6. IN VITRO CULTURE FOR GRAPEVINE IMPROVEMENT
6.4. Haploid plant production and mutatiou breeding
Progress in the genetics of the grapevine is hindered by the high heterozygosity of the genome (Alleweldt, 1997). Availability of homozygous Vitis plants would be of consid- erable interest for grapevine genetics and breeding. Selection of pure lines of Vitis vinif- era cv Pinot noir by repeated self-pollination was developed by Bronner and Oliveira (1990), but homogeneity for numerous characters requires at least 6 generations of self- ing. Searches for haploid plants in polyembryonic seeds from intraspecific or interspeci- fic origin have failed (Bouquet, 1978; Olmo, 1978), as has a biotechnological approach by in vitro anther culture. Though development of multinucleate pollen grains and hap- loid tissues was reported (Gresshoff and Doy, 1974; Rajasekaran and Mullins, 1979;
Bouquet et al., 1982; Altamura et al., 1992; Wei and Ziyi, 1993), all regenerated plant- lets were diploid. Using cytofluorometric determinations of the nuclear DNA content of anther-derived embryos of Vitis vinifera cv Grenache at the torpedo stage, Faure (1992) showed that all were diploid. However, Wei and Ziyi (1993) observed that embryos and plants developed from anther culture of the same variety were mainly triploid. There was strong anatomical evidence that the anther-derived callus originated primarily from so- matic cells of the anther wall, the connective or the filament (Nadel, 1977; Newton and Goussard, 1990). Progeny of selfed anther-derived plants exhibited segregation for leaf characters, indicating heterozygosity and somatic origin. Isoenzyme patterns in anther- derived plants and in the original genotype from which the cultures were established were indistinguishable (Rajasekaran and Mullins, 1983). Although Zou and Li (1981) reported the production of haploid plants by anther culture, this result was questioned and could not be repeated. Recently Sefc et al. (1997) obtained embryoid structures from isolated Vitis microspores, but neither calli nor embryoids regenerated plants. Moreover, due to the poor mitotic activity of the tissues, their ploidy level could not be examined.
As haploid grapevine plants were never reported in nature, it can be assumed that hap- loidy is either lethal or highly deleterious.
Mutation breeding has been applied to grapevine seeds, pollen and somatic tissues, as to other fruit species, using both physical and chemical mutagens (Das and Mukherjee, 1968; Coutinho, 1975). Broertjes and Harten (1988) highlighted the interest of in vitro cultures in mutation for vegetatively propagated crops, particularly ornamentals and fruit trees. But until now, applications of this technique to grapevine improvement are limited (Kim et al., 1986; Rosati et al., 1990; Lima da Silva and Doazan, 1995).
306 L. TORREGROSA et al.
6.5. Somaclonal variation
Genetic variability is a ubiquitous phenomenon associated with tissue culture of plants (Skirvin, 1978; Reisch, 1983). Variation generated by the use of a tissue culture cycle was called somaclonal variation by Larkin and Scowcroft (1981). They defined a tissue culture cycle as a process that involves the establishment of a de-differentiated cell or tissue culture under defined conditions, proliferation for a number of cell generations, and the subsequent regeneration of plants. The nature and causes of somaclonal varia- tion, whether encompassing changes at the cytological, molecular, cytoplasmic or epige- netic levels, are not yet clear, but have far wider significance than the originally expected contribution to plant breeding. There is no single origin but rather a multiplicity of con- tributing processes. Somaclonal variation arises because: (i) the stabilizing influence of organized growth is lost in vitro; (ii) genomic changes are associated with differentiation and de-differentiation of plant cells; (iii) errors in cell division occur when cells are in- duced to divide in culture; (iv) growth regulators influence plant growth and cell divi- sion; (v) plant genomes are not inflexible but are responsive to the environment and ca- pable of rapid changes (Karp, 1991 and 1994).
In grapevine, changes were commonly observed among plants regenerated from so- matic embryos. They involved chlorophyll deficiencies in regenerated plantiets (Bouquet et al., 1990a), variation in morphogenetic development (Bouquet, 1989a), modifications in leaf shape (Bouquet 1989c), and changes in the flower type of rootstock varieties with male flowers (Fig. 12.6). These changes were first observed in Aramon x Rupestris Ganzin hybrids (Mullins, 1987; Bouquet, I 989c), but also in several rootstocks derived from Vilis rupestris: 3309 Couderc, 110 Richter, 140 Ruggeri (Bouquet, unpublished results). Curi-
Figure 12.6. High fruit yield observed on a perfect-flowered somaclonal variant issued from the male rootstock cultivar Aramon x Rupestris Ganzin (from Bouquet, 1989c).
IN VITRO CULTURE AND PROPAGATION OF GRAPEVINE 307 ously, they were never observed in Rupestris cv du Lot. The changes could be related to a modification of the metabolism of endogenous cytokinins induced by the in vitro condi- tions necessary for the regeneration of the somatic embryos. It is well known that expres- sion of the sexual type from male to perfect-flowered can be modified by exogenous cyto- kinins (Negi and Olmo, 1966). Recently, molecular evidence was provided for somaclonal variation in three plants among a population of 47 protoclones, i.e. plants regenerated from protoplasts. The protoclones affected by the mutation possess an altered phenotype, as they show strongly increased formation of axillary shoots compared to control plants and to other protoclones (Schneider et al., 1996).
The variation arising in tissue culture is essentially random in nature and its successful exploitation depends upon the availability of rapid, accurate screening procedures. The development of these methods is relatively straightforward for characteristics such as dis- ease resistance (Daub, 1986). Micropathogenicity tests are available for selection in vitro for resistance to downy mildew (Aldwinckle, 1980; Lee and Wicks, 1982; Barlass et aI., 1986), powdery mildew (Aldwinckle, 1980; Klempka et al., 1984), eUtypa dieback (Mauro et aI., 1988), crown gall disease (Hemstad and Reisch, 1985) and bacterial necrosis (peros et at., 1995). Grzegorcsyk and Walker (1998) developed an in vitro dual culture for phyl- loxera. They tested a wide range of grape species and rootstocks and verified that reactions in this environment mirrored observations from field conditions. Attempts to develop in vitro screening for grey mold resistance have been carried out (Vannel et aI., 1991; Bessis et at., 1992) but the differential response of vitroplants of assayed cultivars to culture fil- trates of the fungus seems to be unrelated to bunch susceptibility to grey mold under field conditions (Fanizza et al., 1995). Selection at the level ofphytoalexin production could be an interesting possibility for future research (Barlass et aI., 1987; Sbaghi et aI., 1995).
In vitro assays were developed to select somaclones of Vitis vinifera cv Ugni-Blanc tolerant to the toxin Eutypine, which participates in the pathogenic action of the fungus responsible for the Eutypa dieback (Soulie et aI., 1993). Previously, protoplasts from two plantlets regenerated from anther-derived somatic embryos of cv Cabemet Sauvignon were found to exhibit a lower susceptibility to the culture filtrates of the fungus Eutypa lata (Mauro, 1986), but these somac1ones did not show any resistance to the fungus in greenhouse tests (Peros and Berger, 1994 and unpublished results). Selection of a somaclone of V vinifera cv Grenache exhibiting a high degree of tolerance in vitro to the bacterial necrosis (Xylophilus arnpelinus) was reported (Bouquet, 1989a; Ride and Bou- quet, unpublished results), but this somaclone proved to be susceptible in greenhouse tests (Peros and Bouquet, unpublished results).
In vitro selection for NaCI tolerance among anther-derived somatic embryos of the rootstock Rupestris cv du Lot was carried out (Lebrun et aI., 1985). Some of the plantlets regenerated from selected embryos in liquid media containing up to 150 mM NaCI were micropropagated and subcultured on salt-free medium. When micrografted with V vinif- era cv Syrah, they showed NaCl tolerance significantly higher than plantlets issuing from non-selected embryos, but this in vitro tolerance disappeared after several cycles of subculture (Bouquet, unpublished results). These results correspond to those of Skene and Barlass (1988), who observed that apparent tolerance to elevated chloride levels in
308 L. TORREGROSA et al.
vitro was a physiological adaptation to culture conditions, which did not persist in whole plants regenerated from culture.
However, in vitro testing could be used for rapid selection of tolerant varieties, ac- cording to Troncoso et al. (1999) who observed a close correlation on II rootstocks be- tween salt tolerance observed under in vitro conditions and that obtained by pot experi- ments or reported in the literature. Netzer et al. (1991) studied the behaviour of four grapevine rootstocks (4IB, Rupestris cv du Lot, 3309 C and 101-14 Mgt) during somatic embryogenesis under conditions of calcareous chlorosis induced by the addition of dif- ferent concentrations of KHC03 in the culture medium. According to their results, the classification of the rootstocks was similar to their classification for sensitivity to cal- careous chlorosis observed in field conditions. But they could not conclude if the toler- ance to high levels of KHC03 observed among selected plantlets was due to physiologi- cal adaptation or genetic somaclonal variation.
Similar attempts were carried out to select tolerance to magnesium deficiency among somaclones of the very susceptible rootstock 44-53 Malegue. Tolerance was observed on somatic embryo-derived plantlets grown in vitro (Bouquet et at., I 990b ), but not recov- ered in field conditions (Bouquet, unpublished results). These disappointing results high- light the need for in vivo testing before accepting the significance of variations observed in vitro, and question the real interest of somaclonal variation in grape improvement.
Moreover, the possibility cannot be excluded that in vitro culture induces shifts toward a higher susceptibility to some diseases and pests among regenerated plants, as shown by Peros et al. (1994) on rust resistance in sugarcane. Bouquet (1989c) reported that somaclones of the rootstock Fercal (V vinifera x V Berlandieri), issued from ovule- derived somatic embryos, exhibited a high susceptibility to phylloxera after acclimatiza- tion in greenhouse and one year growth in nursery. But this susceptibility was only tran- sient, likely related to a phenomenon of rejuvenation (Martinez-Peniche, 1993).
Somaclonal variation was proposed as a way to amplifY clonal variability in Vitis vinif- era (Mullins, 1985), but its usefulness for genetic improvement of wine grapes remains to be proved. The first field trial on somaclones originating from flower cluster-derived so- matic embryos of the hybrid variety Seyval (Krul and Worley, 1977) has shown a dramatic difference between vines propagated in the standard vegetative way from cuttings and those propagated by somatic embryogenesis, particularly in their response to variations induced in berry size and sugar concentration by increased cropping levels. It was obvious that there was considerable heterogeneity within the population of the 64 experimental somaclones insofar as crop production was concerned (Mowbray et al., 1985). Mullins (1987) reported that 12 somaclones originating from ovule-derived somatic embryos of Vitis vinifera cv Cabernet Sauvignon (Mullins and Srinivasan, 1976) were highly variable in growth and cropping, but they have become more uniform with increasing age.
More than 700 somaclones belonging to nine Vilis vinifera and five rootstock varie- ties are currently under evaluation at the INRA grape breeding experiment station of Montpellier (France). Considerable variation in leaf shape was reported among 107 somaclones originating from anther-derived somatic embryos of Vitis vinifera cv Gren- ache noir (Martinez et aI., 1997). Whatever the environment, indentation of the leaves is
IN VITRO CULTURE AND PROP AGA nON OF GRAPEVINE 309 deeper in somaclones than in original clone, even 10 years after planting. But this characteristic, which is also reported in plants obtained from in vitro microcuttings could be related to a rejuvenation phenomenon (Grenan, 1992). However, variability in vigour, cropping level, cluster size, berry mass, sugar concentration and total acidity was con- siderable (Bouquet, unpublished results), and five somaclones selected for their lower berry mass are currently under field trials.
Although several investigations seemed encouraging, Alleweldt and Possingham (1988) found no substantial evidence for the accomplishment of somaclonal variation as a powerful tool in genetic improvement of grapevines. However, its interest could be more evident for characteristics linked to yield and wine quality, which are likely to be under the control of endogenous growth substances, than for characteristics linked to pest and disease tolerance. But as the former cannot be evaluated in vitro, a somaclonal selection would need field trials with hundreds if not thousands of somaclones (Grenan, 1992). In the long term, somaclonal variation could be a solution to the problem of di- minishing genetic variability in the Vitis vinifera cultivars caused by clonal and sanitary selection (Bouquet, 1989b). However, as embryogenic callus cultures are currently used for genetic transformation, there is a need to control the occurrence of somaclonal varia- tion in order to avoid confusion between its effects and the effects of introduced genes.