It is possible that these negative regulators induce parthenocarpy as in the down-regulation of SlARF7 ovary transcript after pollination in tomato .A gene encoding an AOS in the jasmonate biosynthesis pathway was highly down-regulated in seedless transgenic fruits. Since published data associate jasmonates with early stages in climacteric fruit ripening and triggering ethylene production , it is of interest to determine whether seedless transgenic fruits differ in the rate of fruit ripening or in shelf life. Among genes with differential expression in transgenic fruit, some highly down-regulated genes may have important functions in fruit development. Fifteen down-regulated genes were found in parthenocarpic transgenic fruits that were involved in cell wall metabolism . Two of these, GDP-mannose pyrophorylase and b-galactosidase, were highly down-regulated, and a b-1,3 glucan hydrolase was significantly up-regulated in seedless fruits. The effect of these expression changes merits further investigation, since in tomato many genes may cause fruit softening . Indeed, additional cell wall hydrolases and expansins have been associated with tomato fruit softening . Another important functional category among differentially expressed genes was minor CHO metabolism involved in fruit sugar partitioning. The metabolomic analysis found no differences in sugars, but this was not verified in ripe fruit. Many proteins in these pathways are allosterically regulated, so their activity in the fruit may be less affected by changes in transcript level. To validate these data with TaqMan real-time PCR analysis, 17 genes were analysed for correspondence between microarrays and real-time PCR . Twelve of the 17 genes showed a microarray versus real-time RT-PCR correlation of >0.75 and the remaining five genes showed a lower correlation.

These five genes were the high affinity calcium antiporter CAX1 , sugar transporter , L-lactate dehydrogenase , dutch bucket hydroponic shortchain dehydrogenase reductase , and putative vicilin . However, genes involved in auxin and ethylene biosynthesis and signalling were con- firmed to be differentially regulated between transgenic and wild-type fruits.The next sep was to address how the changes in gene expression altered the overall metabolomic profile of parthenocarpic transgenic fruit. The concentrations of >400 metabolites in parthenocarpic fruits transformed with the four different constructs were compared with those from 12 wild-type fruits containing seeds. The acquired data sets were compared using PCA to determine differences and similarities among transgenic seedless fruits and seeded wild-type fruits at the breaker stage. Linear combination of metabolic data generated new vectors or groups to best explain overall variance in the data set without prior assumptions about whether and how clusters might form. It was immediately clear that the overall metabolomic data did not show clear differences among the different fruit genotypes . PCA could not separate the four transgenic lines and the controls: the 95% confidence intervals of the four treatments and two controls overlapped. Principal component 1, which accounted for ;38% of the variance, partly distinguished the control from some parthenocarpic lines such that all negative values were from transgenic lines, but the separation was not complete. Principal component 2 did not clearly separate any treatments from the controls. The relative concentrations of >400 metabolites were determined by peak area in transgenic and control fruits. However, many of them do not have a completely determined structure and could not be identified as a known molecule. Metabolites with known structure were divided into important functional categories and they were compared with transgenic seedless and wild-type seeded fruits using ANOVA univariate analysis .

It was expected that most differences at a metabolomic level induced by the transgene expression might occur at the beginning of fruit set and before fruits reached their final size. However, some changes are also expected when fruitsreach the breaker stage. This stage is physiologically very active, crucial for the ripening process, and important for the development of fruit quality phenotypes. Among 400 compounds analysed, only 16 showed significant differences between transgenic and wild-type fruits . Three of 19 amino acids showed significant differences . INO-rolB-transformed fruits had higher concentrations of these three amino acids than other seedless and seeded fruits. Six of the 18 acids determined revealed significant differences among different fruits. INO-rolB fruits had significantly more glutamate, malate, fumarate, and ascorbate than the other transgenic and seeded wild-type fruits. Among fatty acids, DefH9- iaaM and rolB-transformed fruits had significantly more stearic acid and palmitic acid than seeded wild-type fruits. Linoleic acid was also significantly higher in all transgenic fruits than in seeded wild-type fruits. Among other metabolites, rolB fruits had more oxoproline and ethanolamine and INO-rolB fruits had more putrescine than seeded wild-type fruits. There were no significant differences in sugars among transgenic and wild-type fruits. Despite the differences in gene expression among transgenic and control fruit, PCA analysis of 400 metabolites showed that the overall metabolomic analysis did not distinguish transgenic fruit from untransformed controls . Analysis was performed in fruits at a breaker stage and it would be interesting to determine what occurs also in the ripe fruits. Since only 16 metabolites showed significant differences between transgenic and wild-type fruits, the fundamental metabolism of the fruit seemed to be mostly unchanged. However, some important metabolites were higher in parthenocarpic than in wild-type fruits, especially in INO-rolB fruits, which had the most variability among biological replicates. Although all fruits were harvested at the breaker stage, such biological variability was expected due to unavoidable small differences in fruit developmental stage.

However, it is also possible that these metabolite differences were due to ovule driven expression of rolB regulating rolB-specific fruit metabolic pathways. Fatty acids were significantly higher in DefH9-iaaM and rolB-transformed fruits than in seeded wild-type fruits. Linoleic acid was also significantly higher in all transgenic fruits than in seeded wild-type ones. These data are coincident with differences observed in transcripts related to lipid metabolism. In Arabidopsis, auxins and cytokinins induce FAD3, a desaturase gene that alters fatty acid composition . Several genes involved in auxin metabolism were differentially regulated in our transgenic seedless fruits than in wild-type fruits: some were down-regulated and some up-regulated. Although Yamamoto reported that fatty acid desaturases are auxin regulated in mung bean, auxin regulation of fatty acid biosynthesis is not fully understood.Cracking of the epidermis of harvested fruit destroys the appearance and increases the susceptibility of fruit to infections by opportunistic pathogens. Fruit with cracks are not marketable, and, therefore, have reduced economic value. Fissures of the fruit epidermis often occur prior to harvest, but can also occur after harvest, depending on storage and environmental conditions. The predisposition to form cracks has been correlated with heredity, various fruit traits and external causes, such as cultivation practices and growing environment . Many researchers have attributed cracking predisposition to the thickness of the fruit’s cuticular layer adjacent to the epidermal and sub-epidermal cells. Cracking has also been linked to the loss of flesh firmness and cell wall integrity. Fruit that are susceptible to cracking often have high levels of soluble solids and produce juice with elevated concentrations of osmotically active compounds. As fruit ripen, there is a dramatic increase in their tendency to crack. The production of large, uncracked, ripe fruit in cultivars with thin skins and high soluble solids has proven to be an unmet challenge. The complexity and structural plasticity of the ripening process are challenges for approaches designed to understand the relationship between ripening-associated softening, dutch buckets system sugar accumulation and cracking. Considerable reductions in the incidence and degree of fruit cracking may be achieved by changing cultural or post harvest practices. Consistent watering or exogenous applications of boron, calcium and/or growth promoters, such as GA3, can reduce cracking. Applications of calcium and boron strengthen the linkages between polysaccharides in the cell wall, increasing firmness. Applications of GA3 likely decrease cracking because this treatment increases the deposition of cuticular material in the epidermis and makes it more elastic. Treating plants with abscisic acid increases water movement into and promotes enlargement of the fruit. ABA treatment also increases the tendency of fruit to crack. Application of ABA to “Cabernet Sauvignon” grape berries promotes ripening and the expression of PG1 and proline-rich cell wall protein genes, typically expressed during ripening. Cracking in tomato fruit most commonly begins as they ripen.

During ripening, cell wall modifying proteins, including polygalacturonases and expansins , cooperatively disassemble wall polysaccharide networks and, thereby, contribute to the softening of fruit. Differences in cell wall structure between varieties and between unripe and ripe fruit could be an important factor in fruit tendency to crack. Quantitative and qualitative changes in the sugars in ripe fruit could influence water potential and also contribute to the tendency of the fruit to crack. Over expression of a Golden 2-like transcription factor, SlGLK2, in tomato enhances chloroplast elaboration and photosynthesis gene expression in developing fruit, and results in ripe fruit with elevated soluble solids content. It is desirable to breed or select for varieties whose fruit resist cracking under diverse environmental conditions without hormone treatments. Therefore, we investigated whether reducing the simultaneous expression of SlPG and SlEXP1 genes affects the tendency of fruit to crack. We were also interested to observe cracking of tomato lines with functional or non-functional forms of SlGLK2 to explore the contributions of solutes and sugars to the fruit’s predisposition to form cracks. ABA was used as a tool to enhance cracking incidence of the tomato fruit.A preliminary experiment was conducted in 2012, followed by a similar but more extensive experiment in 2013 with 3 genotypes. The Alisa Craig S. lycopersicum cultivar  expresses functional SlPG, SlEXP1 and SlGLK2 genes. The transgenic line, pg/exp, has suppressed ripe fruit expression of both SlPG and SlEXP1. Suppression of SlPG or SlEXP1 alone did not significantly enhance fruit firmness. However, fruits with suppressed expression of both genes were significantly firmer throughout ripening with a long-term storage and more viscous juice than control fruits. The monogenic u/u mutant of AC, “Craigella” , contains a mutation in SlGLK2 that results in a truncated and, therefore, nonfunctional glk2 protein. In the 2012 experiment, plants of the pg/exp and glk2 genotypes were grown from 15 December 2011 to 3 May 2012 in greenhouses at the University of California, Davis. Prior to germination, seeds were soaked for 3 h in water and for 30 min in a 10% solution of bleach to reduce potential viral contamination, then washed 3 times with deionized water and placed into Petri dishes with 7 mL 30 µM GA3 for 2 days at 4 °C. Subsequently, seeds were germinated in a growth chamber at 25 °C. Seedlings were transplanted and moved to the greenhouse on 16 January. There were 64 plants of each genotype subdivided into two treatments and 4 replications with 8 single plant replicates per treatment. Seedlings were grown in 9.5-L pots containing 33.3% each peat, sand, and red wood compost with 2.6 kg dolomite lime m−3 . The plants were irrigated twice per day with 350 mL of UC Davis nutrient solution containing NH4 +, NO3 − , H2PO4 − , K+ , Ca2+ , Mg2+ , SO4 2− , Fe , Mn , B , Cu , Zn and Mo . Plants were pollinated on 8 March 2012, and were topped on 15 March 2012 when they had 2 clusters of flowers. On 18 April, ABA and control spray treatments began. The plants were sprayed 1× per week for 3 weeks with a backpack applicator until the plants were completely covered with a solution containing deionized water or 0.5 mg L−1 ABA ; each solution also contained 0.5 mL L−1 polysorbate 20 as a surfactant. The cracking fruits were counted and cracking rates were calculated on 26 April. The other characteristics of the fruits were then analyzed. In 2013, WT, pg/exp and glk2 plants were grown from 17 December 2012 to 6 May 2013 in greenhouses. Seed germination protocols were like those used in 2012. Seedlings were transplanted into pots in the greenhouse on 14 January. There were 192 plants in total with 64 plants for each genotype, as in 2012. In the greenhouse, passive ventilation was used to maintain a relative humidity of 26.1–27.4%. The average temperature ranged from 21.5 to 22.7 °C with minimum of 12.8 °C and maximum of 35.0 °C. Cultivation practices were the same as in 2012, although the irrigation schedule was modified due to growth periods. Plants were initially irrigated twice per day with 350 mL of UC Davis nutrient solution.