Classical breeding strategies have successfully driven the production of many desirable cultivars of fruit with improved composition, storage or eating qualities. In postharvest physiology, genetic intervention by conventional breeding has yielded pome- and stone-fruit, citrus, and a range of other subtropical species with improved storage life. Persimmon provides an extreme example of breeding for improved flavour where intense selection of genetic variants has resulted in the non-astringent variant ‘Fuyu’. Can fruit growth, maturation and postharvest physiology be modified even further for human convenience, to produce a new generation of ‘designer’ fruit?

Recent technological advances have driven ‘-omics’-type research to produce more data, more cheaply, in shorter times. Genome sequences are already available for many fruit species (including grape, apple, strawberry, papaya, tomato, pear, melon, banana, date palm and peach), and it is now feasible to obtain complete genome sequence data for individual cultivars or breeding lines, from which the sequence of important alleles can be determined. Genome sequence, together with data on gene expression (transcriptomics), the accumulation of structural proteins and enzymes (proteomics), and changes in the abundance of metabolites (metabolomics) can be integrated to provide a complete picture of ripening-associated or postharvest changes, or the metabolism that underlies a desired trait. This integrated approach is known as ‘systems biology’.

Genes or alleles of genes that have been identified as important in a particular trait can be used as molecular markers to guide targeted conventional breeding efforts. This strategy is known as marker assisted selection and is widely incorporated in breeding of many cereal crops, mainly for disease resistance. In fruit, breeding targets vary between species and between cultivars, and are aimed at improving shelf life, nutritional content, eating quality or disease resistance, or in the alleviation of particular postharvest storage disorders (see Section 11.6.5). Quantitative trait loci (QTLs) for these traits and their underpinning genes are rapidly being identified in many fruit species.

Transgenic strategies have been applied both to investigate gene function and to improve various aspects of fruit quality or production. Genetic engineering was used successfully to suppress ethylene biosynthesis and halt ripening through the silencing of either the ACS or ACO genes (Barry and Giovannoni 2007). Although biologically successful, this technology did not find a commercial use. Suppression of the cell wall-modifying enzyme polygalacturonase in a line of tomato extended fruit shelf life, and when sold to the public in the USA as ‘Flavr Savr’ became the first genetically engineered whole food to go on the market. However, consumer resistance led to its withdrawal a few years later. Transgenic modification has been very successful in papaya, a crop that in Hawaii was devastated by papaya ringspot virus. No natural resistance was available, which meant that classical breeding to combat the problem was not a possibility. A transgenic strategy was the only option, and overexpression of a transgene of the virus coat protein successfully interfered with viral replication and provided resistance (Ferreira et al. 2002). Without the development of transgenic papaya cultivars, the papaya industry in Hawaii would have disappeared. Although consumer concerns, either real or perceived, combined with the high costs of de-regulation, have restricted the use of genetic modification in fresh food crops, virus-resistant papaya provides an example of the successful use of the technology, and consumer acceptance of the resulting product.

In cases where the role of a single plant gene (e.g. encoding an important structural or regulatory protein) can be identified to control a key trait, both transgenic and non-transgenic strategies are available to modify gene functionality. One non-transgenic strategy is known as TILLING (Targeting Induced Local Lesions IN Genomes), where a population of seeds or plants is chemically mutated at random, followed by high-throughput screening to identify individuals where the target gene is affected and where a desired trait has been improved. Although large populations are required for screening, individuals with reduced functionality of the encoded protein or even knockout in any non-essential gene can usually be obtained. The technology has been used successfully in melon to identify lines with a mutated ACO1 gene and improved shelf life (Dahmani-Mardas et al. 2010).