Letters to Nature

Nature 416, 847-850 (25 April 2002) | doi:10.1038/416847a; Received 4 January 2002; Accepted 14 February 2002

Association of dwarfism and floral induction with a grape 'green revolution' mutation

Paul K. Boss1 & Mark R. Thomas

  1. CSIRO Plant Industry and Cooperative Research Centre for Viticulture, PO Box 350, Glen Osmond, SA 5064, Australia
  2. Present address: John Innes Centre, Norwich Research Park, Norwich, Norfolk NR4 7UH, UK.

Correspondence to: Mark R. Thomas Correspondence and requests for materials should be addressed to M.R.T. (e-mail: Email: mark.r.thomas@csiro.au). The GenBank library accession number for VvGAI1 is AF378125.

Top

The transition from vegetative to reproductive growth is an essential process in the life cycle of plants. Plant floral induction pathways respond to both environmental and endogenous cues and much has been learnt about these genetic pathways by studying mutants of Arabidopsis1, 2. Gibberellins (GAs) are plant growth regulators important in many aspects of plant growth and in Arabidopsis they promote flowering3, 4, 5. Here we provide genetic evidence that GAs inhibit flowering in grapevine. A grapevine dwarf mutant derived from the L1 cell layer of the champagne cultivar Pinot Meunier produces inflorescences along the length of the shoot where tendrils are normally formed. The mutated gene associated with the phenotype is a homologue of the wheat 'green revolution' gene Reduced height-1 (ref. 6) and the Arabidopsis gene GA insensitive (GAI)7. The conversion of tendrils to inflorescences in the mutant demonstrates that the grapevine tendril is a modified inflorescence inhibited from completing floral development by GAs.

Grapevine (Vitis sp.) is one of the world's major perennial horticultural crops. It is a vine, and under natural conditions tendrils are used to support a tree-climbing habit to reach high sunlight levels for flowering8. A small number of Vitis vinifera cultivars dominate wine production in the world owing to their reputation for producing premium quality wine, and in France the Champagne region has become famous for its sparkling wine. Pinot Meunier, Pinot noir and Chardonnay are the only three cultivars authorized to be grown for champagne production; together the black berry cultivars, Pinot Meunier and Pinot noir, represent 74% of the planted vines. Pinot Meunier is a cultivar of ancient origins and has long been considered a periclinal mutant of Pinot noir. It is distinguished from Pinot noir in having tomentose (densely covered with trichomes) shoot tips and expanding leaves9, 10. All grapevine cultivar propagation is vegetative, and novel phenotypes, like that of Pinot Meunier, arise by somatic mutation.

The apical meristem of the grapevine shoot is organized into two distinct layers designated L1 (outermost) and L2 (ref. 11). Plants have been regenerated from the L1 and L2 cell layers of Pinot Meunier by passage through somatic embryogenesis, and whereas those from the L2 cell layer were phenotypically indistinguishable from Pinot noir, the plants regenerated from the L1 cell layer displayed the tomentose phenotype of Pinot Meunier and were dwarfed12. When grown under glasshouse conditions favourable for floral induction, the L1 dwarf plants produced inflorescences and bunches along the length of the shoots (Fig. 1a, c) where the L2 plants (and Pinot Meunier) had a normal phenotype and produced tendrils (Fig. 1b, d). Inflorescences and tendrils in grapevines are derived from meristematic structures called uncommitted primordia (Fig. 1e), which develop from shoot meristems and are found opposite two of every three leaves13. Uncommitted primordia formed on actively growing shoots develop into tendrils (Fig. 1b, f), whereas those in latent buds develop into inflorescences. Latent buds are formed during spring and summer and experience a winter dormancy before bud burst and flowering (Fig. 1d). In the L1 dwarf plants this process is not necessary and uncommitted primordia differentiate into inflorescences on actively growing shoots (Fig. 1a, g).

Figure 1: The L1 plant produces inflorescences instead of tendrils.
Figure 1 : The L1 plant produces inflorescences instead of tendrils. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

a, Main shoot of an L1 plant. b, Main shoot of an L2 plant. c, Shoot from a latent bud of an L1 plant (leaves removed). d, Shoot from a latent bud of an L2 plant. e, Scanning electron micrograph of a shoot meristem from a wild-type latent bud showing an uncommitted primordium (UP) that has separated from the shoot apical meristem (SAM). f, Scanning electron micrograph of a tendril tip from an L2 plant. g, Scanning electron micrograph showing flowers at the tip of a tendril-like structure from an L1 plant. Scale bar in eg, 100 microm.

High resolution image and legend (91K)

The dwarf stature of the L1 plants was consistent with altered levels of GAs or an altered response to GAs. The application of GAs and inhibitors of GA biosynthesis has been shown to modify grapevine tendril and inflorescence development14, 15, 16. We concluded, on the basis of the following, that the L1 plants had an altered GA response and that this is associated with a mutated gene similar to the Arabidopsis gene GAI7, 17, a negative regulator of GA response. First, the L1 plants did not respond when GA was applied, indicating that it was not a GA-deficient dwarf. Second, the L1 mutant accumulated fourfold more GA1 and 12-fold more GA4 in leaves than the L2 plant (data not shown). Elevated levels of bioactive GAs were found in the Arabidopsis gai mutant18. Third, inheritance studies confirmed the semidominant nature of the mutation, similar to the Arabidopsis gai mutant, because two sizes of dwarf plants were observed in L1 times L1 progeny (Fig. 2a; see below).

Figure 2: Mutant phenotypes.
Figure 2 : Mutant phenotypes. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

a, The stems of tall (left), semi-dwarf (middle) and dwarf (right) plants obtained from the L1 times L1 cross (leaves and bunches removed). The inset shows a semi-dwarf plant (left) and a dwarf plant (right) 3 months after germination. b, Tendril length on L2 plants (L2 tendril; n = 23), immature tendril-like structures on L1 plants (L1 tendril; n = 35), L1 tendril/inflorescence structures with terminal flowers (L1 tendril + flowers; n = 22) or L1 tendril/inflorescence structures with berries (L1 tendril + berries; n = 26). Results are means plusminus s.e.m.

High resolution image and legend (42K)

GAI-like gene sequences (called VvGAI1, for Vitis vinifera GA-insensitive) were amplified by polymerase chain reaction (PCR) with complementary DNA produced from Pinot Meunier. The deduced amino acid sequence aligned with GA-response members of the Arabidopsis GRAS family of regulatory proteins, and VvGAI1 is 72% identical to GAI (Supplementary Information, Figs S1 and S2). Three similar but distinct VvGAI1 sequences were amplified from Pinot Meunier, indicating that three alleles were present in the chimaera (Supplementary Information, Fig. S3). The L1 and L2 plants possessed only two alleles from Pinot Meunier. One allele was common to both plants, whereas the other differed between L1 and L2 plants solely by the presence of a point mutation in the DELLA domain (Fig. 3; Supplementary Information, Fig. S3). This conserved domain is unique to all GA-response members of the GRAS family of plant regulatory proteins7, 19. Previously described semi-dominant dwarf mutants in Arabidopsis and cereals have DELLA domain mutations that probably result in deletions or truncations of the protein6, 7 (Fig. 3). It is postulated that these modified proteins are no longer able to interact with a GA signal because of the deletions within the DELLA region and remain in a growth-repressing form, resulting in a dwarf phenotype6, 7. The point mutation present in the VvGAI1 allele from both Pinot Meunier and the L1 plant converts a leucine residue of the conserved DELLA domain into histidine (Fig. 3). The change from a small hydrophobic leucine residue into a larger basic histidine residue seems to have altered the GA-response properties of the protein, causing a dwarf phenotype in the grapevine. Future work will determine the importance of this leucine for GA responses and its relative importance to the other amino acids in the DELLA domain.

Figure 3: The Vvgai1 point mutation results in an amino-acid substitution.
Figure 3 : The Vvgai1 point mutation results in an amino-acid substitution. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

The DELLA domains of the predicted proteins of wild-type alleles of the Arabidopsis GAI and homologues from maize (d8), wheat (Rht-B1a, Rht-D1a) and grapevine (VvGAI1) are compared to their mutant alleles6,7 below. Highly conserved regions of this domain, named I and II by Peng et al.6, are shown. Deletions are represented by the dashed line and the position of translation stop codons are represented by an asterisk. A nucleotide substitution results in a leucine to histidine substitution in the I region of the DELLA domain of VvGAI1.

High resolution image and legend (51K)

The L1 plants were self-fertilized or crossed to Vitis riparia and the progeny were genotyped by using analyses with the dCAPS finder program20. All of the tall progeny were homozygous wild-type (VvGAI1/VvGAI1) (Table 1). The dwarf progeny were either heterozygous (VvGAI1/Vvgai1) or homozygous (Vvgai1/Vvgai1) for the point mutation. The Vvgai1/Vvgai1 plants were smaller than the heterozygotes and confirmed the semidominant nature of the mutation (Fig. 2a). Both dwarf genotypes had tomentose shoot tips and produced inflorescences from uncommitted primordia on actively growing shoots, whereas tall progeny were not tomentose and produced only tendrils. These data show that the dwarf, tomentose and prolific flowering phenotypes segregate with the mutated Vvgai1 allele.


The change in GA response of the mutant also had a significant effect on seed dormancy and distorted the ratios of genotypes recovered from the crosses (Table 1). With the use of a typical method21, the seed germination rate was low, and only wild-type plants were recovered. When the seed was treated with GA3 and scarified, higher germination rates and dwarf progeny were obtained. Without this treatment, seed containing the mutation remained dormant under normal germination conditions for more than a year.

Although most uncommitted primordia on Vvgai1-containing plants developed into inflorescences, some did not and resembled immature tendrils. Long tendrils did not form, indicating that a normal GA response is required for complete development (Fig. 2b). Conversion of the tendril to an inflorescence by the development of flowers (Fig. 1g), and later berries, stimulates the elongation of what has now become the rachis (Fig. 2b). This suggests that floral meristems and fruit have a role in determining inflorescence size and structure that is distinct from GA signalling via VvGAI1.

Gibberellin sprays are regularly used in the vineyard to increase the berry size of seedless grapevine varieties artificially. Surprisingly, the weight and volume of mature berries from L1 plants were not smaller than those of L2 or Pinot Meunier berries (data not shown). VvGAI1 expression was not detected in berries (Supplementary Information, Fig. S4). This demonstrates that VvGAI1 has no role in berry development and explains why fruit size is not reduced in the mutant.

Pinot Meunier is a very old cultivar and has been known since the 1500s (ref. 9). The existence of the Vvgai1 mutation in Pinot Meunier pre-dates the introduction of mutant dwarfing alleles into cereals during the 'green revolution' of the 1960s and 1970s (ref. 6). The mutation seems to be cell-autonomous (as is suggested for gai (ref. 22), with only the tomentose phenotype being observed in the Pinot Meunier chimaera and no obvious effect on flowering or plant stature. The full effect of the mutation in grapevine was revealed only when the L1 and L2 layers were separated. In the vineyard, intensive management practices are required to control vine vigour and promote flowering and fruit development. The L1 mutant seems to possess desirable agronomic characteristics of both reduced vigour and increased fruiting potential.

Flowering in Arabidopsis involves a network of floral induction pathways (autonomous, photoperiod, vernalization and GA) that act redundantly to ensure that flowering occurs under appropriate conditions23. Grapevine floral induction in latent buds is stimulated by high light intensity and high temperature rather than by daylength and vernalization24. The association of the Vvgai1 mutation with the marked effect on grapevine flowering indicates a major role for GAs in floral inhibition. Reproductive growth in the shoot is not totally inhibited by GAs because uncommitted primordia are still made, indicating that GA signalling through VvGAI1 acts later by inhibiting the production of floral meristems and creates a new organ—a tendril. In the mutant, floral meristems are produced and normal tendrils are absent.

Vitis vinifera has evolved under very different conditions from those of Arabidopsis. Its natural habitat is forest, where it has adapted to climbing trees to move from a shaded environment below the canopy to full sunlight above. It seems that GAs not only promote stem internode elongation to escape the shaded environment but also convert inflorescences into tendrils for enhanced climbing ability. Our findings suggest that whereas the transmission of the GA signal in the floral genetic pathway promotes flowering in Arabidopsis3, 4, 5, it inhibits floral meristem production in grapevine.

Top

Methods

Plant material and seed germination

Grapevines were grown in commercial potting mix in a glasshouse with temperature and light/dark cycles of 16 h at 30 °C and 8 h at 25 °C. In winter months, the light duration was extended to 16 h through the use of artificial lighting. The Pinot Meunier L1 dwarf and the L2 vines with normal growth were regenerated from embryogenic callus as described12. Self-pollinated L1 dwarf seed and seed obtained from outcrossing the L1 dwarf with Vitis riparia were cold-treated for at least 3 weeks before scarification and germination in light on moistened filter paper. Seedlings were transferred to soil and slowly acclimatized to glasshouse conditions.

PCR amplification of GAI homologues

Degenerate oligonucleotide primers were designed to internal conserved regions of known GAI homologues obtained from GenBank databases. Initial cloning was performed with Shiraz cDNA synthesized from immature inflorescence total RNA that was isolated as described25. After the cloning of an internal GAI-like fragment, 5' and 3' rapid amplification of cloned ends (RACE) reactions26 were used to obtain the full-length sequence for VvGAI1. Primers were used to isolate full-length VvGAI1 alleles from Pinot Meunier, L1 and L2 plants. Sequences were compared from two independent amplifications from DNA and one amplification from cDNA to confirm differences.

DNA extraction and dCAPS analysis

DNA from each grapevine was extracted from leaves with a method described previously27. An oligonucleotide primer was designed that would distinguish between the two species of VvGAI1 sequences found in the Pinot Meunier L1 dwarf cDNA by using the dCAPS finder program20. The primer introduced a XhoI site at the 3' end of the PCR product if the wild-type sequence was the template. After PCR amplification and digestion with XhoI, the presence of either VvGAI1 allele could be determined by electrophoretic separation of the PCR products on a 2.5% agarose gel.

Top

References

------------------

References

1. Hempel, F. D., Welch, D. R. & Feldman, L. J. Floral induction and determination: where is flowering controlled? Trends Plant Sci. 5, 17-21 (2000). | Article | PubMed | ISI |
1. Hempel, F. D., Welch, D. R. & Feldman, L. J. Floral induction and determination: where is flowering controlled? Trends Plant Sci. 5, 17-21 (2000). | Article | PubMed | ISI |
2. Araki, T. Transition from vegetative to reproductive phase. Curr. Opin. Plant Biol. 4, 63-68 (2001). | Article | PubMed | ISI |
2. Araki, T. Transition from vegetative to reproductive phase. Curr. Opin. Plant Biol. 4, 63-68 (2001). | Article | PubMed | ISI |
3. Blazquez, M. A., Green, R., Nilsson, O., Sussman, M. R. & Weigel, D. Gibberellins promote flowering of Arabidopsis by activating the LEAFY promoter. Plant Cell 10, 791-800 (1998). | Article | PubMed | ISI |
3. Blazquez, M. A., Green, R., Nilsson, O., Sussman, M. R. & Weigel, D. Gibberellins promote flowering of Arabidopsis by activating the LEAFY promoter. Plant Cell 10, 791-800 (1998). | Article | PubMed | ISI |
4. Blazquez, M. A. & Weigel, D. Integration of floral inductive signals in Arabidopsis. Nature 404, 889-892 (2000). | Article | PubMed | ISI |
4. Blazquez, M. A. & Weigel, D. Integration of floral inductive signals in Arabidopsis. Nature 404, 889-892 (2000). | Article | PubMed | ISI |
5. Langridge, J. Effect of day-length and gibberellic acid on the flowering of Arabidopsis. Nature 180, 36-37 (1957). | ISI |
5. Langridge, J. Effect of day-length and gibberellic acid on the flowering of Arabidopsis. Nature 180, 36-37 (1957). | ISI |
6. Peng, J. et al. 'Green revolution' genes encode mutant gibberellin response modulators. Nature 400, 256-261 (1999). | Article | PubMed | ISI |
6. Peng, J. et al. 'Green revolution' genes encode mutant gibberellin response modulators. Nature 400, 256-261 (1999). | Article | PubMed | ISI |
7. Peng, J. et al. The Arabidopsis GAI gene defines a signaling pathway that negatively regulates gibberellin responses. Genes Dev. 11, 3194-3205 (1997). | PubMed | ISI |
7. Peng, J. et al. The Arabidopsis GAI gene defines a signaling pathway that negatively regulates gibberellin responses. Genes Dev. 11, 3194-3205 (1997). | PubMed | ISI |
8. Rives, M. Vigour, pruning, cropping in the grapevine (Vitis vinifera L.). I. A literature review. Agronomie 20, 79-91 (2000). | ISI |
8. Rives, M. Vigour, pruning, cropping in the grapevine (Vitis vinifera L.). I. A literature review. Agronomie 20, 79-91 (2000). | ISI |
9. Viala, P. & Vermorel, V. Ampélographie, Tome II (Masson, Paris, 1901).
10. Skene, K. G. M. & Barlass, M. Studies on the fragmented shoot apex of grapevine. IV. Separation of phenotypes in a periclinal chimera in vitro. J. Exp. Bot. 34, 1271-1280 (1983). | ISI |
10. Skene, K. G. M. & Barlass, M. Studies on the fragmented shoot apex of grapevine. IV. Separation of phenotypes in a periclinal chimera in vitro. J. Exp. Bot. 34, 1271-1280 (1983). | ISI |
11. Thompson, M. M. & Olmo, H. P. Cytohistological studies of cytochimeric and tetraploid grapes. Am. J. Bot. 50, 901-906 (1963). | ISI |
11. Thompson, M. M. & Olmo, H. P. Cytohistological studies of cytochimeric and tetraploid grapes. Am. J. Bot. 50, 901-906 (1963). | ISI |
12. Franks, T., Botta, R. & Thomas, M. R. Chimerism in grapevines: implications for cultivar identity, ancestry and genetic improvement. Theor. Appl. Genet. 104, 192-199 (2002). | ISI |
12. Franks, T., Botta, R. & Thomas, M. R. Chimerism in grapevines: implications for cultivar identity, ancestry and genetic improvement. Theor. Appl. Genet. 104, 192-199 (2002). | ISI |
13. Pratt, C. Reproductive anatomy in cultivated grapes: a review. Am. J. Enol. Vitic. 22, 92-109 (1971). | ISI |
13. Pratt, C. Reproductive anatomy in cultivated grapes: a review. Am. J. Enol. Vitic. 22, 92-109 (1971). | ISI |
14. Mullins, M. G. Regulation of inflorescence growth in cuttings of the grapevine (Vitis vinifera L.). J. Exp. Bot. 19, 532-543 (1968). | ISI |
14. Mullins, M. G. Regulation of inflorescence growth in cuttings of the grapevine (Vitis vinifera L.). J. Exp. Bot. 19, 532-543 (1968). | ISI |
15. Utsunomiya, N., Sugiura, A. & Tomana, T. Effect of CCC and pinching on inflorescence induction and development on lateral shoots in grapevine. J. Jpn Soc. Hort. Sci. 47, 151-157 (1978). | ISI |
15. Utsunomiya, N., Sugiura, A. & Tomana, T. Effect of CCC and pinching on inflorescence induction and development on lateral shoots in grapevine. J. Jpn Soc. Hort. Sci. 47, 151-157 (1978). | ISI |
16. Boss, P. K. & Thomas, M. R. Tendrils, inflorescences and fruitfulness: A molecular perspective. Aust. J. Grape Wine Res. 6, 168-174 (2000).
17. Koornneef, M. et al. A gibberellin insensitve mutant of Arabidopsis thaliana. Physiol. Plant. 65, 33-39 (1985). | ISI |
17. Koornneef, M. et al. A gibberellin insensitve mutant of Arabidopsis thaliana. Physiol. Plant. 65, 33-39 (1985). | ISI |
18. Peng, J. et al. Extragenic suppressors of the Arabidopsis gai mutation alter the dose-response relationship of diverse gibberellin responses. Plant Physiol. 119, 1199-1207 (1999). | Article | PubMed | ISI |
18. Peng, J. et al. Extragenic suppressors of the Arabidopsis gai mutation alter the dose-response relationship of diverse gibberellin responses. Plant Physiol. 119, 1199-1207 (1999). | Article | PubMed | ISI |
19. Pysh, L. D., Wysocka-Diller, J. W., Camilleri, C., Bouchez, D. & Benfey, P. N. The GRAS gene family in Arabidopsis: Sequence characterization and basic expression analysis of the SCARECROW-LIKE genes. Plant J. 18, 111-119 (1999). | Article | PubMed | ISI |
19. Pysh, L. D., Wysocka-Diller, J. W., Camilleri, C., Bouchez, D. & Benfey, P. N. The GRAS gene family in Arabidopsis: Sequence characterization and basic expression analysis of the SCARECROW-LIKE genes. Plant J. 18, 111-119 (1999). | Article | PubMed | ISI |
20. Neff, M. M., Neff, J. D., Chory, J. & Pepper, A. E. dCAPS, a simple technique for the genetic analysis of single nucleotide polymorphisms: Experimental applications in Arabidopsis thaliana genetics. Plant J. 14, 387-392 (1998). | Article | PubMed | ISI |
20. Neff, M. M., Neff, J. D., Chory, J. & Pepper, A. E. dCAPS, a simple technique for the genetic analysis of single nucleotide polymorphisms: Experimental applications in Arabidopsis thaliana genetics. Plant J. 14, 387-392 (1998). | Article | PubMed | ISI |
21. Ellis, R. H., Hong, T. D. & Roberts, E. H. A note on the development of a practical procedure for promoting the germination of dormant seed of grape (Vitis spp.). Vitis 22, 211-219 (1983). | ISI |
21. Ellis, R. H., Hong, T. D. & Roberts, E. H. A note on the development of a practical procedure for promoting the germination of dormant seed of grape (Vitis spp.). Vitis 22, 211-219 (1983). | ISI |
22. Peng, J. & Harberd, N. P. Transposon-associated somatic gai-loss sectors in Arabidopsis. Plant Sci. 130, 181-188 (1997). | Article | ISI |
22. Peng, J. & Harberd, N. P. Transposon-associated somatic gai-loss sectors in Arabidopsis. Plant Sci. 130, 181-188 (1997). | Article | ISI |
23. Simpson, G. G., Gendall, A. R. & Dean, C. When to switch to flowering. Annu. Rev. Cell Dev. Biol. 15, 519-550 (1999). | Article | PubMed | ISI |
23. Simpson, G. G., Gendall, A. R. & Dean, C. When to switch to flowering. Annu. Rev. Cell Dev. Biol. 15, 519-550 (1999). | Article | PubMed | ISI |
24. Buttrose, M. S. Climatic factors and fruitfulness in grapevines. Hort. Abstr. 44, 319-326 (1974).
25. Boss, P. K., Davies, C. & Robinson, S. P. Analysis of the expression of anthocyanin pathway genes in developing Vitis vinifera L. cv. Shiraz grape berries and the implications for pathway regulation. Plant Physiol. 111, 1059-1066 (1996). | PubMed | ISI |
25. Boss, P. K., Davies, C. & Robinson, S. P. Analysis of the expression of anthocyanin pathway genes in developing Vitis vinifera L. cv. Shiraz grape berries and the implications for pathway regulation. Plant Physiol. 111, 1059-1066 (1996). | PubMed | ISI |
26. Frohman, M. A., Dush, M. K. & Martin, G. R. Rapid production of full-length cDNAs from rare transcripts: amplification using a single gene-specific oligonucleotide primer. Proc. Natl Acad. Sci. USA 85, 8998-9002 (1988). | PubMed | ISI |
26. Frohman, M. A., Dush, M. K. & Martin, G. R. Rapid production of full-length cDNAs from rare transcripts: amplification using a single gene-specific oligonucleotide primer. Proc. Natl Acad. Sci. USA 85, 8998-9002 (1988). | PubMed | ISI |
27. Thomas, M. R., Matsumoto, S., Cain, P. & Scott, N. S. Repetitive DNA of grapevine: classes present and sequences suitable for cultivar identification. Theor. Appl. Genet. 86, 173-180 (1993). | ISI |
27. Thomas, M. R., Matsumoto, S., Cain, P. & Scott, N. S. Repetitive DNA of grapevine: classes present and sequences suitable for cultivar identification. Theor. Appl. Genet. 86, 173-180 (1993). | ISI |
Top

Supplementary Information

Supplementary information accompanies this paper.

Top

Acknowledgements

We thank R. King and A. Poole for the GA analyses by GC–MS; T. Franks for plant material; S. McClure for help with the scanning electron mictroscopy; and D. MacKenzie and C. Reich for technical assistance. This work was supported in part by the Grape and Wine Research and Development Corporation, Dried Fruits Research and Development Council and the Cooperative Research Centre for Viticulture.

Top

Competing interests statement

The authors declare no competing financial interests.