Comparing Two Varieties of Blood Orange: A Differential Methylation Region Within the Specific Encoding Sequence of a Retrotransposon Adjacent to the Ruby Locus

Abstract

The blood orange arose from the insertion of a retrotransposon adjacent to the Ruby gene, an MYB-type transcriptional activator of anthocyanin production, as reported previously. However, the intricate process of anthocyanin regulation among different varieties of blood orange remains incompletely understood. In this study, mRNA levels of the transcription factors Ruby and TT8 were found to be upregulated in the juice vesicle tissues of a variety with higher concentrations of anthocyanins in the pulp compared with another variety with a lower anthocyanin content. In contrast, comparative analysis of the two varieties using two-dimensional electrophoresis and mass spectrometry did not identify differentially expressed proteins related to anthocyanin biosynthesis in the juice vesicle tissues. Furthermore, higher anthocyanin contents were observed in various tissues of transgenic Arabidopsis thaliana overexpressing the Ruby gene from blood orange compared with the wildtype plant. Moreover, the long terminal repeat (LTR) region of a retrotransposon inserted upstream of the Ruby locus exhibited the ability to drive reporter expression through histochemical assay in a transgenic seedling. Thus, a PCR-based molecular marker was developed, targeting the upstream sequence of the Ruby locus to identify Citrus hybrids with the unique trait of red-fleshed fruit. Intriguingly, bisulfite sequencing revealed differentially methylated regions within a Gag-Pol polyprotein-encoding sequence of a retrotransposon adjacent to Ruby locus when comparing two varieties with different anthocyanin contents. A higher average level of methylation status was observed in the fruit with a lower anthocyanin content. In conclusion, methylation modifications at specific upstream positions on the Ruby locus may influence anthocyanin production in blood oranges.

1. Introduction

Anthocyanins are water-soluble pigments belonging to the flavonoid compound family. They are glycosylated polyhydroxy and polymethoxy derivatives of flavylium salts. ‘Tarocco’ (Citrus sinensis [L.] Osbeck) is a health-promoting blood orange variety widely cultivated across the world. Blood oranges of the ‘Tarocco’ nucellar selection are predominantly cultivated in Zizhong County, Sichuan Province, China, and serve as a natural source of anthocyanins. The early-ripening ‘Tarocco’ variety selected contains 10 mg/L of total anthocyanins at the harvesting stage, with anthocyanin concentrations increasing by more than eight-fold when the fruit is stored at cold temperatures. Furthermore, it has been reported that cyanidin-3-glucoside and cyanidin 3-(6″-malony) glucoside are the main components among active pigments in the fruit juice of blood orange [1]. Anthocyanins have strong antioxidant effects and can improve inflammation and metabolic diseases associated with obesity. Cyanidin-3-glucoside potentially exerts its functions primarily through cyanidin-3-glucoside metabolites; more than 20 types of cyanidin-3-glucoside metabolites have been identified in serum, including protocatechuic acid, phloroglucinaldehyde, vanillic acid, and ferulic acid, due to their antioxidant and anti-inflammatory properties [2]. Furthermore, currently available evidence suggests that acylated anthocyanins may have greater potential for modulating energy metabolism, inflammation, and gut microbiota in type 2 diabetes compared with non-acylated anthocyanins [3].

The blood orange arose from the insertion of a Copia-like retrotransposon adjacent to a gene-encoding Ruby, an MYB-type (v-myb avian myeloblastosis viral oncogene homolog) transcriptional activator of anthocyanin production [4]. A Tcs1 retrotransposon is inserted at the promoter region of the Ruby locus; it must provide the regulatory sequences for initiating Ruby expression since the transcription start site of the transcript maps to the long terminal repeat (LTR) region at the 3’ flanking of Tcs1 [4]. Anthocyanin biosynthesis induced by HY5 (elongated hypocotyl 5) expression and the upregulation of Ruby transcription by light have been observed in blood orange [5]. Furthermore, a novel MYB-regulatory gene, Ruby2, which is adjacent to Ruby, has different alleles that exert opposite effects in the regulation of anthocyanin biosynthesis in Citrus species [6]. Recently, a TT8 transcription factor with a bHLH-type motif was found to interact with Ruby and WD40 to form an MBW (MYB-bHLH-WD40) complex that regulates anthocyanin biosynthesis in blood orange [7].

DNA methylation is an epigenetic modification that plays a central role in plant development and evolution. Methylation near gene promoters varies considerably depending on the cell type, with more methylation in promoters correlating with low or no transcription [8]. Comparisons of methylation levels in the MYB10 promoter in ‘Honeycrisp’ apples with red and green stripes indicated that methylation levels correlate with peel phenotypes, with enriched methylation observed in green stripes [9]. Furthermore, temperature-dependent DNA demethylation is a key factor in postharvest, temperature-dependent anthocyanin accumulation in peach flesh [10]. Intriguingly, hypomethylation levels in promoters and gene bodies were observed in blood oranges treated with 5-azacytidine, resulting in higher anthocyanin contents [11]. In the promoter regions of both DFR and Ruby, the amount of cytosine methylation strongly decreases during cold storage in highly pigmented areas of blood orange fruits, whereas it increases in the low-pigmented areas, possibly causing a partial block of gene transcription [12]. Furthermore, the transcription level of ANTHOCYANIN 1 (ClAN1), a key gene associated with citric acid accumulation, was strongly correlated with DNA methylation levels within its coding sequence, strongly indicating that DNA methylation crucially orchestrates the metabolic synthesis of citric acid during lemon fruit development [13].

Nevertheless, the mechanisms underlying Ruby regulation remain poorly understood in blood orange. To investigate the potential role of epigenetic modifications in the regulation of Ruby, a comparative analysis of methylation levels was conducted via bisulfite sequencing between two varieties exhibiting different anthocyanin contents. In this study, upregulated transcription levels of Ruby and TT8 in high-pigmented fruit suggest that an unknown mechanism may enhance the formation of a regulatory complex via RUBY and other partners, thereby activating downstream genes involved in anthocyanin biosynthesis in blood orange. Furthermore, a PCR (polymerase chain reaction)-based molecular marker was developed to identify a Citrus hybrid with the unique trait of red flesh using a specific amplified fragment with a length of 304 bp. Intriguingly, the demethylation of cytosine in the Gag-Pol polyprotein-encoding sequence within a retrotransposon adjacent to the Ruby locus may influence Ruby expression in blood orange.

2. Materials and Methods

2.1. Fruit Quality Analysis for Two Varieties

‘No.3 Xuecheng’ (defined by B03) and ‘No.9 Xuecheng’ (defined by B09), which were selected through a nucellar mutation in the blood orange ‘Tarocco’, were grafted onto Poncirus trifoliata rootstock and grown in Zizhong County, Sichuan Province. When the total soluble solids content in the fruit juice was more than 10.5%, a random sample of about 50 fruits was harvested from 10-year-old trees, with five replicates. Fruit juice was extracted using a domestic juicer. The fruit weight, peel color, total soluble solids, and reducing sugar contents were determined using previously described methods [14]. The anthocyanin contents of fruit juice were measured at 510 nm via spectrophotometry, as described previously [15]. Additionally, the citrate concentrations in juice vesicle tissues from two varieties were measured using HPLC.

2.2. Differential Expression Gene Identification and Analysis

We assessed the purity, concentration, and integrity of RNA extracted from the juice vesicle tissues of fruit, and then qualified RNA was used for library construction. The library preparations were sequenced on a HiSeq X Ten platform (Illumina, San Diego, CA, USA) and paired-end reads were generated. The reference genome was Citrus sinensis v1.0, selected from the Citrus Pan-genome2breeding Database (http://citrus.hzau.edu.cn/index.php, accessed on 1 October 2024). RNA-sequencing reads were mapped using HISAT2 software version 2.2.1.0.

The quantification of gene expression levels was estimated in fragments per kilobase of transcript per million fragments mapped (FPKM). A differential expression analysis was performed using the DESeq2 version 3.21 [16] based on the read count for each gene to compare the two varieties. Total RNA was isolated from juice vesicles and other tissues using an RNA isolation kit for either B03 or B09, respectively (Huayueyang, Beijing, China). Aliquots of 50 μL cDNA were diluted to 100 μL with ddH₂O, and then 2 μL samples were analyzed via real-time quantitative PCR using an LC480 RT-PCR system (Roche, Basel, Switzerland). The relative quantification of each gene was calculated using the 2^−ΔΔCt method. The PCR program was 95 °C for 3 min, 35 cycles of 95 °C for 15 s, 56 °C for 15 s, and 72 °C for 20 s, followed by a 2 min extension at 72 °C. All primers used in this study are listed in Supplementary Table S1. The relative expression levels of each gene of interest were normalized, and the reference gene EF-1α was used as the internal standard.

2.3. Differential Expression Protein Identification and Analysis

The total proteins were extracted for juice vesicle tissues of harvested fruits of both the B03 and B09 variety, according to a protocol [17]. Isoelectric focusing (IEF) and two-dimensions electrophoresis (2-DE) were implemented to identify differentially expressed proteins between the two varieties. Peptide MS/MS was performed using ABI 5800 MALDI-TOF/TOF Plus mass spectrometry (Applied Biosystems, Foster City, CA, USA). Data were acquired in a positive MS reflector with a CalMix5 standard used to calibrate the instrument Applied Biosystems, Foster City, CA, USA. Both MS and MS/MS data were integrated and processed using GPS Explorer V3.6 software (Applied Biosystems, Rockville, MD, USA) with default parameters. Based on combined MS and MS/MS spectra, proteins were successfully identified based on the 95% or higher confidence interval of their scores in the MASCOT V2.3 search engine, referring to a report [18]. The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of differential expression proteins was performed to enrich high-level functions in the defined biological systems.

2.4. Ectopic Overexpression of Ruby in Arabidopsis Thaliana

The plasmid containing a 35S promoter driving Ruby derived from blood orange ‘Tarocco’ overexpression was constructed using pK2GW7 (VIB, Ghent, Belgium). The destination vector was transformed into Agrobacterium GV3101 via electroporation (BIO-RAD, Hercules, CA, USA). The 250 mL cultures of GV3101 were incubated for 12 h and then centrifuged at 3000 rpm for 10 min. The Agrobacterium pellets were re-suspended in half-strength Murashige and Skoog (1/2 MS) medium without agar. The flowers of A. thaliana were dipped in an Agrobacterium culture supplemented with 0.03% silwet for 15 min. After 2 weeks, A. thaliana seeds were harvested and then grown on 1/2 MS supplemented with kanamycin (50 g/mL) to screen the transgenic plants. The kanamycin-resistant seeds were grown in a chamber at 22 ± 2 °C under long-day conditions (16 h light and 8 h dark). After 10 d, the seedlings were transferred to soil and grown in a greenhouse under the same conditions as described previously. The anthocyanin contents in adult leaves of wildtype (WT) or transgenic lines were determined using spectrophotometry. The total RNA of WT or transgenic lines was extracted using TRIzol (Sigma-Aldrich, St. Louis, MO, USA). cDNA synthesis was performed as previously described. A 2 μL aliquot of cDNA was analyzed for each gene of interest via real-time quantitative PCR using LC480 (Roche, Basel, Switzerland). The relative quantification of each gene was calculated using the 2^−ΔΔCt method. All the primers used in this study are shown in Supplementary Table S1.

2.5. Promoter Function Validation for LTR Sequence from 3’ Flanking of a Retrotransposon

Agrobacterium-mediated transformations were conducted for a specific DNA fragment of interest to drive GUS (β-glucuronidase) reporter expression in transgenic seedlings. In this study, we employed a GUS histochemical assay to conduct a qualitative evaluation of β-glucuronidase activity in transgenic plants. A long terminal repeat (LTR) sequence with a length of 496 bp was obtained from a retrotransposon adjacent to the Ruby locus. Subsequently, a destination vector of pKGWFS7-proLTR:GUS was constructed. The recombinant construct was transformed into the strain GV3101. Flowers of Arabidopsis thaliana were dipped in an Agrobacterium culture supplemented with 0.03% silwet for 15 min as mentioned above. A. thaliana seeds were harvested and then grown in a chamber for 10 days at either 22 °C or 4 °C. The transgenic or wildtype seedlings grown in different growth chamber temperatures were harvested and immersed in GUS dye containing X-Gluc and GUS stain buffer (MKbio, Shanghai, China). After decolorization using 70% alcohol, the histochemical staining in seedlings was observed using a microscope (Leica, Wetzlar, Germany).

2.6. PCR-Based Molecular Marker to Identify Citrus Hybrids with Red-Fleshed Fruit

We obtained a long terminal repeat (LTR) region from the 3’ flanking of a retrotransposon that inserted into a sequence upstream of the Ruby locus. Thus, a pair of primers (Fan-F and Xi-R in Supplementary Table S1) were developed to amplify a specific fragment with a length of 304 bp targeting a sequence upstream of the Ruby gene. Therefore, the red-fleshed offspring of ‘Hime Ruby’, a chimera variety derived from the red-fleshed parent ‘Moro’ grafted onto the blond ponkan ‘Ota’, was employed to validate the accuracy of the primers mentioned above. Additionally, the red-fleshed offspring of ‘Hong Yun Xiang Gan’, a hybrid breed from the blond parent ‘Murcot’ and the red-fleshed parent ‘Tarocco’, was also employed to verify the primers designed above. Genomic DNA from each cultivar was isolated from leaves using an E.Z.N.A. plant DNA DS kit (Omega bio-tek, Norcross, GA, USA). The PCR reaction conditions were 95 °C for 5 min, 35 cycles of 94 °C for 30 s, 59 °C for 30 s, and 72 °C for 30 s, followed by a 5 min extension at 72 °C.

2.7. Methylation Profiling of a Partial Encoding Sequence from a Retrotransposon Adjacent to Ruby

We isolated high-quality genomic DNA from juice vesicle tissues of two varieties of fruits using a protocol [19]. Kismeth is a web-based tool for bisulfite sequencing analysis in any sequence context (CG, CHG, and CHH) and provides a tool for the design of bisulfite primers [20]. To analyze the methylation percentage for the sequence upstream of Ruby, different fragments of the upstream sequence ranging from −2191 to −194 adjacent to the Ruby gene were selected for methylation profiling. Subsequently, we performed a bisulfite conversion, treating methylated or unmethylated cytosines using a kit (ZYMO, Irvine, CA, USA), which converted unmethylated cytosines into uracil. However, methylated cytosines remain unchanged during bisulfite treatment. Once converted, the methylation status of DNA sequence can be calculated by using PCR amplified a fragment to be cloned, followed by DNA sequencing.

2.8. Statistical Analysis

Statistical differences in measurement parameters between varieties were evaluated using ANOVA with SPSS version 23 (IBM, New York, NY, USA). Statistical significance was considered at p < 0.05.

3. Results

3.1. Fruit Quality Measurements

No significant differences were observed in fruit weight between the two varieties when the ripening fruit were harvested in March (Figure 1A). The value of the fruit peel color was measured at 35.8 for B03 and 38.3 for B09, with no significant difference between the two varieties (p > 0.05, Figure 1B). The value for the total soluble solids in fruit juice in B09 was measured at 10.9%, while for B03, it was 10.8% (p > 0.05, Figure 1C). However, higher reducing sugar contents were detected in B03, at 4.1%, compared with B09 (p < 0.05, Figure 1D). Furthermore, the total anthocyanin content in B09 was 14.6 mg/L, but B03 had a lower anthocyanin content of 0.76 mg/L (p < 0.05, Figure 1E). In addition, a higher citrate concentration was measured in B09, at 10.33 g/kg, and in B03, it was lower, at 7.75 g/kg (p < 0.05, Figure 1F). In summary, the B09 variety had higher concentrations of anthocyanins and citrate than B03. Consequently, comparative transcriptomic and proteomic analyses of juice vesicle tissues from two varieties were carried out to explore anthocyanin regulation in blood orange.

3.2. Upregulating Expression of Transcription Factors and Structural Genes in High-Pigmented Fruit

Two transcription factors and an anthocyanin biosynthesis gene were quantitatively analyzed using qPCR with mRNA templates isolated from the juice vesicles, leaves, flower buds, and young shoots of the two varieties (Figure 2A,B). The Ruby and TT8 of vital transcription factors displayed higher fruit-specific expression in the juice vesicle than other tissue types (p < 0.05, Figure 3A,B). In contrast, the mRNA levels of Ruby did not show significant differences in leaves, flowers, or shoot tissues from B09 compared with B03 (p > 0.05, Figure 3A). However, higher expression levels of Ruby and TT8 were observed in the juice vesicle tissues from B09 than in those from B03 (p < 0.05, Figure 3B). Moreover, Ruby, TT8, and WD40 formed a BMW complex to regulate the expression of the downstream genes involved in the anthocyanin biosynthesis pathway in blood orange. However, the mRNA levels of Chalcone synthesis in the juice vesicle tissues from B09 were not significantly different from those from B03 (p > 0.05, Figure 3C). Additionally, a total of 991 differentially expressed genes (DEGs) were identified between the two varieties using a transcriptomic method (Supplementary Table S2). Among these, ten DEGs that were upregulated in B09 were enriched in the flavonoid biosynthesis pathway (Figure 3D). In particular, five out of ten upregulated DEGs in B09, including Cs2g14720 (Chalcone synthesis), Cs9g11190 (Chalcone synthesis), Cs7g28130 (Chalcone isomerase), Cs3g25090 (Dihydroflavonol-4 reductase), and Cs5g09970 (Anthocyanidin synthase), were involved in anthocyanin biosynthesis.

3.3. Differential Expression of Proteins Enriched in Glycolysis or Tricarboxylic Acid Cycle Pathways

A total of 64 differentially expressed proteins were identified through two-dimensional electrophoresis between the two varieties (Supplementary Figure S1A,B). Moreover, a total of 42 differentially expressed proteins were identified using MALDI-TOF-TOF analysis, as shown in Supplementary Table S3. Compared with B03, a total of 29 differentially expressed proteins were upregulated in B09, but three proteins were only presented in B09. In contrast, four proteins were upregulated in B03, and nine proteins were only presented in B03. In comparison with B03, the amounts of Cs2g13410 and Cs5g22920 proteins expressed were increased by 2.9-fold and 41.1-fold in B09, while the protein Cs6g15430 was only observed in B09. Therefore, three proteins mentioned above were enriched in the glycolysis or glyconeogenesis pathway (Supplementary Figure S1C). Compared with B03, another three different upregulated proteins were enriched in the tricarboxylic acid cycle pathway in B09 (Supplementary Figure S1C), including Cs3g16700 (Phosphoenolpyruvate carboxykinase), Cs4g15270 (malate dehydrogenase), and Cs6g11970 (NAD-dependent isocitrate dehydrogenase). Therefore, our findings suggest that differentially expressed proteins associated with glycolysis and tricarboxylic acid cycle pathways were upregulated in the juice vesicle tissues from B09, contributing to a lower sugar-to-acidity ratio. However, no differential expression proteins enriched in the flavonoid (or anthocyanin) biosynthesis pathway were identified between the two varieties. In comparison with the total number of differentially expressed genes (DEGs) mentioned earlier in this article, fewer differentially expressed proteins were observed using two-dimensional electrophoresis, which has a lower detection limit.

3.4. Higher Anthocyanin Content Is Accumulated in Transgenic Seedling Overexpressing Ruby

Two different transgenic Arabidopsis thaliana lines overexpressing Ruby from blood orange are defined as T22 and T13 (Figure 4A–D). We detected higher anthocyanin contents in T13 than in T22 (Figure 4E). The expression levels of Ruby were detected in both T22 and T13 (Figure 4F). However, mRNA levels in Ruby were absent from the wildtype plants. Additionally, higher expression levels of AtTT8 from transgenic seedlings were activated in both T22 and T13 compared with the wildtype (Figure 4G). Furthermore, AtDFR exhibited higher expression levels in both T22 and T13 (Figure 4H). Interestingly, transgenic seedling T13 exhibited greater anthocyanin concentrations and showed higher expression levels of Ruby than T22 or the wildtype. Furthermore, T13 also exhibited higher expression levels of AtTT8 and AtDFR compared with T22. In summary, the upregulated expression of transcription factors and downstream genes involved in anthocyanin production was observed in both high-pigmented blood oranges and transgenic Arabidopsis thaliana.

3.5. Molecular Marker Developed to Identify Citrus Hybrid with Red-Fleshed Fruit

The LTR region from the 3’ flanking of a retrotransposon inserted into a sequence upstream of Ruby may promote the expression of the adjacent Ruby gene (Figure 5A). Therefore, an LTR region with a full length of 496 bp from blood orange was obtained, and its promoter activity was investigated in transgenic seedlings overexpressing proLTR:GUS. Each group included five different transgenic seedlings overexpressing proLTR:GUS and wildtype plants grown at 22 °C and showed different intensities of GUS staining (Figure 5B,C). Furthermore, a total of 13 cold temperature-responsive elements were identified within the LTR sequence in blood orange, given in Supplementary Table S4, meaning that stronger-intensity GUS staining was observed in transgenic seedlings grown at 4 °C for 10 days (Supplementary Figure S2). After that, a retrotransposon-based insertion polymorphism technique was employed to amplify a specific DNA fragment targeting sequences upstream of the Ruby locus to identify Citrus hybrids with red-fleshed fruit. The red-fleshed, anthocyanin-accumulating ‘Hime Ruby’, a chimera selected from the red-fleshed parent ‘Moro’ grafted onto a blond parent, the non-anthocyanin-accumulating, ‘Ota’, produced a specific amplified DNA fragment targeting a region upstream of Ruby with a length of 304 bp (Figure 5D, gel enlargement after longer periods of electrophoresis). Furthermore, an amplified DNA fragment with a length of 304 bp was also observed in the red-fleshed hybrid offspring of ‘Hong Yun Xiang Gan’ and its red-fleshed parent ‘Tarocco’ (Figure 5E). The LTR region from the retrotransposon within the sequence upstream of the Ruby gene plays a crucial role in the promoter function and cold-responsive activity of anthocyanin production in blood oranges.

3.6. Differential Methylation Regions Identified in Sequence Upstream of Ruby Locus in Blood Orange

Compared with B03, the upregulation of Ruby expression in juice vesicles was correlated with higher concentrations of anthocyanin in fruit harvested from B09. Moreover, the average percentage of methylation levels for the first region of the polyprotein encoding sequence within a retrotransposon adjacent to the Ruby locus, ranging from −2191 to −1887, was measured at 80.2% in B03 and 67.5% in B09 through bisulfite sequencing. Moreover, each methylation percentage, for CG, CHG, and CHH, was higher in B03 than in B09 (p < 0.05, Figure 6A). However, the average percentage of the methylation level for the second region of the encoding sequence within the retrotransposon did not differ significantly between the two varieties (ranging from −1914 to −1705, Figure 6B). Furthermore, the average percentage of the methylation level for the third region of the encoding sequence within the retrotransposon was measured as 77.4% in B03 and 65.8% in B09 (ranging from −1732 to −1494). Nevertheless, the percentages of CG, CHG, and CHH in a methylation context did not differ significantly between the two varieties (p > 0.05, Figure 6C). Interestingly, the total average percentage of methylation status calculated from three individual fragments of the encoding sequence within the retrotransposon adjacent to the Ruby locus mentioned above was measured as 85.2% in B03 and 77.4% in B09 (p < 0.05, Figure 6D). Therefore, it is hypothesized that a lower expression of Ruby in blood orange may be attributed to a hypermethylation modification in a specific encoding region within a retrotransposon adjacent to the Ruby locus.

4. Discussion

4.1. RUBY Activation of Genes Involved in Anthocyanin Biosynthesis

Recently, two regulatory complexes formed via a bHLH transcription factor, TT8, interacting with RUBY or other MYB-type partners significantly enhanced the expression of anthocyanin biosynthetic and proton-pumping genes, respectively, resulting in an increased production of anthocyanins and citric acid in blood orange [21]. Meanwhile, many genes involved in the anthocyanin biosynthesis pathway were upregulated, along with the bHLH-type transcription factor and other partners, to form a complex in a transgenic ‘Mexican’ lime overexpressing Ruby derived from the blood orange ‘Moro’ [22].

Meanwhile, the transcript levels of AtTT8 were activated in transgenic Arabidopsis thaliana overexpressing Ruby, seedlings of which accumulated higher concentrations of anthocyanin in this study. Furthermore, higher expression levels of Ruby, TT8, and their downstream genes related to anthocyanin biosynthesis in juice vesicle tissues were observed in B09, meaning that blood orange pulp accumulated more pigments. In this study, no differential expression proteins related to regulatory transcription factors or anthocyanin biosynthesis were identified by using two-dimensional electrophoresis with a lower-resolution detection limit in the comparative proteomic analysis of two varieties with different anthocyanin contents. Nevertheless, the co-expression patterns of each mRNA and their corresponding proteins, derived from a combined transcriptomic and Tandem Mass Tag (TMT) proteomic analysis, revealed differentially expressed genes and their proteins involved in the flavonoid biosynthesis pathway between blood oranges and blonde sweet oranges [14]. In contrast, differentially expressed proteins were enriched in the glycolysis and citrate cycle pathways, resulting in a lower ratio of sugar to acidity in B09. In summary, the transcript levels of genes involved in flavonoid biosynthesis were significantly affected by a regulatory complex formed by RUBY and its partners, resulting in increased anthocyanin production in blood orange. However, the molecular mechanism underlying the regulation of the Ruby gene remains unclear.

4.2. The Molecular Marker to Identify a Citrus Hybrid with High-Pigmented Fruit

Almost all the natural variations in pigmentation due to anthocyanins in Citrus species can be explained by differences in the activity of the Ruby gene caused by point mutations and deletions and insertions of transposable elements [23]. Moreover, the LTR sequence from retrotransposon can impact gene expression and regulatory patterns. LTRs are present at both sides flanking a retrotransposon, making them particularly prone to influencing the expression of adjacent genes. Three dominant alleles—R^D-1, R^D-2, and R^D-3—have arisen due to the insertion of a Copia-like retrotransposon into blood oranges, and the dominant alleles confer red flesh in fruit due to anthocyanin accumulation through the regulation of Ruby expression [4]. Molecular markers for the genotyping of R^D-1 and/or R^D-2 alleles were developed to be specifically amplified among blood orange cultivars [24].

In this study, a long terminal repeat (LTR) with a full length of 496 bp at the 3’flanking of a retrotransposon demonstrated the capacity to drive reporter expression in transgenic seedlings. The insertion of a Tcs retrotransposon with an LTR in a sequence upstream of the Ruby locus led to the production of anthocyanins, resulting in red-fleshed fruit. Therefore, a specific amplified DNA fragment targeting the upstream region of the Ruby locus, corresponding to the genomic sequence from the end of the LTR to the first exon, was developed to identify Citrus hybrids with red-fleshed fruit. In totals, 20 different Citrus accessions with or without the red-fleshed trait were successfully validated using the markers (Supplementary Figure S3), meaning that these specific molecular markers facilitate the early selection of hybrid offspring with the red-fleshed trait.

4.3. Differential Methylation Region Observed in a Specific Positions of Sequence Upstream of Ruby

Methylation levels vary among plant species, and this also depends on the genome content of repetitive and mobile DNA, which has high methylation levels [25]. When environmental stimuli influence the reprogramming of DNA methylation in the germline, these modifications can be transmitted to the next generation [26]. Despite their instability, variations in DNA methylation and the consequent modification of gene expression can induce phenotypic changes that are evolutionarily advantageous [27]. Nevertheless, there are few reports of studies comparing differential methylation statuses for two varieties. The elucidation of a nucellar mutation mechanism through methylation modifications remains a fascinating challenge in Citrus fruits.

A partial retrotransposon sequence, 2164 base pairs in length, was cloned from the region upstream of the Ruby gene in two different varieties and found to be identical in our studies. In this study, the differential methylation percentage observed in the region upstream of the Ruby gene, ranging from −2191 to −1887 bp within the gene body encoding the Gag-Pol polyprotein of a retrotransposon, was 67.5% in B09 compared with 80.2% in B03 (p < 0.05). Furthermore, the specific encoding sequence ranging from −2191 to −1887 bp in the upstream region contained six continuous motifs of the TATA box as a core promoter element for transcription initiation. In contrast, the methylation percentage observed in other encoding regions within a retrotransposon, ranging from −1914 to −1705 and from −1732 to −1494, did not differ significantly between the two varieties (p > 0.05). The Tcs1 retrotransposon is inserted at the Ruby locus; it must provide the regulatory sequences for initiating the expression of the Ruby gene, since the transcription start site maps to the LTR instead of the start codes of Ruby [4]. Furthermore, the hypermethylation status of a partial LTR sequence, ranging from −544 to −194 at the region upstream of Ruby was observed but the two levels of methylation were not significantly different in our preliminary study (p > 0.05). Consequently, the differential methylation status within the encoding sequences, ranging from −2191 to −1887 in the upstream region, may play a role in the regulation of the adjacent gene.

The start of transcription was mapped to an adenine within the LTR sequence, 551 nucleotides upstream of the initiating ATG of the Ruby gene [4]. Consequently, we propose a hypothesis that differential methylation region at a specific encoding sequence within Gag-Pol of a retrotransposon is correlated with the differential expression of Ruby in blood orange. This alteration in methylation status within the upstream region adjacent to the Ruby locus ultimately affects the formation of a regulatory complex comprising RUBY, TT8, and WD40 (Figure 7).

5. Conclusions

In this report, blood orange B09 accumulated a higher concentration of anthocyanins, which was correlated with a higher expression level of Ruby and its downstream genes. A long terminal repeat (LTR) sequence from a retrotransposon in a region upstream of the Ruby locus showed the capacity to exert promoter activity. Consequently, a PCR-based molecular marker was developed to identify Citrus hybrids with red-fleshed fruit by amplifying a specific DNA fragment targeted at the region upstream of Ruby. Interestingly, differential methylation regions were observed through bisulfite sequencing in a specific encoding sequence within a retrotransposon adjacent to the Ruby locus in comparison with two varieties with different concentrations of anthocyanin. Thus, we propose a hypothesis about Ruby regulation, that is, the hypomethylation of the encoding region within a retrotransposon ranging from −2191 to −1887 in the upstream sequence may activate Ruby expression, leading to anthocyanin accumulation in blood oranges.