BackgroundRed colour in kiwifruit results from the presence of anthocyanin pigments. Their expression, however, is complex, and varies among genotypes, species, tissues and environments. An understanding of the biosynthesis, physiology and genetics of the anthocyanins involved, and the control of their expression in different tissues, is required. A complex, the MBW complex, consisting of R2R3-MYB and bHLH transcription factors together with a WD-repeat protein, activates anthocyanin 3-O-galactosyltransferase (F3GT1) to produce anthocyanins. We examined the expression and genetic control of anthocyanins in flowers of Actinidia hybrid families segregating for red and white petal colour.ResultsFour inter-related backcross families between Actinidia chinensis Planch. var. chinensis and Actinidia eriantha Benth. were identified that segregated 1:1 for red or white petal colour. Flower pigments consisted of five known anthocyanins (two delphinidin-based and three cyanidin-based) and three unknowns. Intensity and hue differed in red petals from pale pink to deep magenta, and while intensity of colour increased with total concentration of anthocyanin, no association was found between any particular anthocyanin data and hue. Real time qPCR demonstrated that an R2R3 MYB, MYB110a, was expressed at significant levels in red-petalled progeny, but not in individuals with white petals.A microsatellite marker was developed that identified alleles that segregated with red petal colour, but not with ovary, stamen filament, or fruit flesh colour in these families. The marker mapped to chromosome 10 in Actinidia.The white petal phenotype was complemented by syringing Agrobacterium tumefaciens carrying Actinidia 35S::MYB110a into the petal tissue. Red pigments developed in white petals both with, and without, co-transformation with Actinidia bHLH partners. MYB110a was shown to directly activate Actinidia F3GT1 in transient assays.ConclusionsThe transcription factor, MYB110a, regulates anthocyanin production in petals in this hybrid population, but not in other flower tissues or mature fruit. The identification of delphinidin-based anthocyanins in these flowers provides candidates for colour enhancement in novel fruits.

No abstract

The current study illustrates that fruit breeding should not only target elite fruit that are significantly more liked than existing cultivars, but also target special unique fruit that create major new flavour niches. Breeding targets can be identified in terms of consumer preferences for new and defined flavours. A trained panel was used to assess the flavours of a wide range of kiwifruit, and these characteristics were systematically arranged into flavour and odour wheels. These wheels describe some of the diversity found within the kiwifruit germplasm. Next, consumers from Japan and New Zealand rated their overall liking of fruit from each of 10 genotypes. Consumer preference mapping was used to explore the relationships between consumer liking and flavour. Cluster analysis was used to explore the diverse responses consumers may have to the same fruit. Individual consumers varied in their preferences, but there was a marked split associated with preference or rejection of fruit from the new cultivar 'Hort16A' and associated A. chinensis genotypes. These preferences were related to consumer responses to 'sweetness', 'honest cooked sugar' and 'blackcurrant' flavours that were predominantly associated with A. chinensis genotypes, and absent in previous commercial kiwifruit cultivars. The first significant export of 'Hort16A' fruit occurred in 1998. Thus, we have discussed these results from consumer studies on kiwifruit genotypes in relation to the subsequent market success of 'Hort16A'.

Flowers of diploid Actinidia chinensis (kiwifruit) were hand-pollinated with pollen from either hexaploid A. deliciosa or diploid A. chinensis males and the subsequent fruit were evaluated. Following pollination with A. deliciosa pollen, fruit set, fresh weight, dry matter content, and seed weight and number were reduced. However, the most striking effect was on fruit flesh colour: the proportion of seedlings expressing red pigmentation, the intensity of pigmentation and the anthocyanin concentration were greatly reduced. The effects on maternal fruit tissues were probably indirect consequences of a reduction in the number of fertilized ovules due to partial pollen incompatibility. Effects on seed development could be explained largely by the ploidy difference between the seed and pollen parents. Growers should be cautious about using A. deliciosa pollen to pollinate diploid A. chinensis females, especially red-fleshed cultivars.

Chromosome numbers are reported for the first time for seven taxa ofActinidia: A. arguta var. purpurea, 2n = 8x = c. 232; A. deliciosa var. chlorocarpa, 2n = 6x = 174; A. deliciosa var. coloris, 2n = 6x = 174; A. glaucophylla, 2n = 2x = 58; A. guilinensis, 2n = 2x = 58; A. indochinensis, 2n = 2x = 58 and A. setosa 2n = 2x = 58. Ploidy variation has also been observed in A. melanandra and confinned in A. chinensis var. chinensis: 2n = 2x = 58 and 2n = 4x = 116. Chromosome numbers for another 11 Actinidia taxa were found to be in agreement with those previously reported. Chromosome numbers were the same for male and female plants of the same taxon. Detailed studies of chromosome morphology was not possible under the light micro- B96060Received 30 September 1996; accepted 10 March 1997 scope because of the small size of Actinidia chromosomes.