MOLECULAR AND CONVENTIONAL BREEDING STRATEGIES FOR SEEDLESS CITRUS

Abstract

Seedlessness is a desirable trait in Citrus. We are using conventional breeding methods and molecular techniques to generate new seedless varieties for the Australian citrus industry. Amongst the conventional breeding strategies, diploid hybridisation relies on parents that will transmit parthenocarpic ability, male/female sterility, and pollen incompatibility. Candidate genes to achieve seedless fruit using molecular techniques include those that induce fruit formation in the absence of fertilisation (parthenocarpy), and genes that directly target seed destruction after fertilisation to leave an imperceptible residue. Parthenocarpy occurs naturally in navel orange and Satsuma mandarin but it is not known how it is initiated and controlled at the molecular level. We are using Arabidopsis to study parthenocarpic fruit formation. An Arabidopsis mutant, fruit without fertilisation (fwf), has been identified that forms a seedless fruit (silique) in the absence of pollination. The ovules degenerate in parthenocarpic fwf fruit, and fruit size approaches that of a seeded silique. The fwf fruit forms all of the fruit tissues observed in seeded siliques. It shatters open at maturity, indicating that the seedless fruit matures normally. It is a recessive mutation for which the chromosomal location is known and gene cloning is in progress. Some Citrus cultivars that are normally seedy can set seedless fruit if pollination is prevented. In such cultivars, seeded fruits are often produced because insect-mediated pollination is difficult to avoid. Introduction of a male sterility gene that prevents fertile pollen formation should allow parthenocarpic fruit formation. We have introduced a male sterility gene into West Indian lime. We have also constructed genes that inhibit seed formation in Arabidopsis and tobacco soon after fertilisation and the initiation of seed development. This was done because it is not yet possible to control parthenocarpy effectively. Not all Citrus cultivars possess a parthenocarpic ability, and male-sterile plants are still capable of being cross-pollinated and thus subjected to seed set. The constructs target destruction of various seed tissues early in development, resulting in minimal seed development that leaves behind a soft trace of seed in the fruit. We have observed that in different seed types specific tissues need to be targeted in order to reduce seed size effectively. These genes are also being evaluated in West Indian lime. We are awaiting flowering and fruit set to evaluate the efficacy of their action. Improvement of Citrus in Australia using gene biotechnology began in 1992, and the first transgenic West Indian lime plants containing marker genes were regenerated in. 1996. Transgenic plant lines containing the developed genes have since been generated and individual lines are being mass propagated for use in field trials under the guidelines of the Genetic Manipulation Advisory Committee (GMAC). Our molecular work complements our hybridisation strategies for seedlessness and forms part of the Australian national Citrus breeding and evaluation program supported with funds from the Australian citrus industry via levies collected from growers.