to the modified Eichom-Lorenz (E-L) scheme Çoombe, l9g5) 4 weeks pre-flowering (EL#16). Expanding leaves at stage 3 (Section 3.2.1.) Samples were collected during the 2000-2001 season. % of Control Tissue Flavonols VvFLSI Shaded Leaf, Stage 3 2.8r 5.00 Induced Leaf, Stage 3 38.95 631.25 Shaded EL #16 Inflorescence 20.02 19.63 Induced EL #16 Inflorescence 4s.21 192.52 5.4. Discussion 5. 4. 1. L Shading Treatrnent In order to investigate the effect of light exposure on flavonoid composition in grapes, developing Shiraz berries were shaded in polypropylene boxes designed to exclude light from the developing fruit without otherwise altering bunch microclimate or vine physiology (Figure 2.2). Tials of the boxes used for shading Shiraz berries suggest that the light environment around the developing fruit was substantially reduced. While the possibility exists that some reflected light may have entered the boxes, our measurements indicated that the light intensity inside of the boxes was less than 0.5%o of incident light' 195 This was reinforced by measurements of chlorophyll in shaded and exposed fruit (Section 2.3.13) and supported by visual observation of the shaded bunches, which were yellow compared to the green berries of exposed bunches (Figure 2.35). From these results, we concluded that the levels of light entering the box as reflected light, in the absence of light transmission through the box material, were not sufficient in either quality or quantity to induce chlorophyll biosynthesis. While the boxes were successfully able to exclude light from around the fruit, they did not modi$r the air temperature or relative humidity around the bunches (Section 2.2.1). Thus, it seems reasonable to conclude that differences in fruit composition between treatments would be due to shading rather than to altered temperature or relative humidity. 5.4.1.2. Berry Development in Shaded Grapes In the first experimental season (1999-2000) there was a noticeable difference in berry weight between the shaded and exposed fruit (Figure 5.1A). However, in that season, the phenomenon known in Australia as "hen and chicken" was observed in Shiraz resulting in a high proportion of small under-developed berries as discussed in Section 2.4. However, there was no apparent difference in berry weight between the shaded and exposed fruit in successive seasons (2000-2001 and 2001-2002) it is likely that a disproportionate number of smaller underdeveloped berries were collected from the boxed samples in the first season. The combined data over three seasons (Figure 5.1) suggest that shading of the fruit did not have a consistent impact upon berry development nor on fruit weight at harvest. 196 The accumulation of sugars in the berry as measured by total soluble solids f Brix) was also consistent between treatments and across three successive seasons. While minor differences were observed between samples, the trend and overall levêl was virtually the same in the shaded and exposed ftiit (Figure 5.1). While a slight decrease in sugar accumulation with shading has been reported in the Vitis hybnd Reliance (Gao & Cahoon, tgg4), the hybrid Seyval blanc was unaffected @eynolds et a1.,1986). Previous work in the Vitis vinifera cultivars Shiraz and Merlot also showed no difference in sugar accumulation with bunch shading (Haselgrove et aL.,2000; Spayd et aL.,2002). Ho'trever, differences in sugar accumulation have been reported in the Vitis vinifera cultivars Cabernet Sauvignon and Grenache from the north to south side of the rows (Bergqvist er al., 200I). This may have been a reflection of difference in leaf sun exposure leading to reduced photosynthate assimilation, rather than altered berry exposure to light. The overall indication from the data presented here and that of other workers is consistent with the opinion of Coombe (1989), that sink activity and sugar accumulation in ripening grape berries is largely independent ofexternal events. 5.4.1.3. Anthocyanins in Shaded Grapes In many plants, such as apple and petunia, anthocyanin biosynthesis is light induced and tissues gro\ün in the dark do not accumulate anthocyanins $iegelman & Hendricks, 1958; Lancaster, 1992; Beggs & Welmann, 1994; Dong et al., 1998; Katz & Weiss, lggg). Here we have seen that grapes grown in the absence of light did accumulate anthocyanins, which indicates that light is not an absolute requirement for anthocyanin biosynthesis in the berries of Vitis vinifera cv. Shiraz. 197 Across the three experimental seasons examined here, there was only a difference in total anthocyanins between heatments in one of those seasons. However, there were consistent differences in the composition of individual anthocyanins between treatments across all three seasons. The most obvious difference in anthocyanin composition between treatments was the shift from those anthocyanins with three hydroxyl groups on the "B" ring, the glycosides of malvidin, petunidin and delphinidin, to those with only two hydroxyls, based on peonidin and cyanidin. This observation suggests that the effect of shading is to reduce the activity of the flavonoid-3',5'-hydroxylase, or expression of the gene encoding that enzyme, or altematively to increase the activity of flavonoid-3'- hydroxylase or upregulate expression of that gene. While genes encoding these en4lmes have been cloned in other species (F{olton et al.,l993a; Neilsen & Podivinsky, 1997; Shimada et al., 1999), there are currently no published sequences for these genes in Vitís vinifera and this aspect of anthocyanin biosynthesis in grcpes has yet to be closely examined. The shift in composition based on hydroxyl substitution of the "8" ring suggested that compositional differences based on the methylation pattern of the "8" ring should also be examined. There did appear to be an increase in the percentage of anthocyanins with a single methyl substitution and a decrease in the proportion of anthocyanins with two methyl moieties on the "8" ring across all seasons with shading (Table 5.3). However, this result may not indicate a change in anthocyanin metþltransferase (AMT) activity with shading, as decreased F3'5'H activity or increased F3'H activity could lead to a decrease in substrate for 5' metþlation, which may result in a decrease in 5' methylated product. 198 This shift in anthocyanin composition based on "8" ring substitution with bunch exposure has not previously gained a great deal of attention. However, careful reading of the data presented by Spayd et al. (2002) shows a similar trend in Merlot, while Gao and Cahoon (1994) presented data that suggests a similar shift in the Vitis hybrid Reliance with shading. Previously, reports have tended to focus on shifts in the glycosylation pattern of the anthocyanins (Gao & Cahoon,1994; Haselgrove et a1.,2000), with a proportional shift from non-acylated glucosides to coumaroylglucosides with shading. Such a trend was observed here in the malvidin glucosides and total anthocyanin glucosides only in the second experimental season (Table 5.1). However, because of the shift based on the "8" ring hydroxylation pattern discussed here, it may not be useful to compare the relative proportions of anthocyanin glucosides, unless they are f,rrst separated by their pattern of "B" ring substitution. On this basis, a consistent pattern was again observed. The trihydroxylated anthocyanin non-acylated, acetyl-, and coumaroylglucosides were reduced on both a percentage and a mg/g fresh weight basis with shading, while the dihydroxylated non-acylated, acetyl-, and coumaroyl anthocyanin glucosides increased. The only exception was in the second season (2000-2001), when the percentage of trihydroxylated coumaroyl derivatives increased, although in that season it was observed that all coumaroyl derivatives increased with shading. In the 2000-2001 season, the pattern of compositional change in both malvidin and total anthocyanins based on glycosylation was the same as that reported by Haselgrove et al. (2000). Non-acylated and acetylglucoside derivatives decreased with shading, while coumaroylglucosides increased. Gao & Cahoon (1994) also reported that the proportion of cyanidin non- acylated glucosides in Reliance grapes decreased with shading while the proportion of 199 acylated cyanidin glucosides increased. Haselgrove et al. Q000) suggested that increased light exposure lead to a decreasing proportion of coumaroyl derivatives in the fruit, however our results did not consisteñtly support this conclusion leading us to seek an alternative explanation. Of the three experimental seasons, the second (2000-2001) was significantly warmer than either of the other two seasons. Our data showed an increase in mean daily temperature during berry development and ripening over the next warmest season (1999-2000) of 4.2"C. @:0.0001) in November, 1.2"C (p:0.03) in December, and 3.1"C (p:0.008) in January. Thus, the proportional increase in coumaroyl glucosides in the 2000-2001 season may be due to higher temperatures during that season. Similar results were presented by Spayd e/ al. (2002), who reported that cooling decreased coumaroyl derivatives, while heating substantially increased coumaroylglucosides. The suggestion implicit in this observation is that either activity of the enzyme anthocyanin acyltransferase (AAT) or expression of the gene encoding AAT is increased with temperature. To date there are no Vitis vinifera clones of AAT and thus, no expression studies of the gene encoding this enzyme to determine whether AAT expression reflects these observations. What is more interesting to note is that while anthocyanin coumaroylglucosides were proportionally greater in the warmer season, coumaroylglucosides were also higher in the shaded fruit in that season (2000-2001), even though overall colour was significantly lower in the shaded fruit. While Haselgrove et al. (2000) reported this shift towards coumaroylated anthocyanins with shading, they reported no change in total anthocyanins. It seems likely that the result observed by Haselgrove et al. (2000) was the result of their shading treatment inadvertently increasing berry temperature resulting in the shift 200 towards coumaroylated anthocyanins. Similarly, Price et al. (1995) reported no change in total anthocyanins in the fruit of Vitis vinifera cv. Pinot noir with shading, although this cultivar lacks acylated anthocyanins. In contrast, Gab & Cahoon (1994) observed both a shift in composition and a significant decrease in anthocyanins with shading. However, the absence of acylated glycosides of delphinidin, petunidin, peonidin and malvidin in this cultivar likely clouds the usefulness of these observations. Spayd et al. (2002), working on Vitis vinifera cv. Merlot reported no difference in total anthocyanins with shading in their first season, but a substantial decrease in the second season. If nothing else these observations highlight the ethereality of conclusions based on a single season's observation and indicate the value of a long-term study in this field. Generally, the indication is that higher temperature rather than light results in a shift towards coumaroylated anthocyanins, and that higher temperature results in decreased total anthocyanins. The mechanism for this is uncertain but supports the notion of anthocyanin degradation or inhibition of anthocyanin biosynthesis at high temperature (Buttrose et al., 1971; Pirie, 1977; Markakis, 1982; Cnppen & Morrison, I986a; Yamasaki et al., 1996). There is also a suggestion that at high temperature, the couymaroylated anthocyanins may be more stable than the non-acylated or acetylglucosides or more resistant to such degradative processes as may be taking place. However, the varying stabilities of individual grape anthocyanins has not been well documented. That a decrease in anthocyanin content is more pronounced in the shaded fruit suggests that in response to anthocyanin degradation at high temperature, supplementary anthocyanin biosynthesis is light inducible. Thus, there may be two systems regulating colour accumulation in gtapes, an initial and constitutive system that generates a base level of anthocyanins and an inducible system that is light requiring' 201 The results presented here suggest that the effect of shading is to decrease the proportion of malvidin, petunidin and delphinidin glycosides relative to peonidin and cyanidin glycosides. It remains to be seen whether the effect of increased exposufe is to reverse this trend. Furthermore, there is a strong indication that high temperature shifts the anthocyanin profile from non-acylated to coumaroylglucosides, irrespective of bunch exposure. Thus, we might expect that exposed fruit in a warm climate would have a higher proportion of malvidin, petunidin and delphinidin coumaroyl derivatives, while cool, shaded fruit might have more peonidin and cyanidin non-acylated and acetyl glucosides. What has not yet been firmly established is the relative contribution of each of these components to final wine colour. However, research on the extractability of anthocyanins from the fruit into the wine suggests that the non-acylated and acetylglucosides are more readily extracted from the fruit than the coumaroylglucosides (Leone et al., 1984; Roggero et a|.,1984). 5.4.1.4. Condensed Tannins in Shaded Grapes While condensed tannin accumulation in grape berries has been extensively studied elsewhere (Czochanska et a1.,L979;Katalinic & Males, 1997; de Freitas & Glories, 1999; Kenned¡z et a1.,2000a; Kennedy et a1.,2000b; Kennedy et a1.,2001), the influence of light on tannin accumulation has not previously been investigated. Over two seasons, 1999-2000 and 2000-2001, there was a noticeable difference in total condensed tannins. While in the first season tannin analysis was conducted on whole berry samples it was not possible to determine whether the changes observed were in the seed or the skin of the berry. However, in the subsequent season, these components were 202 analysed separately and it became apparent that there was no effect of shading on seed tannin content and composition. In contrast there were significant differences in the condensed tannin content ofgrape skins in the second season,2000-2001. This suggests that the differences observed during the first season may have been due to changes in skin tannin content. In both seasons, substantial differences were observed in the levels of free monomers and extension subunits from early in berry development until several weeks post-veraison, but as the fruit approached maturity these differences became less pronounced. In the case of free monomers, the level in the exposed skin decreased slightly, while the level in shaded fruit continued to increase, with the levels in shaded and exposed fruit converging around three weeks post-veraison in both seasons. A similar pattem was observed in total extension subunits, whereby the level in the exposed fruit reached a higher total, but subsequently declined to a level similar to that in the shaded fruit at the time of harvest. In the case of the extension subunits, the level in the exposed fruit was much higher than in the shaded fruit and significant decrease was observed from a peak one to two weeks post-veraison until harvest. Previous studies have suggested that the post-veraison decrease in condensed tannin in both seeds and skins is likely the result of reduced extractability (Cheyniet et al', 1997; Saint Cricq de Gaulejac et al.,1997; Kennedy et a1.,2000b; Section 3.4). This trend was apparent in the extension subunits of the exposed fruit, but there was no post-veraison decrease in extension subunits in the shaded fruit. The conclusion to be drawn from this observation is that even though the level of extractable tannin was similar at harvest in both the shaded and exposed fruit, more condensed tannin, particularly extension subunits, was produced in the exposed fruit. The lower level of extension subunits 203 observed in the shaded fruit also resulted in a lower mDP in the skins of shaded berries throughout berry development, It is interesting to note that there was a difference in condensed tannins between treatments in both the 1999-2000 season and the 2000-2001season, while there was only a difference in total anthocyanins between treatments in one of those seasons (2000- 2001). This suggests that the mechanisms that control tannin and anthocyanin biosynthesis in the berry are independent. So saying, it is worth noting that the observed shift in anthocyanin composition towards dihydroxylated anthocyanins based on cyanidin and peonidin and away from the glycosides of delphinidin, petunidin and malvidin was also observed in tannin composition. Specifically, this was observed in extension subunit composition where the proportion of trihydroxylated epigallocatechin was significantly reduced in the shaded fruit (Table 5.4). There was no consistent effect of shading on terminal subunit composition. In the first season, terminal subunits in whole berries were substantially lower throughout most of berry development. However, in the second season there was virrually no difference in terminal subunits in either seeds of skin at any time during berry development. The effect of shading on galloylation was to increase the relative proportion of galloylated epicatechin subunits in the skin. This effect was observed in both the terminal and extension subunits, although a two-fold increase in the percentage of galloylated extension subunits was considerably greater than the increase in the galloylated portion of the epicatechin terminal subunits. 204 While significant differences were observed in condensed tannin accumulation and composition between treatments, it is uncertain what implications these changes might have for winemaking. Potentially, bunch exposure could be a mechanism for managing winegrape tannins, although currently the optimum condensed tannin profile for a wine type has yet to be established. 5.4.1.5. Flavonols in Shaded GraPes The influence of light on flavonol accumulation in plant tissues has previously been studied in a number of plant systems with exposure to UV-light shown to increase the level of flavonol glycosides in vegetative and reproductive tissues (Irazdina & Parsons, 1982; Hartmann et a1.,1998; Ryanet al.,1998; YogJ et al., 1999; Reay & Lancaster, 2001). This effect has also been reported in winegrapes, with exposed fruit higher in flavonol glycosides, while shaded fruit had lower flavonol content (Price et al., 1995; Haselgrove et a1.,2000; Spayd et al., 2002). Because these studies examined fruit at harvest or during ripening, the effect of shading early in berry development remained uncertain. We have previously reported relatively high levels of flavonols in grape berries early in berry development (Section 4.3), so it is not clear whether lower levels of flavonols in the shaded fruit are caused by reduced biosynthesis or by degradation' Here we have examined flavonol accumulation in shaded and exposed (control) fruit over three successive seasons. In each season the level of flavonols was significantly reduced in the shaded fruit, consistent with the previous reports (Price et a1.,1995; Haselgrcve et a1.,2000; Spayd et a1.,2002).In addition, it was observed that from the time that the shading treatment was applied, no further flavonol accumulation was detected in that fruit. This suggests that reduced biosynthesis, rather than degradation is responsible for the low levels of flavonols present in the shaded fruit. FurtheÍnore, in the second 205 experimental season (2000-2001), when the shading treatment was applied five weeks before flowering, there were no detectable flavonols in the developing berries, which suggests that all flavonol biosynthesis in the Shiraz beiries is light induced. This is in sharp contrast to anthocyanin and tannin synthesis, which was little effected by the shading treatments. This suggests that only the branch of the flavonoid pathway leading to flavonol synthesis is light-dependent. Furthermore, while there was significant variation in total and individual anthocyanins between the three seasons studied here, the pattem of flavonol accumulation for the same fruit over the same three seasons showed very little variation (Figure 4.6). What has yet to be established is whether this level of flavonols (0.04-0.05 mg/berry) can be increased in more highly exposed fruit. Other authors have reported substantially greater levels in highly exposed grapes (Price et a\.,1995; Haselgrove et aI.,2000; Spaydet a\.,2002). 5.4.1.6. UFGT andFLS Gene Expression in Shaded Grapes The gene encoding UFGT appears to be a key step in anthocyanin synthesis, its expression correlating well with anthocyanin accumulation in grapes (Boss er a1.,1996). Expression of VvUFGZ was very low before veraison, but increased dramatically after veraison, when anthocyanin accumulation coÍtmenced, in agreement with previous studies (Boss et al., 1996c). The level of expression of VvUFGZ was similar in shaded and exposed fruit (Figure 5.6), consistent with the similar levels of total anthocyanins (Figure 5.2). This suggests that shading had little effect on expression of genes involved in anthocyanin biosynthesis in the grape skin. In contrast, expression of VvFLSI was greatly decreased in shaded leaves, developing inflorescences and ripening fruit (Table 5.5 and Figure 5.6). This gene encodes flavonol 206 s¡mthase (FLS), the first enzyme in the branch of the flavonoid pathway leading to flavonol biosynthesis, and our results indicate that the expression of FLS is light- dependent in Shiraz leaves, inflorescences and fruit. 5.4.1.7. Induction of Flavonol Biosynthesis When shaded leaves and developing inflorescences were exposed to light for a period of three days, these tissues began to accumulate flavonols and VvFLSI expression was upregulated. In fact, there was a 126-fold increase in VvFLSI expression compared to the level in the shaded leaves. Expression of TvFLSI also increased in developing inflorescences upon exposure, but this increase was not as great as in leaves. This is the first report of the induction of genes encoding flavonol synthase by light. Light induction of flavonols has been reported in pea and mustard (Hrazdina & Parsons, 1982; Beggs et al., 1987), but not in grapevines. The results presented here represent clear evidence of the light induction of flavonol biosynthesis and is the strongest indication yet that flavonols may play a lfV-protection role in grapevines. It appears that light differentially regulates expression of different genes in the flavonoid pathway in Shiraz. Further studies are required to elucidate the mechanisms by which light influences flavonol synthesis and expression of VvFLSI in Shiraz grapes while having no effect on parts of the flavonoid pathway that generate anthocyanins and tannins. Nevertheless, flavonols appear to be a good indicator of bunch exposure to light, in agreement with previous studies. Together, the results presented here and previous work suggests that flavonol synthesis in grape berries is highly light-dependent but that light does not greatly alter the total levels of anthocyanins and tannins in ripe Shiraz grapes. Thus, manipulating flavonol levels in grape berries through canopy management should be relatively straightforward. However, at this point it is not been equivocally 207 established whether flavonols contribute significantly to wine grape quality, either as potential copigments or by other mechanisms. 208 Chapter 6. General Discussion The aim of this project (Section 1.5) was to investigate the apparent paradox identified in previous research (Boss et al., 1996a), that genes involved in anthocyanin biosynthesis were expressed at times during grape berry development when there was no synthesis of anthocyanins. The phenylpropanoid pathway (Figure 1.6) gives rise to a number of products other than anthocyanins and it was hypothesised that the accumulation of these compounds would coincide with expression of genes in the pathway early in berry development. Thus, we investigated the accumulation of two groups of these products, the flavonols and tannins in the developing fruit of the grape cultivars Shiraz and Chardonnay. In addition, we wanted to explore the effect of light on berry development and the accumulation of flavonoids in Shiraz grape berries. 6.1. Tannin Accumulation in Shiraz Berries The flavan-3-ol and proanthocyanidin composition of seeds and skin of Shiraz gmpes was determined by reversed-phase HPLC after acetone extraction and acid-catalysis in the presence of excess phloroglucinol. The main period of proanthocyanidin accumulation in grape seeds occurred immediately after fruit-set with maximum levels observed around veraison. Over two seasons there was variation in both proanthocyanidin content and the timing of proanthocyanidin accumulation in seeds. In the skin, proanthocyanidin accumulation occurred from fruit-set until 1-2 weeks after veraison. Proanthocyanidin subunit composition was different in seeds and skin and changed during berry development. Proanthocyanidin levels in both seeds and skin decreased between veraison and harvest. Additional proanthocyanidin subunits were released when the residues remaining after acetone extraction were subjected to direct acid-catalysis in 209 the presence of phloroglucinol. In the seeds, these accounted for much of the post- veraison decrease, but not in grape skin. At harvest, 75yo of extractable berry proanthocyanidin was in the seeds. Accumulation of proanthocyanidins in the seeds appears to be independent of that in the skin, but in both tissues synthesis occurs early in berry development and maximum levels are reached around veraison. Thus, the pattern of proanthocyanidin accurnulation in Shiraz seeds and skin has been clearly established, however a number of areas remain ripe for further investigation. One of these is the post-veraison decline in proanthocyanidins, widely held to represent a decrease in extractability (Amrani-Joutei et a1.,1994; Cheynier et al., 1997; Saint-Cricq de Gaulejac et a.1., 1997; Kennedy et al. 2000b). The release of additional proanthocyanidin subunits by direct acid-catalysis of seed residues suggests these had become associated with other seed coat components. This was not the case with the skin residues and the nature of the decrease in extractable tannins in the skin during ripening remains unclear. The proposition that compounds resistant to acid-catalysis are formed and that these were not identified by this analysis (Section 3.4) suggests that a more thorough investigation of proanthocyanidin-like compounds in the HPLC separation of catalysis products may give an indication of the fate of grape skin tannins post-veraison. While the pattern of proanthocyanidin accumulation has been much clarified, the biochemistry of their formation remains less clear. Little or nothing is known of the process of polymerisation in condensed tannins, although it is generally thought that the chemistry of extension subunits inclines them to polymer formation as a chemical rather than an enzymatic process. Nevertheless, some pundits feel strongly that this is an enz5rmatic process and hold the ratios of subunits with differing stereochemistry as evidence. However, a body of work does indicate the ability of flavan-3-ols and flavan- 210 3,4-diols to polymerise non-enzymatically [utd, 1972; Delcour et a1.,1983; Bakker & Timberlake , 1997; Saucier et al., 1997; Mratous et al., 2002b; Vidal et al., 2002)' This work indicates that subunit ratios in the polymer are a function of favourable conformation in situations of steric hindrance and other constraints of athactive and opposing force. Per contra, the formation of dimers, particularly those comprising the 4-6 dimeric vinculum, in what appears to be a tissue specific manner, suggest that at least the dimerisation process may be enzymatic. Mechanisms for tannin polymer formation have long been mooted, however the formation of flavan-3-ol monomers had until recently been considered a simple reduction of leucocyanidin to catechin or epicatechin, with the formation of the two isomers the result of an as yet unidentified epimerase. However, recent work (Xie et al.,2003) has challenged the conventional wisdom by demonstrating the formation of epicatechin from cyanidin. In addition, a gene encoding this anthocyanidin reduct¿se was also cloned and functionality in carrying out this reaction was demonstrated. Coincidentally, similar work has recently reported cloning and characterisation of a gene encoding leucoanthocyanidin reductase (Tanner et a1.,2003;Harvey et al.,inprep.). Together these studies suggest two paths to condensed tannin formation, and indeed the subunit composition of tannin polymers may be determined by differential flux through the biosynthetic enzymes, leucoanthocyanidin reductase (LAR) and anthocyanidin reductase (ANR). Alternatively, the kinetics of the intermediate enzyme, leucoanthocyanidin dioxygenase (LDOX a.k.a. ANS) may be responsible for the predominance of epicatechin subunits in grape proanthocyanidins. However, much of the enzymology of flavonoid metabolism still remains to be thoroughly investigated. 2tt Despite recent advances in our knowledge, the course of tannin biosynthesis remains unclear. Our thinking previously had terminal subunits synthesised from leucoanthocyanidin by LAR and extension subunits originating from leucoanthocyanidin (Figure 6.14), with the involvement of an epimerase to generate the isomers catechin and epicatechin. The work of Xie et al. (2003) shows this is not the case, but rather that catechin and epicatechin have different origins, leading to a revision of our thinking about proanthocyanidin biosynthesis (Figure 6.18). The results reported by Xie et al. (2003) suggest there may not be an epimerase, and in that light one rnust reconsicler the origin of the extension subunits. Given the different origins of terminal subunits, it hardly seems fanciful that the extension subunits would have similar origins. Thus, the substrates for catechin and epicatechin monomers and terminal subunits, leucocyanidin and cyanidin respectively, would be the source of catechin and epicatechin extension subunits (Figure 6.1C). However, the absence of a hydroxyl moiety at the 4-position of "C" ring of cyanidin significantly weakens the argument for a purely chemical formation of proanthocyanidin polymers. This forces us to revisit the proposition of Haslam (1998) that a quinone methide intermediate in the LAR catalysed reaction is the extension subunit source. This, in the absence of such an intermediate in either the LDOX or ANR catalysed reactions, or a condensing enzyme, leaves Figure 6.18 as the most likely pathway for condensed tannin biosynthesis, based on our current knowledge. t Interestingly, recent research (Abrahams et a1.,2003;Xie et a\.2003) also clarifies one of the enigmas of the pattern of gene expression reported by Boss et al. (1996a), which showed LDOX expressed early in berry development. The conventional pathway (Figure 6.14) does not involve LDOX in tannin biosynthesis, but in the revised pathway (Figure 6.18, 6.1c) LDox and ANR are required for the synthesis of epicatechin. 212 4 A 5 Extasion Subunit Lacocya nidin LDOX Clanidin LAWBanyuls Bond UFGT Interllavd CateóinÆpicdedrin -> C¡anidin3gluoside Temin¿l Subunit B Exteß ion Subunit Lococ¡anidin LDOX Clanidin I.AR Inte rflavå n Bond UFGT I ANR t Cdeó in Tem¡nal Subun¡t Clanidin-3 gluc osi & Ep¡Gtæhin C ExleÉion Subunit Leuæcymidin LDOX Clanidin R hterflavÂnBond UFGT ANR Cdedtin Teminal Subunit Cyanidin-l€lucosiè - Epi€túhin Figure 6.1. Evolution of current thinking in proanthocyanidin biosynthesis: (A) Terminal subunits syáthesised from leucocyanidin by LAR/Banyuls; extension subunits derived directly from léucocyanidin. (B) Subsequent to Xie et at. (2003), with catechin and epicatechin synthesised from leucocyanidin and cyanidin respectively by LAR (leucoanthocyanidin reductase) and ANR (anthoðyanidin reductase). (C) Proposed origin of extension subunits with2,3-trans (catechin) and 2,3-cis (epicatechin) stereochemistry from leucocyanidin and cyanidin respectively. 213 So far, this discussion has considered only the dihydroxylated flavan-3-ols, catechin and epicatechin, but epigallocatechin extension subunits are also detected in the grape skin. Based on the foregoing discussion, it seems likely that that these are derived from the leucoanthocyanidin, delphinidin. It remains a mystery why epigallocatechin terminal subunits are not found in grape tissues, nor why the 2,3-trans isomer, gallocatechin is not detected. Interestingly, the bulk of epigallocatechin subunits accumulate from veraison onwards, suggesting that there is an upregulation of F3'5'H at this time. To date, there has been no published work on the activity of F3'5'H or the expression of the gene encoding F3'5'H in grapevines, nor on the regulation of the F3'5'H gene. once again, LDOX may provide an alternative explanation as a result of the increased expression of LDOX observed post-veraison (Boss et al., 1996a) presumably leading to an increase in LD OX c atalytic activity. The biochemistry of the galloylation of terminal and extension proanthocyanidin subunits also remains unclear. In grapevines, only epicatechin-gallate has been reported. Why only epicatechin is galloylated currently remains an enigma. That galloylation is limited by the stereochemistry of the 2,3-trans isomer catechin and that the timing of epigallocatechin accumulation does not coincide with the activity of an unidentified galloyltransferase, are possible explanations. Thus, there yet remain aspects of proanthocyanidin biochemistry to be elucidated. The number of tannin deficient mutants downstream of flavan-3-ol synthesis identified in Arabidopsis (Shirley et a1.,1995; Abrahams et a1.,2002) lends credence to this notion. The genes identified by these mutants may involve a number of processes including galloylation, dimerisation and polymerisation, the role, if any, of an epimerase, as well as 214 trans-membrane transport, or genes involved in the regulation of proanthocyanidin biosynthesis. Finally, the bulk of this research in grapevines has been on a single cultivar, Shiraz. It is not yet known whether the pattern observed in Shiraz is representative of all of Vitis vinifera or indeed the Vitiaceae. While much of what has been reported here (Chapter 3) is consistent with patterns observed in other cultivars by other workers, substantial deviations from what has previously been reported were also noted. Thus it would be of great viticultural and oenological importance to make a broader survey of tannin accumulation in winegrapes. 6.2. Flavonol Accumulation in Shiraz Berries The predominant flavonols in Shiraz and Chardonnay berries are quercetin-3-glucosides with trace amounts of kaempferol-3-glucosides detected in Shiraz flowers, but not in developing berries. Flavonols were present in the skin of ripening grapes, but were not detected in the seeds or flesh. Flavonols were also present in buds, tendrils, inflorescences, anthers and leaves of the grapevine. The flavonol concentration in flowers was high and decreased between flowering and berry-set then remained relatively constant through berry development. Total flavonols per berry were low until pre- veraison then increased during berry development, particularly before veraison in Chardonnay and during ripening in Shiraz. In addition to determining the pattern of flavonol accumulation during berry development, genes encoding the biosynthetic enzyme flavonol synthase were also cloned. Two cDNA fragments with homology to genes encoding the enzyme flavonol synthase (FLS) were isolated from Shiraz flowers. One of these, VvFLSl, was expressed 2t5 in leaves, tendrils, pedicels, buds and inflorescences as well as in the developing gïapes. Expression was highest between flowering and fruit-set then declined, increasing again during ripening coincident with the increasè in flavonols per berry. The second gene, VvFLS2, was constitutively expressed at a very low level in berries throughout berry development. The pattem of VvFLSI expression indicates two distinct periods of flavonol synthesis during berry development, the first around flowering and the second during ripening of the berries. The pattern of flavonol accumulation and VvFLSI expression in developing berries is consistent with the overall pattern of gene cxpression in the phenylpropanoid pathway (Boss et al., 1996a) and may account for part of the early expression of pathway genes. In this investigation, flavonols were measured as the combined total of quercetin-3- glucoside and quercetin-3-glucuronide. These compounds not being readily separated by the method used here. This was considered adequate, as the aim was to investigate the accumulation of total flavonols and changes in expression of the tr'Z^! gene. What remains unanswered here is whether there were compositional changes between the glucoside and the glucuronide during berry development. These modifications of flavonols subsequent to the formation of the aglycone by FLS have not been investigated. While a range of glycosylation products including, flavonol glucosides, glucuronides, galactosides, glucosylgalactosides, glucosylxylosides and glucosylarabinosides have previously been identified in grapevines (Cheynier & Rigaud, 1986), no flavonol glycoside transferase has yet been identified. To date the only flavonol glycosylhansferase gene characterised has been in Petunia (Miller et al., 1999; Miller et al., 2002). Although flavonol glycosyltransferase activity has been assayed in Prunus x yedoensis and Vigna mungo 216 (Ishikura & Yamamoto, 1990; Ishikura & Mato, 1993) and a putative gene sequence identified tn Arabidopsis. Whether such transferases are specific for flavonols or are similar to or the same as those glycosylating other flavonoids is unknown. However, previous research (Ford & Hoj, 1998; Ford et a1.,1998) indicates that the enzyme glucosylating anthocyanidins to form anthocyanins, UFGT, has a high level of substrate specificity. Furthermore, UFGT is not expressed before veraison when significant synthesis of flavonol glycosides occurs. Thus, it is likely that there are specific glycosyltransferases for different classes of flavonoids, such as flavonols. 6.3. Effect of Shading on Berry Development and Flavonoid Biosynthesis To explore the effects of light on berry development and flavonoid biosynthesis, opaque boxes were applied to bunches of Shiraz grapes prior to flowering. The boxes were designed to exclude light while maintaining airflow to minimise changes in temperature and humidity. There was no significant effect of shading on sugar accumulation and in two of the three seasons studied there was no effect on berry weight. Thus, we concluded that berry development was not significantly effected by shading. The chlorophyll concentration was much lower in the shaded fruit, which appeared pale yellow until veraison. Post-veraison, the fruit coloured normally in the shaded bunches and in two of the three seasons, there was no significant change in anthocyanin content. However, in one season, total anthocyanins were substantially lower in the shaded fruit, around half the level in the exposed fruit, and the cause of this difference remains unclear. Despite these differences in total anthocyanins between shaded and exposed fruit, expression of the gene encoding UDP-glucose flavonoid-3-O-glucosyl transferase (UFGT), a key gene in anthocyanin synthesis, was similar between treatments. That gene expression \ryas so 217 similar while accumulation differed by around 50% suggests a post-translational effect of light on anthocyanin accumulation. This difference in anthocyanin accumulation between shaded and exposed fruit seems difficult to reconcile in both cause and outcome. Two possibilities assert themselves as likely causes, that the difference was caused by the early application of the shading treatment, five weeks pre-flowering as opposed to fruit-set, or by the unusually high temperatures recorded during that season. As for the outcome, early shading or high temperature may cause reduced flux through the biosynthetic pathway, limited by the expression of other gcncs upstream of UFGT, not examined here, resulting in reduced substrate availability for anthocyanin synthesis. Alternatively, high temperature has been shown to cause anthocyanin degradation in grapes and other plant systems (Buthose et al.,l97l;Päe,1977; Markakis, 1982; Crippen & Morrison, r986a; Yamasaki et al., 1996) and that supplementary anthocyanin biosyrthesis in response to degradation is light induced. To explore these possibilities, it would be necessary to test the effect of the timing of shading application as well as the effects of temperature independent of light. While the effect of shading on total anthocyanins exhibited some seasonal variation, the anthocyanin profile of the shaded fruit was consistently altered. The shaded fruit had a greater proportion of dihydroxylated anthocyanins, the glycosides of cyanidin and peonidin, while the exposed fruit had a higher proportion of thc trihydroxylated anthocyanins, the glycosides of delphinidin, petunidin and malvidin. The suggestion implicit in this result is that the enzyme F3'5'H, or the gene encoding F3'5'H, is light sensitive and activity or expression is downregulated in low light conditions. To date this area of flavonoid biochemistry has been little explored other than in Petunia (flolton er al., 1993a). Given the possibility of downregulation of F3'5'H in low light conditions 218 suggested by changes in grape skin composition, the immediately obvious experiment would be to determine if the reverse trend was observed under high light conditions' Further, if F3'5'H is Sensitive to high temperature, the combined effect of shade and temperature could account for the lower colour in the shaded fruit during the hot 2000- 2001 season. Another possibility is that the shift in anthocyanin profile with shading results in a pool of anthocyanins that are less thermally stable than those favoured by high light conditions. Given that generally high light and high temperature would be closely linked under normal environmental and growth conditions, this seems a likely scenario. Thus, exploring the temperature effects on anthocyanin biosynthesis and chemistry both in vivo andin vitro wotldbe of considerable interest. While the glycosylation pattern of anthocyanins was not consistently changed with shading, in the hotter 2000-2001 season when significant differences in total anthocyanins were observed, there were also shifts in composition based on glucoside substitution. Similar changes in composition have been previously reported flaselgrove et a1.,2000; Spayd et al., 2002) and the indication is that temperature rather than light is the primary influencing factor. The general trend with increasing temperature is an increase in coumaroylated anthocyanin glucosides and a decrease in acetylglucosides and non-acylated glucosides. While this effect is yet equivocal, these observations strongly encourage investigation of the effects of temperature on flavonoid biosynthesis in grapevines independent of the influence light. If this effect on anthocyanin composition based on the glycosylation pattern is indeed effected by temperature it would be of interest to examine the biochemical mechanism by which this occurs, whether by altering enzyme kinetics or regulating gene expression. 219 In contrast to anthocyanin accumulation, shading appears to have little effect on grape tannin accumulation. Shading had virtually no effect on the levels of condensed tannins in grape seeds, while the level of condensed tannins in the ripe skin of shaded grape berries was not significantly different to that of the exposed fruit. However, there were significant differences early in berry development, particularly in the levels of extension subunits. The post-veraison decrease in apparent extractability of skin tannins resulted in similar tannin levels in the skins at harvest. In addition to this difference, there were also differences in skin tannin composition bascd on the "8" ring hydroxylation pattern. It is arguable that many of the factors causing similar shifts in anthocyanin composition are applicable to tannin subunits composition in response to light and temperature. Of these factors, the possible light-sensitivity of F3'5'H stands out, however, the changes observed here were only in the order of ten to fifteen percent of extension subunits, five to ten percent of anthocyanins. It is unlikely that such a small shift in t¿nnin composition would represent a quantifiable shift in gene expression. A more noticeable shift in tannin composition with shading was in the proportion of galloylated epicatechin extension subunits. While extension subunits were almost 40yo lower in the shaded fruit (at 24Bdrx), the proportion of galloylated extension subunits doubled. Whether this difference is discernible through molecular techniques remains to be seen, and awaits the cloning and characterisation of a galloylating enzyme. t Of the flavonoids studied here, the greatest, consistent effect of shading was on flavonol biosynthesis. When shading was applied before flowering, shaded fruit had much lower levels of flavonols throughout berry development and at harvest the level of flavonols were less than l0o/o of that in exposed fi:uit. A gene encoding flavonol synthase (VvFLSI) was expressed at flowering and during ripening in exposed grapes but its expression was 220 geatly reduced in shaded fruit. Flavonol accumulation and VvFLSI expression were also greatly reduced in shaded leaves and developing inflorescences. Furthermore, when this shading treatment was removed and tissues sampled three days later, flavonol levels and VvFLSI expression increased in response to exposure to light. In the newly exposed inflorescence the increase from the shaded tissue was around l4-fold, while the increase in the newly exposed leaf was more than 120-fold that in the shaded leaf. This clearly demonstrates that flavonol accumulation and VvFLSI expression in grapevines are light induced. 'What is still unclear is whether increased exposure of the fruit would result in a further increase in flavonol accumulation and VvFLSI expression, is there an upper limit to the flavonol level in a tissue? It is also uncertain whether changes in total flavonols have any impact on flavonol composition. It would be interesting to see if there was a shift from one glycosidic form to the other with changing levels of light exposure. Indeed, in terms of light exposure it would be interesting to compare not only light quantity, but also light quality, by excluding selected wavelengths. 6.4. Implications þr Winemaking The outcomes of this research indicate that shading had little effect on berry development and ripening and while total anthocyanins and tannins were not greatly effected by shading there were consistent changes in composition. Shading did however significantly decrease flavonol synthesis. While much of this is of considerable interest from a physiological and biochemical perspective, the gteater immediate importance is how this new knowledge might impact upon the final wine and upon winemaking techniques. 221 Changes in grape berry composition indubitably impact upon wine composition and therefore wine quality, however many of these effects have not been thoroughly explored or quantified. One of the shortcomings of this, ãnd much other, research into grape berry composition is that the analysis terminates with the fruit at harvest, rather than making the logical progression into winemaking. The role of flavonols as a sunscreen in grape skin is consistent with the observed changes in flavonol accumulation due to shading and exposure. It has also been demonstrated that flavonols have a signifieant copigment effect with malvidin-3-glucoside (Lambert,2002). However, the role of flavonols as copigments either in the berry or in the wine has not been firmly established. Research has shown flavonols and anthocyanins to co-occur in the vacuole of skin cells (Flrazdina & Moskowitz, 1980; Moskowitz & Hrazdina,l981), although the concentrations were considered insufficient for a copigmentation effect. Further, the levels present in the grapes, and must, were also considered inadequate for any appreciable copigmentation effect in the wine (Lambert,2002). Of more importance is the observation that greatest copigmentation effect results from the self-association of malvidin-3-glucoside (Hoshino et a\.,1982; Houbiers, et al.,1998; Lambert, 2002). Thus, while it would be of great interest to make wine from shaded, normally exposed and highly exposed fruit to examine the effects of changes in flavonol content on wine colour, it would be equally interesting to observe the effects in wine from treatments that incline the anthocyanin composition of the fruit towards malvidin-3-glucoside. Based on the work presented here and other recent research (Downey, unpublished) these conditions would be high light and low temperature. While the indication is that high light and cool temperatures are likely to favour the accumulation of malvidin-3-glucoside, these conditions cannot practicably be achieved in 222 many of the world's winegrowing regions. Thus it is also important to determine the combinations of light and temperature that are most favourable to wine colour, as well as mouthfeel and other quality parameters. One of the aspects of wine colour that has not been thoroughly investigated is the relative extractability and stability of different classes of anthocyanins during winemaking, nor indeed of other classes of flavonoids, such as flavonols and tannins. If viticultural management practices are aimed at increasing grape colour and the shift in the anthocyanin profile that increases grape colour results in less extractable or less st¿ble anthocyanins, increased grape colour may translate into decreased wine colour. Thus, it becomes essential to explore the relationship between the anthocyanin profile of the fruit and the colour intensity and stability of the final wine. 6.5. Summary As a result of these investigations \Me can now revise our schematic representation of grape development (Figure 2.1) to include flavonols and tannins. While this was not þerry a new field of investigation, much of the published literature, for whatever reason, had dealt only with individual aspects of flavonoid biosynthesis in Vitis viniferø. Seeds, but not skin or vice versa, anthocyanins, but not tannins, etc. and included a diverse array of varieties, such that while it was possible to get a general impression of flavonoid biosynthesis in grapevines, no definitive study of all flavonoids in the same cultivar existed. In addition, the extant literature included many contradictions arising from analytical, seasonal and varietal differences, which made comparing and contrasting these data problematic. Here we have addressed many of these issues and have presented a body of work that clearly shows the pattern of flavonoid accumulation in Vitis vinifera cv. Shiraz over several seasons. Thus, the revised schematic (Figure 6.2) now includes flavonol and 223 FRUITt-t SET VERAISONr-l RIPENESSf-l BERRY SIZE I I SUGAR ¿ COLOUR FLAVOUR FL ONOLS TANNINS Time after flowering Figure 6.2. Reviserj st:lie¡¡ra.iic representaiion of grape berry deveiopment inciuding the prociucts of this resea,rch. the a,ccuinulatiorr o [ flavonols anel- ta.nnins in the berry tannin accumulation on a whole berry basis, augmented by detaileel eompositional anaiysis'both spatraiiy and temporally (Chapter i;Çhapter $. ':] The second area of investigation presented in this research was the effect of shading on both berry development ancl flavonoid biosynthesis in the grape. while be development was largely unaffected by shading, there were oliserveel dilÍercnces in ihe f,tavonoid accumuiation anci fiavonoid gene expression between shaded and normally cvnneeri firrif fllenerollr¡ fhooo lì-:+^,1 +^ {-I^-.^-^:l^:* ¿.L^ ^r-i.^ ¡srrj, ûarvùu ^ho--vrrcrrõvù vvuru rrr¡riilll Lv rrcvuitulL!ù ili tiiù Ëridpç--^.^^ bKrir. Ct-ritsisteiriiy íia-vr-rnoi biosynthesis was reriueeci in the shacieci fiuit, whiie eha-nges in tctal 224 anthocyanin and tannin accumulation adhered to no discernible pattern. However, closer examination of the data presented here showed that consistent, albeit subtle, changes in anthocyanin and tannin corhposition did occur in response to shading. These changeS had not previously been reported. While the effect of shading has been closely examined here, there arises an interesting counterpoint, whether the effect of increasing light would result in a reversal of the changes observed. Furthermore, as discussed in Chapter 5, it remains uncertain whether or not changes athibuted to bunch exposure are effects of changes in the light environment or in fact, temperature effects. While much of the data here is unequivocal, given that the shading treatment did not have a significant temperature effect, the effect of temperature independent of and interacting with light should be fuither investigated. This research represents a comprehensive analysis of these aspects of flavonoid biosynthesis in gtapevines, however many aspects of flavonoid biogenesis remain to be elucidated. While this research has closely examined the cultivar Shiraz and to a lesser extent Chardonnay, the patterns observed here may not be representative of the whole Vitis genus or Vitaceae family. Indeed, earlier investigations of other cultivars, such as Pinot noir, are not entirely consistent with the observations presented here. Given that broad and fundamental differences in flavonoid composition between cultivars have been extensively documented elsewhere, such inconsistency is to be expected. Indeed, this variability between cultivars forces us to ask whether the trends observed here represent a general pattem or an anomaly in the Vitaceae? Is there genetic variability that could be exploited to improve fruit quality by changed flavonoid biosynthesis/composition? Thus, it is necessary to expand these investigations to other Vitis cultivars to determine whether the phenomena observed in Shiraz, spatially, temporally and with respect to light 225 exclusion, is universal or specific. Descrþtion of a very general trend, or extrapolating a general trend from isolated examples establishes an expectation of a single management practice applicable to wide array species õr cultivars. However, it is becoming increasingly apparent that this is not appropriate and that targeted practices will be required for individual cultivar, particularly with wineries tending more towards payment based on quality parameters, such as colour. Thus, the industry will be driven to take a more targeted approach to viticultural management, which will inevitability highlight many of the shortcomings of our current knowledge. Indeed, it will become increasingly apparent that much of the extant body of viticultural research needs to be revisited using modern analytical techniques. 226 Bibliography A acidln: Plant Flavonoids in Biology and Medicine II: Biochemical, Cellular, and Abrahams, S., Tanner, G.J., Larkin, P.J. and Medicinal Properties (Cody et a/., eds.) pp. Ashton, A.R. (2002) Identification of 235-241. Alan R. Liss, Inc., New York biochemical characterization of mutants in the proanthocyanindin pathway in Arabidopsis, Artaud-V/ild, S.M., Connor, S.L., Sexton, G. and (1993) Differences in coronary P I ant P hys io lo gy 130 ; 561-57 6 Connor, W.E. mortality can be explained by differences in Albert, S., Delseny, M. andDevic, M. (1997) cholesterol and saturated fat intakes in 40 BANYULS, a novel negative regulator of countries but not in France and Finland: A flavonoid biosynthesis in the Arabidopsis seed paradox, C ir culation 88; 27 7 I -27 7 9 coat, Plant Journal ll;289-299 Asen, S. (1975) Factors affecting flower colour, Alonso, E., Bouzeix, M. and Revilla, E. (1991) Acta Horticulturae 4l1' 57 -68 Suitability of water/ethanol mixtures for the extraction of catechins and proanthocyanidins Asen, R., Stewart, R.N. and Norris, K.H. (1972) from Vitis vinifera seeds contained in a Co-pigmentation of anthocyanins in plant winery by-prodlct, Seed Science and tissues and its effect on colour, Technology 19;545-552 P hytoch emistry ll; I 139 -l I 44 Amrani-Joutei, K., Glories, Y. and Mercier, M. Australian Bureau of Statistics (2001) Australian (1994) Localization oftannins in grape berry ll'ine and Grape Industry, Australian Bureau skins, Zilrs 33; 133-138 of Statistics, Canberra. Anderson, O.M. and Francis, G.W. (1985) B analysis of anthocyanins and Simultaneous Bakker, J., Bridle, P., Timberlake, C.F. and thin layers, anthocyanidins on cellulose Arnold, G.M. (1986a) The colours, pigment Journal of Chromatography 318; 450-454 and phenol contents ofyoung port wines: Antcliff, A.J. (1988) Taxonomy - The grapevine Effects of cultivar, season and site, I/itis 25; as a member of the plant kingdom In: 40-52 Yolume l. Resources (Coombe & Viticulture. Bakker, J., Preston, N.W. andTimberlake, C.F. Winetitles, Adelaide Dry, eds.) pp.2ll. (1986b) The determination of anthocyanins in Archer, E. and Strauss, H.C. (1989) Effect of aging red wines: Comparison of HPLC and shading on the performance of Vitis vinifera spectral methods, American Journal of L. cv. Cabemet saunignon, South African Enology and Viticulture 37; 12l-126 ofEnology and Viticulture l0;74-77 Journal Bakker, J. andTimberlake, C.F. (1997) Isolation, Armand, J.P., de Forni, M., Recondo, G., Cals, identihcation, and characterization of new L., Cvitkovic, E. andMunck, J.N. (1988) color-stable anthocyanins occurring in some Flavonoids: A new class ofanticancer agents? red wines, Journal ofAgricultural and Food Preclinical and clinical data of flavone acetic Chemistry 45;35-43 221 Baldi, A., Romani, 4., Mulinacci, N., Vincieri, differential regulation of anthocyanin and F.F. and Ghiselli, A. (1997) The Relative quercetin formation, Planta 1721' 12l-126 Antioxidant Potencies of some Polyphenols in Beggs, C.J. andWellmann, E. (1994) Grapes and Wines. American Chemical Photocontrol of fl avonoid biosynthesis In : Society, ffashington Photomorphogenesîs in Plants (Kendrick & Baranac, J.M., Petranovic, N.A. andDimitric- Kronenberg, eds.) pp. 733-751. Kluwer Markovic, J.M. (1997) Spectroscopic study of Academic Publishers, Dordrecht anthocyanin copigmentation reactions., Bel-Tal, Y. and King, R.W. (1997) Journal of Agricultural and Food Chemistry Environmental factors involved in colouration 44;1333-1336 of flowers of Kangaroo Paw,scientia Barritt, B.H. and Einset, J. (1969) The Horticulturae 7 2 ; 35-48 inheritance of three major fruit colours in Bergqvist, J., Dokoozlian, N. and Ebisuda, N. grapes, Journal oJ the American Societyfor (2001) Sunlight exposure and temperature Hort icultural S cienc e 9 4 ; 87 -89 effects on berry growth and composition of Bate-Smith, E.C. ( I 954) Leuco-anthocyanidins Cabemet Sauvignon and Grenache in the 1 Detection and identiflrcation of central San Joaquin Valley of Califomia, anthocyanidins formed from leuco- American Journal of Enology and Viticulture anthocyanidins in plant tissues, Biochemistry 52; l-7 Journal 58;122-125 Berry, C.J.J. (1987) First Steps in lI/inemaking Bateson, W. (1901) Experiments in plant Amatuer Winemaker, London hybridisation , Journal of the Royal Blanco, Y.2., Ãttw, J.M., Sims, C.A. and Horticultural Society 26; l-32 O'Keefe, S.F. (1998) Effect of processing on Baur, P.S. andWalkinshaw, C.H. (1974) Fine phenolics of wines In: Process-Induced structure of tannin accumulation in callus Chemical Changes in Food (Shahidi et al., cultures of Pinus elliotti (slash pine), eds.) pp. 327-339. Plenum Press, New York Canadían Journal of Botany 52; 615-619 Bors, W., Michel, C. and Stettmaier, K. (2001) Bavage,4.D., Davies, LG., Robbins, M.P. and Structure-activity relationships goveming Morris, P. (1997) Expression of an antioxidant capacities of plant polyphenols In A n t irr hinum dihydrofl avonol reductase gene Flavonoids and other Polyphenols (Packer, results in changes in condensed tannin ed.) pp. 166-180. Academic Press, San Diego structure and accumulation in root cultures of Bors, W. and Saran, M. (1987) Radical L o t u s c o r n i c tt I a t u s (bir d' s f o ot tr ef o iI), P I a n t scavenging by flavonoid antioxidants, Free Mo I e cul ar B iolo gy 3 5 ; 443 -458 Radical Re s e arch Commun icat i o ns 2 ; 289 - Beggs, C.J., Kuhn, K., Bocker, R. and 294 Vy'ellmann, E. (l 987) Phytochrome-induced Boselli, M., Volpe, B. and di Vaio, C. (1995) flavonoid biosynthesis in mustard (Sinapis Effect of seed number per berry on mineral albaL.) cotyledons. Enzymic eontrol and composition of grapevine (Vitis vinifera L.) 228 berries, Jour nal of Horticultural Science 70 ; Boukharta, M., Girardin, M. andMetche, M. 509-5 l5 (1994) Oligomeric procyanidins of vine shoots of cultivar Carignan, Vitis 33; 5-10 Boss, P.K. (1998) The Molecular Biology of Anthocyanin Biosynthesis in Grape Berry Boulton, R.B. (2001) The copigmentation of Ski¡¿s, PhD Thesis, Department of anthocyanins and its fole in the colour ofred Horticulture, Viticulture and Oenology, wine: A critical review, American Journal of University of Adelaide, Adelaide Enology and Viticulture 52;67-87 Boss, P.K., Davies, C. and Robinson, S.P. Boulton, R.B., Singleton, V.L., Bisson, L'F. and (1996a) Analysis ofthe expression of Kunkee, R.E. (1998) Principles and Practices anthocyanin pathway genes in developing of Winemaking. Chapman & Hall, New York L. cv Shiraz grape berries and Vitis vinifera Bourquin, J-C., Sonko, 4., Otten, L. and Walter, pathway rcgulation, Plant the implications for B. (1993) Restriction fragment length 1 1; 1059- 1066 Physiology I polymorphism and molecular taxonomy in Boss, P.K., Davies, C. and Robinson, S.P. Vitis vinifera L,Theoretical and Applied ( I 996b) Expression of anthocyanin Genetics 87; 431-438 pathway genes in red and white biosynthesis Bourziex, M., Weyland, D. andHeredia, N. grapes, Molecular Biology 32;565-569 Plant (l9SO Étude des catéchins etdes Boss, P.K., Davies, C. and Robinson, S.P. procyanidols de la grappe de raisin, du vin et ( I 996c) Anthocyanin composition and d'autre dérivés de la vigne, Bulletin de anthocyanin pathway gene expression in L' O. L Y. 669 -67 0 ; I 17 | -125 4 gtapevine sports differing in berry skin colour, Brossaud, F., Cheynier, V. andNoble, A.C. and lYine Australian Journal ofGrape (2001) Bittemess and astringency of grape Research 2; 163-170 and wine polyphenols, Australian Journal of Botha, J.J., Viviers, P.M., Ferreira, D. and Roux, Grape and ll'ine Research 7 ;33-39 (19S2) tannins: Competing D.G. Condensed Brouillard, R. and Dangles, O. (1994) in biomimetic nucleophillic centres Anthocyanin molecular interactions : the flrrst condensation reactions, Phytochemistry 2l; step in the formation of new pigments during t289-t294 wine aging?, f'o od Chemistry 5l;365-371 Brum-Bousquet, M. (1990) Boucheny, A. and Brouillard, R., George, F. andFougerousse, A. study the leaves of red grapes Biochemical of (1997) Polyphenols produced during red wine (Yitis Plantes Medicinales et viniferaL.), ageing, BioFactors 6; 403-410 Phytotherapie 241' 17 9 -192 Brouillard, R.,Mazza, G., Saad,2., Albrecht- M. andMetche, M. Boukharta, M., Girardin, Gary, A.M. andCheminat, A. (1989) The ( I 98 Galloylated procyanidins of grapevine 8) copigmentation reaction of anthocyanins: A (Vitis shoots. Separation and vinifera) microprobe for the structural study of aqueous by high performance liquid identif,rcation solutions, Journal of the American Chemical chromatography and gas chromatography, Society lll;2604-2610 Journal of Chromatography 455; 406-409 229 Bureau, S.M., Baumes, R.L. andRazungles, A.J under CO2-enriched conditi ons, New (2000) Effects ofvine or bunch shading on Phytolo gist 122; 461-467 glycosylated flavour precursors in grapes if Chalmers, D.J. and Faragher, J.D. (1977) Vitis vinifera L. cv. Syrah, Journal of Regulation of anthocyanin biosynthesis in Agricultural and Food Cheniisiry 48-09A- apple skin. I. Comparison of the effects of 1297 cyclohexamide, ultraviolet light, wounding Butler, L.G. (1982) Relative degree of and maturity, Australian Journal of Plant polymerisation of sorghum tannin during seed Physiology 4;111-l2l development and maturation, Journal of Charrier, B., Coronado, C., Kondorosi, A. and Agricultural and Food Chemistry 30; 1085- Ratet, P. (1995) Molecular characterization I 089 and expression of alfalfa (Medicago sativaL) Buttrose, M.S. (1970) Fruitfulness in grape- fl avanone-3 -hydroxylase and vines: The response ofdifferent cultivars to dihydrofl avonol-4-reductase encoding genes, light, temperature and daylength, Vitis 9;121- Plant Molecular Biology 29;773-86 125 Cheyrier, V., Prieur, C., Guyot, S., Rigaud, J. Buttrose, M.S. and Hale, C.R. (1973) Effect of and Moutounet, M. (1997) The structures of temperature on development of the grapevine tannins in grapes and wines and their inflorescence after bud burst, American interactions with proteins In: Wine: Journal of Enology and Viticulture 24; 14-16 Nutr iti o nal and Therap eutic B en efi ts (Watkins, ed.) pp. 8l-93. American Buttrose, M.S., Hale, C.R. and Kliewer, W.M. Chemical Society, Washington D.C. (1971) Effect of temperature on the composition of Cabernet Sauvignon berries, Cheynier, V. and Rigaud, J. (1986) HPLC American Journal of Enologlt and Viticulture separation and characterisation of flavonols in 22;71-7 5 the skins of Vitis vinifera var. Cinsault, C American Journal of Enology and Viticulture 37;248-252 Cawthorn, D.L. and Morris, J.R. (1982) Cholet, C., Fougere-Rifot, M. Relationship of seed number and maturity to andBouard, J. (1998) Cellular peculiarities berry development, fruit maturation, hormonal of medium size shot grape berries changes, and uneven ripening of'Concord' of Vitis vinifera L. var. MerIot, Journal Internatiottal des Sciences de (Vitis labruscaL.) grapes, Journal o.f the la Vigne et du Vin 32; 193-201 American Society for Horticultural Science 107 ; 1097 -1104 Clarke, H.R., Leigh, J.A. and Douglas, C.J. (1992) Molecular signals in the Chabot, S., Bel-Rhlid, R., Chevenert, R. and interactions between plants and microbes, Piche, Y. (1992)Hyphal growth promotion in Cell 7l;191-9 vitro of the VA mycorrhizal fungu Gigaspora Constantinides, M. andFownes, J.H. (1994) margarita Becker & Hall, by the activity of Tissue-to-solvent ratio and other factors structurally specifie flavonoid compounds affecting determination of soluble polyphenols in tropical leaves, 230 Communications in Soil Science and Plant Coombe, B.G. and McCarthy, M.G. (1997) Analysis 25;3221-3227 Identification and naming of the inception of aroma development in ripening grape berries, Coombe, B.G. (1973) The regulation of set and Australian Journal of Grape and Wine development of the grape berry, Acta Research3; l8-20 H orticulturae 3 4 ; 261 -2'l 3 Coombe, B.G. and McCarthy, M.G. (2000) Coombe, B.G. (1976) The development of fleshy Dynamics of grape berry growth and fruits, Annual Review of Plant Physiology 27; physiology of ripening, Australian Journal of 207-228 Grape and Wine Research 6; l3l-135 Coombe, B.G. (1987a) Distribution of solutes Cork, S.J. andKrockenberger, A.K. (1991) within the developing grape berry in relation Methods and pitfalls of extracting tannins and to its morphology, American Journal of other phenolics from plants: Insights from Enology and Viticulture 38; 120-127 investigations on Eucalyptus leaves, Journal Coombe, B.G. (1987b) Influence of temperature of Chemical Ecology 17 ; 123-134 on composition and quality of grapes, Acta Creasy, L.L. and Coffee, M. (1988) Phytoalexin H o r ticultur ae 20 6; 23 -3 5 production potential of grape benies, Journal B.G. (1988) Grape phenology In: Coombe, of the American Societyfor Horticultural (Coombe Viticulture. Volume L Resources & Science ll3;230-234 Dry, eds.) pp. 139-153. Winetitles, Adelaide Crippen, D.D. and Morrison, J.C. (1986a) The (1989) The grape berry as a sink, Coombe, B.G. effects of sun exposure on compositional Acta Horticulturae 239 ; 1 49- 1 58 development of Cabernet Sauvignon berries, Coombe, B.G. (1992) Research on development American Journal of Enology and Viticulture and ripening of the grape berry, Amerícan 37;235-242 Titiculture 43; Journal ofEnology and l0l- Crippen, D.D.J. and Morrison, J.C. (1986b) The ll0 effects of sun exposure on the phenolic Coombe, B.G. (1995) Adoption of a system for content of Cabemet Sauvignon berries during identifying grapevine growth stages, development, American Journal of Enology Australian j ournal of grape and wine research ønd Viticulture 37; 243-247 1; 104- I 10 Criqui, M.H. and Ringel, B.L. (1994) Does diet Coombe, B.G. and Bishop, G.R. (1980) or alcohol explainthe Frenchparadox? The Development of the grape berry. IL Changes Lancet344; 1719-1723 and during veraison, in diameter deformability Czochanska, Z.,Foo, L.Y. and Porter, L.J. Research Australian Journal of Agricultural (1979) Compositional changes in lower 3l;499-509 molecular weight flavans during grape Coombe, B.G. and Hale, C.R. (1973) The maturation, Phytochemis try 18; 1819-1822 ofripening grape berries and hormone content Czochanska, Z.,Foo, L.Y., Newman, R.N. and growth treatments, the effects of substance Porter, L.J. (1980) Polymeric P I ant P hysio lo gy 5l ; 629 -63 4 proanthocyanidins. Stereochemistry, structural 231 units, and molecular weight, Journal of the grape seeds by liquid secondary ion mass Chemical Society, Perkin Transcations I; spectrometry, Phytochemistry 49 ; 1435-1441 2278-2285 Deikman, J. and Hammer, P.E. (1995) Induction D of anthocyanin accumulation by cytokinins in Arabidopsis thaliana, Plant Physiology 1081. Daigle, D.J. and Conkerton, E.J. (1983) Analysis 47-57 of flavonoids by HPLC,Journal of Liquid Chromato graphy 6; 105-ll8 Delcour, J.4., Ferreira, D. and Roux, D.G. (1983) Synthesis ofcondensed tannins. Part Dakora, F.D., Joseph, C.M. and Phillips, D.A. 9. The condensation sequence (1993) Alfalfa (Medicago sativaL.) root of leucocyanidin with (+)-catechin and with the resultant exudates contain isoflavonoids in the presence procyanidins, Journal of of Rhizobium meliloti, Plant Physiology l0l; the Chemical Society, Perkin 8t9-824 Transactions l; lTll-1717 Dercks, W. and Creasy, L.L. (1989) The Darne, G. (1982) The phenolic compounds of the signihcance of stilbene phytoalexins in internodes of Vitis vinifera L. cv. Ugni blanc the P I as mop ara vitico I a- grapevine interaction, during growing season: Study ofthe evolution Physiological and Molecular Plant of their structure, Vitis 2l; 15-21 Pathology 34;189-202 Darne, G. (1993) New hypotheses on Devic, M., Guilleminot, J., Debeaujon, I., anthocyanin biosyntheses in berries and leaves Bechtold, N., Bensuade, E., Koorneef, of the grapevine, Vitis 32;77-85 M., Pelletier, G. and Delseny, M. (1999) The Darne, G. (1996) Grapevine anthocyanins: BANYULS gene encodes a DFR-like protein Differences between leaves and berries, and is a marker of early seed coat Journal International des Sciences de la development, P lant Jour nal 19 ; 387 -398 Vigne et du Vin,Special Issue; 79-83 Dewick, P.M. and Haslam, E. (1969) Phenol de Freitas, V.A.P. and Glories, Y. (1999) biosynthesis in higher plants: Gallic acid, Concentration and compositional changes of B i oc hemic al Journal ll3 ; 537 -5 42 procyanidins in grape seeds and skin of white Diallinas, G. and Kanellis, A.K. (1994) A Vitis vinifera varieties, Journal of the Science phenylalanine ammonia-lyase gene from of Food and Agriculture 79; ßAl-1606 melon fruit: cDNA cloning, sequence and de Freitas, V.A.P., Glories, Y. and Monique, A. expression in response to development and (2000) Developmental changes of wounding, Plant Molecular Biology 26;473- procyanidins in grapes of rcd Vitis vinifera 479 varieties and their composition in respective Dixon, R.A. and Lamb, C.J. (1990) Molecular wines, American Journal of Enology and communications in interactions between Viticulture 5l; 397 -403 plants and micrcbial pathogens, Annual de Freitas, V.A.P., Glories, Y., Bourgeois, G. Review of Plant Physiology and Plant and Vitry, C. (1998) Characterisation of Mo I ecular Biology 41; 339 -3 6l oligomeric and polymeric procyanidins from 232 Dixon, R.A. andPaiva, N.L. (1995) Stress- berries of two grapevine cultivars, Vitis 4l; induced phenylpropanoid metabolism, The 63-67 PlantCell 7;1085-1097 E M.4., Redmond, J.W., Batley, M. Djordjevic, Ebadi,4., May, P., Sedgley, M. and Coombe, B.G. (1987) Clovers secrete and Rolfe, B.G. (1995a) Fruit set on small Chardonnay specific phenolic compounds which either and Shiraz vines grown under varying or repress nod gene expression in stimulate temperature regimes between budburst and Rhizobium trifolii, EMBO Journal 6; ll'13- fTowering, Australian Journal of Grape and tt79 l[/ine Research l; l-10 Do, C.B. and Cormier, F. (1991) Accumulation Ebadi,4., May, P., Sedgley, M. and Coombe, peonidin 3-glucoside enhanced by osmotic of B.G. (1995b) Effect of low temperature near grape (Vitis vinifera L.) cell stress in flowering time on ovule development and Tissue and Organ suspension, Plant Cell, pollen tube growth in grapevine (Vítis vinifera 24;49-54 Culture L.) cvs Chardonnay and Shiraz, lustralian Dokoozlian, N.K. and Kliewer, W.M. (1996) Journal ofGrape and L\/ine Research l; ll-18 grape berry growth and Influence oflight on Elliger, C.4., Chan, B.C. andWaiss, A.C. Jnr. varies during fruit development, composition (1980) Flavonoids as larval growth inhibitors. ofthe American Society Journal for Structural features governing toxicity, Horticultural Science l2l; 869 -87 4 Naturw is s ens chaften 67 ; 3 5 8-360 Dong, Y.-H., Beuning, L., Davies, K., Mitra, D., Elmqvist, T., Cates, R.G., Harper, J.K. and B. and Kootstra, A. (1998) Expression Morris, Gardfiell, H. (1991) Flowering in males and of pigmentation genes and photo-regulation of females of a Utah will ow , Salix rigida and anthocyanin biosynthesis in developing Royal effects on growth, tannins, phenolic apple flowers, luslralian Journal of Gala glycosides and sugars, Oikos 6l: 65-72 P I ant P hys iolo gy 25 ; 245 -252 Emmett, R.'W., Harris, 4.R., Taylor, R.H. and Reese, J.C. and Jones, K.C. (1981) Dreyer, D.L., McGechan, J.K. (1992) Grape diseases and Aphid feeding deterrents in sorghum, Journal vineyard protection In: Viticulture, Volume 2 of Chemical Ecology 7 ;273-284 Practices (Coombe &Dry, eds.) pp. 232-278 Dry, P.R., Loveys,8.R., Iland, P.G., Botting, Winetitles, Adelaide D.G., McCarthy, M.G. andStoll, M. (1998) Escribano-B ailon, T., Gutierrez-Fern andez, Y ., manipulation to meet fruit specif,lcations. Vine Rivas-Gonzalo, J.C. and Santos-Buelga, C. I:n: Proceedings of the Tenth Australian Wine (l 9 9 2) Characterization of procyanidins of In dus Technical Confer enc e pp. 208 -214, try Vitis vinifera variety Tinta del Pais grape Adelaide. seeds, Journal ofAgricultural and Food Düring, H. andDavtyan, A. (2002) Chemistry 40;1794-1799 Developmental changes of primary processes Escribano-Bailon, M.T., Guerra, M.T., Rivas- ofphotosyntheses in sun- and shade-adapted Gonzalo, J.C. and Santos-Buelga, C. (1995) Proanthocyanidins in skins from different 233 grape varietie s, Z e it s c hrift fur L eb e n s m i t t e I - Foo, L.Y. and Porter, L.J. (1983) Synthesis and Unt e r er suc hun g und- Fors chung 200; 221 -224 conformation of procyanidin diastereoisomers, Journal of Ezzili,B., Darne, G. and Bejaoui, M. (1999) the Chemical Society Perkin Transactions Technique for the production ofanthocyanins 1; 1535-1543 in leaves ofone-eyed cuttings ofCarignan Ford, C.M. andHøj, P.B. (1998) Multiple (Vitis vinifera L.) grown in the laboratory, glucosyltransferase activities in the grapevine Journal Internationaldes Sciences de la Vitis vinifera L.,Australian Journal of Grape Vigne et du Itin 33;9-18 and Wine Research 4; 48-58 F Ford, C.M., Boss, P.K. andHø1, P.B. (1998) Cloning and character ization of Vi t i s v inife r a Fauconneau, B., Waffo-Teguo, P., Huguet, F., UDP-glucose:fl avonoid 3- O- Barrier, L., Decendit, A. and Merillon, J-M. glucosyltransferase, a homologue (1997) Comparative study of radical of the scâvenser and antinxid¡nf nronerfies of enzyme encoded by the maize Bronze- I locus that may primarily serve to glucosylate phenolic compounds from Vitis vinifera cell anthocyanidin s i n viv o, Journ al of B i o I o gical cultures using in vitro tests, Lde Sciences 6l; 2103-10 Chemistry 273; 9224-33 Forkmann, G. (1991) Flavonoids as flower Feeny, P. (1976) Plant apparency and chemical pigments: The formation of the natural defence In: Biochemical Interaction between spectrum and its extension by genetic Plants and Insecls (Wallace & Mansell, eds.) engineering, Plant Breeding pp. 1-40. Plenum Press, New York 106; l-26 Forkmann, G. (1993) Genetics of flavonoids In: Fernández de Simón, 8., Pérez-llzarbe, J., The Flavonoids: Advances in Research Hernández, T., Gómez-Cordovés, C. and since 1986 (Harborne, ed.) pp. 537-564. Chapman Estrella, I. (1990) HPLC study of the & Hall, London efhciency of extraction of phenolic compounds, Chromatographia 30; 35-37 Forkmann, G., de Vlaming, P., Spribille, R., Wiering, H. anel Sch¡am, A.W. (1986) Genetic Firmin, J.L., Wilson, K.E., Rosen, L. an Fougère-Rifot, M., Benharbit El Alami, Flint, S.D., Jordan, P.W. and Caldwell, M.M. N., Brun, O. andBouard, J. (1995a) Gynoecium ( I 985) Plant protective response to enhanced ontogenesis of Vitis vinifera L. UV-B radiation under held conditions: Leaf var. Chardonnay in closed connection optical properties and photosynthesis, with the apparition of tannic vacuoles, Journal Photochemistry a-nd Photobiology 4l; 95-99 International des Sciences de laVigne et du Foo, L.Y. and Porter, L.J. (1980) The Vin 29; 105-130 phytochemistry of proanthocyanid in polymers, Phytochemistry 19 ; 17 47 -17 54 234 Fougère-Rifot, M., Delavaux, J.J., Benharbit El flavan-3-ols by liquid chromatography-ion Alami, N., Bouard, J. and Brun, O. (1993) spray mass spectrometry , Journal of Importance of vacuolar tannins in the Chr o mat o gr aphy A 7 52; 8 5 -9 I of ovaries and ovules in the development Fulcrand, H., Remy, S., Souquet, J-M., Cheynier, grapevine, 127 -13 6 Vitis 32:' V. andMoutounet, M. (1999) Study of wine Fougère-Rifot, M., Park, H.S., Cherrad, M., tannin oligomers by on-line liquid Darné, G. and Bouard, J. (1995b) Plastids, cbromatography electrospray ionization mass starch and chlorophyll iîVitis vinifera L. spectrometry, Journal of Agricultural and berry during its ontogenesis, Journal Food Chemistry 47 ; 1023-1028 International des Sciences de la Vigne et du Fuleki, T. and Ricardo da Silva, J.M. (1997) Vin29;213-225 Catechin and procyanidin composition of Fougère-Rifot, M., Park, H.-S., Benharbit el- seeds from grape cultivars grown in Ontario, Alami, N. and Bouard, J. (1996) Anomalies in Journal of Agricultural and Food Chemistry the structure of grapevine flowers and berries 45;1156-1160 to failure to set fruit and uneven in relation G berry development, Journal International des Galet, P. (L979) A Practical Ampelography: Scíence de la Vigne et du Vin Special Issue; Grap evine IdentiJication. Comstock 47-52 Publishing Associates, Ithaca, N.Y Fritzemeier, K-H. and Kindl, H. (1981) Gao, Y. and Cahoon, G.A. (1994) Cluster Coordinate induction by UV light of stilbene shading effects on fruit quality, skincolor, and synthase, phenylalanine ammonia-lyase and anthocyanin content and composition in cinnamate 4-hydroxylase in leaves of Reliance ( Zllrs hybrid) , Vitis 33;205-209 Vitaceae, P I ant a l5l ; 48-52 Gawel, R. (1998) Red wine astringency: A Frohman, M.4., Dush, M.K. and Martin, G.R. review, Australian Journal of Grape and LV'ine (1988) Rapid production of tull-lengthcDNAs Research 4;'14-95 from rare transcripts: amplif,rcation using a single gene-specific oligonucleotide primer, Gawel, R., Oberholster, A. and Francis, LL. Proceedings ofthe National Academy of (2000) A'Mouth-feel Wheel': Terminology Sciences of the United States of America 85; for communicating the mouth-feel 8998-9002 characteristics of red wine, Australian Journal of Grape and Wine Research 6;203-207 Fukada-Tanaka, S., Hoshino, 4., Hisatomi, Y., Habu, Y., Hasebe, M. and lida, S. (1997) Geissman, T.A. and Crout, D.H.G. (1969) Identiflrcation ofnew chalcone synthase genes Organic Chemistry of Secondary Plant for flower pigmentation in the Japanese and Metabolism. Freeman, Cooper & Co., San common morning glories, Plant and Cell Francisco Physiology 38;754-158 Geissman, T.A. andDittmar, H.F.K. (1965) A Fulcrand, H., Doco, T., Es-Saf,r, N-E., Cheynier, proanthocyanidin from avocado seed, V. andMoutounet, M. (1996) Study of P hyto chemis try 4 ; 3 59 -368 acetaldehyde induced polymerisation of 235 Gerats, A.G.M. and Martin, C. (1992) Flavonoid Guidoni, S., Allara, P. and Schubert, A. (2002) synthesis in Petunia hybrida; genetics and Effect ofcluster thinning on berry skin molecular biology of flower colour In: anthocyanin composition of Vitis vinifera cv. Phenolic Metabolism in Plants (Stafford & Nebbiolo, lz erican Journal of Enology and Ibrahim, eds.) pp. 165-199. Plenum Pres.s, Viti cu ltur e 53; 224-226 New York Guyot, S., le Guernevé, C., Marnet, N. and Glories, Y. (1988) Anthocyanins and tannins Drilleau, J-F. (1999) Methods for determining from wine: Organoleptic properties ln: Plant the degree of polymerisation of condensed IH-NMR Flavonoids in Biology and Medicine II: tannins: A new procedure applied to Biochemical, Cellular, and Medicinal cider apple procyanidins In: Plant Properties (Cody et al., eds.)pp. 123-134. Polyphenols 2. Chemistry, Biology, Alan R. Liss, Inc., New York Pharmacology, Ecology (Gross, et al., eds.) pp. 21 1 -222. Kluwer Academic/Plenum Gnanamanickam, S.S. and Smith, D.A. (1980) Publishers, New York Selective toxicity of isofl avonoid phytoalexins to gram-positive bacteria, Phytopathology 7 0 ; H 894-896 Hagerman, A.E. and Butler, L.G. (1980) Goldberg, D.M., Tsang, E., Karumanchiri, A. Determination of protein in tannin-protein and Soleas, G.J. (1998) Quercetin and p- precipitates, -Iournal of Agricultural and Food coumaric acid concentrations in commercial chemistry 28;944-947 wines, American Journal of Enology and Hagerman, A.E. and Butler, L.G. (1994) Assay Viticulture 49; 1 42-1 5 | of condensed tannins or flavonoid oligomers Goldstein, J.L. and Swain, T. (1963) Changes in and related flavonoids in plants, Methods in tannins in ripening fruits, Phytochemistry 2; Enzymology 234;429-37 37t-383 Harbertson, J.F., Kennedy, J.A. and Adams, D.O Gong,Z.,Yamazaki, M., Sugiyama, M., Tanaka, (2002) Tannin in skins and seeds of Cabemet Y. and Saito, K. (1997) Cloning and Sauvignon, Syrah, anel Pinot noir during molecular analysis of structural genes npening, American Journal of Enology and involved in anthocyanin biosynthesis and Viticulture 53; 54-59 expressed in a forma-specific manner in Harborne, J.B. (1958) Chromatographic Perilla frulescens, Plant Molecular Biology identifi cation of anthocyanin pigments, 35;915-927 It Chromatograp hic Rev iew l; 209-224 Gonzalez-San Jose, M.L., Santa-Maria, G. and Harborne, J.B. ( I 967) C o mp arativ e B io c he mis try Diez, C. ( 1990) Anthocyanins as parameters of the Flavonolds. Academic Press, London for differentiating wines by grape variety, Harborne, (1998) wine-growing region, and wine-making J.B. Phytochemical Methods - A guide to modern methods, Journal of Food Composition and techniques of plant Analysis 3;54-66 analysis. (3rd Edition) Chapman and Hall, London 236 Harbome, J.B. and Grayer, R.J. (1993) cv. Shiraz grape berries, AustralianJournal of Flavonoids and Insects In'. The Flavonoids: Grape and lline Research 6; l4l-149 Advances in Research since 1986 (Harborne, Haslam, E. (1980) In vino veritas: Oligomeric pp. 589-618. Chapman & Hall, London ed.) procyanidins and the ageing of red wines, Harrison, M.J. and Dixon, R.A. (1994) Spatial P hy to chemis try 191, 257 7 -2582 patterns of expression of Haslam, E. (1989) Plant Polyphenols: Vegetable flavonoid/isoflavonoid pathway genes during Tannins Revis ited. Cambtidge University interactions between roots of Medicago Press, Cambridge truncatula and the mycorrhizal fungus Haslam, E. (1998) Practical Polyphenolics: Glomus versiþrme, Plant Journal 6;9-20 From Structure to Molecular Recognition and R.G. (1980) Harrison, B.J. and Stickland, Physiological Action. Cambridge University Precursors and the genetic control of Press, Cambridge pigmentation V. Initiation of anthocyanin Hawker, J.S. (1982) Effect of temperature on synthesis in Antirrhinum maius by B otrytis lipid, starch and enzymes of starch cinerea, Heredity 44; 103- I 09 metabolism in grape, tomato and broad bean Hartmann, U., Valentine, W.J., Christie, J.M., leaves, P hyto chemis try 2l; 33 -3 6 Hays, J., Jenkins, G.I. andWeisshaar, B. Helariutta, Y., Elomaa, P., Kotilainen, M., (1998) Identification of UV/blue light- Seppanen, P. and Teeri, T.H. (1993) Cloning response elements in the Arabidopsis thaliana of cDNA coding for dihydroflavonol-4- chalcone synthase promoter using a reductase (DFR) and characterizationof dfr homologous protoplast transient expression expression in the corollas of Gerbera hybrida system, Plant Molecular Biology 36;741-754 var. Regina (Compositae), Plant Molecular and Robinson, S.P. Harvey, J.S., Downey, M.O. Biology 22;183-193 (in prep.) Expression of Yitis vinifera Heller, Vy'. andForkmann, G. (1993) BANYULS modulates anthocyanin and tannin Biosynthesis of flavonoids ln: The accumulation in tobacco (Tobaccum Flavonoids: Advances in Research since 1986 nicotiana), P lant P hy sio lo gy (Harborne, ed.) pp. 499-535. Chapman & hall, (1991) The effect of manipulated Haselgrove, L. London bunch exposure on the accumulation of Hemingway, R.W. and McGraw, G.V/. (1983) quercetin glycosides and anthocyanins in Kinetics of acid-catalysed cleavage of Vitis vinifera L. cv Shiraz, Masters Thesis, procyanidins, Journal of Wood Chemistry and Department of Horticulture, Viticulture and Technologt 3;421-435 Oenology, University of Adelaide, Adelaide Hemingway, R.W., Foo, L.Y. and Porter, L.J. Haselgrove, L., Botting, D., van Heeswijck, R., (1982) Linkage isomerism in trimeric and Høj, P.8., Dry, P.R., Ford, C. and lland, P.G. polymeric 2,3 < i s -pr o cy anidins, J o ur n a I of t h e (2000) Canopy microclimate and berry Chemical Society Perkin Transactions 1; composition: The effect ofbunch exposure on 1209-t216 the phenolic composition of Vitis vinifera L 231 Herrmam, K.M. (1995) The Shikimate pathway: Hoshino, T., Matsumoto, U. and Goto, T. (1980) Early steps in the biosynthesis of aromatic The stabilizing effect ofthe acyl group on the compounds, The Plant Cell 7 ; 907 -919 co-pigmentation of acylated anthocyanins with c-glucosylflavanols, Hertog, M.G.L. andHollman, P.C.H. (1996) Phytochemistry 19; 663-667 Potential health effects ofthe dietary flavonol quercetin, European Journal of Clinical Hoshino, T., Matsumoto, U., Goto, T. and Nutrition 50;63-71 Harada, N. (1982) Evidence for the self- association of anthocyanins IV. PMR Hipskind, J., Vy'ood, K. and Nicholson, R.L. spectroscopic evidence for (1996) Localised stimulation of anthocyanin the vertical stacking of anthocyanin molecules, accumulation and delineation of pathogen Te trahe d ron Le tt ers. 23 ingress in maize genetically resistant to ; 433 -43 6 Bipolaris maydis race O, Physiological and Houbiers, C., Lima, J.C., Macanita, A.L. and Plant Molecular Pathology 49; 247 -256 Santos, H. (1998) Colour stabilisation of malvidin 3-glucoside: Hmamouchi, M., Es-Safi, N., Lahrichi, M., self aggregation of the flavylium cation and copigmentation Fruchier, A. and Essassi, E.M. (1996) with the Z-chalcone form, Journal of Physical Flavones and flavonols in leaves of some Chemistry B 102; 3578-3585 Moroccan Vítis vinifera cultivars, American Journal ofEnology and Viticulture 47;186- Hrazdina, G. andBorzell, A.J. (1971) 192 Xanthylium derivatives in grape extracts, P hy t o c h e mi s t ry l0; 221 I -2213 Holton, T.A. and Cornish, E.C. (1995) Genetics and biochemistry of anthocyanin biosynthesis, Hrazdina, G. andMoskowitz, A.H. (1980) Plant Cell T; l07l-1083 Subcellular status of anthocyanins in grape skins In: Proceedings of the University Holton, T.4., Brugliera, F., Lester, D.R., Tanaka, of California, Davis, Grape Y., Hyland, C.D., Menting, J.G., Lu, C.Y., and Wine C en t enni a I Symp o s ium 245 -253. University Farcy, E., Stevenson, T.W. and Cornish, E.C. of California Press, Berkeley (1993a) Cloning and expression of cytochrome P450 genes controlling flower Hrazdina, G. and Parsons, G.F. (1982) Induction colour, Nature 366; 276-9 of flavonoid synthesising enzymes by light in etiolated pea (Pisum sativum cv. Holton, T.4., Brugliera, F. and Tanaka, Y. Midfreezer) seedlings, Plunt Phys iology 7 0; (1993b) Cloning and expression offlavonol 506-510 synthase from Petunia hybrida, Plant Journal Hunter, J.J., Ruffner, H.P., Volschenk, C.G. and 4;1003-1010 le Roux, D.J. (1995) Partial defoliation of Vitis vinifera L. cv. Cabernet Sauvignon/99 Horowitz, R.M. (1964) Relations between taste Richter: Effect on root growth, and structure of some phenolic glycosides In: canopy efficiency, grape composition, and wine B io chemistry of P heno lí c Compou nds quality, Amerícan Journal of Enology (Harborne, ed.) pp. 545-571. Academic Press, and London Viticulture 46; 306-3 14 238 Hyams, E. (1965) Dionysus. A Social History of physiology. Cambridge University Press, the lYine Vine. Sedgwick & Jackson Ltd., Cambridge London Jordão, 4.M., Ricardo-da-Silva, J.M. and I Laureano, O. (2001a) Evolution ofcatechins and oligomeric procyanidins during grape Ishikura, N. andMato, M. (1993) Partial maturation Castelão Francês and Touriga purif,rcation and some properties of flavonol Francesa, American Journal of Enology and 3-O-glycosyltransferases from seedlings of Viticultur e 52; 230 -23 4 Tigna mungo, with special reference to the formation of kaempferol 3-O-galactoside and Jordão,4.M., Ricardo da Silva, J.M. and 3-O-glucoside, Plant and Cell Physiology 34; Laureano, O. (2001b) Evolution of 329-335 proanthocyanidins in bunch stems during berry developmerft (Vitis viniferal L.), Vitis Ishikura, N. and Yamamoto, K. (1990) Detection 40; l7-22 and characterizatiort of UDP- glucose: fl avonoid O-glucosyltransferases in Joslyn, M.A. and Mackinney, G. (1938) The rate the leaves of Prunus xyedoensis Matsum' of conversion of chlorophyll to pheophytin, Plant and Cell Physiologl 31; 1 109-l I l5 Journal of the American Chemical Society 60; 1132-1136 Iland, P.G., Ewart, A. and Sitters, J. (1993) Techniques for Chemical Analysis and Joseph, R., Tanner, G. andLarkin, P. (1998) Stability Tests of Grape Juice and Wine. Proanthocyanidin synthesis in the forage Patrick Iland Promotions, Adelaide legume Onobrychis viciþlia. A study of chalcone synthase, dihydroflavonol 4- J reductase and leucoanthocyanidin 4-reductase P.B. (1993) Jackson, D.I. and Lombard, in developing leaves, Australian Journal of and management practices Environment Plant Physiology 251' 27 l-278 affecting grape composition and wine quality Jurd, L. (1972) Anthocyanins and related - Areview,American Journal of Enology and compounds XVI. The dimerisation of Viti cultur e 4 4; 409 - 43 0 flavylium salts in aqueous solutions., M.B. (1988) Climate Jackson, D.I. and Spurling, Tetrahedron. 28; 493 -502 and viticulture in Australia ln: Viticulture. Jurd, L. and Somers, T.C. (1970) The formation Volume l. Resources (Coombe & Dry, eds.) of xanthylium salts from proanthocyanidins, pp. 91-106. Vy'inetitles, Adelaide P hy to chemistry 9 ; 419 -427 Jeandet, P., Bessis, R., Maume, 8.F., Meunier, K P., Peyron, D. andTrollat, P. (1995) Effect of enological practices on the resveratrol isomer Kallithraka, S., Garcia-Viguera, C., Bridle, P. content of wine, Journal of Agricultural and and Bakker, J. (1995) Survey ofsolvents for Food Chemistry 43; 316-319 the extraction of grape seed phenolics, P hy t o ch e mic al Analy sis 6; 265 -267 Jones, H.G. (1992) Plants and Microclimate. A quantitative approach to environmental plant 239 Kantz,K. and Singleton, V.L. (1990) Isolation polyphenols during fruit ripening, and determination of polymeric polyphenols P hytochemistry 551' 7 7 -85 using Sephadex LH-20 and analysis ofgrape Kennedy, J.4., Troup, G.J., Pilbrow, J.R., tissue extracts , American Journal of Enology Hutton, D.R., Hewitt, D., Hunter, C.R., Ristic, a nd V i i ictil tur e 41 ; 223 -228 R., Iland, P.G. and Jones, G.P. (2000b) Katalinic, V. and Males, P. (1997) Development of seed polyphenols in berries Compositional changes in grape polyphenols from Vitis vinifera L. cv. Shiraz, Australian throughout maturation, Journal of Wine Journal of Grape and Wine Research 6;244- Research 8;169-177 254 Katz, A. and Weiss, D. (1999) Light regulation Khajuria, H.N. and Bakhshi, J.C. (1988) Effect of anthocyanin accumulation and chalcone of different growth regulating compounds and synthase gene expression in petunia flowers, girdling on shot berry formation in Perlette Israel Journal of Plønt Sciences 47 ;225-229 (Vitis vinifera L.) grapes, Indian Journal of A gricultur al Res earch 22 : I 50 -l 5 4 Keller, M., Arninþ K.J. and Hrazdina, G. (1998) Interaction of nitrogen availability during Kliewer, W.M. (1970) Effect of day temperature bloom and light intensity during veraison. I. and light intensity on coloration of Vitis Effects on grapevine growth, fruit vinifera L. grapes, Journal of the American development and ripening, I merican Journal Society for Horticultural Science 95; 693-697 of Enolo gy and Viticulture 49; 333 -340 Kliewer, W.M. (1977) Influence of temperature, Keller, M. andHrazdina, G. (1998) Interaction of solar radiation and nitrogen on coloration and nitrogen availability during bloom and light composition of Emperor grapes, American intensity during veraison. II. Effects on Journal ofEnology and Viticulture 28;96-103 anthocyanins and phenolic development Kliewer, V/.M. andAntcliff, A.J. (1970) during grape ripening, American Journal of Influence of defoliation, leaf darkening and Eno lo gy and Viticulture 49 ; 341-349 cluster shading on the growth and Kennedy, J.A. and Jones, G.P. (2001) Analysis composition of Sultana grapes, American of proanthocyanidin cleavage products Journal of Enology and Viticulture 2l;26-36 following acid-catalysis in the presence of Kliewer, Vy'.M., Lider, L.A. and Schultz, H.B. excess phloroglucinol, Journal of Agricultural (1967) Influence ofartificial shading of and Food Chemistry 49;1740-1746 vineyards on the concentration ofsugar and Kennedy, J.4., Hayasaka, Y., Vidal, S., Waters, organic acids in grapes, American Journal of E.J. andJones, G.P. (2001) Composition of Enology and Viticulture 18; 78-86 grape skin proanthocyanidins at different Koes, R.E., Quattrocchio, F. andMol, J.N.M. stages ofberry development, Journal of (1994) The flavonoid biosynthetic pathway in Agricultural and Food Chemistry 49;5348- plants: Function and evolutton, BioEssays 16; 5355 123-132 Kennedy, J.4., Matthews, M.A. and Waterhouse, A.L. (2000a) Changes in grape seed 240 Kristiansen, K.N. (1986) Conversion of (+¡- Lazarus, S.A., Adamson, 9.E., Hammerstone, dihydroquercetin to (+)-2,3- trans-3,4-cis- J.F. and Schmitz, H.H. (1999) High- leucocyanidin and (+)-catechin with an performance liquid chromatography/mass enzyme extract from maturing grains of spectroscopy analysis of proanthocyanidins in barley, C arl s b e r g Re s e arch C ommuni c ations foods and beverages, Journal of Agricultural 51; 5l-60 and Food Chemistry 47 ;3693-3701 L Lea, A.G.H., Bridle, P., Timberlake, C.F. and Singleton, V.L. (1979) The procyanidins of Lambert, S.G. (2002) Copigmentation and its white grapes and wines, American Journal of impact on the stabilisation of red wine Enolo gy and Viticulture 30 ; 289-300 pigments, PhD Thesis; Department of Horticulture, Viticulture and Oenology, Lee, C.Y. andJaworski, A. (1989) Major Adelaide University, Adelaide phenolic compounds in ripening white grapes, American Journal of Enology and Viticulture Lancaster, J.E. (1992) Regulation of skincolor in 40;43-46 apples, Critical Reviews in Plant Sciences l0; 487-502 Lees, G.L., Hinks, C.F. and Suttill, N.H. (1994) Effect ofhigh temperature on condensed Langcake, P. and McCarthy, W.V. (1979) The tannin accumulation in leaf tissues of Big relationship of resveratrol production to Trefoil (Lotus uli ginosøs Schkur), -Io urnal of infection of grapevine leavesby Botrytis the Science ofFood and Agriculture 65;415- ciner ea, Viti s l8 :, 244-253 421 Langcake, P. andPryce, R.J. (1976) The Leng, P., Itamura, H., Yamamura, H. and Deng, production ofresveratrol by Vitis vinifera and X.M. (2000) Anthocyanin accumulation in other members of the Vitaceae as a response apple and peach shoots during cold to infection or injury, Physiological Plant acclimation, Scientia Horticulturae 83 43-50 Pathology 9;77-86 ; Leone, 4.M., la Notte, E. and Gambacorta, G. Langcake, P. andPryce, R.J. (1977a) A new (1984) Gli antociani nelle fasi di macerazione class of phytoalexins from grapevines, e di elaborazione del vino. L'influenza della Experientia 33 ; l5l-152 tecnica diffusiva sulla loro estrazione, Langcake, P. andPryce, R.J. (1977b) The Vignevini 4; 17-31 production ofresveratrol and the viniferins by Leyva,4., Jarillo, J.4., Salinas, J. andMartinez- grapevines in response to ultraviolet Zapater, J.M. (1995) Low temperature induces irradiation, P hytochemistry 16; I 1 93- 1 1 96 the accumulation of phenylalanine ammonia- Lawrence, W.J.C. (1950) l. Genetic control of lyase and chalcone synthase mRNAs of biochemical synthesis as exemplified by plant Arabidopsis thaliana in a light-dependent genetics - flower colour,Biochemicøl Society manner, P lant P hys io I o gy 108; 39 - 46 Symposium No. 4 Biochemical Aspects of Li, Y-G., Tanner, G. and Larkin, P. (1996) The Genetics 4;3-9 DMACA-HCI protocol and the threshold proanthocyanidin content for bloat safety in 241 forage legumes,Journal ofthe Scíence of Markhar4 K.R. (1982) Techniques of Flavonoid Food and Agriculture 70; 89-101 Identification Academic Press, London Liew, C-F., Goh, C-J., Loh, C-S. and Lim, S-H. Markham, K.R. (1989) Flavones, flavonols and (1995) Cloning and nucleotide sequence ofa their glycosides In: Methods in Plant oDNA (Accession No. X89199) encoding Biochemistry Vol. l. (Harborne, ed.)pp 197- flavanone 3-hydroxylase from Bromheadia 235, Academic Press, London lays oni an a (Lindl.) Pichb.f ., P I ant fin Markham, K.R., Tanner, G.J., Caasi-Lit, M., Physiology 109; 339 'Whitecross, M.I., Nayudu, M. and Mitchell, Lo, S-C.C. and Nicholson, R.L. (1998) K.A. (1998) Possible protective role for 3',4'- Reduction of light-induced anthocyanin dihydroxyflavones induced by enhanced UV- accumulation in inoculated sorghum B in a UV-tolerant rice cultivar, mesocotyls. Implications for a compensatory Phytochemistry 49; 1913-1919 role in the defense response, Plant Physíology Marsehner, H. (1995) Míneral Nutrition of 116;919-989 Higher Plants. Academic Press, London Lodish, H., Baltimore, D., Berk, 4., Zipursky, Martin, C. and Gerats, T. (1993) The control of S.L., Matsudaira, P. andDamell, J. (1995) flower colorationln: The Molecular Biology Molecular Cell Biology. W. H. Freeman & of Flowering (Jordan, ed.) pp. 219-255. Co., New York C.A.B. Intemational, Wallingford Lu, Y. and Bennick, A. (1998) Interaction of Mateus, N., de Pascual-Teresa, S., Rivas- tannin with human salivary proline-rich Gonzalo, J.C., Santos-Buelga, C. and de proteins, Archives ofOral Biology 43;717-28 Freitas, Y. (2002a) Structural diversity of Lukacin, R., Wellman, F., Britsch, L., Martens, anthocyanin-derived pigments in port wines, S. andMatern, U. (2003) Flavonol synthase Food Chemisty 7 6; 335-342 from Citrus unshiu is a bifunctional Mateus, N., Marques, S., Goncalves,4.C., -292 dioxygenase, Phytochemistry 62; 287 Machado, J.M. and de Freitas, V. (2001) M Proanthocyanidin composition of red Vitis vinifera varieties from the Douro Valley Macheix, J.J., Fleuriet, A. andBillot, J. (1990) during ripening: Influence of cultivation Fruit Phenolics. CRC Press Inc., Boca Raton, altiÍatde, American Journal of Enology and Florida l¡iticulture 52l' ll 5-l2l Malien-Aubert, C., Dangles, O. andAmiot, M.J. Mateus, N., Silva, A.M.S., Santos-Buelga, C., (2002) Influence ofprocyanidins on the Rivas-Gonzalo, J.C. and de Freitas, V. colour stability ofoenin solutions, Journal of (2002b) Identifi cation of anthocyanin-fl avanol Agricultural and Food Chemistry 50l'3299- pigments in red wines by NMR and mass 3305 spectrometry, Journal of Agricultural and Markham, K.R. (1975) Isolation techniques for Food Chemistry 50; 2ll0-2116 flavonoids In: The Flavonoids (Harborne e/ Matsushima, J., Hiratsuka, S., Taniguchi, N., a/. eds) pp l-44, Chapman and Hall, London 'Wada, R. and Suzaki, N. (1989) Anthocyanin 242 accumulation and sugar content in the skin of induction by fungal elicitor, wounding and grape cultivar' Olympia' treated with ABA, infection, EMBO Journal 6; 1527 -1533 Journal ofthe Japanese Societyfor Meyer, B.J. and Hernandez, R. (1970) Seed Hoticultural Science 58; 55 I -555 tannin extraction in Cabernet Sauvignon, Matthews, S., Isabelle, M., Scalbert, A. and Arherican Journal ofEnology and Viticulture Donnelly, D.M.X. (1997) Extractable and 21;184-188 non-extractable proanthocyanidins in barks, Miller, K.D., Guyon, V., Evans, J.N.S., Phytochemistry 45; 405-410 Shuttleworth, W.A. andTaylor, L.P. (1999) Mazza, G. (1995) Anthocyanins in grapes and Purifrcation, cloning and heterologous grape products, C.rRC Critical Reviews in Food expression of a catalytically efflrcient flavonol Science and Nutrition 35:, 341-37 I 3-O-glactosyltransferase expressed in the male gametophyte of Petunia hybrida, Mazza, G. and Miniati, E. (1993) Grapes In: Journal of Biological Chemistry 27 4; 34011- Anthocyanins in Fruits, Vegetables and 34019 Grains(Mazza & Miniati, eds.) pp. 149-199 CRC Press, Inc., Boca Raton, Florida Miller, K.D., Strommer, J. andTaylor, L.P. (2002) Conservative in deivergent McCarthy, M.G. (1999) Weight loss from solanaceous species ofthe unique gene ripening berries of Shiraz grapevines ( Zi /ls structure and enzyme activity of a vinifera L. cv. Shiraz), Australian Journal of gametophytically-expressed fl avonol 3- O- Grape and Wíne Research 5; 10-16 galactosyltransferase, P I an t Mo I ecul ar McCarthy, M.G. and Coombe, B.G. (1999) Is Biology 48;233-242 weight loss in ripening grape berries cv. Moalem-Beno, D., Tamari, G., Leitner-Dagan, Shiraz caused by impeded phloem transport? Y., Borochov, A. and Weiss, D. (1997) Sugar- Australian Journal of Grape & l|/ine dependent gibberellin-induced chalcone Research 5; 17-21 synthase gene expression in petunia corollas, McCloskey, L.P. and Yengoyan, L.S. (1981) P lant P hy s io lo gy ll3 ; 419 -424 Analysis of anthocyanins inVitis vinifera Mole, S. (1989) Polyphenolics and the nutritional wines and red colour versus aging by HPLC ecology ofherbivores ln: Toxicants ofPlant and spectrometry, American Journal of Origin. Volume IV. Phenolics (Cheeke, ed.) Enolo gy and Viticulture 321, 257 -261 pp. l9l-223. CRC Press, Inc., Boca Raton, McDonald, M.S., Hughes, M., Burns, J., Lean, Florida M.E.J., Matthews, D. and Crozier, A. (1998) Mole, S. and Waterman, P.G. (1988) Light- Survey ofthe free and conjugated myricetin induced variation in phenolic levels in foliage and quercetin content ofred wines ofdifferent ofrain forest plants II. Potential and geographical oigins, Journal of A gricultural significance to herbivores, Journal of and Food Chemistry 46;368- 375 Chemic al E c o I o gy 14 ; 23 -3 4 Mehdy, M.C. and Lamb, C.J. (1987) Chalcone Mole, S., Ross, J.A.M. and'Waterman, P.G. isomerase cDNA cloning andmRNA (1988) Lightinduced variation in phenolic 243 levels in foliage ofrain forest plants L N Chemical changes, Journal of Chemical Nadal, M. andArola, L. (1995) Effects of limited Ecology 14; l-21 irrigation on the composition of must and Moriguchi, T., Kita, M., Ogawa, K., Tomono, wine of Cabernet Sauvignon under semi-arid ' Y., Endo, T. and Omura, M. (2002) Flavonol conditiohs, I/itis 34; I 5 l-l 54 synthase gene expression during citrus fruit Nagabhushan, M., Amonkar,4.J., Nair, U.J., development, Physiologia Plantarum ll4; Santhanam, U., Ammigan, N., D'Souza, A.V. 251-258 and Bhide, S.V. (1988) Catechin as an Morris, P. and Robbins, M.P. (1997) antimutagen: Its mode of action, Journal of Manipulating condensed tannins in forage Cancer Research and Clinical Oncology ll4; legumes In: Biotechnology and the 177-182 Improvement of Forage Legumes (McKersie Nagel, C.W. (1985) Application of high & Brown, eds.) pp. 147-173. CAB performance liquid chromatography to International, New York analysis of flavonoids and phenyl propenoids, Morrison, J.C. and Noble, A.C. (1990) The Cereal Chemistry 621, 144-147 effects ofleafand cluster shading on the Nagel, C.W. andWulf, L.V/. (1979) Changes in composition of Cabernet Sauvignon grapes the anthocyanins, flavonoids and and on fruit and wine sensory properties, hydroxycinnamic acid esters during American Journal of Enology and Viticulture fermentation and aging of Merlot and 4l;193-200 Cabemet Sauvignon, American Journal of Moskowitz, A.H. andHrazdina, G. (1981) Enology and Viticulture 30; 1l l-l l6 Vacuolar contents of fruit sub-epidermal cells Nef, L. (1988) Interactions between the leaf from Vitis species, Plant Physiology 68; 686- mirer, Phyllocnistis suffisella, and poplars In: 692 Mechanisms of Woody Plant Defenses Against Moutounet, M., Rigaud, J., Souquet, J.M. and Insects. Searchfor Pattern (Mattson et al., Cheynier, V. (1996) Structural eds.) pp. 239-251. Springer-Verlag, New York Characterisation of grape berry tannins. Some Negrul, A.M. (1936) Variabilitat und verebung examples of the effect of vine variety, des geschlechts bei der rebe, Gartenbauwiss "terroir" and vine training, Bulletion de l01'215-231 l' O. I. l/. 7 83 -7 84 ; 433 - 443 Nielsen, K.M. andPodivinsky, E. (1997) cDNA Mueller, W.C. and Greenwood, A.D. (1978) The Cloning and endogenous expression ofa ultrastructure of phenolic-storing cells hxed flavonoid 3'5'-hydroxylase from petals of with caffiene, Journal of Experimental Botany lisianthus (Eustoma grandifl orum), P lant 29:'757-764 Science 129;167-174 Mullins, M.G., Bouquet, A. and Williams, L.E. Noble, A.C. (1994) Bitterness in wine, (1992) Biology of the Grapevine. Cambridge P hysiolo gy and B ehaviour 56; l25l -125 5 University Press, Cambridge 244 Northcote, K.H. (1992) Soils and Australian P viticulture, I.n: Titiculture, Volume I Pace-Asciak, C.R., Hahn, S., Diamandis, E.O., Resources (Coombe & Dry, eds.) pp. 6l-90. Soleas, G. and Goldberg, D.M. (1995) The red Winetitles, Adelaide wine phenolics /rdlrs-resveratrol and quercetin Nunan, K.J., Sims, I.M., Bacic, 4., Robinson, block human platelet aggregation and S.P. andFincher, G.B. (1998) Changes in cell eicosanoid synthesis: Implications for wall composition during ripening of grape protection against coronary heart disease, berries, P I ant P hysio lo gy ll8; 7 83 -7 92 Clinica Chimica Acta 235; 207 -219 o Pandey, R.M. and Farmahan, H.L. (197 7) Changes in the rate ofphotosynthesis and Ong, B.Y. andNagel, C.W. (1978a) respiration in leaves and berries of Vitis Hydroxycinnamic acid-tartaric acid ester vinifera grapevines at various stages ofberry content in mature grapes and during the development, I/itis 16;106-l I I maturation of white Riesling grapes, American Journal of Enology and Viticulture Patakas, 4., Noitsakis, B. and Stawakas, D. 29;277-281 (1997) Water relation parameters in Vitis vinifera L. in drought period. Effects ofleaf Ong, B.Y. andNagel, C.V/. (1978b) High- age, Agronomie 17; 129-138 pressure liquid chromatographic analysis of hydroxycinnamic acid-tartaric acid esters and Pekic,El., Kovac, V., Alonso, E. and Revilla, E. their glucose esters in Titis vinifera, Journal (1998) Study ofthe extraction of of C hrom ato gr ap hy 157 ; 345 -3 55 proanthocyanidins from grape seeds, F ood Chemistry 16;201-206 Osawa, Y. (1982) Copigmentation of anthocyanins ln: Anthocyanins as Food Pecket, R.C. and Small, C.J. (1980) Occurrence, Colours (Markakis, ed.) pp. 4l-68. Academic location and development of Press, New York anthocyanopl asts, P hyto chemis try 19 ; 257 l- 2576 Osmond, R. and Anderson, K. (1998) Trends and Cycles in the Australian lline Industry, 1850 Pelletier, M.K. and Shirley,8.1V. (1996) to 2000. Centre for Intemational Economic Analysis of flavanone 3-hydroxylase in Studies, University of Adelaide, Adelaide Arabidopsis seedlings. Coordinate regulation with chalcone synthase and chalcone Oszmianski, J. and Sapis, J.C. (1989) isomerase, P/ant Physiology lll; 339-345 Fractionation and identif,rcation of some low molecular weight grape seed phenolics, Pelletier, M.K., Murrell, J.R. and Shirley, B.W. Journal of Agricultural and Food Chemistry (1997) Characterizalion of flavonol synthase 37;1293-1297 and leucoanthocyanidin dioxygenase genes in Arabidopsis. Further evidence for differential Ough, C.S. (1992) lVinemaking Basics.Food regulation of 'early' and'late' genes, Plant Products Press, New York Physiolo gy ll3; 1437 -1445 Peng,Z.,Hayasaka, Y., Iland, P.G., Sefton, m., Hoj, P. and Vy'aters, E.J. (2001) Quantitative 245 analysis ofprocyanidins (tannins) from grape quercetin in Pinot noir grapes and wine, (Vitis vinifera) seeds by reverse phase high- American Journal of Enology and Viticulture performance liquid chromatography, Journal 46;187-194 of Agricultural and Food Chemisty 49;26-31 Prieur, C., Rigaud, J., Cheynier, V. and Peterson, J.R. and Smart, R.E. (1975) Foliage Moutounet, M. (1994) Oligomeric and removal effects on'Shiraz' grapevines, polymeric procyanidins from grape seeds, American Journal of Enology and Viticulture P hytochemistry 36; 7 8l -7 84 26; l19-124 R Pirie, A. (1977) Phenolics accumulation in red Rankine, B. (1998) Making Good Wine. A wine grapes (Vitis vinifera L.), PhD Thesis; Manual for lí/inemaking Practice in Australia Department of Agronomy and Horticultural and New Zealand. PanMcMillan Publishers, Science, University of Sydney, Sydney Sydney Pirie, A. and Mullins, M.G. (1976) Changes in Raven, P.H., Evert, R.F. and Eichhom, S.E. anthocyanin and phenolics content of (1992) Biology of Plants. Worth Publishers, grapevine leaf and fruit tissues treated with New York sucrose, nitrate, and abscisic acid, Plant Reay, P.F. and Lancaster, J.E. (2001) Physiology 58;468-472 Accumulation of anthocyanins and quercetin Pirie, A.J.G. and Mullins, M.G. (1980) glycosides in'Gala' and'Royal Gala' applc Concentration of phenolics in the skin of fruit skin with UV-B-visible irradiation: grape berries during fruit development and Modiffing effects of fruit maturity, fruit side, ipen:ng,American Journal of Enology and and temperature, Scientia Horticulturae 90; l¡iticulture 3l;34-36 57-68 Plumb, G.W., de Pascual-Teresa, S., Santos- Recourt, K., van Tunen, 4.J., Mur, L.4., van Buelga, C., Cheynier, V. and Williamson, G. Brussel, A.A.N., Lugtenberg, B.J.J. andKijne, ( I 998) Antioxidant properties of catechins and J.W. (1992) Activation of flavonoid proanthocyanidins : Effect of polymerisation, biosynthesis in roots of Vicia søllva subsp. galloylation and glycosylatíon, Free Radical nigra plants by inoculation withRhizobium Research 29;351- 358 le guminos arun biovar v i c i ae, P lant Pool, R.M., Creasy, L.L. andFrackelton, A.S. Mo I e cul ar B io lo gy 19 ; 4l I -420 (1981) Resveratrol and the viniferins, their Renaud, S. and de Lorgeril, M. (1992) Wine, application to screening for disease resistance alcohol, platelets, and the French paradox for in grape breeding programs, Vitis 20;136-145 coronary heart disease, The Lancet 339; 1523- Porter, L.J. (1988) Flavans and 1526 Proanthocyanidins In: The Flavonoids : Revilla, E., Alonso, E. and Kovac, V. (1997) The Advances in Research since 1980 (Harbome, content ofcatechins and procyanidins in ed.) pp. 2l-62. Chapman & Hall, London grapes and wines as affected by Price, S.F., Breen, P.J., Valladao, M. and agroecological factors and technological Watson, B.T. (1995) Cluster sun exposure and practices ln: Wine: Nutritional and 246 Therapeutic BeneJits (Watkins, ed.) pp. 69-80 Rojas-Lara, B.A. and Morrison, J.C. (1989) American Chemical Society, Washington Differential effects of shading fruit or foliage D.C. on the development and composition of grape berries, Vitis 28; 199 -208 Revilla, L, Perez-Magarino, S., Gonzalez-San- Jose, M.L. andBeltran, S. (1999) Romeyer, F.M, Machie¿ J-J. and Sapis, J-C. Identihcation of anthocyanin derivatives in (1986) Changes and importance of oligomeric grape skin extracts and red wines by liquid procyanidins during maturation of grape chromatography with diode array and mass seeds, P hytochemis try 25 ; 219 -221 spectrometric detection, Journal of Romeyer, F.M, Sapis, J-C. and Macheix, J-J. Chromato graphy A 847 ; 83 -90 (1985) Hydroxycinnamic esters and browning Reynolds,4.G., Pool, R.M. andMattick, L.R. potential in mature berries of some grape (1986) Influence ofcluster exposure on fruit varieties,Journal ofthe Science ofFood and composition and wine quality of Seyval blanc Agriculture 36; 7 28-7 32 grapes, Vitis 25;85-95 Rosati, C., Cadic,A., Duron, M., Inguoff, M. and Ribereau-Gayon, P . (1964) Les composés Simoneau, P. (1999) Molecular phénoliques du raisin et duvin, Institute characterisation of the anthocyanin synthase National de la Recherche Agronomique, Paris gene inForsythia x intermedia reveals organ- specific expression during flower Ribereau-Gayon, P. (1974) The chemistry of red development, Plant Science 149; 73-79 wine colour ln: Chemistry of Winemaking @insmoor V/ebb, ed.). pp. 50-87. American Russell, E.tW. (1961) Soil Conditions and Plant Chemical Society, Washington D.C. Growth. Longmans, Green & Co. Ltd., London Ribereau-Gayon, P. and Glories, Y. (1986) Phenolics in grapes and wines. In: Ryan, K.G., Markham, K.R., Bloor, S.J., Proceedings of the 6th Australian Wine Bradley, J.M., Mitchell, K.A. and Jordan, Indus try Technical Co nference.'Winetitles, B.R. (1998) UVB Radiation induced increase Adelaide in quercetin:kaempferol ratio in wild-type and transgenic lines of Petunia, Photochemistry Ricardo da Silva, J.M., Bourzeix, M., Cheynier, and Photobiology 68; 323-330 V. andMoutounet, M. (l99la) Procyanidin composition of Chardonnay, Mauzac and S Grenache blanc grapes, Vitis 30;245-252 Saint-Cricq de Gaulejac, N., Augustin, M., Ricardo da Silva, J.M., Rigaud, J., Cheynier, V., Vivas, N. and Glories, Y. (1997) A Cheminat, A. andMoutounet, M.M. (199lb) biochemical approach to the evolution of Procyanidin dimers and trimers from grape procyanidins in grape seeds during the seeds, P hytochemis try 30 ; 1259 - 1264 ripening of red grapes (Vitis vinifera L. cv. Merlot Noir), Journal of Wine Research 8; Roggero, J.P., Ragonnet, B. and Coen, S. (1984) 159-167 Analyse flrne des anthocyanes des vins etdes pellicules de raisin par la technique HPLC, Saint-Cricq de Gaulejac, N., Vivas, N. and Vignes et vins 327 ; 38-42 Glories, Y. (1998) Phenolic ripening of red 247 grapes. Relationship with wine quality. condensation products in red wine,,4 merican Comparison of Merlot andTempranillo grape Journal ofEnology and Viticulture 48;370- cultivars, Progres Agricole et Viticole ll5; 373 306-3 I 8 Scheffeldt, P. andHrazdina, G. (1978) Saito, N. and Harborne, J.B. (1992) Correlations Copignientation of anthocyanins under between anthocyanin type, pollinator and physiological conditions, Journal of Food flower colour in the Labiatae, Phytochemistry Science 43;517-520 31;3009-3015 Scienza, A. (1976) Causes of sterility, failure of Saito, K., Kobayashi, M., Gong, Z.,Tanaka,Y. fertilisation and shot berry in grape, Yignevini andYamazaki, M. (1999) Direct evidence for 3:31-33 anthocyanin synthase as a 2-oxoglutarate- Shaver, T.N. and Lukefair, M.J. (1969) Effect of dependent oxygenase: Molecular cloning and flavonoid pigments and gossypol on growth functional expression of oDNA from a red and development of bollworm, tobacco forma of Perillafrutescens, Plant Journal 17; budworm, and pink bollworm,Journal of l8l-189 E c onomic Entomo lo gy 62 ; 643-646 Salagoity-Auguste, M-H. and Bertrand, A. Shimada, Y., Nakano-Shimada, R., Ohbayashi, (1984) V/ine phenolics - Analysis of low M., Okinaka, Y., Kiyokawa, S. andKikuchi, molecular weight components by high Y. (1999) Expression of chimeric P450 genes performance liquid chromatography, Journal encoding fl avonoid-3',5'-hydroxylase in ofthe Science ofFood and Agriculture 35; transgenic tobacco, FEBS Letters 461;241- 1241-1247 245 Sambrook, J., Fritsch, E.F. and Maniatis, T. Shirley, B.W. (1996) Flavonoid biosynthesis: (1989) Molecular Cloning: A Laboratory 'new'functions for an'old'pathway, Trends in Manual. Cold Spring Harbour Laboratory Plant Science l; 377 -382 Press, New York Shirley, 8.W., Kubasek, \ry.L., Storz, G., Santos-Buelga, C., Francia-Aricha, E.M. and Bruggemann, E., Koornneef, M., Ausubel, Escribano-Bailon, M.T. (1995) Comparative F.M. and Goodman, H.M. (1995) Analysis of flavan-3-ol composition of seeds from Arabidopsis mutants deficient in flavonoid different grape varieties, Food Chemistry 53; biosynthesis, Plant Journøl 8; 659-71 t97-201 Siegelman, H.Vy'. and Hendricks, S.B. (1958) Sato, M., Ramarathnam, N., Suzuki, Y., Ohkubo, Photocontrol of anthocyanin synthesis in T., Takeuchi, M. and Ochi, H. (1996) Varietal apple skin, Plant Physiology 33; I 85-190 differences in the phenolic content and Singleton, V.L. (1988) Wine phenols ln: Wine superoxide radical scavenging potential of Analysis (Linskens & Jackson, eds.) pp. 173- wines from different sources, Journal of 2 I 8. Springer-Verlag, Berlin Agricultural and Food Chemistry 44;37-41 Singleton, V.L. and Draper, D.E. (1964) The Saucier, C., Little, D. and Glories, Y. (1997) transfer of polyphenolic compounds from First evidence of acetaldehyde-fl avanol 248 grape seeds into wines, American Journal of Somers, T.C. (1978) Interpretations of colour Enology and Viticulture 15;34-40 composition in young red wines, Vitis 17; t6t-167 Singleton, V.L. and Esau, P. (1969) Phenolic Substances in Grapes and Wine, and Their Somers, T.C. and Evans, M.E. (1977) Spectral Significance, Academic Press, New York evaluation of young red wines: Anthocyanin equilibria, total phenolics, free and molecular Singleton, V.L., Timberlake, C.F. aîdLea, SO2, "chemical age", Journal ofthe Science of A.G.H. (1978) The phenolic cinnamates of Food and Agriculture 28;279-287 white grapes and wine, Journal of the Science ofFood and Agriculture 29;403-410 Somers, T.C. and Evans, M.E. (1979) Grape pigment phenomena: Interpretation of major Singleton, V.L. and Trousdale, E. (1983) White colour losses during vinific ation, Journal of wine phenolics: Varietal and processing the Science ofFood and Agriculture 30;623- differences as shown by HPLC,American 633 Journal of Enology and Viticulture 34; 27 -34 Somers, T.C. and Verette, E. (1988) Phenolic Skipp, R.A. and Bailey, J.A. (1977) The composition of natural wine typesln: Wine fungitoxicity of isofl avonoid phytoalexins Analysis (Linskens & Jackson, eds.) pp. 219- measured using different types ofbioassay, 257. Springer-Verlag, Berlin Physiological Plant Pathology ll; l0l-1 12 Somers, T.C., Verette, E. andPocock, K.F. Smart, R. and Sinclair (1976) Solar heating of (1987) Hydroxycinnamate esters of I/itis grape berries and other spherical fruits, vinifera: Changes during white vinifrcation, Agricultural Meteorology 17 : 241 -259 and effects ofexogenous enzymic hydrolysis, Smart, R.8., Smith, S.M. and \Vinchester, R.V. Journal ofthe Science ofFood and (1988) Light quality and quantity effects on Agriculture 40;67-78 fruit ripening for Cabernet Sauvignon, Somers, T.C. andVy'escombe, L.G. (1987) American Journal of Enology and Viticulture Evolution of red wines II. An assessment 39;250-258 of the role ofacetaldehyde, Vitis 26;27-36 Smith, D.A. (1982) Toxicity of Phytoalexins In: Somers, T.C. andZiemelis, G. (1985) Flavonol Phytoalexins (Bailey & Mansflreld, eds.) pp. haze in white wines, Vitis 24;43-50 218-252. Blackie & Son Ltd., Glasgow Souquet, J.-M., Cheynier, V., Brossaud, F. and Smith, G.J. andMarkham, K.R. (1998) Moutounet, M. (1996a) Polymeric Tautomerism of fl avonol glucosides : proanthocyanidins from grape skins, Relevance to plant UV protection and flower P hytochemistry 43; 509-5 12 colotx, Journal of Photochemistry and Photobiology A: Chemistry 118; 99-105 Souquet, J.-M., Cheynier, V., Sarni-Manchado, P. and Moutounet, M. (1996b) Les composés Somers, T.C. (1971) The polymeric nature of phénoliques du raisin, Journal Internationøl wine pigments, Phytochemistry l0; 217 5-2186 des Sciences de la Vigne et du Vin; Special Somers, T.C. (1976) Pigment development Issue,99-107 during ripening of the grape, Vitis 141,269-277 249 Souquet, J.-M., Labarbe, 8., le Guernevé, C., Stafford, H.A. and Lester, H.H. (1984) Flavan-3- Cheynier, V. andMoutounet, M. (2000) ol biosynthesis. The conversion of(+)- Phenolic composition of grape stems, Journal dihydroquercetin and fl avan-3,4-cls-diol of Agricultural and Food Chemistry 48,1076- (leucocyanidin) to (+)-catechin by reductases i080 extracted írom cell suspension çultures of Douglas Fir, Plant Physiology 76; 184-186 Spanos, G.A. and Wrolstad, R.E. (1990) Influence of rpocessing and storage on the Stafford, H.A. and Lester, H.H. (1985) Flavan-3- phenolic composition of Thompson Seedless ol biosynthesis: The conversion of(+)- grape juice,Journal ofAgricultural and Food dihydromyricetin to its flavan-3.4-diol Chemistry 38; 1565-1571 (leucodelphinidin) and to (+)-gallocatechin by reductases extracted from tissue cultures of Sparvoli, F., Martin, C., Scienza, A., Gavazzi, G. Ginkgo bi lo ba and P s eudotsuga menzis ii, and Tonelli, C. (1994) Cloning and molecular Plant Physiology 78; 1 9 I -7 94 analysis of structural genes involved in flavonoid and stilbene biosynthesis in grape Stannard, M.C., Peterson, J.R. and Sproule, R.S. (Vitis viniferaL.), Plant Molecular Biology (1974) Effects ofgibberellic acid on fruit, 24;7 43-755 "hen and chicken" condition, and vine development in the following season of Spayd, S.E., Tarara, J.M., Mee, D.L. and Waltham Cross grapes, Australian Journal of Ferguson, J.C. (2002) Separation of sunlight Exp er imental A gr icultu r e and An imal and temperature effects on the composition of Husbandry 14;256-261 Vitis vinifera cv. Merlot benies, American Journal ofEnology and Viticulture 53; 17l- Stoll, M. (2000) Effects of partial rootzone 181 drying on grapevine physiology andfruit quality, PhD Thesis; Department of Spribille, R. and Forkmann, G. (1984) Horticulture, Viticulture and Oenology, Conversion of dihydroflavonols to flavonols University of Adelaide, Adelaide with enzyme extracts from flower buds of Matthiola incana R. Br., Zeitschrift fur Strack, D. andWray, V. (1993) The Naturþrschung 39c; 7 I 4-7 19 anthocyanins In: The Flavonoids: Advances in Research since 1986 (Harborne, ed.)pp. l-22. St. Leger, S.M., Cochrane, A.L. and Moore, F. Chapman & Hall, London (1979) Factors associated with cardiac mortality in developed countries with Sun,8., Leandro, C., Ricardo da Silva, J.M. and particular reference to the consumption of Spranger, I. (1998) Separation ofgrape and wine, The Lancet 297; l0l7-1020 wine proanthocyanidins according to their degree of polymerization, Journal of Stafford, H.A. (1990) Flavonoid Metabolism. Agricultural and Food Chemistry 46;1390- CRC Press, Inc., Boca Raton, Florida 1396 Stafford, H.A. (1991) Flavonoid evolution: an Suzuki, M. (1995) Enhancement of anthocyanin enzymic approach, Plant Physiology96; 680- accumulation by high osmotic stress and low 685 pH in grape cells (Zllis hybrids), Journal of Plant Physiology 147 ; 1 52-1 55 250 Swain, T. andHillis, W.E. (1959) The phenolic Timberlake, C.F. and Bridle, P. (1976) The effect constituents of Prunus domestica I. The ofprocessing and other factors on the colour quantitative analysis of phenolic constituents, characteristics of some red wines, Vitis 15; Journal ofthe Science ofFood and 37-49 Agricultpre 10;63-68 Tsai-Su, C. and Singlêton, V.L. (1969) T Identif,rcation of three fl avan-3-ols from grapes, Phytochemistry 8; 1553-1 558 Tanaka, Y., Yonekura, K., Fukuchi-Mizutani, M., Fukui, Y., Fujiwara, H., Ashikari, T. and v Kusumi, T. (1996) Molecular and biochemical van Eldik, G.J., Reijnen, W.H., Ruiter, R.K., van characterization of three anthocyanin Herpen, M.M.A., Schrauwen, J.A.M. and synthetic enzymes frorn Gentiana triflora, Wullems, G.J. (1997) Regulation of flavonol Plant and Cell Physiology 37;7ll-6 biosynthesis during anther and pistil Tanner, G.J., Francki, K.T., Abrahams, S., development, and during pollen tube growth 'Watson, J.M., Larkin, P.J. and Ashton, A.R. in Solanum tuberosum, Plant Journal ll; (2003) Proanthocyanidin biosynthesis in 105-l 13 plants, The Journal of Biological Chemistry Vernon, L.P. (l 960) Spectrophotometric 278;31647-31656 determination of chlorophylls and Tanner, G.J., Moate, P.J., Davis, L.H., Laby, pheophytins in plant extracts, Analytical R.H., Yuguang, L. and Larkin, P.J. (1995) Chemistry 32; 1144-1150 Proanthocyanidins (condensed tannin) Vidal, S., Cartlade, D., Souquet, J.-M., Fulcrand, destabilise plant protein foams in a dose H. and Cheynier, V. (2002) Changes in dependent manner, Australian Journal of proanthocyanidin chain length in winelike Agricultural Research 46; I 101-1 109 model solutions, Journal of Agricultural and Taylor, L.P. (1995) Flavonols: Effects on fertility F o od Chemi s try 50 ; 2261 -2266 and fecundity, Crop Science 35; 152l-1526 Vogt, T., Ibdah, M., Schmidt, J.,'Wray, V., Thomas, M.R., Matsumoto, S., Cain, P. and Nimtz, m. and Strack, D. (1999) Light- Scott, N.S. (1993) Repetitive DNA of induced betacyanin and flavonol accumulation grapevine: Classes present and sequences in the bladder cells of Mesembryanthemurn suitable for cultivar identification, Theoretical cry s t al linum, P hyto ch emis t ry 52 ; 583 - 592 and Applied Genetics 86; 173-l 80 Vogt, T., Pollak, P., Taryln, N. and Taylor, L.P. Thorngate, J.H. III (1997) The physiology of (1994) Pollination- or wound-induced human sensory response to wine: A review, kaempferol accumulation in petunia stigmas American Journal of Enology and Viticulture enhances seed production, Plant Cell 6; ll-23 48;271-279 w Thomgate, J.H.III and Singleton, V.L. (1994) Wagner, M.R. (1988) Induced defenses in Localization ofprocyanidins in grape seeds, Ponderosa Pine against defoliating insects In: American Journal of Enology and Viticulture Mechanisms of l{oody Plant Defenses Against 45;259-262 251 Insects. Search for Pattern (NIattson et al., V/illiams, M., Hrazdina, G.,'Wilkinson, M.M., eds.) pp. 141-155. Springer-Verlag, New York Sweeny, J.G. andlacobucci, G.A. (1978) High-pressure liquid chromatographic Walters, T. and Stafford, H.A. (1984) Variability separation of 3-glucosides, 3,5-diglucosides, in accumulation of proanthocyanidins 3 -(6-0 -cotmaroyl)glucosides and 3 -(6-0 -- (condensed tannins) in needles ofDouglas Fir' -p p-coumaroylglucosides)-5 -glucosides of (P s eudots uga menziesii) following long-term anthocyanidins, Journal of Chromatography budworm defoliation, Journal of Chemical 155; 389-398 Ecology l0;1469-1476 .Wan, Winkler,4.J., Cook, J.4., Kliewer, W.M. and C.-Y. and Wilkins, T. (1994) A modiflred Lider, L.A. (1974) General Viticulture (2nd hot borate method signihcantly enhances the Edition), University of Califomia Press, yield of high-quality RNA from cotton Berkeley (Gossypium hirs utum L.), Analytical 'Woltering, Biochemistry 223;7-12 E.J. and Somhorst, D. (1990) 'Waterman, Regulation of anthocyanin synthesis in P.G. and Mole, S. (1994)Analysis of Cymbidiumflowers: Effects of emasculation Phenolic Plant Compounds. Methods in and ethylene, Journal ofPlant Physiology Ecology (Lawton & Likens, eds,) Blackwell 136;295-299 Scientific Publications, Oxford rWulf, L.W. andNagel, C.W. (1978) High- Weiss, M., Schmidt, J., Neumann, D., Wray, V., pressure liquid chromatographic separation of Christ, R. and Strack, D. (1999) anthocyanins of Vitis vinifera, American Phenylpropanoids in mycorrhizas of the Journal ofEnology and Viticulture 291'42-49 Pinaceae, P I ant a 208 ; 49 I -502 White, R.C., Jones, I.D. and Gibbs, E. (1963) x Determination of chlorophylls, Xie, D.-Y., Sharma, S.8., Paiva, N.L., Ferreira, chlorophyllides, pheophytins, and D. and Dixon, R.A. (2003) Role of pheophorbides in plant material, Journal of anthocyanidin reductase, encoded by Food Science 28; 431-436 BANYULS in plant flavonoid biosynthesis, Wicks, A.S. andKliewer, W.M. (1983) Further Science 299;396-399 investigations into the relationship between Y anthocyanins, phenolics and soluble carbohydrates in grape berry skins, American Yamasaki, H., Uefuji, H. and Sakihama, Y. Journal ofEnology and Viticulture 34; ll4- (1996) Bleaching ofthe red anthocyanin 116 induced by superoxide radical, Archives of Biochemistry and Biophysics 332; 183-186 Wildenradt, H.L. and Singleton, V.L. (1974) The production of aldehydes as a result of Ylstra, B., Busscher, J., Franken, J., Hollman, oxidation of polyphenolic compounds and its P.C.H., Mol, J.N.M. and vanTunen, A.J. relation to wine aging, American Journal of (1994) Flavonols and fertilisation inPetunia Enology and Viticulture 25; 119-126 hybrida: Localisation and mode of action 252 during pollen tube growth, Plant Journal 6; 201-212 Ylstra, B., Touraev,,A'., Moreno, R.M.B., Stoger, E., vanTunen, 4.J., Vicente, O., Mol, J.N.M. and Heberle-Bors, E. (1992) Flavonols stimulate development, germination and tube growth of tobacco pollen, Plant Physiology 100;902-907 z Zaat, S. A,.J., Schripsema, J., Wij iffelman, C.4., van Brussel, A.A.N. andLugtenberg, B.J.J. (1989) Analysis of major inducers of the Rhizobíum nodA ptomoler from Yicia sativa root exudate and their activity with different nodD genes, Plant Molecular Biology 13; 175-718 Zucker, M. (197 2) Light and errøymes, Annual Review of P lant Physio lo gy 23; 1 33- 1 56 253 C¿lffil IID llOES¡t cEçtÌatltL{. flATS TI€ t nElùIr À!$ qEsñrsls FR¡¡{rL{. TD LII(E Ib Th{E 'ñrç t{ËxÎ c L{rt{ Dor¡lÌ sf !.H¡T n|E, S\IG¡E(T AT hE. iSE\JE FESoLT¡ÊÞ E.FoRE cr{ÀPfÊÊ, I[¡Àf ìs ll.t LAil I Þ(FÊì{D å¡{I IBÊ E,ì{ÉI6I EXISTEì{ç.E? or1 Tt{\s I IIÀ\IE À qÊsnoÈ,¡ a¡ L o a I o F o a a lr o I c a 2s4