Transcriptome and metabolome analyses reveal pathways associated with fruit color in plum (Prunus salicina Lindl.)

Background In order to reveal the mechanism of fruit color changes in plum, two common plum cultivars Changli84 (Ch84, red fruit) and Dahuangganhe (D, yellow fruit) in Northeast China were selected as plant materials. Transcriptome sequencing and metabonomic analyzing were performed at three different developmental stages: young fruit stage, colour-change stage, and maturation stage. Results “Flavonoid biosynthesis” was significantly enriched in the KEGG analysis. Some DEGs in “Flavonoid biosynthesis” pathway had an opposite trend between the two cultivars, such as CHS, DFR and FLS. Also, transcriptional control of MBW (MYB–bHLH–WD) protein complexes showed a close relationship with plum fruit color, especially the expression of MYBs and bHLHs. In the current study, procyanidin B1 and B2 had the highest level at young fruit stage in Ch84 and the content of procyanidin B2 decreased sharply at the color change stage. Conversely, the content of cyanidin increased with the growth of fruit and reached the peak at the maturation stage. Conclusion The content of procyanidin B1 and B2 in plums at young fruit stage might be the leading factors of the matured fruit color. At the maturation stage, the cyanidin produced by procyanidins keeps the color of the fruit red. Correspondingly, genes in “flavonoid biosynthesis” pathway play critical roles in regulating the accumulation of anthocyanin in plum.


Introduction
Plums (Prunus salicina Lindl.) is a kind of favorite fruit product by consumers for their delicious tastes. Plum have a wide variety of uses and consumers typically prefer to eat fresh plums for their characteristic taste and rich nutrient substance [1]. There are many varieties of plums, which have different characteristics, such as maturity period, different taste, different fruit color and so on. In this study, two plum cultivars with different fruit color were selected to study the molecular mechanism that related to fruit color formation.
Concentration of anthocyanins determines the color of fruits [2,3]. Anthocyanins are responsible for the colors of numerous flowers, fruits, vegetables and even cereals [4], and can strongly contribute to food quality and appeal to consumers. Anthocyanins  The molecular mechanism of anthocyanin and proanthocyanidin biosynthesis in plum is inadequate.
Therefore, more research is needed to reveal it.
In this study, we selected two common plum cultivars Changli84 (Ch84) and Dahuangganhe (D) in Northeast China. The fruit color of Ch84 is red, while D is yellow. In order to reveal the mechanism of fruit color changes between the two cultivars, transcriptome sequencing and metabonomic analyzing were performed at three different developmental stages: young fruit stage, colour-change stage, and maturation stage. We hypothesize that anthocyanin synthesis related genes and anthocyanin metabolites play an important role in fruit color formation. This study might provide evidence for the correlation between anthocyanins and plum fruit color.

Plant materials
All plum fruit samples were collected from the experimental farm of the Experimental farm of Jilin Academy of Agricultural Sciences, Gongzhuling City, Jilin Province. Two varieties of plum trees, Changli 84 (Ch84) and Dahuangganhe (D), were used in this study. All plum trees (six years old) grow under natural conditions, using conventional irrigation and fertilization strategies. The climate of the experimental area is temperate and monsoonal with a mean annual temperature of 5.6°C. The average annual precipitation is 594.8 mm, of which 80% falls from May to September. The fruit tissues were collected at young fruit stage (abbreviated as Y, 18 th May, 30 days after anthesis), colour-change stage (abbreviated as C, 17 th Jun, 60 days after anthesis) and maturation stage (abbreviated as D, 17 th July, 90 days after anthesis). Three individuals were selected for each variety, and three fruits were randomly selected from one individual for each sampling stage, and then pooled as a repeat. A total of 18 samples were collected (two varieties, three stages, three repetitions / stage / variety). For molecular analysis, tissue samples were directly snap-frozen in liquid nitrogen and kept at −80 °C.

RNA isolation and sequencing
Total RNA extraction, library construction and RNA-Seq were performed by Genedenovo Biotechnology Corporation (Guangdong, China). Briefly, total RNA of these 18 samples were

LC-MS and LC-MS/MS Analysis
In order to study the difference of metabolomics between the two cultivars at different developmental stages, untargeted metabolomics was carried out using a previously described   Table S1.

Metabolomic data analysis
To produce a matrix containing fewer biased and redundant data, peaks were filtered to remove the redundant signals caused by different isotopes, in-source fragmentation, K+, Na+, and model was applied to rank the metabolites that best distinguished between two groups. The threshold of VIP was set to 1. In addition, T-test was also used as a univariate analysis for screening differential metabolites. Those with a P value of T test ＜0.05 and VIP ≥ 1 were considered differential metabolites between two groups. Then, metabolites were mapped to KEGG metabolic pathways for pathway analysis and enrichment analysis.

Fruit performance on two plum varieties at the developmental stage.
After 90 days of growth, plums gradually matured. There were obvious differences between the two varieties. Ch84 had bigger fruit size than D. The fruit flesh color of ch84 was red and D was yellow ( Figure 1A). The fruit weight, longitudinal diameter and transverse diameter of Ch84 were significantly higher than those of D (Figure 1B to D). The most striking difference is the color of the fruit. repeat samples was small.

Transcripts annotation
Due to lacking of a reference genome in plum, all the assembled transcripts were blasted against six public databases (Nr, Nt, SwissProt, KOG, Pfam, GO and KEGG) using search tools.

Analysis of differentially expressed genes (DEGs) between the two plum varieties
According to false discovery rate (FDR) < 0.05 and |log2 foldchange| ≥ 2, DEGs were identified using the DESeq software package. As shown in Figure 2A Figure   3C.

Figure 2 Statistics of differentially expressed genes (DEGs)
A shows the number of DEGs between the two plum varieties at different developmental stage. B shows the Venn analysis results of different comparisons.

Figure 3 The GO and KEGG enrichment results of DEGs and the expression of candidate DEGs
A and B shows the GO and KEGG enrichment results of all the DEGs, respectivly. The redder the color, the higher the significance. C shows the expression of candidate DEGs of the two plum at different developmental stage.

Transcriptional control of MBW (MYB-bHLH-WD) protein complexes
In order to further explore the DEGs that related to fruit color, we identified the gene expression level of MBW protein complexes. We found that 6 MYB related DEGs were highly expressed in Ch84 but lowly in D. Nine bHLHs had a significantly higher expression in Ch84 than D.
However, there was no WD40 differently expressed between Ch84 and D. By mapping the gene expression heat map of MBW complex, MYB and bHLH may be the key genes regulating anthocyanin synthesis in plum (Figure 4).

Figure 4 Expression of MBW protein complexes
A simplified model depicting the transcription factors of MYB, bHLH and WD40. The redder the color of the box, the higher the expression, and the bluer the color, the lower the expression.

Analysis of differential metabolites (DMs) between the two plum varieties
To distinguish the classes and to assess the global metabolism variations, the unsupervised PCA was performed in both positive and negative spectra. The PCA score plot of positive spectra indicated that there was a clear classification of observations of Ch84 and D ( Figure 5A). To further distinguish the Ch84 and D and to identify differential variants, a supervised OPLS-DA was conducted. In Figure 5B, C and D, a remarkable separation of LC-MS data in Ch84 and D groups at different developmental stages was observed in the OPLS-DA score plot, indicating that this OPLS-DA model was non-overfitting. The results of these assessments mentioned above suggested that the LC-MS data quality was reliable. Fifty-four differential metabolites were identified and the heatmap of the differential metabolites were shown in Figure 6A. To uncover the most relevant biological pathways of osimertinib resistance, the KEGG enrichment analysis was performed. The enrichment and topology analysis demonstrated that main difference of metabolites between the two varieties was related to "Cysteine and methionine metabolism" (ko00270), "Biosynthesis of amino acids" (ko01230) and "Aminoacyl-tRNA biosynthesis" (ko00970) (Figure 6B). Interestingly, we found significant differences in the content of several anthocyanin related metabolites between the two cultivars, including Cyanidin 3-glucoside, Cyanidin-3-O-alpha-arabinopyranoside, Procyanidin B2 and Procyanidin B1. The level of these mentioned metabolites were shown in Figure 6C.

Figure 5 Summary of the metabolomic data
A shows the PCA results, different colors represent different groups. B to D shows the OPLS-DA analysis results between the two plum varieties at the young fruit stage, color changing stage and fruit mature stage.

Figure 6 Differential metabolites and functional enrichment analysis
A shows the level of differential metabolites. The redder the color, the higher the level. B shows the KEGG enrichment results based on these differential metabolites. C shows the level of candidate differential metabolites of the two plum at different developmental stage.

Discussion
There are obvious differences in fruit color between the two plum varieties. Finding out the difference can help us understand the mechanism of fruit color. As shown in Figure 1, The fruit color of Ch84 is dark red, while that of D is yellow. The main reason for this difference may be caused by different anthocyanin content in the plum fruit. The transcriptome analysis results showed that " Flavonoid biosynthesis" was significantly enriched in the KEGG analysis.
Flavonoid biosynthesis is one of the most extensively studied secondary metabolic pathways in The results of metabolomics also give us a lot of hints about the mechanism of plum fruit color.
Polyphenols including yellow flavonoids, procyanidins (B1 and B2) and cyanidin-3-Oglucoside in substantial amounts have been characterized in different palm fruits [42]. In our results, procyanidin B1 and B2 had the highest level at young fruit stage in Ch84 and the content of procyanidin B2 decreased sharply at the color change stage. Conversely, the content of cyanidin increased with the growth of fruit and reached the peak at the maturation stage. While for D, the metabolites mentioned above did not change significantly at all developmental stages.
As we know, procyanidins are members of the proanthocyanidin (or condensed tannins) class of flavonoids. They are oligomeric compounds, formed from catechin and epicatechin molecules. They yield cyanidin when depolymerized under oxidative conditions [43].
Therefore, we speculated that the content of polyphenols, like procyanidin B1 and B2, in plums at young fruit stage might be the leading factors of the matured fruit color.

Conclusion
Anthocyanins content is the main factor for the difference of fruit color between the two varieties. The content of procyanidin B1 and B2 in plums at young fruit stage might be the leading factors of the matured fruit color. At the maturation stage, the cyanidin produced by procyanidins keeps the color of the fruit red. Correspondingly, genes in "flavonoid biosynthesis" pathway are active. DEGs like CHS, CHI3, DFR had higher expression level in Ch84 than those in D, and these genes regulated the accumulation of anthocyanin in plum. Also, we speculated that the expression levels of MYBs and bHLHs might dominate the color change of plum fruit, while the role of WD40 is limited.

Data Availability Statement
The transcriptional datasets analyzed during the current study are available in the National Center for Biotechnology information (NCBI) repository, and the BioProject accession number was PRJNA726302(https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA726302). The Prunus salicina transcriptome shotgun assembly (TSA) project has the project accession GJEK00000000.1. This version of the project (01) has the accession number GJEK00000000.1, and consists of sequences GJEK01000001-GJEK01035577.