The potato population F₁RM was obtained by crossing with the heterozygous diploid yellow peel yellow flesh potato RH (S. tuberosum) as the female parent and the yellow peel yellow flesh diploid inbred clone M6 (S. chacoense) as the male parent. F₁RM tubers are also yellow peel yellow flesh, but the peel of the tubers will turn purple gradually after exposure to light. Among them, the tubers of line RM-210 were selected to study the regulation of anthocyanin accumulation during light exposure. The RM-210 plants were grown in 24-cm-diameter plastic pots in greenhouse with 16 h light per day supplied by lamps. For about 80 days after planting, the mature tubers were harvested and dark storage for 10 days at 20°C. Then they were incubated under the continuous light intensity of 2000 LX, and samples were collected at 0, 6, 12, 36, 48, and 120 h. Five tubers were used for sampling at each time point, and the peels (1 mm thick) were collected and stored at –70°C for molecular and biochemical analyses.

Nicotiana benthamiana L. for dual luciferase assays was grown under greenhouse conditions until 5–6 leaves were available for Agrobacterium infiltration.

Anthocyanin mixtures were extracted according to the agricultural standard of China (NY/T 2640-2014). 100mg powder was extracted with 1.0 ml 1% formic acid methanol overnight at 4°C. Extracts were centrifuged at 10,000 ×g for 10 min and filtrated with 0.22-μm microporous membrane. Anthocyanin quantification was measured by High Performance Liquid Chromatography (HPLC) as a previous description (Zhou et al., 2008). The anthocyanin standard consisted of peonidin, delphinidin, malvidin, petunidin, pelargonidin, and cyanidin dissolved in methanol.

Total RNA was isolated according to previous description (Zhang et al., 2022). The RNA integrity was assessed using the Fragment Analyzer 5400 (Agilent Technologies, CA, USA). Sequencing libraries were generated using NEBNext^® Ultra™ RNA Library Prep Kit for Illumina^® (NEB, USA) in accordance with the manufacturer’s recommendations and index codes were added to attribute sequences to each sample.

The original files obtained from the Illumina platform are transformed into raw data by base calling. The raw reads were trimmed by removing adapters, reads containing ploy-N more than 5% and low-quality sequences. Raw sequences were transformed into clean reads after processing. Then the clean reads were mapped to the transcriptome of the potato DM reference genomes (6.1 version, http://spuddb.uga.edu/dm_v6_1_download.shtml) with Salmon software, version 1.3.0 (Patro et al., 2017).

Transcripts Per Kilobase Million (TPM) were used to estimate gene expression level. The R package DESeq2 was used to perform the differential gene expression among different samples. Thresholds for significantly differential expression were false discovery rate (FDR) ≤ 0.05 and fold change ≥2 or ≤ 0.5. The KEGG pathway enrichment analysis of differentially expressed genes (DEGs) was performed by TBtools at the adjusted p <  0.05 by using a hypergeometric test (Chen et al., 2020). The cluster analysis was performed as described previously (Guo et al., 2016).

Reverse transcription quantitative real-time PCR (RT-qPCR) primers were designed with the NCBI Primer-BLAST and listed in Table 1 . The RT-qPCR was performed on a CFX96™ real-time PCR system (Bio-Rad, USA) with the TransStart Top Green qPCR SuperMix kit (Transgen, Beijing, China). The potato housekeeping genes ef1α (GenBank: AB061263) and actin (GenBank: XM_006345899.2) were selected as internal reference genes. Relative expression of the individual gene was calculated with calculated multiple internal control method (Vandesompele et al., 2002).

According to the Soltu.DM.10G020840.1 sequence from the Potato Genome Database, the open reading frame (ORF) of the StMYBA1 was amplified from the cDNA of RM-210 tuber peel. The ORF of StMYBA1 was cloned into the pSAK-277 vector through EcoRI restriction sites, and the recombinant plasmid pSAK-StMYBA1 was sequenced (Sangon Biotech, Shanghai, China). The primer sequences pSAK-StMYBA1-F/R were listed in Table 1 .

The promoter sequences of StMYBA1, StCHS, StCHI, StF3H, StF3’H, StDFR, StF3’5’H, StANS and StGST were isolated from M6 (the tubers peel will turn purple gradually after exposure to light) and cloned into the pGreenII0800-LUC vector with the KpnI and NcoI restriction sites. The primer sequences were listed in Table 1 . Dual-luciferase assays were performed in tobacco, as previously described (Liu et al., 2016; Zhang et al., 2022), and the pGreenII0800-LUC vector without a promoter was used as a negative control. pSAK-StMYBA1 was infiltrated into tobacco leaves with pGreenII0800-StCHS, pGreenII0800-StCHI, pGreenII0800-StF3H, pGreenII0800-StF3’H, pGreenII0800-StF3’5’H, pGreenII0800-StDFR, pGreenII0800-StGST and pGreenII0800-StANS in a second Agrobacterium strain, respectively. To analyze the light response of the StMYBA1 gene promoter, tobacco leaves were infected with Agrobacterium containing fusion vectors pGreenII0800-StMYBA1, and the pGreenII0800-LUC vector without a promoter was selected as a negative control. After infiltration, the tobacco plants were kept in the dark for 12 h. Then some of the plants were moved to the culture chamber with 16 h of light per day, and the other plants were still in the dark. The leaves were sampled after three days of infiltration and stored at –70°C. The LUC and REN activities were tested using the Dual-Luciferase^® reporter Assay System (Promega, USA).

The overexpression (OE) vector pSAK-StMYBA1 was transformed into the potato diploid clone AC142 as previously (Si et al., 2003; Zhang et al., 2014). The transgenic plants and untransformed control were grown in 12-cm-diameter plastic pots in the greenhouse. For one month after planting, the leaves were sampled, frozen in liquid nitrogen, and stored at –70°C until use.

Tubers of potato population F₁RM are yellow peel yellow flesh, while the peel of the tubers will turn purple gradually after exposure to light ( Figure 1A ). Among them, the tubers of line RM-210 were selected to study the regulation of anthocyanin accumulation during light exposure. Purple can be seen at the top of the RM-210 tubers at 48 h light exposure ( Figure 1B ), and the whole tuber peels appear to be purple at 120 h ( Figure 1A ). Thus, the time points of sampling for anthocyanin content and RNA-seq analysis were 0, 6, 12, 36, 48, and 120 h.

The HPLC chromatograms of the extracts from the peel of tubers are shown in Figure S1 , and one peak was detected at approximately 0.91 min. According to the comparison analysis with the standard, the component represented by the spectrogram is delphinium. The results showed that delphinium could not be detected in the tuber peel with 0 and 12 h light exposure, and the content of delphinium in the tuber peel for 36, 48, and 120 h was 6.36, 34.08, and 43.07 mg/kg, respectively ( Figure 1C ). This indicates that light can promote the accumulation of delphinium in RM-210 tuber peel and show purple.

To understand the key pathways and regulatory genes in light regulation of anthocyanin biosynthesis, transcriptomics based on RNA-seq was performed on RM-210 tuber peel response to light. mRNA-sequencing libraries were generated from tuber peel samples obtained at 0, 6, 36, and 48 h light exposure, respectively, and except for two replicates from samples obtained at 36 h, all other samples had three replicates. The relationship among the samples was analyzed by the Principle Component Analysis (PCA) approach, which shows that the three replicates of the same sample are polymerized together with good repeatability ( Figure 2A ). Therefore, it was suggested that the RNA-seq data was suitable for further differential gene expression (DGE) analysis. To better understand the overall expression patterns of the transcripts in four samples, clustering analysis of the expression patterns in different samples was performed using the R package “TCseq”. There were six clusters in total from the four samples ( Figure 3 ). Interesting, the expression levels of genes from cluster 2 were higher at 6h than other samples, indicating that the genes in this cluster were the early light response. The expression of genes from class 6 increased with the light duration and anthocyanin accumulation, indicates that the function of these genes may related to anthocyanin accumulation.

To explore the genes contributing to anthocyanin biosynthesis in response to light, comparisons 6h vs 0h, 36h vs 0h, and 48h vs 0h were made. Based on fold change ≥ 2 and FDR < 0.01 criteria, 3224, 3976, and 3701 genes up-regulated were identified in comparisons 6h vs 0h, 36h vs 0h, and 48h vs 0h respectively, and 2642, 2969, and 2602 genes respectively were significantly down-regulated ( Figure 2B ). The list of DEGs in each comparison is shown in Table S1 , and the KEGG pathway enrichment analysis was performed. There were 26 pathways enriched in 6h vs 0h, the top 5 of which were photosynthesis-antenna proteins, photosynthesis, porphyrin and chlorophyll metabolism, ubiquinone and other terpenoid-quinone biosynthesis, and glutathione metabolism. There were 37 pathways enriched in 36h vs 0h, the top 5 of which were photosynthesis, photosynthesis-antenna proteins, phenylalanine metabolism, oxidative phosphorylation, and porphyrin and chlorophyll metabolism. In 48h vs 0h, there were 34 pathways enriched, of which the first 5 were photosynthesis, photosynthesis-antenna proteins, porphyrin and chlorophyll metabolism, phenylalanine metabolism, and carbon fixation in photosynthetic organisms ( Figure 2C ).

Delphinidin in potato is mainly synthesized by structural genes ( Figure 4A ) and regulated transcription factors. The expression levels of most structural genes in the four samples were different. For example, the expression of most PAL genes were high at 0h or 6h, C4H and 4CL genes were high at 6h, CHS and CHI genes were high at 36h or 48h, and the expression of F3H, F3’5’H, DFR, and ANS were high at 48h ( Figure 4A ). This result indicated that the expression of these early biosynthesis genes of anthocyanin is earlier than that of late biosynthesis genes.

To further explore the key transcription factors regulating structural genes, 114 differentially expressed structural genes, and TFs ( Table S2 ) were subjected to co-expression cluster analysis. The results showed that there were three clusters with a strong correlation between major structural genes and TFs ( Figure 4B ). The structural genes PAL (Soltu.DM.09G005700.1, Soltu.DM.10G020990.1), C4H (Soltu.DM.06G032850.1, Soltu.DM.06G032860.1), 4CL (Soltu.DM.03G020790.1) and MYB transcription factors (Soltu.DM.03G034270.1, Soltu.DM.11G026620.2), bZIP (Soltu.DM.09G003280.1) and B3 (Soltu.DM.09G002750.1) transcription factors were included in cluster 1. Above structural genes, PAL, C4H and 4CL had the highest expression in the sample exposed to light for 6 h, so it was speculated that the transcription factors in cluster 1 might be early light response factors. Cluster 2 included two CHS genes (Soltu.DM.09G028560.1, Soltu.DM.05G023610.1), two CHI genes (Soltu.DM.05G001950.1, Soltu.DM.07G023150.1) and one MYB gene (Soltu.DM.10G020840.1, StMYBA1), whose expression level is gradually increased with the extension of treatment time. And structural genes F3H (Soltu.DM.02G023850.1), F3’5’H (Soltu.DM.11G020990.1), DFR (Soltu.DM.02G024900.2), ANS (Soltu.DM.08G026700.1), GST (Soltu.DM.02G020850.1), and MYB (Soltu.DM.12G023200.1), bHLH (Soltu.DM.09G019660.1, StAN1), and WRKY (Soltu.DM.10G023760.1, StWRKY13) transcription factors were included in cluster 3. Expression levels of all the five structural genes in cluster 3 increased rapidly in the sample exposed to light for 48 h, indicating that the transcriptional factors associated with them may also function at the late stage of anthocyanin biosynthesis. To assess the repeatability of the sequencing data, the genes in cluster1 and cluster2 were selected to determine their expression levels using RT-qPCR. The results showed that most of them had a similar expression trend with transcriptome analysis results ( Figure S2 ).

The StMYBA1 gene in cluster 2 was previously confirmed to promote anthocyanin accumulation in tobacco leaves, and light is required for its function on anthocyanin accumulation (Liu et al., 2017). Analysis of the promoter sequences of StMYBA1 showed that it contained multiple cis-acting regulatory elements involved in light responsiveness, such as Box4 (ATTAAT) and G-box (CACGTC/CACGTC/TACGTG) ( Figure 5A , Table S3 ). To further analyze the response of the StMYBA1 promoter to light, tobacco leaves were infected with Agrobacterium containing fusion vectors pGreenII0800-StMYBA1. The dual luciferase activity in tobacco leaves was evaluated after 3d incubation in darkness or light. The result showed that the promoter activity of StMYBA1 in light conditions was significantly higher than that in dark, and no differences were observed in control ( Figure 5B ), suggesting that the activity of StMYBA1 promoter could be induced by light.

The 2Kb promoter sequences of key structural genes in potato anthocyanin biosynthesis (StCHS, Soltu.DM.05G023610.1; StCHI, Soltu.DM.07G023150.1; StF3H, Soltu.DM.02G023850.1; StF3’H, Soltu.DM.03G029340.1; StF3’5’H, Soltu.DM.11G020990.1; StDFR, Soltu.DM.02G024900.2; StANS, Soltu.DM.08G026700.1; StGST, Soltu.DM.02G020850.1) were selected. The sequence analyzing showed that each of the structural gene promoter sequences contained at least one MYB binding elements ( Figure 5A , Table S3 ). Therefore, transient luciferase assays in tobacco were used to test the activation of these structural gene promoters by StMYBA1. The results showed that compared to the negative control, StMYBA1 can significantly activate the promoters of StCHS, StCHI, and StF3H, but not the promoters of the other five structural genes ( Figure 5C ).

To explore the function of StMYBA1 in potato anthocyanin biosynthesis, the potato genotype AC142 was transformed with expression vector pSAK-StMYBA1 ( Figure S3A ). After rooting screening ( Figure S3B ) and gene expression analysis, nineteen overexpression transgenic lines with increased StMYBA1 transcripts were generated ( Figure S3C ). Three independent transgenic lines (OE-7, OE-8, and OE-10) were selected for function analysis. The transgenic lines showed normal plant morphology under greenhouse growing conditions comparing to wild type ( Figure S3D ). The leaf phenotype of transgenic lines was observed after 30 days of planting. Leaves of transgenic lines were purple, whereas those of wild-type were green ( Figure 6A ). Anthocyanin content in transgenic leaves was analyzed. The contents of delphinidin, pelargonidin, and cyanidin in the leaves of the transgenic lines were 5.37–6.15, 2.01–2.74, and 3.98–5.05 times than those in AC142 leaves, respectively ( Figure 6B ). These results strongly demonstrated that StMYBA1 could promote the accumulation of anthocyanin in potato leaves. The regulatory role of StMYBA1 was further investigated by analyzing the expression of the structural genes. The results showed that overexpression of the StMYBA1 gene could increase the expression of structural genes (StCHS, StCHI, StF3H, StF3’H, StF3’5’H, StDFR, StANS, and StGST) significantly ( Figures 6C–K ), implying that the StMYBA1 gene promotes anthocyanin biosynthesis in potato leaves by upregulating the expression of structural genes.