Horticulture Research, 2022, 9: uhac206 https://doi.org/10.1093/hr/uhac206 Article An LTR retrotransposon in the promoter of a PsMYB10.2 gene associated with the regulation of fruit f lesh color in Japanese plum Arnau Fiol1 , Sergio García1 , Christian Dujak1 , Igor Pacheco2 , Rodrigo Infante3 and Maria José Aranzana 1,4, * 1 Centre for Research in Agricultural Genomics (CRAG) CSIC-IRTA-UAB-UB, Campus UAB, Bellaterra, 08193 Barcelona, Spain 2 Institutode Nutrición y Tecnología de Alimentos (INTA), Universidad de Chile, El Líbano 5524, Santiago, Chile 3 Departamento de Producción Agrícola, Facultad de Ciencias Agronómicas, Universidad de Chile, Santa Rosa 11315, Santiago, Chile 4 IRTA (Institut de Recerca i Tecnologia Agroalimentàries), Barcelona, Spain *Corresponding author: E-mail: mariajose.aranzana@irta.cat Abstract Japanese plums exhibit wide diversity of fruit coloration. The red to black hues are caused by the accumulation of anthocyanins, while their absence results in yellow, orange or green fruits. In Prunus, MYB10 genes are determinants for anthocyanin accumulation. In peach, QTLs for red plant organ traits map in an LG3 region with three MYB10 copies (PpMYB10.1, PpMYB10.2 and PpMYB10.3). In Japanese plum the gene copy number in this region differs with respect to peach: there are at least three copies of PsMYB10.1, with the expression of one of them (PsMYB10.1a) correlating with fruit skin color. The objective of this study was to determine a possible role of LG3-PsMYB10 genes in the natural variability of the f lesh color trait and to develop a molecular marker for marker-assisted selection (MAS). We explored the variability within the LG3-PsMYB10 region using long-range sequences obtained in previous studies through CRISPR-Cas9 enrichment sequencing. We found that the PsMYB10.2 gene was only expressed in red f lesh fruits. Its role in promoting anthocyanin biosynthesis was validated by transient overexpression in Japanese plum fruits. The analysis of long-range sequences identified an LTR retrotransposon in the promoter of the expressed PsMYB10.2 gene that explained the trait in 93.1% of the 145 individuals analyzed. We hypothesize that the LTR retrotransposon may promote the PsMYB10.2 expression and activate the anthocyanin biosynthesis pathway. We propose for the first time the PsMYB10.2 gene as candidate for the f lesh color natural variation in Japanese plum and provide a molecular marker for MAS. Introduction cyanin levels in the fruit [4, 13]. R2R3-MYB proteins regulate anthocyanin biosynthesis through interaction with a basic helix– Anthocyanins are plant pigments synthesized from the phenyl- loop–helix (bHLH) transcription factor and a WD-repeat protein, propanoid pathway and they confer red, purple or almost black which results in the MBW complex that can bind to the pro- coloration based on their structural conformation and accumula- moters of several anthocyanin biosynthesis structural genes and tion levels [1]. These f lavonoids protect plants against biotic and enhance their expression [14–16]. Other R2R3-MYBs downregulate abiotic stress, and also enhance animal-mediated f lower pollina- anthocyanin biosynthesis by diverging to other branches of the tion and seed dispersion [2]. Some edible fruits are a good source of phenylpropanoid pathway or by repressing structural genes [17– anthocyanins and their color is considered a fruit quality parame- 19]. However, the MYB10 genes remain the main factors leading to ter, not only because of the visual attractiveness to the consumer, but also because anthocyanin ingestion has been correlated to red-to-blue fruit coloration, and DNA polymorphisms with direct several beneficial effects on human health [3]. The Rosaceae fam- impact on its expression or protein function have been identified ily includes species with edible fruits such as apples, pears, straw- and correlated with fruit color variation in strawberry [5], apple berries, peaches, cherries, apricots and plums, among others. [20, 21], pear [7, 22], peach [6, 23], sweet cherry [8, 24] and Japanese Some of them have tremendous intraspecific fruit color variabil- plum [25]. Transposons have also been reported to play a role ity, with anthocyanins conferring color to the whole fruit or single in this variability [5, 26]. In addition, anthocyanin levels can be tissues, or not contributing to the final coloration at all [4–10]. modulated by ethylene production [27] and environmental factors The R2R3-MYB genes are among the largest families of tran- such as light and temperature [28–30], adding an extra layer of scription factor in plants. They mostly regulate plant-specific complexity to their regulation. processes, including anthocyanin biosynthesis by acting as tran- Anthocyanins are synthesized in many plant tissues, constitu- scriptional activators as well as repressors [11, 12]. In rosaceous tively or under certain conditions. In the Prunus genome, several species, R2R3-MYB10 genes have been found to regulate antho- QTLs for fruit color have been mapped in its eight linkage groups Received: 26 January 2022; Accepted: 8 September 2022; Published: 13 September 2022; Corrected and Typeset: 1 December 2022 © The Author(s) 2022. Published by Oxford University Press on behalf of Nanjing Agricultural University. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 2 | Horticulture Research, 2022, 9: uhac206 (LG), some of them in agreement with the position of a cluster al. [45], we identified a copia-like LTR retrotransposon inserted of MYB10 genes in LG3 [31–33]. In peach, this cluster contains in the promoter of the PsMYB10.2 expressed genes only. The three MYB10 genes (MYB10.1, MYB10.2 and MYB10.3) [34], while an presence/absence of the LTR retrotransposon explained 93.1% increased number of MYBs have been annotated in the syntenic of the red f lesh color trait acquired during the last stage of region of the sweet cherry, apricot and Japanese plum genomes ripening, making it an efficient molecular marker for MAS. This [35–37]. molecular marker, together with that for skin color, will greatly The multiple loci mapped in the Prunus genome for antho- assist the early selection for fruit color in Japanese plum breeding cyanin accumulation and the complexity of the phenotype high- programs. Moreover, we describe a protocol for transiently light the complex regulation of synthesis of this pigment, in which overexpressing genes in Japanese plum fruits, which to our several genes, differing in their spatial and temporal expression knowledge has not yet been reported in this species and will assist patterns, promote or repress the phenylpropanoid pathway. In validating the function of other candidate genes for fruit quality peach, the recessive blood f lesh (bf ), mapped in LG4, promotes the in Prunus. accumulation of anthocyanins in the f lesh after the pit hardening stage, i.e. long before fruit maturity [38], while the dominant blood f lesh (DBF), mapped in LG5, controls the blood-f lesh trait Results at the final stages of fruit maturation [39]. The f lesh color around The MYB10-H2 haplotype explains most of the the stone (Cs) mapped in LG3 [40] explains a localized pigment flesh color variability accumulation, unlike the highlighter (h) in the same LG that We studied the correlation between the LG3-MYB10 haplotypes suppresses anthocyanin accumulation in the whole fruit [23, 41]. described by Fiol et al. [25] and fruit f lesh color in a sample of Some of these traits, such as DBF, are regulated through the acti- 103 Japanese plum selections (Figure 1) (Supplementary Data 1). vation of MYB10 gene expression [9], while others are determined Using a χ 2 test, we found haplotype H2 highly correlated with red by polymorphisms in the MYB10 genes themselves [6, 23, 42]. f lesh (p-value = 8.06 × 10−12 ) (Supplementary data Table 1). Forty- Japanese plum (Prunus salicina L.) is one of the rosaceous crops one produced fruits with red f lesh, 33 of them (80.5%) with H2 in with more fruit color variability. Fiol et al. [25] studied the allelic either homozygosis or heterozygosis. Haplotype H2 was present in variability in the LG3-PsMYB10 region. They found a triplication seven non-red selections (11.29% in the non-red group), including of the PsMYB10.1 gene in some of the varieties and identified one that was homozygous (C57, H2/H2). Five of them had an infre- an allele in one of the gene copies highly associated with the quent allele (a467) not observed in the H2 of red f lesh selections skin anthocyanin color. In addition, they characterized the vari- (see Fiol et al. [25] for more details), indicating that there may be ability into haplotypes, providing an efficient molecular marker two different H2 haplotypes. for marker-assisted selection (MAS) of skin color in breeding programs. However, they found no allele significantly associated The PsMYB10.2 gene is expressed in the red with the red f lesh color. To our knowledge, only two studies have flesh found polymorphisms, mapped in LG6 and LG3, associated with The haplotype H2 defined in Fiol et al. [25] was characterized the fruit f lesh color in Japanese plum [43, 44]. The number of by three PsMYB10 alleles: a470 homologous to PsMYB10.1, a466 individuals used in both studies was reduced, which limits the homologous to PsMYB10.2, and a492 homologous to PsMYB10.3. transfer of the markers to broader and more variable collections None of these alleles were exclusive to H2. We tested their expres- and motivates the search for markers more tightly linked to the sion in the f lesh of mature fruits of five varieties with different trait. f lesh color and haplotype combinations. Only the PsMYB10.2 The well-known role of MYB10 genes in rosaceous fruit color allele (a466) was expressed in the f lesh, and only in the red- variation, the weak association found between LG3 makers and f leshed fruits (Figure 2). The PsMYB10.2 sequence amplified from f lesh color in Japanese plum and the diverse number of alleles mature fruit f lesh cDNA had an approximate length of 800 bp, and gene copies identified by Fiol et al. [25] motivated further while the same primers amplified a fragment of around 2 kb in exploration of the genetic variability of the LG3-PsMYB10 region. genomic DNA. The expression of the MON gene was used as an For that, Fiol et al. [45] used CRISPR-Cas9 target enrichment internal control, which amplified in each of the tested cDNAs sequencing in a pool of Japanese plum varieties. This confirmed (Supplementary data Figure 1). the high level of variability in the MYB10 region between Japanese plum varieties compared to other Prunus genomes, and they found The predicted PsMYB10.2 protein sequence has that the level of homology between Japanese plum varieties was low variability comparable to that between Prunus species and their wild rel- The RT-PCR bands, as well as the PsMYB10.2 bands from atives. The CRISPR-Cas9 target sequencing approach yielded an genomic DNA of H1 to H4, were sequenced from C3, C31, C6 excellent source of polymorphisms (SNPs and SVs) that can be and C12, respectively. There was a complete match of the RT- used to find additional variants associated with fruit color, espe- PCR sequences with the exon sequences of the PsMYB10.2-H2 cially in the f lesh. allele. The alignment of the predicted protein sequences of the The objective of this study was to explore the LG3-PsMYB10 PsMYB10.2 alleles in the haplotypes H1 to H4 as well as that pre- variability in the search for polymorphisms associated with the dicted from the NCBI PsMYB10.2 GenBank sequence EU155161.1 red f lesh trait and to design a molecular marker for marker- revealed variation in some amino acids (Figure 3). While the assisted selection (MAS) in Japanese plum. With the MYB10 PsMYB10.2 alleles in the H2 haplotpye (PsMYB10.2-H2 allele) and molecular marker provided in Fiol et al. [25], we found the in the H4 haplotype (PsMYB10.2-H4 allele) coded for proteins association of one haplotype (H2) whose PsMYB10.2 gene was with 243 amino acids; the PsMYB10.2-H1 and H3 alleles had three expressed only in the mature f lesh of red fruits. We validated the additional amino acids in their C-terminal sequence. The identity role of PsMYB10.2 in the activation of anthocyanin biosynthesis of all the predicted proteins was over 95%, the proteins of H2 by its transient expression in yellow fruits. After exploring the and H3 being the closest, with 99.18% identity, while H4 was de novo sequences from the CRISPR-Cas9 experiment in Fiol et slightly more distant to all the others, being 95.06% homologous Fiol et al. | 3 Figure 1. Graphic representation of haplotype combinations (genotypes) and f lesh color (with or without anthocyanin-related coloration) in 103 Japanese plum selections. Red or yellow sections indicate the fruit f lesh color with the given haplotype combination (ex. H1/H2). Their frequency is indicated in the section. Section size is proportional to the number of individuals. Ten of the haplotype combinations were found in both red and yellow fruits (∗ ). Haplotype H2 was associated with the red f lesh color in 84.47% of the tested genotypes; individuals with phenotype not explained by H2 (15.53%) (i.e. have H2 and are yellow or do not have H2 and are red) are displaced out of the circle. Figure 2. Expression of the PsMYB10.2 gene in the mature f lesh of five Japanese plum varieties (three fruits per variety). H1, H2, H3, H4, H5 and H9 are LG3-MYB10 haplotypes defined in Fiol et al. [25]. Well L: 50 bp ready-to-use DNA Ladder (GeneON); well C: no template control reaction. with H1 and 95.88% with H2 and H3 (Supplementary Data 2). acid polymorphisms between the PsMYB10.2 proteins occurred at The amino acid sequences were also compared with the peach these positions. PpMYB10.2 translated protein (Prupe.3G163000) (Figure 3), where the functionality is significantly reduced due to the simultaneous The transient expression of PsMYB10.2 promotes occurrence of a lysine (K) and an arginine (R) at positions 63 and anthocyanin biosynthesis in the flesh of 90, respectively [46]. In these positions, all the PsMYB10.2 proteins Japanese plum fruits were monomorphic, with an arginine (R) in both positions. Motifs To determine whether the expression of PsMYB10.2 promoted and other key amino acids reported by Stracke et al. [12], Lin-Wang anthocyanin biosynthesis, the genomic sequence obtained from a et al. [13] and Yang et al. [47] were identified; none of the amino red H2/H2 genotype was transiently overexpressed in yellow f lesh 4 | Horticulture Research, 2022, 9: uhac206 Figure 3. Alignment of the in silico proteins of the PsMYB10.2 genes with haplotypes H1 to H4, the Japanese plum PsMYB10.2 from the NCBI database (EU155161.1) and the peach MYB10.2 (PpMYB10.2) from the Peach Genome v2 [48]. The conserved R2 and R3 domains are shown. The motifs from orthologous MYBs (red rectangles) and the conserved residues from anthocyanin-promoting R2R3-MYBs from the Rosaceae family (shadowed residues) were monomorphic in the four haplotypes. Red arrows indicate key function residues reported by (∗ ) Zhou et al. [46] and (∗∗ ) Yang et al. [47]. Japanese plum fruits. We did the same with the PsMYB10.2-H4 sequence against the “Sanyueli” and “Zhongli No.6” reference sequence, which coded for the most distant PsMYB10.2 protein genomes [37,49] resulted in hits distributed along the eight among those cloned. The fruits agroinfiltrated with either the chromosomes, none of them in the LG3-MYB10 region. Further PsMYB10.2-H2 or the PsMYB10.2-H4 sequences had visible red search in the Rosaceae plants transposable elements database stains on the yellow f lesh in the region surrounding the injec- (RPTEdb) identified homology (E-value = 0.0) with two Prunus tion (Figure 4), indicating that both coded proteins were func- mume Ty1-copia LTR retrotransposons: RLCopia_1_485_pmu and tional. These red stains were not observed in other areas, in RLCopia_1_510_pmu. With the LTR_FINDER software, conserved fruits injected with the negative control vector (pBI121), nor in sequences of long terminal repeat (LTR) retrotransposons were noninjected fruits. The red stains in the f lesh corresponded to a identified in the negative strand: the primer-binding site (PBS), 20 to 103 fold increase of anthocyanin levels compared with the the polypurine tract (PPT), the two LTR regions (762 bp and 772 bp fruits injected with the control vector. long, with 97.4% sequence identity), and the conserved 5 TG and 3’ CA nucleotide pairs on them. The LTR retrotransposon insertion An LTR retrotransposon is inserted in the in the PsMYB10.2 gene promoter was validated by long-range PsMYB10.2-H2 allele promoter PCR. The amplification produced a band of 5 kb in a red f lesh To identify polymorphisms in the promoter that could be H2/H2 and of 2 kb in a yellow f lesh H1/H1 selection (C31 and responsible for the lack of expression of the PsMYB10.2 gene C27, respectively) (Figure 5b). The short band was also amplified in haplotypes other than H2, we did a BLAST search of the H1 in an H4/H4 selection (C12) and in C57, which was the only and H2 allele sequences against the de novo assembly of the yellow f lesh selection homozygous for the H2. Sequencing of variety “Black Gold” (H1/H2) obtained by Fiol et al. [45] through the bands confirmed that the difference in size corresponded CRISPR-Cas9 enrichment sequencing. Only one contig for the to the insertion of the LTR retrotransposon in H2 (Figure 5c). PsMYB10.2-H2 sequence was identified in “Black Gold” (with A characteristic of transposon insertions is the generation of 100% identity), with a sequence extending 3142 bp upstream a duplicated sequence at the genomic integration site, called of the start codon. The BLAST search of this contig against all Target Site Duplication (TSD). In the H2 sequence of the red f lesh “Black Gold” contigs identified an additional one 5004 bp long. selection, a 5 bp TSD sequence was identified f lanking the two Both contigs overlapped 2458 bp with 99.02% identity; the two LTR regions. This 5 bp sequence was found without duplication in sequences were merged to obtain a consensus sequence of the PsMYB10.2 gene promoter of yellow f lesh selections with H2 5522 bp upstream of the PsMYB10.2-H2 gene start codon. For and H4, and was missing in the PsMYB10.2-H1 allele promoter. the PsMYB10.2-H1, we identified a contig with 99.7% identity in “Black Gold”, which extended 9339 bp upstream of the gene start The presence of the LTR retrotransposon codon. Alignment of the isolated H1 and H2 contigs revealed a correlates with red flesh color 2.8 kb insertion in H2, 633 bp upstream of the gene start codon The insertion of the LTR retrotransposon in the PsMYB10.2 gene (Figure 5a). The BLAST search of this 2.8 kb sequence against promoter was evaluated in the collection of 103 Japanese plum the LG3-PsMYB10 region assemblies of the five varieties from selections with a set of PCR primers, giving a 150 bp band when Fiol et al. (submitted) (“Angeleno” H1/H3, “Black Gold” H1/H2, the LTR retrotransposon was present and a 300 bp band when “Fortune” H3/H6, “Golden Japan” H4/H9, and “TC Sun” H4/H5) gave it was absent (Figure 5d; Supplementary Data 1). Among the 41 positive hits in “Black Gold” only. The BLAST search of the same selections with red f lesh, the LTR retrotransposon was found Fiol et al. | 5 Figure 4. Transient overexpression of the PsMYB10.2-H2 and PsMYB10.2-H4 alleles in yellow f lesh fruits and anthocyanin content. (a) fruits agroinfiltrated with the PsMYB10.2-H2 allele; (b) fruits agroinfiltrated with the PsMYB10.2-H4 allele. Negative controls: (c) fruits injected with the pBI121 vector and (d) noninjected fruits. (e) The anthocyanin content in the f lesh of three fruits per treatment, error bars correspond to the standard deviation of each sample analyzed. in either homozygosis or heterozygosis in all the 32 genotyped Discussion with H2 plus one H4/H8 individual (C43). The eight red selections Activation of the PsMYB10.2 gene promotes without the LTR retrotransposon acquired the red color at the very anthocyanin synthesis in the fruit flesh late stage of ripening. These included selection 98–99 and five In a previous study [25] we isolated PsMYB10 alleles and found seedlings of the same parent (“Sweet Pekeetah”, with yellow f lesh). one highly associated with the red skin color and none with the In the yellow f lesh group, only C4 (H1/H2) had the LTR retrotrans- f lesh color. A PCR primer combination was then developed for poson insertion. RT-PCR analysis showed that the PsMYB10.2 gene marker-assisted selection (MAS), which relied on the organization was expressed in the mature f lesh of C4, indicating that, in this of the alleles into haplotypes. Here we increased the sample selection, anthocyanin biosynthesis was repressed downstream in size and looked for correlation between the haplotypes rather the pathway. Our results indicate that there are two H2 haplotypes which than the alleles, with the red/non-red f lesh trait, identifying a differ in the insertion of the LTR. While H2 selections lacking the strong correlation of the H2 haplotype with the red f lesh phe- LTR produce fruits without anthocyanins in their f lesh, those with notype (p-value = 8.06 × 10−12 ). This provides another example of the LTR produce red-f leshed fruits. The LTR explained f lesh color the advantages of using haplotype data rather than only single in 91.26% of the germplasm analyzed here. In all cases, the LTR polymorphisms [50–52]. However, the association did not explain insertion was linked to the PsMY10.2 gene expression, suggesting the f lesh color in 15.53% of the germplasm tested (consisting of a role of this polymorphism in the promotion of the gene expres- advanced breeding lines and a few seedlings): these produced sion. The LTR was also inserted in C43 (H4/H8) suggesting that red f lesh fruits without carrying the H2 haplotype or had the H2 one of the two haplotypes also has two forms: one with the LTR haplotype but had yellow, orange or white f lesh fruits. In Fiol et al. retrotransposon inserted, producing red fruits, and one without. [25] we found that one allele, a467, was always present in H6 and The association of the LTR retrotransposon with red f lesh color in a few individuals with H2. This allele was in the H2 of most of was further validated in a collection of 42 commercial varieties the yellow f lesh individuals. We consider that H2 could have two (eight with red f lesh and 34 with yellow or orange f lesh). The LTR versions: one associated with the red f lesh color, and one (hereby was able to predict the presence or lack of anthocyanin f lesh color defined as H2∗ ) non-associated. in 97.6% of them. The prediction only failed for “Black Beauty” The expression of the PsMYB10.2 gene in red f lesh individuals (yellow f lesh, H1/H2), which, as C4, had the LTR insertion in carrying the H2 haplotype suggested that this gene had a func- heterozygosis (Supplementary Data 1). Considering all the indi- tional role in the synthesis of anthocyanins in the f lesh. This is viduals of the study, the retrotransposon explained the f lesh color in agreement with Fang et al. [53], who found that anthocyanin of 93.1% of the 145 genotypes analyzed. biosynthetic genes and the PsMYB10.2 transcription factor were 6 | Horticulture Research, 2022, 9: uhac206 Figure 5. (a) Graphic representation of the 2.8 kb LTR retrotransposon insertion in the PsMYB10.2 gene promoter in H2. (b) PCR amplification of the region in the PsMYB10.2 gene promoter in the C27 (H1/H1) and C31 (H2/H2) selections. The ladder in the first lane corresponds to Lambda DNA fully digested with HindIII and EcoRI enzymes. (c) Sequence of the retrotransposon insertion site in different haplotypes, the red and yellow circles correspond to red f lesh and lack of red f lesh, respectively. The TSD (target site duplication) produced by the mechanism of the retrotransposon insertion is indicated. (d) Agarose gel visualization of a representative set of Japanese plum selections, the 150 bp band corresponds to the presence of the LTR retrotransposon and the 300 bp band to its absence. upregulated in a red f lesh radiation-induced mutant of a yellow cyanin biosynthesis [12, 13, 46, 47]. In Japanese plum, there were f lesh Japanese plum variety. Here we show that the MYB10.2 no relevant differences in expressed and non-expressed alleles of transcription factor is accountable for part of the natural varia- the PsMYB10.2 gene at the translated amino acid level, suggesting tion of the trait. Fang et al. [53] validated the gene function by that the differences in their functionality could be due to transient overexpression in Nicotiana, co-infiltrating the MYB10.2 polymorphisms in the promoter or in the gene non-coding regions. and the PsbHLH3 genes (required to activate anthocyanin syn- Consistent with this hypothesis, the overexpression of PsMYB10.2- thesis through the R2R3-MYB-bHLH-WD-repeat complex, MBW). H2 and PsMYB10.2-H4 allelic sequences in yellow f lesh fruits The transient overexpression was in yellow Japanese plum fruits, initiated the synthesis of anthocyanins. Therefore, since the most which did not require co-expression of the bHLH3 gene with the distant alleles were functional, we may conclude that all the PsMYB10.2 to observe the pigment accumulation in the f lesh. This PsMYB10.2 alleles cloned may be per se functional and that the indicates that, in the samples tested, the mechanism required for observed differences in f lesh coloration were due to activation activation of the anthocyanin biosynthetic genes was functional of the PsMYB10.2-H2 allele. We noted that the anthocyanin and that the main determinant for anthocyanin accumulation in content in the fruits expressing either the PsMYB10.2-H2 or the f lesh was the expression of the PsMYB10.2 gene. Fang et al the PsMYB10.2-H4 alleles differed between samples, being [54] found that PsMYB10.2 gene expression increased in the late always higher for the latter allele. This might be attributable to stages of maturation in red f lesh cultivars, correlating with the differences in the texture and size of the fruits used, causing expression of anthocyanin biosynthetic genes and with the con- unequal dispersion of the injected Agrobacterium tumefaciens centration of the pigment. Here, we also found that the gene was solution resulting in areas with different concentrations. In expressed in the late stages of fruit development in the red f lesh consequence, while the protocol and experiment design described fruit of ´ Black Gold’, but not during development of the yellow here can successfully validate the function of certain genes f lesh fruits of the “Golden Japan” cultivar (results not shown). qualitatively, a different design including a larger number of fruits Functional amino acid modifications and key residues have of the same variety at the same maturity stage should be used for been described in R2R3-MYB10 proteins that promote antho- quantitative evaluations. Fiol et al. | 7 An LTR retrotransposon in the PsMYB10.2 gene gene, iv) the accumulation of mutations which subdivide promoter region may activate the gene their functionality, complementing each other to achieve the expression function of the ancestral gene (sub-functionalization) [61, 62]. The CRISPR-Cas9 enrichment strategy applied in the PsMYB10 This sub-functionalization can be generated by changes in the region [45] provided a reliable source of data to search for poly- gene regulatory sequences, altering their spatial and temporal morphisms in the promoter of the PsMYB10.2-H2 allele that could expression [60]. This is likely the case in the PsMYB10.1 and explain the gene expression. One of the advantages of long-read PsMYB10.2 genes which greatly differ in their upstream regulatory over short-read high-throughput sequencing technologies is the sequences, causing the former to be expressed in the skin and the ability to detect structural variants [55], which served here to latter in the f lesh, localizing anthocyanin synthesis in different discover the insertion of a 2.8 kb retrotransposon. This large poly- fruit tissues. The haplotypes observed by Fiol et al. [25], based morphism involving a highly repeated sequence in the genome on the segregation of the alleles of the PsMYB10.1, PsMYB10.2 might have been difficult, or even impossible, to identify by other and PsMYB10.3 gene copies, dominant for the red skin (H1, H3; means. both with the PsMYB10.1–356 allele) are recessive for yellow f lesh, Transposable elements can modulate gene expression by while the haplotype dominant for the red f lesh color (H2; with introducing transcription factor binding sites, transcription an LTR in the promoter of the PsMYB10.2 gene) is recessive for initiation sequences or cis-regulatory motifs that respond to the yellow skin (Figure 6). As such, the presence of H2 with either environmental factors. Together with other genetic and epigenetic H1 or H3 is required to confer anthocyanin coloration in both mechanisms they have been found to be responsible for several tissues. Okie (2008) [63] described how plums with yellow skin cases of phenotypic variation in plants [56]. The retrotransposon were transparent and if accompanied by red f lesh they were identified here belongs to the long terminal repeat (LTR) family, mottled and bronze-like, matching the fruits in our study with which has previously been reported to play a role in MYB10 background skin color (or mottled-type) and red f lesh, which variability and the shift of anthocyanin fruit color in other have H2 along with any except H1 or H3. rosaceous crops. The case reported in this study might be similar to that for red-skin apples, where an LTR retrotransposon in Not all the fruit flesh color variability in Japanese the MdMYB10 gene promoter was associated and functionally plum is attributable to the identified MYB10 validated with increased gene expression [26]. In strawberry, an haplotypes LTR retrotransposon disrupting the MYB10 gene sequence pro- Fiol et al. [45] studied the synteny between the LG3-MYB10 duced a causative of the white-f lesh phenotype, while in another region in the genome assemblies of Prunus salicina (two genomes), allele, a CACTA-like transposon inserted in the promoter was Prunus avium (two genomes), Prunus persica (two genomes), P. associated with the enhanced gene expression observed in the mira, Prunus armeniaca (five genomes), P. sibirica (two genomes) red fruits [5]. Outside the Rosaceae family, cases of transposons and P. mandshurica, demonstrating the high complexity of this enhancing the expression of the orthologous R2R3-MYB gene have region in Japanese plum, with homology levels between varieties been reported in purple caulif lower, blood orange and pepper comparable to those between Prunus species. This, together with [57–59]. the identification here of two H2 haplotypes (H2 and H2∗ ) plus In the advanced breeding line collection studied here, we the insertion of the LTR in either H4 or H8 in C43 suggest that observed that the LTR retrotransposon was dominant for the there may be additional polymorphisms in the LG3-PsMYB10 red f lesh color and improved the initial correlation of H2: the region that could account for the variability in the f lesh color LTR explained the phenotype of 91% of the lines versus the 84% not explained by the LTR. explained by the H2 haplotype. The number of alleles shared In addition, similarly to what has been reported in peach, a between H2 and H2∗ and the TSD identified in H2 suggest that single genomic region may not be responsible for all the red the latter originated after insertion of the LTR retrotransposon in f lesh color variability in Japanese plum. In the 103 selections, C4 the promoter region of the original PsMYB10.2-H2∗ allele. Further (H1/H2) was the only one non-anthocyanic with the LTR. When functional studies are required to verify if the retrotransposon is this line was crossed with another yellow selection (C14 [H3/H4]), the main determinant enhancing PsMYB10.2-H2 gene expression some offspring (all with H2) had red f lesh while, as with C4, some and to locate regulatory sequences within the transposable offspring with H2 were yellow [25]. We observed a similar segrega- element or within the disrupted insertion site. tion in the first year of fruiting of F1 progeny of C4 with C11 (yellow f lesh, H3/H6) (data not shown), indicating that, despite being yel- PsMYB10 genes make a great contribution to the low, C4 can be a donor of the red f lesh color trait through the LTR Japanese plum skin and flesh colors retrotransposon. Interestingly, the PsMYB10.2 gene is expressed As for the advanced breeding lines, the fruit skin and f lesh color in the C4 f lesh, which could be explained by the presence of a in the panel of 42 commercial varieties were largely explained by downstream regulator of the anthocyanin biosynthesis pathway. the variability in the LG3-MYB10 genes (only one outlier for the The color prediction in the commercial cultivar collection failed f lesh color, none for the skin color markers), which confirmed the only in “Black Beauty”, also H1/H2, which could be a similar case. strong association of the region with the fruit color trait and the A detailed observation of the trait showed that most of the red usefulness of the molecular markers designed. The MYB10 region individuals without the LTR insertion had different phenotype contains several gene copies in tandem [45], a conformation that to in those with the insertion. While selections and varieties could have been generated by unequal crossing over by either such as “Black Gold” and “Black Splendor” (H1/H2) accumulated homologous or non-homologous recombination [60]. the anthocyanin color progressively and homogeneously in the After a gene duplication event, the subsequent fate of f lesh during ripening (Supplementary data Figure 2a), those the copy can be i) conservation of the gene function (main- without the insertion acquired color in the very late stages tenance), ii) degeneration into a pseudogene or its removal of fruit ripening, and firstly around the skin. One example (non-functionalization), iii) very rarely, the gain of a novel is selection 98–99 considered here as red f lesh as in [43] function (neo-functionalization) or, together with the original although it has been described as yellow f lesh [33, 64, 65] 8 | Horticulture Research, 2022, 9: uhac206 Figure 6. Japanese plum fruits showing the presence or absence of anthocyanin color in skin and f lesh for every possible combination of the seven observed and most frequent MYB10 haplotypes found in the collections of Japanese plum selections and commercial varieties. The combinations H1/H2 and H2/H3 produce anthocyanins in the skin and f lesh. We did not observe combinations with H5 and H6, but we expect they lack anthocyanins in both tissues. (Supplementary data Figure 2b). Other five red-f leshed individ- of anthocyanin is activated and the red pigment accumulates uals without the LTR are offspring of “Sweet Pekeetah”, which in the f lesh. Here, we provide strong evidence for the PsMYB10.2 does not have the retrotransposon and has yellow f lesh: the as candidate for the natural variability of the red f lesh trait in late red f lesh color might segregate through other genomic Japanese plum. As has been demonstrated in other rosaceous region/s. We noted that, in some cases, the red pigmentation species, the insertion of an LTR retrotransposon in the promoter of the f lesh was only localized near the red or black skin of the gene may promote its expression. Apart from contributing (Supplementary data Figure 2c). While this could be due to the to the knowledge of the genetic mechanisms behind the natural direct effect of a genomic region not studied here, a possible variation of Japanese plum fruit color, here we provide for the first contribution of the skin color to that of the f lesh should not time i) an improved protocol for the transient gene overexpression be discarded. Also not explained by the LTR retrotransposon mediated by Agrobacterium in the f lesh of Japanese plum fruits, insertion is the synthesis of anthocyanins induced by chilling that can be used for the validation of other candidate genes; ii) stress (5◦ C). This phenotype has been reported in some yellow an efficient molecular marker for the f lesh color which, together f lesh commercial cultivars studied here: “Friar” (H1/H5), “Black with the one previously developed by our group for skin color, Amber” (H1/H3) and “Royal Diamond” (H1/H3) [66–68]. can be used for marker-assisted breeding. However, as previously Additional QTL studies are required to identify more genomic discussed, the anthocyanin synthesis is complex, regulated by regions involved in the regulation of f lesh coloration. Due to a network involving additional loci and environmental factors, appearing earlier in the maturation process, the homogenous red and not all the f lesh color variability may be determined by a f lesh color trait, if expressed, could have an epistatic effect on late single region. Therefore, further studies are needed to capture coloration. As also noted by other authors, it will be interesting phenotypic variation caused by other loci. to study not only the presence of red f lesh color but when [43] and where it appears, which would facilitate identifying other genomic regions controlling fruit coloration localized in specific Materials and methods fruit tissues and appearing in different ripening stages and/or Plant material and nucleic acid isolation environmental conditions. A collection of 103 Japanese plum selections from two breeding programs was analyzed, consisting of 81 trees cultivated in Huelva (Spain) and 22 trees cultivated in Rinconada de Maipú (Chile) Conclusion (Supplementary Data 1). Fruit color was visually phenotyped from In this study we found that most of the natural variability several fruits at maturity for a minimum of two seasons. Flesh observed in the Japanese plum fruit f lesh color is explained by the color was categorized as white, green, yellow, orange or red; skin variability in the LG3-MYB10 region and triggered by expression color was classified as mottled, green, yellow, red, purple or black. of the PsMYB10.2 gene. We have shown that when the PsMYB10.2 The fruits from five cultivars with different f lesh color were gene is expressed in the f lesh of mature fruits, the biosynthesis obtained in triplicate at maturity from the local market. These Fiol et al. | 9 were “Angeleno” (pale yellow f lesh), “Golden Japan” (yellow f lesh), using the methodology described above. The selections were “Rose” (orange f lesh), “Black Gold” (red f lesh) and a mottled-type C3 (H1/H1), C31 (H2/H2), C6 (H3/H3) and C12 (H4/H4), being plum (dark red f lesh). The skin was removed and the f lesh tissue C31 the unique with red f lesh color. The PCR purified products was used for RNA and DNA extraction. The mature f lesh of three were ligated to the pCR™2.1-TOPO® vector (Invitrogen), following fruits from the selection C4 was also used for RNA extraction. the manufacturer’s instructions, and sequenced. The obtained Transient gene expression assays were carried out using 30 yellow sequences were aligned to the P. salicina MYB10 gene from the f lesh fruits of “Golden Japan” cultivar, harvested one week before NCBI database (EU155161.1) and the PsMYB10.2 cDNA sequence full maturity stage. from “Black Gold” using Sequencher 5.0 software (Gene Codes To further validate the results, a DNA collection of 42 com- Corporation, Ann Arbor, MI, USA). The introns were removed, mercial varieties was analyzed (Supplementary Data 1). The DNA and the in-silico coded protein sequences obtained with the was extracted from the leaves in all cases except for “Rose” and ExPASy Translate tool [71] were aligned to the peach PpMYB10.2 the mottled-type plum, where the DNA was extracted from the (Prupe.3G163000) amino acid sequence using Clustal Omega [72] fruit. Fruit color descriptors for this collection were obtained from and visualized using Jalview v2 [73]. breeders’ descriptions. DNA was extracted following the CTAB procedure [69]. The RNA Vector construction was extracted with the Maxwell RSC Plant RNA Kit (Promega) and The genomic 2 kb sequences from PsMYB10.2-H2 and PsMYB10.2- was further DNase-treated with the TURBO DNA-free Kit (Invitro- H4 were PCR-amplified from the C31 and C12 selections, respec- gen). The quality of the extracted nucleic acids was assessed using tively, using the Phusion Green High-Fidelity DNA Polymerase a Nanodrop ND-1000 Spectrophotometer and 0.8% agarose gels. kit (Thermo Scientific) with primers that included restriction All the studied material, including their fruit color phenotype and enzyme sites (Supplementary data Table 2). The reactions were their genotyping results, can be found in Supplementary Data 1. purified with the High Pure PCR product purification kit (Roche) and digested with SacI and XbaI enzymes in M buffer (TaKaRa). MYB10 genotyping and association with flesh The gene sequence included a SacI restriction site in the second color intron and, after 1 hour of XbaI digestion, the amplicon was All extracted DNAs were analyzed with the MYB10 markers partially digested with SacI by limiting the reaction to 2 minutes (MYB10F2/MYB10NR2 and MYB10F2/MYB10NR4 primer pairs) to at 37◦ C. The pBI121 vector was fully digested with the same obtain their haplotype combination [10, 25]. A χ 2 test was run on enzymes to remove the GUS gene, then the genic sequences the collection of 103 Japanese plum selections to correlate the H1 were ligated next to the CaMV 35S promoter using T4 DNA to H6 haplotypes with the presence or absence of anthocyanin- Ligase (Thermo Scientific) to construct the 35S:MYB10.2-H2 and based coloration of the fruit f lesh. For this, it was assumed that 35S:MYB10.2-H4 vectors. The ligation products were transformed the red phenotype category indicated anthocyanin presence and into Escherichia coli JM109 competent cells and grown in LB the remaining categories (white, green, yellow and orange) its media supplemented with 50 mg/l of kanamycin; screening was absence. performed by direct colony PCR using M102_ex3_F and M13fwd primers (Supplementary data Table 2). The vectors from positive Expression of the MYB10 genes in the fruit flesh colonies were extracted with the GenJet Plasmid Miniprep kit The RNA samples extracted from the fruit f lesh were reverse (Thermo Scientific) and sequenced before transformation into A. transcribed using the PrimeScript RT-PCR Kit (TaKaRa) and the tumefaciens EHA105 competent cells. oligo(dt)20 primer. The expression of the MON reference gene [70] was used for quality control of the synthesized cDNAs. PsMYB10.1, Transient gene expression in Japanese plum PsMYB10.2 and PsMYB10.3 gene expression in the f lesh of the fruits red “Black Gold” and the yellow “Golden Japan” cultivars was The MYB10.2 function was validated by its transient overexpres- evaluated by PCR. Primers and annealing temperatures used for sion in ripe fruits. The procedure was based on previous studies each gene are described in Supplementary data Table 2. The PCR describing the genetic transformation in plum and the transient reactions were carried out using 1.5 mM MgCl2 and 1X NH4 buffers expression in fruits of closely related species [74–77], with some for 1 U of BioTaq polymerase (Bioline), 0.2 mM dNTP mix, 0.2 μM modifications to increase its efficiency. Specifically, A. tumefaciens of each primer, 2 μl of cDNA and MilliQ water adjusted to a 10 μl EHA105 cells transformed with the pBI121, the 35S:MYB10.2-H2 reaction. PCR conditions were 1 minute at 94◦ C, 35 cycles of 94◦ C and the 35S:MYB10.2-H4 vectors were grown for 24 hours in for 30 seconds, the annealing temperature of each primer pair for liquid LB media with kanamycin and rifampicin (50 mg/l), then 20 seconds and 72◦ C for 2 minutes, followed by a final elongation 1 ml of each culture was used to inoculate 15 ml of LB liquid step of 5 minutes at 72◦ C. Genomic DNA extracted from the fruit media containing 20 μM of acetosyringone and 10 mM MES (2-(N- of each variety was added as amplification control. Amplification Morpholino)ethanesulfonic acid) at pH 5.6, and grown overnight. results were visualized in 2% agarose TAE gels after running The final OD600 of the culture was quantified in a UV-2600 spec- 45 minutes at 120 V. Amplification from the cDNAs only occurred trophotometer (Shimadzu), then cells were pelleted by centrifuga- with “Black Gold” PsMYB10.2. The same reaction was repeated tion and resuspended to a OD600 of 0.8 in injection buffer (10 mM adding the cDNAs of “Angeleno”, “Rose” and the mottled-type magnesium chloride hexahydrate, 200 μM acetosyringone, 10 mM plums, and later repeated with the cDNA of C4. The PsMYB10.2 MES at pH 5.6) and kept in the dark at room temperature for gene product from “Black Gold” was purified using PureIT ExoZAP 2–4 hours with occasional shaking. The 30 “Golden Japan” fruits Kit (Ampliqon) and sequenced by capillary electrophoresis with used for the experiment were rinsed 2 minutes in a 30% bleach the same primers used to amplify the fragment. solution, washed with Tween-20 (0.05%) and completely dried. Ten fruits were injected with the Agrobacterium solution carrying PsMYB10.2 sequencing on different haplotypes the 35S:MYB10.2-H2 vector, ten more with the 35S:MYB10.2-H4, The PsMYB10.2 alleles on haplotypes H1 to H4 were PCR-amplified five with the pBI121, and the remaining five fruits were not from the genomic DNA of PsMYB10 homozygous selections injected. Using 1 ml hypodermic syringes, a dose of 1 ml was 10 | Horticulture Research, 2022, 9: uhac206 evenly distributed on one side of each fruit by following the suture Acknowledgments line, with a minimum of three injection points per fruit. Fruits AF is recipient of grant BES-2016-079060 funded by MCIN/AEI/ were kept at room temperature in the dark and opened in halves 10.13039/501100011033 and by “ESF Investing in your future”. CD after 13 days by cutting from the suture line. Anthocyanin was was supported by “DON CARLOS ANTONIO LOPEZ” Abroad Post- quantified using the differential pH spectrophotometry method graduate Scholarship Program, BECAL-Paraguay. This research as used in Fang et al. [54] in Japanese plum fruits, weighing was supported by project RTI2018-100795-B-I00 and by project approximately 0.1 g of f lesh tissue extracted from the observable PID2021-128885OB-I00 both funded by MCIN/AEI/10.13039/ red f lesh patches. The same approximate quantity was extracted 501100011033 and by “ERDF A way of making Europe”. We from the injection sites in the case of the pBI121-treated fruits acknowledge support from the CERCA Programme (“Gen- and randomly in the noninjected group of fruits, which served as eralitat de Catalunya”), and the “Severo Ochoa Programme negative controls. for Centres of Excellence in R&D” 2016-2019 (SEV-2015-0533) and 2020-2023 (CEX2019-000902-S) both funded by MCIN/AEI/ Identification of polymorphisms in the 10.13039/501100011033. PsMYB10.2 gene promoter We thank PLANASA company and their breeders Mario Ortiz To identify polymorphisms in the PsMYB10.2 gene promoter, the and Antonio García for providing plant material and phenotype CRISPR-Cas9 MYB10-enriched MinION sequencing data from Fiol information. We also wish to thank Jorge Naranjo (Tany Nature), et al. [45] was investigated. This data included sequences from Alfonso Guevara (IMIDA) and David Ruiz (CEBAS-CSIC) for provid- the red f lesh cultivar “Black Gold” (H1/H2) and from the non-red ing plant material from commercial varieties, and L. Maria Lois f lesh varieties “Angeleno” (H1/H3), “Fortune” (H3/H6), “Golden for providing the pBI121 vector. Japan” (H4/H9) and “TC Sun” (H4/H5). The blastn software [78] was run with the H1 and H2 PsMYB10.2 input sequences in the “Black Gold” assembly, selecting the contigs with the best Author contributions score hits that extended upstream of the gene. The homologous PsMYB10.2-H2 contig ended prematurely and was extended The experiments were conceived and designed by AF and MJA; by selecting the only contig that overlapped the upstream AF and SG conducted the gene transient expression experiments; sequence after a blast analysis. By aligning and comparing AF and CD performed the gene expression analysis; IP and RI the sequences in Sequencher 5.0, a large insertion on the H2 provided plant materials and contributed to phenotypic data contig compared to the H1 was identified. This sequence was collection; AF and MJ wrote the manuscript. All authors have crit- blast searched in the CRISPR-MinION sequences of the five ically revised the manuscript and approved the final document. varieties, and then on the P. salicina “Sanyueli” v2.0 [49] and P. salicina “Zhongli No 6” v1.0 [37] genomes using blastn software from GDR. The Blast search was repeated in the Rosaceae Data availability transposable elements database from the RPTEdb [79] and the Correspondence requesting materials or data should be addressed sequence was further analyzed with LTR_FINDER software [80] to to M.J.A (mariajose.aranzana@irta.es). identify conserved structures from the LTR-type retrotransposon identified. Conflict interests Validation of the insertion with long range PCR The authors declare no competing interests and marker design To validate the large insertion, primers pM102F and pM102R (Supplementary data Table 2) were designed and used in a long- Supplementary data range PCR reaction using the LongAmp® Taq DNA Polymerase (NEB) to amplify 40 ng of genomic DNA following the kit instruc- Supplementary data is available at Horticulture Research online. tions. The red f lesh selection used was C31 (H2/H2), the yellow f lesh selections were C27 (H1/H1), C57 (H2/H2) and C12 (H4/H4). The PCR amplicons were run in a 2% agarose gel for 45 min- References utes at 120 V. The PCR products were purified and sequenced 1. Saigo T, Wang T, Watanabe M et al. Diversity of anthocyanin and by capillary electrophoresis as described above. The sequencing proanthocyanin biosynthesis in land plants. Curr Opin Plant Biol. results were aligned with Sequencher 5.0 to validate the insertion 2020;55:93–9. site and used in Primer3 software [81] to design two primers 2. Winkel-Shirley B. Biosynthesis of flavonoids and effects of f lanking the 2.8 kb insertion on each side (LTR_F, LTR_R) and stress. Curr Opin Plant Biol. 2002;5:218–23. one within it (LTR_i) (Supplementary data Table 2). The PCR reac- 3. Khoo HE, Azlan A, Tang ST et al. Anthocyanidins and antho- tion for the LTR marker was carried out using 0.2 μM, 0.4 μM cyanins: colored pigments as food, pharmaceutical ingredients, and 0.15 μM of LTR_F, LTR_i and LTR_R primers, respectively, and the potential health benefits. Food Nutr Res. 2017;61:1361779. mixed with 1.5 mM MgCl2 and 1X NH4 buffers, 1 U of BioTaq 4. Espley RV, Hellens RP, Putterill J et al. Red colouration in apple polymerase, 0.2 mM dNTPs and MilliQ water to 10 μl. Thermo- fruit is due to the activity of the MYB transcription factor, cycler conditions were 94◦ C for 1 minute, 35 cycles of 94◦ C for MdMYB10. Plant J. 2007;49:414–27. 15 seconds, 56◦ C for 30 seconds and 72◦ C for 15 seconds, followed 5. Castillejo C, Waurich V, Wagner H et al. Allelic variation of MYB10 by one step at 72◦ C for 5 minutes. This reaction was used to is the major force controlling natural variation in skin and genotype for the absence or presence of the retrotransposon in flesh color in strawberry (Fragaria spp.) fruit. Plant Cell. 2020;32: all the extracted DNA material, loading the 10 μl PCR results 3723–49. into TAE gels with 2% agarose which run for 40 minutes at 6. Tuan PA, Bai S, Yaegaki H et al. The crucial role of PpMYB10.1 in 120 V. anthocyanin accumulation in peach and relationships between Fiol et al. | 11 its allelic type and skin color phenotype. BMC Plant Biol. structural variation and polymorphisms correlating with fruit 2015;15:280. skin color. Front Plant Sci. 2021;12:655267. 7. Feng S, Wang Y, Yang S et al. Anthocyanin biosynthesis in pears 26. Zhang L et al. A high-quality apple genome assembly reveals is regulated by a R2R3-MYB transcription factor PyMYB10. Planta. the association of a retrotransposon and red fruit colour. Nat 2010;232:245–55. Commun. 2019;10:1–13. 8. Jin W, Wang H, Li M et al. The R2R3 MYB transcription factor 27. Cheng Y, Liu L, Yuan C et al. Molecular characterization of PavMYB10.1 involves in anthocyanin biosynthesis and deter- ethylene-regulated anthocyanin biosynthesis in plums during mines fruit skin colour in sweet cherry (Prunus avium L.). Plant fruit ripening. Plant Mol Biol Rep. 2016;34:777–85. Biotechnol J. 2016;14:2120–33. 28. Takos AM, Jaffé FW, Jacob SR et al. Light-induced expression of 9. Zhou H, Lin-Wang K, Wang H et al. Molecular genetics of blood- a MYB gene regulates anthocyanin biosynthesis in red apples. fleshed peach reveals activation of anthocyanin biosynthesis by Plant Physiol. 2006;142:1216–32. NAC transcription factors. Plant J. 2015;82:105–21. 29. Niu J, Zhang G, Zhang W et al. Anthocyanin concentration 10. Fiol A, Howad W, Surya A et al. Development of molecular depends on the counterbalance between its synthesis and markers for fruit skin color in Japanese plum (Prunus salicina degradation in plum fruit at high temperature. Sci Rep. 2017;7: Lindl.). Acta Hortic. 2021;1307:221–6. 1–16. 11. Albert NW, Davies KM, Lewis DH et al. A conserved network of 30. Fang Z, Lin-Wang K, Jiang C et al. Postharvest temperature and transcriptional activators and repressors regulates anthocyanin light treatments induce anthocyanin accumulation in peel of pigmentation in eudicots. Plant Cell. 2014;26:962–80. ‘Akihime’plum (Prunus salicina Lindl.) via transcription factor 12. Stracke R, Werber M, Weisshaar B. The R2R3-MYB gene family in PsMYB10. 1. Postharvest Biol Technol. 2021;179:111592. Arabidopsis thaliana. Curr Opin Plant Biol. 2001;4:447–56. 31. Sooriyapathirana SS, Khan A, Sebolt AM et al. QTL analysis and 13. Lin-Wang K, Bolitho K, Grafton K et al. An R2R3 MYB tran- candidate gene mapping for skin and flesh color in sweet cherry scription factor associated with regulation of the anthocyanin fruit (Prunus avium L.). Tree Genet Genomes. 2010;6:821–32. biosynthetic pathway in Rosaceae. BMC Plant Biol. 2010;10:50. 32. Donoso JM, Picañol R, Serra O et al. Exploring almond genetic 14. Koes R, Verweij W, Quattrocchio F. Flavonoids: a colorful model variability useful for peach improvement: mapping major genes for the regulation and evolution of biochemical pathways. Trends and QTLs in two interspecific almond x peach populations. Mol Plant Sci. 2005;10:236–42. Breeding. 2016;36:1–17. 15. Heppel SC, Jaffé FW, Takos AM et al. Identification of key amino 33. Salazar JA, Pacheco I, Shinya P et al. Genotyping by sequencing acids for the evolution of promoter target specificity of antho- for SNP-based linkage analysis and identification of QTLs linked cyanin and proanthocyanidin regulating MYB factors. Plant Mol to fruit quality traits in Japanese plum (Prunus salicina Lindl.). Biol. 2013;82:457–71. Front Plant Sci. 2017;8:476. 16. Lai Y, Li H, Yamagishi M. A review of target gene specificity 34. Rahim MA, Busatto N, Trainotti L. Regulation of anthocyanin of flavonoid R2R3-MYB transcription factors and a discussion biosynthesis in peach fruits. Planta. 2014;240:913–29. of factors contributing to the target gene selectivity. Front Biol. 35. Shirasawa K, Isuzugawa K, Ikenaga M et al. The genome 2013;8:577–98. sequence of sweet cherry (Prunus avium) for use in genomics- 17. Ravaglia D, Espley RV, Henry-Kirk RA et al. Transcriptional regu- assisted breeding. DNA Res. 2017;24:499–508. lation of flavonoid biosynthesis in nectarine (Prunus persica) by 36. Jiang F, Zhang J, Wang S et al. The apricot (Prunus armeniaca a set of R2R3 MYB transcription factors. BMC Plant Biol. 2013;13: L.) genome elucidates Rosaceae evolution and beta-carotenoid 68–8. synthesis. Hortic Res. 2019;6:128. 18. Zhou H, Lin-Wang K, Wang F et al. Activator-type R2R3-MYB 37. Huang Z, Shen F, Chen Y et al. Chromosome-scale genome genes induce a repressor-type R2R3-MYB gene to balance assembly and population genomics provide insights into the anthocyanin and proanthocyanidin accumulation. New Phytol. adaptation, domestication, and flavonoid metabolism of Chi- 2019;221:1919–34. nese plum. Plant J. 2021;108:1174–92. 19. Yan H, Pei X, Zhang H et al. MYB-mediated regulation of antho- 38. Werner DJ, Okie W. A history and description of the Prunus cyanin biosynthesis. Int J Mol Sci. 2021;22:3103. persica: plant introduction collection. HortScience. 1998;33: 20. Espley RV, Brendolise C, Chagné D et al. Multiple repeats of a 787–93. promoter segment causes transcription factor autoregulation in 39. Shen Z, Confolent C, Lambert P et al. Characterization and red apples. Plant Cell. 2009;21:168–83. genetic mapping of a new blood-flesh trait controlled by the 21. Telias A, Lin-Wang K, Stevenson DE et al. Apple skin patterning is single dominant locus DBF in peach. Tree Genet Genomes. 2013;9: associated with differential expression of MYB10. BMC Plant Biol. 1435–46. 2011;11:93–3. 40. Yamamoto T, Yamaguchi M, Hayashi T. An integrated genetic 22. Pierantoni L, Dondini L, de Franceschi P et al. Mapping of an linkage map of peach by SSR, STS, AFLP and RAPD. J Jpn Soc Hort anthocyanin-regulating MYB transcription factor and its expres- Sci. 2005;74:204–13. sion in red and green pear, Pyrus communis. Plant Physiol Bioch. 41. Beckman T, Alcazar JR, Sherman W et al. Evidence for qualitative 2010;48:1020–6. suppression of red skin color in peach. HortScience. 2005;40: 23. Bretó M, Cantín C, Iglesias I et al. Mapping a major gene for 523–4. red skin color suppression (highlighter) in peach. Euphytica. 42. Guo J, Cao K, Deng C et al. An integrated peach genome struc- 2017;213:14. tural variation map uncovers genes associated with fruit traits. 24. Starkevič P, Paukštytė J, Kazanavičiūtė V et al. Expression Genome Biol. 2020;21:1–19. and anthocyanin biosynthesis-modulating potential of sweet 43. Salazar JA, Pacheco I, Silva C et al. Development and applicability cherry (Prunus avium L.) MYB10 and bHLH genes. PLoS One. of GBS approach for genomic studies in Japanese plum (Prunus 2015;10:e0126991. salicina Lindl.). J Hortic Sci Biotechnol. 2019;94:284–94. 25. Fiol A, García-Gómez BE, Jurado-Ruiz F et al. Characteriza- 44. Acuña CV et al. Characterization of genetic diversity in tion of Japanese plum (Prunus salicina) PsMYB10 alleles reveals accessions of prunus salicina lindl: keeping fruit flesh color 12 | Horticulture Research, 2022, 9: uhac206 ideotype while adapting to water stressed environments. Agron- 63. Okie W. Prunus domestica (European Plum)-Prunus salicina omy. 2019;9:487. (Japanese plum). In: Janick J, Paull RE, eds. The Enciclopedia of Fruit 45. Fiol A, Jurado-Ruiz F, Lopez-Girona E et al. An efficient CRISPR- and Nuts. Cambridge University press, 2008,694–705. Cas9 enrichment sequencing strategy for characterizing com- 64. Salazar JA, Pacheco I, Zapata P et al. Identification of loci con- plex and highly duplicated genomic regions. A case study trolling phenology, fruit quality and post-harvest quantitative in the Prunus salicina LG3-MYB10 genes cluster. Plant Methods. parameters in Japanese plum (Prunus salicina Lindl.). Postharvest 2022;18:105. Biol Technol. 2020;169:111292. 46. Zhou H, Liao L, Xu S et al. Two amino acid changes in the 65. Valderrama-Soto D, Salazar J, Sepúlveda-González A et al. Detec- R3 repeat cause functional divergence of two clustered MYB10 tion of quantitative trait loci controlling the content of phenolic genes in peach. Plant Mol Biol. 2018;98:169–83. compounds in an Asian plum (Prunus salicina L.) F1 population. 47. Yang J, Song HD, Chen Y et al. A single amino acid substitution in Front Plant Sci. 2021;12:1331. the R2R3 conserved domain of the BrPAP1a transcription factor 66. Crisosto CH, Garner D, Crisosto GM et al. Increasing impairs anthocyanin production in turnip (Brassica rapa subsp. ‘Blackamber’plum (Prunus salicina Lindell) consumer rapa). Plant Physiol Biochem. 2021;162:124–36. acceptance. Postharvest Biol Technol. 2004;34:237–44. 48. Verde I, Jenkins J, Dondini L et al. The peach v2.0 release: high- 67. Manganaris GA, Vicente AR, Crisosto CH et al. Effect of delayed resolution linkage mapping and deep resequencing improve storage and continuous ethylene exposure on flesh reddening chromosome-scale assembly and contiguity. BMC Genomics. of ‘Royal Diamond’plums. J Sci Food Agric. 2008;88:2180–5. 2017;18:225. 68. Wang L, Sang W, Xu R et al. Alteration of flesh color and 49. Liu C, Feng C, Peng W et al. Chromosome-level draft genome of enhancement of bioactive substances via the stimulation of a diploid plum (Prunus salicina). GigaScience. 2020;9:giaa130. anthocyanin biosynthesis in ‘friar’ plum fruit by low tempera- 50. N’Diaye A, Haile JK, Cory AT et al. Single marker and haplotype- ture and the removal. Food Chem. 2020;310:125862. based association analysis of semolina and pasta colour in elite 69. Doyle JJ, Doyle JL. A rapid DNA isolation procedure for small durum wheat breeding lines using a high-density consensus quantities of fresh leaf tissue. Phytochemistry. 1987;19:11–15. map. PLoS One. 2017;12:e0170941. 70. Kim H-Y et al. RNA-seq analysis of spatiotemporal gene expres- 51. Maldonado C, Mora F, Scapim CA et al. Genome-wide haplotype- sion patterns during fruit development revealed reference genes based association analysis of key traits of plant lodging and for transcript normalization in plums. Plant Mol Biol Rep. 2015;33: architecture of maize identifies major determinants for leaf 1634–49. angle: hap LA4. PLoS One. 2019;14:e0212925. 71. Artimo P, Jonnalagedda M, Arnold K et al. ExPASy: SIB bioinfor- 52. Calle A, Wünsch A. Multiple-population QTL mapping of matu- matics resource portal. Nucleic Acids Res. 2012;40:W597–603. rity and fruit-quality traits reveals LG4 region as a breeding 72. Sievers F, Wilm A, Dineen D et al. Fast, scalable generation target in sweet cherry (Prunus avium L.). Hortic Res. 2020;7: of high-quality protein multiple sequence alignments using 1–13. Clustal omega. Mol Syst Biol. 2011;7:539. 53. Fang Z-Z, Lin-Wang K, Zhou D-R et al. Activation of PsMYB10. 73. Waterhouse AM, Procter JB, Martin DM et al. Jalview version 2— 2 transcription causes anthocyanin accumulation in flesh of a multiple sequence alignment editor and analysis workbench. the red-fleshed mutant of ‘Sanyueli’(Prunus salicina Lindl.). Front Bioinformatics. 2009;25:1189–91. Plant Sci. 2021;12:1167. 74. Canli F, Tian L. Assessment of regeneration and transient 54. Fang Z-Z, Zhou D-R, Ye X-F et al. Identification of candi- expression factors for agrobacterium-mediated transformation date anthocyanin-related genes by transcriptomic analysis of of Prunus salicina Lindl. Eur J Hortic Sci. 2009;74:66. ‘Furongli’plum (Prunus salicina Lindl.) during fruit ripening using 75. Xi W, Feng J, Liu Y et al. The R2R3-MYB transcription fac- RNA-seq. Front Plant Sci. 2016;7:1338. tor PaMYB10 is involved in anthocyanin biosynthesis in apri- 55. Amarasinghe SL, Su S, Dong X et al. Opportunities and challenges cots and determines red blushed skin. BMC Plant Biol. 2019; in long-read sequencing data analysis. Genome Biol. 2020;21:1–16. 19:287. 56. Wei L, Cao X. The effect of transposable elements on phenotypic 76. Spolaore S, Trainotti L, Casadoro G. A simple protocol for tran- variation: insights from plants to humans. Sci China Life Sci. sient gene expression in ripe fleshy fruit mediated by agrobac- 2016;59:24–37. terium. J Exp Bot. 2001;52:845–50. 57. Jung S, Venkatesh J, Kang M-Y et al. A non-LTR retrotransposon 77. Zhao Y, Mao W, Chen Y et al. Optimization and standardization activates anthocyanin biosynthesis by regulating a MYB tran- of transient expression assays for gene functional analyses in scription factor in Capsicum annuum. Plant Sci. 2019;287:110181. strawberry fruits. Hortic Res. 2019;6:1–13. 58. Butelli E, Licciardello C, Zhang Y et al. Retrotransposons control 78. Altschul SF, Gish W, Miller W et al. Basic local alignment search fruit-specific, cold-dependent accumulation of anthocyanins in tool. J Mol Biol. 1990;215:403–10. blood oranges. Plant Cell. 2012;24:1242–55. 79. Ma K, Zhang Q, Cheng T et al. Identification of transposons 59. Chiu L-W, Zhou X, Burke S et al. The purple cauliflower arises near predicted lncRNA and mRNA pools of Prunus mume from activation of a MYB transcription factor. Plant Physiol. using an integrative transposable element database constructed 2010;154:1470–80. from Rosaceae plant genomes. Mol Gen Genomics. 2018;293: 60. Lallemand T, Leduc M, Landès C et al. An overview of duplicated 1301–16. gene detection methods: why the duplication mechanism has to 80. Xu Z, Wang H. LTR_FINDER: an efficient tool for the pre- be accounted for in their choice. Genes. 2020;11:1046. diction of full-length LTR retrotransposons. Nucleic Acids Res. 61. Otto SP, Yong P. The evolution of gene duplicates. Adv Genet. 2007;35:W265–8. 2002;46:451–83. 81. Kõressaar T, Lepamets M, Kaplinski L et al. Primer3_masker: 62. Prince VE, Pickett FB. Splitting pairs: the diverging fates of dupli- integrating masking of template sequence with primer design cated genes. Nat Rev Genet. 2002;3:827–37. software. Bioinformatics. 2018;34:1937–8.