Transcriptomic Analyses Reveal the Mechanism by Which Different Light Qualities and Light Duration Induce Anthocyanin Biosynthesis in ‘Kyoho’ Grapes

Abstract

The light plays a key role in regulating anthocyanin biosynthesis in plants. However, the molecular basis of anthocyanin synthesis in grape skins irradiated at night with supplemental white light (W), red light (R), blue light (B), and blue light for 3 h, 6 h, 9 h, and 12 h (B3, B6, B9, and B12) is not known. In the present study, the total anthocyanin content in grape skins was significant under different light (DL) and blue-light time (DT) treatments, and the best result was obtained with B9. Analysis of RNA-seq data determined that the key genes in the anthocyanin synthesis pathway, PAL, CHS, DFR, GT, CFIP, VIT_14s0068g00920, VIT_06s0009g02860, and VIT_06s0004g08150, were significantly up-regulated under night-time supplemental light treatments, which led to the significant expression of the phenylpropanoid biosynthesis, phenylalanine metabolism, flavonoid biosynthesis, flavone and flavonol biosynthesis, and the anthocyanin biosynthesis pathway, which were significantly expressed, promoting the accumulation of anthocyanin. The B caused significant expression of PAL (VIT_08s0040g01710), CFIP (VIT_13s0067g03820, VIT_13s0067g02870) and DFR (VIT_18s0001g12800), which may be one of the reasons for the better effect of B compared to W anthocyanin enrichment in grape skins. The DT treatment resulted in significant expression of GT (VIT_11s0052g01600), Peroxidase, VIT_09s0018g01190, and VIT_11s0037g00570. In addition, many TFs such as bHLH, MYB, ERF, WRKY, C2H2, MYB-related, and NAC were found to be involved in the synthesis of anthocyanins under light regulation. These results provide new insights into plants’ nocturnal supplemental-light regulation of anthocyanin accumulation.

1. Introduction

The grapes (Vitis vinifera L.), from a fruit tree widely cultivated worldwide, are enjoyed by a wide range of people. As a fresh fruit for consumption, grapes are also used as a raw material for the production of various products such as wine, grape juice, sultanas, and other functional grape by-products [1,2]. They contain various bioactive compounds such as proanthocyanidins, anthocyanins, and flavonols, the content of which varies considerably in different tissues of grapes [3]. Anthocyanins are mainly found in grape skins and their composition and concentration determine the color of grape skins [4]. There are numerous human health benefits of anthocyanins in grapes, such as antioxidant, anti-inflammatory, regulation of gut microbiota, protection of eyesight, and prevention of cardiovascular diseases [5].

The accumulation of anthocyanins is induced by environmental signals (abiotic stresses such as light, temperature, salinity, oxygen concentration, etc.) and phytohormones (ABA, ETH, TAA, GA, CTK, BR, etc.) and regulated by plant transcription factors (TFs) [6,7,8,9]. These external and internal factors can regulate anthocyanin accumulation by modulating the expression of genes involved in anthocyanin biosynthesis, such as Phenylalanine ammonia-lyase (PAL), chalcone synthase (CHS), chalcone isomerase (CHI), dihydroflavonol 4-reductase (DFR), flavanone 3-hydroxylase (F3H), and glucosyltransferase (GT) [10]. Among the environmental factors, light significantly affects the synthesis of grape anthocyanins and phenolics [11]. The modes of cultivation of grapes are open-field cultivation and facility cultivation. The facility’s greenhouse cultivation protects the grape berries from many diseases, but there will be a lack of light exposure to the grapes. Compared to open-air cultivation, greenhouse cultivation reduces the synthesis of most phenolic compounds such as anthocyanins in grape berries [12]. The use of artificial light supplementation in the greenhouses can be a good way to increase the content of anthocyanins in grape skins and make the skins of grapes uniformly colored.

In recent years, the regulation of anthocyanin synthesis by light-emitting diode (LED) supplemental light treatment has been widely used in various plants. Zhang et al. [13] investigated the effects of red and blue light on strawberry coloring using LEDs. It was found that blue light (B) could rapidly stimulate the accumulation of anthocyanins in the fruits, and red light (R) had a proanthocyanidin synthesis promotional effect. The flavonoid and anthocyanin contents of strawberry fruits were significantly increased by red–blue mixed light (RBL) treatment and ultraviolet B (UV-B) irradiation, respectively, and the UV-B treatment was the most effective [14]. Supplementary B enhances the anthocyanin content in Mini Red Romaine Lettuce, and RBL is most suitable for cultivating Mini Red Romaine Lettuce. [15]. The RBL treatment was the most beneficial for the accumulation of anthocyanins and sugars in grapes, and the B treatment had the highest content of volatile compounds in grapes [16]. During grape storage, the R treatment resulted in higher levels of anthocyanin biosynthesis-related enzymes than green and B treatments [17]. The B was the most favorable treatment for the accumulation of total phenolics in grapevine healing tissues. White, blue, and violet light induced increased anthocyanin accumulation. Mixed wavelengths of light favored flavonoid accumulation [18].

Nowadays, the method of delayed night-time supplemental light has been used in some studies in species such as lettuce, tomatoes, grapes, and peppers [19,20]. Supplemental light treatment at night increased the antioxidant capacity of lettuce and the concentration of phytonutrients such as anthocyanins, carotenoids, and total phenolics [19]. At night, using supplemental light treatment of grape seedlings, it was found that blue light was still a better light source for the increase in anthocyanin content in ripe grapes [21]. However, there are few studies reported on the night-time supplemental light treatment of grapes under the facility’s greenhouses when they are in the color-change stage. Therefore, in this paper, we subjected grapes to nocturnal supplemental light treatment during color change and performed transcriptome analyses to characterize potential transcripts involved in anthocyanin accumulation in grape pericarp under supplemental light irradiation. These findings can improve our understanding of the molecular mechanisms by which night-time supplemental light regulates anthocyanin biosynthesis in grapes.

2. Materials and Methods

2.1. Plant Materials

The field experiment was conducted in the facilities of a cultivated vineyard in Jiangbei Vineyard Area, Wuhu City, Anhui Province, China (118°12′36″ E; 31°27′14″ N), which belongs to the subtropical humid monsoon climate, with an average annual temperature of 16 °C, sunshine hours of about 2000 h, rainfall of 1200 mm, and a frost-free period of about 245 days, with a trellis of 82 m in length, a ridge height of 10 m, and a span of 10 m. The supplemental light experiments were conducted in the greenhouse. The variety used for the test was the ‘Kyoho’ grape planted in 2015, shaped in a ‘V’-type grapevine, with an age of 8 years, planting density of 2.8 m × 2.5 m, and east–west direction. Supplemental light was applied during the color-change period and stopped when the grapes were ripe (soluble solids > 16 °Brix). Every day at night, white, red, and blue light was used as continuous supplemental light 6 h (20:00–2:00), with blue and red wavelengths of 450 nm and 650 nm, respectively, with no supplemental light treatment as a control (CK); in addition to the blue light, the supplemental-light time was divided into four time periods (3 h, 6 h, 9 h, and 12 h). The supplemental light is provided by Blue Shark Lighting Co., Ltd., with a tube length of 1.2 m and a power of 18 W. The light intensity is about 18.5 (µmol m⁻² s⁻¹). Three replicates of each light source were sampled at 90, 95, 100, and 105 days after flowering (DAF). Thirty grapes were picked from each treatment, from each of the 10 bunches of grapes, with one grape per bunch from the top, middle, and bottom of each bunch, respectively. The collected grape samples were immediately snap-frozen with liquid nitrogen and placed in a −80° refrigerator for storage.

2.2. Determination of Physicochemical Specifications of Berries

The soluble solid content was determined using a PAL-1 portable digital refractometer (ATAGO, Tokyo, Japan); the pH value was measured using a SX-620 pen-type pH meter (Shanghai Sanshin Instrumentation Co., Ltd., Shanghai, China); and the total acidity (in terms of tartaric acid) was determined using the acid–base titration method. The total anthocyanin content of grape skins was estimated based on the pH differential method [22], and transcriptome sequencing was performed for different light-quality (DL) groups (N, W, and B) and different light time (DT) groups (CK, B3, and B9). The N and CK were without supplemental light treatment in the DL and DT groups, respectively.

2.3. Transcriptomic Sequencing

2.3.1. RNA Extraction and cDNA Library Construction

The total RNA was extracted using the Agilent RNA Isolation Kit, and RNA quality was assessed by an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA) and detected using a UV spectrophotometer and 1% agarose gels. As well as enriching mRNA with polyA structure in total RNA by Oligo(dT) magnetic beads, mRNA was broken down into short fragments by using ion interruption. The mRNA was used as a template, and cDNA was synthesized using a 6-base random primer and reverse transcriptase. After the library construction was completed, the library fragments were enriched by PCR amplification, and the library was quality checked by Agilent 2100 Bioanalyzer, and then the total concentration of the library and the effective concentration of the library were detected.

2.3.2. Transcriptome Sequencing and Data Quality Control

Based on the Illumina sequencing platform, these libraries were subjected to Paired-end (PE) sequencing. In the raw data obtained from sequencing, some reads with junctions and low quality need to be removed. Fast-Q was used to remove the sequences with junctions at the 3′ end, and the reads with an average quality score lower than Q20 (bases with an accuracy of 99% or more) were removed; the raw data were filtered, and the distribution of the sequencing error rate and the GC content were checked to obtain clean data for the subsequent analyses. The filtered reads were then aligned to a reference genome (Vitis_vinifera.12X.dna.toplevel.fa) using the HISAT 2.2.1 (http://ccb.jhu.edu/software/hisat2/index.shtml (accessed on 4 December 2023)) software.

2.3.3. Analysis of Differentially Expressed Genes (DEGs) and Function Annotation

To explore the changes in anthocyanins of grape skins by DL as well as DT, we evaluated gene expression levels using the fragments per kilobase of transcript per million fragments mapped reads (FPKM) method, and we considered that FPKM > 1 of a gene is generally expressed. After quantification of gene expression was completed, the expression was corrected using hypothesis testing probability (p-value), and genes satisfying |log2FoldChange| > 1 and p-value < 0.05 were defined as differentially expressed genes (DEGs). The concatenated sets and samples of DEGs from all comparison groups were analyzed by bidirectional clustering using the R 4.4.1 language Pheatmap software package, and based on the results of hierarchical clustering, the differentially expressed genes were classified into different clusters according to their expression patterns (genes in the same cluster had similar expression trends), and subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis. Each treatment was treated as a genome and subjected to gene set enrichment analysis (GSEA). The TF prediction was performed by comparing the plants with the Plant TFDB (Plant TFs Database) to obtain the prediction of the TFs and the information of the families to which the TFs belong.

2.4. Quantitative Real-Time Polymerase Chain Reaction Assay (qRT-PCR)

To validate the results obtained from transcriptome sequencing, we selected six genes related to anthocyanin biosynthesis for qRT-PCR analysis. The specific primer sequences are shown in Table S1. qRT-PCR was performed using a Roche Light Cycler 480 fluorescent PCR system (Roche, Basel, Switzerland) and SYBR premixture fluorescent dye kit (Vazyme Biologicals Co., Ltd., Nanjing, China) for real-time PCR assay. The reaction mixture contained 8.2 µL of RNase-free dH2O, 0.8 µL of primers (10 µmol/L), 1 µL of cDNA, and 10 µL of 2 × SYBR real-time PCR premixture. The experiments were performed in 3 biological replicates with Actin as an internal control, and the data were analyzed using the 2^−ΔΔct method for each gene.

2.5. Statistical Analysis

The experimental data were processed using SPSS 21.0 software (SPSS Inc, Chicago, IL, USA) with three replicates for each sample and Duncan’s test, using one-factor ANOVA. The significance level was p < 0.05.

3. Results

3.1. Effect of DL and DT on Physiochemical Characteristics of Grape Berries

The changes in soluble solids of grapes by DL are shown in Figure 1A. The soluble-solid content of grapes increased from 11.85 °Brix to 16.25 °Brix during ripening. During the same period, W, R, and B were all effective in increasing the soluble-solid content compared to the no-supplemental-light-treatment group (N), with B being the most effective at different harvesting stages (90DAF, 95DAF, 100DAF, and 105DAF); B increased 18.99%, 11.07%, 4.89%, and 8.62% to 14.1, 16.05, 17.15, and 17.65 °Brix, respectively. Figure 1B shows the effect of DL on the total acidity of grapes. The acidity of grape berries gradually decreased as they ripened. W, R and B reduced the acidity of grape berries, to a certain extent, compared to group N. The acidity of grape berries was reduced to a certain extent when compared to group B. The lowest value of acidity was observed under B irradiation at the time of grape ripening. The changes made by DL on the pH value of grapes are shown in Figure 1C. The PH value gradually increased during grape ripening, and at the same period, W, R and B increased the PH value of grape berries compared with group N, with B having the best effect. The effects of R and B on PH were significant at 90 DAF and 95 DAF, and the effects of W and B on PH were significant at 100 DAF and 105 DAF. The total anthocyanin content (TAC) was one of the key indicators in this experiment. Figure 1D shows the effect of DL on total anthocyanin in grapes. During grape color change, TAC was accumulating and the TAC was higher under W, R and B treatments than in group N, and the effect of B was more significant compared to R and W. The TAC was higher under W, R and B treatments than in group N, and the effect of B was more significant compared to R and W. At 95 DAF, W, R and B treatments resulted in significantly higher TAC (B > R > W), by 23.23% (0.31 mg/g), 55.56% (0.39 mg/g), and 86.87% (0.47 mg/g), respectively, as compared to group N. At 100 DAF, W, R and B treatments resulted in significantly higher TAC (B > R > W), by 23.23% (0.31 mg/g), 55.56% (0.39 mg/g), and 86.87% (0.47 mg/g), respectively. At 100 DAF, the effect of W and B on TAC was significant and the increase in TAC by R was not significant. At 105 DAF, W, R and B treatments resulted in a significant increase in TAC (B > W > R), by 40.21% (0.69 mg/g), 29.90% (0.64 mg/g) and 71.65% (0.84 mg/g), respectively. Overall, the B treatment showed the best improvement in the quality of grape berries.

We further divided the B, which was the most effective in improving the quality of grape berries, into different irradiation times (DT), and explored the changes in the quality of grape berries with DT. The effect of DT on the soluble solids of grapes is shown in Figure 1E. In the CK group, the soluble-solid content increased from 12.05 °Brix to 16.5 °Brix as the grapes ripened. Compared with the CK group, B irradiation for 3 h, 6 h, 9 h, and 12 h increased the soluble-solid content, with the best effect from B9, which was the highest, at 18.25 °Brix. At the 90DAF, 95DAF, 100DAF, 105DAF stages, B9 increased by 24.07%, 14.88%, 20% and 10.61%, respectively. The effects of B12 and B9 on the content of soluble solids of grapes were approximately the same until 100 DAF, and at 105 DAF, only B9 showed better results, followed by B6. Figure 1F shows the change in total grape acidity by DT. At 90 DAF and 100 DAF, the decrease in total acidity was not significant; at 95 DAF, B6, B9, and B12 all significantly reduced the total acidity of the grapes; and at the time of ripening, B9 had the greatest effect on the total acidity, reaching 5.25%. The effect of DT on the pH of grapes is shown in Figure 1G. At 90 DAF and 95 DAF, only B9 and B12 showed significant improvement in pH; at 100 DAF, only B12 showed significant improvement in pH; when the grapes were ripening, only B9 maintained a better performance, and this performance correlated with the total acidity and soluble-solid content of the grapes at ripening. The changes in grape TAC by DT are shown in Figure 1H. Under DT treatment, grape TAC had different effects on different DAFs. At 90 DAF, all DT treatments significantly improved TAC, and the effects of B6, B9 and B12 were better than that of B3; at 95 DAF, the performance of DT was B9 > B6 > B12 > B3; at 100 DAF, the effect of B3 on TAC was not significant, and B9 > B6 > B12; and at 105 DAF, the effect of B12 on TAC was not significant, and B9 > B6 > B3. In conclusion, both B6 and B9 treatments improved the relevant parameters (soluble solids, PH, total acidity, and TAC) of grape berries more, and B9 was the most effective. The performance of B6 in the DT group was almost the same as that in the DL group, which, to some extent, indicates that the data from the DL group had better reproducibility.

3.2. Transcriptome Data Analysis and DEG Analysis

To investigate the mechanism of regulation of anthocyanin synthesis in grapevine pericarp by DL and DT treatments during the turnaround period, its DEGs were analyzed by mRNA sequencing. There were nine sequenced samples in the DL groups, and each sample produced an average of 42,978,562 clean reads, which accounted for an average of 98% of the sequenced reads. The distribution of the Q30 values ranged from 94.45 to 95.61%. The clean reads were localized to the specified grape reference genomes, and 37,485,798.44 clean reads (uniquely mapped) with only one position in the reference genome were screened; the average uniquely mapped localization rate of these sequenced samples was 95.71% (Table S2). In the DT groups, each sample generated an average of 44,286,377.56 clean reads, which accounted for an average of 98.04% of the sequencing reads. The Q30 values were distributed between 94.68 and 95.55%. There were 38,717,488.89 uniquely mapped, and the average localization rate of these sequencing samples was 95.71% (Table S2). There were 37,485,798.44 uniquely mapped. Those which were mapped had an average localization rate of 95.70% (Table S2), which indicates that the selected reference genome is sufficient for subsequent analysis.

In the DL group (N/W, N/B and W/B), we screened a total of 770, 698 and 427 DEGs, of which 247, 264 and 145 were up-regulated genes, and 532, 434 and 282 were down-regulated genes (Figure 2A), which contained nine common DEGs (Co-DEGs). They contained Co-DEGs 271 (N/B vs. N/W), 110 (N/B vs. W/B) and 96 (N/W vs. W/B) in a two-way comparison (Figure 2B). Our cluster analysis of the DEGs in the N, W and B groups revealed that there were large differences in DEGs among the three treatment groups and that the intra-group differences in each group were small enough to be well-clustered together (Figure 2C). These clustered DEGs were divided into a total of nine clusters (Figure S1A) according to their expression patterns, and supplemental light treatments were transcriptionally down-regulated in clusters 1–5, but the opposite results were observed in clusters 6.7 and 9, suggesting that they may play a role in the metabolic regulation associated with grape skins.

In the DT group (CK/B3, CK/B9, and B3/B9), we screened a total of 623, 1404, and 950 DEGs, of which 212, 966, and 648 were up-regulated genes, and 411, 438, and 302 were down-regulated genes (Figure 2D), which contained common DEGs (Co-DEGs) 34; two-by-two comparisons showed that they contained Co-DEGs 226 (CK/B3 vs. CK/B9), 130 (CK/B3 vs. B3/B9) and 454 (CK/B9 vs. B3/B9) (Figure 2E). We clustered the DEGs in CK, B3, and B9 groups with large differences and small within-group differences (Figure 2F). The DEGs were similarly divided into nine clusters (Figure S1B), with the B3 and B9 treatments being transcriptionally up-regulated in cluster 9, and showing the opposite result in cluster 4.

3.3. GO and KEGG Analysis of DEGs

3.3.1. GO and KEGG Analysis under DL Treatments

Functional annotation of DEGs in N/W, N/B, and W/B through the GO database resulted in 264, 241, and 235 functional categories of differential expression (p value < 0.05), including biological process (BP), molecular function (MF), and cellular component (CC). We filtered the functional terms of the top 10 in BP, MF, and CC, respectively (Figure 3). In N/W, GO terms were predominantly enriched on terms such as extracellular region (GO:0005576), iron ion binding (GO:0005506), and oxidation-reduction process (GO:0055114) (Figure 3A). The enriched pathways of GO were primarily associated with polyketide metabolic process (GO:0030638), polyketide biosynthetic process (GO:0030639), trihydroxystilbene synthase activity (GO:0050350), secondary metabolite biosynthetic process (GO:0044550) and other pathways in N/B (Figure 3B). Anthocyanins are secondary metabolites; they are important parts of the metabolic pathway. In W/B, the GO pathway is primarily involved in terms such as response to hydrogen peroxide (GO:0042542), protein complex oligomerization (GO:0051259), protein self-association (GO:0043621), and response to reactive oxygen species (GO:0000302) (Figure 3C). These DEGs were used for KEGG pathway-enrichment analysis, which revealed that flavonoid biosynthesis (vvi00941), flavonoid and flavonol biosynthesis (vvi00944), phenylalanine metabolism (vvi00360), and Phenylpropanoid biosynthesis (vvi00940) were significantly enriched in N/W and N/B and that these pathways are all closely related to anthocyanin biosynthesis. In addition, in N/W, plant–pathogen interaction (vvi04626), photosynthesis—haptoglobin (vvi00196), and anthocyanin biosynthesis were closely related. In addition, in N/W, plant–pathogen interactions (vvi04626), photosynthesis—haptoglobin (vvi00196), and anthocyanin biosynthesis are closely related (Figure 3D). These pathways are important for developmental ripening, light response, and anthocyanin accumulation in grapes, respectively. Anthocyanins protect plants against stress and promote significant expression of plant–pathogen interaction. In N/B, the biosynthesis of astragaloids, diarylheptanes, and gingerols (vvi00945), and amino acid and nucleotide sugar metabolism (vvi00520) were significantly enriched (Figure 3E). Protein processing (vvi04141), diterpene biosynthesis (vvi00904), and phytohormone signaling (vvi04075) in the endoplasmic reticulum were significantly enriched in W/B (Figure 3F).

3.3.2. GO and KEGG Analysis under DT Treatments

GO functional enrichment analysis of DEGs in CK/B3, CK/B9, and B3/B9 revealed 287, 359, and 325 differentially expressed functional categories (p-value < 0.05), respectively, and in CK/B3, GO terms were mainly enriched in DNA-binding transcription factor activity (GO:0003700), transcription regulator activity (GO:0140110), regulation of cellular macromolecule biosynthetic process (GO:2000112), and so on. (Figure 4A). In CK/B9, GO terms were mainly enriched in terms such as ADP binding (GO:0043531), protein binding (GO:0005515), and photosystem I (GO:0009522) (Figure 4B). In B3/B9, GO terms were mainly enriched in protein self-association (GO:0043621), RNA modification (GO:0009451), protein complex oligomerization (GO:0051259), and in the other terms (Figure 4C). Pathway enrichment was analyzed by the KEGG database for the three comparison groups, and in CK/B3, Plant–pathogen interaction (vvi04626), Plant hormone signal transduction (vvi04075), Photosynthesis—antenna proteins (vvi00196), Zeatin biosynthesis (vvi00908), Phenylalanine metabolism (vvi00360), Diterpenoid biosynthesis (vvi00904), Phenylpropanoid biosynthesis (vvi00904), and Phenylpropanoid biosynthesis (vvi00940) were significantly enriched (Figure 4D). In CK/B9, Photosynthesis—antenna protein (vvi00196), Photosynthesis (vvi00195), Flavonoid biosynthesis (vvi00941), and Protein processing in the endoplasmic reticulum (vvi04141) were significantly enriched (Figure 4E). In B3/B9, Protein processing in endoplasmic reticulum (vvi04141), Photosynthesis (vvi00195), Photosynthesis—antenna proteins (vvi00196), Plant hormone signal transduction (vvi04075), Flavonoid biosynthesis (vvi00941), Phenylpropanoid biosynthesis (vvi00940), and Anthocyanin biosynthesis (vvi00942) were significantly enriched (Figure 4F). These pathways also have key roles in light response and anthocyanin accumulation in grape.

3.3.3. Gene-Set Enrichment Analysis (GSEA) under DL and DT Treatments

We analyzed the GSEA enrichment for each group and screened the top 20 relevant pathways in the results. We found that Flavonoid biosynthesis (vvi00941) was significantly different in both DL and DT groups (Figure 5). In the DL group, based on the core enrichment information, 11 genes were identified (CHI2, VIT_02s0025g04720, VIT_04s0023g03370, VIT_05s0136g00260, VIT_06s0004g08150, VIT_14s0068g00920, VIT_16s0022g01020, VIT_18s0001g03470, VIT_18s0001g03510, VIT_18s0001g12800, VIT_18s0001g14310), which were all significantly expressed in the Flavonoid biosynthesis pathway (Table S3) and this pathway was activated. In the DT group, seven genes (VIT_06s0061g01430, VIT_08s0040g00780, VIT_09s0018g01190, VIT_11s0037g00570, VIT_14s0068g00930, VIT_18s0001g03470, VIT_ 18s0001g14310) were all significantly expressed in this pathway. In addition, the gene sets of all groups except two groups, W/B and CK/B9, were also significantly different in the Phenylpropanoid biosynthesis (vvi00940) pathway (Figure 5). Interestingly, the Phenylpropanoid biosynthesis (vvi00940) pathway was activated only in B3/B9 and repressed in all other groups.

3.4. Transcription Factor Analysis

The structural genes involved in anthocyanin biosynthesis are regulated by TFs, and the TFs associated with light signaling and anthocyanin biosynthesis are also significantly differentially expressed through light treatment. In our study, 200 (54 up-regulated, 146 down-regulated) and 183 (66 up-regulated, 117 down-regulated) DEGs were identified as TFs in the N/W and N/B groups, respectively, divided into 37 and 38 IF families (Table S4). We screened the top 19 TF families (Figure 6) and found that there were mainly ERF, bHLH, NAC, WRKY, MYB, and MYB-related TF families (Figure 6A,B), which had 29 (4 up-regulated, 25 down-regulated), 22 (9 upward, 13 downward), 17 (3 upward, 14 downward), 17 (2 upward, 15 downward), 7 (4 upward, 3 downward), and 13 (3 upward, 10 downward) DEGs in the N/W group; and 18 (3 upward, 15 downward), 21 (7 upward, 14 downward), 15 (7 upward, 8 downward), 11 (2 upward, 9 downward), 13 (9 up-regulated, 4 down-regulated) and 11 (2 up-regulated, 9 down-regulated) DEGs. A total of 257 (163 up-regulated, 94 down-regulated) DEGs were recognized as TFs in the B3/B9 group, which were classified into 43 TF families (Table S5), mainly containing TFs such as bHLH, ERF, NAC, MYB, C2H2 and MYB-related families (Figure 6C); these TF families had 30 (13 up-regulated, 17 down-regulated), 26 (15 up-regulated, 11 down-regulated), 25 (20 up-regulated, 5 down-regulated), 24 (19 up-regulated, 5 down-regulated), 19 (8 up-regulated, 11 down-regulated), and 13 (10 up-regulated, 3 down-regulated) DEGs, respectively.

3.5. qRT-PCR Validation of Relevant Differential Genes

To further validate the RNA-seq results and to investigate the expression patterns of DEGs, we screened six genes related to anthocyanin biosynthesis in the DL and DT groups, respectively (VIT_02s0033g00390, VIT_04s0023g03370, VIT_04s0079g00690, VIT_ 06s0004g02620, VIT_13s0019g04460, VIT_14s0068g00930), which were subjected to qRT-PCR validation (Figure S2). The results showed that the FPKM values of these genes were generally consistent with the expression levels of their related genes. This indicates the reliability of the RNA-seq results and that the identified DEGs can be used for further analysis.

4. Discussion

The production of anthocyanins is influenced by various environmental factors such as light, low temperature, drought, and salinity. Light is a particularly important environmental factor that induces the regulatory pathway of anthocyanin biosynthesis in plants [23]. However, the promotion of anthocyanin accumulation in ‘Kyoho’ grape pericarp by supplemental light has not been well explained in terms of specific gene-expression mechanisms. In this study, we treated ‘Kyoho’ berries with supplemental light and identified differentially expressed genes by transcriptome analysis. Our results identified more than 1000 and 2000 DEGs under DL treatment and DT treatment, respectively, suggesting that light induces the expression of more genes in the berry skin of grapes, which may be responsible for the accumulation of anthocyanin in the berry skin. The anthocyanin biosynthesis is mainly divided into three stages: (1) Phenylpropanoid biosynthesis pathway; (2) Flavonoid biosynthesis pathway; and (3) Anthocyanin biosynthesis pathway. Enrichment analyses by GO, KEGG, and GSEA showed that the pathways associated with Flavonoid biosynthesis, Flavone and flavonol biosynthesis, Phenylalanine metabolism, Phenylpropanoid biosynthesis and Anthocyanin biosynthesis-related DEGs are responsible for anthocyanin accumulation. In addition, transcription factors such as ERF, bHLH, NAC, WRKY, MYB and MYB-related also have an impact on anthocyanin biosynthesis. These pathways and TFs that regulate anthocyanin accumulation are discussed below.

4.1. Phenylpropanoid Biosynthesis and Phenylalanine Metabolism

Phenylpropanoid biosynthesis is catalyzed by PAL, C4H, and 4CL; PAL catalyzes the biosynthesis of cinnamic acid and provides the initial substrate for other phenylpropanoids and phenolic compounds, and it has a key role in controlling the biosynthesis of anthocyanins and other phenolic compounds [24,25]. Increased phenylalanine metabolism also contributes to anthocyanin accumulation [26]. The accumulation of phenolic compounds in red-grape berries and skin is widely promoted at the transcriptional level through the active phenylpropanoid pathway [24]. Our results show that the phenylpropanoid biosynthesis pathway is up-regulated by Phenylalanine ammonia-lyase (PAL: VIT_06s0004g02620, VIT_13s0019g04460), and up-regulated by PAL (VIT_ 08s0040g01710) which was down-regulated; Peroxidase (VIT_07s0191g00050, VIT_06s0004g01180, VIT_07s0191g00050, VIT_02s0012g00540, VIT_01s0010g00960) was up-regulated and Peroxidase (VIT_07s0129g00360, VIT_18s0001g06840, VIT_14s0066g01850, VIT_02s0012g00540, VIT_08s0058g00990, VIT_06s0004g07770, VIT_ 08s0058g00970, VIT_18s0001g01140) were down-regulated; interestingly, Peroxidase (VIT_02s0012g00540) was down-regulated in N/B but up-regulated in B3/B9; unknown DEGs (VIT_11s0037g00570) were also up-regulated in B3/B9 and CK/B3 and down-regulated in CK/B3 (Figure 7). This suggests that differential expression of PAL and Peroxidase favors the enrichment of phenylpropanoid biosynthesis and promotes the accumulation of anthocyanins in grape skins. It has been shown that high temperatures can stimulate peroxidase activity, which reduces anthocyanins in ripe grape berries [27]. The R treatment increased superoxide dismutase (SOD) and peroxidase (POD) activities, and antioxidant enzyme activities were negatively correlated with anthocyanin [28]. Peroxidase (VIT_02s0012g00540) and unknown DEGs (VIT_11s0037g00570) may have inhibited peroxidase activity under different durations of B treatments, increasing grape-skin anthocyanins. PAL (VIT_08s0040g01710) was down-regulated only under B treatment, which may be due to the specific regulation of this gene by B. PAL (VIT_06s0004g02620, VIT_13s0019g04460, VIT_08s0040g01710) was also involved in Phenylalanine metabolism, and Aspartate aminotransferase (VIT_08s0058g01000) and Amine oxidase (VIT_05s0020g03310) are down-regulated in this pathway. In addition, we identified unknown-DEGs involved in the Phenylpropanoid biosynthesis pathway (VIT_06s0004g08150, VIT_11s0037g00570, VIT_04s0044g00190, VIT_10s0003g05420, VIT_ 09s0018g01190, VIT_02s0025g02920, VIT_11s0052g01090, VIT_03s0063g00140, VIT_04s0023g02900, VIT_03s0038g02030) (Table S6), and these DEGs may also contribute significantly to the pathway’s enrichment.

4.2. Flavonoid, Flavone and Flavonol Biosynthesis

Supplemental light treatments significantly enriched the flavonoid biosynthesis pathway in strawberry fruits, allowing the accumulation of flavonoids and anthocyanins [14]. Most DEGs associated with the Phenylpropanoid/Flavonoid pathway are preferentially expressed in red grape skins compared to those of white grape varieties [29]. We showed that in the flavonoid biosynthesis pathway, the Chalcone-flavonone isomerase family protein (CFIP: VIT_13s0067g03820, VIT_13s0067g02870), the dihydroflavonol reductase (DFR: VIT_18s0001g12800) were significantly up-regulated under B treatment, with no significant difference under W treatment; Chalcone synthase (CHS: VIT_05s0136g00260, VIT_14s0068g00930) was significantly up-regulated under both B and W treatments and were significantly up-regulated (Figure 7). DFR catalyzes the biosynthesis of anthocyanins from dihydroflavonoids and is a key regulatory enzyme for anthocyanin biosynthesis in plants, and the expression of this enzyme is positively correlated with the content of anthocyanins [30,31]. Anthocyanin synthesis in pepper fruits under different light conditions was associated with the genetic involvement of CHS, DFR, and CHI [32]. In our study, we found that B could promote anthocyanin accumulation through CFIP, DFR, and CHS, while W could only regulate anthocyanin content through CHS. Unknown DEGs (VIT_09s0018g01190, VIT_11s0037g00570) were up-regulated only in B3/B9 (Table S6), which may be one of the reasons that DT treatments enriched the flavonoid biosynthesis pathway to promote the increase in anthocyanin content in the skins of grapes. The flavone and flavonol biosynthesis is closely related to plant color, and also plays an important role in regulating plant growth and development and resistance to various stresses [33]. Flavonols, as precursor substances of anthocyanin, can be converted to anthocyanin [34]. The unknown DEGs (VIT_06s0009g02970, VIT_06s0009g02860, and VIT_06s0009g02840) were significantly up-regulated under B and W treatments in not only the flavone and flavonol biosynthesis pathways, but also in flavonoid biosynthesis. VIT_14s0068g00920 and VIT_06s0009g02860 were even more up-regulated in N/W, N/B, CK/B9, and B3/B9 (Table S6). These DEGs cause the accumulation of anthocyanin via flavone and flavonol biosynthesis with DL and DT treatments. Increasing light intensity increases flavonol and anthocyanin content, but ultra-high light intensity decreases flavone and proanthocyanidin accumulation [35].

4.3. Anthocyanin Biosynthesis

The anthocyanin biosynthesis pathway converts colorless anthocyanins to colored anthocyanins through the action of ANS, a process that also requires the involvement of many enzymes (GT, BZ1, 3AL, etc.). Glycosyltransferase (GT) is active on a variety of anthocyanins and flavonols, as well as phenolic acids, and its expression promotes the accumulation of anthocyanins during the ripening process of red cherry fruits [36]. GT1 can promote maize anthocyanin expression through the modulation of the activity of ANS [37]. The supplemental light treatment increased anthocyanin accumulation and positively regulated light-induced anthocyanin biosynthesis through up-regulated expression of GT [38]. Our study is consistent with that of Wang et al. [38]. GT (VIT_11s0052g01630, VIT_16s0039g02230, VIT_11s0052g01600) was highly expressed in the Anthocyanin biosynthesis pathway, GT (VIT_11s0052g01630) was both N/W and B3/B9, GT (VIT_16s0039g02230, VIT_11s0052g01600) was significantly up-regulated in both N/W and B3/B9, and GT (VIT_16s0039g02230, VIT_11s0052g01600) was significantly expressed in N/W and B3/B9, respectively. This suggests that supplemental light treatment can significantly enrich the anthocyanin biosynthesis pathway by inducing the expression of GT (VIT_11s0052g01630), and increase the accumulation of anthocyanin in grape skins. The difference in anthocyanin content between W and B may be related to the selective expression of GT (VIT_16s0039g02230, VIT_11s0052g01600). The expression levels of GT, CHI, and photosynthesis-antenna proteins are highly correlated with the concentrations of most flavonoid compounds [39]. The mixed-light (UV-B) treatment significantly induced the expression of GT and promoted the glycosylation of monomeric anthocyanins to form flavonoid glucosides in poplar plums [40].

4.4. TFs That Regulate Anthocyanin Accumulation

The TFs play an important role in regulating the light-induced accumulation of anthocyanins in the grape berries (Figure 8). The significant expression of TFs (bHLH, MYB, ERF, C2H2, MYB-related, and NAC) in grape tissues has been associated with flavonoids and anthocyanin biosynthesis, and showed a high degree of positive correlation with phenolic accumulation in grapes under different light conditions [4,41,42,43,44]. MYBA1 highly activates the ANS2 promoter in grape berries, and EFR23 and bHLH93 activate the DFR genes to participate in the regulation of grape berry color [43]. MYB can also interact with bHLH to promote UFGT promoter activity and anthocyanin biosynthesis. The NAC can promote anthocyanin accumulation by binding to the promoter of MYB [45]. WRKY5 can interact with MYBA1 by promoting jasmonic acid biosynthesis to positively regulate grape anthocyanin synthesis [46]. WRKY is predominantly found in the nucleus, WRKY75, which contains the typical WRKYGQK heptapeptide sequence and C2H2 -zinc-finger structure [47]. UV-B can induce the expression of WRKY71-L, and, by interaction with MYB1 and UFGT promoter interactions, promote anthocyanin accumulation in apple healing tissues [48]. TFs (bHLH, MYB, ERF, WRKY, C2H2, MYB-related, and NAC) were significantly expressed under both DL and DT treatments, suggesting that these TFs can regulate the expression levels of anthocyanins in the skin of grape berries.

5. Conclusions

The transcriptome data were used to analyze the regulation of the pathways related to anthocyanin synthesis in the pericarp of ‘Kyoho’ grapes under DL and DT treatments, and the reliability of the RNA-seq results was verified by Q-PCR. The key genes in the anthocyanin synthesis pathway, PAL, CHS, DFR, GT, CFIP, and unknown genes (VIT_14s0068g00920, VIT_06s0009g02860, VIT_06s0004g08150) were significantly up-regulated under the night-time supplemental light treatment, which resulted in the significant up-regulation of Phenylpropanoid biosynthesis; the Flavonoid, Flavone and flavonol biosynthesis, and Anthocyanin biosynthesis pathway were significantly expressed, promoting the accumulation of anthocyanin. Blue light caused significant expression of PAL (VIT_08s0040g01710), CFIP (VIT_13s0067g03820, VIT_13s0067g02870) and DFR (VIT_18s0001g12800), which may be one of the reasons for the better effect of blue light compared to white light in ‘Kyoho’ grape pericarp anthocyanin enrichment. The DT treatments resulted in significant expression of GT(VIT_11s0052g01600), Peroxidase, VIT_09s0018g01190, and VIT_11s0037g00570. In addition, many TFs such as bHLH, MYB, ERF, WRKY, C2H2, MYB-related, and NAC were found to be involved in the synthesis of anthocyanins under light regulation. These results provide valuable information for the study of anthocyanin accumulation and candidate genes involved in anthocyanin synthesis in ‘Kyoho’ grape skins.