Activation of PsMYB10.2 Transcription Causes Anthocyanin Accumulation in Flesh of the Red-Fleshed Mutant of ‘Sanyueli’ (Prunus salicina Lindl.)

Plum is one of the most important stone fruits in the world and anthocyanin-rich plums are increasingly popular due to their health-promoting potential. In this study, we investigated the mechanisms of anthocyanin accumulation in the flesh of the red-fleshed mutant of the yellow-fleshed plum ‘Sanyueli’. RNA-Seq and qRT-PCR showed that anthocyanin biosynthetic genes and the transcription factor PsMYB10.2 were upregulated in the flesh of the mutant. Functional testing in tobacco leaves indicated that PsMYB10.2 was an anthocyanin pathway activator and can activate the promoter of the anthocyanin biosynthetic genes PsUFGT and PsGST. The role of PsMYB10.2 in anthocyanin accumulation in the flesh of plum was further confirmed by virus-induced gene silencing. These results provide information for further elucidating the underlying mechanisms of anthocyanin accumulation in the flesh of plum and for the breeding of new red-fleshed plum cultivars.


INTRODUCTION
Plum is one of the most consumed stone fruits in the world and is a rich source of health-promoting compounds such as anthocyanins (Lee et al., 2009;Johnson et al., 2011;Valero et al., 2013;Fanning et al., 2014;Wang et al., 2020). Anthocyanins are responsible for the red pigmentation in plum skin and in the flesh of red-fleshed varieties (Cheng et al., 2016). It has been widely reported that the consumption of anthocyanin-rich foods promotes health (Santhakumar et al., 2015;Igwe et al., 2017;Jamar et al., 2017;Bakuradze et al., 2019). Increasing evidences have indicated that anthocyanins can delay the onset of diseases such as cancers, diabetes, cardiovascular and inflammatory diseases (He and Giusti, 2010). For example, Liu et al. (2017) reported that cyanidin, which is the major type of anthocyanin in plum fruits (Tomás-Barberán et al., 2001;Usenik et al., 2009;Wang R. et al., 2018), alleviates inflammation in vivo by blocking the binding of the cytokine interleukin-17A to the IL-17RA subunit, which suggested that cyanidin derived small molecules could be developed into drugs for the treatment of IL-17A-dependent inflammatory diseases and cancer.
China is the largest plum producer in the world (Milošević and Milošević, 2018), with a production >700 million tons in 2019 1 . While red-fleshed fruits are rich in anthocyanin and favored by consumers (Espley et al., 2013;Silvestri et al., 2018), they represent only one-fifth of the plum accessions in China (Zhang and Zhou, 1998). Most of the cultivated cultivars are red-skinned but yellow-fleshed. Therefore, it is of great interest to elucidate the mechanisms controlling anthocyanin accumulation in flesh of plum and develop new red-fleshed cultivars.
Previously, we identified a red-fleshed and late-ripening mutant (named 'Fuhongli') from the radiation-induced mutant population obtained from 60 Co-γ ray treated buds of 'Sanyueli' plum (Fang et al., 2016a). Due to its largely similar genetic origins, the mutant provides opportunities to investigate the molecular mechanisms that modulate anthocyanin accumulation in flesh of plum. This study aims to clarify the mechanisms underlying the accumulation of anthocyanins in the flesh of 'Fuhongli'. The results indicated that anthocyanin content in the flesh of 'Fuhongli' increased rapidly during fruit ripening, but no anthocyanin was accumulated in the flesh of 'Sanyueli'. RNA-Seq, dual-luciferase assays and transient color assays in tobacco leaves and plum fruits identified PsMYB10.2, which activated the promoters of PsUFGT and PsGST, as a key transcription factor contributing to anthocyanin accumulation in the flesh of 'Fuhongli'. The present study suggests that PsMYB10.2 is a weak anthocyanin activator and is responsible for anthocyanin biosynthesis in flesh of plum fruits.

Plant Materials
All fruits were harvested from 5-years old 'Sanyueli' (SY) plum (Prunus salicina Lindl.) tree and its red-fleshed mutant 'Fuhongli' (MT) grown in Fruit Research Institute, Fujian Academy of Agricultural Sciences, Fuzhou, Fujian Province, China. Fruit samples of 'Sanyueli' were collected at 95, 105, 115 days after flowering (DAF) and Fruit samples of red-fleshed mutant were collected at 95, 105, 115, and 125 DAF. Fruit maturity in 'Sanyueli' plums was reached 115 days after flowering (DAF), while 'Fuhongli' plums ripened 150 DAF. Fruits sampled from three different branches were used as biological replicates. Ten representative fruits sampled from each branch at each developmental stage were peeled and sliced into small pieces. The sliced fruit fleshes were pooled into three replicates and immediately frozen in liquid nitrogen and kept at −80 • C for further experiments.
(PT-1-Y), sucrose (ZHT-1-Y) assay kits (Comin Biotechnology Co., Ltd., Suzhou, China) according to the manufacturer's instructions. Titratable acid was measured as described by Chen and Zhu (2011). For the determination of malic acid content, 0.1 g powder was suspended with 1 ml ice-cold water and extracted by ultrasonic treatment for 1 h. The extracts were centrifugated at 8,000 × g for 10 min. The supernatant was filtrated ussing a 0.22 µm membrane filter and then analyzed with HPLC using an Agilent 1100 system (Agilent Technologies, Palo Alto, CA, United States). Separation was performed using a Kromasil C18 250 × 4.6mm, 5µm column (AkzoNobel, Sweden).

Determination of Anthocyanin and Carotenoid Content in Plum Flesh
For quantification of anthocyanin, approximately 2 g of fruit flesh was ground to fine powder in liquid nitrogen and extracted in 10 mL methanol with 0.05% HCl at 4 • C for 24 h. Anthocyanin content was quantified by pH differential method as described previously (Fang et al., 2016b). Three measurements for each biological replicate sample were performed. The total carotenoid content was measured using the carotenoid assay kit (Comin Biotechnology Co., Ltd., Suzhou, China.) as previously described (Hou et al., 2020).
For HPLC analysis, approximately 2 g of the flesh of mature fruits of 'Fuhongli' was ground in liquid nitrogen and extracted in 20 ml methanol with 1% HCl at 4 • C for 12 h in the dark. Samples were brought to room temperature then centrifuged for 10 min at 5,866 × g. HPLC analysis was performed using an Agilent 1120 system (Agilent Technologies, Palo Alto, CA). Separation was performed using a C-18 4.6 × 150 mm column (InfinityLab Poroshell 120 EC-C18, Agilent, United States). The mobile phase was 0.3% phosphoric acid in water (solvent A) and acetonitrile (solvent B) at a flow rate of 1 ml min −1 . The linear gradient of phase B was as follows: 0-20 min, 5%; 20-30 min, 15%; 30-40 min, 25%; 40-40 min, 25%; 48-50 min, 5%. The injection volume was 10 µL. Anthocyanin components were detected at 510 nm and peaks identified by comparison of peaks to authentic standards.

RNA Preparation, Library Construction and RNA-Seq
Extraction of total RNA, library construction and RNA-Seq were performed by staff at Beijing BioMarker Technologies (Beijing, China) as described previously (Fang et al., 2016b). Sequencing of the cDNA library was carried out on the Illumina HiSeq TM X Ten sequencing platform. RNA-seq was performed in three replicates. The RNA-seq reads have been deposited in the NCBI Short Read Archive and are accessible under PRJNA690967.

Analysis of RNA-Seq Data
The RNA-seq raw reads were processed with in-house perl scripts to remove reads containing adapter, reads containing ploy-N and low-quality reads. All generated clean reads were mapped to the genome sequence of 'Sanyueli' (Fang et al., 2020) using HISAT2 (Kim et al., 2015). The expression level of genes was estimated by the fragments per kilobase of transcript per million fragments mapped (FPKM) values using the StringTie software package (Pertea et al., 2015). DESeq2 (Love et al., 2014) was employed to perform differential gene expression analysis. Genes with fold change ≥2 and a false discovery rate <0.01 were considered significantly differentially expressed. One sample of 'Sanyueli' at 105DAF (SY105d-2) which was not well correlated with other biological replicates at the same stage was excluded from further analysis. GO and KEGG pathway enrichment analysis was carried out as reported previously (Fang et al., 2020). The heat maps for differentially expressed genes were constructed using TBtools (Chen et al., 2020).

Prediction of Transcription Factors
All transcripts were subjected to the PlantTFDB v5.0 2 to predict transcription factors using the Transcription Factor Prediction module.

Real-Time Quantitative RT-PCR (qRT-PCR) Analysis
The expression of 15 differentially expressed genes were validated using qRT-PCR. The extraction of total RNA from fruit flesh was carried out using a modified CTAB method (Xuan et al., 2015). All primers used for qRT-PCR were provided in Supplementary Table 1. The synthesis of cDNA was carried out using the PrimeScript R RT reagent kit with gDNA Eraser (Takara, Dalian, China). qRT-PCR was performed using the Eppendorf Realplex 4 real-time PCR system (Hamburg, Germany) as described previously (Fang et al., 2016b). qRT-PCR conditions were 5 min at 95 • C, followed by 40 cycles of 5 s at 95 • C, 15 s at 60 • C, and 30 s at 72 • C, followed by 60-95 • C melting curve detection. Actin (PsSY0026636) gene was used as the reference. The expression levels were calculated as described previously (Fang et al., 2016b). Three biological and three technical replications were performed.

Vector Construction
cDNA was generated from red pigmented flesh of 'Fuhongli' using a RevertAid First-Strand cDNA synthesis kit (Thermo Fisher Scientific, Waltham, MA). PsMYB10.2 cDNA and PsbHLH3 cDNA were isolated using 2 × Phanta Max Master Mix (Vazyme, Nanjing, China) and inserted into pSAK277. Genomic DNA was isolated from young plum leaves using a Plant Genomic DNA kit (Tiangen, China) and used as a template for promoter amplification. Promoter sequences of PsUFGT (2523bp) and PsGST (2339bp) were amplified and inserted into pGreenII 0800-LUC vector using ClonExpress Ultra One Step Cloning Kit (Vazyme, Nanjing, China). A fragment of PsMYB10.2 (296-645 bp) was amplified using specific primers (TRV2-PsMYB10.2F and pTRV2-PsMYB10.2R) and inserted into pTRV2 to generate pTRV2-PsMYB10.2. Primer sequences used for the vector construction are listed in Supplementary Table 2.

Transient Color Assay in Tobacco Leaf and Quantification of Anthocyanins
Transient overexpression assays were performed in young leaves of 2-week-old seedlings of Nicotiana tabacum grown in the greenhouse as reported by Zhou et al. (2019) using Agrobacterium tumefaciens strain GV3101. A. tumefaciens GV3101 carrying constructs were incubated at 28 • C for 2 days and resuspended in infiltration buffer containing 10 mM MgCl 2 , 100 µM acetosyringone (pH = 5.7) to an OD 600 of 0.5. Separate strains containing PsMYB10.2 and PsbHLH3 fused to the 35S promoter in the pSAK277 vector and empty pSAK277 vector were infiltrated or co-infiltrated into the abaxial leaf surface. Peach PpMYB10.1  co-infiltrated with PsbHLH3 was used as a positive control. Photographs were taken 7 days after infiltration.
For quantification of anthocyanins, 10 mg of freeze-dried tobacco leaves from the infiltrated area was mixed in 1 mL of methanol: water: formic acid (80: 19.5: 0.5, v/v/v) and shaken for 4 h at room temperature. The tube containing the mixture was centrifuged at 10,000 g for 15 min. The supernatants were filtered through a 0.45 µm PTFE syringe filter and submitted for high performance liquid chromatography (HPLC) analysis according to a method reported by Andre et al. (2012) with a few modifications. Briefly, quantification of the anthocyanins was performed using a Dionex Ultimate 3000 system (Sunnyvale, CA) equipped with a diode array detector (DAD). A 5 µL aliquot was injected onto a Thermo C18 Acclaim PolarAdvantage II column (150 × 2.1 mm i.d.; 3 µm particle size) (Waltham, MA). The mobile phases were (A) H 2 O with 5% formic acid and (B) MeCN with 0.1% formic acid. The flow rate was 0.35 mL min −1 , and the column temperature was 35 • C. The 28 min gradient was as follows: 0-5 min, 7% B constant; 5-10 min, 7-12% B; 10-20 min, 12-25% B; 20-21 min, 25-100% B; 21-24 min, 100% B constant; 24 min, 7% B; 24-28 min, 7% B re-equilibration time. Monitoring was set at 520 nm for quantification. Anthocyanins were identified by their spectral data and were quantified as cyanidin-3-glucoside using fivepoint calibration curves. A validation standard was injected after every 10th injection.

Dual-Luciferase Assay
Dual-luciferase assays were conducted in N. benthamiana leaves as reported previously (Lin- . Promoter constructs were individually transformed into A. tumefaciens GV3101 with the pSoup plasmid. Agrobacteria cultivation was performed as described above for transient color assay. For dualluciferase assays, Agrobacteria strain GV3101 were transformed with the constructs and incubated at 28 • C for 2 days and resuspended in infiltration buffer containing 10 mM MgCl2, 100 µM acetosyringone (pH = 5.7) to an OD 600 of 0.5. Two Agrobacteria suspensions carrying the overexpression vectors (empty vector, PsMYB10.2 and PsbHLH3) were mixed equally and then combined with one of the dual-luciferase promoter construct (pPsUFGT:LUC and pPsGST: LUC) suspensions in a 5:1 ratio. The mixture of Agrobacteria suspension carrying an empty vector control construct and the dual-luciferase promoter construct, again suspensded in a 5:1 ratio, was infiltrated as control. The final mixture was infiltrated into the abaxial leaf surface of N. benthamiana using a needleless 1 ml syringe. Four days after infiltration, 3-mm diameter leaf punches from infiltrated patches were harvested and subjected to dualluciferase assay.
RNAi Transient Assay of Plum Fruit Using the Virus-Induced Gene Silencing (VIGS) System pTRV1, pTRV2, and pTRV2-PsMYB10.2 were individually transformed into Agrobacterium strain GV3101. These Agrobacterium strains were suspended in infiltration buffer (10 mM MgCl 2 , 10 mM MES, pH 5.6, 150 µM acetosyringone) and kept at room temperature for 3 h. Fruit infiltration was performed using maturing fruit of 'Fuhongli' as previously described . Flesh tissues around the infiltration site were harvested and used for quantification of anthocyanin concentration and qRT-PCR analysis. Quantification of anthocyanin content was performed as described previously (Fang et al., 2016b).

Accumulation of Anthocyanin in 'Fuhongli'
The dynamic changes in content of total soluble sugar, fructose, sucrose, glucose, total acid and malic acid were analyzed in the 'Sanyueli' and 'Fuhongli' during ripening. The accumulation pattern of total soluble sugar, fructose, sucrose and glucose were similar in 'Sanyueli' and 'Fuhongli' (Figures 1A-D). The content of total acid and malic acid decreased during the stages of fruit ripening in both 'Sanyueli' and 'Fuhongli' (Figures 1E,F). The flesh of unripe 'Sanyueli' fruit was green and changed into yellow during the late stages of ripening (Figure 2A). The flesh color of 'Fuhongli' was similar to that of 'Sanyueli' during the early stages of fruit development. However, the flesh of 'Fuhongli' showed weak red pigmentation at 105DAF and rapidly turned red during the late stages of ripening (Figure 2A). Analysis of fruit flesh anthocyanin contents demonstrated that no anthocyanin was accumulated in the flesh of 'Sanyueli' (Figure 2B). In contrast, a small amount of anthocyanins was detected in the flesh of 'Fuhongli' at 105DAF, when the fruit skin was still green, and the content of anthocyanin increased significantly during the late stages of ripening ( Figure 2B). The carotenoid content increased in the flesh of 'Sanyueli' , but slightly decreased in the flesh of 'Fuhongli' (Figure 2C). According to HPLC analysis, the anthocyanins accumulated in the flesh of mature fruits of 'Fuhongli' were cyanidin-3-rutinoside and cyanidin-3glucoside ( Figure 2D).

RNA-Seq and Identification of Differentially Expressed Genes(DEGs) During Fruit Ripening of 'Sanyueli' and 'Fuhongli'
Flesh of 'Sanyueli' and 'Fuhongli' at different stages were subjected to RNA-seq analysis. A total of 163.32Gb clean data were obtained from these samples after discarding low-quality reads and adaptor sequences (Supplementary Table 3). The clean reads were mapped to the genome of 'Sanyueli' (Fang et al., 2020). The mapping ratio to the reference genome ranged between 81.29% and 85.35% (Supplementary Table 3). Correlation analysis showed that the biological replicates were highly correlated within each stage (Supplementary Figure 1). A total of 2,091 putative new genes that have not appeared in the genome of 'Sanyueli' were identified and 1,641 of them was annotated by databases including NCBI non-redundant protein database (NR), Swiss-Prot protein database (Swiss-Prot), Gene Ontology (GO), Clusters of Orthologous Groups database (COG), euKaryotic Ortholog Groups of proteins database (KOG), protein family database (Pfam) and Kyoto Encyclopedia of Genes and Genomes (KEGG) (Supplementary Table 4).
A total of 9,594 genes were identified to be differentially expressed by pairwise comparison (Supplementary Figure 2). Among them, 3,073 and 3,000 genes were differentially expressed during ripening of 'Sanyueli' and 'Fuhongli' , respectively. The comparison of transcriptome between 'Sanyueli' and 'Fuhongli' at different stages enabled the identification of 8477 DEGs (Supplementary Figure 2).

Identification of Differentially Expressed Structural Genes Involved in Anthocyanin Accumulation
To investigate in which step of anthocyanin synthesis was activated in the red flesh of 'Fuhongli' , we identified differentially expressed structural genes of anthocyanin biosynthesis according to the KEGG and GO annotation results. In total, 249 genes were assigned to phenylpropanoid biosynthesis and flavonoid biosynthesis pathway according to KEGG enrichment analysis (Supplementary Table 5). Ten of them, including PAL (PsSY0020810), C4H (PsSY0029112), 4CL (PsSY0028380), CHS (PsSY0019243 and PsSY0011918), CHI (PsSY0003309), F3H (PsSY0015860), F3 H (PsSY0012653), DFR (PsSY0000799), and ANS (PsSY0019761), were identified as DEGs (Figure 3). PsSY0019615 was annotated as CHI and encodes a protein with molecular function designated as chalcone isomerase activity (GO:0045430) and involved in anthocyanin-containing compound biosynthetic process (GO:0009718). PsSY0020267 was annotated as F3 H and predicted to have flavonoid 3 ,5hydroxylase activity (GO:0033772). PsSY0022880 was annotated as anthocyanidin 3-O-glucosyltransferase having anthocyanidin 3-O-glucosyltransferase activity (GO:0047213). In addition, PsSY0017871 was predicted to encode a GST which was shown to be upregulated in the flesh of red-fleshed cultivar 'Furongli' plum during fruit ripening in our previous study (Fang et al., 2016b). All of these anthocyanin-related genes, except 4CL, were upregulated during fruit ripening of 'Fuhongli' (Figure 3 and Supplementary Table 5). There was no significant change in the expression of PAL (PsSY0020810), C4H (PsSY0029112), CHS (PsSY0003309), F3H (PsSY0015860), F3 H (PsSY0012653), and ANS (PsSY0019761) in the flesh of 'Sanyueli' during fruit ripening, while that of CHS (PsSY0019243 and PsSY0011918), CHI (PsSY0019615), and DFR (PsSY0000799) was significantly downregulated (Figure 3 and Supplementary Table 5). It is noteworthy that the expression of C4H (PsSY0029112), CHS (PsSY0003309) and F3 H (PsSY0012653) in the flesh of 'Sanyueli' was comparable to the highest level in the flesh of 'Fuhongli' , but UFGT (PsSY0022880) and GST (PsSY0017871) were only expressed at extremely low levels in the flesh of 'Sanyueli' (Figure 3 and Supplementary Table 5). The expression profiles of 11 differentially expressed structural genes were validated using qRT-PCR. The results showed that the expression patterns detected by qRT-PCR experiments of the selected genes were consistent with the expression patterns investigated by RNA-seq analysis (Figure 4).

Identification of Transcription Factors
Since the structural genes involved in anthocyanin accumulation are transcriptionally regulated, we examined identify transcription factor potentially related to anthocyanin accumulation in the flesh of 'Fuhongli'. From this, 1687 genes were predicted to encode transcription factors and 536 of them were differentially expressed (Supplementary Table 6). A total of 29 transcription factor genes, including 19 genes (Cluster A) that were increased in the flesh of 'Fuhongli' but expressed at low level or decreased in the flesh of 'Sanyueli' and 10 genes (Cluster B) that were expressed at low level in the flesh of 'Fuhongli' but increased or expressed in higher level in the flesh of 'Sanyueli' , were identified (Figure 5 and Supplementary  Table 6). These include five MYBs and the best hit in Arabidopsis to four of them is AtMYB113 (Figure 5 and Supplementary  Table 6). PsSY0018232, which was decreased in both 'Sanyueli' and 'Fuhongli' but more abundant in 'Fuhongli' during fruit ripening, has high sequence similarity with MYB repressors such as PpMYB6. Intriguingly, the expression of a gene annotated as LOB domain-containing protein 4 (LBD4, PsSY0017735) was significantly enhanced in the flesh of 'Fuhongli' during pigmentation but only expressed at a low level in the flesh of 'Sanyueli' (Figures 4, 5 and Supplementary Table 6). In addition, many other TF types including NACs, Dof, AP2, G2-like, FAR1, and MADS were also included in the differentially expressed transcription factor sets (Figure 5 and Supplementary Table 6). The expression pattern of PsICE1 and PsYABBY4, which are lowly expressed in the flesh of 'Sanyueli' , was also confirmed by qRT-PCR (Figure 4).

PsMYB10.2 Is a Positive Regulator of Anthocyanin Biosynthesis
Nucleotide sequence alignment results demonstrated that the coding region of the four PsMYB genes (PsSY0021637, PsSY0023440, PsSY0011279, and PsSY0011777) that were upregulated in the flesh of 'Fuhongli' were nearly identical (Supplementary Figure 3). The genomic DNA sequence of PsMYB genes was also nearly identical. However, a 7,045 nt insertion was detected in the second intron of PsSY0011777 (Supplementary Figure 4). The PsSY0021637, PsSY0023440, and PsSY0011279 encode a 732 bp coding sequence and predicted protein with 243 amino acid residues while PsSY0011777 encode a 741 bp coding sequence and predicted protein with 247 amino acid residues. Phylogenetic analysis showed that all of them fell in the anthocyanin clade (SG6) and were clustered into a subgroup with previously reported anthocyanin-activating MYB proteins from other Rosacea species (Figure 6A). The two proteins most closely related to the PsMYBs were PcMYB10.2 and PpMYB10.2 from cherry plum and peach ( Figure 6A). Multiple sequence alignment showed the PsMYBs contain conserved R2R3 domain and ' ANDV' and SG6 motif for anthocyanin-promoting MYBs. They are highly homologous to PcMYB10.2 and PpMYB10.2, but PpMYB10.2 had an 18-amino acid deletion and PsMYBs (PsSY0021637 and PsSY0011279) had a 3-amino acid deletion in the C-terminal region ( Figure 6B). These results suggest that the PsMYBs are likely activators of anthocyanin biosynthesis.
The cDNA sequence of PsSY0011279 was cloned from the red flesh of 'Fuhongli' and designated as PsMYB10.2 (GenBank accession No. MK340932). The coding sequence of PsMYB10.2 is identical to the differentially expressed R2R3 MYB gene identified from red-fleshed 'Furongli' plum in our previous study (Fang et al., 2016b). PsMYB10.2 was nearly undetectable in the flesh of 'Sanyueli' but its expressed increased in the flesh of 'Fuhongli' during pigmentation (Figure 4). Tobacco leaf transient expression assays were carried out to test the function of PsMYB10.2. Coinfiltration of PsMYB10.2 and PsbHLH3 resulted in faint red coloration at infiltration sites 7 days after transformation, but no pigmentation was found when PsMYB10.2 or PsbHLH3 were infiltrated alone ( Figure 7A). Quantification of anthocyanins in leaves of N. tabacum indicated that no anthocyanin was detected in the infiltrated areas transformed with empty vector, PsMYB10.2 or PsbHLH3 (Figure 7B). Anthocyanin was detected in tobacco leaves infiltrated with PsMYB10.2 and PsbHLH3, but the content was significantly lower than that in tobacco leaves infiltrated with PpMYB10.1 and PsbHLH3 (Figure 7B). In addition, promoter structure analysis revealed there are multiple potential MYB binding sites in the promoter of PsUFGT and PsGST (Figure 7C). The promoters of PsUFGT and PsGST were isolated and fused to a luciferase reporter and dual luciferase assays were employed to investigate the interaction of PsMYB10.2 with the promoter sequences of structural genes involved in anthocyanin accumulation. Infiltration of PsMYB10.2 was able to activate the promoters of both PsUFGT and PsGST, and co-infiltration with PsbHLH3 enhanced the activity of PsMYB10.2 ( Figure 7D). These results confirmed the role of PsMYB10.2 as a positive regulator of anthocyanin accumulation.

Silencing of the PsMYB10.2 Reduces Anthocyanin Accumulation in Flesh of 'Fuhongli'
To further validate the role of the PsMYB10.2, the VIGS system was used to transiently suppress the expression of PsMYB10.2 in the flesh of 'Fuhongli'. Coinjection with pTRV1 and pTRV2-PsMYB10.2 reduced pigmentation in infiltration sites ( Figure 8A). The anthocyanin content in the flesh around the injection site of pTRV1/pTRV2-PsMYB10.2 was significantly lower than the corresponding opposite non-injected site ( Figure 8B). Anthocyanin content in the flesh tissues around the infiltration site of pTRV1/pTRV2 was reduced, but significantly higher than that of pTRV1/pTRV2-PsMYB10.2. In addition, the expression level of PsMYB10.2 and PsUFGT at the injection site of pTRV1/pTRV2-PsMYB10.2decreased by approximately 57 and 66%, respectively, compared with the injection site of pTRV1/pTRV2 (Figure 8C). These results further confirm that PsMYB10.2 is involved in the regulation of anthocyanin accumulation in flesh of 'Fuhongli'.

DISCUSSION
The association of anthocyanins with fruit quality and potential health benefits, has driven efforts to produce anthocyanin-rich FIGURE 3 | Expression of structural genes involved in the phenylpropanoid and anthocyanin biosynthesis, as evaluated by RNA-sequencing in the plum flesh of 'Sanyueli' and 'Fuhongli'. FPKM values of all genes were normalized with maximum FPKM values of each gene. The expression level of genes was indicated using filled circle with different size and color. The larger the circle, the higher the expression. High expression was indicated in red, while low expression was indicated in blue. The values on the right indicates the highest FPKM value of each gene. Abbreviations for pathway genes as described in the Introduction.
plums through breeding and postharvest treatments in recent years (Cevallos-Casals et al., 2006;Fanning et al., 2014;Santhakumar et al., 2015;Fang et al., 2016a;Wang et al., 2020). To our knowledge, 'Fuhongli' is the first red-fleshed mutant induced by γ-radiation. In this study, the mechanism involved in anthocyanin accumulation in the flesh of 'Fuhongli' was investigated using HPLC, RNA-Seq, dual-luciferase assays and transient color assays in tobacco leaves and plum fruits.
'Fuhongli' is a red-fleshed and late-ripening mutant of 'Sanyueli' plum (Fang et al., 2016a). A delayed sugar accumulation was observed in the flesh of 'Fuhongli' , together with delayed reduction of organic acid concentration through ripening (Figure 1), corroborating its late-ripening trait. Our results indicated also that anthocyanin content increased rapidly in fruit flesh of 'Fuhongli' from 115 DAF. The presence of cyanidin-3-rutinoside and cyanidin-3-glucoside, which have been identified as the main anthocyanins in Chinese plum (Fanning et al., 2014), was confirmed in 'Fuhongli' , while no anthocyanin was detected in flesh of 'Sanyueli'. Our previous study indicated that the structural genes involved in anthocyanin accumulation were upregulated during fruit ripening (Fang et al., 2016b). It is possible that 'Fuhongli' has gained the ability to accumulate anthocyanins due to the activation of anthocyanin biosynthetic genes. In this study, comparison of 'Sanyueli' and 'Fuhongli' fruit flesh transcriptomes indicated that most genes involved in biosynthesis and accumulation of anthocyanin were also upregulated in flesh of 'Fuhongli'. On the other hand, most biosynthetic enzymes were lowly expressed or decreased in flesh of 'Sanyueli' flesh during ripening. The transcripts of PsUFGT and PsGST were barely in 'Sanyueli' flesh. This suggests a direct link to the lack of anthocyanin accumulation in flesh of 'Sanyueli'.
The upregulation of structural genes suggested that anthocyanin accumulation in 'Fuhongli' flesh might be transcriptionally regulated. Several studies have suggested R2R3 MYBs as key genes for anthocyanin accumulation in Rosaceae species Rahim et al., 2014;Gu et al., 2015). Our previous study showed that PsMYB10.2 was likely the key gene involved in anthocyanin biosynthesis in flesh of 'Furongli' plum (Fang et al., 2016b). Results from the RNA-seq and qRT-PCR analyses performed in this study further indicated that transcripts encoding PsMYB10.2 were significantly upregulated in flesh of 'Fuhongli' but not expressed in flesh of 'Sanyueli'. Sequence analysis indicated that PsMYB10.2 belongs to the anthocyanin activator clade, similar to peach PpMYB10.2 (Zhou et al., 2016 and shows a high amino acid sequence homology to the cherry plum activator PcMYB10.2. Ectopic expression of PsMYB10.2 and PsbHLH3 resulted in anthocyanin accumulation and red pigmentation in tobacco leaves, but no pigmentation was observed when PsMYB10.2 was infiltrated alone. PsMYB10.2 silencing by VIGS reduced the expression of PsUFGT and the concentration of anthocyanins in the flesh of 'Fuhongli'. These results suggested that PsMYB10.2 is a key anthocyanin activator responsible for red pigmentation in the flesh of 'Fuhongli'. The upregulation of PsMYB10.2 in flesh of 'Fuhongli' suggests mutations are responsible for its transcriptional activation. For example, changes in the upstream regulators of PsMYB10.2 or a loss of repression. Consistent with this, several transcription factors that were more abundant in the flesh 'Fuhongli' or 'Sanyueli' were identified. These include the transcription factors that have been suggested to regulate the expression of anthocyanin MYB activators such as NAC  and MADS (Jaakola et al., 2010). Interestingly, our results demonstrated that the expression level of PsLBD4 was upregulated and more abundant in the flesh of 'Fuhongli'. LBD genes were first identified for their role in lateral organ development in Arabidopsis (Shuai et al., 2002) and have been reported to be involved in regulation of anthocyanin biosynthesis but they act as repressors. Rubin et al. (2009) demonstrated that AtLBD37, AtLBD38, and AtLBD39 are anthocyanin repressors. They showed that overexpression of these genes inhibited nitrogen deficiency induced anthocyanin via suppressing the expression of anthocyanin activators PAP1 and PAP2, while loss of function enhanced anthocyanin accumulation. Li et al. (2017) reported a nitrate-induced LBD gene MdLBD13, a homolog of Arabidopsis LBD37/38, repressed anthocyanin biosynthesis by reducing expression of the flavonoid pathway genes in apple. However, Zhang et al. (2019) found that CsLOB_3 and CsLBD36_2 can bind to and activate the promoter of structural genes involved in flavonoid biosynthesis. However, it remains to be investigated as to whether these transcription factors are involved in regulating the expression of PsMYB10.2 and anthocyanin biosynthesis in the flesh of 'Fuhongli'.
The MYB responsible for anthocyanin biosynthesis in peach flesh is PpMYB10.1 , while PpMYB10.2 is involved in peach flower coloration (Zhou et al., 2016). Tuan et al. (2015) showed that PpMYB10.2 was expressed in peach leaves that do not contain anthocyanin. Gu et al. (2015) suggested that PcMYB10.2 is involved in purple coloration in sepals of cv. Ziyeli. They showed that PcMYB10.2 was expressed in green-colored flesh of cherry plum, but it was undetectable in purple-colored flesh . High expression of PcMYB10.2 was also observed in both purple and green fruit skin, but its expression was much higher in green fruit skin than in purple fruit skin . These results suggest divergence of the function of MYB10.2 gene in Rosaceous species.
Our results showed that PsMYB10.2 was closely related to peach PpMYB10.2, the anthocyanin-promoting ability of which is much weaker than PpMYB10.1. Zhou et al. (2018) showed only induced faint anthocyanin production in tobacco leaves. It is noteworthy that the red coloration induced by PsMYB10.2 and PsbHLH3 was observed one a week after infiltration and The prediction of the cis-acting elements in the promoter of PsUFGT and PsGST was performed using the PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) database. Color ellipses represent predicted MYB binding sites. (D) Analysis of the interaction of the PsMYB10.2 gene with the promoters of PsUFGT and PsGST genes using dual-luciferase reporter assay Nicotiana benthamiana leaves. Error bars represent the standard errors for three replicates. Different lowercase letters indicate significant differences among treatments according to one-way ANOVA test (P < 0.05). Means with the same letter are not significantly different at the 0.05 level.
was stable, while PpMYB10.2 and PpbHLH3 induced faint coloration in tobacco leaves 2 weeks after infiltration and the faint pigmentation was only observed about one-third of the replicates . Our results demonstrated that PsMYB10.2 was weaker than PpMYB10.1 and possibly stronger than PpMYB10.2. Sequence alignment indicated that there was an 18-amino acids deletion in C-terminal of PpMYB10.2. Whether the 18-amino acids deletion in PpMYB10.2 is responsible for the difference of anthocyanin-promoting activity between PsMYB10.2 and PpMYB10.2 need further investigation. In most cases, the anthocyanin content is low (often <20 mg per 100 g fresh weight) in the flesh of red-fleshed plums and is much lower than that in the peel of red or black skin plums (Fanning et al., 2014;Zhou D. et al., 2015). Similarly low anthocyanin content was also observed in the flesh of 'Furongli' (Fang et al., 2016b) and 'Fuhongli' in this study. As PsMYB10.2 was responsible for anthocyanin accumulation in the flesh of plum (Fang et al., 2016b), the weak anthocyanin-promoting ability of PsMYB10.2 may explain the low anthocyanin concentration in flesh of redfleshed plums. An anthocyanin-rich plum cultivar Queen Garnet, which was shown to contain 107 mg per 100 g fresh weight in flesh, was developed in Queensland (Netzel et al., 2012;Fanning et al., 2014). It would be interesting to identify anthocyanin activators and elucidate underlying mechanisms responsible for anthocyanin biosynthesis in the flesh of anthocyanin-rich plum cultivars such as Queen Garnet.
Our study shows that the direct mechanism for the presence of red color in the flesh of 'Fuhongli' plum fruit is via the upregulation of the anthocyanin biosynthetic and transport pathway genes. This is mediated by the transcriptional activator PsMYB10.2. Further work will be required to establish the causation of PsMYB10.2 upregulation in the mutagenized 'Fuhongli' plum. Since this is unlikely to be due to actual changes in the coding sequences of the activating MYBs, then it may be due to changes in upstream regulators of the MYB or some loss of repression. Future studies should focus on investigating mutations that are responsible for activating the expression of PsMYB10.2 in flesh of 'Fuhongli'. Taken together, the results of our study provide useful information for the development of new red-fleshed plums.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm. nih.gov/genbank/, PRJNA690967.

AUTHOR CONTRIBUTIONS
Z-ZF and X-FY supervised the project. Z-ZF designed and performed the whole experiments and wrote the manuscript. D-RZ, Y-JL, and C-CJ participate in the experiments and data analysis. KL-W and RE provided scientific suggestion and revised the manuscript. SL-P prepared and handled plum samples. CA performed quantification of anthocyanins in tobacco leaves and revised the manuscript. All authors read and approved the final manuscript.

ACKNOWLEDGMENTS
This was a short text to acknowledge the contributions of specific colleagues, institutions, or agencies that aided the efforts of the authors.