The RNA Directed DNA Methylation (RdDM) Pathway Regulates Anthocyanin Biosynthesis in Crabapple (Malus cv. spp.) Leaves by Methylating the McCOP1 Promoter

The synthesis of anthocyanin pigments in plants is known to be regulated by multiple mechanisms, including epigenetic regulation; however, the contribution of the RNA-directed DNA methylation (RdDM) pathway is not well understood. Here, we used bisulfite sequencing and Real Time (RT)-quantitative (q) PCR to analyze the methylation level of the promoter of constitutively photomorphogenic 1 (McCOP1) from Malus cv. spp, a gene involved in regulating anthocyanin biosynthesis. The CHH methylation level of the McCOP1 promoter was negatively correlated with McCOP1 RNA expression, and inhibiting DNA methylation caused decreased methylation of the McCOP1 promoter and asymmetric cytosine CHH methylation. We observed that the McCOP1 promoter was a direct target of the RdDM pathway argonaute RISC component 4 (McAGO4) protein, which bound to a McCOP1 promoter GGTTCGG site. Bimolecular fluorescence complementation (BIFC) analysis showed that RNA-directed DNA methylation (McRDM1) interacted with McAGO4 and another RdDM protein, domains rearranged methyltransferase 2 (McDRM2), to regulate the CHH methylation of the McCOP1 promoter. Detection of CHH methylation and COP1 gene expression in the Arabidopsis thaliana atago4, atdrm2 and atrdm1 mutants showed that RDM1 is the effector of the RdDM pathway. This was confirmed by silencing McRDM1 in crabapple leaves or apple fruit, which resulted in a decrease in McCOP1 CHH methylation and an increase in McCOP1 transcript levels, as well as in anthocyanin accumulation. In conclusion, these results show that the RdDM pathway is involved in regulating anthocyanin accumulation through CHH methylation of the McCOP1 promoter.


Introduction
Anthocyanins are flavonoid pigments that contribute much of the rich coloration in the plant kingdom, and they are also involved in defense against biotic [1][2][3] and abiotic stresses [4]. Higher anthocyanin levels are also often associated with improved agronomic traits and greater consumer preference and they are recognized human phytonutrients, with benefits as dietary components that include anti-colorectal cancer [5] and gastrointestinal tract cancer [6] properties. Accordingly, there is considerable interest in understanding the molecular processes that affect anthocyanin accumulation.
Anthocyanin biosynthesis is promoted by members of the large MYB transcription factor family, which recognize and bind to response cis-elements in the promoter regions of flavonoid biosynthesis genes [7][8][9][10][11][12]. In studies of apple (Malus domestica), a correlation was observed between the expression pattern of four MYB genes (MdMYB1, MdMYB10, MdMYBA and MdMYB110a) and anthocyanin biosynthesis in fruit peels following exposure to sunlight and low temperatures, both of which elicit anthocyanin accumulation, consistent with a regulatory role in anthocyanin biosynthesis [13][14][15][16]. The MdMYB10 promoter in red-fleshed apples has six tandem repeats of a MdMYB10 binding cis-element, which allow autoregulation of MdMYB10 expression and anthocyanin accumulation [14]. Studies have also shown that MdMYB1 interacts with the ubiquitin E3 ligase, MdCOP1, causing it to be ubiquitinated and degraded via the proteasome pathway, providing another mechanism to regulate apple fruit coloration [17]. Additionally, overexpression of pear (Pyrus communis) PbCOP1.1 was found to reduce red coloration in the fruit peel of the 'Red Bartlett' pear cultivar [18]. These findings further indicate that MdCOP1 is involved in ubiquitination and degradation of MdMYB1, thereby affecting anthocyanin biosynthesis. However, the transcriptional regulation of COP1 in Malus crabapple, which is a widely used model for the study of flavonoid biosynthesis, due to its diversely colored germplasm collections and horticultural importance, is still unclear [17,19]. Indeed, crabapple represents an interesting experimental target to test for other potential regulatory systems that might influence anthocyanin accumulation, including potential epigenetic regulation.
DNA methylation represents an epigenetic mechanism for regulating gene expression at the transcription and post-translation levels in eukaryotes [20]. The principle RNAmediated epigenetic pathway in plants is the RNA Directed DNA Methylation (RdDM) pathway, which mediates de novo DNA methylation of cytosine bases, utilizing small interfering RNAs (siRNAs) [21]. An important element of the RdDM pathway is the DRM2 (DOMAINS REARRANGED METHYLASE 2)-RDM1 (RNA-directed DNA methylation)-AGO4 (argonaute 4) protein complex [22], in which AGO4 binds to small RNAs, including siRNAs or long non-coding RNAs (lncRNAs) [23,24], and cleaves the target RNA transcripts. Moreover, AGO4 catalytic activity is important for the creation of secondary siRNAs that strengthen its repressive effects [25]. DRM2 guides DNA methylation at homologous loci through binding of small RNAs to AGO4 [26,27]. RDM1 binds to singlestranded methylated DNA and controls the production of a 24 nucleotide siRNA, which acts in the RdDM pathway [22]. The RDM1 and AGO4 protein complex regulates the methylation of downstream genes, including ubiquitination genes, MYB-type transcription factors and siRNA regulatory genes, by binding to their promoters [28]. Studies suggest that the RdDM pathway participate widely in various physiological activities in plants, such as seed development, dormancy, fruit ripening, sexual reproduction and, notably for this study, coloration [29][30][31][32][33][34].
A recent study showed that RNA-mediated AGO4 binding to the promoters of certain ubiquitination genes (AT5G26960, AT2G05600, AT3G52030, AT5G41630) directed asymmetric DNA methylation of their genomic regions [28]. Furthermore, it is known that MdAGO4 binds to, and methylates, the MdMYB1 promoter and thus regulates anthocyanin biosynthesis via the RdDM pathway [35]. We therefore hypothesized that the RdDM pathway coordinates the expression of some ubiqitination genes, thereby indirectly regulating anthocyanin biosynthesis in Malus plants.
In this study, the crabapple cultivar 'Radiant', which has young leaves and buds that are red to purple and mature leaves that are green, was used to investigate the mechanism by which the RdDM pathway methylates the promoter of McCOP1 and regulates anthocyanin biosynthesis.

There Is a Negative Correlation between McCOP1 Promoter Methylation Levels and Transcription in Crabapple Leaves
We measured anthocyanin levels in ten developmental stages of 'Radiant' leaves by high-performance liquid chromatography (HPLC) and observed a gradual decrease ( Figure 1A,B). To determine whether methylation of the McCOP1 promoter might play a part in regulating anthocyanin accumulation, we use two PCR primer pairs (BP1 and BP2) to detect the methylation level of the McCOP1 promoter ( Figure 2A). The results showed that CG, CHG, and especially CHH methylation declined during leaf development ( Figure 2B,C).
Two McCOP1 sequences, McCOP1-1 and McCOP1-2 which are paralogs, were cloned from 'Radiant' leaves based on homologous sequences in the apple (M. domestica) mRNA database (https://www.rosaceae.org/, accessed on 26 July 2021). The full-length coding sequences (1941 bp and 1872 bp, respectively) (Supplementary Figure S1) showed a high degree of nucleotide sequence identity (98.4%) (Supplementary Figure S1). Real time (RT)-quantitative (q)PCR analysis revealed a gradual increase in the expression of both genes during leaf development ( Figure 1C). Furthermore, a correlation analysis showed that the degree of CG, CHG and CHH methylation was negatively correlated with

DNA Methylation Levels of the McCOP1 Promoter Change after Treatment with Methylation Inhibitors
To further confirm that the asymmetric cytosine CHH sites were the main methylation loci in the McCOP1 promoter, the 'Radiant' cultivar was separately treated with the methylation inhibitors 5-azacytidine and curcumin. As shown in Figure 2, as inhibitor concentrations increased, leaf development and coloration were significantly inhibited ( Figure 3A). Compared with the control, the relative expression of McCOP1-1 in leaves increased after 0.1, 1 and 10 µmol·L −1 5-azacytidine or curcumin treatment. We also noted that the methylation inhibitors promoted McCOP1-2 expression, and that this was especially evident in the 10 µmol·L −1 curcumin-treated sample. Overall, the use of DNA methylation inhibitors promoted the expression of the McCOP1 genes ( Figure 3B). In addition, the treatments also resulted in changes in CG, CHH and CHG methylation, and the degree of CHH methylation was significantly lower after treatment, as determined using the BP2 primer set ( Figure 3C,D). We therefore concluded that CHH represents the main methylation site in the McCOP1 promoter.

AGO4 Recognize the SINE Site in the McCOP1 Promoter
In A. thaliana, AGO4 is known to be the key RdDM pathway effector, and to regulate the expression levels of target genes by binding to their promoters and mediating promoter methylation [36,37]. To understand how the RdDM pathway affects methylation of the McCOP1 promoter, we investigated the relationship between the AGO4 protein and the  ) were amplified from the leaves of the 'Royalty' crabapple cultivar and an alignment of the predicted proteins sequences revealed conserved domains and/or motifs and showed that they were highly similar to the corresponding apple sequences. A phylogenetic analysis, based on predicted amino acid sequences, showed that each crabapple protein grouped to the same branch as the corresponding apple protein (Supplementary Figure S4).  Previous studies using co-immunoprecipitation (co-IP) and pull-down assays have shown that MdAGO4, MdRDM1 and MdDRM2 from apple interact with each other [35]. We also tested the relationship of these three proteins in crabapple using bimolecular fluorescent complementation (BiFC) and confirmed that MdAGO4, MdRDM1 and MdDRM2 can form a protein complex (Supplementary Figure S2). We conclude that RDM1, DRM2 and AGO4 form a complex, and that AGO4 binds promoter of target genes, while RDM1 catalyzes the methylation.

McRDM1 Silencing Inhibits Anthocyanin Biosynthesis in Crabapple
To identify the main factor affecting CHH methylation of the AtCOP1 (AT2G32950) promoter, we used the A. thaliana atago4, atrdm1 and atdrm2 mutants. Bisulfite analysis showed that CHH methylation in the AtCOP1 promoter in leaves of the atrdm1 mutant was significantly lower compared with the promoter of the other mutants (Supplementary Figure S3). We therefore hypothesized that RDM1 plays a role in the RdDM pathway.
To verify McRDM1 function and confirm that the RdDM pathway is involved in Mc-COP1 promoter methylation and anthocyanin biosynthesis, the McRDM1 gene was silenced using virus induced gene silencing (VIGS). To this end, the TRV-GFP-McRDM1 vector was introduced into 'Royalty' tissue culture seedlings using vacuum infiltration. Green fluorescence was detected in transgenic seedlings transformed with the TRV-GFP-McRDM1 and TRV-GFP (empty vector) constructs, and green coloration was seen in the TRV-GFP-McRDM1-infected leaves, but not in other plants ( Figure 5A). HPLC analysis showed that the anthocyanin levels were substantially lower in the TRV-GFP-McRDM1 leaves compared to leaves transformed with the empty vector ( Figure 5B). Compared with the control,  . Relative expression levels in crabapple leaves around the infiltration sites were determined using RT-qPCR. Bar = 2 cm. RT-qPCR and HPLC were performed with three biological replicates. Error bars indicate the standard error of the mean ± SE of three replicate measurements. Different letters above the bars indicate significantly different values (p < 0.05), calculated using one-way analysis of variance (ANOVA) followed by a Duncan's multiple range test. Scale bar = 1 cm.
We also injected apple fruit with an Agrobacterium tumefaciens culture containing TRV-GFP-McRDM1, and observed fading of the red coloration, as well as lower anthocyanin levels, compared with the empty vector control ( Figure 6B). Furthermore, RT-qPCR results showed that the expression of RdDM pathway genes and anthocyanin related genes in these fruit was similar to that in the McRDM1-silenced leaves ( Figure 6C,D). These results confirm the hypothesis that McRDM1 plays a role in the RdDM complex during Malus fruit and leaf coloration. . Relative expression levels around the infiltration sites were determined using RT-qPCR. Bar = 2 cm. RT-qPCR and HPLC were performed with three biological replicates. Error bars indicate the standard error of the mean ± SE of three replicate measurements. Different letters above the bars indicate significantly different values (p < 0.05), calculated using one-way analysis of variance (ANOVA) followed by a Duncan's multiple range test. Scale bar = 1 cm.

AtRDM1 Contributes to Anthocyanin Accumulation in A. thaliana under Stress Conditions
Anthocyanins have been shown to confer a degree of tolerance to abiotic stress, such as low temperature, low nitrogen, high salinity, and drought stress [38,39], and some abiotic stresses have also been reported to affect the RdDM pathway [40]. Since silencing McRDM1 in crabapple seedlings enhanced anthocyanin accumulation and McRDM1 and AtRDM1 have similar functions in the RdDM pathway, we next measured the expression of methylation-related and anthocyanin-related genes in A. thaliana mutant seedlings grown under low temperature and high glucose conditions. In response to these treatment, the petioles and leaves of the wild type, and the atdrm2 and atago4 mutants became red, while there was no color difference in the atrdm1 mutant when grown under low temperature or high glucose stresses (Figure 7). HPLC analysis indicated that the anthocyanin accumulation in the atrdm1 mutant was significantly lower than in wild type, or in the atdrm2 and atago4 mutants under low temperature or high glucose stresses. Using RT-qPCR, we also detected a decrease in the transcription of anthocyanin biosynthetic genes in the atrdm1 mutants (Figure 7). We therefore concluded that the atrdm1 mutant has impaired anthocyanin accumulation under the imposed stresses, and that RDM1 promotes anthocyanin accumulation under stress conditions. Expression analysis further showed that AtCOP1 expression in atrdm1 increased significantly under low temperature or high glucose stress compared to the control, while the expression of AtMYB75 decreased after the stress treat-ments. We observed no significant difference between the atdrm2 and atago4 mutants and WT in the expression level of anthocyanin related genes (Figure 7). The same result was also found in A. thaliana mutant rosette leaves s (Figure 7). and HPLC were performed with three biological replicates. Error bars indicate the standard error of the mean ± SE of three replicate measurements. Different letters above the bars indicate significantly different values (p < 0.05), calculated using one-way analysis of variance (ANOVA) followed by a Duncan's multiple range test.

Discussion
DNA methylation is known to strongly affect chromatin structure and silence gene expression in both plants and mammals, and it is known that epigenetic modifications can cause somatic mutations in plants [41], including some resulting in color alterations. In floral tissues of the orchid Oncidium, the cytosine loci of the OgCHS 5 -upstream promoter are methylated, leading to a decrease in OgCHS expression and lower levels of anthocyanins [42], and in 'Max Red Bartlett' pear, a correlation between PcMYB10 promoter hypermethylation and a green skin phenotype suggested that DNA methylation silenced expression of the PcMYB10 transcription factor, resulting in lower anthocyanin abundance [43]. In apple fruit, it has been shown that the MdAGO4 proteins MD07G1052200 and MD07G1052400 interact with MdRDM1 (MD16G1197500) and MdDRM2 (MD17G1031900 and MD09G1029900), forming a protein complex [35]. Moreover, MdAGO4 was found to recognize and bind to the MdMYB1 promoter sequence containing the ATATCAGA site responsible for CHH methylation to promote anthocyanin biosynthesis [35]. Thus, there are multiple lines of evidence suggesting that DNA methylation plays an extensive role in anthocyanin biosynthesis.
In Malus spp., MYB10 functions to promote anthocyanin biosynthesis by regulating the expression of anthocyanin biosynthesis genes [44,45]. The ubiquitin E3 ligase, MdCOP1 affects apple fruit coloration by causing ubiquitination and degradation of MdMYB1 via the proteasome pathway [17]. Additionally, McMYB10 modulates its own expression by regulating McCOP1-1 and McCOP1-2 expression to coordinate McCOP1-mediated ubiquitination of McMYB10 [46]. In this study, we found that CHH methylation in the McCOP1 promoter negatively correlated with the expression of McCOP1 during 'Radiant' leaf development, and that CHH is the main methylation site in the McCOP1 promoter. These results provide new insights into the mechanisms by which anthocyanin biosynthesis is regulated.
RDM1, a key effector in the RdDM pathway, can form complexes with AGO4 and DRM2 to jointly regulate gene transcription and physiological development [22]. Mutations in the RDM1 gene were reported to damage the generation of 24-nucleotide siRNAs, resulting in a decrease in DNA methylation, and a consequent reduction in transcriptional gene silencing at the RdDM target loci [22]. In addition, RDM1 has a role in linking siRNA production with pre-existing or de novo cytosine methylation [22]. Of the two complexes involved in the RdDM pathway that promote the accumulation of siRNAs, RDM1 forms a dimer, which interacts with DRD1 and meristem silencing 3 (DMS3), resulting in a complex with a stoichiometry of 1 DRD1:4 DMS3:2 RDM1 [47]. Moreover, the ability to form homodimers is essential for RDM1 to function fully in the RdDM pathway, which is particularly important during the de novo methylation step [48]. These results are consistent with RDM1 playing a major role in methylation through the RdDM pathway, and our study showed a significant reduction in CHH methylation in the AtCOP1 promoter in atrdm1 mutants. Furthermore, anthocyanin accumulation was significantly reduced in crabapple and apple in which McRDM1 had been transiently silenced, indicating the importance of RDM1 in anthocyanin biosynthesis in Malus plants. Finally, the atrdm1 mutant showed less induction of anthocyanin accumulation under low temperature and high glucose conditions than did wild type and other mutant plants, consistent with RDM1 regulating plant anthocyanin biosynthesis through the RdDM pathway under stress.
Twenty four-nucleotide small interfering RNA (siRNAs) can bind to the effector protein AGO4 and direct de novo DNA methylation by the methyltransferase DRM2 [49][50][51][52]. There are ten AGO proteins (AGO1-AGO10) in A. thaliana, of which AGO4, AGO6 and AGO9 act in the RdDM pathway [1,53], and AGO4 in particular has been extensively studied. In plants, the maintenance of CHH methylation is controlled by DRM2, a domain-rearranging methyltransferase from the RdDM pathway [54]. In thi pathway, AGO4 can bind two classes of non-coding RNAs: small interfering RNAs (siRNAs) or long non-coding RNA (lncRNA), to generate complexes that can guide DRM2 to specific genomic locations [55]. Nonsymmetrical CHH (where H is a base other than G) methylation is thought to require de novo methylation by DRM2 after AGO4 binding to chromatin [56]. In A. thaliana, AGO4 binding sites are often located distant from genes that are differentially expressed in the atago4 mutant [57], and it has been shown by chromatin immunoprecipitation (ChIP)-sequencing that AGO4 preferentially binds to a region between 200 to 500 bp upstream of target gene transcription start sites [28]. Recently, a study also showed that MdAGO4 proteins can directly bind to the MdMYB1 promoter through the ATATCAGA sequence in vivo [35]. In this study, we found that McAGO4 can bind to the McCOP1 promoter at the SINE site (GGTTCGG), but not the above mentioned ATATCAGA site, suggesting that there are several AGO4 binding sites, and that they lack conservation among the target genes. We here provide evidence of binding by McAGO4 to the promoter of McCOP1 and the formation of AGO4-RDM1-DRM2 complexes in crabapple leaves.
Plants have evolved mechanisms to promote anthocyanin biosynthesis in response to variation in environmental conditions, such as long day, low temperature and low pH conditions [58][59][60]. Recent studies have suggested that M. domestica BTB/POZ and TAZ domain-containing protein 2 (MdBT2) negatively regulate hormonal and environmental signal-induced anthocyanin biosynthesis in apple fruit [61]. Under drought conditions, the ethylene response factor ERF family protein 38 (MdERF38) interacts with the positive anthocyanin biosynthesis modulator MdMYB1 and facilitates its binding to anthocyanin biosynthesis genes [62]. Additionally, the apple transcription factor bHLH3 (MdbHLH3), has been identified as a positive regulator of cold tolerance and anthocyanin biosynthesis due to its activation of C-repeat-binding factor 2 (MdCBF2) and Dihydroflavonol 4-reductase (MdDFR) expression [63]. Some abiotic stresses have also been reported to affect methylation and demethylation in plants. For example, salt stress enhances DNA methylation in poplar (Populus), than in wild type plants grown under normal conditions involving histone H3 and 5.8 rDNA loci [64]. In A. thaliana, heat stress induces DRM2, DNAdirected RNA polymerase IV subunit 1 (NRPD1) and DNA-directed RNA polymerase V subunit 1 (NRPE1) expression, demonstrating that the RdDM pathway is mobilized by high temperature to strengthen stabilization of DNA methylation [40]. Multiple studies have shown that DNA methylation is also involved in stress responses leading to anthocyanin accumulation. As an example, peach (Prunus persica) stored at low temperatures were reported to show higher transcript levels of genes involved I anthocyanin biosynthesis, as well as the transcription factor PpbHLH3, and this was associated with lower methylation levels in the promoters of these genes [65]. Such results indicate that plants promote anthocyanin accumulation by regulating promoter methylation under stress conditions.
Here, we report that the RdDM pathway regulates anthocyanin biosynthesis through the formation of an AGO4s (McAGO4-A and McAGO4-like)-RDM1-DRM2 complex, and that McAGO4 binds to the promoter of McCOP1 in crabapple leaves. We propose that RDM1 is the main effector mediating DNA methylation during stress-induced anthocyanin accumulation.

Plant Materials and Growth
Eight-year-old trees (Malus cv 'Radiant') were grafted onto Malus hupehensis and planted at the Crabapple Germplasm Resources Nursery at the Beijing University of Agriculture (40.1 • N, 116.6 • E). Young leaves of the 'Radiant' cultivar used in this study are red and turn green during maturation. One tree of each cultivar showing consistent growth was selected. Leaf samples were collected from annual branches growing on the fringe of the canopy and in the same compass directions. Leaves of 'Radiant' were collected at 10 developmental stages (3,6,9,12,15,18,21,24,27

Anthocyanin Quantification
Leaf and fruit samples (1.0 g fresh weight) were extracted with 10 mL extraction solution (methanol: water: formic acid: trifluoroacetic acid = 70:27:2:1) at 4 • C in the dark for 72 h, shaking every 6 h. The mixture was then centrifuged at 4 • C at 12,000× g for 15 min and the supernatant filtered through a 0.22 µm MilliporeTM filter (Billerica, MA, USA). HPLC analysis was performed as previously described [66], with detection at 520 nm [67]. All samples were analyzed in triplicate (extracted from three different pools of leaves).  Supplementary Table S1.

Measurement of Methylation Levels
We measured the level of methylation through BSP (Bisulfite Sequencing PCR) analysis [68]. A DNA Bisulfite Conversion Kit (TIANGEN BIOTECH, Beijing, China) was used to purify the bisulfite-treated genomic DNA. Digested DNA was used in a Methylation-Specific PCR (TIANGEN BIOTECH, Beijing, China) reaction following the manufacturer's instructions. The methylation level of each fragment was calculated. Primer sequences are listed in Supplementary Table S1.

Expression Analysis
RT-qPCR was performed with the SYBR ® Premix Ex Taq TM II (Takara, Ohtsu, Japan) in a CFX96TM Real-Time PCR System (BIO-RAD, Hercules, CA, USA). Relative transcript levels were quantified by normalizing to the 18S ribosomal RNA gene (M. domestica, GenBank DQ341382) using the 2 −∆∆CT method [69]. RT-qPCR analysis was performed with three technical replicates and three biological replicates. Primers are listed in Supplementary Table S1.

Yeast One-Hybrid Assay
The full-length McAGO4-A, McAGO4-Like and McDRM2 sequences were cloned into the pJG4-5 (pB42AD) vector [70] for protein expression, and the McCOP1 promoter sqeunce was ligated into the placZi vector [70]. Constructs were co-transformed into yeast strain EYG48 and selected on a SD/Trp-Ura plate at 28 • C. Yeast culture solutions were used for the color reaction on a SD/Trp-Ura plate (with X-gal added) at 28 • C. All transformations and screenings were performed three times. The primer sequences used are listed in Supplementary Table S1.

EMSA
The full-length McAGO4-A, McAGO4-Like and McDRM2 cDNA PCR products were subcloned into the pMAL-C2X expression vector [71]. Bacterial growth and protein induction were performed as described by the manufacturer (Novagen). Following induction of protein expression in the cells with 0.3 mM/100 mL isopropyl-β-D-thiogalactoside (IPTG), and washing with and suspending in PBS solution, 5 mL bacterial aliquots (12 mg/mL) were stored at −80 • C.

BiFC
The McAGO4-A, McAGO4-Like, McDRM2, McRDM1, MdCOP1-1, MdCOP1-2 and McMYB10 coding sequences were cloned into the p35SYCE and p35SYNE vectors, which contain DNA encoding the C-terminal or N-terminal regions of yellow fluorescent protein (YFP) (YFPN or YFPC), respectively. Tobacco (Nicotiana benthamiana) leaves were transiently co-transformed using an A. tumefaciens (GV3101) infection method with different combinations of these constructs [73]. YFP-dependent fluorescence was detected 24 h after transfection using a confocal laser-scanning microscope (Zeiss LSM 510 Meta). All images correspond to single optical slices of epidermal cells. Excitation and emission wavelengths were 514 and 527 nm, respectively, for enhanced YFP. Specific primers are listed in Supplementary Table S1.

RDM1 Silencing in Crabapple Plantlets and Apple Fruit
The McRDM1 full-length cDNA was designed with XbaI and BamHI sites for insertion into the TRV2 vector, which was transformed into A. tumefaciens (GV3101) and used for transient co-transformation with the TRV1 vector into cv 'Royalty' seedlings and 'Red Fuji' fruit via vacuum infiltration and injection, respectively [67]. A handheld UV instrument (SUNLONGE SUPER UV LED LAMP) was to detect and record the GFP fluorescence of the successfully transformed apple seedlings and fruit 5 days after transformation for seedlings and 3 days for fruit [67]. Plant leaves showing GFP fluorescence were used to detect gene expression levels and anthocyanin content after UV detection.  Figure S3: AtCOP1 promoter methylation level determined by bisulfite sequencing using different primers. A. The phenotype of WT, atago4, atdrm2 and atrdm1 seedlings. B. The position of BP3 and BP4 primers in the promoter sequence. C. AtCOP1 promoter methylation level using BP3 primers. D. AtCOP1 promoter methylation level using BP4 primers. Error bars indicate the standard error of the mean ± SE of three replicate measurements. Different letters above the bars indicate significantly different values (p < 0.05) calculated using one-way analysis of variance (ANOVA) followed by a Duncan's multiple range test. Figure S4 Table S1: Primer sequences used in this study. Table S2