Investigating phenotypic relationships in persimmon accessions through integrated proteomic and metabolomic analysis of corresponding fruits

Abstract

Together with phenological and genomic approaches, gel-based and label-free proteomic as well metabolomic procedures were separately applied to plants to highlight differences between ecotypes, to estimate genetic variability within/between organism populations, or to characterize specific mutants/genetically modified lines at metabolic level. To investigate the possible use of tandem mass tag (TMT)-based quantitative proteomics in the above-mentioned contexts and based on the absence of combined proteo-metabolomic studies on Diospyros kaki cultivars, we here applied integrated proteomic and metabolomic approaches to fruits from Italian persimmon ecotypes with the aim to characterize plant phenotypic diversity at molecular level. We identified 2255 proteins in fruits, assigning 102 differentially represented components between cultivars, including some related to pomological, nutritional and allergenic characteristics. Thirty-three polyphenols were also identified and quantified, which belong to hydroxybenzoic acid, flavanol, hydroxycinnamic acid, flavonol, flavanone and dihydrochalcone sub-classes. Heat-map representation of quantitative proteomic and metabolomic results highlighted compound representation differences in various accessions, whose elaboration through Euclidean distance functions and other linkage methods defined dendrograms establishing phenotypic relationships between cultivars. Principal component analysis of proteomic and metabolomic data provided clear information on phenotypic differences/similarities between persimmon accessions. Coherent cultivar association results were observed between proteomic and metabolomic data, emphasizing the utility of integrating combined omic approaches to identify and validate phenotypic relationships between ecotypes, and to estimate corresponding variability and distance. Accordingly, this study describes an original, combined approach to outline phenotypic signatures in persimmon cultivars, which may be used for a further characterization of other ecotypes of the same species and an improved description of nutritional characteristics of corresponding fruits.

Keywords: biodiversity; fruit; metabolomics; persimmon (Diospyros kaki Thunb.); phenotypic relationships; proteomics.

Figure 1

Morphological characteristics of persimmon fruits of Kaki tipo, Cioccolatino, Vaniglia and Lampadina cultivars. (A) Kaki tipo; (B) Cioccolatino; (C) Vaniglia; (D) Lampadina. Bars correspond to a 1 cm-length.

Figure 2

Heat-map showing the variable abundance of differentially represented proteins in persimmon fruits of Kaki tipo, Cioccolatino, Vaniglia and Lampadina cultivars. Each column corresponds to a cultivar, whereas each row represents a single protein. Increasing brightness towards red indicates higher protein responses (measured as summed peak areas) and green indicates lower protein responses. Protein abundance values were grouped for three replicates and scaled before clustering. The dendrograms from the unsupervised hierarchical cluster analysis of the columns and the rows (using the Euclidean distance function and average linkage method) illustrate the similarity of the cultivars and proteins. Protein accession codes reported on the right side of the figure (102 in number) are vertically ordered according to figure appearance. Detailed information on differentially represented proteins reported in this figure are described in Supplementary Table S3 .

Figure 3

Functional distribution of differentially represented proteins in persimmon fruits of Kaki tipo, Cioccolatino, Vaniglia and Lampadina cultivars. Identified protein species were initially assigned with Mercator software, followed by a functional group cataloguing including information from recent literature data.

Figure 4

Heat-map reporting the variable abundance of differentially represented proteins involved in response to external stimuli and allergenicity in persimmon fruits of Kaki tipo, Cioccolatino, Vaniglia and Lampadina cultivars. Each column corresponds to a cultivar, whereas each row represents a single protein. Increasing brightness towards red indicates higher protein responses (measured as summed peak areas) and green indicates lower protein responses. Protein abundance values were grouped for three replicates and scaled before clustering. Known allergenic components present in persimmon and other fruits (https://www.allergome.org/) included isoflavone reductase (dlo_pri0042f.1_g01950.1), some thaumatin-like proteins (dlo_pri0360f.1_g00050.1, dlo_pri0037f.1_g02010.1 and dlo_pri0037f.1_g01930.1), some pathogenesis-related proteins (dlo_pri0138f.1_g01030.1 and dlo_pri0112f.1_g00900.1), some endochitinases (dlo_pri0469f.1_g00680.1, dlo_pri0469f.1_g00500.1, dlo_pri0580f.1_g00070.1, and dlo_alt0990f-001-01.1_g00020.1), glyceraldehyde 3-phosphate dehydrogenase (dlo_pri0045f.1_g00320.1), Bet v1-like protein (dlo_pri0004f.1_g05100.1), profilin (dlo_pri0284f.1_g00530) and some chitinases (dlo_pri0990f.1_g00080.1 and dlo_pri2008f.1_g00030.1). Detailed information on the remaining differentially represented proteins reported in this figure are described in Supplementary Table S3 . Asterisk indicates allergenic proteins whose differential representation was assigned based on the significance of change in at least one comparison (p-value ≤ 0.01).

Figure 5

PCA score plot of differentially represented proteins in persimmon fruits of Kaki tipo, Cioccolatino, Vaniglia and Lampadina cultivars. This plot demonstrates the separation of the protein level data (normalized abundances) and the genotype dependent relationship. Each point in the chart describes a biological replicate. PCA explained about 64.8% of variability among persimmon cultivars.

Figure 6

PCA score plot of polyphenols in persimmon fruits of Kaki tipo, Cioccolatino, Vaniglia and Lampadina cultivars. PCA explained about 88.2% of variability among the four cultivars, considering the first two components on x and y axes. Spatial 2D distribution of samples was achieved by considering loading plots of secondary metabolites sub-classes.

Figure 7

Heat-map reporting the variable abundance of polyphenols in persimmon fruits of Kaki tipo, Cioccolatino, Vaniglia and Lampadina cultivars. Each column corresponds to a cultivar, whereas each row represents a compound. Heat-map encompassed centered and reduced intensities (area counts) moving from red to blue, through white, including Euclidean distance function and Ward linkage method. The dendrograms from the hierarchical cluster analysis of the columns and the rows illustrate the similarity of the cultivars and the distribution of polyphenolic compounds between aglycones and glycosides (top and bottom section), respectively.