Towards marker-assisted breeding of important physiological and fruit quality traits in papaya (Carica papaya L.)

Abstract

Papaya is one of the top five tropical fruit crops grown worldwide and an important fruit crop in Australia. An increasing demand for higher quality fruit has led to the development of new cultivars with improved physiology and flavour-related traits, together with consistency in productivity. The new commercial cultivars that comply with market requirements in terms of fruit quality would likely lead to the expansion of the Australian papaya industry. However, many of the desirable traits are complex, likely conditioned by multiple genetic components. To aid in the speed and accuracy for trait selection, molecular markers designed to the specific genomic regions that are significantly associated with the target trait characteristics are sought. The use of markers to evaluate seedling populations for fruit quality traits has the potential to greatly reduce time and the number of trees required for field planting and evaluation. Also, the identification of trait-associated genomic regions may uncover the underlying candidate genes and their functionalities for further potential manipulation and strategic selection. This research project described and demonstrated a strategy for developing molecular markers for fruit quality trait selection in papaya following the development of stable measures for elite phenotypic characters. The study was based on a F2 segregating population of a cross between an Australian cultivar ‘RB2’ and a Hawaiian cultivar ‘Sunrise Solo’. These two commercial cultivars are different in several of key fruit quality characters. ‘RB2’ has superior skin quality, thick and firm flesh and the low flesh sweetness. In contrast, ‘Sunrise Solo’ has low skin quality, flesh firmness and flesh thickness but has high sugar content. Initially, a standardised systematic, and statistically robust procedure for fruit evaluation of papaya was established to quantitatively assess the desirable fruit traits and compare them among the populations for associated genotypic (molecular) differences. The influence of harvest maturity on 14 fruit quality traits was investigated to elucidate the optimum harvest time. The fruit of each parental cultivar were harvested at five different maturity stages; mature green, three colour break stages (25%, 50%, 75%) and fully ripe, and stored at ambient temperature. Fruit evaluation was performed at 2 day intervals during the period 0-14 days after harvest (DAH). Significant differences among the fruit characters were detected (P-value < 0.001) between ‘RB2’ and ‘Sunrise Solo’ in all of fruit size characteristics assessed (fruit weight, fruit length, fruit width, cavity width and cavity length), skin characteristics (skin gloss and freckle) and flesh characteristics (flesh colour, firmness, thickness and sweetness). Significant changes in skin and flesh characteristics were also revealed within each cultivar over the five post-harvest maturity stages (P-value < 0.001). ‘RB2’ cultivar had redder flesh colour than ‘Sunrise Solo’, and developed earlier skin and flesh colouration. ‘Sunrise Solo’ cultivar had a greater proportion of skin freckle and a faster rate of decrease in fruit firmness. Furthermore, comparison of on-tree ripened to postharvest ripened fruit revealed the optimum harvest maturity to be at stage 3 (fruit with colour break of 50%). Harvest indices were derived from several of the fruit parameters and trait correlations were developed for application to standardise harvest timing. The hybridity and heritability of fruit quality traits was assessed within the breeding populations for the development of efficient selection strategies. In total, 9 ‘RB2’, 16 ‘Sunrise Solo’, 24 F1 and 226 F2 plants were included. Fruit evaluation on 14 fruit quality traits was performed following standard methods across two harvests; in December 2016 and April 2017. The result confirmed that breeding materials used in this study were true hybrids. Significant differences between the two parental lines were detected for all characteristics, except for cavity length, cavity width and consistency of flesh colour. F1 and F2 indicated similar means (at mid parent value) for most fruit quality traits. Moderate to high heritability was detected for most of the fruit quality components except consistency of flesh colour, skin gloss, skin colour and fruit cavity. High heritability (over 70%) was detected for flesh colour, fruit firmness, flesh firmness and flesh sweetness. These traits can further be exploited for improving fruit quality in papaya. Among 14 fruit quality traits, an important papaya fruit quality to preserve at point of sale is sweetness, and in particular sweetness that is derived from sucrose. This is directly regulated by a suite of genes involved in sucrose synthesis (ss) and it is unknown how the expression of these genes is affected during fruit maturation, presumably leading to differences in sweetness of the harvested fruit. To facilitate marker development for flesh sweetness, gene predictions and functional annotations of sugar-related genes were integrated by using genome sequence analysis and gene expression analysis. Eleven genes, potentially involved in sugar metabolic pathways during fruit development, were investigated along with sugar accumulation in ‘RB2’ and ‘Sunrise Solo’. The expression profiles were observed in mature leaf and eight different fruit tissues (30-120 DAA and 25%-100% colour break) using relative quantitative PCR. The expression values of each gene in leaves and fruit at different developmental stages were normalized with multiple stable reference genes (18S rRNA, CYPC and TBP2), and compared between ‘RB2’and ‘Sunrise Solo’. Both genotypes had a similar pattern of sugar accumulation with sucrose as the major sugar in ripe fruit (40-60% of total fruit sugar). During fruit development, a positive correlation was detected between fruit sugar accumulation and expression of genes cpSPS1, cpSPS2, cpCWINV1, cpAVIN2 and cpSUS3 (correlation 0.5-0.8, p-value 0.05). Expression of cpAVIN2 indicated the highest increase at full–ripe stage with a 6-fold increase in ‘RB2’ and a 10-fold increase in ‘Sunrise Solo’. Differential expression of genes was detected mainly in sucrose phosphate synthase (cpSPS1, cpSPS2, cpSPS3 and cpSPS4) and invertase (cpCWINV1and cpAVIN2) genes. The relative expression of cpSPS2 indicated it’s importance in sugar accumulation in papaya fruit, and with all tissue types of ‘Sunrise Solo’ showing greater expression of cpSPS2 than in ‘RB2’ (ranged from 0.5 to 7 fold). The maximum expression of cpSPS2 was observed in mature fruit at 120 days after anthesis. The other important sugar-related genes were cpCWINV1 and cpAVIN2, which were highly expressed in ‘Sunrise Solo’ during ripening stages. The locations of these functional genes on the papaya genome was then sought to potentially identify candidate markers to be used for the selection of high sugar-content papaya. A genetic map based QTL analysis was performed to identify important genomic regions linked to flesh sweetness and other key fruit traits. For this, whole-genome single nucleotide polymorphisms (SNP) were discovered using a genotyping-bysequencing (GBS) technique in 226 F2 individuals of cross between ‘RB2 and ‘Sunrise Solo’. The final linkage map was constructed based on 219 SNP markers and comprised 11 linkage groups covering a total distance of 509 cM. A total of 21 QTL were identified for seven key fruit quality traits including flesh sweetness, fruit weight, fruit length, fruit width skin freckle, flesh thickness and fruit firmness characters. These QTL independently explained a significant amount of the phenotypic variation (up to 19.5% each) of each trait. Eight major QTL related to flesh sweetness, fruit weight, length, width and firmness were discovered with stability across two harvest years, and candidate genes for these major QTL were further investigated. Several candidate genes, including RNA polymerase, ultraviolet-B receptor, ethylene-responsive transcription factor, pectin acetyl esterase, glutamate receptor, NAC domaincontaining protein and exopolygalacturonase which had effects for regulation of developmental growth, photosynthesis activity and carbohydrate metabolic process, were mapped within the major QTLs. Significantly associated markers to these stable major QTLs present an opportunity for the application of marker-assisted breeding. In summary, this thesis has extended knowledge on the inheritance and genetic control for key papaya physiological and fruit quality traits. Several of the putative candidate genes conditioning sweetness have also been uncovered. These, together with associated SNP markers represent a valuable resource for the future of strategic selective breeding of elite Australian papaya cultivars.